provided for non-commercial research and education use ... · briana hagen, guelph, ontario sandra...

The Canadian Journal of Veterinary Research is published by Canadian Veterinary Medical Association. The attached copy is

furnished to readers for personal, non-commercial research and education use. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party Web sites are prohibited. Those who require further

information regarding reprints or archiving and manuscript policies are encouraged to contact [email protected].

Provided for non-commercial research and education use. Not for reproduction, distribution or commercial use.

VETERINARY RESEARCHRECHERCHE VÉTÉRINAIREREVUE CANADIENNE DE

JULY/JUILLET 2019, VOL. 83, NO. 3

CANADIAN JOURNAL OFFOR PERSONAL USE ONLY

Published quarterly by the Canadian Veterinary Medical AssociationPublication trimestrielle de l’Association canadienne des médecins vétérinaires

C a n a d i a n J o u r n a l o fVe t e r i n a r y R e s e a r c h

Formerly CANADIAN JOURNAL OF COMPARATIVE MEDICINE

The journal publishes the results of original research in veterinary and comparative medicine. Manuscripts must be as concise as pos-sible, and the research described must represent a significant con-tribution to knowledge in veterinary medicine. Full-length papers, short communications, and review papers are welcome. All manu-scripts will be reviewed for scientific content and editorial accuracy.

Manuscripts must conform to the Instructions for Authors found on the CVMA website. Consultation of a recent issue of the journal to ensure that the manuscript conforms to current style is also recom-mended. Please use the website (www.canadianveterinarians.net) to submit articles. A publication charge of $90 per published page for CVMA Members and $130 per published page for Non-members, plus a cost of $25 per table, $25 per illustration, and $150 per color illustration is payable by contributors. Instructions for ordering reprints are sent with the galley proofs. Color illustrations will only be published if the author(s) agree to pay the extra cost.

R e v u e c a n a d i e n n e d er e c h e r c h e v é t é r i n a i r e

Remplace la REVUE CANADIENNE DE MÉDECINE COMPARÉE

La revue publie les résultats de recherches originales en médecine vétérinaire et comparative. Les manuscrits doivent être aussi brefs que possible et la recherche décrite doit apporter une contribution impor-tante à la médecine vétérinaire. Les exposés complets, les communi-cations brèves et les comptes rendus sont bienvenus. Nous véri fierons le fonds scientifique et la forme de tous les manuscrits.

Les manuscrits doivent être conformes aux directives aux auteurs qui se trouvent sur le site Web de l’ACMV. La consultation d’un récent numéro de la revue est aussi profitable, afin de s’assurer qu’un manuscrit se conforme au style courant. Veuillez utiliser le site Web (www.veterinairesaucanada.net) pour soumettre les articles. Les auteurs doivent acquitter des frais de 90 $ par page publiée pour les membres de l’ACMV et 130 $ par page publiée pour les non-membres, 25 $ pour chaque tableau, 25 $ pour chaque illustration, et 150 $ pour chaque illustration en couleur. Un formulaire de commande de tirés à part accompagne l’en voi des épreuves d’un manuscrit. Les illustrations en couleurs ne seront publiées que si les auteurs en assument les frais supplémentaires.

Established in 1937 ISSN 0830-9000

Address all correspondence to/Adresser toute correspondance à :Canadian Journal of Veterinary Research Revue canadienne de recherche vétérinaire339, rue Booth Street, Ottawa, Ontario K1R 7K1Tel./Tél. : (613) 236-1162 — Fax/Télécopieur : (613) 236-9681e-mail/Courriel : [email protected]© Canadian Veterinary Medical Association 2019

L’Association canadienne des médecins vétérinaires 2019

Typesetting/Typographie AN Design Communications www.an-design.ca

Canadian Journal of Veterinary ResearchRevue canadienne de recherche vétérinaireEditor — Rédacteur Éva Nagy, Guelph, OntarioAssociate Editor — Rédacteur adjoint Faizal Careem, Calgary, AlbertaAssistant Editors — Assistants à la rédaction Briana Hagen, Guelph, Ontario Sandra McConkey, Charlottetown, Prince Edward Island Serge Messier, Saint-Hyacinthe (Québec)Managing Editor — Directrice de la rédaction Heather Broughton, Ottawa, OntarioAssistant Managing Editor — Directrice adjointe à la rédaction Stella Wheatley, Ottawa, OntarioAdvertising and Sponsorship Consultant — Consultante, publicité et commandites Laima Laffitte, Wendover, OntarioEditorial Coordinator/Coordonnatrice de la rédaction Kelly Gray-Sabourin, Ottawa, Ontario

Abstracted and/or indexed in:/Résumé et/ou catalogué dans :Biological Abstracts Index VeterinariusChemical Abstracts Nutrition AbstractsCurrent Contents (Agric. Section) Science Citation IndexEMBASE/Excerpta Medica Veterinary BulletinIndex Medicus Derwent Veterinary Drug File

Subscription/Abonnement 2019Personal Subscription/ $180 $ (1 GST/HST as Abonnement personnel applicable; TPS/TVH en (Canada) vigueur)Personal Subscription/ $200 $ US/É.-U. Abonnement personnel (Foreign/Étranger)Institutional Subscription/ $250 $ (1 In Canada, GST/HST Abonnement institutionnel as applicable; Au Canada, TPS/TVH selon le cas)Institutional Subscription/ $250 $ US/É.-U. Abonnement institutionnel (Foreign/Étranger)

FOR PERSONAL USE ONLY

ARTICLES

Molecular characterization of Mycobacterium avium subspecies paratuberculosis C-type and S-type isolated from sheep and goats by using a combination of MIRU-VNTR lociJagdip Singh Sohal, Julie Arsenault, Anne Leboeuf, Pierre Hélie, Sébastien Buczinski, Yves Robinson, Olivia Labrecque, Virginie Lachapelle, Gilles Fecteau, Yvan L’Homme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160

Molecular variability of Streptococcus uberis isolates from intramammary infections in Canadian dairy farms from the Maritime regionJulian Reyes, Juan C. Rodriguez-Lecompte, Adam Blanchard, J.T. McClure, Javier Sánchez . . . . . . . 168

Salmonella shedding and seropositivity and its association with in-feed flavophospholipol in nursery pigsSaranya Nair, Jane Newman, Abdolvahab Farzan, Robert M. Friendship . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177

Accuracy of external measurements of 3-dimensional (3D) printed biomodels of the canine radius used in an in-hospital settingSebastian Mejia, Nikia Stewart, Angel Miller, Roman Savicky, Christopher Monarski, George E. Moore, Dennis Keith . . . . . . . . . . . . . . . . . . . . 181

Pilot study comparing serum chemotherapy levels after intra-arterial and intravenous administration in dogs with naturally occurring urinary tract tumorsMeghan Kirsch, Chick Weisse, Allyson Berent, Craig Clifford, Nicole Leibman, Luke Wittenburg, Stephen B Solomon, Kenneth Lamb . . . . . . . . . . . . . . . . 187

Comparison of 3 blind brachial plexus block techniques during maintenance of anesthesia and postoperative pain scores in dogs undergoing surgical procedures of the thoracic limbAlicia Skelding, Alexander Valverde, Rodrigo Aguilera, Noel M. Moens, Melissa Sinclair, Jeffrey J. Thomason . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197

Effects of various dietary supplements on inflammatory processes in primary canine chondrocytes as a model of osteoarthritisEman A. AlRaddadi, Tanja Winter, Harold M. Aukema, Donald W. Miller . . . . . . . . . . . . . . 206

Relationship between reproductive management practices and fertility in primiparous and multiparous dairy cowsIstván Fodor, György Gábor, Zsolt Lang, Zsolt Abonyi-Tóth, László Ózsvári . . . . . . . . . . . . . . . . . 218

SHORT COMMUNICATIONS/COMMUNICATIONS BRÈVES

Feeding altrenogest during late lactation improves fertility of primiparous sows nursing smaller littersNutthee Am-in, Roy N. Kirkwood . . . . . . . . . . . . . . . . . 228

Investigating the ability of methicillin-resistant Staphylococcus pseudintermedius isolates from different sources to adhere to canine and human corneocytes Nathita Phumthanakorn, Nuvee Prapasarakul . . . . . . . 231

Effects of 3 morphine doses, in combination with acepromazine, on sedation and some physiological parameters in dogsEduardo R. Monteiro, Thiago A. Rabello, Julia P.P. Rangel, Juarez S. Nunes Jr, Cesar D. Freire, Daniela Campagnol . . . . . . . . . . . . . . . 235

JULY/JUILLET 2019, VOL. 83, NO. 3


Article

160 The Canadian Journal of Veterinary Research 2019;83:160–167

I n t r o d u c t i o nMycobacterium avium subspecies paratuberculosis (Map) is respon-

sible for paratuberculosis, a chronic granulomatous enteritis of domestic and wild ruminants. The disease is prevalent in domestic animals worldwide and is associated with substantial productivity losses in cattle and small ruminants (1,2). Diagnosis and control of paratuberculosis is challenging due to the long incubation period and to the environmental resistance of Map. In domestic ruminants, paratuberculosis is generally characterized by a subclinical phase of several months or years, followed by a terminal stage eventu-ally leading to culling or death (3). Growing evidence suggests that

infection severity, the ability to survive in macrophages and the response to the host’s immune system are all influenced by the type of Map strain (427).

Historically, Map strain types were named after the species where they were first isolated, which were the sheep (S) type and cattle (C) type (8). The C-type Map strains are the predominant strain type isolated from dairy cattle whereas S-type strains have been mainly isolated from small ruminants (9). Over the last few decades, new typing techniques were developed, mostly based on investigation of tandem-repeats (TR), which are polymorphic minisatellite sequences dispersed throughout the genome of Mycobacteria (10). Included in the TR family, the mycobacterial interspersed repetitive units

Molecular characterization of Mycobacterium avium subspecies paratuberculosis C-type and S-type isolated from sheep and goats

by using a combination of MIRU-VNTR lociJagdip Singh Sohal, Julie Arsenault, Anne Leboeuf, Pierre Hélie, Sébastien Buczinski, Yves Robinson,

Olivia Labrecque, Virginie Lachapelle, Gilles Fecteau, Yvan L’Homme

A b s t r a c tMycobacterium avium subspecies paratuberculosis (Map) is the etiological agent of paratuberculosis of domestic and wild ruminants. Map strains are segregated into 2 main groups or strain types referred to as sheep (S) type and cattle (C) type. Few small ruminant Map strains have been genetically characterized to date. The present study was undertaken to genetically characterize a panel of 30 small ruminant Map strains in the province of Quebec, Canada. Mycobacterial Interspersed Repetitive Units — Variable-Number Tandem Repeat analysis (MIRU-VNTR) were used as genetic markers in addition to IS1311 PCR-REA. S-type and C-type strains were found in both sheep and goats, although C-type strains were more frequently isolated from goats and S-type strains were more common in sheep. A total of 12 distinct Map genotypes were uncovered in the present collection of strains using these markers. Considering the genetic diversity reported here, molecular characterization of Map stains in small ruminants using MIRU-VNTR markers represent an interesting avenue for both epidemiological investigations regarding the sources of herd infection and association studies between Map strains and their virulence, persistence and host-specific adaptation characteristics.

R é s u m éMycobacterium avium subspecies paratuberculosis (Map) est l’agent étiologique de la paratuberculose affectant les ruminants sauvages et domestiques. Les souches de Map se répartissent dans deux grands groupes ou types appelés ‘sheep (S)’ et ‘cattle (C)’. Très peu de souches de Map provenant des petits ruminants ont été caractérisées génétiquement jusqu’à présent. Cette étude a été initiée afin de caractériser un ensemble de 30 souches de Map provenant de 5 troupeaux de moutons et 8 troupeaux de chèvres situés dans la province de Quebec, Canada, et d’évaluer leur diversité génétique. Une analyse répétée en tandem des unités répétitives alternées des mycobactéries (MIRU-VNTR) a été utilisée comme marqueurs génétiques en plus du marqueur IS1311 PCR-REA. Les souches de type S et C ont été retrouvées chez les isolats ovins et caprins, avec une prédominance des souches de type C chez les isolats provenant de chèvres tandis que les souches de type S étaient plus fréquentes chez les moutons. Un total de 12 génotypes distincts de Map ont été retrouvés parmi les isolats d’après les marqueurs utilisés. Considérant la diversité génétique observée, la caractérisation moléculaire des isolats de Map représente une avenue intéressante pour investiguer les sources potentielles d’infection des troupeaux et pour étudier les associations entre les caractéristiques génétiques et pathogéniques des isolats.

(Traduit par les auteurs)

Canadian Food Inspection Agency, 3400 Casavant, St-Hyacinthe (Sohal, Robinson, Lachapelle, L’Homme); Faculté de Médecine Vétérinaire, Université de Montréal, 3200 Sicotte, St-Hyacinthe (Arsenault, Hélie, Buczinski, Fecteau, L’Homme); Ministère de l’Agriculture, des pêcheries et de l’alimentation du Quebec, 200 chemin Ste-Foy (Leboeuf); Laboratoire de santé animale du Quebec, Ministère de l’Agriculture, des pêcheries et de l’alimentation du Quebec, 3220 Sicotte, St-Hyacinthe, Quebec (Labrecque).

Address all correspondence to Dr. Yvan L’Homme; telephone: (418) 688-8310; e-mail: [email protected]

Received March 21, 2018. Accepted June 6, 2018.


2000;64:0–00 The Canadian Journal of Veterinary Research 161

(MIRU) and variable number tandem repeats (VNTR) methods have gained popularity due to their ease of use and improved discrimina-tory powers over traditional methods (11215). Field application of these PCR-based typing methods has started to unveil interspecies transmission events of C-type and S-type strain (16,17). However, most of these studies have focused on the characterization of C-type strains due to the difficulties inherent with the isolation of S-type strains in culture.

The present study was undertaken to characterize a panel of small ruminant Map strains from different herds in the province of Quebec, Canada, and to describe their relatedness to bovine strains previously characterized in the province of Quebec. Since Map has limited genetic diversity, we used a total of 13 MIRU-VNTR mark-ers to characterize the genome of Map strains isolated from small ruminants.

M a t e r i a l s a n d m e t h o d s

Source of Map strainsSheep — A collection of 20 strains isolated from 20 individual

sheep originating from a project conducted from 2010 to 2012 involving meat sheep herds located across the province of Quebec was used. Map isolates were recovered from intestinal tissues at the ileocecal junction (ICJ) and from mesenteric lymph nodes (MLN) of sheep over 2 y of age, suspected of having paratuberculosis. Tissue sections from ICJ and MNL (0.5 cm3) were finely chopped and suspended in 10 mL of sterile distilled water (SDW). Tissue samples were then placed in a plastic bag and homogenized using a stomacher at maximum speed for 5 min. The homogenate was mixed with 10 mL of 0.9% hexadecyle pyridinium chloride (HPC) for decontamination and allowed to stand at room temperature for 16 h. The suspension was centrifuged at 3800 3 g for 30 min at 4°C. The pellet was resuspended in 1 mL of SDW and centrifuged at 6500 3 g for 5 min. The final pellet was resuspended in 1 mL SDW. Aliquots of 100 mL were inoculated on 5 replicate slants of modified agar medium (Modified MiddleBrook 7H10 agar medium; BD, Mississauga, Ontario) supplemented with mycobactin and 5 replicate slants of Lowenstein-Jensen medium (BD, Mississauga, Ontario) supplemented with mycobactin (18). Slants were incubated at 37°C and observed every 2 wk for the appearance of colonies for a maximum of 36 wk. Molecular confirmation of Map strains was done using PCR (IS900 PCR) as indicated (19).

Goat — Goat Map isolates (n = 10) originating from 10 individual goats were obtained from a collection originating from a distinct project investigating the main causes of death in goats, in Quebec in 2010 (20). A section of the ileum was collected for culture from goats presenting with gross lesions suggestive of paratuberculosis at necropsy. Granulomatous ileitis and mesenteric lymphadenitis with intra-histiocytic acid-fast bacilli, consistent with paratubercu-losis, was later confirmed in all selected goats. Primary isolation of Map was done as follows: on day 1, 1 mm of ileum tissue was cut into small pieces and 400 mg were placed in a 2 mL tube contain-ing 1 mL of saline and 1 mm glass beads. The tubes were placed for 5 min in a tissue homogenizer (Tissue lyser; Qiagen, Toronto, Ontario), and then centrifuged for 2 min at 8500 3 g. The superna-

tant was harvested and mixed 50:50 with Brain heart infusion (BHI) broth containing 0.9% hexadecylpyridinium chloride monohydrate (HPC). The suspensions were incubated at 35°C for 24 h. On day 2, the samples were centrifuged for 30 min at 900 3 g. The pellet was resuspended in 1.0 mL of sterile saline. For each sample, 6 tubes of mycobacteria growth indicator tube (MGIT Para TB Medium; BD) containing a different antibiotic cocktail were seeded with 100 mL of suspension. To increase the likelihood of small ruminant Map strain isolation and following discussions with the manufacturer, the tubes were incubated for 3 wk at 35°C and then transferred to the MGIT 960 apparatus for 2 consecutive cycles of 7 wk of incuba-tion. Presumptive positive vials were confirmed by acid fast stain and by PCR (TaqMan Map Johne’s Reagents; Life Technologies, Mississauga, Ontario). For isolating colonies, subcultures were made as follows: a total of 500 mL of growth medium from positive MGIT vials was diluted 1:5 in sterile phosphate-buffered (saline) solution (PBS). Aliquots of 100 mL of each dilution were then inoculated on 5 replicate slants of modified medium (MiddleBrook 7H10) with mycobactin (18) and egg yolk medium (Herrold’s Egg Yolk medium; BD with mycobactin). Slants were incubated at 37°C and observed every 2 wk for the appearance of colonies for a maximum of 36 wk.

Group typing of Map strains using IS1311 PCR-REA

Group typing was done using IS1311 polymerase chain reaction restriction enzyme analysis (PCR-REA) directly from Map colonies. IS1311 PCR reactions were done as described previously with slight modifications (21,22). Briefly, single colonies were harvested with a sterile toothpick and diluted in 10 mL of PCR grade water before being heated at 95°C for 10 min. The mixture was transferred to 40 mL of a PCR reaction mix consisting of 5 mL of 103 PCR buffer, 200 mM dNTPs, 2.5 mM of MgCl2, 0.6 mM of primers M-56 and M-119, and 2 U of Taq polymerase (Invitrogen, Mississauga, Ontario). Amplification was done under the following conditions: one cycle of 3 min at 94°C and 35 cycles of 30 s at 94°C, 15 s at 62°C, and 1 min at 72°C. Amplification reactions were analyzed on a capillary electro-phoresis instrument (Qiaxcel; Qiagen). The expected amplicon size was 608 nucleotides. Digestion of the amplicons using restriction enzymes (REA reactions) were carried out in a volume of 30 mL, containing 20 mL of IS1311 amplicons, 3 mL of reaction 103 buf-fer, and 10 U of each endonucleases HinfI and MseI (New England Biolabs, Whitby, Ontario). Reactions were incubated at 37°C for 2.5 h. Restriction reactions were analyzed using a capillary electrophoresis instrument (Qiaxcel; Qiagen). Band patterns were interpreted as previously described (21).

Genomic DNA extractionFollowing months of incubation, once visible colonies were appar-

ent on a slant, all colonies were harvested in mycobacterial storage medium (20% glycerol solution containing 3.0% w/v tryptone soya broth) by up and down pipetting and stored at 270°C. A total of 200 mL of this Map solution was used to isolate genomic DNA using a DNA extraction kit (QIAamp DNA mini kit; Qiagen) following the recommended protocol with the following slight modifications: an initial enzymatic lysis step was done using 200 mL of 20 mg/mL of lysozyme, then incubated for 3 h at 37°C before proceeding with the



manufacturer’s protocol. The isolated DNA was stored at 220°C until needed.

MIRU-VNTR typingThe MIRU-VNTR typing was carried out by PCR amplification of

a total of 13 different loci. A total of 8 loci [VNTR 292 (also known as MIRU-2), VNTR X3 (also known as MIRU-3), VNTR 25, VNTR 47, VNTR 3, VNTR 7, VNTR 10, and VNTR 32] described by Thibault et al (12); 2 loci (MIRU-1 and MIRU-4) described by Bull et al (24); 1 loci (VNTR 259) described by Castellanos et al (11); and finally 2 loci (VNTR 1067 and VNTR 3527) described by Overduin et al (23). For each locus, the primers and PCR conditions were exactly as described by the authors (11,15,23,24). The amplicons were analyzed on a capillary electrophoresis instrument (Qiaxcel; Qiagen). The number of repeats at each locus was determined according to the sizes of the amplicons, based on a discrimination using the number of repeats as previously described (11,23,24). The Institut National de Recherche Agronomique, France (INRA) MIRU-VNTR nomen-clature (i.e., INMV profile number) was used to analyze the results for the 8 MIRU-VNTR loci: 292, X3, 25, 47, 3, 7, 10, and 32 using the INRA online typing tool Mycobaterium avium complex MAC-INMV, (homepage on the internet: http://mac-inmv.tours.inra.fr/). For the remaining 5 loci (MIRU 1, MIRU 4, VNTR 259, VNTR 1067, and VNTR 3527) a new nomenclature (Mycobacterium VNTR -MV- typ-

ing) was proposed, sequentially following the 3 different MV types reported by Sohal et al (13). The number of repeats at each of the 13 marker loci was recorded. The combination of repeats for each marker formed a specific profile for every strain, which was used as a determinant to attribute Map genotypes.

Data analysisThe allelic diversity was calculated at the different loci as pre-

viously described (25), and separately for the C-type and S-type strains. A single isolate producing the same genotype per herd was included in the calculation (25). The allelic diversity represents the probability that 2 alleles randomly selected from the population are different from each other. The distribution of genotypes at the farm level was descriptively compared between sheep and goat isolates, and also with dairy cattle strains, which were previously reported in the same province and similar period of time (13).

Re s u l t s

Strain typingA total of 20 strains were obtained from 20 individual sheep

belonging to 5 distinct flocks. A total of 16 strains were S-type and 4 strains were C-type according to IS1311 PCR-REA results (Table I).

Table I. Sheep Map isolate typing using MIRU-VNTR analysis.

Isolate Farm Groupa INMV profileb INMV Typec MV Profiled MV Typee Genotypef

JY 1 1 S 41331218 INMV 129 31112 MV 7 G 8JY 2 2 S 41331218 INMV 129 31112 MV 7 G 8JY 3 2 S 71331218 INMV 119 31112 MV 7 G 11JY 4 2 C 32332228 INMV 2 31222 MV 1 G 2JY 84 2 S 41332128 INMV 56 32222 MV 8 G 12JY 87 2 S 41332128 INMV 56 32222 MV 8 G 12JY 88 2 S 41332128 INMV 56 32222 MV 8 G 12JY 91 2 S 41332128 INMV 56 32222 MV 8 G 12JY 97 2 S 41332128 INMV 56 32222 MV 8 G 12JY 7 3 S 41331218 INMV 129 31222 MV 1 G 9JY 26 3 C 32332228 INMV 2 31222 MV 1 G 2JY 37 3 C 32332228 INMV 2 31222 MV 1 G 2JY 9 4 S 41331218 INMV 129 31212 MV 3 G 10JY 77 4 S 40331118 INMV 61 31112 MV 7 G 5JY 78 4 S 40331118 INMV 61 31112 MV 7 G 5JY 79 4 S 40331118 INMV 61 31112 MV 7 G 5JY 80 4 S 40331118 INMV 61 31112 MV 7 G 5JY 81 4 S 40331118 INMV 61 31112 MV 7 G 5JY 20 5 S 41331118 INMV 72 31212 MV 3 G 7JY 46 5 C 32332218 INMV 3 31272 MV 6 G 6a S — Sheep type; C — Cattle type.b INMV profile based on loci: 292, X3, 25, 47, 3, 7, 10, and 32.c INMV type based on the INRA MAC-INMV database for MIRU-VNTR typing.d MV Profile based on loci: MIRU 1, MIRU 4, VNTR 259, VNTR 1067, and VNTR 3527.e MV type sequentially following the nomenclature proposed by Sohal et al (13).f Genotype based on combined INMV and MV profiles.INMV; MV; INRA — Institut National de Recherche Agronomique, France; MAC; MIRU — mycobacterial interspersed repetitive units; VNTR — variable number tandem repeats.



The S-type strains grew on both MB7H10 and LJ media, whereas C-type strains only grew on MB7H10. Growth rate of S-type strains was much slower than C-type strains, taking an average of 12 wk longer before the appearance of visible colonies.

A total of 12 strains from 12 individual goats originating from 8 flocks were isolated in MGIT and confirmed by PCR. From this total, 10 samples from 8 flocks gave visible colonies on MB7H10 and HEYM media and are referred to as strains in Table II. Group typing of these Map strains using IS1311 PCR-REA revealed that the C-type represented the predominant group (8/10 strains; Table II), whereas 2 strains originating from the same flock, were S-type.

SubtypingA total of 12 C-type strains were obtained from 8 goats and

4 sheep. The INMV typing segregated the C-type strains into 4 dif-ferent INMV types: INMV 1 (n = 5), INMV 2 (n = 5), INMV 3 (n = 1), and INMV 5 (n = 1) (Tables II and III). Two distinct MV types were observed for C-type strains: MV 1 (n = 11) and MV 6 (n = 1). By combining the results of INMV and MV types, a total of 4 distinct genotypes were observed: G1 (n = 5), G2 (n = 5), G3 (n = 1), and G6 (n = 1). Only genotype G2 was shared between sheep and goats.

A total of 18 S-type strains were isolated from 16 sheep and 2 goats. A total of 6 different INMV types were observed (Tables II and III): INMV 56 (n = 5) and INMV 61 (n = 5), INMV 72 (n = 1), INMV 119 (n = 1), INMV 129 (n = 4), and INMV 130 (n = 2). Among the INMV types, INMV 56, and INMV 61 had never been reported before. Five distinct MV types were observed for S-type strains: MV 1 (n = 1), MV 3 (n = 2), MV 4 (n = 2), MV 7 (n = 8), and MV 8 (n = 5). Combined diversity analysis of INMV and MV types segregated S-type strains into 8 different genotypes: G4 (n = 2), G5 (n = 5), G7 (n = 1), G8 (n = 2), G9 (n = 1), G10 (n = 1), G11 (n = 1), and G12 (n = 5). No genotype was shared between sheep and goats.

Allelic diversityAs shown in Table III, most INMV and MV markers were mono-

morphic for C-type strains; only loci INMV 10, INMV 292, and MV 1067 showed heterogeneity and a total of 4 distinct genotypes were revealed. The gene diversity ranged from 0.18 to 0.54 for these 3 loci. This is in sharp contrast with S-type strains, which were poly-morphic at most loci, revealing a total of 8 distinct genotypes. Only loci INMV 32 and INMV 47 were monomorphic for S-type strains. The allelic diversity ranged from 0.22 to 0.55 for the polymorphic loci.

Comparison with dairy cattle strainsThe comparison of small ruminant Map strains with dairy cattle

strains from the same province and similar time period revealed G2 as the shared genotype between the 3 species. The G2 genotype was also the most common among dairy cattle (Table IV). Genotype G1, which was the most commonly found genotype in goats, was also shared between goats and dairy cattle.

D i s c u s s i o nCompared to Map strains originating from cattle, which are almost

exclusively C-type, characterization of the genetic diversity of small ruminants Map strains is not well-documented (15). Furthermore, few studies have reported molecular characterization of S-type Map strains, most likely due to the challenge associated with their isolation and their slower growth rate than C-type strains in culture (15,26–28). In this study, a total of 30 small ruminant Map strains were isolated and genotyped, including 18 S-type strains. These strains were from a convenience sample of a limited number of farms, limiting our ability to apply the results to the entire province. However, the farms of origin were distributed across the province of Quebec, were independently owned, and typical of the small

Table II. Goat Map isolate typing using MIRU-VNTR analysis.

Isolate Farm Groupa INMV profileb INMV typec MV profiled MV typee Genotypef

C7 a C 42332218 INMV 5 31222 MV 1 G 3C14 b C 32332228 INMV 2 31222 MV 1 G 2C29 c C 42332228 INMV 1 31222 MV 1 G 1C51 c S 51130218 INMV 130 21131 MV 4 G 4C60 c S 51130218 INMV 130 21131 MV 4 G 4C30 d C 42332228 INMV 1 31222 MV 1 G 1C41 e C 42332228 INMV 1 31222 MV 1 G 1C42 f C 32332228 INMV 2 31222 MV 1 G 2C56 g C 42332228 INMV 1 31222 MV 1 G 1C64 h C 42332228 INMV 1 31222 MV 1 G 1a S — Sheep type; C — Cattle type.b INMV profile based on loci: 292, X3, 25, 47, 3, 7, 10, and 32.c INMV type based on the INRA MAC-INMV database for MIRU-VNTR typing.d MV profile based on loci: MIRU 1, MIRU 4, VNTR 259, VNTR 1067, and VNTR 3527.e MV type sequentially following the nomenclature proposed by Sohal et al (13).f Genotype based on combined INMV and MV profiles.INMV; MV; INRA — Institut National de Recherche Agronomique, France; MAC; MIRU — mycobacterial interspersed repetitive units; VNTR — variable number tandem repeats.



ruminant industry, which could be considered representative of the province.

Group typing revealed that both goats and sheep can be infected with either group of strain (S-type or C-type) (4). However, the C-type was predominant in goats, whereas the S-type was most often isolated from sheep; these results are consistent with current knowledge (17,21,29,30). However, it should be noted that in our study, the protocol used for primary isolation of Map in goats was adapted for C-type strains. Considering that S-type strains have dif-ferent culture requirements and often need longer incubation time (31), they may have been under detected in goat herds in this study. It would, therefore, be advisable in future studies to include isolation media that supports the growth of both C-type and S-type strains.

Two of the 3 sheep farms with C-type strains had a G2 genotype profile (INMV 2-MV1), which was also found to be the most common profile among dairy cattle strains in this province (13). Genotype G2 was also found in 2 of the 8 goat farms in the present study. Likewise, the most common genotype found among goat strains was INMV 1-MV1 (G1), which was previously determined to be common among dairy cattle in the province (13). The 2 VNTR types, INMV 1 and INMV 2, were also found to be the most common in cattle from other Canadian provinces (32) and in dairy goats from Ontario (33). Taken together, these findings could suggest the pres-ence of a common source of infection or inter-species transmission. Further studies are warranted to characterize epidemiological links between cattle, goats, and sheep farms in this country. The presence

Table III. Relative frequency (%) of selected isolates from sheep and goats according to the specific allele copy number and allelic diversity for each locus, by type (C versus S).

Number Specific allele copy number AllelicLocus of isolates 0 1 2 3 4 5 6 7 8 $ 9 diversityType = C INMV X3 11 100 0.00 3 11 100 0.00 7 11 100 0.00 10 11 18 82 0.32 25 11 100 0.00 32 11 100 0.00 47 11 100 0.00 292 11 45 55 0.54

MV MIRU 1 11 100 0.00 MIRU 4 11 100 0.00 259 11 100 0.00 1067 11 91 9 0.18 3527 11 100 0.00

Type = S INMV X3 9 11 89 0.22 3 9 11 78 11 0.41 7 9 33 67 0.50 10 9 89 11 0.22 25 9 11 89 0.22 32 9 100 0.00 47 9 100 0.00 292 9 78 11 11 0.41

MV MIRU 1 9 11 89 0.22 MIRU 4 9 89 11 0.22 259 9 56 44 0.55 1067 9 67 22 11 0.55 3527 9 11 89 0.22INMV; MV; MIRU — mycobacterial interspersed repetitive units.



of multiple Map strain types was observed in all 5 farms harboring more than 1 isolate, suggests multiple sources of infection. This finding underscores the importance of including multiple isolates from each farm when conducting molecular epidemiology studies of Map in small ruminants.

Markers VNTR 32 and 47 were found to be monomorphic in this study in both C-type and S-type strains isolated from both species. This is consistent with a number of studies, including a recent one from our group, where nearly 200 cattle Map strains were genotyped (13,15,28). However, others have found variations at those 2 loci in populations of different animal species (12,25). Of note, the MV loci were highly monomorphic for C-type strains of sheep and goats of this study with a single locus exhibiting genetic variability. This is similar to our previous findings regarding C-type strains from cattle where only locus MV 1067 showed genetic variability (13). However, this is in sharp contrast with the genetic variability of S-type strains from this study where all 5 MV loci were found to be polymorphic. With regards to the discriminatory hierarchy, locus INMV 292 showed the highest diversity for C-type strains whereas locus MV 1067 displayed the highest variability for S-type strains. Together, these findings justify a careful evaluation of different markers and geographical specificities regarding Map strains and genetic variations. Otherwise, there is a potential risk to lose impor-tant genetic and epidemiological information. In a recent study,

24 S-type strains from various geographical areas were analyzed and only 3 INMV loci (292, X3, and 25) were found to be both polymor-phic and sufficient to describe the entire variability of their strains (15). Conversely, our findings revealed that multiple markers were polymorphic and informative. Overall, MIRU-VNTR genotyping proved to be a powerful technique for discriminating Map strains in this study, particularly for S-type. According to our findings, it could be used for future molecular epidemiological studies. The dis-criminatory power can be further improved, by using SSR markers in conjunction to MIRU-VNTR as shown previously (13).

Overall, the results presented in this study provide additional information about the diversity of Map strains in small ruminants in Quebec. By using informative molecular markers applied to isolates infecting different host species, this study reports on the genetic diversity of a number of Map strains isolated from small ruminants. The methods presented here, in which Map strains can be relatively easily segregated based on their genotype, could serve as a starting point for investigating transmission patterns, virulence, persistence, and clinical significance of distinct Map strains in Quebec. With the recent advances and availability of whole genome sequencing technologies and the underlining promise of increased discrimina-tory power, future molecular epidemiology studies, as well as host-pathogen investigations, will undoubtedly unveil unprecedented findings about this important pathogen.

Table IV. Distribution of Map genotypes in 65 dairy cattle, 5 sheep and 8 goat farms.

INMV MV Number of positive farma

Profile Type Profile Type Genotype Cattleb Sheep GoatGroup = C 42332228 INMV 1 31222 MV 1 G1 5 — 5 32332228 INMV 2 31222 MV 1 G2 42 2 2 42332218 INMV 5 31222 MV 1 G3 — — 1 32332218 INMV 3 31272 MV 6 G6 — 1 — 32332218 INMV 3 31222 MV 1 G13 17 — — 32332228 INMV 2 31212 MV 3 G14 2 — — 32332328 INMV 8 31212 MV 3 G15 1 — — 22332228 INMV 13 31222 MV 1 G16 3 — —

Group = S 51130218 INMV 130 21131 MV 4 G4 — — 1 40331118 INMV 61 31112 MV 7 G5 — 1 — 41331118 INMV 72 31212 MV 3 G7 — 1 — 41331218 INMV 129 31112 MV 7 G8 — 2 — 41331218 INMV 129 31222 MV 1 G9 — 1 — 41331218 INMV 129 31212 MV 3 G10 — 1 — 71331218 INMV 119 31112 MV 7 G11 — 1 — 41332128 INMV 56 32222 MV 8 G12 — 1 —

Group = B 22532228 INMV 68 31232 MV 2 G17 2 — —a A farm was considered as positive for a genotype when the genotype was found in at least 1 animal (and/or environmental sample for dairy cattle) from the farm. A total of 191, 20, and 10 isolates were genotyped from dairy cattle, sheep, and goat samples, respectively.b From Sohal et al (13).INMV; MV — microbacterial interspersed repetitive units.



A c k n o w l e d g m e n t sThis work was partly supported by the sector councils of Quebec

and Ontario who administer the Canadian Agricultural Adaptation Program (CAAP) for Agriculture and Agri-Food Canada. The authors thank all those who made this work possible by contributing to the creation of a strain collection, including our collaborators from the Centre d’Expertise en Production Ovine du Quebec (CEPOQ), Société des Éleveurs de Chèvres Laitières de Race du Quebec (SECLRQ), Regroupement des Éleveurs de Chèvres de Boucherie du Quebec (RECBQ), as well as the participating slaughterhouse, veterinary practitioners, and producers.

Re f e r e n c e s 1. Garcia AB, Shalloo L. Invited review: The economic impact and

control of paratuberculosis in cattle. J Dairy Sci 2015;98:5019–5039. 2. Ruggiero S, Pieragostini E, Rubino G, Petazzi F. Impact of

Mycobacterium avium subspecies paratuberculosis on profit effi-ciency in semi-extensive dairy sheep and goat farms of Apulia, southern Italy. Prev Vet Med 2017;136:56–64.

3. Benedictus A, Mitchell RM, Linde-Widmann M, et al. Transmis-sion parameters of Mycobacterium avium subspecies paratubercu-losis infections in a dairy herd going through a control program. Prev Vet Med 2008;83:215–227.

4. Stevenson K. Genetic diversity of Mycobacterium avium subspe-cies paratuberculosis and the influence of strain type on infection and pathogenesis: A review. Vet Res 2015;46:64.

5. Gollnick NS, Mitchell RM, Baumgart M, Janagama HK, Sreevatsan S, Schukken YH. Survival of Mycobacterium avium subsp. paratuberculosis in bovine monocyte-derived macrophages is not affected by host infection status but depends on the infecting bacterial genotype. Vet Immunol Immunopathol 2007; 120:93–105.

6. O’Brien R, Mackintosh CG, Bakker D, Kopecna M, Pavlik I, Griffin JF. Immunological and molecular characterization of susceptibility in relationship to bacterial strain differences in Mycobacterium avium subsp. paratuberculosis infection in the red deer (Cervus elaphus). Infect Immun 2006;74:3530–3537.

7. Verna AE, Garcia-Pariente C, Munoz M, et al. Variation in the immuno-pathological responses of lambs after experimental infection with different strains of Mycobacterium avium subsp. paratuberculosis. Zoonoses Public Health 2007;54:243–252.

8. Collins DM, Gabric DM, de Lisle GW. Identification of two groups of Mycobacterium paratuberculosis strains by restriction endonuclease analysis and DNA hybridization. J Clin Microbiol 1990;28:1591–1596.

9. Windsor PA. Paratuberculosis in sheep and goats. Vet Microbiol 2015;181:161–169.

10. Supply P, Mazars E, Lesjean S, Vincent V, Gicquel B, Locht C. Variable human minisatellite-like regions in the Mycobacterium tuberculosis genome. Mol Microbiol 2000;36:762–771.

11. Castellanos E, Romero B, Rodriguez S, et al. Molecular charac-terization of Mycobacterium avium subspecies paratuberculosis Types II and III isolates by a combination of MIRU-VNTR loci. Vet Microbiol 2010;144:118–126.

12. Thibault VC, Grayon M, Boschiroli ML, et al. New variable-number tandem-repeat markers for typing Mycobacterium avium subsp. paratuberculosis and M. avium strains: Comparison with IS900 and IS1245 restriction fragment length polymorphism typing. J Clin Microbiol 2007;45:2404–2410.

13. Sohal JS, Arsenault J, Labrecque O, et al. Genetic structure of Mycobacterium avium subsp. paratuberculosis population in cattle herds in Quebec as revealed by using a combination of multi-locus genomic analyses. J Clin Microbiol 2014;52:2764–2775.

14. Mobius P, Luyven G, Hotzel H, Kohler H. High genetic diversity among Mycobacterium avium subsp. paratuberculosis strains from German cattle herds shown by combination of IS900 restriction fragment length polymorphism analysis and mycobacterial interspersed repetitive unit-variable-number tandem-repeat typing. J Clin Microbiol 2008;46:972–981.

15. Biet F, Sevilla IA, Cochard T, et al. Inter- and intra-subtype genotypic differences that differentiate Mycobacterium avium subspecies paratuberculosis strains. BMC Microbiol 2012;12: 264.

16. Fritsch I, Luyven G, Kohler H, Lutz W, Mobius P. Suspicion of Mycobacterium avium subsp. paratuberculosis transmission between cattle and wild-living red deer (Cervus elaphus) by multi target genotyping. Appl Environ Microbiol 2012;78:1132–1139.

17. Verdugo C, Pleydell E, Price-Carter M, et al. Molecular epide-miology of Mycobacterium avium subsp. paratuberculosis isolated from sheep, cattle and deer on New Zealand pastoral farms. Prev Vet Med 2014;117:436–446.

18. Whittington RJ, Marsh IB, Saunders V, et al. Culture pheno-types of genomically and geographically diverse Mycobacterium avium subsp. paratuberculosis isolates from different hosts. J Clin Microbiol 2011;49:1822–1830.

19. Irenge LM, Walravens K, Govaerts M, et al. Development and validation of a triplex real-time PCR for rapid detection and specific identification of M. avium sub sp. paratuberculosis in faecal samples. Vet Microbiol 2009;136:166–172.

20. Debien E, Hélie P, Buczinski S, Leboeuf A, Bélanger D, Drolet R. Proportional mortality: A study of 152 goats submitted for necropsy from 13 goat herds in Quebec, with a special focus on caseous lymphadenitis. Can Vet J 2013;54:581–587.

21. Sevilla I, Singh SV, Garrido JM, et al. Molecular typing of Mycobacterium avium subspecies paratuberculosis strains from different hosts and regions. Rev Sci Tech 2005;24:1061–1066.

22. Singh SV, Sohal JS, Singh PK, Singh AV. Genotype profiles of Mycobacterium avium subspecies paratuberculosis isolates recov-ered from animals, commercial milk, and human beings in North India. Int J Infect Dis 2009;13:e221–227.

23. Overduin P, Schouls L, Roholl P, et al. Use of multilocus variable-number tandem-repeat analysis for typing Mycobacterium avium subsp. paratuberculosis. J Clin Microbiol 2004;42:5022–5028.

24. Bull TJ, Sidi-Boumedine K, McMinn EJ, Stevenson K, Pickup R, Hermon-Taylor J. Mycobacterial interspersed repetitive units (MIRU) differentiate Mycobacterium avium subspecies paratuber-culosis from other species of the Mycobacterium avium complex. Mol Cell Probes 2003;17:157–164.

25. Ronai Z, Csivincsik A, Gyuranecz M, Kreizinger Z, Dan A, Janosi S. Molecular analysis and MIRU-VNTR typing of Mycobacterium



avium subsp. paratuberculosis strains from various sources. J Appl Microbiol 2015;118:275–283.

26. Whittington RJ, Marsh I, Turner MJ, et al. Rapid detection of Mycobacterium paratuberculosis in clinical samples from ruminants and in spiked environmental samples by modified BACTEC 12B radiometric culture and direct confirmation by IS900 PCR. J Clin Microbiol 1998;36:701–707.

27. Sevilla I, Garrido JM, Geijo M, Juste RA. Pulsed-field gel elec-trophoresis profile homogeneity of Mycobacterium avium subsp. paratuberculosis isolates from cattle and heterogeneity of those from sheep and goats. BMC Microbiol 2007;7:18.

28. Mobius P, Fritsch I, Luyven G, Hotzel H, Kohler H. Unique genotypes of Mycobacterium avium subsp. paratuberculosis strains of Type III. Vet Microbiol 2009;139:398–404.

29. Motiwala AS, Li L, Kapur V, Sreevatsan S. Current understand-ing of the genetic diversity of Mycobacterium avium subsp. para-tuberculosis. Microbes Infect 2006;8:1406–1418.

30. de Juan L, Mateos A, Dominguez L, Sharp JM, Stevenson K. Genetic diversity of Mycobacterium avium subspecies paratuber-culosis isolates from goats detected by pulsed-field gel electro-phoresis. Vet Microbiol 2005;106:249–257.

31. Whittington RJ. Factors affecting isolation and identification of Mycobacterium avium subsp. paratuberculosis from fecal and tis-sue samples in a liquid culture system. J Clin Microbiol 2009;47: 614–622.

32. Ahlstrom C, Barkema HW, Stevenson K, et al. Limitations of vari-able number of tandem repeat typing identified through whole genome sequencing of Mycobacterium avium subsp. paratubercu-losis on a national and herd level. BMC Genomics 2015;16:161.

33. Bauman CA, Jones-Bitton A, Ahlstrom C, et al. Identification of Mycobacterium avium subspecies paratuberculosis strains isolated from dairy goats and dairy sheep in Ontario, Canada. Can J Vet Res 2017;81:304–307.


Article


I n t r o d u c t i o nStreptococcus uberis is a commensal microorganism of the bovine

gut (1) reported to be one of the main causes of bovine mastitis in many countries (2,3). In Canada, S. uberis is one of the most fre-quently isolated mastitis pathogens, especially in tie-stalls herds (4). Intramammary infections (IMIs) with S. uberis can produce subclinical and clinical mastitis events during the lactating and the non-lactating periods (5). Streptococcus uberis colonizes extra-mammary locations, including different body sites (6), straw bedding (7), and pasture (8). It has long been recognized that S. uberis can

have contagious and environmental modes of transmission (6,9). Although, S. uberis is often called an “environmental Streptococcus” (10), it is now recognized that contagious transmission may be the most common mode of transmission in some countries (3).

Different bacterial typing methods have been used to differenti-ate the sources of infections, including multilocus sequence typing (MLST) (3,11,12). Multilocus sequence typing can be used to cata-logue bacterial diversity and group sequence types (STs) in clonal complexes (CCs) of related isolates based on allelic profiles (12). Analysis of STs of S. uberis isolates allows us to gain an understand-ing of potential transmission for bovine mastitis. A herd with a lot of

Molecular variability of Streptococcus uberis isolates from intramammary infections in Canadian dairy farms from the Maritime region

Julian Reyes, Juan C. Rodriguez-Lecompte, Adam Blanchard, J.T. McClure, Javier Sánchez

A b s t r a c tThe primary objective of this study was to explore the variability of Streptococcus uberis (S. uberis) isolates by extracting multilocus sequence typing (MLST) data from whole-genome sequencing. The secondary objective was to determine the distribution of the phenotypic antimicrobial resistance (AMR) and the associated AMR genes as well as the virulence gene profiles among sequence types (STs). Sixty-two isolates were recovered from 16 herds in 3 Canadian Maritime Provinces: New Brunswick (14.5%), Nova Scotia (48.3%), and Prince Edward Island (37.1%). Of these, 9, 30, and 23 were recovered from post-calving, lactational samples, and post-mastitis samples, respectively. These 62 S. uberis isolates belonged to 34 STs; 11 isolates were typed to 9 known STs and 51 isolates were classified as one of 25 new STs. Thirteen isolates were part of major clonal complexes (CCs). Post-mastitis isolates contained 10 unique STs, lactational isolates contained 11 unique STs, and post-calving isolates had 3 STs. Each farm had only 1 isolate that was a unique ST except for STs 233, 851, 855, 857, 864, and 866, which were found in multiple cows per herd on more than one farm. ST851 and ST857 were found in each of the 3 sample types, with ST857 found in cows from all 3 Maritime provinces. These results indicate that S. uberis is a diverse non-clonal pathogen with specific STs residing in clonal clusters, carrying multiple AMR genes and virulence, with a diverse phenotypic AMR.

R é s u m éL’objectif premier de la présente étude était d’explorer la variabilité d’isolats de Streptococcus uberis en extrayant les données de typage génomique multilocus (MLST) du séquençage du génome entier. L’objectif secondaire était de déterminer la distribution des phénotypes de résistance antimicrobienne (AMR) et les gènes d’AMR associés ainsi que les profils des gènes de virulence parmi les types de séquence (STs). Soixante-deux isolats ont été obtenus de 16 troupeaux dans trois provinces maritimes canadiennes : le Nouveau-Brunswick (14,5 %), la Nouvelle-Écosse (48,3 %), et l’Ile-du Prince-Édouard (37,1 %). Parmi ces isolats, 9, 30, et 23 ont été obtenus d’échantillons post-vêlage, en lactation, et post-mammite, respectivement. Ces 62 échantillons appartenaient à 34 STs; 11 isolats ont été typés comme appartenant à 9 STs connus et 51 isolats ont été classifiés dans un des 25 nouveaux STs. Treize isolats faisaient partis de complexes clonaux majeurs (CCs). Les isolats post-mammite contenaient 10 STs uniques, les isolats de la période de lactation contenaient 11 STs uniques, et ceux de la période post-vêlage avaient 3 STs. Chaque ferme n’avait seulement qu’un isolat qui avait un ST unique sauf pour les STs 233, 851, 855, 857, 864, et 866 qui ont été retrouvés chez plusieurs vaches par troupeau sur plus d’une ferme. Les ST851 et ST857 ont été trouvés dans chacun des trois types d’échantillons, avec ST857 retrouvé dans des vaches des trois provinces maritimes. Ces résultats indiquent que S. uberis est un agent pathogène diversifié non-clonal avec des STs spécifiques résidant dans des groupements clonaux, arborant de multiples gènes de résistance antimicrobienne et de virulence, avec des phénotypes d’AMR variés.

(Traduit par Docteur Serge Messier)

Department of Health Management (Reyes, McClure, Sánchez) and Department of Pathology and Microbiology (Rodriguez-Lecompte), Atlantic Veterinary College, University of Prince Edward Island, Charlottetown, Prince Edward Island C1A 4P3; School of Veterinary Medicine and Science, University of Nottingham, Sutton Bonington Campus, Leicestershire LE12 5RD, England (Blanchard).

Address all correspondence to Dr. J. Reyes; telephone: 11-902-566-0995; fax: 11-902-566-0823; e-mail: [email protected]

Received April 18, 2018. Accepted June 8, 2018.



ST diversity, among it S. uberis isolates, indicates an environmental origin of IMIs, whereas ST homogeneity suggests a contagious transmission or point source of infection (13). Differences among S. uberis CCs have been previously reported to be associated with pathogenicity potential. For example, CC-5 has been associated with clinical mastitis, CC-143 with subclinical mastitis, and CC-86 with latent infections (14).

With the decreasing cost of high-throughput sequencing for bacte-rial genomes, MLST typing can be performed from whole-genome sequences (WGS) using available analytical pipelines (15). In addi-tion to bacterial typing, WGS identifies other genes of interest such as virulence genes and antimicrobial resistance (AMR) genes (16). The distribution of phenotypic and genotypic AMR in S. uberis isolates from cattle in the Maritime Provinces has been reported, including reduced susceptibility to penicillin and low susceptibil-ity to tetracyclines (10,17). Moreover, multiple mobile elements conferring AMR have been detected including those for resistance to erythromycin and lincomycin (18).

The primary objective of this study was to identify S. uberis STs using MLST from WGS data and determine their epidemiological distribution among dairy herds in the Maritime Provinces. A sec-ondary objective was to explore the distribution of phenotypic or genotypic AMR profiles as well as select virulence genes across STs in S. uberis isolates.


Population characteristicsFifty-eight cows from 16 dairy herds located in the Canadian

Maritime Provinces (Nova Scotia, New Brunswick, and Prince Edward Island) that had S. uberis isolated from an IMI were included in this analysis. Table I presents the distribution of cow and S. uberis isolate characteristics by herd and province. The isolates were recov-ered from different types of fresh milk samples, according to the fol-lowing collection criteria from the National Cohort of Dairy Farms: i) the lactational isolates were recovered from non-clinical cows from the first 2 wk of lactation until pre-dry off, ii) the post-calving samples were recovered between 0 and 14 d in milk, and iii) post-mastitis samples were collected between 2 and 5 wk after a recorded clinical mastitis event (10,19). From the total number of S. uberis isolates, 48.3% were recovered from lactational samples, 37.1% from post-mastitis samples, and 14.5% from post-calving samples.

Bacterial collectionThe S. uberis isolates were obtained from the pathogen collection

of the Canadian Bovine Mastitis Research Network cohort study database (19). Enviromental streptococci from this initial collection were used for a phenotypic study (10) and at least one bacterial species including S. uberis was targeted in each quarter and fresh milk sample type (10). Streptococcus uberis and S. dysgalactiae isolates were selected from this collection based on their resistant and pan susceptibility characteristics and used in a subsequent WGS AMR characterization study (17). Of these, 62 S. uberis isolates with phe-notypic and WGS information where included in the present report for further bacterial typing.

Microbiology and phenotypic resistance determination

Species identification of all the isolates was performed using a Bruker microflex MALDI-TOF MS System and Biotyper 3.0 soft-ware (Bruker Daltonik GmbH, Bremen, Germany), as reported by Cameron et al (10). Phenotypic antimicrobial susceptibility informa-tion for the isolates in the current study were part of an earlier study by Cameron et al (10), which determined the minimun inhibitory concentration (MIC) and isolate sensitivity for 8 antimicrobials relevant for mastitis treatment (ampicillin, ceftiofur, cephalothin, erythromycin, penicillin, penicillin/novobiocin combination, pirli-mycin, and tetracycline) using a Sensititre microdilution system and bovine mastitis MIC plate (CMV1AMAF; Trek Diagnostic Systems, Cleveland, Ohio, USA) (10).

Nucleic acid procedures and whole-genome sequencing

The DNA extraction was performed using the QIAamp DNA Mini Kit (Qiagen, Toronto, Ontario) and the DNA quantification was performed using the NanoDrop 1000 spectrophotometer (Thermo Scientific, Wilmington, Delaware, USA), as previously reported (17). The Illumina sequencing was performed at the British Columbia Genome Sciences Centre, using polymerase chain reaction free genome library construction, where different stages of library con-struction were completed on an Illumina HiSeq 2500 platform with an average coverage of 3673 (17). The WGS were assembled de novo using the Pathosystems Resource Integration Center (PATRIC) auto strategy option, which included 3 separate assemblies for base call-ing correction (BayesHammer), assembly (Velvet, Spades, IDBA), and scoring (ARAST), giving the best assembly solution. The annota-tion was conducted using the RAST tool kit PATRIC 3.2.9 (20). The genomes were compared to the Comprehensive Antibiotic Resistance Database (21) and the Antibiotic Resistance Genes Database (22). For the virulence genes, the genomes were compared with the virulence factor database and virulence factors from Michigan State University (23) within the PATRIC resource, using a sequence identity threshold of 90% for the sequence query of virulenge genes. The genomes are available via PATRIC.

Multilocus sequence typing analysisSequence types were extracted from the WGS data using the

MLST 1.6 tool (15) from the Center for Genomic Epidemiology (CGE) through the batch upload service (24) and the Bacterial Isolate Genome Sequence Database. Seven housekeeping genes (arc, ddl, gki, recP, tdk, tpi, and yqiL) were used to perform the MLST analysis. In order to determine the population structure of these isolates, an eBURst analysis was performed using the up-to-date MLST profiles database of the S. uberis database, representing 875 MLST profiles.

Statistical analysisDescriptive analyses were carried out using STATA 13 statistical

software to characterize STs, their phenotypic and genotypic AMR characteristics, and their virulence factors, as well as provide epide-miological data including sample time and distribution among cows, herds, and provinces. The eBURST representations of the phenotypic



characteristics of S. uberis isolates were generated using PHYLOVIZ 1.1 software (25).

Re s u l t s

Multilocus sequence typing and eBURST analysisOut of the 62 isolates, only 10 were classified as a previously

known STs; no ST was assigned to the remaining 52 isolates by the CGE tool. Furthermore, the WGS of the 62 isolates were submitted to the PubMLST sequence query resource and compared with the database, from which all the sequences were classified into known and new STs. Three isolates were assigned to a different ST by the

CGE resource in comparison with the PubMLST curation. Within this population of isolates, 34 STs were found. Eleven isolates belonged to 9 known STs and 51 isolates belonged to 25 new STs (ST850 to ST874). Table II shows the number of isolates by ST and their respective allelic profile and major CC. Thirteen isolates belonged to 3 major clonal clusters: CC5, CC86, and CC143. Of the remaining isolates, 37 belonged to smaller eBURST groups and 12 were singletons. Figure 1 depicts the population snapshot of S. uberis, including the isolates from the current study.

Distribution of STsTwenty-one isolates had unique STs across herds; however, in

5 herds, some specific STs (233, 851, 855, 857, 864, and 866) were

Table I. Number of isolates in all 34 sequence types (STs) and their respective allelic profile and clonal complex in 62 Streptococcus uberis isolates from Canadian dairy farms.

MLST allelic profile Locus Major clonal Number ofMLST ST arcc ddl gki recp tdk tpi yqil complex isolates 59 6 2 3 2 3 2 3 1233 1 1 2 2 2 1 3 ST-5 complex 3287 1 1 2 2 16 1 3 ST-5 complex 1351 3 2 3 2 3 2 3 ST-86 complex 1394 3 25 3 2 64 2 3 1446 5 1 27 2 3 1 3 1455 3 8 5 2 69 4 3 1810 3 28 41 2 3 4 3 1815 56 1 3 2 2 18 3 1850 3 1 4 1 3 2 3 1851 3 1 4 2 5 1 3 5852 3 2 3 4 9 4 3 ST-86 complex 1853 3 7 5 3 3 1 3 ST-86 complex 1854 3 1 37 2 13 1 2 1855 9 1 27 2 39 1 3 6856 25 1 27 2 17 2 3 ST-143 complex 1857 1 1 6 2 16 1 3 4858 6 24 5 19 16 1 3 2859 44 25 5 2 5 2 3 2860 9 1 4 2 4 1 3 2861 9 1 4 2 2 2 3 1862 9 28 27 2 39 1 3 1863 9 1 4 2 2 1 3 2864 2 1 6 2 17 2 3 ST-143 complex 4865 3 28 3 2 5 2 3 3866 4 1 43 2 2 1 3 4867 3 2 3 2 5 4 3 ST-86 complex 1868 3 1 3 2 3 2 3 1869 3 8 3 4 3 4 17 1870 9 1 4 2 4 2 3 2871 9 7 5 3 13 1 10 1872 1 1 37 2 1 1 6 2873 25 1 27 2 2 18 3 1874 1 1 5 2 1 1 3 1Total 62MLST — multilocus sequence typing.



present more than once either in the same cow or in different cows from the same herd (Table II). Of the 13 STs detected more than once, 11 were found in multiple herds. Moreover, 9 STs were found in more than one province. Figure 2 presents the distribution by sample type. Eleven STs were only present in lactation samples, 10 STs were only present in the post-mastitis samples, and 3 unique STs were exclu-sively present among post-calving samples. Only 2 STs were shared between all 3 sample types (851 and ST 857) and there were no STs shared between lactation and post-calving isolates.

Phenotypic antimicrobial resistanceThe distribution of AMR phenotypes is presented in Table III.

The most frequent single AMR among STs was for penicillin and the most common resistance phenotype to multiple antimicrobials was for ampicillin, penicillin, and tetracycline. The phenotypic resis-tance to at least 1 antimicrobial group was present in 31 of 34 STs. In 2 STs (855 and 874), ceftiofur and cephalexin resistance was present. Figure 3 represents the AMR profile of isolates with a minimum spanning tree. There were STs exhibiting phenotypic resistance to different antimicrobials ranging from 0 to 5 different products from the 8 antimicrobials tested.

Resistance and virulence gene distribution among STs

The supplementary data show the presence of virulence and AMR across STs. A frequent finding among STs was the presence of AMR genes, including the beta-lactam resistance genes Tem157 and bl2b,

the lincosamide resistance genes LnuB and LinB, and the tetracy-cline resistance gene TetM. Moreover, the combination of virulence genes purB, lepA, gidA, fba, Spy 1633, and ccpA, was conserved in the majority of STs.

D i s c u s s i o nStreptococcus uberis is an important udder pathogen in Canadian

dairy herds (19). It is a difficult pathogen to eradicate through tra-ditional management practices due to its ubiquity in the herd envi-ronment (9). Our results show that S. uberis was widely distributed among selected herds and had high strain variation from different sample categories. Environmental transmission can explain the strain heterogeneity within a herd (13). Contagious transmission of this pathogen has also been documented (26,27) and reports have shown the dominance of one particular strain persistently infecting the cows during lactation (28,29). In the present project, the high frequency of specific STs within a herd could be showing the dominance of a particular strain with contagious potential or a common source of transmission. Strain homogeneity is a consequence of contagious transmission; however, it is also a consequence of enviromental trasnsmission (as an alterative mechanim) (13) and can be the result of the transmission of one dominant strain present in the herd environment. In most herds in our study, there were multiple STs within a herd, with some STs isolated 2 and 3 times from different cows. These results are in agreement with other reports on the high diversity of strains determined by MLST195 sequence types (STs,

Table II. Distribution of the dairy population (cows and herd characteristics) of 62 Streptococcus uberis isolates recovered from 16 dairy herds in Canada’s Atlantic region.

Cow characteristics Isolate characteristics Mean Median Number of Province Herd Cow cow age1 cow DIM Cows isolates STNew Brunswick 401 1 2.0 153.0 1 1 858C

403 2 6.0 221.5 2 2 865B, 870BC

404 3 2.0 25.0 3 3 857BC, 865BC, 872BC

405 3 3.0 160.0 3 3 351, 850, 871

Nova Scotia 406 5 5.0 61.0 5 5 810, 859BC, 860B, 868, 873 407 8 2.0 29.5 8 8 59, 851AB, 854, 857ABC, 866ABC

408 5 2.0 58.0 5 5 233ABC, ST852, 853, 872BC

410 9 3.0 23.5 9 12 851A, 855ABC, 856, 860B, 861, 862, 863A

Prince Edward Island 411 1 1.0 102.0 1 1 857B

412 5 2.0 104.0 5 5 455, 858C, 865BC,866B, 870BC

413 1 7.0 246.0 1 1 874 414 2 3.5 98.0 2 2 864B, 869 415 3 2.0 1.0 3 3 287, 446, 866BC

416 5 4.0 120.0 5 5 233BC, 855AC, 859BC

417 4 2.0 346.0 4 5 394, 815, 864AB

418 1 4.0 8.0 1 1 8671 Age signifies the number of cow parturitions.A ST present in different cows.B ST present in different herds.C ST present in different provinces.DIM — days in milk of the cow; ST — sequence type.



with some indication of contagious events or the acquisition from the same environmental source (28).

The presence of STs 851 and 857 in all sample types indicates that these strains can cause infections across all stages of lactation and can originate from persistent infections. However, these isolates were not collected as follow-up samples from the same cow. In previous reports, different S. uberis STs were reported to cause persistent IMIs (3). The lack of STs shared between post-calving and lactation sam-ples could be caused by the limited sample size. The STs present in post-calving samples could be derived from environmental exposure, most S. uberis infections can occur in dry cows previous to calving

(30). Moreover, the majority of infections during the post-calving period could have originated from IMIs that occurred 2 wk prior to the calving or 2 wk before the post drying off period. Nonetheless, due to the limited number of on-farm isolates, the present results should be interpreted carefully.

The distribution of STs among Atlantic Provinces showed some more frequent STs among herds in each province, such as 851 in Nova Scotia, 864 Prince Edward Island, and 865 in New Brunswick. Other studies have also examined the geographic variation of the S. uberis population (31). Together, these observations on the dominance of particular STs within the herds of a province could be indicating potential regional inter-herd transmission.

The population structure was consistent with other reports that S. uberis does not follow a strict clonal structure (11,31). The CC groups found in our study were consistent with major CCs reported from the United Kingdom and New Zealand, specifically CC-5, CC-86, and CC-143 (31). However, most isolates in our study were not derived from these primary founders and were either classified as new STs or belonged to different CCs. The detection of a non-clonal population, in our case, suggests that most of the S. uberis isolated from IMIs were commensal and not a host or adapted organ (32).

The combination of individual genes (bl2b, Tem157, linb, lnuB, and TetM) was mainly present. Most STs exhibited phenotypic resistance to one or more antimicrobials. However, the genotypic resistance was more frequently present than the phenotypic resistance, potentially due to the lack of gene expression (33). This can be explained by the selection of isolates used in the present research, according to a previous environmental streptococci phenotypic study (10), which intended to deepen knowledge of phenotypic characteristics of AMR on S. uberis and S. dysgalactiae isolates. Streptococcus uberis’s capac-ity to develop penicillin resistance has been demonstrated through

Figure 1. Population snapshot of 62 Streptococcus uberis isolates in dairy herds from the Maritime region com-pared with known sequences from the PubMLST global database. Red squares represent new sequence types (STs).

Figure 2. Distribution of Streptococcus uberis sequence types (STs) per sample type in Canadian dairy herds from the Atlantic Provinces.

Lactation

Post-calvingPost-masttis

2 an

d 5

wee

ks a

fter

a

clin

ical

mas

titi

s ev

ent

0 and 14 days in milk

Lactational and pre-dry-off

10

11

3

ST-850

ST-815 ST-446 ST-874ST-871

ST-287 ST-351 ST-394ST-873

ST-59 ST-870ST-856

ST-867ST-861

ST-862

ST-455 ST-810 ST-852

ST-853 ST-854 ST-858

ST-859

ST-854 ST-869

ST-868

ST-863

ST-872 ST-866

ST-864ST-855

ST-865ST-233

ST-860

ST-851

ST-857

ST-858



a penicillin-binding protein-mediated mechanism (34); therefore, the phenotypic AMR could be caused by the usage of penicillin in these herds. The horizontal gene transfer of AMR genes can occur between environmental bacteria including bacteria of different genus (35). Previous reports in Canadian dairy herds have shown that IMI treatment with penicillin and first generation cephalosporin based

products was frequently used to treat clinical mastitis (36). Other reports have also shown the resistance of S. uberis to erythromycin, tetracycline (37), aminoglycoside, macrolide-lincosamide (MLS) (38), streptomycin, and enrofloxacin antimicrobials (39). This may reflect the fact that MLS, tetracycline, and other combinations of products have been used widely in dairy cattle for the treatment of mastitis and other diseases (36).

The combination of virulence genes purB, lepA, gidA, fba, Spy 1633, and ccpA, was present in most STs, showing that this multigene arrangement is conserved among S. uberis, independent of the strain variation, and could be a species characteristic. Similar to another study (32), there were no S. uberis clones associated with increased pathogenicity characteristics. Previous reports have shown that a “complex multigene arrangement” can influence infection potential or clinical outcome, more than the presence or absence of a particular virulence gene (3,40). However, the present study did not include any host-pathogenicity indicator. The population structure of the present S. uberis isolates was in agreement with other studies about the non-clonal structure of this pathogen, with some specific STs showing some degree of clonality. Additionally, S. uberis isolates showed high variability among dairy herds in Atlantic provinces. This could be a reflection on the diverse origins of S. uberis isolates (from different cows and herds) that were not part of the same chain of transmission, or were not epidemiologically related.

A c k n o w l e d g m e n t sThe bacterial isolates were provided by the Canadian Bovine

Mastitis and Milk Quality Research Network. We would like to acknowledge Dr. Ruth Zadoks for the scientific review of the paper. We would also like to acknowledge Dr. Derek Price for the assistance with the graphic material of the paper and Bill Chalmers for editing the manuscript. Preliminary results of this research were presented at the 98th Annual Conference of Research Workers in Animal Diseases in Chicago, Illinois, in December 2017.

Table III. Frequency of the isolates tested for phenotypic antimicrobial resistance and the distribution of the phenotypes classified according to the minimum inhibitory concentrations.

ST Antimicrobial phenotype Number of isolates/ST59 Pen1 Pirl1 1233 Amp1 Pen1 Tet1 3287 Sensible 1351 Pen1 1394 Amp1 Pen1 Tet1 1446 Amp1 Pen1 Tet1 1455 Pen1 Tet1 1810 Pen1 1815 Amp1 Pen1 Tet1 1850 Amp1 Pen1 Pnov1 1851 Amp1 Pen1 Pnov1 Ery1 Pirl1 5852 Sensible 1853 Pen1 1854 Amp1 Pen1 1855 Amp1 Pen1 Cep1 Pirl1 Tet1 6856 Pen1 Pirl1 Ery1 Tet1 1857 Amp1 Pen1 Tet1 4858 Sensible 2859 Tet1 2860 Pen1 2861 Pen1 1862 Pen1 Tet1 1863 Amp1 Pen1 Pnov1 2864 Amp1 Pen1 Ery1 Pirl1 Tet1 4865 Amp1 Pen1 3866 Amp1 Pen1 Tet1 4867 Pen1 1868 Sensible 1869 Pen1 Tet1 1870 Amp1 Pen1 Pirl1 2871 Amp1 Pen1 1872 Pen1 2873 Tet1 1874 Amp1 Pen1 Xnl1 1Total 62ST — sequence type; Pen — penicillin; Pirl — Pirlymicin; Amp — ampicillin; Tet — tetracycline; Pnov — Penicillin-Novobiocin; Ery — Erythromycin; Cep — Cephalexin; Xnl — Ceftiofur.

Figure 3. Distribution of the phenotypic resistance of 62 Streptococcus uberis sequence types (STs) to different antimicrobial products (1 to 5) using the full MLST goEBURST distance to represent the relationship between STs.

864

871

856

815

446 862 850

853 394852

866

851

855

865863

873

455

869867

810

59861

854

872

874

233

351868

860

870

859 0

1

2

3

4

5

287 858

857



Supplementary data. Presence and absence of antimicrobial resistance and virulence genes in the whole genome sequences by sequence types on 62 Streptococcus uberis isolates recovered from dairy herds from the Atlantic region of Canada grouped by sequence type (ST).

ST

Antimicrobial resistance genes Virulence genes

tem

1

tem

127

tem

136

tem

157

tem

163

tem

47

tem

71

tem

89

tem

95

bl2b

linB

lnuB

tets

tetm

SP_0

121

SP_0

251

SP_0

494

SP_0

829

SPy_

1633

ccpA

fba

gidA

guaA

hasC

lepA

leuS

purB

59

233

233

233

287

351

394

446

455

810

815

850

851

851

851

851

851

852

853

854

855

855

855

855

855

855

856

857

857

857

857

858

858

859

859

860



Re f e r e n c e s 1. Zadoks RN, Middleton JR, McDougall S, Katholm J, Schukken YH.

Molecular epidemiology of mastitis pathogens of dairy cattle and comparative relevance to humans. J Mammary Gland Biol Neoplasia 2011;16:357–372.

2. Zadoks R, Fitzpatrick J. Changing trends in mastitis. Ir Vet J 2009; 62:S59–70.

3. Davies PL, Leigh JA, Bradley AJ, Archer SC, Emes RD, Green MJ. Molecular epidemiology of Streptococcus uberis clinical mastitis in dairy herds: Strain heterogeneity and transmission. J Clin Microbiol 2016;54:68–74.

4. Olde Riekerink RG, Barkema HW, Kelton DF, Scholl DT. Incidence rate of clinical mastitis on Canadian dairy farms. J Dairy Sci 2008;91:1366–1377.

5. Leigh JA. Streptococcus uberis: A permanent barrier to the control of bovine mastitis? Vet J 1999;157:225–238.

6. Zadoks RN, Tikofsky LL, Boor KJ. Ribotyping of Streptococcus uberis from a dairy’s environment, bovine feces and milk. Vet Microbiol 2005;109:257–265.

7. Ericsson Unnerstad H, Lindberg A, Persson Waller K, et al. Microbial aetiology of acute clinical mastitis and agent-specific risk factors. Vet Microbiol 2009;137:90–97.

Supplementary data. Presence and absence of antimicrobial resistance and virulence genes in the whole genome sequences by sequence types on 62 Streptococcus uberis isolates recovered from dairy herds from the Atlantic region of Canada grouped by sequence type (ST) (continued).

ST

Antimicrobial resistance genes Virulence genes

tem

1

tem

127

tem

136

tem

157

tem

163

tem

47

tem

71

tem

89

tem

95

bl2b

linB

lnuB

tets

tetm

SP_0

121

SP_0

251

SP_0

494

SP_0

829

SPy_

1633

ccpA

fba

gidA

guaA

hasC

lepA

leuS

purB

860

861

862

863

863

864

864

864

864

865

865

865

866

866

866

866

867

868

869

870

870

871

872

872

873

874



8. Lopez-Benavides MG, Williamson JH, Pullinger GD, Lacy-Hulbert SJ, Cursons RT, Leigh JA. Field observations on the variation of Streptococcus uberis populations in a pasture-based dairy farm. J Dairy Sci 2007;90:5558–5566.

9. Zadoks RN, Gillespie BE, Barkema HW, Sampimon OC, Oliver SP, Schukken YH. Clinical, epidemiological and molecular characteristics of Streptococcus uberis infections in dairy herds. Epidemiol Infect 2003;130:335–349.

10. Cameron M, Saab M, Heider L, McClure JT, Rodriguez-Lecompte JC, Sanchez J. Antimicrobial susceptibility patterns of environ-mental streptococci recovered from bovine milk samples in the Maritime provinces of Canada. Front Vet Sci 2016;3:1–14.

11. Coffey TJ, Pullinger GD, Urwin R, et al. First insights into the evolution of Streptococcus uberis: A multilocus sequence typing scheme that enables investigation of its population biology. Appl Environ Microbiol 2006;72:1420–1428.

12. Maiden MC, Jansen van Rensburg MJ, Bray JE, et al. MLST revisited: The gene-by-gene approach to bacterial genomics. Nat Rev Microbiol 2013;11:728–736.

13. Klaas IC, Zadoks RN. An update on environmental mastitis: Challenging perceptions. Transbound Emerg Dis 2017;65:166–185.

14. Tomita T, Meehan B, Wongkattiya N, et al. Identification of Streptococcus uberis multilocus sequence types highly associated with mastitis. Appl Environ Microbiol 2008;74:114–124.

15. Larsen MV, Cosentino S, Rasmussen S, et al. Multilocus sequence typing of total-genome-sequenced bacteria. J Clin Microbiol 2012; 50:1355–1361.

16. Quainoo S, Coolen JPM, van Hijum SAFT, et al. Whole-genome sequencing of bacterial pathogens: The future of nosocomial outbreak analysis. Clin Microbiol Rev 2017;30:1015–1063.

17. Vélez JR, Cameron M, Rodríguez-Lecompte JC, et al. Whole-genome sequence analysis of antimicrobial resistance genes in Streptococcus uberis and Streptococcus dysgalactiae isolates from Canadian dairy herds. Front Vet Sci 2017;4:1–11.

18. Haenni M, Saras E, Bertin S, Leblond P, Madec J-Y, Payot S. Diversity and mobility of integrative and conjugative elements in bovine isolates of Streptococcus agalactiae, S. dysgalactiae subsp. dys-galactiae, and S. uberis. Appl Environ Microbiol 2010;76:7957–7965.

19. Reyher KK, Dufour S, Barkema HW, et al. The National Cohort of Dairy Farms — A data collection platform for mastitis research in Canada. J Dairy Sci 2011;94:1616–1626.

20. Wattam AR, Davis JJ, Assaf R, et al. Improvements to PATRIC, the all-bacterial Bioinformatics Database and Analysis Resource Center. Nucleic Acids Res 2017;45:D535–D542.

21. McArthur AG, Waglechner N, Nizam F, et al. The Comprehensive Antibiotic Resistance Database. Antimicrob Agents Chemother 2013;57:3348–3357.

22. Liu B, Pop M. ARDB — Antibiotic Resistance Genes Database. Nucleic Acids Res 2009;37:D443–D447.

23. Mao C, Abraham D, Wattam AR, et al. Curation, integration and visualization of bacterial virulence factors in PATRIC. Bioinformatics 2015;31:252–258.

24. Thomsen MC, Ahrenfeldt J, Cisneros JL, et al. A bacterial analysis platform: An integrated system for analysing bacterial whole genome sequencing data for clinical diagnostics and surveil-lance. PLoS One 2016;11:e0157718.

25. Francisco AP, Vaz C, Monteiro PT, Melo-Cristino J, Ramirez M, Carriço JA. PHYLOViZ: Phylogenetic inference and data visual-ization for sequence based typing methods. BMC Bioinformatics 2012;13:87.

26. Peeler EJ, Green MJ, Fitzpatrick JL, Morgan KL, Green LE. Risk factors associated with clinical mastitis in low somatic cell count British dairy herds. J Dairy Sci 2000;83:2464–2472.

27. Zadoks RN, Allore HG, Barkema HW, Sampimon OC, Gröhn YT, Schukken YH. Analysis of an outbreak of Streptococcus uberis mastitis. J Dairy Sci 2001;84:590–599.

28. Phuektes P, Mansell PD, Dyson RS, Hooper ND, Dick JS, Browning GF. Molecular epidemiology of Streptococcus uberis isolates from dairy cows with mastitis. J Clin Microbiol 2001;39:1460–1466.

29. Pullinger GD, Coffey TJ, Maiden MC, Leigh JA. Multilocus-sequence typing analysis reveals similar populations of Streptococcus uberis are responsible for bovine intramammary infections of short and long duration. Vet Microbiol 2007;119: 194–204.

30. Todhunter DA, Smith KL, Hogan JS. Environmental streptococcal intramammary infections of the bovine mammary gland. J Dairy Sci 1995;78:2366–2374.

31. Pullinger GD, López-Benavides M, Coffey TJ, et al. Application of Streptococcus uberis multilocus sequence typing: Analysis of the population structure detected among environmental and bovine isolates from New Zealand and the United Kingdom. Appl Environ Microbiol 2006;72:1429–1436.

32. Zadoks RN, Schukken YH, Wiedmann M. Multilocus sequence typing of Streptococcus uberis provides sensitive and epide-miologically relevant subtype information and reveals positive selection in the virulence gene pauA. J Clin Microbiol 2005; 43:2407–2417.

33. Gao J, Yu FQ, Luo LP, et al. Antibiotic resistance of Streptococcus agalactiae from cows with mastitis. Vet J 2012;194:423–424.

34. Haenni M, Galofaro L, Ythier M, et al. Penicillin-binding pro-tein gene alterations in Streptococcus uberis isolates present-ing decreased susceptibility to penicillin. Antimicrob Agents Chemother 2010;54:1140–1145.

35. Aminov RI. Horizontal gene exchange in environmental micro-biota. Front Microbiol 2011;2:1–19.

36. Saini V, McClure JT, Léger D, et al. Antimicrobial use on Canadian dairy farms. J Dairy Sci 2012;95:1209–1221.

37. Pitkälä A, Koort J, Björkroth J. Identification and antimicrobial resistance of Streptococcus uberis and Streptococcus parauberis iso-lated from bovine milk samples. J Dairy Sci 2008;91:4075–4081.

38. Rato MG, Bexiga R, Florindo C, Cavaco LM, Vilela CL, Santos-Sanches I. Antimicrobial resistance and molecular epidemiology of streptococci from bovine mastitis. Vet Microbiol 2013;161: 286–294.

39. Petrovski KR, Grinberg A, Williamson NB, et al. Susceptibility to antimicrobials of mastitis-causing Staphylococcus aureus, Streptococcus uberis and Str. dysgalactiae from New Zealand and the USA as assessed by the disk diffusion test. Aust Vet J 2015;93:227–233.

40. Hossain M, Egan SA, Coffey T, et al. Virulence related sequences; Insights provided by comparative genomics of Streptococcus uberis of differing virulence. BMC Genomics 2015;16:1–13.


Article


I n t r o d u c t i o nNon-typhoidal Salmonella spp. (Salmonella) are estimated to be

the fourth leading cause of enteric illness in Canada (1). Pigs are a potential source for human infection and the emergence of multi-drug resistant strains of Salmonella in pigs presents an increased public health concern (2,3). Pigs are often subclinical Salmonella carriers and may shed bacteria during periods of stress, such as weaning, thus promoting transmission among pigs (4,5). One of the most common serotypes of Salmonella on Canadian swine farms, Salmonella Typhimurium (3,6,7), is also a commonly reported cause of salmonellosis in humans (8,9). Previous studies in Europe have shown that public health risks can be mitigated through pre-harvest reductions of Salmonella in swine (10).

Flavophospholipol, a phosphoglycolipid antimicrobial agent produced by Streptomyces species (11,12), may have the ability to reduce Salmonella shedding and colonization in pigs. It functions by hindering bacterial cell wall synthesis through the inhibition of transglycolase activity, therefore functioning predominately against Gram-positive bacteria (11,13,14). Flavophospholipol is not as effec-tive against Gram-negative bacteria because of its inability to reach target intracellular elements (13,15,16). Despite that, studies have shown some activity against members of the Enterobacteriaceae family,

including Salmonella and Escherichia coli (17–20). This is presumed to be a result of increased susceptibility to flavophospholipol in Gram-negative bacteria containing R-plasmids, in conjunction with a speculated ability to enter the bacterial cells via sex pili and pilin protein precursors (15). As a result, flavophospholipol may alter the microflora in favor of beneficial bacteria and decrease available intestinal binding sites or reduce intestinal pH, leading to inhibition of Salmonella colonization (17,19,21).

Previous studies have found flavophospholipol effective in reduc-ing Salmonella (17,18). A recent study by Nair et al (22), however, found that flavophospholipol was ineffective in reducing Salmonella shedding in naturally infected grower-finisher pigs, although it may be more effective if applied at an earlier stage in pig production. The objective of this study was to investigate Salmonella shedding and colonization as well as antibody response to Salmonella in weaned pigs receiving 4 parts per million (ppm) flavophospholipol in feed compared with control pigs.

M a t e r i a l s a n d m e t h o d sThe project was approved by the Animal Care Committee of

the University of Guelph, in accordance with the guidelines of the Canadian Council of Animal Care.

Salmonella shedding and seropositivity and its association with in-feed flavophospholipol in nursery pigs

Saranya Nair, Jane Newman, Abdolvahab Farzan, Robert M. Friendship

A b s t r a c tThe objective of this study was to assess the impact of in-feed flavophospholipol on Salmonella shedding and antibody response in nursery pigs. Weaned pigs were fed either a diet containing 4 ppm flavophospholipol (n = 16) or a non-medicated feed (n = 16) for 36 d. All pigs were orally challenged with a 2-mL dose of 108 colony-forming units (CFUs)/mL of Salmonella Typhimurium on Days 7 and 8 of the trial. On Day 36, all pigs were euthanized and samples were collected from the liver, spleen, and ileocecal lymph nodes. Fecal and tissue samples were quantitatively cultured for Salmonella and serum samples were tested for the presence of the Salmonella antibody by enzyme-linked immunosorbent assay (ELISA). There was no difference between the 2 groups in antibody response and the presence of Salmonella in feces and tissue (P . 0.05). Medicating nursery diets with flavophospholipol at 4 ppm did not appear to reduce Salmonella infection in nursery pigs.

R é s u m éL’objectif de la présente étude était d’évaluer l’impact de l’ajout de flavophospholipol dans l’aliment sur l’excrétion de Salmonella et la réponse en anticorps chez des porcs en pouponnière. Des porcs sevrés ont été nourris avec soit une diète contenant 4 ppm de flavophospholipol (n = 16) ou une diète non-médicamentée (n = 16) pendant 36 j. Tous les porcs ont reçu oralement une dose de 2 mL de 108 unités formatrices de colonies (UFC)/mL de Salmonella Typhimurium aux Jours 7 et 8 de l’essai. Au Jour 36, tous les porcs ont été euthanasiés et on préleva des échantillons de foie, rate, et nœuds lymphatiques iléo-caecaux. Des échantillons de fèces et de tissus ont été cultivés pour quantifier le nombre de Salmonella et des échantillons de sérum furent testés pour la présence d’anticorps contre Salmonella par épreuve immuno-enzymatique (ELISA). Il n’y avait pas de différence entre les deux groupes quant à la réponse en anticorps et la présence de Salmonella dans les fèces et les tissus (P . 0,05). L’ajout de 4 ppm de flavophospholipol à la diète en pouponnière ne semble pas réduire l’infection par Salmonella chez les porcs en pouponnière.


Department of Population Medicine, Ontario Veterinary College, University of Guelph, 50 Stone Road East, Guelph, Ontario N1G 2W1.

Address all correspondence to Saranya Nair; telephone: (905) 912-4927; fax: (519) 763-3117; e-mail: [email protected]

Received February 15, 2018. Accepted June 18, 2018.



Pigs and sample collectionThe trial was conducted in the isolation unit at the Ontario

Veterinary College, University of Guelph. Four-week-old pigs (N = 32), obtained from the Arkell Swine Research Facility in Guelph, were randomly assigned to 1 of 4 separate rooms (8 pigs per room) (Day 0). Pigs in 2 rooms (treatment group) received 4 ppm in-feed flavophospholipol (Flavomycin; Huvepharma, Mitchell, Ontario, Canada) from Day 1, 24 h after arriving at the isolation unit, until end of trial (Day 36). Pigs in the other 2 rooms (control) were fed an identical ration, but without flavophospholipol. All pigs were orally challenged with a 2-mL dose of 108 colony-forming units (CFUs)/mL of Salmonella Typhimurium DT 104, with resistance to nalidixic acid, on Day 7 and Day 8. Fecal samples were collected from individual pigs before challenge on Days 0 and 6 and after challenge on Days 8, 9, 12, 14, 19, 21, 26, 28, and 36. Blood samples were collected at the same time except on Day 26. At Day 36, the pigs were euthanized and samples were collected from the liver, ileocecal lymph node, and spleen.

Salmonella isolation and colony-forming unitsAll fecal and tissue samples were cultured for Salmonella and

colony-forming units/gram of feces and tissue were determined. Fecal and tissue samples were enriched in tetrathionate broth (TTB) (Becton Dickinson, Sparks Glencoe, Maryland, USA) in a 1-to-9 ratio, homogenized for 30 s with a Seward Stomacher 400 Circulator (Seward, Norfolk, England), and incubated at 37°C for 18 to 24 h. Then, 0.1 mL of TTB was transferred to a 9.9-mL Rappaport-Vassiliadis Broth (RVB) tube (Becton Dickinson) and incubated at 41°C for 18 to 24 h. Lastly, a loop-full (10 mL) of RVB was plated on Xylose Lactose Tergitol 4 Agar (Remel XLT-4; Thermo Fisher Scientific, Lenexa, Kansas, USA) containing nalidixic acid and incubated at 37°C for 18 to 24 h to 72 h. In addition, 1 g of feces was diluted with 0.1% buffered peptone water (BPW) (Becton Dickinson) to make serial dilutions. Dilutions of 1021, 1022, and 1023 were plated on XLT-4 agar and Brilliant Green Agar (BGA) (Thermo Fisher Scientific), both containing nalidixic acid. The BGA and XLT-4 plates were incubated at 37°C for 24 h and 24 to 72 h, respectively and the number of colonies per plate was counted. Fecal samples collected on Days 0 and 6, i.e., before challenge, were cultured using the same method, except the samples were plated on XLT-4 agar without nalidixic acid.

Salmonella antibody detectionAn indirect enzyme-linked immunosorbent assay (ELISA) (Pigtype

Salmonella Ab Kit; QIAGEN, Leipzig, Germany) was used to deter-mine the level of antibody to Salmonella as per the kit manual. A sample-to-positive ratio (S/P) value was determined using the equation provided by the kit manufacturer. Samples with S/P ratio of 0.3 were considered seropositive based on the QIAGEN pigtype Salmonella Ab kit.

Statistical analysisData were entered into Microsoft Excel for Mac 2011, Version

14.5.5 (Microsoft, Redmond, Washington, USA) and, after cleaning, were imported to Stata/SE 14.1 for Mac (StataCorp, College Station,

Texas, USA). A mixed-effects linear regression method with “pig” (repeated measurements) as a random effect was used to compare Salmonella CFU (fecal and tissue samples), antibody titer (S/P ratio), and rectal temperature in both the treatment and control groups. In addition, a mixed-effects logistic regression method with “pig” as a random effect was used to compare seropositivity (yes/no) in pigs in both the treatment and control groups.

Re s u l t sExcept for 1 pig in the treatment group that tested positive for

Salmonella by culture on Day 0, all pigs tested negative on Day 0 and Day 6, i.e., before challenge. During the post-challenge period, on Days 8, 9, 12, 14, 19, 21, 26, 28, and 36, Salmonella was recovered from all fecal samples collected from pigs in both groups, except for 1 pig in the treatment group that tested negative on Days 14 and 28. No significant difference was found in colony-forming units/gram of Salmonella in feces of pigs receiving flavophospholipol compared to pigs receiving the non-medicated diet post-challenge (Figure 1).

In the 31 pigs euthanized, Salmonella was isolated from 7 tissue samples collected from 5 pigs (16%). These samples were from the lymph nodes (2 samples from control and 2 from treatment group), the liver (1 sample from control group), and the spleen (1 sample from control and 1 sample from treatment group). Salmonella was recovered in all 3 types of tissue collected from 1 control pig. No significant difference was found in colony-forming units/gram of Salmonella recovered from lymph node, liver, and spleen tissue samples of pigs in the treatment group and those in the control group (P . 0.05).

One control pig that was lethargic and had a markedly swol-len and firm hock was euthanized on Day 14 of the trial. During the histopathology examination conducted at the Animal Health Laboratory, University of Guelph, this pig had a diffuse, marked hemorrhagic enteritis, bursitis, and multifocal lung abscesses, with mild pneumonia. Moderate quantities of Salmonella and Trueperella pyogenes were isolated from lung tissue and large quantities of Trueperella pyogenes were isolated from the swollen hock. This pig was Salmonella seronegative at each stage of testing. Another control pig that was diagnosed with mild bronchopneumonia and mesen-teric torsion/volvulus in the histopathology examination died at Day 24. This pig was Salmonella seropositive on Days 1 and 6, i.e., before challenge, but seronegative during the post-challenge period.

Salmonella seropositivity in the medicated and control pigs is shown in Figure 2. Although 66% of pigs were seropositive on Day 1 of the trial, no significant difference in Salmonella antibody response (S/P ratio) and seropositivity was found on any sampling day among pigs in the treatment and control groups. Only 1 treated pig and 4 control pigs were Salmonella seropositive during the entire trial.

D i s c u s s i o nThe findings from this trial indicate that feeding a diet containing

4 ppm of flavophospholipol over a span of 36 d had no impact on the prevalence of Salmonella shedding and antibody response in nursery pigs. The impact of flavophospholipol on Salmonella shedding in pigs and broiler chickens has previously been investigated, but the



findings were inconsistent (4,17,18,20,22). For example, in 1 study, Salmonella shedding decreased in pigs fed 4 ppm flavophospholipol over 49 d and challenged with a 2.5 3 1011 CFUs/mL of Salmonella at the same time as they started on the diet (18). In another study, Salmonella shedding was reduced after 6 wk in broiler chickens receiving a higher dose of flavophospholipol (9 ppm) and challenged with Salmonella enteritidis (17).

On the other hand, Letellier et al (4) placed 12-day-old pigs on a diet containing 0.5 ppm of in-feed flavophospholipol for 14 d before challenging them with 107 CFUs/mL of Salmonella Typhimurium and found the treated pigs showed a reduction in Salmonella Typhimurium for only a few days post-challenge and no significant difference in Salmonella shedding (4). In a more recent study, no difference was seen in Salmonella shedding during the 10 wk that finisher pigs were fed 4 ppm of flavophospholipol compared to control pigs, although this could have been due to the presence of multiple Salmonella serotypes in these naturally infected pigs (22). These contradictory findings may in part be due to dif-ferent dosages and feeding durations, different animal species and challenge strains, or the presence of multiple Salmonella serotypes in naturally infected pigs, as well as the length of post-challenge monitoring (4,17,18,20,22).

In countries that approve flavophospholipol for use in swine feed, the label states that dosages of 2 to 4 ppm in-feed will improve feed efficiency and increase rate of weight gain in grower-finisher pigs, but does not mention reducing Salmonella. Based on this trial, dosages of 4 ppm do not appear to decrease Salmonella infection in weanling pigs, although a higher dose or a longer period of use may produce a different result. Further research is therefore required.

It should be noted that two-thirds of the nursery pigs herein were seropositive, probably due to maternal Salmonella antibodies. In terms of the Salmonella challenge, however, this passive immunity in the early nursery stage did not protect these pigs against the challenge strain. The high levels of Salmonella Typhimurium DT 104 shedding in pigs post-challenge indicate the dominance of the strain and the effectiveness of the challenge protocol.

In the present study, Salmonella was recovered from at least 1 sample collected from the liver, spleen, and ileocecal lymph nodes in only 3 controls and 2 treated pigs. Salmonella may be able to colonize in the gut when there is a reduced colonization resistance by commensal bacteria, normally found in the healthy gut, for essential nutrients and mucosal surfaces (23). Internal

colonization of Salmonella depends on gut microbiota, genetics, and the challenge protocol (dosage and strain) (23). Although passive immunity seemed to be depleted by Day 12 of the trial, it may have protected the pigs from further colonization of the challenge strain in the tissues.

Consistent with previous studies, the present study showed that serology lags behind the bacterial culture results for shedding during Salmonella infection (24–26). Two pigs were positive for Salmonella shedding at all sampling periods, but were never seropositive. This could be due to laboratory error or the development of disease tolerance. Kogut and Arsenault (27) hypothesized that in poultry, after initial Salmonella infection in the caecum, which can minimize host defenses, Salmonella may have the ability to initiate an intense reprogramming of the immune and metabolic systems that alters the host’s defense to disease tolerance.

In conclusion, feeding nursery-aged pigs a diet containing 4 ppm flavophospholipol did not reduce Salmonella shedding and Salmonella antibody response compared to controls. It would have been ben-eficial, however, to have another control group that was not treated with the medicated feed and not challenged with Salmonella in order to assess the changes in the basal level of Salmonella shed-ding and antibody level throughout the study period. In addition, although all the pigs were Salmonella negative on Day 6, 1 pig was found Salmonella positive on Day 0. This presented the possibility of Salmonella co-infection among pigs that went unobserved from Day 0 to Day 6. Furthermore, conducting Salmonella serotyping on the isolates found on Day 0 and those from tissue samples would have been vital in identifying both existing serotypes in the popu-lation and whether the tissue was colonized with the Salmonella challenge strain. The selective recovery of the Salmonella challenge strain introduced diagnostic bias by limiting identification of any other existing Salmonella infection in this population.

Despite these limitations, this study identified the inability of flavophospholipol to prevent Salmonella infection when adminis-tered at an early stage in pig production. Further studies are needed, however, to determine whether a higher concentration of flavophos-pholipol would protect pigs against Salmonella infection.

A c k n o w l e d g m e n t sFunding for this study was provided by the Food Safety

Research Program of the Ontario Ministry of Agriculture, Food and

Figure 1. Salmonella colony-forming units (CFUs)/gram of feces in 32 nursery pigs challenged orally with Salmonella Typhimurium DT 104 on Days 7 and 8 fed a medicated diet with 4 ppm flavophospholipol (treat-ment) or non-medicated diet (control) from Day 1 to Day 36.

Day of sampling

CFU

/g (l

og10

)

6.00

5.00

4.00

3.00

2.00

1.00

0.00

Flavophospholipol Control

9 12 14 19 21 26 28

Figure 2. Salmonella seropositivity in 32 nursery pigs challenged orally with Salmonella Typhimurium DT 104 on Days 7 and 8 fed a medicated diet with 4 ppm flavophospholipol (treatment) or a non-medicated diet (control) for the duration of the trial (Day 1 to Day 36).

100%

80%

60%

40%

20%

0%

Ser

opos

itive

pig

s

1 6 8 9 12 14 21 28 36

Day of sampling

Flavophospholipol Control



Rural Affairs (OMAFRA), Swine Innovation Porc, Huvepharma Animal Health, and the University of Guelph-OMAFRA Research Partnership.

Re f e r e n c e s 1. Thomas M, Murray R, Flockhart L, et al. Estimates of the bur-

den of foodborne illness in Canada for 30 specified pathogens and unspecified agents, circa 2006. Foodborne Pathog Dis 2013; 10:639–648.

2. Farzan A, Friendship RM, Dewey CE, Poppe C, Funk J. Evalua-tion of the risk factors for shedding Salmonella with or without antimicrobial resistance in swine using multinomial regression method. Zoonoses Public Health 2010;57:85–93.

3. Farzan A, Friendship RM, Cook A, Pollari F. Occurrence of Salmonella, Campylobacter, Yersinia enterocolitica, Escherichia coli O157 and Listeria monocytogenes in swine. Zoonoses Public Health 2010;57:388–396.

4. Letellier A, Messier S, Lessard L, Quessy S. Assessment of vari-ous treatments to reduce carriage of Salmonella in swine. Can J Vet Res 2000;64:27–31.

5. Nollet N, Houf K, Dewulf J, et al. Distribution of Salmonella strains in farrow-to-finish pig herds: A longitudinal study. J Food Prot 2005;68:2012–2021.

6. Wilkins W, Rajic A, Waldner C, et al. Distribution of Salmonella serovars in breeding, nursery, and grow-to-finish pigs, and risk factors for shedding in ten farrow-to-finish swine farms in Alberta and Saskatchewan. Can J Vet Res 2010;74:81–90.

7. Farzan A, Friendship RM, Dewey CE, Muckle AC, Gray JT, Funk J. Distribution of Salmonella serovars and phage types on 80 Ontario swine farms in 2004. Can J Vet Res 2008;72:1–6.

8. Public Health Agency of Canada. Canadian Antimicrobial Resistance Surveillance System — Report 2015. Available from: http://researchid.com/wp-content/uploads/2018/02/2015_antimicrobial-surveillance-antimicrobioresistance-eng.pdf Last accessed March 21, 2019.

9. Butaye P, Michael GB, Schwarz S, Barrett TJ, Brisabois A, White DG. The clonal spread of multidrug-resistant non-typhi Salmonella serotypes. Microbes Infect 2006;8:1891–1897.

10. Wegener HC, Hald T, Wong LF, et al. Salmonella control programs in Denmark. Emerg Infect Dis 2003;9:774–780.

11. Huber G. Moenomycin and related phosphorus-containing anti-biotics. In: Hahn FE, ed. Mechanism of Action of Antibacterial Agents. Berlin-Heidelberg: Springer-Verlag, 1979:135–153.

12. Bause E, Legler G. The effect of flavomycin on the synthesis and transfer of lipid-linked saccharides in pig brain. Biochem J 1982;201:481–487.

13. Huber G, Nesemann G. Moenomycin, an inhibitor of cell wall synthesis. Biochem Biophys Res Commun 1968;30:7–13.

14. Vanderwel D, Ishiguro EE. Properties of cell wall peptidoglycan synthesized by amino acid deprived re1A mutants of Escherichia coli. Can J Microbiol 1984;30:1239–1246.

15. Watanabe T, Ogata Y, Sugawara K. Increase of flavomycin sensitivity of bacteria by R-factors. In: Kremery V, Rosiwal I, Watanabe T, eds. Bacterial Plasmids and Antibiotic Resistance. New York, New York: Springer Verlag, 1972:105–113.

16. Kissel A. Determination of the in-vitro susceptibility of selected bacterial strains carrying resistance plasmids to the feed addi-tives salinomycin–sodium (SAL) and flavophospholipol (FPL) and to various active substances used for therapeutic pur-poses — MIC determination before and after 5 or 10 passages in additive-free and in SAL- or FPL- containing liquid media. Ger Hoechst Roussel Vet 1998.

17. Bolder NM, Wagenaar JA, Putirulan FF, Veldman KT, Sommer M. The effect of flavophospholipol (Flavomycin) and salinomycin sodium (Sacox) on the excretion of Clostridium perfringens, Salmonella enteritidis, and Campylobacter jejuni in broilers after experimental infection. Poult Sci 1999;78:1681–1689.

18. Dealy J, Moeller MW. Influence of bambermycins on Salmonella infection and antibiotic resistance in swine. J Anim Sci 1976;42: 1331–1336.

19. van den Bogaard AE, Hazen M, Hoyer M, Oostenbach P, Stobberingh EE. Effects of flavophospholipol on resistance in fecal Escherichia coli and enterococci of fattening pigs. Antimicrob Agents Chemother 2002;46:110–118.

20. Dorr PM, Gebreyes WA. Flavophospholipol: Effect on multi-drug resistant Salmonella enterica serovar Typhimurium in swine. J Swine Health Prod 2009;17:308–317.

21. Pfaller MA. Flavophospholipol use in animals: Positive impli-cations for antimicrobial resistance based on its microbiologic properties. Diagn Microbiol Infect Dis 2006;56:115–121.

22. Nair S, Farzan A, O’Sullivan TL, Friendship RM. A clinical trial investigating the impact of in-feed flavophospholipol on Salmonella shedding and antimicrobial resistance in pigs. Can Vet J 2018;59:59–65.

23. Kim HB, Isaacson RE. Salmonella in swine: Microbiota interac-tions. Annu Rev Anim Biosci 2017;5:43–63.

24. Kranker S, Alban L, Boes J, Dahl J. Longitudinal study of Salmonella enterica serotype Typhimurium infection in three Danish farrow-to-finish swine herds. J Clin Microbiol 2003; 41:2282–2288.

25. Nielsen B, Baggesen D, Bager F, Haugegaard J, Lind P. The sero-logical response to Salmonella serovars typhimurium and infantis in experimentally infected pigs. The time course followed with an indirect anti-LPS ELISA and bacteriological examinations. Vet Microbiol 1995;47:205–218.

26. Nair S, Farzan A, O’Sullivan TL, Friendship RM. Time course of Salmonella shedding and antibody response in naturally infected pigs during grower-finisher stage. Can J Vet Res 2018; 82:139–145.

27. Kogut MH, Arsenault RJ. Immunometabolic phenotype altera-tions associated with the induction of disease tolerance and persistent asymptomatic infection of Salmonella in the chicken intestine. Front Immunol 2017;8:372.


Article


Accuracy of external measurements of 3-dimensional (3D) printed biomodels of the canine radius used in an in-hospital settingSebastian Mejia, Nikia Stewart, Angel Miller, Roman Savicky, Christopher Monarski,

George E. Moore, Dennis Keith

A b s t r a c tThe objective of this study was to determine if biomodels printed on a fused deposition modeling (FDM) device from computed tomography (CT) data are accurate by comparing external measurements to the native bone, considering that the clinical usefulness of the printed biomodels in an in-hospital setting depends on their verified accuracy and consistency. Using canine cadaveric radii previously stripped of all soft tissues, 7 parameters of the actual bone and the 3-dimensional (3D) printed biomodels were measured and compared to determine how accurately the models represent the cadaveric bone. A total of 28 canine radii were collected, in which the landmarks for measurements were established. Radiographs were then taken to determine the frontal center of rotation of angulation (CORA) and CT scans were carried out. Finally, a 3D virtual reconstruction was done and converted into a stereolithography (STL) format file, from which 2 biomodels were printed per bone. Measurements for biomodels were compared for equivalence to cadaveric measurements. For the 7 measured parameters, the mean difference between biomodel and cadaveric parameters ranged from an increase of 10.66% in cranial-caudal proximal (CrCdP)-CORA to a decrease of 21.32% in distal width of the radius. For all 7 measured parameters, measurements for biomodels were statistically equivalent to their corresponding cadaveric bone (P , 0.001). The 7 measured parameters in the 3D models printed with an FDM device were not significantly different than those in the original bone. In fact, these measurements closely approximated original bone measurements (within 1.5%); therefore, validating their application in future presurgical planning for various orthopedic procedures.

R é s u m éL’objectif de la présente étude était de déterminer si les biomodèles imprimés par un appareil de modélisation par dépôt en fusion (MDF) à partir de données obtenues par tomodensitométrie (TMD) sont précis en comparant les mesures externes à l’os naturel, considérant que l’utilité clinique des biomodèles imprimés dans un milieu hospitalier dépend de leur précision vérifiée et de leur constance. Utilisant des radius canins provenant de cadavres et dont on a retiré tous les tissus mous, sept paramètres de l’os naturel et des biomodèles imprimés en trois dimensions (3D) ont été mesurés et comparés afin de déterminer jusqu’à quel point les modèles représentent l’os cadavérique. Vingt-huit radius canins ont été amassés, et à partir desquels les critères de mesure ont été établis. Des radiographies ont été prises afin de déterminer le centre frontal de rotation d’angulation (CORA) et des TMDs effectuées. Finalement, une reconstruction virtuelle 3D a été faite et convertie en fichier de format de stéréolithographie, à partir desquels deux biomodèles ont été imprimés par os. Les mesures des biomodèles ont été comparées pour équivalence aux mesures des os cadavériques. Pour les sept paramètres mesurés, la différence moyenne entre les paramètres des biomodèles et ceux des os cadavériques variait d’une augmentation de 10,66 % du CORA-crânial-caudal proximal à une diminution de 21,32 % dans la largeur distale du radius. Pour les sept paramètres mesurés, les mesures des biomodèles étaient statistiquement équivalentes (P , 0,001) à leurs mesures correspondantes sur les os cadavériques. Les sept paramètres mesurés sur les modèles 3D imprimés avec un appareil à MDF n’étaient pas différents statistiquement de ceux de l’os original. En fait, ces mesures étaient très près des mesures des os naturels (en dedans de 1,5 %), validant ainsi leur application dans la planification pré-chirurgicale future de différentes procédures orthopédiques.


Research Fellow, College of Veterinary Medicine and Biomedical Sciences, Colorado State University, Fort Collins, Colorado, USA (Mejia); Diagnostic Imaging Department, College of Veterinary Medicine, University of Minnesota, St. Paul, Minnesota, USA (Stewart); Surgery Department, VCA Animal Specialty Group, San Diego, California, USA (Miller); Surgery Department (Savicky, Monarski) and Diagnostic Imaging Department (Keith), VCA Animal Referral and Emergency Centre of Arizona, Mesa, Arizona, USA; Department of Veterinary Administration, College of Veterinary Medicine, Purdue University, West Lafayette, Indiana, USA (Moore).

Address all correspondence to Dr. Sebastian Mejia; telephone: (970) 297-5000; fax: (970) 297-1254; e-mail: [email protected]

Received April 4, 2018. Accepted June 27, 2018.



I n t r o d u c t i o nIn human medicine, and more recently in veterinary medicine,

3-dimensional (3D) printing has proven to be a promising tech-nological advancement, with applications in multiple industries and fields, and more specifically, in both soft tissue and orthopedic surgery (1–6). Among potential applications are a variety of clinical settings, including research and development of novel diagnostic and therapeutic techniques, preoperative complex surgical plan-ning, and teaching students and residents to understand disease processes (1–5,7,8).

Rapid prototyping (RP) is a manufacturing technology used in multiple industries to develop high-fidelity 3D structures from source image data (3). Developing an RP model involves creating a physical 3D model from a computer-based model (7,9) obtained through advanced diagnostic imaging. Computed tomography (CT) and magnetic resonance imaging (MRI) have proven superior to radiographs for evaluating and diagnosing bone abnormalities (10,11). Medical-grade software is then used to convert these images into a “printable” file, known as a stereolithography (STL) file (1).

There are 2 broad categories of methods for generating or printing the 3D physical model. The first category is subtractive technologies, i.e., milling, in which the model is carved from a preexisting block of the desired material. The second category is additive techniques, of which there are several major types used for rapid prototyping, i.e., stereolithography, selective laser sintering, fused-deposition modeling (FDM), and multi-jet modeling (4). For the purpose of this study, we evaluated the FDM method (7).

Three-dimensional printing has been successfully used for pre-surgical planning in oral and maxillofacial surgery in both humans and animals (12,13), as well as for correcting complex angular limb deformity (ALD) in veterinary medicine (14–16). Among the reported benefits was the ability to preplan the surgical approach, select the appropriate locking implant in advance to ensure that the necessary implants are available, and allow precontouring of the implant, which decreases surgical and anesthetic time and associated complications (15,17).

Considering that the clinical usefulness of printed biomodels depends on verified accuracy and consistency, the objective of this study was to compare measurements of biomodels printed on an FDM device with polyactic acid (PLA) using the CT scans of canine radii that were previously stripped of the soft tissues with the actual bones to determine if the biomodels accurately represented the original source of the imaging. Fitzwater and coworkers evalu-ated the effect of the amount of radiation used during the CT scan on the accuracy of biomodels printed out of canine femora (18,19). To the authors’ knowledge, however, there are no reports evaluat-ing the accuracy of 3D printed biomodels on the canine radius. We hypothesized that the measurements of external dimensions would be nearly identical and there would not be a scaling effect, i.e., increasing or decreasing of dimensions.

M a t e r i a l s a n d m e t h o d sA prospective cadaveric study design was used to determine the

accuracy of values obtained from biomodels printed as 3D virtual

reconstructions from CT scan studies of canine radii previously stripped of soft tissues. Fourteen paired radii were harvested from 14 skeletally mature dogs with no known radius/ulna pathology. The dogs were euthanized at the local humane society for reasons unrelated to this study. The radii were dissected free of all soft tis-sue by a taxidermist.

A system was developed to create specific landmarks on the canine radii, which would allow the investigator to obtain 7 dif-ferent objective values for comparison between the radii and the biomodels. Before the CT scan, specific points were marked on the bone in order to generate and facilitate repeatable measurements of lengths and circumferences. The landmarks for determining length were drilled holes made with a burr and surgical drill in the cranial aspect of the radial head to create the cranial-caudal proximal point (CrCdP) and in the groove for the external carpi radialis to create the cranial-caudal distal point (CrCdD) (Figure 1A). A caudal-cranial ex-vivo radiograph was taken to determine the frontal center of rotation of angulation (CORA) once the CrCdP and CrCdD points had been drilled in the bone. A frontal CORA was determined on the radiographic image, using the proximolateral edge of the radial head and the medial-most aspect of the articular fovea as landmarks for the proximal joint orientation line and the lateral-most aspect of the articular surface and the medial aspect of the articular face, ignoring the styloid process for the distal radial joint orientation line (Figure 2). All CORA determinations were carried out by a single investigator (SM) to preserve homogeneity and eliminate inter-observer variation.

The CORA plane and location were determined; the magnitude was not recorded, as it had no clinical significance for this study. For the subjects in which the proximal and distal anatomical axis would not intersect within the cortical confines of the bone, an intermedi-ate anatomic axis segment was calculated, which resulted in both a proximal and distal CORA. In these cases, the proximal CORA was chosen as the landmark to measure distances to CrCdP and CrCdD (20,21). Once determined, a Kirschner pin was used in a cranial-to-caudal fashion on the frontal plane at the level of the CORA to create a mark for measuring circumference and distance from this point to CrCdP and CrCdD (Figure 1B).

Marks were also created at 3 different locations from medial to lateral (or vice versa) on the sagittal plane using a surgical drill and

Figure 1. A — Determination of the cranial-caudal proximal (CrCdP) and cranio-caudal distal (CrCdD) landmarks. B — Determination of medial-lateral landmarks for measurement of proximal radial circumference.



Kirschner pins. Proximally, the selected landmark was the radial tuberosity on the medial aspect of the bone, at a perpendicular angle to the long axis of the radius that was used to measure the circumfer-ence. Distally, the ulnar notch was used on the lateral aspect of the bone also in a perpendicular manner to the long axis to determine distal width. Finally, the mid-diaphysial circumference point was marked in the same manner as previously described at an equidis-tant point between CrCdP and CrCdD, at a perpendicular angle to the long axis of the radius.

Each radius was imaged ex vivo with a 64-slice multi-detector CT scanner with a protocol of 120 kVp, 200 mA, rotation time of 1 s, 0.625-mm slice thickness, and 0.312-mm slice interval. The scans were then imported in digital imaging and communications in medi-cine (DICOM) format into a biomodeling medical software (Mimics; Materialise, Plymouth, Michigan, USA) and 3D reconstructions were created in an STL format. DICOM data contains volumetric density values for each voxel, and in order to be able to print this data, it must be processed as a continuous surface instead of a volume. The medical-grade software (Mimics) was used to obtain surface render-

ings for further processing (Figures 3A, B) which is a necessary step for reconstruction and 3D printer interpretation (17,22,23). Once the definition of target geometry into an STL mesh was achieved, the file was uploaded to the FDM device proprietary software MakerWare (MakerBot; Brooklyn, New York, USA). This software assigned the materials and orientation of the model and divided the model into a series of horizontal 2- dimensional (2D) slices to be sent to the device for deposition (17). Each slide included precise X-Y plane information for build and support material, with an average preset slice thickness of 40 mm.

The FDM device was a MakerBot Replicator Z18 (MakerBot) and the material used was polyactic acid (PLA). Two biomodels of each radius were printed and then 7 precise measurements were carried out on the cadaveric bones and printed biomodels for comparison. A digital caliper was used to measure length and a medical measuring tape was used to measure circumference.

The following 7 values were measured: proximal circumference; distal width; mid-diaphysial circumference; length (established as the distance between CrCdP and CrCdD); distance from a proximal point to frontal CORA (CrCdP-CORA); distance from a distal point to frontal CORA (CrCdD-CORA); and circumference at the CORA. All measurements were taken by a single investigator (SM) for con-sistency and to eliminate any possibility of inter-observer variation.

Statistical analysisMeasurements from biomodels and cadaveric radii were tested for

equivalence, for each of 7 anatomic parameters, using the Anderson and Hauck test of bioequivalence. A P-value , 0.05 was designated for statistical significance, rejecting the assumption of a 10% differ-ence between measurements.

Re s u l t sA total of 28 radii from 14 dogs was used. Dogs had a mean body

weight of 28.06 kg (range: 19.9 to 36.8 kg). Estimated average age was 1.89 y (range: 1 to 3 y). Breeds included 11 pit bulls, 1 Rottweiler, and 1 mixed breed dog. There were 4 intact females, 1 spayed female, 5 intact males, and 4 neutered males. The descriptive statistics for both the 28 dissected canine radii and the 56 printed biomodels are presented in Table I. For the 7 measured parameters, the mean dif-ference between biomodel and cadaveric parameters ranged from an increase of 10.66% in CrCdP-CORA to a decrease of 21.32% in distal width of the radius. For all 7 measured parameters, measurements for biomodels were statistically equivalent to their corresponding cadaveric bone (P , 0.001).

D i s c u s s i o nOur hypothesis was supported by the finding that all printed

biomodels were bioequivalent to their original bone sources, which rejects the assumption of a 10% difference between measurements. The slight variation between bones and biomodels could be attrib-uted to intra-observer variation due to different ways of using the measuring tools and placing of the caliper within the marked point. Despite creating specific landmarks, which were replicated onto the printed biomodels to use for measuring in order to eliminate the

Figure 2. Radiographic determination (ex vivo) of frontal center of rota-tion of angulation (CORA) after creating CrCdP and CrCdD landmarks.



subjectivity involved in determining the precise anatomic landmarks, i.e., ulnar notch, radial tuberosity, the investigator noted that a slight alteration in positioning the measuring device could generate a slightly different measurement on the same sample (, 0.5 mm). An alternative approach would be to measure the cadaveric bones and biomodels through CT scans and software using multiplanar reconstructions (MPR) or a coordinate measuring machine and specialized software (24,25).

Another possible source of variation could be model fabrication reproducibility, and to a lesser degree, variation could be related to the settings of the CT scan. During the creation of the STL file, there is a subjective component to the thresholding process. Thresholding is the

determination of the number of pixels that are used when 3D rendering is built from the DICOM data. While there was no linear association between threshold variations and external biomodel accuracy, it has been reported that the accuracy of biomodels created by FDM increased significantly with higher radiation settings (18,19). Measurement accuracy also necessitates that all supporting building materials and filaments be removed before the biomodel is used, as their presence may create an artifact effect at the time of clinical application.

More recent and advanced FDM devices can use filaments with different attributes during the printing process for the biomodel (based on organic tissue, such as bone) and the additional vertical support structures. This second filament used for support may be dissolvable or it may be a different color. Either technique would facilitate the complete removal of the necessary support material and lead to a more accurate surface on the model.

This prospective study and previous retrospective case reports and studies have proven that 3D printing used in conjunction with imaging may be useful in the characterization and planning of complicated orthopedic procedures. For example, complex fracture repairs and angular limb deformity (ALD) corrections, which can be extremely challenging to diagnose with traditional diagnostic imag-ing alone (14,26,27), could benefit from 3D reconstructed models for surgical planning.

Finally, several factors that affect the functional properties of PLA need to be taken into consideration. Among them, the processing temperature and sample size significantly affect the morphology of the final components of the biomodel (28). While these factors did not seem to have a significant impact on the final result and accuracy of the biomodel in our study, different clinical applications will require different levels of fidelity and will have a lower toler-able margin of error. It is therefore important to interpret results and

Table I. Descriptive statistics for the 28 radii and 56 biomodels.

Biomodels Dimensions Radii (N = 28) (N = 56)(mm) Mean SD Mean SD1. Prox circ 40.89 4.32 40.66 4.452. Dis wid 22.75 2.36 22.45 2.313. MD circ 41.18 4.23 41.01 4.254. Length 144.73 15.89 145.31 15.645. CrCdP-CORA 63.36 27.29 63.78 27.106. CrCdD-CORA 80.50 32.33 80.72 32.187. CORA circ 40.45 4.55 40.38 4.54SD — standard deviation; Prox circ — proximal circumference; Dis wid — distal width; MD circ — mid-diaphysial circumference; CrCdP — cranial-caudal proximal; CrCdD — cranial-caudal distal; CORA circ — circumference at the center of rotation of angulation.

Figure 3. A — Segmentation and 3D virtual reconstruction process (conversion to STL format). B — Stereolithography file of canine radius.



apply them to the clinical setting with caution in regard to the level of accuracy expected.

The clinical relevance and significance of 3D printing technology in veterinary and human medicine is innate and obvious. While computer-aided design-based (CAD) software tools have proven useful in measuring 3D morphometric parameters in polygonal mesh models, which allow the file to be manipulated and its structures to be observed in a 3-dimensional way (29,30), the added advantage of the tactile appreciation of the structure combined with the abil-ity to preset and contour implants before surgery is invaluable as a clinical and teaching tool.

The in-hospital setup used during this study (FDM device) cost approximately $6000 USD and each printing took from 6 to 12 h, depending on the size of the biomodel. The printing material (PLA) is inexpensive (approximately $48 USD per 1 kg spool), while the software costs vary from free to several thousand dollars per year. The cost and time needed to generate this biomodel in a practical in-hospital setting makes it an easy and accessible option for private specialty practices that carry out advanced imaging and complex orthopedic surgical procedures on a regular basis.

Re f e r e n c e s 1. Eltorai AE, Nguyen E, Daniels AH. Three-dimensional printing

in orthopedic surgery. Orthopedics 2015;38:684–687. 2. Choi JW, Kim N. Clinical application of three-dimensional print-

ing technology in craniofacial plastic surgery. Arch Plas Surg 2015;42:267–277.

3. Frame M, Huntley JS. Rapid prototyping in orthopaedic surgery: A user’s guide. Sci World J 2012;doi:10.1100/2012/838575.

4. McMenamin PG, Quayle MR, McHenry CR, Adams JW. The production of anatomical teaching resources using three- dimensional (3D) printing technology. Anat Sci Educ 2014;7: 479–486.

5. Mok S, Nizak R, Fu SC, et al. From the printer: Potential of three-dimensional printing for orthopaedic applications. J Ortho Translat 2016;6:42–49.

6. Crosse KR, Worth AJ. Computer-assisted surgical correction of an antebrachial deformity in a dog. Vet Comp Orthop Traumatol 2010;23:354–361.

7. Esses SJ, Berman P, Bloom AI, Sosna J. Clinical applications of physical 3D models derived from MDCT data and created by rapid prototyping. AJR Am J Roentgenol 2011;196:W683–688.

8. Mulford J, MacKay N, Babazadeh S. Three dimensional printing in orthopaedic surgery. A review of current and future applica-tions. Orthop J Sports Med 2016;doi:10.1177/2325967116S60002.

9. Bauermeister AJ, Zuriarrain A, Newman MI. Three-dimensional printing in plastic and reconstructive surgery: A systematic review. Ann Plast Surg 2016;77:569–576.

10. Savio G, Baroni T, Concheri G, et al. Computation of femoral canine morphometric parameters in three-dimensional geo-metrical models. Vet Surg 2016;45:987–995.

11. Chia HN, Wu BM. Recent advances in 3D printing of biomateri-als. J Biol Eng 2015;9:4.

12. Winder J, Bibb R. Medical rapid prototyping technologies: State of the art and current limitations for application in oral

and maxillofacial surgery. J Oral Maxillofac Surg 2005;63: 1006–1015.

13. Mavili ME, Canter HI, Saglam-Aydinatay B, Kamaci S, Kocadereli I. Use of three-dimensional medical modeling meth-ods for precise planning of orthognathic surgery. J Craniofac Surg 2007;18:740–747.

14. Dismukes DI, Fox DB, Tomlinson JL, Essman SC. Use of radio-graphic measures and three-dimensional computed tomographic imaging in surgical correction of an antebrachial deformity in a dog. J Am Vet Med Assoc 2008;232:68–73.

15. DeTora MD, Boudrieau RJ. Complex angular and torsional defor-mities (distal femoral malunions). Preoperative planning using stereolithography and surgical correction with locking plate fixa-tion in four dogs. Vet Comp Orthop Traumatol 2016;29:416–425.

16. Kaiser SM, Harms O, Konar M, et al. Clinical, radiographic, and magnetic resonance imaging findings of gastrocnemius musculotendinopathy in various dog breeds. Vet Comp Orthop Traumatol 2016;29:515–521.

17. Winer JN, Verstraete FJM, Cissell DD, Lucero S, Athanasiou KA, Arzi B. The application of 3-dimensional printing for preopera-tive planning in oral and maxillofacial surgery in dogs and cats. Vet Surg 2017;46:942–951.

18. Fitzwater KL, Marcellin-Little DJ, Harrysson OL, Osborne JA, Poindexter EC. Evaluation of the effect of computed tomogra-phy scan protocols and freeform fabrication methods on bone biomodel accuracy. Am J Vet Res 2011;72:1178–1185.

19. Fitzwater KL. Rapid prototyping and computed tomography optimization of canine long bone modeling [PhD dissertation]. Columbus, Ohio: The Ohio State University, 2010.

20. Knapp JL, Tomlinson JL, Fox DB. Classification of angular limb deformities affecting the canine radius and ulna using the center of rotation of angulation method. Vet Surg 2016;45:295–302.

21. Fox DB, Tomlinson JL, Cook JL, Breshears LM. Principles of uniapical and biapical radial deformity correction using dome osteotomies and the center of rotation of angulation methodol-ogy in dogs. Vet Surg 2006;35:67–77.

22. Doney E, Krumdick LA, Diener JM, et al. 3D printing of pre-clinical X-ray computed tomographic data sets. J Vis Exp 2013; doi:10.3791/50250.

23. Hespel AM, Wilhite R, Hudson J. Invited review — Applications for 3D printers in veterinary medicine. Vet Radiol Ultrasound 2014;55:347–358.

24. Markose E, Vikraman B, Veerabahu M. Three dimensional CT reconstruction: A comparison between 2D, 3D CT and original anatomical structures. J Maxillofac Oral Surg 2009;8:8–12.

25. Salmi M, Paloheimo KS, Tuomi J, Wolff J, Makitie A. Accuracy of medical models made by additive manufacturing (rapid manufacturing). J Craniomaxillofac Surg 2013;41:603–609.

26. Piras LA, Peirone B, Fox D. Effects of antebrachial torsion on the measurement of angulation in the frontal plane: A cadaveric radiographic analysis. Vet Comp Orthop Traumatol 2012;25: 89–94.

27. Kwan TW, Marcellin-Little DJ, Harrysson OL. Correction of biapical radial deformities by use of bi-level hinged circular external fixation and distraction osteogenesis in 13 dogs. Vet Surg 2014;43:316–329.



28. Drummer D, Cifuentes-Cuéllar S, Rietzel D. Suitability of PLA/TCP for fused deposition modeling. Rapid Prototyp J 2012;18:500–507.

29. Victor J. Rotational alignment of the distal femur: A literature review. Orthop Traumatol Surg Res 2009;95:365–372.

30. Lopez MJ, Lewis BP, Swaab ME, Markel MD. Relationships among measurements obtained by use of computed tomography and radiography and scores of cartilage microdamage in hip joints with moderate to severe joint laxity of adult dogs. Am J Vet Res 2008;69:362–370.


Article


Pilot study comparing serum chemotherapy levels after intra-arterial and intravenous administration in dogs with naturally occurring

urinary tract tumorsMeghan Kirsch, Chick Weisse, Allyson Berent, Craig Clifford, Nicole Leibman,

Luke Wittenburg, Stephen B. Solomon, Kenneth Lamb

A b s t r a c tThe proposed advantages of intra-arterial chemotherapy (IAC) are based on the premises of local dose escalation to the tumor and reduced availability of systemic drugs. There is a lack of objective pharmacokinetic data to confirm the advantage of IAC in dogs with naturally occurring urogenital tumors. The objective of this study was to determine if IAC administration in urogenital tumors would result in decreased systemic drug exposure when compared to intravenous routes. Twenty-two dogs with naturally occurring urogenital tumors were enrolled in this prospective case-controlled study. Mitoxantrone, doxorubicin, or carboplatin were administered by IAC and intravenous routes [intravenous awake (intravenous chemotherapy — IVC) and under general anesthesia (IVGAC)] 3 weeks apart. Serum assays were used to determine the extent of systemic drug exposure. Dose-normalized peak systemic serum concentration (Cmax) and area under the serum drug concentration-time curve (AUC) were used to quantify systemic exposure. A total of 26 mitoxantrone treatments were administered to 10 dogs. While there was no significant difference in Cmax, the AUC was significantly lower after IAC compared with IVGAC. Ten doxorubicin treatments were administered to 5 dogs. There were no significant differences in Cmax or AUC. A total of 14 carboplatin treatments were administered to 7 dogs. The Cmax was significantly lower for IAC compared to IVC, while the AUC values were equivocal. This study demonstrates certain lower serum values may be achieved after IAC delivery of carboplatin and mitoxantrone. These chemotherapy agents may have a preferred pharmacological profile for regional chemotherapy delivery in dogs with urogenital tumors.

R é s u m éLes avantages proposés de la chimiothérapie intra-artérielle (CIA) sont basés sur les prémisses d’une escalade de la dose locale à la tumeur et d’une disponibilité réduite des drogues systémiques. Il y a un manque de données pharmacocinétiques objectives pour confirmer l’avantage de l’administration de CIA chez les chiens avec des tumeurs urogénitales se produisant naturellement. L’objectif de la présente étude était de déterminer si l’administration de CIA lors de tumeurs urogénitales résulterait en une diminution de l’exposition systémique aux drogues lorsque comparé aux voies intraveineuses. Vingt-deux chiens avec des tumeurs urogénitales d’occurrence naturelle participèrent à cette étude cas-témoin prospective. De la mitoxantrone, de la doxorubicine, ou de la carboplatine furent administrées par CIA et voies intraveineuses [intraveineuse éveillée (chimiothérapie intraveineuse — CIV) et sous anesthésie générale (CIVAG)] à 3 sem d’intervalle. Des analyses du sérum furent utilisées afin de déterminer l’étendue de l’exposition systémique aux drogues. Le pic de la concentration sérique systémique normalisé pour la dose (Cmax) et la surface sous la courbe de la concentration sérique de la drogue-temps (SSC) furent utilisés pour quantifier l’exposition systémique. Un total de 26 traitements à la mitoxantrone fut administré à 10 chiens. Bien qu’il n’y ait pas de différence significative dans le Cmax, la SSC était significativement plus basse après la CIA comparativement à la CIVAG. Dix traitements de doxorubicine furent administrés à cinq chiens. Il n’y avait pas de différence significative dans le Cmax ou ls SSC. Un total de 14 traitements de carboplatine fut administré à sept chiens. Le Cmax était significativement plus bas pour la CIA comparativement à la CIV, alors que les valeurs de SSC étaient équivoques. Cette étude démontre que certaines valeurs sériques plus faibles peuvent être obtenues après CIA avec la carboplatine et la mitoxantrone. Ces agents de chimiothérapie pourraient avoir un profil pharmacologique préférentiel pour livraison régionale de chimiothérapie chez les chiens avec des tumeurs urogénitales.


Animal Medical Center, New York, New York 10065, USA (Kirsch, Weisse, Berent, Leibman); Hope Veterinary Specialists, Malvern, Pennsylvania 19355, USA (Clifford); University of California-Davis, School of Veterinary Medicine, Davis, California 15616, USA (Wittenburg); Memorial-Sloan Kettering Cancer Center, New York, New York 10065, USA (Solomon); Lamb Consulting, West St. Paul, Minnesota 55118, USA (Lamb).

Address all correspondence to Dr. Meghan Kirsch; telephone: (929) 294-6968; e-mail: [email protected]

Dr. Weisse and Dr. Berent are consultants for Infiniti Medical, LLC. Dr. Clifford and Dr. Weisse were members of the scientific advisory board for the Puppy Up Foundation. The grant from Puppy Up to support this study was awarded before they became members of its advisory board.

The authors declare no off-label use of antimicrobials.

Received April 24, 2018. Accepted August 14, 2018.



I n t r o d u c t i o nCanine malignant urogenital tumors typically carry a poor prog-

nosis and traditional treatments include surgery and/or radiation therapy (1). The current standard-of-care therapy involves nonste-roidal anti- inflammatory drugs (NSAIDs) and chemotherapy (1). Chemotherapeutic agents are generally well-tolerated in companion animals, with standard protocols designed to result in less than 5% severe adverse events requiring hospitalization and a less than 1% mor-tality rate (2). Unfortunately, tumor responses to these chemotherapeutic protocols are often short-lived and inconsistent, with an approximately 30% initial tumor response rate that is not lasting (2,3). Due to these his-torically poor response rates, the authors and others have investigated novel approaches to improve the biological response rates to commonly available chemotherapy agents in veterinary patients (3,4).

Intra-arterial chemotherapy (IAC) is carried out to introduce an anti-cancer drug directly into the arteries that supply the tumor in the hope of providing a local dose escalation to the tumor without increasing the toxicity experienced during subsequent systemic dis-tribution. Intra-arterial chemotherapy is conducted via super-selective catheter placement under fluoroscopic guidance and is used in humans to decrease adverse systemic side effects and achieve greater tumor chemotherapy concentrations and response rates under certain circumstances (3). This technique was first reported in veterinary patients for treating canine urinary bladder carcinomas in combina-tion with radiotherapy (3,5). In the 2 dogs that received intra-arterial chemotherapy by way of the terminal aorta in that study, decreases in tumor dimensions were noted in reference to the pretreatment mea-surements. While that study was conducted in only 2 dogs and was combined with radiation therapy, further evaluations have confirmed similar findings using regional therapies with limited systemic effects (3,5–8). The authors and their colleagues have recently investigated these techniques and demonstrated initial improved tumor response rates and reduced adverse events in veterinary patients receiving IAC when compared with a similar group receiving intravenous chemotherapy (IVC) (3,7,8). The uptake of chemotherapeutics into tumors is thought to depend not only on the route of administration, but also on the rate of infusion, local blood flow, permeability of the blood vessels, and chemical structure of the compound, as well as the rate of drug metabolism and elimination (6,9).

The objective of this study was to determine from a pharmaco-kinetic standpoint if serum chemotherapy measurements could demonstrate an advantage of intra-arterial chemotherapy versus traditional intravenous routes of administration in terms of lower systemic drug exposure. The study used chemotherapy agents with established linear pharmacokinetic profiles and efficacy in treating urogenital tumors (6,10–14). The investigators hypothesized that measured serum chemotherapy levels would be lower with IAC than with intravenous (IVC) administration of the same chemotherapy agent and dose in the same patient.


Patient selectionFrom October 2013 to May 2017, 22 dogs with urinary tract

carcinomas (prostatic carcinoma and urinary bladder/urethral

transitional cell carcinoma) considered amenable to chemotherapy treatment at the Animal Medical Center [intra-arterial chemotherapy (IAC), intravenous chemotherapy (IVC) and intravenous under gen-eral anesthesia chemotherapy (IVGAC)] and at the Hope Veterinary Specialists (IVC treatments only) were recruited to participate in the study.

Eligibility criteria for patients were as follows: a measurable tumor of the urinary tract with a cytological or histopathological diagnosis; no previous surgery or radiation therapy; no evidence of serious renal or hepatic comorbidities that would preclude the use of general anesthesia or potentially skew results; and an expected survival time of at least 3 mo. If a patient received chemotherapy before inclusion in the study, a standard washout period of 21 d was instituted as determined by the medical oncologist. Patients were not excluded if they received nonsteroidal anti-inflammatory drugs (NSAIDs) at any point.

The mean age of the dogs was 9.7 y (range: 6 to 12 y). There were 10 castrated males, 1 intact male, and 11 spayed females. Purebred dogs accounted for 68% (15/22) of the study population and the remaining 32% (7/22) were mixed-breed dogs. Purebred dogs included 5 Labrador retrievers, 3 Scottish terriers, 2 fox terriers, 1 German shorthaired pointer, 1 Belgian terrier, 1 Vizsla, 1 French bulldog, and 1 shih tzu. Median weight at diagnosis was 20.3 kg (range: 5.9 to 40.1 kg) and the mean surface area of the patient (surface area of the patient not the tumor) at diagnosis was 0.72 m2 (range: 0.32 to 1.17 m2). There were 7 dogs weighing less than 15 kg and only 1 dog weighing less than 9 kg.

Types of cancer, which included 19 transitional cell carcinomas and 3 unclassified carcinomas, were diagnosed based on cytology in 20 dogs and histopathology in 2 dogs. Thoracic radiographs were conducted before treatment in all cases and all were reported as normal by a Board-certified radiologist. Abdominal ultraso-nography was carried out on all dogs in the study. A mass of the bladder/ urethra was present in 16 dogs and a mass involving the prostate was found in 6 dogs. Medial iliac lymphadenopathy was confirmed on abdominal ultrasonography in 10 dogs, none of which was submitted for cytological evaluation. All dogs were receiving cyclooxygenase inhibitors, carprofen (Novox or Rimadyl), deracoxib (Deramaxx), firocoxib (Previcox), or meloxicam (Metacam) at the time of enrollment in the study.

Each dog’s medical history was obtained, a physical examination was conducted, including weight, and standard staging procedures were followed. All investigations were conducted according to the protocol described here, which was evaluated and approved by the Animal Medical Center Institutional Animal Care and Use Committee.

ProceduresStandard, routine guidelines for evaluating oncology patients

and administering chemotherapy were followed. Patients were recruited into the study by either the Animal Medical Center or Hope Veterinary Specialists. If recruited by the Animal Medical Center Medical Oncology Service or Hope Veterinary Specialists, the IVC was administered first and further treat-ments were continued 3 wk later with the Animal Medical Center Interventional Radiology Service. If recruited by the Animal Medical Center Interventional Radiology Service, dogs



received IAC first and continued treatments were administered at either the Animal Medical Center or Hope Veterinary Specialists.

Each individual patient entering the study received the same dose of chemotherapy (mg/m2) as prescribed by the medical oncologist (CC or NL) via alternate routes (IAC, IVC, or IVGAC). Serial 2-mL peripheral blood samples were taken from the forelimbs or jugular veins before treatment, immediately after chemotherapy delivery (time 0), and at subsequent predetermined times of 10, 30, 60, and 180 min after administration. Each patient therefore acted as his/her own control when comparing systemic chemotherapy levels achieved through the different routes of administration. Treatments were given approximately 3 wk apart and a complete blood count was obtained on each patient 7 to 10 d after treatment and before chemotherapy was administered.

The order of treatment (IAC, IVC, and IVGAC) was determined at the clinician’s discretion, but was typically based on clinician availability at the different testing sites. Chemotherapy drugs and doses were determined by the primary medical oncologist. Full standard systemic doses were administered using mitoxantrone (5 mg/m2), doxorubicin [30 mg/m2 or 1 mg/kg body weight (BW) if , 15 kg], or carboplatin (300 mg/m2 or 10 mg/kg BW if , 15 kg), as these substances could be readily quantified in canine serum. Intravenous chemotherapy was administered according to standard intravenous chemotherapy guidelines and the same protocol was followed during intra-arterial delivery. Adverse events were graded in accordance with the Veterinary Cooperative Oncology Group — common terminology criteria for adverse events (VCOG-CTCAE) (2) and the medical oncologist’s recommendations if grade-III to

Figure 1. Serial lateral caudal abdominal digital subtraction (DSA) fluoroscopic images of dogs with urinary tract tumors. In all images the dog positioned with the head to the left of the image. A — Female dog with a microcatheter (white arrows) positioned in vaginal artery (Vag). A DSA image demonstrates filing of the vaginal and caudal vesical (CdVes) arteries with tumor vascular enhancement in the urinary bladder apex and body (*). B — Female dog with a microcatheter (white arrows) positioned in caudal vesical atery (CdVes). The DSA image demonstrates increased vascularity and timor blush of the trigonal tumor (*). C, D — Male dog with prostate tumor (*) and a microcatheter (white arrows) positioned in the internal pudendal artery (IP). A DSA image demonstrates filling of the IP, CdVes and prostate (Pros) arteries with reflux filling of the deep gluteal artery (DG). D — As the Pros artery was too small to be selectively catheterized, the IP artery has been coil embolized (black arrows) to divert contrast and subsequent chemotherapy into the prostatic tumor as demonstrated by the tumor blush (*).



grade-IV hematologic or gastrointestinal toxicity occurred. Adverse events were defined as complications that resulted in a dose reduc-tion or removal from the study. Dose-normalized values of peak systemic serum concentration (Cmax) and area under the serum drug concentration-time curve (AUC) were used to allow comparison of these parameters in dogs receiving different doses.

Following initial evaluation of the pilot data comparing only IAC and IVC, a third treatment arm of intravenous chemotherapy under general anesthesia (IVGAC) was added to the study to control for the potential effects of general anesthesia on drug clearance and volume of distribution. This cohort of patients, which enrolled after May 2014, received a third treatment with IVGAC. This treatment was always carried out last in order to preserve the initial study design and allow for direct comparison to participants that received only 2 of the procedure types (IAC and IVC).

Protocol for intra-arterial chemotherapy (IAC)The protocol used for intra-arterial delivery was similar to that

described previously (4). The patient was placed under general anesthesia and a standard aseptic preparation was made over the femoral or carotid artery. Surgical cut-down and catheterization of the femoral or external carotid artery was conducted and a combi-nation of catheters, guide wires, microcatheters, and microwires were manipulated under fluoroscopic guidance until the catheter reached the desired position. Super-selective catheterization of the terminal feeding artery was attempted in order to deliver the chemo-therapy as close to the bulk of the tumor as possible. Intraoperative fluoroscopy and angiography were used to determine positioning and blood flow through the target vessels (Figure 1). In general, for more unilateral tumors, the more tumor-burdened side was targeted. Otherwise, the better-perfused side was chosen, based on angiography. Only unilateral infusions were done, however, in order to mimic the IVC-targeted times as closely as possible. After the procedure, the access artery was ligated and the surgical site was closed routinely. Intravenous fluids were maintained at 5 mL/kg BW

per hour during general anesthesia and for 3 additional hours until the final blood sample was collected.

Protocol for intravenous chemotherapy (IVC)Intravenous chemotherapy was administered according to stan-

dard intravenous chemotherapy guidelines at the institution as a slow bolus or infusion, depending on the drug. Before intravenous chemotherapy was started, intravenous crystalloids were adminis-tered at 5 mL/kg BW per hour and continued until the 3-hour blood sample collection was obtained to help mimic the fluid volumes received during general anesthesia.

Protocol for intravenous chemotherapy under general anesthesia (IVGAC)

For the subset of patients that received the third treatment with intravenous chemotherapy under general anesthesia, the anesthetic protocol mimicked that of the IAC group, with the caveat that these patients remained under general anesthesia only for the time neces-sary to administer the chemotherapy agent. The infusion duration in this group was the same as for the other groups. Once the standard chemotherapy administration was complete, recovery was accom-plished and blood collection and intravenous fluid rates proceeded as with the other groups. Additionally, the mean procedure times, anesthesia times, and lowest mean indirect blood pressure readings were compared between IAC and IVGAC routes of administration.

Sample evaluationThe samples were analyzed at the Pennsylvania Animal Diagnostic

Laboratory System (PADLS) New Bolton Center Toxicology Laboratory, University of Pennsylvania School of Veterinary Medicine in Kennett Square, Pennsylvania, USA. Carboplatin was analyzed by measuring platinum using inductively coupled serum mass spec-trometry (ICP-MS) (Nexion 300D; Perkin Elmer, Shelton, Connecticut, USA). Doxorubicin and mitoxantrone were analyzed using liquid chromatography (Shimadzu, Columbia, Maryland, USA), coupled

Table I. Statistical summary of comparative bioavailability data for dose-normalized mitoxantrone in 10 dogs.

IAC versus IVC IAC IVC P-valueCmax (ng/mL) 88.94 6 19 84.63 6 20 0.78AUC (min 3 ng/mL/mg) 3209.71 6 628 3127.25 6 659 0.90IAC versus IVGAC IAC IVGAC P-valueCmax (ng/mL) 84.95 6 20 100.22 6 21 0.34AUC (min 3 ng/mL/mg) 2999.16 6 263 3718.14 6 263 0.04*IVC versus IVGAC IVC IVGAC P-valueCmax (ng/mL) 76.51 6 20 104.17 6 21 0.07AUC (min 3 ng/mL/mg) 2759.47 6 773 3669.76 6 792 0.15Results are presented as least square geographic means standard deviation using a repeated measures analysis of variance (ANOVA). Corrected doses of mitoxantrone (5 mg/m2) were used for least square geographic mean calculations. Statistical analysis was carried out using repeated measures ANOVA (P , 0.05).IAC — intra-arterial chemotherapy; IVC — intravenous chemotherapy; IVGAC — intravenous under general anesthesia chemotherapy; Cmax — peak systemic serum concentration; AUC — area under the serum drug concentration-time curve.* Significant result.



with triple quad mass spectrometry (API 4000 LC/MS/MS; Sciex, Framingham, Massachusetts, USA).

Statistical analysisDescriptive statistics were expressed as mean and standard devia-

tion. The error residuals were deemed normal by visual inspection and by Kolmogorov-Smirnoff testing. A crossover design with 3 treatment routes was used to analyze the data using repeated measures analysis of variance (ANOVA) with subject nominated as a random effect and a compound symmetry covariance structure. The subject within route was used to evaluate the sequence effects, which suggested a non-significant carryover in all cases, and was subsequently omitted from the model. In the case of improper tim-ing or missing data points, the entire procedure was removed from statistical analysis and the remaining procedures were compared for that patient. Categorical variables expressed as percentages were analyzed using chi square or Fisher’s exact test when appropriate. Power analyses from the repeated measures mixed model were analyzed as minimum, maximum, and mean standard deviations for the post-hoc analyses (10). Analyses were considered significant at P , 0.05 and were carried out using SAS software (SAS Institute, Cary, North Carolina, USA) (10).

Re s u l t sTwenty-two dogs with urinary tract carcinomas were enrolled

in the study from October 2013 to May 2017. These dogs received 50 separate chemotherapy deliveries.

Mitoxantrone groupA total of 26 mitoxantrone treatments, including 10 IAC, 9 IVC,

and 7 IVGAC procedures, were conducted in 10 dogs. Three patients received mitoxantrone by IAC first, 7 received treatment with IVC first, and all IVGAC treatments were carried out last. There was no significant difference among the mean doses administered (IAC 4.48, IVC 4.7, and IVGAC 4.3 mg/m2) considering treatment type [(IAC versus IVC) P = 0.43, (IAC versus IVGAC) P = 0.57, (IVC versus IVGAC) P = 0.20]. Dose-normalized AUC for mitoxantrone was significantly lower (P = 0.04) after IAC (2999 ng/mL) when compared with IVGAC (3718 ng/mL) (Table I). This difference was not demonstrated when comparing IAC versus IVC (P = 0.86) or IVC versus IVGAC (P = 0.15). There was no significant difference in dose-normalized Cmax among the procedure types. The mean chemotherapy infusion times for the IAC (11.25 min) and IVGAC groups (11 min) were not statistically different (P = 0.95). There was a significant difference in the total anesthesia time between the IAC (82 min) and IVGAC (17 min) groups (P = 0.0001). The lowest mean indirect arterial blood pressures among groups were not sta-tistically different (IAC = 61 mmHg, IVGAC = 70 mmHg, P = 0.14) (Table II).

Doxorubicin groupA total of 10 treatments, including 5 IAC, 3 IVC, and 2 IVGAC

procedures, were conducted in 5 dogs in the doxorubicin group. The first treatment was IAC in 3 patients and IVC in 2 and all patients underwent IVGAC as the last procedure. The mean IVGAC dose

Table II. Statistical summary of comparative anesthesia data for mitoxantrone, doxorubicin, and carboplatin between procedures (IAC versus IVGAC).

Mitoxantrone (10 dogs) IAC (10 procedures) IVGAC (7 procedures) P-valueLowest MAP (mmHg) 61 70 0.14Total infusion time (min) 11.25 11 0.95Total anesthesia time (min) 82 17 0.0001*Doxorubicin (5 dogs) IAC (5 procedures) IVGAC (2 procedures) P-valueLowest MAP (mmHg) 65 73 0.43Total infusion time (min) 20.8 15 0.28Total anesthesia time (min) 65 31 0.18Carboplatin (7 dogs) IAC (7 procedures) IVGAC (4 procedures) P-valueLowest MAP (mmHg) 62 78 0.04*Total infusion time (min) 14 12 0.52Total anesthesia time (min) 90.6 15 0.00*Combined mitoxantrone, carboplatin, and doxorubicin (22 dogs) IAC (22 procedures) IVGAC (13 procedures) P-valueLowest MAP (mmHg) 62.7 73.8 0.023*Total infusion time (min) 15.3 12.5 0.25Total anesthesia time (min) 79.2 20.1 , 0.00*Results are presented as least square geographic means standard deviation. Statistical analysis was carried out using Chi-square or Fisher’s exact test when appropriate (P , 0.05). Lowest reported indirect mean arterial pressure (MAP) during anesthesia is reported in millimeters of mercury and obtained by indirect oscillometric readings.IAC — intra-arterial chemotherapy; IVGAC — intravenous under general anesthesia chemotherapy; MAP — mean arterial pressure.* Significant result.



administered was significantly lower than the mean dose admin-istered to the IAC group (IVGAC 23 mg/m2, IAC 27.8 mg/m2; P = 0.03) and the IVC group (IVC 29 mg/m2; P = 0.01). However, there were no significant differences in the dose-normalized Cmax or AUC values among the procedures (Table III). The mean che-motherapy infusion time was not statistically different between groups (IAC = 20.75 min, IVGAC = 15 min, P = 0.28). There was no significant difference in total anesthesia time between the groups (IAC = 65 min, IVGAC = 31 min, P = 0.18). The lowest mean indi-rect arterial blood pressures between the groups were not statisti-cally different (IAC = 65 mmHg, IVGAC = 73 mmHg, P = 0.43) (Table II).

Carboplatin groupA total of 14 carboplatin treatments, consisting of 7 IAC, 3 IVC, and

4 IVGAC procedures, were carried out in 7 dogs. Four dogs received IAC first, 3 received IVC first, and all patients underwent IVGAC as the last procedure. There was no significant difference among groups in terms of mean dose administered (IAC 270 mg/m2 versus IVC 271 mg/m2, P = 0.98; IAC 270 mg/m2 versus IVGAC 256 mg/m2, P = 0.49; IVC 271 mg/m2 versus IVGAC 256 mg/m2, P = 0.53).

Dose-normalized AUC values among treatment types were com-parable (Table IV). The dose-normalized Cmax was significantly lower (P = 0.02) for the intra-arterial group (IAC 0.16 ng/mL), however,

Table III. Statistical summary of comparative bioavailability data for dose-normalized doxorubicin in 5 dogs.

IAC versus IVC IAC IVC P-valueCmax (ng/mL) 40.00 6 10 32.93 6 12 0.75AUC (min 3 ng/mL/mg) 1067.72 6 238 886.03 6 291 0.72IAC versus IVGAC IAC IVGAC P-valueCmax (ng/mL) 40.00 6 11 35.17 6 11 0.55AUC (min 3 ng/mL/mg) 1067.72 6 268 871.91 6 269 0.10IVC versus IVGAC IVC IVGAC P-valueCmax (ng/mL) 0.19 6 0 0.23 6 0 0.46AUC (min 3 ng/mL/mg) 9.99 6 2 11.68 6 2 0.41Results are presented as least square geographic means standard deviation using a repeated measures analysis of variance (ANOVA). Corrected doses of doxorubicin (30 mg/m2) were used for least square geographic mean calculations. Statistical analysis was carried out using repeated measures ANOVA (P , 0.05).IAC — intra-arterial chemotherapy; IVC — intravenous chemotherapy; IVGAC — intravenous under general anesthesia chemotherapy; Cmax — peak systemic serum concentration; AUC — area under the serum drug concentration-time curve.

Table IV. Statistical summary of comparative bioavailability data for dose-normalized carboplatin in 7 dogs.

IAC versus IVC IAC IVC P-valueCmax (ng/mL) 0.16 6 0 0.19 6 0 0.02*AUC (min 3 ng/mL/mg) 9.68 6 1 9.96 6 1 0.73IAC versus IVGAC IAC IVGAC P-valueCmax (ng/mL) 0.16 6 0 0.21 6 0 0.19AUC (min 3 ng/mL/mg) 9.68 6 1 11.34 6 2 0.24IVC versus IVGAC IVC IVGAC P-valueCmax (ng/mL) 0.19 6 0 0.23 6 0 0.46AUC (min 3 ng/mL/mg) 9.99 6 2 11.68 6 2 0.41Results are presented as least square geographic means standard deviation using a repeated measures analysis of variance (ANOVA). Corrected doses of carboplatin (300 mg/m2) were used for least square geographic mean calculations. Statistical analysis was carried out using repeated measures ANOVA (P , 0.05).IAC — intra-arterial chemotherapy; IVC — intravenous chemotherapy; IVGAC — intravenous under general anesthesia chemotherapy; Cmax — peak systemic serum concentration; AUC — area under the serum drug concentration-time curve.* Significant result.



than for the intravenous group (IVC 0.19 ng/mL). This difference was not demonstrated when IAC was compared to IVGAC (P = 0.19). The mean chemotherapy infusion times for the IAC (14 min) and IVGAC groups (12 min) were not significantly different (P = 0.52). There was a significant difference in total anesthesia time among the groups (IAC = 90.6 min, IVGAC = 15 min, P , 0.001). The lowest mean indirect arterial blood pressures among the groups were significantly different (IAC = 62 mmHg, IVGAC = 78 mmHg, P = 0.04) (Table II).

Procedure variablesNo adverse events occurred during any types of procedures in

any of the chemotherapy groups. There was no difference among the types of procedures in hematologic variables (neutrophil count P = 0.06 and platelet count P = 0.07) 7 to 10 d after treatment (Table V). Eight of the total 22 patients in the study (36%) had adverse events that resulted in a dose reduction or would have resulted in a dose reduction in the case of a treatment being admin-istered last (Table VI). The most common adverse event was a grade-III to grade-IV febrile neutropenia (3 with IAC, 4 with IVC, and 1 with IVGAC) and grade-II or grade-IV thrombocytopenia (3 with IAC, 3 with IVC, and 1 with IVGAC) 7 to 10 d after treat-ment. Most serious adverse events were noted in the mitoxantrone group, which accounted for 7 of the 8 or 88% of the total adverse events. A single adverse event was noted after IAC doxorubicin due to grade-II thrombocytopenia. There were no adverse events in the carboplatin group.

D i s c u s s i o nThe results of this study demonstrate weakly significant differ-

ences in peak serum concentrations between intravenous and intra-arterial carboplatin treatments. These results were not consistently seen across all tumor types or across all chemotherapy agents. This is not overly surprising as systemic drug concentrations do not neces-sarily represent the local tumor environment. Unfortunately, from a clinical standpoint, obtaining local tumor concentration values is not a realistic goal in client-owned animals. As previously stated, the way in which a tumor takes up a chemotherapeutic compound is governed by many factors that were not directly investigated in this study and therefore definitive conclusions about drug concentration cannot be made. This point is further strengthened by the limitations of the patient population. This study should therefore be considered as a pilot study on which to base further studies.

The peak serum concentrations of carboplatin were weakly but significantly higher in the awake intravenous group than in the intra-arterial group (P = 0.02). Serial peripheral blood sampling was carried out at consistent times and Cmax was determined as the measured concentration immediately after infusion for all che-motherapeutic agents. These results were not consistent across all intravenous groups (awake and under general anesthesia) as would be expected from our hypothesis. While lower means were seen with IAC (0.16 ng/dL) when compared with IVGAC (0.21 ng/dL), these results were not significant (P = 0.19). Additionally, the duration of anesthesia before delivery of chemotherapy was considerably shorter

Table V. Median hematologic variables (absolute neutrophil and platelet count) 7 to 10 d after treatment with mitoxantrone, doxorubicin, and carboplatin for each procedure.

Median absolute Median grade Median absolute Median grade neutrophil count (mL) neutropenia platelet count (mL) thrombocytopeniaMitoxantrone (26 procedures in 10 dogs) IAC (n = 10) 2593 1.7 184 000 1.7 IVC (n = 9) 2348 2 164 000 1.2 IVGAC (n = 7) 2600 1.6 214 000 1.2 Total (n = 26) 2494 1.8 182 000 1.4

Doxorubicin (10 treatments in 5 dogs) IAC (n = 5) 4488 1 261 000 1.3 IVC (n = 3) 2310 1 176 000 1 IVGAC (n = 2) 4066 1 125 000 1 Total (n = 10) 3968 1 216 000 1

Carboplatin (14 treatments in 7 dogs) IAC (n = 7) 4937 1 330 000 1 IVC (n = 3) 6383 1 281 000 1 IVGAC (n = 4) 3343 2.5 390 000 1 Total (n = 14) 4647 1.5 342 000 1Hematologic variables are graded in accordance with the Veterinary Cooperative Oncology Group — common terminology criteria for adverse events (VCOG-CTCAE) (2).IAC — intra-arterial chemotherapy; IVC — intravenous chemotherapy; IVGAC — intravenous under general anesthesia chemotherapy.



for the IVGAC group than for the IAC group, as it takes time to gain vascular access and for catheter placement before IAC is adminis-tered. Even in light of this consideration, which would be expected to increase serum levels in the IAC group due to reduced systemic clearance and smaller volume of distribution, IAC consistently dem-onstrated lower dose-normalized Cmax values. These results suggest that regional delivery of carboplatin should be investigated further if Cmax values correlate with enhanced tumor response and/or reduced adverse events.

Interestingly, much of the published data about carboplatin refers to the AUC as it correlates to efficacy and toxicity. Joerger et al (11) reported that toxicity of carboplatin, thrombocytopenia in this case, was best correlated with carboplatin AUC, but could also correlate with overall exposure when Cmax was included as well. A target AUC has been established in humans to produce better outcomes and studies have shown that the number of platinum-DNA adducts increases with increasing plasma AUC. As platinum-DNA adducts are the described mechanism of action for carboplatin for most cancers, a link can be drawn between plasma AUC and efficacy and

toxicity (11,15). In a phase-I evaluation of a dosing strategy provided by AUC, a targeted/maximum tolerated AUC for carboplatin could be determined in cats (16). In humans, efficacy of carboplatin in ovarian and testicular cancer was described using a targeted AUC of 5 to 7 mg/mL per minute and was associated with better treat-ment outcomes (11). It may be more reasonable to consider AUC as a parameter to guide further studies, although a lower Cmax has also been linked to a reduction in adverse events during administration in humans (11).

Area under the curve (AUC) is often considered the most impor-tant calculation in terms of systemic drug exposure and subsequently the most valuable in predicting toxicity after exposure to chemo-therapeutic drugs. The AUC has been demonstrated to correlate most significantly with bone marrow suppression and, more specifically, the degree of neutropenia (16).

Mitoxantrone AUC concentrations were significantly different for IAC than for IVGAC, but not for any other route. A higher AUC for IVGAC would support our hypothesis that IVGAC provides higher systemic drug concentrations based on logical assumptions of drug

Table VI. Hematologic adverse events that led to dose reduction or removal from study.

Adverse events First procedure Second procedure Third procedure Grade Grade Dose Grade Grade Dose Grade Grade Dose (mg/m2) neutropenia thrombocytopenia (mg/m2) neutropenia thrombocytopenia (mg/m2) neutropenia thrombocytopeniaMitoxantrone (10 dogs, 26 procedures) Patient 1 IAC IVC 5 3 4 3.75 1 1 Patient 2 IVC IAC IVGAC 4.6 3 1 3.68 1 1 3.68 1 1 Patient 3 IVC IAC IVGAC 4 4 1 3.4 4 1 3 1 1 Patient 4 IVC IAC IVGAC 5 4 2 4 1 2 4 4 2 Patient 5 IVC IAC IVGAC 5 4 2 3.75 1 1 3.75 1 1 Patient 6 IAC 4.5 4 4 Patient 7 IVC IAC IVGAC 5 4 2 3.75 1 1 3.75 1 1Doxorubicin (5 dogs, 10 treatments) Patient 8 IAC IVGAC 30 1 2 24 1 1Hematologic adverse events are graded in accordance with the Veterinary Cooperative Oncology Group — common terminology criteria for adverse events (VCOG-CTCAE) (2). Adverse events were defined as complications that resulted in a dose reduction (grade-III to grade-IV neutropenia and/or grade-II to grade-IV thrombocytopenia) or removal from the study.IAC — intra-arterial chemotherapy; IVC — intravenous chemotherapy; IVGAC — intravenous under general anesthesia chemotherapy.



distribution under general anesthesia, although we cannot make this determination. While the mitoxantrone group experienced the most adverse events, the population was too small to interpret statistically significant differences between the routes of administration. There are few data in terms of serum mitoxantrone levels (Cmax and AUC) regarding efficacy and toxicity (adverse events). For mitoxantrone, information pertaining to Cmax relates to acute toxicity, whereas AUC has traditionally correlated better with efficacy and, in some reports, with adverse events such as gastrointestinal signs and bone marrow suppression (13). Information provided by human studies regard-ing Cmax and AUC is reported after standard intravenous routes of delivery. From a toxicity standpoint, it would be important to determine if targeted delivery permitted similar efficacy with lower systemic toxicity due to the reduced AUC. While IVGAC would not be routinely administered clinically, the comparison to IAC reduces the conflicting effects of general anesthesia. The informa-tion gained from this comparison could suggest that placement of the intra-arterial port for IAC in awake patients is superior to IVC. This requires future investigation.

Anti-cancer drugs have a narrow therapeutic index and the impact of general anesthesia on metabolism, distribution, and clearance is largely unknown. After intravenous administration, the peak blood concentration of the chemotherapeutic agent is determined by the dose, rate of administration, and cardiac output. Higher cardiac output and subsequent higher volume of blood delivered to the tissues results in lower peak systemic levels. General anesthesia often reduces glomerular filtration rate and therefore clearance of chemotherapeutic agents that rely on renal clearance, such as carbo-platin, which may result in higher AUC. This effect may not pertain to drugs cleared through metabolism by hepatic mixed function oxidase enzymes (9,17). The third arm of the study incorporating the IVGAC group was ultimately added to control for these factors as much as possible.

For carboplatin, the lower Cmax values after intra-arterial che-motherapy (IAC) might be explained by higher initial local uptake by the tumor tissue. While this supports the hypothesis that IAC would result in lower systemic chemotherapy levels, similar sta-tistically significant values were not found for the mitoxantrone or doxorubicin groups. The failure to identify a statistical difference could be complicated by the relatively small tumors being perfused comparable to the volume of distribution of the entire systemic circulation, systemic circulatory changes under anesthesia over time, tumor vascularity and microenvironment, or physiochemical properties of the individual drugs. In order to determine if intra-arterial chemotherapy resulted in greater total drug exposure to the tumor itself, a preferable approach would be to take tumor samples after chemotherapy is administered in order to obtain tissue drug concentrations, although this would have been difficult in this client-owned patient population.

A previous study validated the use of AUC for tumor tissue as being far more sensitive an indicator of drug absorption than plasma concentrations (17). A research study conducted with laboratory rabbits with bladder tumors determined that internal iliac artery infusion of carboplatin and pirarubicin resulted in significantly higher intratumoral chemotherapy levels, although no significant differences in the plasma levels were demonstrated (17). In a study

by Chen et al (6), while no significant difference was noted in plasma AUC for dogs administered carboplatin by intravenous or internal iliac intra-arterial delivery, platinum AUC for uterine tissue was significantly higher for the intra-arterial group. The Cmax for the peripheral vein administration group was significantly higher than the abdominal aorta and internal iliac artery administration groups at time zero, 15 and 30 min after carboplatin was administered (6). That study did not use super-selective arterial catheterization as we did in the current study. In-vivo pharmacokinetic studies in humans have used dynamic positron emission tomography (PET) as an alternative to invasive tissue sampling, although this modality is not commonly used in the veterinary clinical setting (18).

As blood flow in tumors is the primary determinant of exposure to the chemotherapeutic agent, it will be maximized if the tumor is well-vascularized and blood flow is slow (9,17). The degree of tumor perfusion among patients was not evaluated or characterized in the current study. Intra-arterial chemotherapy uses super-selective cath-eters that can deliver chemotherapy to the arteries feeding the tumor (3). In the current study, the chemotherapy agent within the local arteries was visualized by use of iodinated contrast and fluoroscopic guidance-conferring exposure. It would be reasonable to speculate that longer exposure times would result in more chemotherapy drugs being retained by the tumor itself.

Although not used in this study, methods to increase tempo-ral exposure of the chemotherapy being infused could result in additional uptake by the tumor. This could be accomplished by temporary balloon occlusion of proximal arteries while chemo-therapy was being administered or by vascular embolization with chemotherapy (chemoembolization). In a study by Song et al (17), transarterial chemotherapy (TAC) with carboplatin was compared with a transarterial chemoembolization (TACE) procedure using a lipiodol-carboplatin mixture (TACE group) administered to the internal iliac arteries in normal dogs. This study found that TACE had a pharmacokinetic advantage compared to TAC in the urogenital system as there was a 2-fold increase in uterine AUC concentrations and lower peak plasma concentrations in the TACE group (17). These animals were not recovered to determine if any detrimental effects of embolization of this vessel resulted in morbidity.

It is not clear if the dogs with naturally occurring tumors that received super-selective delivery in the present study could safely tolerate a chemoembolization procedure, as non-target embolization would likely occur to the caudal vesical artery, internal pudendal artery, and possibly, the middle rectal arteries in certain circum-stances. In a canine prostatic hyperplasia model, super-selective catheterization of the prostatic branches of the internal pudendal arteries was successfully carried out using a 3-French microinfusion catheter, which was shown to induce prostatic infarction and sub-sequent reduction in prostate size (3). Chemoembolization for liver tumors has been demonstrated to result in 10- to 50-fold increases in intra-tumoral drug concentrations and to be generally well-tolerated, presumably due to the dual blood supply present in this organ (7,8).

To the authors’ knowledge, this is the first study to compare the pharmacokinetic profiles of intra-arterial chemotherapy with intravenous routes, awake and under general anesthesia, in a canine urogenital tumor model. Intra-arterial administration of carboplatin



and mitoxantrone demonstrated potentially lower Cmax and AUC values, respectively, when compared to certain intravenous coun-terparts. This points to a possible advantageous “first pass effect” after regional delivery of carboplatin and mitoxantrone and both should be considered for future larger comparative pharmacokinetic investigations in dogs with naturally occurring urogenital tumors. A natural next step for investigation would be to confirm the pharma-cokinetic advantages of using arterial ports to deliver these agents intra-arterially in awake patients (19).

A c k n o w l e d g m e n tThe study was supported by a grant from the Puppy Up Foundation.

Re f e r e n c e s 1. Knapp DW, Ramos-Vara JA, Moore GE, Dhawan D, Bonney PL,

Young KE. Urinary bladder cancer in dogs, a naturally occur-ring model for cancer biology and drug development. ILAR J 2014;55:100–118.

2. Veterinary cooperative oncology group — Common terminology criteria for adverse events (VCOG-CTCAE) following chemo-therapy or biological antineoplastic therapy in dogs and cats v1.1. Vet Comp Oncol 2016;14:417–446.

3. Culp WTN, Weisse C, Berent AC, et al. Early tumor response to intraarterial or intravenous administration of carboplatin to treat naturally occurring lower urinary tract carcinoma in dogs. J Vet Intern Med 2015;29:900–907.

4. Rocha TA, Mauldin GN, Patnaik AK, Bergman PJ. Prognostic factors in dogs with urinary bladder carcinoma. J Vet Intern Med 2000;14:486.

5. McCaw DL, Lattimer JC. Radiation and cisplatin for treatment of canine urinary bladder carcinoma: A report of two case histories. Vet Radiol 1988;29:264–268.

6. Chen C, Wang W, Zhou H, et al. Pharmacokinetic comparison between systemic and local chemotherapy by carboplatin in dogs. Reprod Sci 2009;16:1097–1102.

7. Weisse C. Hepatic chemoembolization: A novel regional therapy. Vet Clin North Am Small Anim Pract 2009;39:627–630.

8. Weisse C, Clifford CA, Holt D, Solomon JA. Percutaneous arterial embolization and chemoembolization for treatment of benign and malignant tumors in three dogs and a goat. J Am Vet Med Assoc 2002;221:1430–1436.

9. Chen HS, Gross JF. Intra-arterial infusion of anticancer drugs: Theoretic aspects of drug delivery and review of responses. Cancer Treat Rep 1980;64:31–40.

10. Shieh G. A comparative study of power and sample size calcula-tions for multivariate general linear models. Multivar Behav Res 2003;38:285–307.

11. Joerger M, Huitema ADR, Richel DJ, et al. Population pharmaco-kinetics and pharmacodynamics of paclitaxel and carboplatin in ovarian cancer patients: A study by the European organization for research and treatment of cancer-pharmacology and molecu-lar mechanisms group and new drug development group. Clin Cancer Res Off J Am Assoc Cancer Res 2007;13:6410–6418.

12. Gustafson DL, Rastatter JC, Colombo T, Long ME. Doxorubicin pharmacokinetics: Macromolecule binding, metabolism, and excretion in the context of a physiologic model. J Pharm Sci 2002;91:1488–1501.

13. An G, Morris ME. A physiologically based pharmacokinetic model of mitoxantrone in mice and scale-up to humans: A semi-mechanistic model incorporating DNA and protein binding. AAPS J 2012;14:352–364.

14. Frazier DL, Price GS. Use of body surface area to calculate che-motherapeutic drug dose in dogs: II. Limitations imposed by pharmacokinetic factors. J Vet Intern Med 1998;12:272–278.

15. Duffull SB, Robinson BA. Clinical pharmacokinetics and dose optimisation of carboplatin. Clin Pharmacokinet 1997;33:161–183.

16. Bailey DB, Rassnick KM, Dykes NL, Pendyala L. Phase I evalu-ation of carboplatin by use of a dosing strategy based on a tar-geted area under the platinum concentration-versus-time curve and individual glomerular filtration rate in cats with tumors. Am J Vet Res 2009;70:770–776.

17. Song T, Wang W, Liu P, Chen C. Pharmacokinetic comparison between pelvic transarterial chemoembolization and transcath-eter arterial chemotherapy in an animal model. Mol Med Rep 2009;2:663–667.

18. Cho SY, Szabo Z. Molecular imaging of urogenital diseases. Semin Nucl Med 2014;44:93–109.

19. Sinn M, Nicolaou A, Ricke J, et al. Interventionally implanted port catheter systems for hepatic arterial infusion of chemo-therapy in patients with primary liver cancer: A phase II-study (NCT00356161). BMC Gastroenterol 2013;13:125.


Article


Comparison of 3 blind brachial plexus block techniques during maintenance of anesthesia and postoperative pain scores in dogs

undergoing surgical procedures of the thoracic limbAlicia Skelding, Alexander Valverde, Rodrigo Aguilera, Noel M. Moens,

Melissa Sinclair, Jeffrey J. Thomason

A b s t r a c tThe degree of analgesia provided by blind techniques for brachial plexus blocks (BPBs) has not been compared in clinical cases undergoing surgery of the thoracic limb. The objective of this study was to evaluate the anesthetic conditions and postoperative analgesia resulting from 3 different BPB local anesthetic techniques in canine patients undergoing such surgery. Twenty-four client-owned dogs received a standardized premedication/induction protocol (hydromorphone and acepromazine/propofol), maintained with isoflurane in oxygen using mechanical ventilation, in a prospective, randomized, blinded clinical trial. Before surgery, dogs received 1 of 3 anatomical BPB techniques: traditional, perpendicular, or axillary, with 0.2 mL/kg body weight (BW) of bupivacaine 0.5%. Cardiorespiratory variables and isoflurane end-tidal concentrations were recorded throughout anesthesia. Scores for anesthetic maintenance (0–best to 4–responsive), recovery quality (0–not responsive to 3–responsive), and Glasgow pain scale were recorded for up to 24 h postoperatively. All dogs recovered uneventfully from anesthesia and no differences in the measured variables or scores were noted among groups, during and after anesthesia. When thoracic limb amputations in each of the 3 groups (n = 9; 4 in traditional, 3 in perpendicular, 2 in axillary) were compared to the other surgical procedures (n = 15); however, scores for anesthetic maintenance were higher [0 (0 to 1) versus 0 (0 to 0); median (interquartile range)], recovery [1 (0 to 2) versus 0 (0 to 0)], and pain [2.4 (2.4 to 3.0) versus 1.6 (1.4 to 2.2)] in the first 3 h post-extubation. Surgery times were also longer with amputations [115 min (100 to 138 min) versus 50 min (41 to 90 min)]. The 3 BPB techniques provided similar anesthesia and postoperative pain scores. Despite higher pain scores in thoracic limb amputations than in less invasive surgeries, the BPB appeared to provide significant comfort.

R é s u m éLe degré d’analgésie fournie par les blocs à l’aveugle du plexus brachial (BPBs) n’a pas été comparé lors de cas cliniques soumis à une chirurgie du membre thoracique. La présente étude visait à évaluer les conditions anesthésiques et l’analgésie post-opératoire résultant de trois techniques différentes d’anesthésie locale par BPB chez des patients canins soumis à une chirurgie. Vingt-quatre chiens appartenant à des propriétaires ont reçu une prémédication/induction standardisée (hydromorphone et acépromazine/propofol), avec maintien à l’isoflurane dans de l’oxygène en utilisant une ventilation mécanique, dans une étude clinique prospective, randomisée, et à l’aveugle. Avant la chirurgie, les chiens ont reçu un BPB par une des trois techniques anatomiques : traditionnelle, perpendiculaire, ou axillaire, avec 0,2 mL/kg de poids corporel de bupivacaïne 0,5 %. Les variables cardiorespiratoires et les concentrations d’isoflurane en fin d’expiration furent enregistrées tout au long de l’anesthésie. Les pointages pour le maintien de l’anesthésie (0-meilleur à 4-réactif), la qualité du rétablissement (0-non réactif à 3-réactif), et l’échelle de la douleur de Glasgow ont été notés jusqu’à 24 h post-opération. Tous les chiens ont récupéré sans problème de l’anesthésie et aucune différence mesurable dans les variables mesurées ou les pointages ne fut notée parmi les groupes, durant ou après l’anesthésie. Toutefois, lorsque les amputations du membre thoracique dans chacun des trois groupes (n = 9; 4 dans la traditionnelle, 3 dans la perpendiculaire, 2 dans l’axillaire) furent comparées aux autres procédures chirurgicales (n = 15), les pointages pour le maintien de l’anesthésie étaient plus élevés [0 (0 à 1) versus 0 (0 à 0); médiane (écart interquartile)], rétablissement [1 (0 à 2) versus 0 (0 à 0)], et douleur [2,4 (2,4 à 3,0) versus 1,6 (1,4 à 2,2)] dans les 3 h suivant l’extubation. La durée des chirurgies étaient également plus longues lors des amputations [115 min (100 à 138 min) versus 50 min (41 à 90 min)]. Les trois techniques de BPB produisaient des pointages similaires d’anesthésie et de douleur postopératoire. Malgré des pointages de douleur plus élevés lors d’amputations du membre thoracique que lors de chirurgies moins invasives, les BPB semblent fournir un confort significatif.


Department of Clinical Studies (Skelding, Valverde, Aguilera, Moens, Sinclair) and Department of Biomedical Sciences (Thomason), Ontario Veterinary College, University of Guelph, Guelph, Ontario N1G 2W1.

Address all correspondence to Dr. Alexander Valverde; telephone: (226) 924-5879; fax: (519) 767-1580; e-mail: [email protected]

Received August 13, 2018. Accepted October 13, 2018.



I n t r o d u c t i o nLocoregional anesthesia is commonly carried out in veterinary

medicine as an adjunct to general anesthesia to decrease the require-ment for anesthetic drugs and improve patient comfort intra- and postoperatively. In canine patients, surgical procedures of the tho-racic limb are commonplace and can include, but are not limited to, fracture repair, arthroscopy, mass removal, and limb amputa-tion. Major procedures such as these require general anesthesia, usually using balanced anesthetic techniques that include local nerve blocks.

The brachial plexus is a network of nerves that is formed from the ventral spinal nerve rami of C6 through T1 in dogs and provides most of the innervation to the thoracic limb (1–3). The brachial plexus is commonly blocked using local anesthetic in combination with gen-eral anesthesia in canine patients undergoing surgical procedures of the thoracic limb and, when done correctly using bupivacaine or ropivacaine, can result in analgesia of the thoracic limb for more than 6 h (4).

A commonly used technique for carrying out a brachial plexus block (BPB) in canine patients is a blind approach [traditional approach (TA)] to the plexus that employs specific anatomical land-marks for needle positioning (5,6). Although techniques for carrying out the BPB in dogs using ultrasonography (7) or electrolocation (8) have been described, the blind approach to the brachial plexus is often still used because it does not require additional equipment or an advanced skill level.

In previous descriptions of the blind technique (5,6), it was difficult to ascertain the final location of the needle tip to ensure adequate contact of the local anesthetic with the nerves of the brachial plexus and subsequently form an appropriate block. In a recent comparison of 3 blind techniques [TA, perpendicular approach (PA), and axillary approach (AA)] for BPB in cadaver dogs (3), there were no differences in nerve staining among techniques. More importantly, however, 3 key anatomical landmarks were identified that could be used to improve the distribution of the dye on the brachial plexus by more precisely positioning the final location of the needle. Regardless of the technique used to approach the brachial plexus, these landmarks were the transverse process of C6, the point of the shoulder, and the first rib. These sites are used to help position the tip of the needle at a level midway between the transverse process of C6 and the point of the shoulder and advanced to the cranial aspect of the first rib, where the brachial plexus is located.

Blind techniques for BPB have not been compared in dogs for their efficacy in controlling intra- and postoperative pain in clinical cases undergoing surgical procedures of the thoracic limb. In cats undergoing distal thoracic limb surgery, BPB has an inhalant-sparing effect and improves postoperative comfort (9). The objective of the current study was to evaluate the anesthetic conditions and post-operative analgesia resulting from 3 different blind BPB techniques (TA, PA, AA) in a clinical setting in canine patients undergoing surgical procedures of the thoracic limb. It was hypothesized that the 3 techniques would be equally effective in providing intra- and postoperative comfort.


Animals and anesthesiaTwenty-four client-owned dogs of various breeds presented

to the Ontario Veterinary College Health Sciences Centre from March 2016 to March 2017 for surgical procedures of the thoracic limb, and for whom owner consent was obtained, were used to evaluate 3 blind techniques for blocking the brachial plexus. Surgical procedures of the thoracic limb included amputation (n = 9) and soft-tissue or orthopedic repair (n = 15). Dogs were classified as status I or II according to the American Society of Anesthesiologists (ASA). The study was carried out in accordance with the guidelines of the Canadian Council on Animal Care and was approved by the Institutional Animal Care Committee at the University of Guelph.

All anesthetic procedures were carried out by 1 of 2 anesthesia residents, except for the brachial plexus blocks, which were done by the same Board-certified anesthesiologist. Dogs had not received any analgesic drugs at least 3 h before anesthesia. Dogs were premedicated with intramuscular (IM) hydromorphone (Sandoz Canada, Boucherville, Quebec), 0.05 mg/kg body weight (BW) and acepromazine (Atravet; Boehringer Ingelheim, Burlington, Ontario), 0.02 to 0.05 mg/kg BW. The preanesthetic medication was delivered through an intravenous (IV) catheter in 1 dog. A peripheral catheter (BD Insyte-W; BD Infusion Therapy Systems, Sandy, Utah, USA) was placed in the right or left saphenous vein, depending on the surgical procedure. Twenty-five minutes after premedication, the dogs were scored for degree of sedation at the time of catheteriza-tion using a scale from 0 to 3, with 0–indicating no obvious sedation and 3–indicating profound sedation (10) (Appendix I). Anesthesia was induced with propofol (pms-Propofol 1%; Pharmascience, Montreal, Quebec), 2 to 3 mg/kg BW, IV titrated to effect to allow orotracheal intubation and maintained with isoflurane (Aerrane; Baxter, Mississauga, Ontario) vaporized in 100% oxygen. Dogs were immediately ventilated using a volume-cycled ventilator (EMC 2002IE; Hallowell, Pittsfield, Massachusetts, USA) with a tidal vol-ume of 10 to 15 mL/kg BW, to maintain an end-tidal carbon dioxide (ETCO2) of 35 to 45 mmHg and an end-tidal isoflurane (ETIso) of 1.5%. All dogs were administered isotonic fluids (Plasma-Lyte A; Baxter) IV at a rate of 5 to 10 mL/kg BW per hour.

Instrumentation for monitoring was completed within 5 min after induction and included a Doppler (Benson Medical Industries, Markham, Ontario) for indirect systolic blood pressure and a multi parameter monitor (Carescape B650 Monitor; GE Healthcare, Helsinki, Finland) for heart rate and rhythm, respiratory rate, peripheral capillary oxygen saturation (SpO2), ETIso, ETCO2, and esophageal temperature. Systolic arterial pressure was maintained above 90 mmHg by adjusting the IV fluid rate and by administer-ing a dopamine (Baxter) infusion, 3 to 7 mg/kg BW per minute, as required. Anesthetic depth was adjusted in accordance with the following protocol: after surgery was initiated, if the patient dem-onstrated no response to surgical stimulation, the vaporizer setting was adjusted to decrease the ETIso to 1.3%. Dogs were administered hydromorphone, 0.03 mg/kg BW, IV every 2 h during surgery and 1 dose of meloxicam (Metacam; Boehringer Ingelheim, Burlington, Ontario), 0.1 mg/kg BW, IV was administered just before extubation.



The overall quality of anesthetic maintenance was scored at the end of the procedure using a scale from 0 to 4, with 0–indicating the best plane of anesthesia and 4–indicating the least desirable plane of anesthesia (Appendix II). If dogs demonstrated response to sur-gical stimulation, i.e., sudden increases in systolic blood pressure of . 20% from previous value and/or signs of bucking the ventila-tor, the vaporizer setting was adjusted to deliver an ETIso of 1.5% and not readjusted thereafter. If the animal continued bucking the ventilator, lidocaine (Alveda Pharmaceuticals, Toronto, Ontario), 2 mg/kg BW was administered by IV a maximum of 2 times, with 5 min between doses for its minimum alveolar concentration (MAC)-sparing effect. If the response was still not controlled, an additional dose of lidocaine and a dose of hydromorphone, 0.03 mg/kg, BW were administered at , 2 h from the previous hydromorphone dose for a return to a stable anesthetic depth. A circulating warm water blanket and forced warm air were used to maintain body tempera-ture. At the completion of surgery, dogs were transferred to recovery for extubation and pain assessment. Dogs were extubated when they were spontaneously breathing, developed a brisk palpebral reflex, and the eye position had returned from a ventral to a more central position. Immediately after extubation, the dogs were scored for quality of recovery using a scale from 0 to 3, with 0–indicating profound sedation and 3–indicating no sedation to an excitable recovery (Appendix III).

Pain assessments were completed at 0, 15, and 30 min and at 1, 2, 3, 6, 9, 12, and 24 h after extubation using the shortened Glasgow Composite Measure Pain Scale (CMPS-SF) (11) (Appendix IV). Postoperative analgesia with hydromorphone, 0.03 to 0.05 mg/kg BW, IV or IM was scheduled routinely at 3 to 6 h post-extubation and repeated q6 to q8h, as deemed appropriate. If the dog received a pain score of 5/20 or 6/24 earlier than the postoperative scheduled analgesia, rescue analgesia with hydromorphone was administered. For dogs that underwent thoracic limb amputation and had pain dif-fusion catheters placed during surgery, bupivacaine (Marcaine 0.5%; Hospira Healthcare, Montreal, Quebec) infusion was also provided at 4 h after surgery was completed or as part of the rescue analgesia if required by the dog.

Brachial plexus blockDogs were assigned to 1 of the 3 techniques [TA, PA, or AA (n = 8/

group)] using a randomization scheme (http://www.randomization.com). After instrumentation, the brachial plexus of the affected limb was blocked by a Board-certified anesthetist (AV) and the anesthetist (AS, RA) carrying on the anesthesia of the dog remained unaware of the type of block used.

The blocks were carried out using previously described land-marks (3). For the TA, the dog was placed in lateral recumbency with the limb to be blocked up in a relaxed position and extended away from the body, perpendicular to the long axis of the torso, at a standing angle. A 20-gauge, 2.5- to 3.5-inch spinal needle (BD Medical, Franklin Lakes, New Jersey, USA) was introduced medial to the limb between the scapula and caudal section of the jugular groove, in a craniocaudal direction, from the midpoint between the transverse process of C6 and the point of the shoulder, and advanced beyond the scapulohumeral joint and just cranial to the first rib. For the PA, the dog was also placed in lateral recumbency with the limb Ta

ble

I. D

escr

ipti

on o

f do

gs a

nest

heti

zed

wit

h is

oflu

rane

und

ergo

ing

thor

acic

lim

b su

rger

y [n

o am

puta

tion

(n

= 1

5) v

ersu

s am

puta

tion

(n

= 9

)] t

hat

rece

ived

0.

2 m

L/kg

BW

of

bupi

vaca

ine

at t

he b

rach

ial p

lexu

s be

fore

sur

gery

.

In

tra-o

pera

tive

Anes

thes

ia

N

o

Brac

hial

Ag

e Bo

dy w

eigh

t Ac

epro

maz

ine

Prop

ofol

hy

drom

orph

one

time

Surg

ery

ampu

tatio

n/

pl

exus

blo

ck

(y)

(BW

, kg)

(m

g/kg

BW

) (m

g/kg

BW

) do

ses

(min

) tim

e (m

in)

ampu

tatio

n Su

rgic

al p

roce

dure

Trad

ition

al

5.5

26.8

0.

03

2.3

2 23

5 12

3 4/

4 4

ampu

tatio

ns, 1

radi

us/u

lna

(1.9

to 9

.5)

(24.

6 to

31.

9)

(0.0

25 to

0.0

35)

(2.0

to 3

.2)

(1 to

2)

(173

to 2

45)

(83

to 1

39)

fra

ctur

e, 1

sho

ulde

r arth

rosc

opy,

1

mas

s re

mov

al

Perp

endi

cula

r 2.

6 20

.5

0.02

5 3.

1 1

185

85

5/3

3 ra

dius

/uln

a fra

ctur

es,

(0

.4 to

7.8

) (4

.25

to 3

2.1)

(0

.02

to 0

.03)

(2

.4 to

3.3

) (1

to 1

.5)

(118

to 2

09)

(47

to 1

08)

2

ampu

tatio

ns, 1

hum

eral

co

ndyl

ar fr

actu

re, 1

sho

ulde

r/el

bow

arth

rosc

opy,

1 e

lbow

ar

thro

scop

y

Axill

ary

2.9

25.8

0.

03

2.8

1 16

5 52

6/

2 3

ampu

tatio

ns, 3

radi

us/u

lna

(0

.8 to

6.5

) (9

.9 to

39.

4)

(0.0

2 to

0.0

3)

(2.3

to 3

.6)

(1 to

1)

(150

to 2

08)

(46

to 9

0)

fra

ctur

es, 1

hum

eral

frac

ture

, 1

digi

t am

puta

tion

Dat

a ex

pres

sed

as m

edia

n (lo

wer

-upp

er q

uarti

le).

No

sign

ifica

nt d

iffer

ence

s de

tect

ed a

mon

g gr

oups

.



to be blocked up and in the same position as for the TA. The spinal needle was introduced perpendicular to the scapula, caudal to the scapulohumeral joint, and aimed cranial to the first rib, at a level that corresponded to the midpoint between the transverse process of C6 and the point of the shoulder, until the tip of the needle was just cranial to the first rib. For the AA, the dog was placed in dorsal recumbency with the limb to be blocked at a standing angle. The spinal needle was introduced at the level of the thoracic inlet in a ventral-to-dorsal direction, lateral to the manubrium and medial to the limb, parallel and cranial to the first rib, and advanced to the level of the midpoint between the transverse process of C6 and the point of the shoulder, until the tip of the needle was just cranial to the first rib. Each dog was administered 0.2 mL/kg BW of bupiva-caine 0.5% for the block.

Statistical analysisThe number of dogs in the study was determined by considering a

2-point difference between the means of any 2 groups and a standard deviation of 1, in the pain scores determined using the shortened Glasgow Pain Score, or by considering a 0.2% difference between the means of any 2 groups and a standard deviation of 0.12% in the ETIso. For a type-I error (a) of 0.05 and a desired power of 0.90, a sample size of 7 dogs per group was required.

Comparisons among the 3 groups were completed using a Kruskal-Wallis test. Because of the number of amputations, the data from the 3 groups were also rearranged and analyzed using a Wilcoxon rank sum test for an additional comparison of dogs under-going amputation versus no amputation, which included dogs from all of the 3 groups. A Monte Carlo estimate was used for significance (P , 0.05). Statistical analysis was carried out using Version 9.4, SAS program (SAS Institute, Cary, North Carolina, USA).

Re s u l t sDogs underwent a variety of surgical procedures of the thoracic

limb, which included 10 fracture repairs, 9 amputations, 3 arthro-scopies, 1 mass removal, and 1 fourth digit amputation. There were no differences in characteristics among the 3 groups (Table I). Dogs were an average of 3.5 y old [range: 0.9 to 6.9 y; median and 95% confidence interval (CI)] and weighed an average of 26.8 kg (18.0 to 31.1 kg).

There were no differences in the acepromazine dose used for preanesthetic sedation, the propofol dose required for orotracheal intubation, or the total anesthesia time among groups (Table I). There were no differences in sedation, anesthetic, or recovery scores, or cardiorespiratory variables and body temperature among groups (Table II). Five dogs in the TA group, 3 in the PA group, and 4 in the AA group required dopamine support for hypotension. Six of these dogs were treated with glycopyrrolate for initial bradycardia and hypotension (4 in the TA group, 1 in the PA group, and 1 in the AA group). Three dogs only required a single dose of glycopyrrolate and no further cardiovascular support (2 in the PA group and 1 in the AA group).

Only 1 dog from the PA group received 2 doses of lidocaine for rescue analgesia in the intra-operative period. Based on the dura-tion of surgery and the planned frequency of hydromorphone Ta

ble

II. A

nest

heti

c pa

ram

eter

s in

dog

s an

esth

etiz

ed w

ith

isof

lura

ne u

nder

goin

g th

orac

ic li

mb

surg

ery

that

rec

eive

d 0.

2 m

L/kg

BW

of

bupi

vaca

ine

at t

he b

rach

ial

plex

us b

efor

e su

rger

y us

ing

1 of

3 a

nato

mic

al a

ppro

ache

s (n

= 8

per

gro

up).

Sy

stol

ic

Brac

hial

En

d-tid

al

End-

tidal

ar

teria

l H

eart

rate

plex

us

Pain

Se

datio

n An

esth

esia

Re

cove

ry

isof

lura

ne

CO2

pres

sure

(b

eats

per

Te

mpe

ratu

re

bloc

k sc

ore

scor

e sc

ore

scor

e (%

) (m

mH

g)

(mm

Hg)

m

in)

(°C)

Trad

ition

al

2.1

2 0

0.5

1.41

40

95

83

36

.5

(1

.6 to

2.8

) (0

.5 to

3)

(0 to

0.5

) (0

to 1

) (1

.37

to 1

.45)

(3

9 to

40)

(9

2 to

103

) (7

8 to

94)

(3

6.4

to 3

6.9)

Perp

endi

cula

r 1.

7 2

0 0

1.40

40

10

3 10

0 36

.1

(1

.5 to

3.3

) (1

.5 to

2)

(0 to

0.5

) (0

to 2

) (1

.35

to 1

.44)

(3

7 to

43)

(9

9 to

122

) (9

1 to

104

) (3

5.3

to 3

7.0)

Axill

ary

1.6

2 0

0 1.

38

41

93

87

36.2

(1

.2 to

2.0

) (1

.5 to

2)

(0 to

0)

(0 to

0)

(1.3

6 to

1.3

9)

(38

to 4

4)

(87

to 1

05)

(74

to 9

7)

(35.

8 to

36.

9)D

ata

expr

esse

d as

med

ian

(low

er-u

pper

qua

rtile

). N

o si

gnifi

cant

diff

eren

ces

dete

cted

am

ong

grou

ps.



administration during anesthesia, 5 dogs received a single dose and 3 dogs received 2 doses in the TA group. In the PA group, 5 dogs received a single dose, 2 dogs received 2 doses, and 1 dog did not receive intra-operative analgesia and in the AA group, all dogs received a single dose (Table I).

The recommendation for the CMPS–SF is to use rescue analge-sia if postoperative pain scores are $ 5/20 (non-ambulatory) or $ 6/24 (ambulatory) (11). The number of doses of hydromorphone received in the 24-h postoperative period was similar for each group: 3 (1.6 to 4.0; median and 95% CI) for the TA group; 3 (2.0 to 3.2) for the PA group; and 2 (0.8 to 3.0) for the AA group. For the TA group, 4 dogs had scores $ 5 for a total of 8 assessments; 3 dogs had amputations with 5 scores $ 5, including 1 score before anes-thesia; and 1 young dog (3 mo of age) had a fractured radius/ulna repair with 3 scores $ 5. For the PA group, 6 dogs had scores $ 5 for a total of 12 assessments; 2 dogs had amputations with 5 scores $ 5; whereas the other 4 dogs did not have amputations and had 7 scores $ 5, including 2 scores before anesthesia. For the AA group, 2 dogs had 2 scores $ 5; 1 dog had an amputation and the other did not have an amputation and showed this score before anesthesia (Table III).

Nine of 24 dogs had thoracic limb amputations: 4 dogs in the TA, 3 dogs in the PA, and 2 dogs in the AA group. Dogs undergoing tho-racic limb amputations had higher pain scores in the first 3 h post-extubation, higher recovery scores, higher anesthesia scores, longer anesthetic times, longer surgery times, were older, and weighed more than dogs not undergoing thoracic limb amputations (Table IV).

D i s c u s s i o nIn this study, the anesthetic maintenance score, recovery score,

and postoperative analgesia resulting from 3 blind brachial plexus block techniques in canine patients undergoing surgical procedures of the thoracic limb were compared and evaluated. The results of this study supported the hypothesis that the analgesia provided by the 3 techniques was not different if landmarks are properly identified to facilitate the location of the brachial plexus (3). Blind techniques are inexpensive and, if anatomical landmarks are accurately iden-

tified, they can be clinically effective. A dog cadaver study using blind techniques for staining of the saphenous, obturator, and lateral cutaneous femoral nerves also showed consistent and optimal nerve staining when anatomical landmarks were followed closely and this should result in effective clinical blocks with local anesthetics (12).

Pain evaluation in veterinary species is challenging, particularly when comparing the degree of pain across individuals. The CMPS-SF scale is validated for clinical use in dogs undergoing surgery (11). We detected differences in analgesic comfort when comparing dogs that underwent thoracic limb amputation with those that did not. Amputation of the thoracic limb in dogs includes a skin incision that begins at the most dorsal aspect of the scapular spine and dissection of the muscles participating in synsarcosis to reflect the scapula from the body wall (13). Innervation to the skin of the caudal cervical region is provided by the dorsal rami of the C5, C6, T2, and T3 spinal nerves and innervation to the trapezius muscle, omotransversarius muscle, and part of the brachiocephalicus muscle is provided by cranial nerve 11 (spinal accessory nerve) (2,6), none of which would be anesthetized by the brachial plexus block. This innervation of the dorsal shoulder and surrounding skin would explain why dogs undergoing amputation of the thoracic limb did not experience the same degree of postoperative comfort as dogs undergoing surgical procedures of the antebrachium and distal forelimb, as demonstrated by higher postoperative pain scores. The BPB appeared to be use-ful, however, since the maximum number of dogs requiring rescue analgesia was 3 out of 9 in the first 15 min post-surgery.

Since the dogs in this study were undergoing major surgical procedures, the anesthetic protocol for these dogs used a balanced anesthetic technique. Such techniques have become popular across species in veterinary medicine and refer to the use of a combination of analgesic and anesthetic drugs to exploit the benefits of each drug, while minimizing the adverse effects that would occur from large doses of a single anesthetic agent, which are typically inhal-ant anesthetics (14). While inhalant anesthetics are most commonly used in veterinary medicine to maintain general anesthesia, large doses can result in severe cardiovascular and respiratory depression (14). In dogs, opioids are known to decrease inhalant requirements to maintain a surgical plane of anesthesia. Hydromorphone is a

Table III. Number of dogs with Glasgow composite pain scores equal to or higher than 5/20 (non-ambulatory) or 6/24 (ambulatory) indicating need for rescue analgesia after thoracic limb surgery [no amputation (n = 15) versus amputation (n = 9)] that received 0.2 mL/kg BW of bupivacaine at the brachial plexus before surgery.

Number of dogs before Number of dogs at time post-extubation (h)Surgery anesthesia 0.25 0.5 1 2 3 6 9 12 16 24No amputation 3 1 1 1 2 1 0 1 1 0 0 (5,6,7) (7) (7) (7) (5,7) (7) (6) (6) [PA, AA, PA] [PA] [PA] [PA] [PA, TA] [PA] [TA] [TA]

Amputation 1 3 2 1 2 1 0 1 0 0 0 (6) (8,9,9) (8,9) (8) (5,6) (7) (11) [TA] [PA, TA, TA] [PA, TA] [PA] [PA, TA] [PA] [AA]Value in brackets indicates the specific score for each dog. Letters in square brackets indicate the group (TA — traditional approach; PA — perpendicular approach; AA — axillary approach).



m-opioid receptor agonist that was chosen as part of the balanced anesthetic protocol in this study because it has been demonstrated to both reduce isoflurane requirements (15) and to provide effective postoperative analgesia in dogs (16,17).

The minimum alveolar concentration (MAC) is a measure of potency of volatile anesthetic agents and is defined as the concen-tration of inhalant anesthetic required to prevent purposeful move-ment in response to a noxious stimulus in 50% of a population (18). In order to prevent purposeful movement in 95% of a population in response to a noxious stimulus (a surgical plane of anesthesia), values of 1.2 to 1.4 times MAC are typically required (14,18). The MAC of isoflurane in dogs varies depending on the study, but is typically reported to be 1.2% to 1.4% (15,18,19). For consistency in this study, the initial ETIso was initially set at 1.5%, considering the reported range of MAC of isoflurane for dogs and the use of anesthetic drugs and techniques that would decrease their inhalant requirements, including acepromazine (20), hydromorphone (15), and the brachial plexus block. Dogs were kept at an ETIso of approx-imately 1.4% in the 3 groups, which corresponds to 1.0 to 1.2 MAC. Cardiorespiratory parameters were similar among the 3 groups and values were considered optimal for anesthetized dogs. The number of dogs requiring blood pressure support with hypotension was also similar among groups. The hypotension that occurred during anesthesia may be the result of the dose-dependent vasodilatory effects of isoflurane (21), since the initial ETIso of 1.5% exceeded the 1.0 MAC value reported for isoflurane in dogs. In addition, the use of hydromorphone and acepromazine in the premedication may have exacerbated the cardiovascular depression of isoflurane from synergistic effects.

This study had several limitations that are associated with con-ducting clinical research, including compliance of owners and clinicians responsible for the cases. The dogs included in this study were administered a variety of analgesics intraoperatively for pain management, including an opioid, a non-steroidal anti-inflammatory, and a local block technique with a local anesthetic. The anesthetic protocol was the same for all dogs included in the study and thus the authors believe that the differences detected, or lack thereof, have to do with the block technique rather than inclusion of other analgesics. A placebo group was not included in this study because the authors considered it unethical. As the objective of the study was to compare the analgesic comfort that resulted from the 3 BPB techniques, it was anticipated that it would be difficult to get owners to enroll in the study if analgesic coverage was restricted.

Another limitation was that dogs were also administered post-operative analgesia according to a pre-scheduled format, follow-ing the standards of our hospital and compliance with clinicians responsible for the cases. Analgesia was restricted in the first 3 h post-extubation to allow for a more critical assessment of the degree of analgesia that resulted from the brachial plexus block. Thereafter, analgesia was included as deemed necessary, which includes regu-lar dosing of hydromorphone for this type of surgeries q6 to q8h, sometimes regardless of the pain score, to prevent discomfort, to facilitate the handling of the patient, and to comply with clinicians’ routine orders for their cases. While this type of analgesic manage-ment could have influenced the pain scores, the contribution of the BPB to analgesia should have been significant, based on studies that Ta

ble

IV. A

nest

heti

c pa

ram

eter

s in

dog

s an

esth

etiz

ed w

ith

isof

lura

ne u

nder

goin

g fo

relim

b su

rger

y [n

o am

puta

tion

(n =

15)

ver

sus

ampu

tati

on (n

= 9

)] t

hat

rece

ived

0.

2 m

L/kg

BW

of

bupi

vaca

ine

at t

he b

rach

ial p

lexu

s be

fore

sur

gery

.

Pa

in s

core

Pa

in s

core

Post

-ope

rativ

e

Anes

thes

ia

Surg

ery

0

to 3

h

0 to

4 h

Re

cove

ry

hydr

omor

phon

e An

esth

esia

tim

e tim

e Ag

e Bo

dy w

eigh

t Su

rger

y po

st-e

xtub

atio

n po

st-e

xtub

atio

n sc

ore

(dos

es)

scor

e (m

in)

(min

) (y

) (k

g)N

o am

puta

tion

1.6*

1.

5 0*

3*

0*

16

0*

50*

1.0*

20

.2*

(1

.4 to

2.2

) (1

.2 to

2.2

) (0

to 0

) (3

to 3

) (0

to 0

) (1

40 to

230

) (4

1 to

90)

(0

.5 to

3.0

) (5

.2 to

28.

3)

Ampu

tatio

n 2.

4 2.

0 1

4 0

203

115

8.0

30.4

(2.4

to 3

.0)

(1.8

to 2

.8)

(0 to

2.0

) (4

to 5

) (0

to 1

) (1

86 to

235

) (1

00 to

138

) (6

.5 to

10.

0)

(27.

6 to

36.

2)

P-va

lue

0.01

1 0.

081

0.00

5 .

0.0

001

0.04

8 0.

071

0.00

6 0.

0001

0.

01D

ata

expr

esse

d as

med

ian

(low

er-u

pper

qua

rtile

).*

Sign

ifica

ntly

low

er th

an fo

r am

puta

tion.



have demonstrated long-lasting analgesic effects of the BPB for up to 11 h after bupivacaine and epinephrine administered at a dose of 4 mg/kg (volume of 1.07 mL/kg BW) (22) and up to 3 h using 1 mg/kg (volume of 0.4 mL/kg BW) (23).

In conclusion, the 3 anatomical approaches evaluated here to block the brachial plexus in canine patients undergoing surgical procedures of the thoracic limb provided a similar degree of comfort in the post-operative period. In addition, while this study demon-strated that dogs undergoing thoracic limb amputations exhibited less comfort postoperatively than dogs not undergoing thoracic limb amputations, this type of block offers a significant degree of comfort based on pain scores obtained in the postoperative period.

A c k n o w l e d g m e n t sThis article represents a portion of a thesis submitted by

Dr. Skelding to the Department of Clinical Studies, University of Guelph as partial fulfillment of the requirements for a Doctor of Veterinary Science degree. The study was supported by a grant from Pet Trust of the Ontario Veterinary College.

Appendix I. Sedation score at the time of catheterization or induction.

0 Bright and alert, no apparent sedation and/or excitable-dysphoric (excited, anxious, difficult to restrain in recumbency, very interactive and responsive, vocalizing, reactive to voice and touch).

1 Calm, minimal sedation (quiet but still alert and aware of surroundings, mild resistance to restraint for lateral recumbency, moderate response to voices and touch, mild resistance to catheterization).

2 Very calm, with moderate sedation (quiet, relaxed, minimal restraint required for lateral recumbency, mild response to voices or touch, no resistance to catheterization).

3 Profound sedation (quiet, very relaxed, no restraint necessary for lateral recumbency, does not respond to voice or touch).

Modified from Liao et al (10).

Appendix II. Anesthetic maintenance quality score.

0 No adjustments required in the depth of anesthesia during maintenance. No autonomic responses (less than 20% increase in BP from previous value and/or signs of bucking the ventilator).

1 Dog shows autonomic responses (BP increases suddenly . 20% from previous value and/or signs of bucking the ventilator), which require increasing the end-tidal isoflurane to 1.5% for return of a stable anesthetic depth.

2 Dog shows autonomic responses (BP increases suddenly . 20% increase from previous value and/or signs of bucking the ventilator), which require increasing the end-tidal isoflurane to 1.5% and administration of 1 to 2 boluses of lidocaine (2 mg/kg BW, IV) for return of a stable anesthetic depth.

3 Dog shows autonomic responses (BP increases suddenly . 20% increase from previous value and/or signs of bucking the ventilator), which require increasing the end-tidal isoflurane to 1.5%, administration of . 2 boluses of lidocaine (2 mg/kg BW, IV), and repeat of hydromorphone (0.03 mg/kg BW) , 2 h from the previous dose for return of a stable anesthetic depth.

4 As 3, but there is no return of a stable anesthetic depth despite all actions.


Appendix III. Recovery score from the time of extubation to the first 15 min.

0 Profound sedation (quiet, very relaxed, no restraint necessary for lateral recumbency, does not respond to voice or touch).

1 Very calm, with moderate sedation (quiet, relaxed, minimal restraint required for lateral recumbency, mild response to voices or touch).

2 Calm, however, minimal sedation (quiet but still alert and aware of surroundings, mild resistance to restraint for lateral recumbency, moderate response to voices and touch).

3 Minimal or no apparent sedation and/or excitable-dysphoric (excited, anxious, difficult to restrain in recumbency, very interactive and responsive, vocalizing, reactive to voice and touch).




Re f e r e n c e s 1. Allam MW, Lee DG, Nulsen FE, Fortune EA. The anatomy of the

brachial plexus of the dog. Anat Rec 1952;114:173–179. 2. Dyce KM, Sack WO, Wensing CJG. Textbook of Veterinary

Anatomy. 4th ed. St. Louis, Missouri: Saunders Elsevier, 2010:321–323.

3. Skelding A, Valverde A, Sinclair M, Thomason J, Moens N. Ana-tomical characterization of the brachial plexus in dog cadavers and comparison of three blind techniques for blockade. Vet Anaesth Analg 2018;45:203–211.

4. Sakonju I, Maeda K, Maekawa R, Maebashi R, Kakuta T, Takase K. Relative nerve blocking properties of bupivacaine and ropivacaine in dogs undergoing brachial plexus block using a nerve stimulator. J Vet Med Sci 2009;71:1279–1284.

5. Campoy L, Read M. Small Animal Regional Anesthesia and Analgesia. 1st ed. Ames, Iowa: Wiley-Blackwell, 2013:141–165.

6. Campoy L, Read M, Peralta S. Canine and feline local anes-thetic and analgesic techniques. In: Grimm KA, Lamont LA, Tranquilli WJ, Greene SA, Robertson SA, eds. Veterinary Anes-thesia and Analgesia. 5th ed. Hoboken, New Jersey: John Wiley & Sons, 2015:827–857.

7. Campoy L, Bezuidenhout AJ, Gleed RD, et al. Ultrasound-guided approach for axillary brachial plexus, femoral nerve, and sciatic nerve blocks in dogs. Vet Anaesth Analg 2010;37:144–153.

8. Mahler SP, Adogwa AO. Anatomical and experimental studies of brachial plexus, sciatic, and femoral nerve-location using peripheral nerve stimulation in the dog. Vet Anaesth Analg 2008;35:80–89.

Appendix IV.

A. Look at dog in Kennel Is the dog (i) Quiet 0 Crying or whimpering 1 Groaning 2 Screaming 3

In the case of spinal, pelvic, or multiple limb fractures, or where assistance is required to aid locomotion do not carry out section B and proceed to CPlease tick if this is the case then proceed to C

B. Put lead on dog and lead out of the kennel When the dog rises/walks is it? (iii) Normal 0 Lame 1 Slow or reluctant 2 Stiff 3 It refuses to move 4

D. Overall Is the dog? (v) Happy and content or happy and bouncy 0 Quiet 1 Indifferent or non-responsive to surroundings 2 Nervous or anxious or fearful 3 Depressed or non-responsive to stimulation 4

Total Score (i1ii1iii1iv1v1vi) = ______________________________

Reid et al (11).

(ii) Ignoring any wound or painful area 0 Looking at wound or painful area 1 Licking wound or painful area 2 Rubbing wound or painful area 3 Chewing wound or painful area 4

C. If it has a wound or painful area including abdomen, apply gentle pressure 2 inches round the site Does it? (iv) Do nothing 0 Look round 1 Flinch 2 Growl or guard area 3 Snap 4 Cry 5

Is the dog? (vi) Comfortable 0 Unsettled 1 Restless 2 Hunched or tense 3 Rigid 4



9. Mosing M, Reich H, Moens Y. Clinical evaluation of the anaes-thetic sparing effect of brachial plexus block in cats. Vet Anaesth Analg 2010;37:154–161.

10. Liao P, Sinclair M, Valverde A, et al. Induction dose and recovery quality of propofol and alfaxalone with or without midazolam coinduction followed by total intravenous anesthesia in dogs. Vet Anaesth Analg 2017;44:1016–1026.

11. Reid J, Nolan AM, Hughes JML, Lascelles D, Dawson P, Scott EM. Development of the short-form Glasgow Composite Measure Pain Scale (CMPS-SF) and derivation of an analgesic intervention score. Anim Welf 2007;16:97–104.

12. Echeverry-Bonilla DF, Pelaez JT, Buriticá EF, Laredo FG. Assessment of the potential efficacy of blind perineural injec-tion techniques for blockade of the saphenous, obturator, and lateral cutaneous femoral nerves in dog cadavers. Am J Vet Res 2017;78:412–420.

13. Smeak DD, Lisano S. Step-by-step forelimb amputation. Clinician’s Brief. Available from: http://www.cliniciansbrief.com/article/step-step-forelimb-amputation Last accessed May 1, 2019.

14. Ilkiw JE. Balanced anesthetic techniques in dogs and cats. Clin Tech Small Anim Pract 1999;14:27–37.

15. Machado CE, Dyson DH, Grant Maxie M. Effects of oxymor-phone and hydromorphone on the minimum alveolar concentra-tion of isoflurane in dogs. Vet Anaesth Analg 2006;33:70–77.

16. Hoelzler MG, Harvey RC, Lidbetter DA, Millis DL. Comparison of perioperative analgesic protocols for dogs undergoing tibial plateau leveling osteotomy. Vet Surg 2005;34:337–344.

17. Bateman SW, Haldane S, Stephens JA. Comparison of the anal-gesic efficacy of hydromorphone and oxymorphone in dogs and cats: A randomized blinded study. Vet Anaesth Analg 2008; 35:341–347.

18. Eger EI, Saidman LJ, Brandstater B. Minimum alveolar concen-tration: A standard of anesthetic potency. Anesthesiology 1965; 26:756–763.

19. Valverde A, Morey TE, Hernández, J, Davies W. Validation of several types of noxious stimuli for use in determining the mini-mum alveolar concentration for inhalation anesthetics in dogs and rabbits. Am J Vet Res 2003;64:957–962.

20. Boyd CJ, McDonell WN, Valliant A. Comparative hemodynamic effects of halothane and halothane-acepromazine at equipotent doses in dogs. Can J Vet Res 1991;55:107–112.

21. Mutoh T, Nishimura R, Kim HY, Matsunaga S, Sasaki N. Cardiopulmonary effects of sevoflurane, compared with halo-thane, enflurane, and isoflurane, in dogs. Am J Vet Res 1997;58: 885–890.

22. Futema F, Fantoni DT, Auler JO Jr, Cortopassi SR, Acaui A, Stopiglia AJ. A new brachial plexus block technique in dogs. Vet Anaesth Analg 2002;29:133–139.

23. Akasaka M, Shimizu M. Comparison of ultrasound- and electrostimulation- guided nerve blocks of brachial plexus in dogs. Vet Anaesth Analg 2017;44:625–635.


Article


I n t r o d u c t i o nOsteoarthritis (OA), also known as degenerative joint disease, is a

common clinical disease in humans, dogs, cats, and other companion animals and is one of the leading causes of disability and morbidity around the world (1,2). OA is characterized clinically by joint pain, stiffness, and functional disability and radiographically by narrow-ing of joint spaces and formation of osteophytes (bone spurs) (3). While OA can affect any joint in the body, it most commonly affects joints in knees, feet, hips, and spine (4).

OA results in the loss of articular cartilage within the joint of a bone. The articular cartilage is made up of a sparse distribution of

chondrocytes within a dense extracellular matrix (ECM) (5). Under normal conditions, chondrocytes maintain a balance between the synthesis and the degradation of the ECM (6). In OA, however, altered chondrocyte activity leads to a gradual loss of the articular cartilage (6). Inflammatory mediators, such as interleukin-1b (IL-1b), tumor necrosis factor alpha (TNFa), IL-6, IL-8, and prostaglandins, which are also produced and released by the chondrocytes, activate the degradation of the joint, leading to collagen and proteoglycan breakdown (7).

There is no known cure for OA and medical therapy has focused on providing symptomatic relief and maintaining joint function. For symptomatic relief, non-steroidal anti-inflammatory drugs

Effects of various dietary supplements on inflammatory processes in primary canine chondrocytes as a model of osteoarthritis

Eman A. AlRaddadi, Tanja Winter, Harold M. Aukema, Donald W. Miller

A b s t r a c tThe use of dietary supplements as an alternative treatment for joint-related pathologies such as osteoarthritis (OA) is increasing. However, there is little scientific evidence to support the intended use. The aim of this study was to evaluate the anti-inflammatory effects of creatine- and amino acid-based supplements in primary cultured canine chondrocytes (CnCs) as an in-vitro model of OA and compare the effects to more commonly used agents, such as the non-steroidal anti-inflammatory drug (NSAID), carprofen, and the joint supplement, glucosamine (GS). CnCs were stimulated with interleukin-1b (IL-1b) and the subsequent release of prostaglandin E2 (PGE2) and tumor necrosis factor alpha (TNFa) was measured using an enzyme-linked immunosorbent assay (ELISA). Changes in oxylipins were also assessed using high-performance liquid chromatography/tandem mass spectrometry (HPLC/MS/MS). All compounds examined were able to significantly reduce the release of PGE2 and TNFa and were associated with reductions in cyclooxygenase-2 (COX-2) expression and nuclear factor-kappaB (NF-kB) phosphorylation. The creatine- and amino acids-based supplements also altered the profile of oxylipins produced. All compounds examined were less effective at reducing the release of PGE2 than carprofen. Carprofen significantly increased release of TNFa from CnCs, however, while the other agents reduced TNFa release. This study suggests that creatine- and amino acid-based supplements may have a beneficial role in preventing inflammation within the joint and that further studies are warranted.

R é s u m éL’utilisation de suppléments alimentaires à titre de traitement alternatif pour les pathologies associées aux articulations telle que l’arthrose (OA) est en augmentation. Toutefois, il y a peu d’évidences scientifiques qui supportent l’utilisation proposée. L’objectif de la présente étude était d’évaluer les effets anti-inflammatoires de suppléments à base de créatine et d’acides aminés sur des cultures primaires de chondrocytes canins (CnCs) utilisés comme modèle in vitro d’OA et de comparer les effets à des agents plus communément utilisés, tel que l’agent anti-inflammatoire non-stéroïdien (AINS) carprofen, et le supplément articulaire, glucosamine (GS). Les CnCs furent stimulés avec de l’interleukine-1b (IL-1b) et la libération subséquente de prostaglandine E2 (PGE2) et le facteur nécrosant de tumeur alpha (TNFa) fut mesurée par épreuve immuno-enzymatique (ELISA). Les changements dans les oxylipines furent également mesurés par chromatographie en phase liquide à haute performance/spectrométrie de masse tandem (HPLC/MS/MS). Tous les composés examinés étaient en mesure de réduire significativement la libération de PGE2 et de TNFa et étaient associés avec des réductions d’expression de cyclooxygénase-2 (COX-2) et de phosphorylation du facteur nucléaire kappaB (NF-kB). Les suppléments à base de créatine et d’acides aminés ont également altéré le profil des oxylipines produits. Tous les composés examinés étaient moins efficaces que le carprofen pour réduire la libération de PGE2. Le carprofen augmentait significativement la libération de TNFa par les CnCs, alors que les autres agents la réduisaient. La présente étude suggère que les suppléments à base de créatine et d’acides aminés pourraient avoir un rôle bénéfique dans la prévention de l’inflammation dans l’articulation et que des études supplémentaires sont requises.


Department of Pharmacology and Therapeutics, The Rady Faculty of Health Sciences, University of Manitoba, Winnipeg, Manitoba R3E 3J7 (AlRaddadi, Miller); Department of Food and Human Nutritional Sciences, University of Manitoba, Winnipeg, Manitoba R3T 2N2 (Winter, Aukema).

Address all correspondence to Dr. Donald W. Miller; telephone: (204) 789-3278; fax: (204) 789-3932; e-mail: [email protected]

Received July 4, 2018. Accepted September 24, 2018.



(NSAIDs) are a mainstay of OA therapy. Due to the adverse effects of long-term NSAIDs therapy, including gastrointestinal irritation and bleeding (3,8), alternative approaches to relieving joint pain have been explored. Nutraceuticals, such as glucosamine (GS) and chondroitin, represent the largest category of natural supplements for veterinary medicine (9,10). In humans, these nutraceuticals have been reported to reduce pain and stiffness of the joints in patients suffering from OA (3). Interest in creatine and amino acid supple-ments as potential adjuvants in the treatment of OA is also growing (11,12). Clinical trials examining the use of GS and chondroitin in OA have conflicting results, however, which highlights the need for more basic fundamental studies on the effects of nutraceuticals in OA (3,13).

The present study examined the anti-inflammatory effects of sev-eral different dietary supplements in cultured canine chondrocytes (CnCs) as an in-vitro model of OA. The dietary supplements exam-ined included creatine monohydrate (CM), creatine hydrochloride (CHCl) (14,15), GS, and the conjugated amino acid supplement, Ethyl-a(guanido-methyl)ethanoate (AlphaGEE). Creatinine (CRN), a stable metabolite of creatine, was also examined. Effects on tra-ditional inflammatory mediators, such as prostaglandin E2 (PGE2) and TNFa, were evaluated along with a more extended oxylipin analysis. The oxylipins are produced by oxygenation of polyunsatu-rated fatty acids and have both inflammatory and anti-inflammatory activity (16).

While all the dietary supplements examined were able to inhibit the release of PGE2, AlphaGEE was the most potent, with significant effects on PGE2 and TNFa release observed at 10 mM. The dietary supplements examined also altered the oxylipin profile of chon-drocytes after inflammatory activation. Interestingly, CRN, which is generally considered to be a biologically inactive metabolite of creatine, also displayed anti-inflammatory activity in multiple assays. These results demonstrate that inflammatory processes within chondrocytes can be modified by treatment with selected amino acids and GS and suggest that further studies with these nutraceuticals as potential alternative or complementary treatments for OA are warranted.

M a t e r i a l s a n d m e t h o d sPrimary CnCs derived from normal canine articular cartilage

were purchased from Cell Applications (San Diego, California, USA) along with complete culture media and culturing reagents. Cell culture flasks, plates, and dishes were obtained from either Corning Incorporated (Corning, New York, USA) or Thermo Fisher Scientific (Waltham, Massachusetts, USA). Recombinant canine IL-1b was purchased from R&D Systems (Minneapolis, Minnesota, USA). CHCl and AlphaGEE were provided by Vireo Systems (Madison, Tennessee, USA). CM, CRN, and carprofen (Rimadyl) were obtained from Acros Organics (Geel, Belgium) and Fluka Analytical (St. Gallen, Switzerland), respectively. Canine TNFa ELISA kit and Prostaglandin E2 Parameter Assay kit were purchased from Raybiotech (Norcross, Georgia, USA) and R&D Systems, respectively. Pierce protease inhibitor mini-tablets and the Pierce bicinchoninic acid (BCA) protein assay kit were obtained from Pierce Biotechnology (Rockford, Illinois, USA). PageRuler Plus

Prestained protein ladder, 10 to 250 kDa, was obtained from Thermo Scientific. Immuno-blot polyvinyldifluoride (PVDF) membrane was purchased from Bio-Rad Laboratories (Mississauga, Ontario). Primary rabbit polyclonal antibody for cyclooxygenase-2 (COX-2) was obtained from Abcam (Cambridge, UK). Mouse monoclonal antibody for B-actin was purchased from Sigma Aldrich (St. Louis, Missouri, USA). The secondary horseradish peroxidase (HRP)-linked antibodies were purchased from Novus Biologicals (Littleton, Colorado, USA). All other reagents and chemicals were purchased from Sigma Aldrich.

Canine chondrocyte culturePrimary CnCs were received as passage-1 cells and were plated

in a T-75 flask and grown in CnC complete medium. When reach-ing approximately 80% confluency, CnCs were passaged at a seed-ing density of 8000 to 15 000 cells/cm2. Cells were maintained in a humidified environment at 37°C with 5% carbon dioxide (CO2), with media replacement every other day until the cells reached 60% to 70% confluence and every day upon reaching confluency. All experiments were conducted on confluent monolayers of CnC (passage number 4).

Cell viability studiesCnCs were seeded in 96-well plates with 5000 cells/well and

incubated for 24 h in culture media. Cells were treated with culture media containing IL-1b (10 ng/mL) in the presence or absence of CM, AlphaGEE, CRN, CHCL, or GS (10 to 1000 mM), or with car-profen (1 to 100 mM). Cells were exposed to these treatments for 72 h. Additional cells treated with culture media alone were used as control. For every control and experimental treatment, 4 wells were used. Cell viability was determined using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. After 72 h, 25 mL of the MTT substrate [5 mg/mL in phosphate-buffered saline (PBS)] were added to each well and incubated for 3 h at 37°C. Viable cells with active metabolism convert MTT into a purple-colored formazan product with an absorbance maximum of approximately 570 nm. After the MTT substrate was added, dimethylsulfoxide (DMSO) was added to solubilize the cells and the resulting absorbance of the formazan product in the cell lysates was measured at a wavelength of 570 nm using a Synergy HT plate reader (BioTek, Winooski, Vermont, USA). Cell viability was expressed as a percentage of the absorbance levels observed in the control cells (media alone).

ELISA for PGE2 and TNFa analysisCnCs were seeded in 96-well plates with 5000 cells/well and incu-

bated for 24 h in culture media. Cells were treated with culture media containing IL-1b (10 ng/mL) alone or in combination with CM, CRN, CHCL, AlphaGEE, GS (10 to 1000 mM), or carprofen (1 to 100 mM). Additional cells were treated with culture media alone. For every control and experimental treatment, 4 wells were used. Culture media was collected at various time points (4 to 72 h) and stored at 280°C until further analysis. Quantitative analysis of PGE2 and TNFa in culture media samples was determined using enzyme-linked immunosorbent assay (ELISA). The ELISA assays were carried out as detailed in the manufacturer’s protocol and the optical density was determined using a Synergy HT plate reader (BioTek) set at 450 nm.



Western blot assays for COX-2 and NF-kBChanges in COX-2 and phosphorylated nuclear factor-kappaB

(NF-kB) protein expression in response to inflammatory stimuli were examined using Western blot analysis. Confluent monolay-ers of primary CnCs grown on 100-mm petri dishes were treated with culture media containing IL-1b (10 ng/mL), alone or with 1000 mM CM, CHCl, CRN, AlphaGEE, or GS, or with 100 mM car-profen. Additional cells were treated with culture media alone. For every control and experimental treatment, 3 petri dishes were used. At various treatment times, cell media was aspirated and the cells were washed twice with ice-cold PBS. Cells were lysed with 100 mL radioimmunoprecipitation assay (RIPA) buffer [150 mM sodium chloride (NaCl), 1.0% IGEPAL CA-630, 0.5% sodium deoxycho-late, 0.1% sodium docecyl sulfate (SDS), and 50 mM Tris, pH 8.0], supplemented with a protease inhibitor cocktail. The cell lysate was sonicated and centrifuged at 15 000 rpm for 10 min at 4°C. Total cell protein supernatant was collected and prepared for electrophoresis by adding to 20 mg of total protein in a 1:1 ratio. Samples were then denatured by heating at 95°C for 5 min and loaded onto a 10% poly-acrylamide gel for electrophoresis at 120 V at room temperature. The protein bands were transferred to an immunoblot PVDF membrane by applying a constant current of 200 mA for 2 h at 4°C.

Subsequently, membranes were washed with 5% (w/v) non-fat skim milk in tris-buffered saline with 0.1% tween-20 (TBS-T) to block non-specific binding sites under constant agitation for 1 h at room temperature. Membranes were then incubated with rabbit polyclonal primary antibody for COX-2 (1 mg/mL) or phospho-NF-kappaB p65 pSer536 antibody (1:1000 dilution) (Thermo Fisher Scientific) overnight at 4°C. Membranes were washed 3 times and incubated with mouse anti-rabbit immunoglobulin G-horseradish peroxidase (IgG-HRP) secondary antibody, diluted to 1:10 000 in blocking buffer for 1 h at room temperature. The specific bands were visual-ized using a chemiluminescence kit (Bio-Rad) and densitometry analysis was carried out using the Bio-Rad ChemiDoc MP imaging system and ImageLab 5.2.1 build 11 software. Protein expression was normalized to b-actin by stripping the membranes using a re-blot solution (Chemicon International, Billerica, Massachusetts, USA). The stripped membranes were then re-blotted with mouse monoclonal primary antibody for b-actin (0.5 mg/mL) and rabbit anti-mouse IgG-HRP.

Oxylipin analysisThe production of oxylipins in response to inflammatory

mediators and the various treatments was examined using high- performance liquid chromatography/tandem mass spectrometry (HPLC/MS/MS). Confluent monolayers of primary CnCs were grown on 6-well plates and treated with culture media containing IL-1b (10 ng/mL) in the presence and absence of CM, CHCl, CRN, or AlphaGEE or GS (1000 mM), or with 100 mM of carprofen. Additional cells were treated with culture media alone. For every control and experimental treatment, 4 wells were used. After treatment, cells were placed in a humidified CO2 incubator maintained at 37°C. After 24 h, 3-mL samples were collected from culture media and antioxidant solution [0.2 mg/mL butylated hydroxytoluene (BHT), 0.2 mg/mL ethylenediamine tetra-acetic acid (EDTA), 2 mg/mL

triphenylphosphine, and 2 mg/mL indomethacin in methanol/ethanol/water (2:1:1, v/v/v)] was added to each sample to protect it from degradation and the samples were stored at 280°C until further analysis.

Oxylipin concentrations in media samples were determined as described by Aukema et al (17). Briefly, after adding 10 ng of deuterated internal standards (Cayman Chemical, Michigan, USA), samples were adjusted to pH , 3. Solid-phase extraction (SPE) was carried out with Strata-X SPE columns (Phenomenex, California, USA) that were preconditioned with methanol and pH 3 water, loaded with sample, rinsed with 10% methanol, and eluted with methanol. Samples were dried and resuspended in solvent for analysis by HPLC/MS/MS (API 6500; Sciex, Concord, Ontario). Quantification of oxylipins was determined using the stable isotope dilution method (17).

Statistical analysisAll data were expressed as mean 6 standard error of the mean

(SEM). Comparison among the groups of data was evaluated by either 1-way analysis of variance (ANOVA), if only 1 variable was present, or 2-way ANOVA, if more than 1 variable was present, using Fisher’s least significant difference test for multiple comparisons of the means. The tests were done using GraphPad Prism, Version 6 software (San Diego, California, USA), with P-values , 0.05 con-sidered significant.

Re s u l t sCell viability of primary cultured CnCs was evaluated after

exposure to both the inflammatory stimulant (IL-1b) and the vari-ous treatment agents (Figure 1). No significant effects on cell viabil-ity were observed after CNCs were exposed to IL-1b for 72 h. Furthermore, none of the treatments examined changed cell viability within a 72-h exposure period.

Release of PGE2 from primary CnC monolayers under control and stimulated conditions

Treatment of cultured CnCs with IL-1b produced a time- dependent increase in PGE2 release, with significant increases observed from 8 to 48 h compared to unstimulated cells (Figures 2 and 3). Treatment of CnCs with the COX inhibitor, carprofen (1 to 100 mM), substantially attenuated the release of PGE2 induced by IL-1b. The reductions in PGE2 release were observed as early as 8 h after carprofen and at all concentrations examined (Figure 2A). Exposure of cells to GS produced significant decreases in PGE2 release from CnC monolayers. The effects were both time- and concentration-dependent, with 58% reduction at 8 h after exposure at the highest concentration examined and more modest reductions at 48 h (Figure 2B). Treatment of CnCs with CM resulted in a significant decrease in PGE2 concentrations compared to IL-1b alone at both 8 and 48 h, up to 55% and 38%, respectively (Figure 3A). Similar reductions in PGE2 release were obtained with CHCl (Figure 3B). The creatine metabolite, CRN, was also able to reduce PGE2 release, with greatest effects observed at 8 h after exposure (Figure 3C). Interestingly, of the dietary supplements examined, AlphaGEE appeared to be the most efficacious and potent at reducing PGE2



Figure 1. Effects of various treatments on cell viability. MTT assay to assess cell viability of primary cultured canine chondrocytes (CnCs) after 72-hour exposure to the inflammatory stimulant (IL-1b) and carprofen (A), CM (B), AlphaGEE (C), CRN (D), CHCL (E), or glucosamine (GS) (F). Values represent the mean 6 SEM of 8 CnC monolayers.

% V

iabi

lity

(ver

sus

Con

trol

)%

Via

bilit

y (v

ersu

s C

ontr

ol)

% V

iabi

lity

(ver

sus

Con

trol

)

% V

iabi

lity

(ver

sus

Con

trol

)%

Via

bilit

y (v

ersu

s C

ontr

ol)

% V

iabi

lity

(ver

sus

Con

trol

)



release from CnCs, with responses observed at concentrations as low as 10 mM and maximal inhibition of 76% at the 8-hour time point (Figure 3D). It should be noted that, while all the dietary supple-ments examined were able to reduce PGE2 release in response to the inflammatory stimulus, the level of reduction was not as great as with carprofen.

COX-2 expression in primary CnC monolayersThe expression of COX-2 was also examined in primary CnC

monolayers under control conditions and after stimulation with IL-1b alone or in combination with various potential anti- inflammatory agents (Figure 4A). There was no detectable COX-2 in CnC lysates under control conditions. In contrast, COX-2 expression

increased significantly when the monolayers were exposed to the pro-inflammatory cytokine, IL-1b (Figure 4A). All the dietary supple-ments examined were able to significantly attenuate COX-2 expres-sion in response to IL-1b (Figure 4A). This was in contrast to the effects of carprofen exposure in CnCs, which significantly enhanced COX-2 expression in response to IL-1b stimulation (Figure 4A).

Release of TNFa and activation of NF-kB in primary CnC monolayers under control and stimulated conditions

Under control conditions, little TNFa was detected in the culture media (Figure 2C and 2D). After stimulation with IL-1b, how-ever, significant increases in TNFa release (2-fold) were observed

Figure 2. Effects of carprofen and glucosamine (GS) on the release of inflammatory mediators from primary cultured canine chondrocytes (CnCs). Release of PGE2 (A and B) or TNFa (C and D) under control conditions and after stimulation with IL-1b with and without carprofen (A and C) or GS (B and D). Values represent the mean 6 SEM of 4 monolayers per treatment group.*P , 0.05, **P , 0.01, ***P , 0.001, compared to IL-1b treatment group (2-way ANOVA), ****P , 0.0001.

PG

E2

Con

cent

ratio

n (p

g/m

L)TN

Fa C

once

ntra

tion

(pg/

mL)

PG

E2

Con

cent

ratio

n (p

g/m

L)TN

Fa C

once

ntra

tion

(pg/

mL)



at the various time points examined (Figure 2C and 2D). Those CnC monolayers stimulated with IL-1b and receiving carprofen had even greater TNFa release than IL-1b alone (Figure 2C). In contrast, treatment of primary cultured CnCs with GS produced slight reductions in TNFa release in response to IL-1b stimula-tion, which was most apparent after 48 h (Figure 2D). For the cre-atine compounds and AlphaGEE, there was a reduction in TNFa release at 24 h at the highest concentration of CM examined and an even more pronounced reduction at all concentrations at 48 h (Figure 5A). Similar effects were observed with CHCL, although maximal effects occurred at 24 h (Figure 5B). Significant reduc-tions in TNFa release after IL-1b stimulation were also observed at 24 and 48 h in the CRN treatment group, but only at the highest

concentration of CRN examined (Figure 5C). Treatment of cells with AlphaGEE led to significant reductions in TNFa release in response to IL-1b stimulation, with maximal responses observed at 24 h (Figure 5D).

As TNFa produces its effects by activating NF-kB transcription factor, the expression of the phosphorylated form of NF-kB was examined under normal conditions and after IL-1b stimulation. As shown in Figure 4B, treating the cells with IL-1b alone (10 ng/mL) for 1 h induced significant expression of phosphorylated-NF-kB compared to untreated control cells. All the creatine compounds, as well as AlphaGEE, GS, and carprofen, significantly reduced the expression of phosphorylated-NF-kB compared to the positive control, with CRN causing the most reduction (85%) (Figure 4B).

Figure 3. Release of PGE2 from primary cultured canine chondrocytes (CnCs) under control conditions and after stimulation with IL-1b with and without CM (A), CHCL (B), CRN (C), or AlphaGEE (D). Values represent the mean 6 SEM of 4 monolayers per treatment group.*P , 0.05, **P , 0.01, ***P , 0.001, compared to IL-1b treatment group (2-way ANOVA), ****P , 0.0001.

PG

E2

Con

cent

ratio

n (p

g/m

L)P

GE

2 C

once

ntra

tion

(pg/

mL)

PG

E2

Con

cent

ratio

n (p

g/m

L)P

GE

2 C

once

ntra

tion

(pg/

mL)



Oxylipin production in CnCs under control and stimulated conditions

Of the 135 oxylipins analyzed, 90 were present at or above the quantitation level of the assay and over 50 were present at a level of $ 0.1 ng/mL in media taken from control cells and those stimulated with IL-1b. Of those oxylipins that were present in measurable amounts in the culture media, 14 showed at least a 2-fold change after a 24-h IL-1b challenge, including PGE1, d17-6k PGF1a, and dihomo PGF2a derived from dihomo gamma linolenic acid, eicosapentaenoic acid, and adrenic acid precursors, respec-tively, in addition to PGE2, 11 PGE2, and 6-keto Prostaglandin F1a, 6,15-diketo-,dihydro-PGF1a (6,15-dk-dh-PGF1a) originating from arachidonic acid metabolism.

The remaining oxylipins with at least a 2-fold elevation after IL-1b exposure were from the lipooxygenase (LOX) metabolic path-ways and included 11-hydroxy-eicosatetraenoic acid (11-HETE),

12-HETE, and 13-hydroxy-docosahexaenoic acid (13-HDoHE). In addition, 4 oxylipins showed at least a 2-fold decrease in release after IL-1b stimulation, including 14,15 epoxy-eicosatrienoic acid (14,15 EpETrE), 20-hydroxy-docosahexaenoic acid (20-HDoHE), 16,17 epoxy-docosapentaenoicacid (16,17 EpDPE), and 18-hydroxy-eicosapentaenoic acid (18-HEPE). All the oxylipins showing a reduction after IL-1b were derived from the cytochrome P450 (CYP) metabolic pathway. Those oxylipins that displayed significant changes in response to the various treatment groups compared to IL-1b stimulation alone were 12-HETE, 18-HEPE, 20-HDoHE, and dihomo PGF2a (Figure 6). In the case of 12-HETE and dihomo PGF2a, the increases observed in response to IL-1b stimulation were significantly reduced with all treatments examined (Figure 6). The various treatments were able to restore 18-HEPE and 20-HDoHE, which were significantly reduced after IL-1b stimulation alone, to levels similar to unstimulated control groups (Figure 6).

Figure 4. Western blot analysis of COX-2 (A) and NF-kB (B) in stimulated cells after 48 h for COX-2 and 1 h for NF-kB under the various treatment conditions. Values were expressed as mean 6 SEM of 3 samples per treatment group.*P , 0.05, **P , 0.01, ***P , 0.001, compared to IL-1b treatment group (1-way ANOVA).

CO

X-2

/Act

in d

ensi

ty

NF-

kB/A

ctin

den

sity



D i s c u s s i o nWhile OA has long been considered a non-inflammatory joint

disease, increasing evidence supports the role of inflammatory mediators produced by the articular tissue in the early initiation and progression of the disease (7,18,19). OA is therefore best char-acterized as a condition involving chronic, low-grade inflammation within the afflicted joint. The inflammatory mediators associated with both pain and cartilage degeneration in OA include the inflam-matory cytokines, IL-1b, and TNFa, as well as prostaglandins and other oxylipins produced from the metabolism of arachidonic acid (20–22). Agents capable of modulating the multiple cytokine

and oxylipin pathways activated within the cells of the affected joint are of interest and may potentially influence the progression of OA.

In the current study, primary CnCs derived from normal canine articular cartilage were used as an in-vitro model of OA for examin-ing the anti-inflammatory effects of various nutraceutical products. Primary CnCs are considered a well established in-vitro model for studying the inflammatory mechanisms of OA and have been used to identify the key inflammatory targets and to evaluate the poten-tial effectiveness of therapeutic agents (23–25). Chondrocytes from patients with OA show elevated levels of IL-1b and TNFa release compared to chondrocytes isolated from non-OA patients, which helps to establish the role of these cytokines in disease (26). Primary

Figure 5. Release of TNFa from primary cultured canine chondrocytes (CnCs) under control conditions and after stimulation with IL-1b with and without CM (A), CHCL (B), CRN (C), or AlphaGEE (D). Values represent the mean 6 SEM of 4 monolayers per treatment group.*P , 0.05, **P , 0.01, ***P , 0.001, ****P , 0.0001, compared to IL-1b treatment group (2-way ANOVA).

TNFa

Con

cent

ratio

n (p

g/m

L)TN

Fa C

once

ntra

tion

(pg/

mL)

TNFa

Con

cent

ratio

n (p

g/m

L)TN

Fa C

once

ntra

tion

(pg/

mL)



chondrocyte cultures have also been used to screen for natural products that could be important in reducing inflammation in OA (24,25). The ability of cultured chondrocytes to activate inflamma-tory pathways after exposure to IL-1b, which is a key inflammatory mediator in the pathogenesis of OA, is important for the present study (24,25).

Complementary approaches for treating OA are of interest due in part to the absence of disease-modifying therapies for the disease. In the present study, we examined several dietary supplements for their potential anti-inflammatory and disease-modifying activity. The supplements selected were based on both manufacturer’s claims and published studies supporting their potential anti-inflammatory

effects. In the case of creatine compounds, previous studies have reported an anti-inflammatory effect in rodent models of inflamma-tion (27). More recent studies have shown that high-dose CM supple-mentation reduced both plasma TNFa and PGE2 in athletes (28,29). As both the rodent and human studies involved a TNFa component, we hypothesized that creatine compounds would also be effective anti-inflammatory agents in our in-vitro OA model. Previous studies have also reported anti-inflammatory effects of several amino acid and modified, i.e., branched chain, amino acid compositions (30–33). While the mechanisms proposed for the anti-inflammatory effects of various amino acids differ based on the inflammatory model examined, there is evidence to suggest that the cellular pathways

Figure 6. Levels of 12-HETE (A), 18-HEPE (B), dihomo PGF2a (C), and 20-HDoHE (D) in stimulated cells after 24 h under the various treatment condi-tions. Values were expressed as mean 6 SEM of 4 samples per treatment group. Asterisks above the bars indicate significance compared to IL-1b treatment group; square brackets indicate significance compared to control (untreated group).*P , 0.05, **P , 0.01, ***P , 0.001, ****P , 0.0001.ns — nonsignificant (1-way ANOVA).

12-H

ETE

Con

c. (n

g/m

L)di

hom

o P

GF2

a C

onc.

(ng/

mL)

18-H

EP

E C

onc.

(ng/

mL)

20-H

DoH

E C

onc.

(ng/

mL)



involved affect those examined in the present study, i.e., oxylipins and TNFa (30–32).

All the nutritional supplements examined in this study displayed some ability to reduce the release of PGE2 and TNFa into the cul-ture media after IL-1b stimulation. The effects were both time- and concentration-dependent, with significant decreases in the concentra-tions of PGE2 observed at 8 and 48 h after stimulation and significant reductions in TNFa release occurring at 24 and 48 h. The reason(s) for the biphasic effects observed with the creatine compounds is(are) not known. These compounds could affect PGE production through multiple pathways and thus different exposure times could result in different effects on PGE release. The studies examining the pharma-cological effects of creatine in a rodent carrageenan inflammation model suggested a lysosomal membrane stabilization effect and potential downstream effects on formation of prostaglandin (27). Such effects may be responsible for the early reductions in PGE2 release, i.e., at 8 h, observed in the present study, while the later effects may be driven by decreases in TNFa release.

The increased PGE2 found in chondrocytes from cartilage of human patients suffering from OA appears to be associated with increased expression of COX-2 (26). In addition, COX-2 expression is influenced by transcription factors activated through IL-1b and TNFa (25,27). In the present study, the expression of COX-2 was significantly increased after the cells were stimulated with IL-1b compared to control cells. Furthermore, with the exception of carprofen, all the tested compounds were able to decrease COX-2 expression at 48 h. A similar effect was observed when examining TNFa release in the media, with all the tested compounds, except carprofen, significantly reducing the release of TNFa.

The fact that carprofen increased the expression of COX-2 and the release of TNFa is consistent with the findings of other studies on the effects of COX-2 inhibitors. Page et al (34) showed that NSAIDs, specifically celecoxib and rofecoxib, which are COX-2-specific NSAIDs, and diclofenac, which is a COX-1-and COX-2-bispecific NSAID, significantly increased the spontaneous release of TNFa in arthritic human synovial membrane cultures. The same increase of TNFa with NSAIDs was observed in primary human monocytes after stimulating the cells with lipopolysaccharide (LPS) (34). These findings might explain the ability of NSAIDs to relieve joint pain and tenderness and their lack of disease-modifying anti-rheumatic activity, as increased TNFa release would be expected to exacer-bate the inflammation and the pathologic processes in arthritic patients.

As the inflammatory effects of TNFa are associated with activa-tion of NF-kB, the effects of the compounds on NF-kB activation were also examined. As expected, given the effects of the various nutraceutical compounds on TNFa release, there was also a reduc-tion in NF-kB activation observed compared to IL-1b stimulation alone. The ability of the nutraceutical compounds to influence multiple pathways in the inflammation process suggests that the anti-inflammatory properties may be a result of the beneficial effects of creatine on the mitochondria and its antioxidant proper-ties. A growing body of evidence suggests an interplay between inflammation and mitochondrial dysfunction in OA. For example, Kim et al (35) have shown that pro-inflammatory cytokines such as IL-1b and TNFa disturb normal mitochondrial function in primary

human osteoarthritic chondrocyte cultures by damaging mitochon-drial DNA (mtDNA), decreasing energy production, and increasing production of reactive oxygen species (ROS). It has also been shown that mitochondrial dysfunction induced by commonly used inhibi-tors, such as oligomycin and antimycin A, which inhibit mitochon-drial respiratory chain complexes V and III, respectively, increases inflammatory responsiveness to IL-1b and TNFa, i.e., increases IL-8, COX-2, and PGE2 production, in normal human chondrocytes (36). This suggests that any therapeutic intervention that has the ability to stabilize and improve mitochondrial function could also decrease the inflammation observed in OA.

Although the role of prostaglandins and specifically PGE2 in the pathophysiology of OA has been the subject of extensive research, the role of other oxylipins is not well-understood. Since inflam-mation is an important component in the pathophysiology of OA, changes in oxylipin release in response to inflammation in the CnCs might provide important insight into the development of OA, as well as provide potential biomarkers for treatment response. In the present study, the production of oxylipins in CnCs was examined in response to exposure to IL-1b. It was demonstrated that exposure of the cells to IL-1b, which is an early endogenous pro-inflammatory mediator, increased the levels of 10 oxylipins formed via COX and LOX pathways and decreased the levels of 4 oxylipins derived from the CYP pathway. The pro-inflammatory PGE2 was the COX-derived oxylipin present at the highest level upon incubation with IL-1b. Several other COX-derived oxylipins showed substantial elevations (up to 32-fold) after IL-1b exposure, although in smaller concentra-tions. While the role of these oxylipins in OA and inflammation remains to be determined, recent studies point to their potential value as biomarkers for symptomatic OA (37).

Two oxylipins that were increased after IL-1b stimulation were 11-HETE and 12-HETE. Both of these oxylipins are pro-inflamma-tory, derived via the LOX pathway from AA (16), and are strongly associated with inflammation (16). A role for these oxylipins in OA is suggested by Wong et al (37) who reported increased levels of 12-HETE in knee joints of a rat model of OA compared to healthy controls. It should be noted that all the creatine- and amino-acid based supplements examined in this study were able to reduce the release of 12-HETE in CnCs stimulated with IL-1b.

In addition to increasing the release of pro-inflammatory oxylipin, stimulating the cells with IL-1b decreased the levels of some anti-inflammatory oxylipins derived via the CYP pathway, including 18-HEPE (16). Although the importance of this particular oxylipin in OA is unknown, 18-HEPE has the ability to decrease LPS-induced TNFa secretion in the murine macrophage cell line (38). The ability of CM, CHCl, CRN, and AlphaGEE to restore 18-HEPE after IL-1b stimulation in order to control levels may therefore be involved in the effects of these agents on TNFa release reported here.

The finding that CRN also has anti-inflammatory properties was unexpected since this compound has been considered to be the inactive metabolite of creatine, which is excreted in the urine (39). Previous studies using the rodent carrageen inflammation model have reported anti-inflammatory actions for both creatine and CRN (27,30). The finding of anti-inflammatory effects of CRN in the CnC model in the present study is also consistent with the more recent findings of Leland et al (40). This study reported that short-term



exposure (10 to 60 min) of a mouse macrophage cell line (RAW 264.7) to 100 mM of CRN significantly down-regulated levels of toll-like receptor 2 (TLR2), TLR3, TLR4, and TLR7 messenger ribonucleic acid (mRNA) in the cells, which suggests that CRN may have immuno-suppressive properties (40).

In summary, the present study demonstrates that various creatine- and amino acid-based compounds can produce a significant anti-inflammatory effect. The exact mechanism(s) by which the creatine and amino acid compounds produce their anti-inflammatory effects are not known. Unlike the COX-2 inhibitor, the dietary supplements examined here appeared to affect multiple pathways in the process of inflammation in the CnC model. The creatine- and amino acid-based compounds examined interfered with NF-kB and TNFa release, lead-ing to reductions in COX-2 expression and the release of PGE2, as well as alterations in both pro-inflammatory and anti-inflammatory oxylipins in response to IL-1b activation.

An important consideration for translating these initial findings in a cell culture model of OA to in-vivo animal models and human trials is the concentrations required to obtain the effects observed. While the concentrations required for anti-inflammatory response to GS in the present study were much higher than the peak plasma levels reported with standard dietary supplementation with GS (41), the concentrations required for creatine- and amino acid-based compounds were certainly within the range of blood levels observed after dietary supplementation (42). This study suggests that supple-mentation with creatine- and amino acid- based supplements may provide beneficial effects in OA and warrants further research into their clinical effectiveness and mechanism of response.

A c k n o w l e d g m e n t sFunds for this study were provided by Vireo Systems (Madison,

Tennessee, USA). Dr. Donald Miller is an inventor of intellectual property involving CHCL and AlphaGEE, which has been licensed to Vireo Systems. Eman AlRaddadi is a recipient of Health Science Scholarships for post-graduate studies from the Ministry of Higher Education of the Kingdom of Saudi Arabia and King Saud bin Abdulaziz University.

Re f e r e n c e s 1. Dahaghin S, Bierma-Zeinstra SM, Ginai AZ, Pols HA, Hazes JM,

Koes BW. Prevalence and pattern of radiographic hand osteoar-thritis and association with pain and disability (the Rotterdam study). Ann Rheum Dis 2005;64:682–687.

2. Rychel JK. Diagnosis and treatment of osteoarthritis. Top Com-panion Anim Med 2010;25:20–25.

3. Lane NE, Schinitzer TJ. Osteoarthritis. In: Goldman L, Schafer IA, eds. Goldman’s Cecil Medicine. 24th ed. Philadelphia, Pennsylvania: Saunders Elsevier, 2011:1672–1676.

4. Xia B, Di C, Zhang J, Hu S, Jin H, Tong P. Osteoarthritis patho-genesis: A review of molecular mechanisms. Calcif Tissue Int 2014;95:495–505.

5. Sophia Fox AJ, Bedi A, Rodeo SA. The basic science of articular cartilage: Structure, composition, and function. Sports Health 2009;1:461–468.

6. Hedbom E, Hauselmann HJ. Molecular aspects of pathogenesis in osteoarthritis: The role of inflammation. Cell Mol Life Sci 2002;59:45–53.

7. Berenbaum F. Osteoarthritis as an inflammatory disease (osteo-arthritis is not osteoarthrosis!). Osteoarthritis Cartilage 2013;21: 16–21.

8. Filardo G, Kon E, Longo UG, et al. Non-surgical treatments for the management of early osteoarthritis. Knee Surg Sports Traumatol Arthrosc 2016;24:1775–1785.

9. Beale BS. Use of nutraceuticals and chondroprotectants in osteoarthritic dogs and cats. Vet Clin North Am Small Anim Pract 2004;34:271–289.

10. Bauer JE. Evaluation of nutraceuticals, dietary supplements, and functional food ingredients for companion animals. J Am Vet Med Assoc 2001;218:1755–1760.

11. Neves M Jr, Gualano B, Roschel H, et al. Beneficial effect of creatine supplementation in knee osteoarthritis. Med Sci Sports Exerc 2011;43:1538–1543.

12. Roy BD, de Beer J, Harvey D, Tarnopolsky MA. Creatine monohydrate supplementation does not improve functional recovery after total knee arthroplasty. Arch Phys Med Rehabil 2005;86:1293–1298.

13. Petersen SG, Beyer N, Hansen M, et al. Nonsteroidal anti- inflammatory drug or glucosamine reduced pain and improved muscle strength with resistance training in a randomized con-trolled trial of knee osteoarthritis patients. Arch Phys Med Rehabil 2011;92:1185–1193.

14. Gufford BT, Sriraghavan K, Miller NJ, et al. Physicochemical characterization of creatine N-methylguanidinium salts. J Diet Suppl 2010;7:240–252.

15. Dash AK, Miller DW, Huai-Yan H, Carnazzo J, Stout JR. Evalua-tion of creatine transport using Caco-2 monolayers as an in vitro model for intestinal absorption. J Pharm Sci 2001;90:1593–1598.

16. Gabbs M, Leng S, Devassy JG, Monirujjaman Md, Aukema HM. Advances in our understanding of oxylipins derived from dietary PUFAs. Adv Nutr 2015;6:513–540.

17. Aukema HM, Winter T, Ravandi A, Dalvi S, Miller DW, Hatch GM. Generation of bioactive oxylipins from exogenously added arachidonic, eicosapentaenoic and docosahexaenoic acid in primary human brain microvessel endothelial cells. Lipids 2016;51:591–599.

18. Abramson SB, Attur M, Yazici Y. Prospects for disease modifica-tion in osteoarthritis. Nat Clin Pract Rheumatol 2006;2:304–312.

19. Attur M, Belitskaya-Levy I, Oh C, et al. Increased interleukin-1b gene expression in peripheral blood leukocytes is associated with increased pain and predicts risk for progression of symptomatic knee osteoarthritis. Arthritis Rheum 2011;63:1908–1917.

20. Stannus O, Jones G, Cicuttini F, et al. Circulating levels of IL-6 and TNF-a are associated with knee radiographic osteoarthritis and knee cartilage loss in older adults. Osteoarthritis Cartilage 2010;18:1441–1447.

21. Sohn DH, Sokolove J, Sharpe O, et al. Plasma proteins present in osteoarthritic synovial fluid can stimulate cytokine production via toll-like receptor 4. Arthritis Res Ther 2012;14:R7.

22. Attur M, Krasnokutsky S, Statnikov A, et al. Low-grade inflam-mation in symptomatic knee osteoarthritis: Prognostic value



of inflammatory plasma lipids and peripheral blood leukocyte biomarkers. Arthritis Rheumatol 2015;67:2905–2915.

23. Sunaga T, Oh N, Hosoya K, Takagi S, Okumura M. Inhibitory effects of pentosan polysulfate sodium on MAP-kinase pathway and NF-B nuclear translocation in canine chondrocytes in vitro. J Vet Med Sci 2012;74:707–711.

24. Shakibaei M, Allaway D, Nebrich S, Mobasheri A. Botanical extracts from rosehip (Rosa canina), willow bark (Salix alba), and nettle leaf (Urtica dioica) suppress IL-1b-induced NF-kB activa-tion in canine articular chondrocytes. Evid Based Complement Alternat Med 2012; Epub 2012.

25. Martinez SE, Chen Y, Ho EA, Martinez SA, Davies NM. Pharmacological effects of a C-phycocyanin-based multicom-ponent nutraceutical in an in-vitro canine chondrocyte model of osteoarthritis. Can J Vet Res 2015;79:241–249.

26. Kapoor M, Martel-Pelletier J, Lajeunesse D, Pelletier JP, Fahmi H. Role of proinflammatory cytokines in the pathophysiology of osteoarthritis. Nat Rev Rheumatol 2011;7:33–42.

27. Khanna NK, Madan BR. Studies on the anti-inflammatory activ-ity of creatine. Arch Int Pharmacodyn Ther 1978;231:340–350.

28. Bassit RA, Curi R, Costa Rosa LF. Creatine supplementation reduces plasma levels of pro-inflammatory cytokines and PGE2 after a half-ironman competition. Amino Acids 2008;35:425–431.

29. Santos RV, Bassit RA, Caperuto EC, Costa Rosa LF. The effect of creatine supplementation upon inflammatory and muscle sore-ness markers after a 30 km race. Life Sci 2004;75:1917–1924.

30. Saxena RN, Pendse VK, Khanna NK. Anti-inflammatory and analgesic properties of four amino-acids. Indian J Physiol Phar-macol 1984;28:299–305.

31. Mine Y, Zhang H. Calcium-sensing receptor (CaSR)-mediated anti-inflammatory effects of L-amino acids in intestinal epithelial cells. J Agric Food Chem 2015;63:9987–9995.

32. Lee JH, Park E, Jin HJ, et al. Anti-inflammatory and anti-genotoxic activity of branched chain amino acids (BCAA) in lipopolysaccharide (LPS) stimulated RAW 264.7 macrophages. Food Sci Biotechnol 2017;26:1371–1377.

33. Cruzat VF, Bittencourt A, Scomazzon SP, Leite JS, de Bittencourt PI Jr, Tirapegui J. Oral free and dipeptide forms of glutamine supplementation attenuate oxidative stress and inflammation induced by endotoxemia. Nutrition 2014;30:602–611.

34. Page TH, Turner JJ, Brown AC, et al. Nonsteroidal anti- inflammatory drugs increase TNF production in rheumatoid synovial membrane cultures and whole blood. J Immunol 2010; 185:3694–3701.

35. Kim J, Xu M, Xo R, et al. Mitochondrial DNA damage is involved in apoptosis caused by pro-inflammatory cytokines in human OA chondrocytes. Osteoarthritis Cartilage 2010;18:424–432.

36. Vaamonde-Garcia C, Riveiro-Naveira RR, Valcarcel-Ares MN, Hermida-Carballo L, Blanco FJ, Lopez-Armada MJ. Mitochon-drial dysfunction increases inflammatory responsiveness to cytokines in normal human chondrocytes. Arthritis Rheum 2012; 64:2927–2936.

37. Wong A, Sagar DR, Ortori CA, Kendall DA, Chapman V, Barrett DA. Simultaneous tissue profiling of eicosanoid and endocan-nabinoid lipid families in a rat model of osteoarthritis. J Lipid Res 2014;55:1902–1913.

38. Weylandt KH, Krause LF, Gomolka B, et al. Suppressed liver tumorigenesis in fat-1 mice with elevated omega-3 fatty acids is associated with increased omega-3 derived lipid mediators and reduced TNF-alpha. Carcinogenesis 2011;32:897–903.

39. Wyss M, Kaddurah-Daouk R. Creatine and creatinine metabo-lism. Physiol Rev 2000;80:1107–1213.

40. Leland KM, McDonald TL, Drescher KM. Effect of creatine, creatinine, and creatine ethyl ester on TLR expression in mac-rophages. Int Immunopharmacol 2011;11:1341–1347.

41. Persiani S, Roda E, Rovati LC, Locatelli M, Giacovelli G, Roda A. Glucosamine oral bioavailability and plasma pharmacokinetics after increasing doses of crystalline glucosamine sulfate in man. Osteoarthritis Cartilage 2005;13:1041–1049.

42. Jager R, Harris RC, Purpura M, Francaux M. Comparison of new forms of creatine in raising plasma creatine levels. J Int Soc Sports Nutr 2007;4:17.


Article


I n t r o d u c t i o nReproductive performance has a huge impact on the productivity

and profitability of a dairy herd. Fertility, in turn, is related to the parity of cows (1).

Primiparous cows often experience lower energy balance than their multiparous counterparts because, in addition to the energy and nutrient demand of lactation, they usually eat less and require energy for growth, which compromises their reproductive perfor-mance (2,3). As multiparous cows are more adaptable to reinitiating

postpartum cyclicity, the interval from calving to first ovulation is longer in primiparous than in multiparous cows (4–6). While dystocia is more common in primiparous cows, multiparous cows are more likely to suffer from retained placenta and ovarian cysts, which extend the time to conception (1,7). Moreover, the likelihood of conception was 8% lower in cows in parity 3 and above compared to primiparous cows (1).

Certain physiological phenomena may directly influence the effectiveness of reproductive management practices. Pregnancy-associated glycoprotein (PAG) levels were higher in both milk

Relationship between reproductive management practices and fertility in primiparous and multiparous dairy cows

István Fodor, György Gábor, Zsolt Lang, Zsolt Abonyi-Tóth, László Ózsvári

A b s t r a c tThe aim of this study was to analyze the impact of reproductive management practices on fertility in primiparous and multiparous dairy cows and to identify the differences by parity. Data on reproductive performance and management of cows were gathered from 34 large commercial Holstein-Friesian dairy herds in Hungary. Individual data on 23 781 cows that calved in the studied herds in 2014 were collected from the national milk recording database and used to calculate the reproductive indices. Farm managers and veterinarians were interviewed in person from May 22 to November 6, 2015 using a questionnaire to collect information about reproductive management practices. The data were statistically analyzed by mixed-effects models. Our results showed that the use of a voluntary waiting period (VWP) was linked to a greater increase in calving-to-conception interval (CCI, P , 0.05) and a greater decline in the chance of pregnancy at 200 days in milk (P200, P , 0.001) in multiparous cows. Estrus synchronization was related to a larger reduction in days-to-first service (DFS, P , 0.001), breeding interval (IBI, P , 0.05), and CCI (P , 0.01) and a greater improvement in P200 (P , 0.001) in primiparous cows. Early pregnancy diagnosis and pregnancy recheck improved IBI (P , 0.01 for both practices), CCI (P , 0.01 and P , 0.001, respectively), and P200 (P , 0.001 for both practices) to a larger extent in primiparous than in multiparous cows. In conclusion, estrus synchronization, early pregnancy diagnosis, and pregnancy recheck led to a greater improvement in fertility in primiparous cows than in multiparous cows.

R é s u m éL’objectif de la présente étude était d’analyser l’impact des pratiques de gestion de la reproduction sur la fertilité de vaches laitières primipares et multipares et d’identifier les différences par parité. Les données sur les performances de reproduction et la gestion des vaches ont été amassées auprès de 34 fermes laitières Holstein-Friesian commerciales importantes en Hongrie. Les données individuelles de 23 781 vaches ayant mise-bas en 2014 dans les troupeaux à l’étude ont été obtenues de la base de données nationale d’enregistrement de la production laitière et furent utilisées pour calculer les indices de reproduction. Les gérants de ferme et les vétérinaires ont été interviewés en personne entre le 22 mai et le 6 novembre 2015 à l’aide d’un questionnaire pour obtenir des informations sur les pratiques de gestion de la reproduction. Les données ont été analysées statistiquement par des modèles à effets mixtes. Nos résultats ont démontré que l’utilisation d’une période volontaire d’attente (VWP) était liée à une plus grande augmentation de l’intervalle vêlage-conception (CCI, P , 0,05) et à un plus grand déclin de la probabilité de gestation à 200 jours en lait (P200, P , 0,001) chez les vaches multipares. La synchronisation des œstrus était reliée à une plus grande réduction du nombre de jours avant la première saillie (DFS, P , 0,001), de l’intervalle de reproduction (P , 0,05), et de CCI (P , 0,01), et une meilleure amélioration du P200 (P , 0,001) chez les vaches primipares. Des diagnostics de gestation hâtifs et des vérifications de la gestation ont amélioré IBI (P , 0,01 pour ces deux pratiques), CCI (P , 0,01 et P , 0,001, respectivement), et P200 (P , 0,001 pour les deux pratiques) d’une manière plus importante chez les primipares que chez les multipares. En conclusion, la synchronisation des œstrus, les diagnostics de gestation hâtif, et la vérification de la gestation ont conduit à une plus grande amélioration de la fertilité chez les vaches primipares que chez les vaches pluripares.


Department of Veterinary Forensics, Law and Economics (Fodor, Ózsvári) and Department of Biomathematics and Informatics (Lang, Abonyi-Tóth), University of Veterinary Medicine Budapest, István utca 2, H-1078 Budapest, Hungary; Research Institute for Animal Breeding, Nutrition and Meat Science, National Agricultural Research and Innovation Centre, Gesztenyés utca 1, H-2053 Herceghalom, Hungary (Gábor).

Address all correspondence to Dr. István Fodor; telephone: 136-1-478-4185; fax: 136-1-478-4186; e-mail: [email protected]

Received July 9, 2018. Accepted October 26, 2018.



and plasma samples from primiparous cows than in samples from multiparous cows. This can potentially affect the accuracy of a PAG test used for early pregnancy diagnosis (8,9). Response to hormonal synchronization differs by parity, as demonstrated in a meta-analysis comparing Double-Ovsynch and Presynch-Ovsynch. Borchardt et al (10) concluded that Double-Ovsynch was beneficial in terms of preg-nancy per artificial insemination compared to Presynch-Ovsynch in primiparous, but not in multiparous cows. Astiz and Fargas (11) suggested that different hormonal synchronization protocols should be used in high-yielding dairy cows from different parities.

Parity-wise differences in the efficiency of reproduction manage-ment are increasingly being studied in dairy herds. Based on previ-ous studies, we hypothesized that parity may affect the relationship between reproductive management and fertility parameters. The aim of this study was to analyze the relationship between reproductive management practices and fertility in primiparous and multiparous dairy cows.


Data collectionThe survey of reproductive management practices was carried

out from May 22 to November 6, 2015 in a total of 34 large-scale Hungarian dairy herds. The inclusion criteria for the farms were as follows: computerized on-farm data recording; continuous partici-pation in milk recording at least since January 1, 2014; herd size of more than 250 cows; and a willingness to provide data to the authors. Farm managers and veterinarians were interviewed during farm visits and a questionnaire was used to collect information about reproductive management practices. The farms were located in all regions of Hungary, where a continental climate consists of warm, dry summers and cold winters. Mean 6 standard deviation (SD) size of the studied herds was 755 6 470 cows (range: 291 to 2502 cows) and the 305-day milk yield of the herds amounted to 10 014 6 965 kg (range: 8330 to 12 541 kg).

Reproductive data for individual cows were collected from the national milk recording database (Livestock Performance Testing Ltd., Gödöllo’’, Hungary). The data of 23 781 cows that calved from January 1 to December 31, 2014 on the farms were analyzed retrospectively. For each cow, data were gathered regarding herd, animal ID, date of birth, date and number of inseminations, results of pregnancy diagnoses, calving dates, parity, 305-day milk yields, and culling date. Data were managed in MS Excel 2013 (Microsoft Corporation, Redmond, Washington, USA).

Management practices and reproductive parameters analyzed

The definitions of the management practices studied, the respec-tive number of farms, and the number of primiparous and multipa-rous cows are shown in Table I. As part of the questionnaire, farms were asked whether they use a voluntary waiting period (VWP), i.e., the period after calving when they do not inseminate cows, and the length of the VWP. In those herds that applied a VWP, its average length was 50.2 6 11.3 d. Two categories were created based on the questionnaire: farms that applied a $ 50-day-long VWP

(“$ 50 d”) and those that applied a VWP of , 50 d or did not use VWP (“, 50 d”).

Estrus detection was carried out in every herd, using either estrus detection aids, such as pedometers, activity meters, and tail chalk, or visual observation alone. A combination of hormonal synchroni-zation and estrus detection was applied in 79.4% of the herds. The most widely adopted synchronization protocol was Ovsynch, which was used by 74.1% of the herds that applied synchronization. This involves giving injections of a gonadotropin-releasing hormone (GnRH), followed by prostaglandin F2alpha (PGF2a) 7 d later, a second injection of GnRH 48 to 56 h later, then insemination after 16 to 24 h. Cosynch, which was used by 18.5% of the herds, involves insemination at the time of the second injection of GnRH during the Ovsynch protocol. And finally, 14.8% of the herds used Presynch-Ovsynch, which involves 2 PGF2a injections 14 d apart, followed by an Ovsynch protocol 10 to 14 d later.

Analysis of pregnancy diagnosis policy included the application of early pregnancy diagnosis methods, such as transrectal ultraso-nography or PAG tests, and the frequency of pregnancy diagnosis (Table I). The first pregnancy diagnosis was conducted 35.1 6 8.5 d after insemination, on average. Both transrectal ultrasonography and PAG tests were used from days 28 to 30 onwards. Use of transrectal ultrasonography was analyzed separately because this technique can be used not only to diagnose pregnancy, but also to examine the ova-ries and diagnose pathological conditions of the reproductive tract. The information gathered during the ultrasonographic examinations was used in the treatment of reproductive disorders in these herds. It was also evaluated whether early pregnancy diagnosis before day 60 after artificial insemination (AI), followed by a pregnancy recheck around day 60, was associated with better reproductive performance. Pregnancy status was rechecked 1.6 6 0.8 times, on average, typically around day 60 after insemination and around dry-off. Cows in all herds were fed total mixed ration (TMR).

The relationship between parity and management practices and the following reproductive indices was analyzed: days to first service (DFS), breeding interval (IBI), calving-to-conception interval (CCI), first-service conception risk (CR1), and pregnancy at 200 days in milk (P200). Days to first service (n = 19 794) and IBI (n = 14 811) were calculated for the inseminated cows, whereas CCI (n = 15 707) was calculated for those cows that were diagnosed pregnant after the date of last insemination. First-service conception risk was quanti-fied for cows with at least 1 insemination followed by a known pregnancy diagnosis outcome (n = 19 794). Cows that stayed in the herd at least until 200 days in milk (DIM) were included in the analysis of P200 (n = 19 134).

Statistical analysesThe associations of parity with the continuous variables, i.e.,

DFS, IBI, CCI, under different management practices were exam-ined with linear mixed-effects models using the lme4 package in R (12). The continuous dependent variables were log-transformed to eliminate skewness of the data. The associations of parity with the dichotomous dependent variables, i.e., CR1, P200, under different reproductive management practices were analyzed with logistic regression, including mixed-effects using the lme4 package in R (12). In each mixed model, one of the reproductive indices was the



Table I. Reproductive management practices studied and the corresponding number of herds and cows by parity (the total number of farms is 34, unless otherwise specified).

Farms Primiparous Multiparous Definition (n) cows (n) cows (n)Voluntary waiting period (VWP) $ 50 d The length of VWP reported by the farm was $ 50 d 19 5107 7758

, 50 d The length of VWP reported by the farm was , 50 d or the farm did not apply a VWP 15 4239 6677

Estrus detection aids Yes Estrus detection aids, e.g., pedometers, activity meters, tail chalk, were applied 25 6585 10 386

No Visual observation was used to detect cows in estrus 9 2761 4049

Estrus synchronization Yes Hormonal synchronization protocols, e.g., Ovsynch, Cosynch, or Presynch-Ovsynch, 27 7589 11 901 were applied

No Hormonal synchronization was not applied 7 1757 2534

Early pregnancy diagnosis Yes Pregnancy was diagnosed by transrectal ultrasonography or pregnancy-associated 24 6982 10 897 glycoprotein (PAG) tests

No Pregnancy was diagnosed solely by rectal palpation 10 2364 3538

Reproductive ultrasonography Yes Transrectal ultrasound was used for pregnancy diagnosis and other reproductive 23 6545 10 214 examinations

No Transrectal ultrasound was not used in the herd 11 2801 4221

Pregnancy recheck Yes Pregnancy diagnosis before day 60 post-insemination was followed by a pregnancy 18 5113 8511 recheck around day 60

No Pregnancy status was not rechecked within 60 d post-insemination 16 4233 5924

Frequency of pregnancy diagnosis (n = 33) . Once a week Pregnancy diagnosis was done more than once a week 10 3101 4563

Weekly Pregnancy diagnosis was done once a week 16 4501 7135

, Once a week Pregnancy was diagnosed less frequently than weekly 7 1521 2346



outcome variable. Parity, a management practice, and the interac-tion between parity and the management practice were included as fixed explanatory variables. Herd was a random effect. Multiple comparisons within a given management practice, e.g., pregnancy diagnosis frequency, were carried out by Tukey’s post-hoc test, using the multcomp package in R (13). The associations between reproduc-tive indices and management by parity were quantified by contrasts using the multcomp package in R (13). The statistical analyses were carried out in R version 3.4.0. (14).

Re s u l t s

Reproductive performance and parityIn the overall population, mean (6 SD) and median days to first

service (DFS) were 75.6 d (6 35.8) and 68.0 d, respectively. Mean (6 SD) and median breeding interval (IBI) were 42.8 d (6 24.9) and 37.3 d, respectively, whereas the mode of the IBI was 23 d. Mean (6 SD) and median calving-to-conception interval (CCI) were 157.5 d (6 93.1) and 134.0 d. In the entire studied population, CR1 was 20.5%, whereas 59.6% of the cows were pregnant at 200 days in milk (DIM) in the herds.

In general, primiparous cows had a shorter breeding interval (42.2 versus 43.2 d, P , 0.001, Tables II and III), needed less time to con-ceive postpartum (152.3 versus 161.8 d, P , 0.001), conceived better for first service (24.8% versus 17.3%, P , 0.001, Table IV), and were more likely to be in calf at 200 DIM (65.2% versus 55.4%, P , 0.001) compared to multiparous cows. No differences between parities were found for time from calving to first insemination (75.7 versus 75.6 d, P . 0.05).

Voluntary waiting period, estrus detection, and synchronization

In primiparous cows, the interval from calving to conception was 8.24% longer when VWP was $ 50 d compared with those animals that were subject to a shorter or no VWP (Table V). Similarly, in mul-tiparous cows, the CCI was 14.56% longer when VWP was $ 50 d compared with those cows that were subject to a shorter or no VWP. In primiparous cows, the odds of pregnancy at 200 DIM was smaller when VWP was $ 50 d than , 50 d (OR = 0.834, Table VI). In multiparous cows, the difference between the VWP of $ 50 d and , 50 d was greater (OR = 0.603). The benefit of short or no VWP was, therefore, greater in multiparous cows than in primiparous cows in terms of both time to conception and probability of pregnancy at 200 DIM (P = 0.015 and P , 0.001).

The use of estrus detection aids was associated with a shorter time to first insemination in primiparous cows than in multiparous cows (P = 0.015, Table V). When estrus synchronization was used, the reduction in DFS (P , 0.001, Table V), IBI (P = 0.026), and CCI (P = 0.005) and the improvement in probability of pregnancy at 200 days in milk (P200) (P , 0.001, Table VI) were greater in pri-miparous cows than in multiparous cows.

Pregnancy diagnosis policyWhen early pregnancy diagnosis was performed, the reduction

in breeding interval (P = 0.001, Table V), CCI (P = 0.006), and the

improvement in the probability of pregnancy at 200 days in milk (P , 0.001, Table VI) were greater in primiparous cows than in multiparous cows. The use of transrectal ultrasonography was associated with larger reduction in IBI and a greater improvement in P200 in primiparous cows than in multiparous cows (P , 0.001). When pregnancy status was rechecked in a herd, IBI (P = 0.002), CCI (P , 0.001), and P200 (P , 0.001) were slightly better in primiparous cows than in multiparous cows.

The magnitude of herd effect ranged from 19% to 21%, 13% to 16%, and 13% to 14% in the models of DFS, IBI, and CCI, respec-tively. In the models of CR1 and P200, the magnitude of herd effect ranged from 30% to 43% and 43% to 51%, respectively.

D i s c u s s i o nMultiparous cows experienced larger improvement in time to con-

ception postpartum when the voluntary waiting period (VWP) was short or not applied compared with primiparous cows. This suggests that not having a minimum waiting period of at least 50 d is better for multiparous cows. It should be noted, however, that days to first service (DFS) was remarkably higher than the VWP reported by the farm. Stangaferro et al (15) found that the median days to concep-tion was 6 d shorter (85 versus 91 d) in the 60-day-long VWP group compared to the 88-day-long VWP group in primiparous cows and 28 d shorter (104 versus 132 d) in multiparous cows. Inchaisri et al (16) found that the economically optimal duration of the VWP was , 8 wk in 63% of the cows and that primiparous cows experienced lower losses when VWP was extended with a longer period com-pared with multiparous cows.

In our study, primiparous cows experienced a greater advantage than multiparous cows in terms of time to first insemination when estrus detection aids were used. The secondary signs of estrus, such as walking activity, decrease, while the primary sign of estrus, i.e., standing still, increases with higher parity (17,18). Several studies indicate that the performance of activity meters may be compro-mised by higher parity (3,19). Each additional lactation number decreased walking activity at estrus by 21.4% in a study conducted in northeastern Spain (20). In Israel, the average herd rate of unde-tected estrus in 2008 was higher in multiparous than in primiparous cows (38.9% versus 43.9%) (21). Based on activity data, inseminating multiparous cows once a day may yield acceptable results, however, primiparous cows should be scheduled for insemination at least twice a day to improve conception rates (22). It is possible that the lack of associations among the method of estrus detection, reproduc-tive performance, and parity in our study could be attributed to the widespread use of hormonal synchronization programs.

The use of hormonal synchronization protocols was related to a greater advantage in terms of time to first insemination, breed-ing interval, time to conception, and probability of pregnancy at 200 days in milk (P200) in primiparous than in multiparous cows. This suggests that parity is associated with the response of the cow to the synchronization protocol. In a study evaluating differences between primiparous and multiparous cows after Ovsynch based on 6 trials, it was found that Ovsynch was more effective in primiparous animals (CR1: 37.9% versus 31.6%; P200: 81.8% versus 75.4%) (23). Machado et al (24) compared insemination only at completion of the



Tabl

e II

. D

escr

ipti

ve s

tati

stic

s of

day

s to

fir

st s

ervi

ce (

DFS

), b

reed

ing

inte

rval

(IB

I),

and

calv

ing-

to-c

once

ptio

n in

terv

al (

CC

I) b

y vo

lunt

ary

wai

ting

per

iod

(VW

P),

estr

us d

etec

tion

, and

hor

mon

al s

ynch

roni

zati

on a

nd b

y pa

rity

.

D

ays

to fi

rst s

ervi

ce

Bree

ding

inte

rval

(d)

Ca

lvin

g-to

-con

cept

ion

inte

rval

(d)

Pr

imip

arou

s M

ultip

arou

s Pr

imip

arou

s M

ultip

arou

s Pr

imip

arou

s M

ultip

arou

s

Mea

n 6

SD

Mea

n 6

SD

Mea

n 6

SD

Mea

n 6

SD

Mea

n 6

SD

Mea

n 6

SD

(Med

ian)

n

(Med

ian)

n

(Med

ian)

n

(Med

ian)

n

(Med

ian)

n

(Med

ian)

n

Volu

ntar

y w

aitin

g

perio

d (V

WP)

$

50

d 78

.6 6

37.

1 46

42

79.5

6 3

7.5

6138

44

.3 6

27.

5 33

55

46.3

6 2

5.8

4749

15

7.6

6 9

6.7

3979

17

2.3

6 9

6.7

4464

(7

1)

(7

1)

(3

8)

(4

0)

(1

30)

(1

50)

,

50

d 72

.1 6

35.

3 37

42

71.0

6 3

1.8

5272

39

.5 6

24.

3 26

87

39.6

6 2

0.9

4020

14

5.6

6 9

1.0

3188

15

0.4

6 8

4.5

4076

(6

3)

(6

4)

(3

5)

(3

5)

(1

19)

(1

30)

Estru

s de

tect

ion

aids

Ye

s 74

.8 6

35.

1 59

31

75.4

6 3

4.5

8182

41

.7 6

25.

9 43

08

42.9

6 2

3.4

6319

15

2.2

6 9

5.1

5094

16

3.1

6 9

2.4

6137

(6

8)

(6

9)

(3

6)

(3

8)

(1

25)

(1

41)

N

o 77

.9 6

39.

6 24

53

76.1

6 3

7.1

3228

43

.3 6

27.

1 17

34

44.1

6 2

5.3

2450

15

2.4

6 9

2.6

2073

15

8.5

6 9

0.0

2403

(6

8)

(6

7)

(3

7)

(3

8)

(1

26)

(1

38)

Estru

s sy

nchr

oniza

tion

Ye

s 72

.6 6

32.

8 68

03

73.5

6 3

3.0

9394

41

.4 6

24.

3 49

62

42.9

6 2

3.3

7289

15

0.0

6 9

2.9

5909

16

0.8

6 9

0.8

7006

(6

7)

(6

7)

(3

6)

(3

8)

(1

23)

(1

39)

N

o 89

.2 6

47.

0 15

81

85.4

6 4

2.8

2016

45

.9 6

33.

6 10

80

44.9

6 2

6.8

1480

16

3.2

6 1

00.0

12

58

166.

6 6

95.

8 15

34

(76)

(74)

(37)

(39)

(136

)

(144

)



Tabl

e II

I. D

escr

ipti

ve s

tati

stic

s of

day

s to

fir

st s

ervi

ce (

DFS

), b

reed

ing

inte

rval

(IB

I),

and

calv

ing-

to-c

once

ptio

n in

terv

al (

CC

I) b

y pr

egna

ncy

diag

nosi

s po

licy

and

pari

ty.

D

ays

to fi

rst s

ervi

ce

Bree

ding

inte

rval

(d)

Ca

lvin

g-to

-con

cept

ion

inte

rval

(d)

Pr

imip

arou

s M

ultip

arou

s Pr

imip

arou

s M

ultip

arou

s Pr

imip

arou

s M

ultip

arou

s

Mea

n 6

SD

Mea

n 6

SD

Mea

n 6

SD

Mea

n 6

SD

Mea

n 6

SD

Mea

n 6

SD

(Med

ian)

n

(Med

ian)

n

(Med

ian)

n

(Med

ian)

n

(Med

ian)

n

(Med

ian)

n

Early

pre

gnan

cy

diag

nosi

s

Yes

71.3

6 3

1.4

6264

72

.1 6

31.

2 86

06

40.4

6 2

3.1

4567

42

.2 6

22.

5 66

47

148.

8 6

93.

0 54

79

159.

6 6

91.

0 64

62

(66)

(67)

(36)

(37)

(121

)

(137

)

N

o 88

.6 6

46.

1 21

20

86.2

6 4

3.8

2804

47

.7 6

33.

6 14

75

46.6

6 2

7.5

2122

16

3.6

6 9

7.8

1688

16

8.9

6 9

3.7

2078

(7

7)

(7

5)

(4

0)

(4

0)

(1

39)

(1

49)

Repr

oduc

tive

ul

traso

und

Ye

s 71

.6 6

31.

9 58

86

71.8

6 3

1.4

8089

40

.4 6

23.

1 43

39

42.4

6 2

2.6

6296

15

0.7

6 9

3.7

5163

16

0.9

6 9

1.5

6091

(6

7)

(6

7)

(3

6)

(3

8)

(1

23)

(1

39)

N

o 85

.4 6

44.

0 24

98

84.9

6 4

1.8

3321

46

.6 6

32.

4 17

03

45.4

6 2

6.9

2473

15

6.3

6 9

6.2

2004

16

4.2

6 9

2.4

2449

(7

3)

(7

4)

(3

9)

(3

9)

(1

30)

(1

43)

Preg

nanc

y re

chec

k

Yes

73.5

6 3

3.0

4651

74

.8 6

33.

6 68

21

41.6

6 2

4.7

3338

43

.6 6

23.

9 52

74

150.

2 6

95.

3 40

39

162.

5 6

92.

7 50

63

(67)

(68)

(37)

(120

)

(140

)

N

o 78

.5 6

40.

2 37

33

76.8

6 3

7.4

4589

42

.9 6

28.

0 27

04

42.7

6 2

4.0

3495

15

5.0

6 9

3.2

3128

16

0.9

6 9

0.3

3477

(7

0)

(6

9)

(3

6)

(3

7)

(1

30)

(1

40)

Freq

uenc

y of

pr

egna

ncy

diag

nosi

s

. O

nce

a w

eek

80.7

6 4

2.9

2773

77

.2 6

40.

8 35

28

43.2

6 2

8.9

2007

43

.5 6

26.

3 27

45

157.

9 6

94.

4 22

87

160.

2 6

91.

1 25

32

(71)

(67)

(37)

(37)

(133

)

(137

)

W

eekl

y 76

.0 6

32.

8 40

25

78.3

6 3

2.8

5620

42

.4 6

25.

8 28

05

43.8

6 2

3.5

4144

14

7.7

6 9

2.4

3582

15

9.7

6 9

0.4

4344

(7

0)

(7

1)

(3

7)

(1

20)

(1

38)

,

Onc

e a

wee

k 63

.7 6

27.

5 13

84

64.2

6 2

7.1

1958

40

.1 6

22.

6 11

05

41.3

6 2

0.2

1639

15

6.7

6 9

9.0

1120

17

0.5

6 9

6.0

1412

(5

7)

(5

8)

(3

6)

(3

8)

(1

28)

(1

48)



Presynch-Ovsynch program with insemination after detected estrus during the Presynch-Ovsynch protocol. While no differences were found in the effectiveness of the 2 strategies in primiparous animals, inseminating at completion of the synchronization program yielded better CR1 (41.2% versus 35.3%) in multiparous cows. Compared to inseminating at spontaneous estrus, another study found that timed insemination after the Presynch-Ovsynch protocol substan-tially improved pregnancy rate in primiparous cows as opposed to multiparous cows (25).

The application of early pregnancy diagnosis methods was asso-ciated with larger improvement of the reproductive parameters in primiparous cows. However, the accuracy of pregnancy diagnosis by means of transrectal ultrasonography is not affected by age of the cow (26). The accuracy of PAG tests showed minor variations by parity according to the results in 519 cows (27). Although significant, the link between early pregnancy diagnosis and shorter days to first service (DFS) in our study is not biologically plausible. This apparent association may be attributable to other management factors, e.g.,

it is possible that cows subjected to early pregnancy diagnosis were more likely to be involved in a hormonal synchronization protocol, which was associated with shorter DFS.

Early pregnancy diagnosis followed by a pregnancy recheck was associated with a greater improvement in CCI and P200 in primipa-rous cows than in multiparous cows. Pregnancy recheck is done to identify when pregnancy loss has occurred after the early pregnancy diagnosis (28). Some studies found no effect of parity on pregnancy loss (29–31), while others found higher incidence of pregnancy wast-age in multiparous cows than in primiparous cows (32–36). A study of 6396 cows from 4 dairy herds found that multiparous cows were almost twice as likely to lose pregnancy 30 to 58 d after insemination as their primiparous counterparts (36). The odds of pregnancy loss were 40% lower in primiparous cows than in cows in parity 3 and above based on data from 1001 pregnant cows in Korea (37).

In addition to reproductive management and parity, herd also played a role in determining fertility results. The magnitude of herd effect was comparable to that of reproductive management practices

Table IV. Descriptive statistics of first-service conception risk (CR1) and pregnancy status at 200 days in milk (DIM) by reproductive management and parity.

Pregnant at 200 days First service conception risk in milk (P200) Primiparous Multiparous Primiparous Multiparous % n % n % n % nVoluntary waiting period (VWP) $ 50 d 24.9 4642 16.6 6138 63.4 4580 49.8 5975

, 50 d 24.7 3742 18.2 5272 67.5 3634 62.2 4945

Estrus detection aids Yes 24.8 5931 17.1 8182 65.7 5806 55.1 7820

No 24.9 2453 18.0 3228 64.0 2408 56.2 3100

Estrus synchronization Yes 24.7 6803 16.7 9394 66.9 6649 55.7 8958

No 25.6 1581 20.2 2016 58.0 1565 54.0 1962

Early pregnancy diagnosis Yes 24.7 6264 16.9 8606 68.3 6084 57.2 8100

No 25.3 2120 18.7 2804 56.3 2130 50.2 2820

Reproductive ultrasound Yes 23.9 5886 16.4 8089 67.9 5727 56.8 7631

No 27.0 2498 19.5 3321 59.2 2487 52.2 3289

Pregnancy recheck Yes 25.6 4651 16.9 6821 67.0 4540 55.1 6484

No 23.9 3733 18.0 4589 63.1 3674 55.8 4436

Frequency of pregnancy diagnosis . Once a week 22.9 2773 16.0 3528 60.7 2750 52.3 3462

Weekly 28.4 4025 20.4 5620 69.9 3933 58.5 5303

, Once a week 17.1 1384 11.2 1958 60.5 1327 52.7 1840



Table V. Differences in reproductive parameters between users and non-users of each management practice by parity based on mixed-effects models (%).

Days to first service (DFS) Breeding interval (IBI) Calving-to-conception interval (CCI) Primiparous Multiparous P-value Primiparous Multiparous P-value Primiparous Multiparous P-valueVoluntary waiting period 19.02 111.97 0.113 112.15 116.92 0.062 18.24 114.56 0.015 (VWP)

Estrus detection aids 23.98 20.92 0.015 23.70 22.72 0.353 20.13 12.90 0.171

Estrus synchronization 218.61 213.93 , 0.001 29.80 24.45 0.026 28.09 23.48 0.005

Early pregnancy diagnosis 219.53 216.36 , 0.001 215.30 29.44 0.001 29.05 25.51 0.006

Reproductive ultrasound 216.16 215.43 0.159 213.30 26.61 , 0.001 23.58 22.01 0.123

Pregnancy recheck 26.37 22.60 , 0.001 23.03 12.11 0.002 23.10 10.99 , 0.001

Frequency of pregnancy diagnosis . Once a week 16.18 21.40 0.983 11.89 20.68 0.999 16.91 10.31 0.905 versus weekly

. Once a week 126.69 120.25 0.084 17.73 15.33 0.596 10.77 26.04 0.867 versus , once a week

Weekly versus 119.31 121.96 0.044 15.74 16.05 0.542 25.74 26.33 0.648 , once a weekWithin parity, the numbers indicate the difference between the users and non-users of a management practice, e.g., calving-to-conception interval was 9.05% and 5.51% shorter in primiparous and multiparous cows, respectively, when pregnancy was diagnosed early. P-values correspond to the difference in the responses to the given management practice between primiparous and multiparous cows. Significant differences, i.e., P , 0.05, are indicated in italics.

Table VI. Odds ratio of conception (first-service conception risk) or pregnancy (pregnant at 200 days in milk) between the users and non-users of each management practice by parity based on the mixed-effects models.

Pregnant at 200 days First-service conception risk in milk (P200) Primiparous Multiparous P-value Primiparous Multiparous P-valueVoluntary waiting period (VWP) 1.011 0.895 0.139 0.834 0.603 , 0.001

Estrus detection aids 0.995 0.940 0.596 1.077 0.956 0.158

Estrus synchronization 0.953 0.792 0.071 1.464 1.071 , 0.001

Early pregnancy diagnosis 0.969 0.884 0.327 1.672 1.326 , 0.001

Reproductive ultrasound 0.849 0.810 0.579 1.458 1.204 , 0.001

Pregnancy recheck 1.096 0.926 0.020 1.187 0.972 , 0.001

Frequency of pregnancy diagnosis . Once a week versus weekly 0.749 0.743 0.068 0.665 0.778 0.197 . Once a week versus , once a week 1.440 1.510 0.012 1.008 0.984 0.862 Weekly versus , once a week 1.923 2.032 , 0.001 1.516 1.265 0.668Within parity, the numbers indicate the odds ratio of conception or pregnancy between the users and non-users of a management practice, e.g., the odds of pregnancy at 200 days in milk (P200) was 1.464-fold greater in primiparous cows and 1.071-fold greater in multiparous cows when estrus synchronization was carried out. P-values correspond to the difference in the responses to the given management practice between primiparous and multiparous cows. Significant differences, i.e., P , 0.05, are indicated in italics.



and parity of the cow. Our results are in agreement with a Canadian study, which found that the effect of reproductive management on fertility was different by farm (38).

In conclusion, estrus synchronization, early pregnancy diagnosis, and pregnancy recheck led to a greater improvement in fertility in primiparous cows than in their multiparous herdmates. Our study has shown that the relationship between reproductive management prac-tices and fertility is different in primiparous and multiparous cows.

A c k n o w l e d g e m e n t sThe authors thank the farm managers and veterinarians who took

part in the survey for providing farm data and information. The Project was supported by the European Union and co-financed by the European Social Fund: (1) EFOP-3.6.1-16-2016-00024 ‘Innovations for Intelligent Specialisation on the University of Veterinary Science and the Faculty of Agricultural and Food Sciences of the Széchenyi István University Cooperation’; (2) EFOP-3.6.2-16-2017-00012 ‘Development of a product chain model for functional, healthy and safe foods from farm to fork based on a thematic research network’; and (3) EFOP-3.6.3-VEKOP-16-2017-00005 ‘Strengthening the sci-entific replacement by supporting the academic workshops and programs of students, developing a mentoring process’.

Re f e r e n c e s1. Gröhn YT, Rajala-Schultz PJ. Epidemiology of reproductive per-

formance in dairy cows. Anim Reprod Sci 2000;60–61:605–614.2. Lucy MC. Reproductive loss in high-producing dairy cattle:

Where will it end? J Dairy Sci 2001;84:1277–1293.3. Walsh SW, Williams EJ, Evans AC. A review of the causes of

poor fertility in high milk producing dairy cows. Anim Reprod Sci 2011;123:127–138.

4. Adrien ML, Mattiauda DA, Artegoitia V, et al. Nutritional regu-lation of body condition score at the initiation of the transition period in primiparous and multiparous dairy cows under graz-ing conditions: Milk production, resumption of post-partum ovarian cyclicity and metabolic parameters. Animal 2012;6: 292–299.

5. Saqib MN, Qureshi MS, Khan RU. Changes in postpartum metabolites and resumption of ovarian cyclicity in primiparous and multiparous dairy cows. Appl Biol Chem 2018;61:107–111.

6. Zhang J, Deng LX, Zhang HL, et al. Effects of parity on uterine involution and resumption of ovarian activities in postpartum Chinese Holstein dairy cows. J Dairy Sci 2010;93:1979–1986.

7. Citek J, Nemeckova D, Stadnik L, Stupka R. Ovarian diseases in Holstein dairy cows, factors determining their occurrence and impact on reproduction traits and breeding costs. Indian J Anim Sci 2017;87:810–813.

8. Ricci A, Carvalho PD, Amundson MC, Fourdraine RH, Vincenti L, Fricke PM. Factors associated with pregnancy-associated glycoprotein (PAG) levels in plasma and milk of Holstein cows during early pregnancy and their effect on the accuracy of preg-nancy diagnosis. J Dairy Sci 2015;98:2502–2514.

9. Mercadante PM, Ribeiro ES, Risco C, Ealy AD. Associations between pregnancy-associated glycoproteins and pregnancy out-

comes, milk yield, parity, and clinical diseases in high-producing dairy cows. J Dairy Sci 2016;99:3031–3040.

10. Borchardt S, Haimerl P, Pohl A, Heuwieser W. Evaluation of prostaglandin F2a versus prostaglandin F2a plus gonadotropin-releasing hormone as Presynch methods preceding an Ovsynch in lactating dairy cows: A meta-analysis. J Dairy Sci 2017; 100:4065–4077.

11. Astiz S, Fargas O. Pregnancy per AI differences between primip-arous and multiparous high-yield dairy cows after using Double Ovsynch or G6G synchronization protocols. Theriogenology 2013;79:1065–1070.

12. Bates DM, Machler M, Bolker BM, Walker SC. Fitting linear mixed-effects models using lme4. J Stat Softw 2015;67:1–48.

13. Hothorn T, Bretz F, Westfall P, Heiberger RM. Simultaneous inference in general parametric models. Biometrical J 2008;50: 346–363.

14. The R Project for Statistical Computing, Vienna, Austria. 2017. Available from: https://www.r-project.org/ Last accessed March 29, 2019.

15. Stangaferro ML, Wijma RW, Masello M, Thomas MJ, Giordano JO. Extending the duration of the voluntary waiting period from 60 to 88 days in cows that received timed artificial insemination after the Double-Ovsynch protocol affected the reproductive performance, herd exit dynamics, and lactation performance of dairy cows. J Dairy Sci 2018;101:1–19.

16. Inchaisri C, Jorritsma R, Vos PL, van der Weijden GC, Hogeveen H. Analysis of the economically optimal voluntary waiting period for first insemination. J Dairy Sci 2011;94:3811–3823.

17. Roelofs J, López-Gatius F, Hunter RH, van Eerdenburg FJ, Hanzen Ch. When is a cow in estrus? Clinical and practical aspects. Theriogenology 2010;74:327–344.

18. Madureira AM, Silper BF, Burnett TA, et al. Factors affecting expression of estrus measured by activity monitors and concep-tion risk of lactating dairy cows. J Dairy Sci 2015;98:7003–7014.

19. Yániz JL, Santolaria P, Giribet A, López-Gatius F. Factors affecting walking activity at estrus during postpartum period and subse-quent fertility in dairy cows. Theriogenology 2006;66:1943–1950.

20. López-Gatius F, Santolaria P, Mundet I, Yániz JL. Walking activ-ity at estrus and subsequent fertility in dairy cows. Theriogenol-ogy 2005;63:1419–1429.

21. Galon N. The use of pedometry for estrus detection in dairy cows in Israel. J Reprod Develop 2010;56:S48–S52.

22. LeRoy CNS, Walton JS, LeBlanc SJ. Estrous detection inten-sity and accuracy and optimal timing of insemination with automated activity monitors for dairy cows. J Dairy Sci 2018;101:1638–1647.

23. Tenhagen BA, Surholt R, Wittke M, Vogel C, Drillich M, Heuwieser W. Use of Ovsynch in dairy herds — Differences between primiparous and multiparous cows. Anim Reprod Sci 2004;80:1–11.

24. Machado VS, Neves R, Lima FS, Bicalho RC. The effect of Presynch-Ovsynch protocol with or without estrus detection on reproductive performance by parity, and the long-term effect of these different management strategies on milk production, repro-duction, health and survivability of dairy cows. Theriogenology 2017;93:84–92.



25. El-Tarabany MS, El-Tarabany AA, Roushdy EM. Impact of par-ity on the efficiency of ovulation synchronization protocols in Holstein cows. Theriogenology 2016;86:2230–2237.

26. Szenci O, Gyulai G, Nagy P, Kovács L, Varga J, Taverne MA. Effect of uterus position relative to the pelvic inlet on the accu-racy of early bovine pregnancy diagnosis by means of ultraso-nography. Vet Q 1995;17:37–39.

27. Dufour S, Durocher J, Dubuc J, Dendukuri N, Hassan S, Buczinski S. Bayesian estimation of sensitivity and specificity of a milk pregnancy-associated glycoprotein-based ELISA and of transrectal ultrasonographic exam for diagnosis of pregnancy at 28–45 days following breeding in dairy cows. Prev Vet Med 2017;140:122–133.

28. Santos JE, Thatcher WW, Chebel RC, Cerri RL, Galvão KN. The effect of embryonic death rates in cattle on the efficacy of estrus synchronization programs. Anim Reprod Sci 2004;82–83:513–535.

29. Starbuck MJ, Dailey RA, Inskeep EK. Factors affecting retention of early pregnancy in dairy cattle. Anim Reprod Sci 2004;84: 27–39.

30. Chebel RC, Santos JE, Reynolds JP, Cerri RL, Juchem SO, Overton M. Factors affecting conception rate after artificial insemination and pregnancy loss in lactating dairy cows. Anim Reprod Sci 2004;84:239–255.

31. Labérnia J, López-Gatius F, Santolaria P, López-Béjar M, Rutllant J. Influence of management factors on pregnancy attrition in dairy cattle. Theriogenology 1996;45:1247–1253.

32. Alnimer MA, Ababneh MM. Effect of estrus expression prior to ovulation synchronization protocols on reproductive efficiency of lactating dairy cow. Livest Sci 2014;163:172–180.

33. Bonneville-Hébert A, Bouchard E, Du Tremblay D, Lefebvre R. Effect of reproductive disorders and parity on repeat breeder status and culling of dairy cows in Quebec. Can J Vet Res 2011; 75:147–151.

34. Gábor G, Kastelic JP, Abonyi-Tóth Z, Gábor P, Endro’’di T, Balogh OG. Pregnancy loss in dairy cattle: Relationship of ultrasound, blood pregnancy-specific protein B, progesterone and production variables. Reprod Domest Anim 2016;51:467–473.

35. Keshavarzi H, Sadeghi-Sefidmazgi A, Kristensen AR, Stygar AH. Abortion studies in Iranian dairy herds: I. Risk factors for abortion. Livest Sci 2017;195:45–52.

36. Santos JE, Rutigliano HM, Sá Filho MF. Risk factors for resump-tion of postpartum estrous cycles and embryonic survival in lactating dairy cows. Anim Reprod Sci 2009;110:207–221.

37. Lee JI, Kim IH. Pregnancy loss in dairy cows: The contributing factors, the effects on reproductive performance and the eco-nomic impact. J Vet Sci 2007;8:283–288.

38. Neves RC, Leslie KE, Walton JS, LeBlanc SJ. Reproductive per-formance with an automated activity monitoring system versus a synchronized breeding program. J Dairy Sci 2012;95:5683–5693.


Short Communication Communication brève


A primary constraint to consistent weaner output is the availabil-ity of enough service-ready gilts and sows when required for breed-ing (1), which is influenced by weaning-to-estrus intervals (WEIs). It is established that long WEIs ($ 6 d) result in shorter estrus dura-tions, lower farrowing rates, and smaller litter sizes (2,3), probably due to poor timing of insemination relative to ovulation. Very short WEIs (# 3 d) are similarly associated with poorer fertility (2) and, while not specifically investigated, this is also presumably due to poor timing of insemination. Causes of variation in WEIs have pre-viously been reviewed and it was noted that WEI was impacted by differences in the size of the ovarian follicles at weaning (4). Smaller follicles at weaning, which are presumably due to variations in sow nutrient intake caused by the temperature of their environment dur-ing lactation, were associated with longer WEIs. This would account for the usually longer WEI of primiparous sows.

Ovarian follicles grow and regress in waves during late lacta-tion and WEI and oocyte quality can be influenced by follicle size at weaning and when sows are bred relative to the stage of this

wave (4). Feeding altrenogest (AT) to primiparous sows from the day before weaning suppresses the release of luteinizing hormone (LH), which limits follicle growth to 4 to 6 mm and blocks estrus, thus allowing a longer period of metabolic recovery after lactation. The LH pulsatility suppression lasts for about 5 h after 2 daily AT feeds (5) and for more than 12 h after feeding AT for 13 d (6). When pulsatile release of LH resumes, the follicular phase ensues and sows return to estrus in about 6 d. In addition to metabolic recovery, the effect of AT likely involves synchronization of follicle waves with mating.

Follicular growth may start earlier in sows with small litters (# 8 piglets) as they are subject to lower metabolic demands during lactation. These sows may experience late lactation estrus or a very short WEI, which impairs the predictability of the WEI and poten-tially reduces their fertility after breeding. Feeding AT during late lactation to sows with normal litter sizes has been associated with a tendency to increase subsequent litter sizes (7,8). Therefore, we hypothesized that feeding AT to primiparous sows nursing small

Feeding altrenogest during late lactation improves fertility of primiparous sows nursing smaller litters

Nutthee Am-in, Roy N. Kirkwood

A b s t r a c tThe objective of this study was to determine whether feeding altrenogest (AT) to primiparous sows with smaller litters during the last week of lactation would improve their fertility. At day 21 of a 28-day lactation, 40 primiparous sows nursing # 8 pigs were assigned equally to 2 groups, either to be fed 20 mg per day of AT for the last 7 days of lactation or to serve as untreated controls. To detect estrus, sows had daily fence-line contact with a mature boar. At estrus detection, sows were subject to an ovarian examination with transrectal ultrasound and preovulatory follicles ($ 0.6 mm) were counted. Sows were artificially inseminated at estrus detection and again 24 hours later. Compared to controls, the AT-fed sows had longer weaning-to-estrus intervals (WEIs; P , 0.001), more pre-ovulatory follicles (P , 0.001), and larger subsequent litter sizes (P = 0.03). Farrowing rates were unaffected by treatment. These data suggest that sows nursing small litters are more likely to initiate a follicular phase during lactation, but that feeding AT prevents this and increases ovulation rate and subsequent litter size.

R é s u m éL’objectif de la présente étude était de déterminer si en fournissant de l’altrenogest (AT) à des truies primipares avec des petites portées durant la dernière semaine de lactation il y aurait amélioration de leur fertilité. Au jour 21 d’une période de lactation de 28 jours, 40 truies primipares allaitant # 8 porcelets ont été assignées également à deux groupes, pour soit être nourries avec 20 mg par jour d’AT pour les sept derniers jours de la lactation ou de servir de témoins négatifs. Afin de détecter l’œstrus, les truies venaient en contact derrière une barrière avec un verrat mature. Au moment de la détection de l’œstrus, les truies étaient soumises à un examen des ovaires par échographie transrectale et les follicules pré-ovulatoires ($ 0,6 mm) étaient dénombrés. Les truies furent inséminées artificiellement au moment de la détection de l’œstrus et encore une fois 24 h plus tard. Comparativement aux témoins, les truies nourries avec de l’AT avaient un intervalle sevrage-œstrus plus long (WEIs; P , 0,001), avaient plus de follicules pré-ovulatoires (P , 0,001) et des tailles de portées subséquentes plus grandes (P = 0,03). Les taux de mise-bas n’étaient pas affectés par le traitement. Ces résultats suggèrent que les truies allaitant des petites portées sont plus susceptibles d’initier une phase folliculaire durant la lactation, mais qu’en leur donner de l’AT prévient ceci et augmente le taux d’ovulation et la taille subséquente des portées.


Department of Obstetrics, Gynaecology and Reproduction, Faculty of Veterinary Science, Chulalongkorn University, Bangkok, Thailand (Am-in); School of Animal and Veterinary Sciences, University of Adelaide, Roseworthy 5371, Australia (Kirkwood).

Address all correspondence to Dr. Roy N. Kirkwood; telephone: 161 8 8313 7617; e-mail: [email protected]

Received May 7, 2018. Accepted June 26, 2018.



litters during the last week of lactation will prevent resumption of follicle growth during lactation, resulting in normal and synchro-nized WEI and increased subsequent litter sizes.

This pilot study was conducted on a 1200-sow commercial farrow-to-finish farm in Nakhon Pathom, Thailand from August 2014 to May 2015. It was approved by the Institutional Animal Care and Use Committee at Chulalongkorn University. At day 21 of a 28-day lactation, 40 primiparous Landrace-Yorkshire sows were selected that were nursing 8 or fewer piglets. Sows were divided equally into 2 groups to either receive orally 20 mg/d of altrenogest (AT) (Virbagest; Virbac Animal Health, Nonthaburi, Thailand) during the last 7 d of lactation, including the day of weaning (n = 20), or to serve as non-treated controls (n = 20).

To facilitate estrus detection, sows had daily fence-line contact with a mature boar from the day after weaning. At estrus detec-tion, all sows were subject to an ovarian examination with tran-srectal ultrasound using a 5-MHz convex probe (HS-2000; Honda Electronics, Tokyo, Japan) and the preovulatory follicles ($ 0.6 mm) were counted. Sows were inseminated at estrus detection and again 24 h later with 3 3 109 sperm in 80-mL of extender. From weaning, sows were housed individually in gestation stalls in an open-sided barn and were fed 3 kg/d of a commercially formulated gestation diet. During gestation, feed allowance was 2 kg/d. Water was freely available from nipple drinkers. Data recorded were WEI, number of preovulatory follicles immediately before insemination, and subse-quent farrowing rates and litter sizes. All analyses were conducted with SAS 9.1 (SAS Institute, Cary, North Carolina, USA) and data are presented as mean 6 SD. The WEI, number of preovulatory follicles, and litter sizes were tested for normality using the UNIVARIATE procedure option NORMAL. As the number of preovulatory follicles did not follow a normal distribution, they were compared using Mann-Whitney U-test. Weaning-to-estrus intervals (WEIs) and total-born piglets were compared using t-test and farrowing rates were compared using Fisher’s exact test.

All sows exhibited estrus and were inseminated. The control group had a shorter WEI than the AT-fed group (2.7 6 1.0 versus 5.7 6 1.4 d, respectively; P # 0.0001). The farrowing rates were not significantly different between treated and control groups (95% versus 90%, respectively), which reflected the excellent breeding management on this farm. Compared to the control sows, however, the AT-fed sows had more preovulatory follicles (22.7 6 3.6 versus 17.3 6 3.5; P # 0.0001) and more total-born piglets (11.9 6 1.8 versus 10.6 6 2.0; P = 0.03).

Our data show that primiparous sows nursing small litters exhibit a shorter WEI, which suggests that follicle growth resumes during late lactation. These data also support our hypothesis that feeding altrenogest (AT) during late lactation suppresses the resumption of follicle growth and results in a normal WEI. If we accept that follicu-lar phases will be of similar length in both treatments, it suggests that the follicular phase in AT-fed sows tended to be initiated 3 d before weaning and, by extension, the AT-fed sows were subject to approxi-mately 3 d of follicular growth suppression. Previous studies with weaned sows have shown that a short duration of AT feeding can result in poorer performance in sows that had large follicles at wean-ing, probably because these follicles lose their estrogenic capacity and become atretic (6). In the present study, however, this adverse effect of

short-term AT feeding on fertility was not evident when the ovaries had small follicles at weaning, as would be expected in late lactation.

It has been demonstrated that follicles continue to grow to about 4 to 6 mm even under AT suppression (5,9). Therefore, it is appar-ent that, while AT suppresses both luteinizing hormone (LH) and follicle-stimulating hormone (FSH), the degree of suppression, par-ticularly for FSH, is not sufficient to prevent follicles from growing to a medium size and sufficient gonadotrophic activity remains to allow limited follicle growth (5). This may explain, at least in part, why the feeding of AT to gilts results in increased ovulation rates and larger litters for gilts and young sows (10–12). In a scenario analogous to the mechanism of estrus synchronization, at the initia-tion of AT feeding, the growth of medium follicles is arrested, while smaller follicles grow to medium size. When AT is withdrawn, the follicle pool would contain the original medium follicles and other small follicles that had time to achieve medium size. If these follicles have not initiated apoptosis and oocyte degeneration, an increased ovulation rate and potentially larger litters will ensue.

In conclusion, this study showed that primiparous sows nursing small litters are likely to initiate a follicular phase during lactation, which leads to shorter and unpredictable weaning-to-estrus intervals (WEIs). The feeding of AT prevented this follicular phase during lactation, normalized the WEI, and increased ovulation rate and subsequent litter size.

A c k n o w l e d g m e n t sThe authors are grateful for the financial support of the Thailand

Research Fund (TRG5780251) and the Ratchadaphiseksomphot Endowment Fund, Chulalongkorn University.

Re f e r e n c e s 1. Dial GD, Duangkaew C, King V, Rademacher C. The influence

of the gilt pool on weaned pig output. Proc AD Leman Swine Conf, Minneapolis, MN 1996;23:39–41.

2. Wilson MR, Dewey CE. The associations between weaning-to-estrus interval and sow efficiency. J Swine Health Prod 1993; 1:10–15.

3. Steverink DW, Soede NM, Groenland GJ, van Schie FW, Noordhuizen JP, Kemp B. Duration of estrus in relation to repro-duction results in pigs on commercial farms. J Anim Sci 1999;77: 801–809.

4. Lucy MC, Liu J, Boyd CK, Bracken CJ. Ovarian follicular growth in sows. Reprod Suppl 2001;58:31–45.

5. van Leeuwen JJ, Martens MR, Jourquin J, et al. Follicle size and reproductive hormone profiles during a post-weaning altrenogest treatment in primiparous sows. Reprod Fertil Dev 2015;27:304–312.

6. van Leeuwen JJ, Williams SI, Martens MR, et al. The effect of different post-weaning altrenogest treatments of primiparous sows on follicular development, pregnancy rates, and litter sizes. J Anim Sci 2011;89:397–403.

7. Kirkwood RN, Smith WC, Lapwood KR. Influence of oral administration of allyl trenbolone on subsequent litter size of primiparous sows. NZ J Exp Agric 1986;14:477–480.



8. Kirkwood RN, Aherne FX. The influence of feeding allyl tren-bolone during late lactation on the reproductive performance of primiparous and multiparous sows. Can Vet J 1988;29:68–69.

9. van Leeuwen JJ, Martens MR, Jourquin J, Driancourt MA, Kemp B, Soede NM. Variation in LH pulsatility during 24h after a post-weaning altrenogest treatment in relation to follicle development in primiparous sows. Anim Reprod Sci 2011;126:101–107.

10. Marinat-Botté F, Bariteau F, Forgerit Y, et al. Control of oestrus in gilts II. Synchronization of oestrus with a progestogen, altreno-gest (Regumate): Effect on fertility and litter size. Anim Reprod Sci 1990;22:227–233.

11. Marinat-Botté F, Bariteau F, Forgerit Y, Macar C, Poirier P, Terqui M. Synchronization of oestrus in gilts with altrenogest: Effects on ovulation rate and foetal survival. Anim Reprod Sci 1995;39:267–274.

12. Koutsotheodorus F, Hughes PE, Parr RA, Dunshea FR, Fry RC, Tilton JE. The effects of post-weaning progestogen treatment (Regumate) of early-weaned primiparous sows on subsequent reproductive performance. Anim Reprod Sci 1998;52:71–79.




Methicillin-resistant Staphylococcus pseudintermedius (MRSP) strains are commonly found on canine skin after treatment with antibiotics and are emerging pathogens in human patients (1). These organisms colonize humans who have contact with dogs, and they can contami-nate the environment in veterinary hospitals and households (2). The initial step in colonization and infection is adhesion to skin cells (corneocytes), which may be influenced by bacterial cell wall-associated (CWA) proteins encoded by S. pseudintermedius surface protein genes (sps) (3). The ability of S. pseudintermedius to adhere to both canine and human corneocytes has been used as evidence for zoonotic potential (4). The adhesion ability of individual strains may relate to their source of isolation. This ability was previously evaluated for isolates from dogs and humans, but environmental isolates had not yet been investigated (4,5). Thus, to estimate their pathogenic potential, their ability to adhere to human and canine cor-neocytes needed to be assessed. Moreover, the association between adherence and genes encoding CWA proteins is still not clear. This study aimed to determine the adherence ability of 5 MRSP isolates

and whether there were correlations between the presence of genes for CWA proteins and adherence.

Five MRSP isolates were obtained from the culture collection at Chulalongkorn University, Bangkok, Thailand (Table I). They included 2 isolates each from dogs and the environment and 1 isolate from a human. The sequence type (ST) of these isolates in multilocus sequence typing and their profile of genes encoding CWA proteins had previously been established (1,6).

Corneocytes were collected from dogs and humans according to the Animal Care and Use Protocol (no. 1731009) of Chulalongkorn University and with approval of the Research Ethics Review Committee for Research Involving Human Research Participants, Health Science Group, Chulalongkorn University (no. 225.1/59). The corneocytes were collected from the abdominal area of 5 dogs and from the medial biceps area of 5 human volunteers who had little contact with dogs, as previously described (5,7). All dogs and humans were clinically healthy and did not have a history of skin disease. Debris and normal microbiota were removed from the

Investigating the ability of methicillin-resistant Staphylococcus pseudintermedius isolates from different sources to adhere to canine

and human corneocytesNathita Phumthanakorn, Nuvee Prapasarakul

A b s t r a c tAssays were done to assess the ability of 5 methicillin-resistant Staphylococcus pseudintermedius (MRSP) isolates from difference sources to adhere to canine and human corneocytes. Cell wall-associated (CWA) protein gene profiles were examined to look for associations with adherence. Five MRSP strains were studied: 3 with the same CWA protein gene profile (14 genes) and belonging to sequence type (ST) 45 were isolated from a dog, a human, and the environment. The other 2 were an environmental isolate belonging to ST433 that had the lowest number of CWA protein genes (12) and a canine clinical isolate belonging to ST733 that had the greatest number of CWA protein genes (18). The 3 isolates of MRSP ST45, a major clone in Thailand, had the greatest ability to adhere to canine and human corneocytes. Nevertheless, MRSP adherence ability could not be predicted from the profile of genes encoding CWA proteins.

R é s u m éDes analyses ont été effectuées afin de déterminer la capacité de cinq isolats de Staphylococcus pseudintermedius résistants à la méthicilline (SPRM) provenant de différentes sources à adhérer à des cornéocytes canins et humains. Les profils de gènes des protéines associées à la paroi cellulaire (APC) furent examinés afin de vérifier des associations avec l’adhérence. Cinq souches de SPRM furent étudiées : trois avec le même profil de gène de protéines APC (14 gènes) et appartenant au type de séquence (ST) 45 isolées d’un chien, un humain et l’environnement. Les deux autres souches étaient un isolat provenant de l’environnement et appartenant à ST433 et qui avait le plus petit nombre de gènes de protéines APC (12) et un isolat clinique canin appartenant au ST733 et qui avait le plus grand nombre de gènes de protéines APC (18). Les trois isolats de ST45, un clone majeur en Thaïlande, avait la plus grande capacité d’adhérer aux cornéocytes canins et humains. Toutefois, la capacité d’adhérence des SPRM ne pouvait être prédite par le profil de gènes codant pour les protéines APC.


Department of Microbiology, Faculty of Veterinary Science, Chulalongkorn University, Henri-Dunant Road, Pathumwan, Bangkok 10330, Thailand (Phumthanakorn, Prapasarakul); Diagnosis and Monitoring of Animal Pathogens Research Unit, Faculty of Veterinary Science, Chulalongkorn University, Henri-Dunant Road, Pathumwan, Bangkok 10330, Thailand (Prapasarakul).

Address all correspondence to Dr. Nuvee Prapasarakul; telephone: 166(0) 22189581-3; fax: 166(0) 22511656; e-mail: [email protected]

Received May 7, 2018. Accepted July 22, 2018.



sampling sites by application of 5 successive adhesive tape strips (Scotch Tape; 3M, St. Paul, Minnesota, USA), and then medical adhesive disks 22 mm in diameter (D-Squame; CuDerm Corporation, Dallas, Texas, USA) were placed on the skin to collect corneocytes. The confluence of corneocytes in the disk was examined by light microscopy, and disks with confluent corneocytes were selected for the adherence assays.

The adherence assay was done in duplicate with some minor modifications from previous protocols (5). Briefly, the isolates were cultured overnight in brain–heart infusion (BHI) broth (Oxoid; Thermo Scientific, Basingstoke, England) at 37°C. An aliquot of the MRSP suspension (100 mL) was transferred into 30 mL of BHI broth with shaking, and 25 mL of the culture at mid-exponential phase was centrifuged at 1500 3 g and 4°C for 5 min. The bacterial pellet was centrifuged at 800 3 g and 4°C for 10 min and washed 3 times with sterile phosphate-buffered saline (PBS). The bacterial suspension was adjusted to an optical density at 600 nm of 0.15, which represented approximately 3 3 107 colony-forming units per milliliter. Two cor-neocyte disks were incubated with 500 mL of bacterial suspension for 45 min, and 1 disk from each individual was incubated with PBS as a negative control. After incubation the disks were washed with sterile PBS for 20 s, stained with 0.5% crystal violet for 20 s, and then washed with water for 20 s. Ten fields of confluent corneocytes per disk were examined under a microscope (3 1000) to determine bac-terial adhesion counts (BACs). Microscopic pictures were captured with Motic Images Plus software, Version 2.0 (Motic China Group Company, Xiamen, China).

The Mann–Whitney U-test was used to compare the median BAC of individual strains for different types of corneocytes. The Kruskal–Wallis test was used to find the difference between strains, with SPSS software, Version 22.0 (IBM Corporation, Armonk, New York, USA).

The 3 MRSP strains of ST45, the main clone in Thailand, had significantly higher BACs than the 2 strains of ST733 and ST433 (P , 0.001) (Figure 1). The ST45 isolates from the environment and from a human showed the greatest adherence to canine and human corneocytes (Figure 2). No positive association was found between adherence and the sps gene profiles: MRSP SW1F4O2, belonging to ST433, had the highest number of sps genes (18/18) and the least adherence to both canine and human corneocytes (Figures 1 and 2). Interestingly, the 2 isolates from dogs, belonging to ST45 and ST733, had significantly higher BACs with human corneocytes than with canine corneocytes (P = 0.007 and P , 0.001, respectively). The iso-late from a human, belonging to ST45, had a significantly greater BAC with human corneocytes than with canine cells (P , 0.001), suggesting some host specificity. The 2 environmental strains showed no difference in attachment to canine or human cells (Figure 1).

The adherence assay is a reproducible and rapid method for quan-titative measurement of staphylococcal adherence to corneocytes (8,9). The low number of representative isolates might be a limita-tion of this study. Nevertheless, it is difficult to perform this assay with a high number of isolates because of the large number of host corneocytes needed. Accordingly, many studies using the adherence assay have had limited numbers of isolates (4,9–11).

In this study, MRSP derived from different sources adhered to canine and human corneocytes. The environmental isolate SW1F4O1, belonging to ST45, showed an adhesion ability resembling that of Ta

ble

I. M

ethi

cilli

n–re

sist

ant

Sta

phyl

ococ

cus

pseu

dint

erm

ediu

s is

olat

es u

sed

in t

his

stud

y.

S.

pse

udin

term

ediu

s su

rface

pro

tein

gen

esa

Stra

in

Sour

ce

Site

ST

SC

Cmec

sp

sA

spsB

sp

sC

spsD

sp

sE

spsF

sp

sG

spsH

sp

sI

spsJ

sp

sK

spsL

sp

sM

spsN

sp

sO

spsP

sp

sQ

spsR

AP65

D

og

Perin

eum

45

SCCm

ec57

395

1

1

1

2

1

2

1

1

1

1

1

1

1

1

1

2

2

1

VB88

H

uman

N

asal

cav

ity

45

SC

Cmec

5739

5 1

1

1

2

1

2

1

1

1

1

1

1

1

1

1

2

2

1

SW1F

4O1

Envi

ronm

ent

Der

mat

olog

y

45

SC

Cmec

5739

5 1

1

1

2

1

2

1

1

1

1

1

1

1

1

1

2

2

1

un

it flo

or

SW1F

4O2

Envi

ronm

ent

Der

mat

olog

y 4

33

Non

type

able

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

un

it flo

or

MIC

460

Dog

Py

oder

ma

733

V 1

1

1

2

1

2

1

1

2

1

1

1

1

1

2

2

2

1

a D

etec

ted

as d

escr

ibed

pre

viou

sly

(6).

ST —

seq

uenc

e ty

pe; S

CCm

ec —

sta

phyl

ococ

cal c

asse

tte c

hrom

osom

e m

ec.



isolates from a dog and a human in this clonal lineage. The MRSP SW1F4O2, belonging to ST433, also adhered to both types of corneo-cytes, but it had the weakest adherence. The results also showed that strains from the major Thai clone, ST45, had the greatest potential

to adhere to canine and human corneocytes, reflecting the situation with the major European type, ST71. This suggests that strains of the predominant STs are better adapted to attaching to canine and human skin cells than are strains belonging to other STs (5). The

Figure 1. Box plots representing bacterial adhesion counts (BACs) of 5 methicillin-resistant Staphylococcus pseudintermedius (MRSP) isolates on canine corneocytes (CC) and human corneocytes (HC). The top and bottom of the box represent lower quartile and upper quartile. The line in the box is the median and the whisker is the range of the data. The asterisk indicates a statistically significant difference at P , 0.05.

Figure 2. Adhesion of MRSP to canine and human corneocytes. A — MRSP SW1F4O1-ST45 adhering to canine corneocytes. B — MRSP VB88-ST45 adhering to human corneocytes. C and D — MRSP SW1F4O2-ST433 adhering to canine and human corneocytes, respectively. Bar — 10 mm.

A

C

B

D



3 strains that showed greater adhesion to human than to canine cor-neocytes might have a greater ability to adapt to human corneocytes in vivo than other strains (5). The strain variation may also influence the adhesion ability. It has been suggested that clinical isolates might adhere to corneocytes more than isolates from healthy dogs (11). In this study, we did not observe this variation since a single clinical isolate was used. Comparison of the adhesion ability of isolates belonging to the dominant clones from clinical and healthy dogs would be useful. In this study no association was found between sps profile and BAC. So far, only 3 out of 18 sps genes — spsD, spsL, and spsO — have been proven to adhere to corneocytes (12). Nevertheless, MRSP SW1F4O2, belonging to ST433 and harboring spsD, spsL, and spsO, had the lowest ability to adhere to canine and human corneocytes in this study. Therefore, the link between sps profile and adhesion ability still needs to be clarified. The adher-ence of S. pseudintermedius is possibly influenced by the colonization status of the corneocyte donors, and strains with greater adhesion capacity might reside in persistent carriers (10). Moreover, bacterial colonization is a multifactorial process; therefore, products encoded by other gene types may be required for host binding, as described for S. aureus (9). A combination of in vivo experiments, whole genome comparisons, and analysis of surface protein gene expression using canine and human MRSP should provide more understanding of basic bacterial host adaptation (9,13).

In this study, MRSP ST45 isolates from a dog, a human, and the environment adhered to both canine and human corneocytes. The clone type of MRSP may influence the initial stage of colonization and infection but without a direct relationship to the surface protein gene profile.

A c k n o w l e d g m e n t sThis work was supported by a scholarship from the Graduate

School, Chulalongkorn University. Bangkok, Thailand, to com-memorate the 72nd anniversary of his Majesty King Bhumibala Aduladeja, the 100th Anniversary of the Chulalongkorn University Fund for Doctoral Scholarship, and the 90th Anniversary of the Chulalongkorn University Fund (Ratchadaphiseksomphot Endowment Fund). We thank Professor David J. Hampson of City University of Hong Kong, Kowloon, for his kind editorial assistance during preparation of this manuscript.

Re f e r e n c e s 1. Chanchaithong P, Perreten V, Schwendener S, et al. Strain typ-

ing and antimicrobial susceptibility of methicillin-resistant coagulase- positive staphylococcal species in dogs and people associated with dogs in Thailand. J Appl Microbiol 2014;117: 572–586.

2. van Duijkeren E, Catry B, Greko C, et al. Review on methicillin- resistant Staphylococcus pseudintermedius. J Antimicrob Chemother 2011;66:2705–2714.

3. Bannoehr J, Ben Zakour NL, Reglinski M, et al. Genomic and surface proteomic analysis of the canine pathogen Staphylococcus pseudintermedius reveals proteins that mediate adherence to the extracellular matrix. Infect Immun 2011;79:3074–3086.

4. Woolley KL, Kelly RF, Fazakerley J, Williams NJ, Nuttall TJ, McEwan NA. Reduced in vitro adherence of Staphylococcus species to feline corneocytes compared to canine and human corneocytes. Vet Dermatol 2008;19:1–6.

5. Latronico F, Moodley A, Nielsen SS, Guardabassi L. Enhanced adherence of methicillin resistant Staphylococcus pseudintermedius sequence type 71 to canine and human corneocytes. Vet Res 2014; 45:70.

6. Phumthanakorn N, Chanchaithong P, Prapasarakul N. Develop-ment of a set of multiplex PCRs for detection of genes encoding cell wall-associated proteins in Staphylococcus pseudintermedius isolates from dogs, humans and the environment. J Microbiol Methods 2017;142:90–95.

7. Saijonmaa-Koulumies LE, Lloyd D. Adherence of Staphylococcus intermedius to canine corneocytes in vitro. Vet Dermatol 2002;13: 169–176.

8. Lu Y-F, McEwan NA. Staphylococcal and micrococcal adherence to canine and feline corneocytes: Quantification using a simple adhesion assay. Vet Dermatol 2007;18:29–35.

9. Moodley A, Espinosa-Gongora C, Nielsen SS, McCarthy AJ, Lindsay JA, Guardabassi L. Comparative host specificity of human- and pig-associated Staphylococcus aureus clonal lineages. PLoS One 2012;7:e49344.

10. Paul NC, Latronico F, Moodley A, Nielsen SS, Damborg P, Guardabassi L. In vitro adherence of Staphylococcus pseudinter-medius to canine corneocytes is influenced by colonization status of corneocyte donors. Vet Res 2013;44:52.

11. McEwan NA, Kalna G, Mellor D. A comparison of adherence by four strains of Staphylococcus intermedius and Staphylococcus homi-nis to canine corneocytes collected from normal dogs and dogs suffering from atopic dermatitis. Res Vet Sci 2005;78:193–198.

12. Bannoehr J, Brown JK, Shaw DJ, Fitzgerald RJ, van den Broek AH, Thoday KL. Staphylococccus pseudintermedius surface pro-teins SpsD and SpsO mediate adherence to ex vivo canine corneocytes. Vet Dermatol 2012;23:119–124.

13. Uhlemann AC, Porcella SF, Trivedi S, et al. Identification of a highly transmissible animal-independent Staphylococcus aureus ST398 clone with distinct genomic and cell adhesion properties. MBio 2012;3:e00027–12.




Morphine is one of the opioids most often used for relief of acute pain, including surgical pain, in dogs. When administered alone to this species, morphine induced mild to moderate seda-tion (1). Opioids are also administered in combination with tran-quilizers, such as acepromazine, to induce neuroleptoanalgesia (2). Acepromazine is a phenothiazine derivative that causes mild to moderate sedation in dogs (3,4). More pronounced sedation was observed in dogs when acepromazine was combined with opioid analgesics compared with administration of each drug alone (3). Reported side effects of such combinations in dogs were decreased arterial blood pressure and heart rate; however, the magnitude of these effects was not greater than when each drug was administered alone (3,4).

The degree of sedation induced by acepromazine–opioid combi-nations has been shown to be influenced by the choice of opioid. In a previous study in dogs (4), sedation scores were compared after

administration of acepromazine [0.05 mg/kg body weight (BW)] combined with methadone (0.5 mg/kg BW), morphine (0.5 mg/kg BW), butorphanol (0.2 mg/kg BW), or tramadol (2 mg/kg BW). Sedation was greatest when acepromazine was combined with methadone, lowest when it was combined with tramadol, and inter-mediate when it was combined with either morphine or butorphanol. In a recent study in dogs (5), methadone at doses of 0.25, 0.5, and 0.75 mg/kg BW improved the sedation induced by acepromazine, the highest dose providing a longer duration of effect. Conversely, another study (6) showed that tramadol at doses of 3 and 5 mg/kg BW failed to improve acepromazine-induced sedation in dogs.

Despite the widespread use of acepromazine–morphine (AM) combinations in dogs, to the authors’ knowledge no study has com-pared the influence on sedation level in dogs of different morphine doses in combination with acepromazine. The present study aimed to do that and to study the effects on some physiological parameters.

Effects of 3 morphine doses, in combination with acepromazine, on sedation and some physiological parameters in dogsEduardo R. Monteiro, Thiago A. Rabello, Julia P.P. Rangel, Juarez S. Nunes Jr,

Cesar D. Freire, Daniela Campagnol

A b s t r a c tThis study evaluated the effects of 3 morphine doses combined with acepromazine, on sedation and physiological parameters in 5 clinically healthy dogs. Four treatments were administered intramuscularly in a randomized, blinded, crossover design: acepromazine, 0.05 mg/kg, alone (ACP) and acepromazine plus morphine at doses of 0.25, 0.5, and 1.0 mg/kg body weight (BW) (AM0.25, AM0.5, and AM1.0, respectively). Sedation scores and cardiorespiratory variables were evaluated for 120 min after drug administration. The sedation scores were significantly higher with the AM0.25 and AM1.0 treatments than with the ACP treatment. At 30 min the scores were 36% to 66% higher with AM1.0 than with AM0.25 and AM0.5, respectively, but these differences were not significant. The physiological variables remained acceptable for dogs. The results of this study do not support the use of AM0.5 over AM0.25 to improve sedation in dogs, but they do indicate that sedation may be greater with AM1.0 than with AM0.25 and AM0.5. Studies with a greater number of samples are warranted to confirm this statement.

R é s u m éCette étude visait à évaluer les effets de trois doses de morphine combinées à de l’acépromazine, sur la sédation et des paramètres physiologiques chez cinq chiens cliniquement en santé. Quatre traitements furent administrés par voie intramusculaire dans un design croisé randomisé à l’aveugle : acépromazine, 0,05 mg/kg seule (ACP) et acépromazine plus morphine à des doses de 0,25, 0,5, et 1,0 mg/kg de poids corporel (AM0,25, AM0,5, et AM1,0, respectivement). Les pointages de sédation et des variables cardiorespiratoires furent évalués pour 120 min après l’administration des drogues. Les pointages de sédation étaient significativement plus élevés avec les traitements AM0,25 et AM1,0 qu’avec le traitement ACP. À 30 min, les pointages étaient 36 % et 66 % plus élevés avec AM1,0 qu’avec AM0,25 et AM0,5, respectivement, mais ces différences n’étaient pas significatives. Les variables physiologiques sont demeurées acceptables pour les chiens. Les résultats de cette étude ne militent pas en faveur de l’utilisation d’AM0,5 par rapport à AM0,25 pour améliorer la sédation chez les chiens, mais ils indiquent que la sédation peut être plus grande avec AM1,0 qu’avec AM0,25 et AM0,5. Des études avec un plus grand nombre d’échantillons sont requises pour confirmer cet énoncé.


School of Veterinary Medicine, University of Vila Velha, Vila Velha, ES, Brazil.

Address all correspondence to Dr. Eduardo R. Monteiro; telephone: 55-51-33087856; e-mail: [email protected]

Dr. Monteiro’s current address is Department of Animal Medicine, Faculty of Veterinary Medicine, Federal University of Rio Grande do Sul, Porto Alegre, RS, Brazil.

Received April 19, 2018. Accepted August 14, 2018.



The hypothesis was that sedation and physiological changes would be greater with higher morphine doses.

This study was approved by the institutional Animal Research Ethical Committee (protocol 336/2014). Six adult crossbreed dogs (3 males and 3 females) were used. The dogs were considered clinically healthy according to the results of physical examination, a complete blood (cell) count, and serum biochemistry analyses. Each dog underwent 4 treatments, on different occasions, with 1-week washout intervals, in a randomized, blinded, crossover design. All treatments were administered intramuscularly. For treatment with acepromazine alone (ACP), Acepran 0.2% (Vetnil; Louveira, São Paulo, Brazil) was administered at a dose of 0.05 mg/kg BW. For combination treatments, designated AM0.25, AM0.5, and AM1.0, acepromazine (same dose) was mixed in the same syringe with morphine [Dimorf (Cristália, Itapira, São Paulo, Brazil)] at doses of 0.25, 0.5, and 1.0 mg/kg BW, respectively. On all occasions the final volume of the injection was 0.125 mL/kg BW, obtained by adding physiologic saline as appropriate. A person not involved in other phases of the study prepared the solutions.

Food, but not water, was withheld for 12 h before treatment. A 20-gauge catheter was percutaneously introduced into a dor-sal pedal artery after local infiltration of lidocaine. The catheter was connected to a noncompliant tubing system and a pressure transducer (TruWave; Edwards Lifesciences, Draper, Utah, USA) filled with heparinized saline. The transducer was connected to a multiparameter monitor (Lifewindow 6000Vet; Digicare Animal Health, Boynton Beach, Florida, USA), zeroed at the level of the manubrium of the sternum, to allow measurement of systolic, mean, and diastolic arterial pressures (SAP, MAP, and DAP). Thereafter, the dogs were left undisturbed for 30 min before baseline variables were recorded. Heart rate and respiratory rate were counted by auscultation with a stethoscope and by observation of chest wall movements, respectively. Rectal temperature was measured with a digital thermometer. Blood samples from the arterial catheter were collected for measurement of arterial pH, oxygen partial pressure (PaO2), carbon dioxide partial pressure (PaCO2), and bicarbonate (HCO3

2) concentration with a portable analyzer (i-STAT 1; Abbott Point of Care, Princeton, New Jersey, USA). On all occasions blood gas analysis was done immediately after sampling.

Sedation was evaluated with the use of 2 scoring systems: a visual analogue scale (VAS) and a numeric descriptive scale (NDS). The VAS consisted of a 10-cm line on which the left end, 0, represented no sedation and the right end, 10, represented the most sedation possible. On each occasion the distance between the left end and the mark assigned by the observer provided the VAS score for that occasion. The NDS ranged from 1 to 6 as follows: 1 — alert: no motor deficits, equivalent to the state before anesthesia; 2 — faint sedation: stands, walks, some ataxia and disorientation; 3 — slight sedation: stands but ataxic, can remain sternal; 4 — mild sedation: can stay sternal, cannot stand, may struggle; 5 — moderate sedation: can raise head, usually laterally recumbent; and 6 — heavy sedation: nonresponsive, cannot raise head (7). Data collection was standard-ized. Between assessments the dogs were left on the floor. At each time point the dog was first observed undisturbed and lying on the floor. Then the dog was gently positioned in lateral recumbency on a table, and the physiological parameters were measured. The dog

was then returned to the floor and encouraged to stand and walk. Finally, the observer recorded the VAS and NDS scores. A single observer, unaware of the assigned treatment, evaluated sedation on all occasions. This person was familiar with the dogs’ behavior and the scoring systems.

For all variables a baseline value was recorded before treatment. Values for SAP, MAP, heart rate, and respiratory rate, as well as seda-tion scores, were reassessed at 15, 30, 45, 60, 90, and 120 min after drug administration. Rectal temperature was recorded at 30, 60, 90, and 120 min, and a 2nd sample for blood gas analysis was collected 45 min after drug administration.

Sample size was calculated by means of G*Power statistical soft-ware for Windows (Version 3.1.6; Universität Kiel, Kiel, Germany). Six dogs per group were considered necessary to detect a 3.0-cm difference in VAS sedation scores with an a error of 0.05 and a power of 80%. The assumption of normality of data was examined by a Kolmogorov–Smirnov test. For cardiorespiratory and temperature data a 1-way repeated-measures analysis of variance (ANOVA) fol-lowed by Dunnett’s test was used to compare differences between each time point (15 to 120 min) and baseline. For blood gas variables a paired t-test was used to compare data at 45 min with baseline values. Differences among treatments were identified by means of a 2-way repeated-measures ANOVA. When a difference was signifi-cant, the Bonferroni correction was used to identify the time point at which the difference existed. For subjective data (VAS and NDS scores) Friedman and Dunn’s tests were done to compare differences over time and between treatments. For all analyses, differences were considered significant at a P-value , 0.05.

Signs of disease developed in 1 male dog just before the beginning of the study, and the animal was excluded. Therefore, only 5 dogs completed the study. Mean 6 standard deviation weight of dogs was 15.6 6 3.2 kg. As Table I shows, with ACP treatment the heart rate was increased above baseline at all time points, but significant dif-ferences were detected only at 60 and 90 min (P = 0.03). With AM0.25, AM0.5, and AM1.0 treatments there was a trend toward a decrease in heart rate over time, but a significant difference was observed only with AM0.5 treatment at 90 min (P = 0.04). With ACP treatment the SAP was decreased below baseline at 15 (P = 0.04), 30 (P = 0.03), and 90 (P = 0.04) min and the MAP was decreased at 15 (P = 0.03) and 30 (P = 0.02) min. With AM0.25 treatment the SAP was decreased at 30 (P = 0.0009), 45 (P , 0.0001), 60 (P , 0.0001), 90 (P , 0.0001), and 120 (P = 0.004) min, the MAP was decreased at 30 (P = 0.006), 45 (P = 0.0001), 60 (P = 0.0003), and 90 (P = 0.002) min, and the DAP was decreased at 45 (P = 0.005) and 60 (P = 0.004) min. With AM0.5 treatment the SAP was decreased at 45 (P = 0.04) min, the MAP was decreased at 15 (P = 0.03), 30 (P = 0.0004), 45 (P = 0.0003), and 60 (P = 0.0005) min, and the DAP was decreased at 45 (P = 0.04) and 60 (P = 0.02) min. With AM1.0 treatment the SAP was decreased from 15 to 120 min (P-values: 0.005 at 15 min, , 0.0001 from 30 to 60 min, 0.0011 at 90 min, and 0.005 at 120 min), and the MAP was decreased at 30 (P = 0.002), 45 (P = 0.0006), 60 (P = 0.01), and 90 (P = 0.02) min.

The respiratory rate (Table I) decreased significantly below base-line with ACP treatment, from 45 to 120 min (P-values at 45, 60, 90, and 120 min: 0.02, 0.002, 0.007, and 0.02, respectively). With AM0.5 and AM1.0 treatments the respiratory rate was significantly decreased at 45 (P = 0.03) and 120 (P = 0.02) min and at 120 (P = 0.04) min,



respectively. Significant decreases from baseline in pH and increases in PaCO2 concentration (Table II) were observed at 45 min with AM0.5 treatment (P = 0.008 and P = 0.005, respectively) and with AM1.0 treatment (P = 0.009 and P = 0.02, respectively). Compared with ACP treatment the treatments with AM0.5 and AM1.0 resulted in a lower pH (P = 0.008 and P = 0.0003, respectively) at 45 min, and treatment with AM1.0 resulted in a higher PaCO2 (P = 0.02) at 45 min. The HCO3

2 concentration was increased from baseline with AM0.5 treatment at 45 min (P = 0.04) but did not differ significantly from the values with the other treatments (Table II). The mean PaO2 values (Table II) were . 80 mmHg with all the treatments at baseline and

at 45 min, but with AM0.5 treatment the value was below baseline at 45 min (P = 0.046).

Rectal temperature (Table I) was decreased significantly below baseline with AM0.5 at 30 min (P = 0.03) and from 60 to 120 min (P = 0.0001 at 60 min and P , 0.0001 at 90 and 120 min). It was also decreased significantly below baseline from 60 to 120 min with AM0.25 (P = 0.002 at 60 min and P , 0.0001 at 90 and 120 min) and with AM1.0 treatment (P = 0.004 at 60 min and P , 0.0001 at 90 and 120 min). Compared with ACP treatment the combination treatments resulted in significantly lower temperatures, at 90 min with AM0.5 (P = 0.004) and AM1.0 (P = 0.0001) and at 120 min with

Table I. Mean values [6 standard deviation (SD)] for heart rate (HR), systolic, mean, and diastolic arterial pressures (SAP, MAP, and DAP), respiratory rate (RR), and rectal temperature (Temp.) in 5 clinically healthy dogs before (Baseline) and after intramuscular administration of 0.05 mg/kg body weight (BW) of acepromazine alone (ACP) or in combination with morphine, 0.25 mg/kg BW (AM0.25), 0.5 mg/kg BW (AM0.5), or 1.0 mg/kg BW (AM1.0).

Time (min) after drug administrationParameter and treatment Baseline 15 30 45 60 90 120HR (beats/min) ACP 73 6 13 86 6 23 84 6 24 85 6 25 89 6 22a 88 6 25a 84 6 23 AM0.25 73 6 18 80 6 16 72 6 16 70 6 19 67 6 21 63 6 16 67 6 19 AM0.5 79 6 9 82 6 12 73 6 10 72 6 14 69 6 16 67 6 13a 74 6 20 AM1.0 76 6 10 88 6 17 75 6 12 72 6 13 69 6 11 69 6 16 69 6 14

SAP (mmHg) ACP 142 6 22 127 6 24a 126 6 31a 130 6 35 129 6 28 127 6 24a 134 6 22 AM0.25 149 6 10 140 6 19 133 6 17a 127 6 17a 129 6 16a 129 6 20a 135 6 17a

AM0.5 150 6 14 131 6 18 126 6 17 124 6 15a 127 6 13 141 6 6 139 6 14 AM1.0 147 6 12 132 6 17a 125 6 13a 120 6 11a 122 6 16a 129 6 16a 131 6 17a

MAP (mmHg) ACP 85 6 5 75 6 6a 74 6 11a 78 6 13 78 6 11 78 6 10 82 6 8 AM0.25 87 6 2 84 6 8 78 6 6a 74 6 7a 75 6 6a 77 6 7a 82 6 6 AM0.5 86 6 7 76 6 7a 70 6 6a 70 6 8a 71 6 5a 79 6 9 83 6 10 AM1.0 86 6 5 80 6 7 72 6 5a 71 6 4a 75 6 12a 76 6 9a 78 6 11

DAP (mmHg) ACP 61 6 6 55 6 4 54 6 5 58 6 9 58 6 9 57 6 6 61 6 7 AM0.25 64 6 5 62 6 6 58 6 3 55 6 4a 55 6 2a 58 6 4 61 6 2 AM0.5 60 6 6 55 6 7 51 6 4 51 6 6a 50 6 4a 57 6 6 60 6 12 AM1.0 61 6 4 59 6 5 54 6 3 53 6 5 54 6 6 57 6 8 60 6 14

RR (breaths/min) ACP 21 6 2 18 6 3 18 6 3 16 6 3a 15 6 2a 16 6 3a 16 6 3a

AM0.25 22 6 10 21 6 7 18 6 5 17 6 5 17 6 3 17 6 4 18 6 5 AM0.5 30 6 20 20 6 8 18 6 5 16 6 5a 18 6 6 18 6 6 15 6 5a

AM1.0 27 6 12 23 6 3 24 6 2 22 6 2 22 6 5 21 6 6 18 6 4a

Temp. (°C) ACP 38.3 6 0.4 — 38.0 6 0.3 — 38.0 6 0.4 38.0 6 0.7 38.1 6 0.7 AM0.25 38.3 6 0.5 — 38.0 6 0.4 — 37.7 6 0.3a 37.4 6 0.3a 37.3 6 0.2a,b

AM0.5 38.6 6 0.3 — 38.0 6 0.4a — 37.5 6 0.5a 37.2 6 0.7a,b 37.0 6 0.8a,b

AM1.0 38.4 6 0.3 — 38.0 6 0.4 — 37.5 6 0.3a 36.9 6 0.6a,b 36.7 6 0.8a,b

a Significantly different from baseline.b Significantly different from the value with ACP treatment (P , 0.05).



AM0.25 (P = 0.0073), AM0.5 (P = 0.0001), and AM1.0 (P , 0.0001). No significant difference in temperature was detected among the com-bination treatments. The maximum decreases in mean temperature were 0.2°C, 1.0°C, 1.6°C, and 1.7°C with ACP, AM0.25, AM0.5, and AM1.0, respectively.

Administration of all treatments increased the sedation scores above the baseline values (Table III). With ACP treatment the increases were observed only with the VAS (P = 0.006). Compared with the baseline values the VAS scores with AM0.25 treatment were higher at 30 (P = 0.008), 45 (P = 0.006), and 60 (P = 0.008) min, and the NDS scores were higher at 15 (P = 0.016), 30 (P = 0.016), and 45 (P = 0.016) min. With AM0.5 treatment the scores were higher than at baseline at 30 min (P = 0.0009 and P = 0.0046 for VAS and NDS, respectively) and 45 min (P = 0.0015 and P = 0.0046 for VAS and NDS, respectively). With AM1.0 treatment the scores were higher than at baseline at 30, 45, and 60 min (P = 0.0407, P = 0.0099, and P = 0.0060 for VAS and P = 0.0206, P = 0.0206, and P = 0.0206 for NDS, respectively). Compared with ACP treatment the AM0.25 (P = 0.042) and AM1.0 (P = 0.009) treatments, but not the AM0.5 treatment, resulted in higher VAS scores. However, the NDS scores were significantly higher only for AM1.0 treatment (P = 0.020) in comparison with ACP treatment. There was no significant difference

in sedation scores among the AM0.25, AM0.5, and AM1.0 treatments at any time point during the observation period.

In the present study AM0.25 and AM1.0 treatment improved the sedation induced by acepromazine in dogs. Although no significant difference was observed in sedation among the 3 morphine doses, the sedation scores were 36% to 66% higher with AM1.0 treatment than with AM0.25 and AM0.5 treatment at 30 min.

The overall results of this study agree with previous reports that combinations of acepromazine with opioid analgesics result in greater sedation than acepromazine alone in dogs (3–5,8). However, the present study did not identify a dose–response relationship between morphine dose and degree of sedation induced by the combination. Similarly, in a recent study (5) the combination of acepromazine (0.05 mg/kg BW administered intramuscularly) and incremental doses of methadone (0.25, 0.50, and 0.75 mg/kg BW) was not associated with significant improvement in sedation. Nevertheless, in this previous study the higher methadone doses resulted in prolonged sedation compared with the lower doses, which is different from the results in our study, in which the sedation scores were significantly higher than at baseline until 45 to 60 min for the AM0.25, AM0.5, and AM1.0 treatments. At 120 min the dogs still presented some degree of sedation with all the treatments, but no significant difference in sedation scores compared with baseline or among the combination groups was detected.

These results must be interpreted with caution considering a major limitation. According to the preliminary sample size calculation, 6 dogs per group would be necessary to detect a 3.0-cm difference in VAS scores with adequate statistical power. At peak sedation (30 min) the maximum difference in VAS scores between the AM1.0, AM0.5, and AM0.25 treatments was 2.1 cm. At 60 min the scores with the AM1.0 and AM0.5 treatments were 7.7 and 4.7 cm, respectively. Besides a 3.0-cm difference in VAS score, no significant difference was detected, which might indicate that the study was underpow-ered because the sample size was only 5 dogs. Another point to consider is that the VAS and NDS scores were significantly higher with the AM1.0 treatment than with the ACP treatment at several time points, but most of the time such differences were not detected between the ACP treatment and the AM0.25 and AM0.5 treatments. These results may indicate that sedation was greater with the AM1.0 treatment and that differences from the AM0.25 and AM0.5 treatments were not detected because of the small sample size. Therefore, a study with a larger sample may yield different results. Although this study might lack the power to support the absence of a signifi-cant difference in sedation between the highest morphine dose and the 2 other doses of the opioid, our results provide evidence that increasing the dose of morphine from 0.25 to 0.50 mg/kg BW in the combination does not improved sedation: the scores determined with both sedation scales were nearly the same with AM0.25 and AM0.50 treatment throughout the observation period.

The cardiovascular effects observed in this study are typical of combinations of acepromazine with opioids in dogs (3–5,8). The heart rate decreased by 15% and the SAP and MAP by 15% to 19%, respectively, compared with baseline values but were still within the clinically acceptable range for dogs. The decrease in arterial blood pressure resulted from acepromazine-induced blockage of a-adrenergic receptors in the vascular beds, as reported in a previous

Table II. Mean values (6 SD) for arterial pH, carbon dioxide partial pressure (PaCO2), oxygen partial pressure (PaO2), and bicarbonate concentration (HCO3

2) in the 5 dogs before (Baseline) and 45 min after drug administration.

45 min after drug Parameter and treatment Baseline administrationpH ACP 7.39 6 0.02 7.39 6 0.02 AM0.25 7.37 6 0.03 7.36 6 0.03 AM0.5 7.38 6 0.03 7.35 6 0.03a,b

AM1.0 7.39 6 0.02 7.33 6 0.02a,b

PaCO2 (mmHg) ACP 34 6 4 34 6 2 AM0.25 34 6 4 36 6 2 AM0.5 34 6 2 39 6 2a

AM1.0 33 6 3 40 6 3a,b

PaO2 (mmHg) ACP 91 6 3 91 6 6 AM0.25 84 6 8 89 6 4 AM0.5 89 6 2 86 6 1a

AM1.0 88 6 4 86 6 5

HCO32 (mEq/L)

ACP 20.9 6 2.2 20.5 6 1.8 AM0.25 19.9 6 1.8 20.3 6 1.6 AM0.5 20.3 6 1.4 21.3 6 1.4a

AM1.0 20.0 6 1.0 21.1 6 0.8a Significantly different from baseline.b Significantly different from the value with ACP treatment (P , 0.05).



study in dogs (9). However, decreases in myocardial contractility and cardiac output have also been reported as a cause of the decrease in arterial pressure in dogs administered 0.1 mg/kg BW of aceproma-zine intravenously (8). The decrease in heart rate with the AM0.25, AM0.5, and AM1.0 treatments is likely associated with a centrally mediated effect induced by the opioid that results in an increase in vagal tone (10). The reduction in heart rate was not influenced by the morphine dose, as was the case in another study in dogs admin-istered acepromazine–methadone combinations (5). However, our dogs were well-conditioned to the laboratory environment and had low baseline heart rates. In stressed dogs with a high heart rate, one might expect decreases in the rate greater than the 15% reduction observed herein.

In this study, arterial samples were collected for blood gas analy-sis 45 min after administration of the drugs. The moment for blood sampling was chosen in order to evaluate blood gases at the peak sedative effect. According to previous studies the peak sedative effect after intramuscular administration of acepromazine– opioid combinations occurs within 30 to 45 min after injection (3,5). Compared with baseline values, the respiratory rate decreased with all treatments except AM0.25. The reduction might be attributable to a calming effect of the acepromazine. Another reason could be respiratory depression induced by morphine. Opioid agonists at m-receptors can decrease minute volume and increase PaCO2 (11). Despite the significant decreases in respiratory rate found in this study, the influence on blood gases was minimal and not of clinical relevance.

A reduction in body temperature among dogs given aceproma-zine was reported in 1 study (3) but not observed in another (5). In the present study, body temperature was not decreased in the dogs given acepromazine alone during the 120-minute observation period. Conversely, administration of morphine and methadone alone resulted in temperature decreases in unanesthetized dogs (1). Moreover, the decrease was similar in dogs given methadone alone (0.5 mg/kg BW) or in combination with acepromazine (0.05 mg/kg BW) (3). These findings indicate that the opioid is responsible for the

drop in body temperature in dogs given an acepromazine–opioid combination.

In conclusion, this study has provided evidence that increasing the morphine dose in a combination with acepromazine from 0.25 to 0.5 mg/kg BW does not improve sedation, but increasing the dose to 1.0 mg/kg BW might result in greater sedation than with the lower doses of the opioid. Studies with a greater number of samples are warranted to confirm this statement.

Re f e r e n c e s1. Maiante AA, Teixeira Neto FJ, Corrente JE, Pedroso CEBP.

Comparison of the cardio-respiratory effects of methadone and morphine in conscious dogs. J Vet Pharmacol Ther 2008;32: 317–328.

2. Hall LW, Clarke KW, Trim CM. Principles of sedation, analgesia and premedication. In: Veterinary Anaesthesia. 10th ed. London, England: WB Saunders, 2001:75–112.

3. Monteiro ER, Figueroa CDN, Choma JC, Campagnol D, Bettini CM. Effects of methadone, alone or in combination with acepromazine or xylazine, on sedation and physiologic values in dogs. Vet Anaesth Analg 2008;35:519–527.

4. Monteiro ER, Junior AR, Assis HMQ, Campagnol D, Quitzan JG. Comparative study on the sedative effects of morphine, metha-done, butorphanol or tramadol, in combination with aceproma-zine, in dogs. Vet Anaesth Analg 2009;36:25–33.

5. Bitti FS, Campagnol D, Rangel JP, Nunes Junior JS, Loureiro B, Monteiro ER. Effects of three methadone doses combined with acepromazine on sedation and some cardiopulmonary variables in dogs. Vet Anaesth Analg 2017;44:237–245.

6. Monteiro ER, Lobo RB, Nunes Junior JS, Rangel JPP, Bitti FS. Tramadol does not enhance sedation induced by acepromazine in dogs. Can J Vet Res 2016;80:323–328.

7. Buback JL, Boothe HW, Carroll GL, Green RW. Comparison of three methods for relief of pain after ear canal ablation in dogs. Vet Surg 1996;25:380–385.

Table III. Median (and interquartile range) of sedation scores measured on 2 scales in the 5 dogs before and after drug administration.

Time (min) after drug administrationScore and treatment Baseline 15 30 45 60 90 120VAS (cm) ACP 0 1.3 (0.8–3.0) 2.5 (1.0–3.0)a 2.0 (1.0–3.1)a 2.0 (1.0–3.4)a 2.1 (0.6–3.0) 1.2 (0.3–2.8) AM0.25 0 5.8 (4.4–7.7)b 5.8 (4.6–8.3)a 5.1 (4.3–7.8)a 4.9 (4.3–8.4)a,b 5.0 (3.3–7.8) 3.8 (2.5–7.6) AM0.5 0 5.0 (4.4–6.0) 5.7 (4.8–8.1)a 5.7 (4.9–7.7)a 4.7 (3.7–7.1) 4.8 (3.7–7.2) 4.7 (3.2–6.9) AM1.0 0 6.0 (5.7–8.1)b 7.9 (6.9–8.4)a,b 7.3 (5.8–9.4)a,b 7.7 (4.8–9.5)a,b 7.5 (4.5–9.4)b 6.5 (4.0–9.1)b

NDS ACP 1.0 (1.0–1.0) 2.0 (2.0–2.0) 2.0 (2.0–2.0) 2.0 (2.0–2.0) 2.0 (2.0–2.0) 2.0 (2.0–2.0) 2.0 (1.5–2.0) AM0.25 1.0 (1.0–1.0) 3.0 (3.0–5.0)a 3.0 (3.0–5.0)a 3.0 (3.0–5.0)a 3.0 (2.5–5.0) 3.0 (2.0–5.0) 2.0 (2.0–5.0) AM0.5 1.0 (1.0–1.0) 3.0 (3.0–4.0) 3.0 (3.0–5.5)a 3.0 (3.0–5.5)a 3.0 (2.5–4.5) 3.0 (2.5–4.0) 3.0 (2.0–4.0) AM1.0 1.0 (1.0–1.0) 3.0 (3.0–5.0)b 5.0 (3.5–5.5)a,b 4.0 (3.0–6.0)a,b 4.0 (3.0–6.0)a,b 3.0 (3.0–6.0)b 3.0 (2.5–6.0)a Significantly different from baseline.b Significantly different from the value with ACP treatment (P , 0.05).VAS — Visual analogue scale (range: 0 to 10 cm); NDS — Numeric descriptive scale (range: 0 to 6).



8. Stepien RL, Bonagura JD, Bednarski RM, Muir WW. Cardio-respiratory effects of acepromazine maleate and buprenor-phine hydrochloride in clinically normal dogs. Am J Vet Res 1995;56:78–84.

9. Ludders JW, Reitan JA, Martucci R, Fung DL, Steffey EP. Blood pressure response to phenylephrine infusion in halothane-anesthetized dogs given acetylpromazine maleate. Am J Vet Res 1983;44:996–999.

10. Reitan J, Stengert K, Wymore M, Martucci RW. Central vagal control of fentanyl-induced bradycardia during halothane anes-thesia. Anesth Analg 1978;57:31–36.

11. Kukanich B, Wiese AJ. Opioids. In: Grimm KA, Lamont LA, Tranquilli WJ, Greene SA, Robertson SA, eds. Veterinary Anesthe-sia and Analgesia. Ames, Iowa: Wiley Blackwell, 2015:207–226.


Supplies and services Produits et services

RENOUVELEZ VOTRE ABONNEMENT

DÈS MAINTENANT!

RENEW YOUR SUBSCRIPTION

NOW!

Four issues of the Canadian Journal of Veterinary Research for 2019 (online)

Quatre numéros de la Revue canadienne de recherche vétérinaire pour 2019

Subscription/Abonnement 2019Personal Subscription/ $180 $ (1 GST/HST as Abonnement personnel applicable; TPS/TVH en (Canada) vigueur)

Personal Subscription/ $200 $ US/É.-U. Abonnement personnel (Foreign/Étranger)

Institutional Subscription/ $250 $ (1 In Canada, GST/HST Abonnement institutionnel as applicable; Au Canada, TPS/TVH selon le cas)

Institutional Subscription/ $250 $ US/É.-U. Abonnement institutionnel (Foreign/Étranger)

Contact/Communiquez avec : Kelly Gray-Sabourin Editorial Coordinator/Coordonnatrice de la rédactionCanadian Journal of Veterinary ResearchRevue canadienne de recherche vétérinaire339 rue Booth Street Ottawa, Ontario K1R 7K1Canada

Tel./Tél. : (613) 236-1162, x117Fax/Télécopieur : (613) 236-9681E-mail/Courriel : [email protected]

w w w . c a n a d i a n v e t e r i n a r i a n s . n e t

w w w . v e t e r i n a i r e s a u c a n a d a . n e t

Tables of contentsComing events

Article abstractsClassified advertising

Subscription informationInstructions for authors

Tables des matières Événements à venirRésumés d’articles Annonces classées Information sur l’abonnement Directives à l’intention des auteurs

Canadian Journal of Veterinary Research

available online now!

Articles are now easily accessible over the Internet at

PubMed Centralwww.pubmedcentral.com

Reaching Canada’s VeterinariansGet your message into

The Canadian Veterinary JournalFor more information contact:Laima LaffitteAdvertising and Sponsorship ConsultantTel.: (613) 673-2659Fax: (613) 673-2462e-mail: [email protected]


CANADIAN JOURNAL OF VETERINARY RESEARCH / REVUE CANADIENNE DE RECHERCHE VÉTÉRINAIRE


provided for non-commercial research and education use ... · briana hagen, guelph, ontario sandra...

Documents