Evaluation of bovine lactoferrin as a topical therapy forchemotherapy-induced mucositis in the golden Syrian hamster

Julie Clarke, Ben Edwards, Leanne Srpek, Geo� Regester*Women's and Children's Hospital, Child Health Research Institute, Inc., 72 King William Road, North Adelaide SA 5006, Australia

Received 20 May 1998; accepted 4 June 1998

Abstract

Bovine lactoferrin was applied topically to the oral mucosa of Syrian hamsters and assessed for its ability to decrease the severity ofchemotherapy-induced oral mucositis. Results indicated that the chemotherapy agent 5-¯uorouracil (5-FU) administered to hamsters on

days 0 and 2 produced severe leukopenia between days 4 and 7 of the trial, and that severity of oral mucositis coincided with the sup-pressed immune state in these animals. Bovine lactoferrin applied continuously to oral wounds in hamsters induced by a combination ofchemotherapy treatment and mild abrasion of the cheek pouch, failed to decrease the severity of mouth ulcers relative to a groupreceiving BSA as a control protein source. Hamster cheek pouches treated twice daily with lactoferrin had a signi®cantly worse condition

score between days 6 and 8, and days 12 and 13 ( p<0.05 to p<0.001), a higher ulcer score between days 6 and 15 ( p<0.05 top<0.001) and larger ulcer area between days 7 and 14 ( p<0.05 to p<0.001) compared to animals administered the control protein.Body weight changes between treatment and control groups showed no signi®cant di�erence over the trial period. In contrast to the pre-

study hypothesis, we report a detrimental e�ect from topical administration of bovine lactoferrin to the wounded oral mucosa ofimmunocompromised hamsters.# 1999 Published by Elsevier Science Ltd. All rights reserved.

Keywords: Lactoferrin; Mucositis; Ulceration; Therapy; Antimicrobial; Immune

1. Introduction

Oral mucositis is a major dose limiting toxicity ofchemotherapy for which there is no e�ective pre-ventative or treatment therapy. Currently 800,000patients receive chemotherapy in the United States eachyear while approximately 400,000 of these patientsdevelop oral complications associated with their treat-ment [1]. Chemotherapeutic dosage has been identi®edas a major predictor of treatment outcome in a varietyof cancer therapy programs, however, overcoming thechemotherapy dose-limiting toxicity ``mucositis'' hasbecome a major goal for oncologists. Thus to intensifycancer therapy, it will be necessary to reduce the asso-ciated toxicity arising from such indiscriminate drugtargeting of fast dividing cells lining the mouth andgastrointestinal tract.

Presently, there are no established treatments for oralmucositis, although administration of antibiotics andantiseptic mouthwashes have been useful to a limited

degree in controlling secondary infection [2,3]. Recently,biologically active proteins have been assessed for theiractivity in preventing and treating oral mucositis.Granulocyte macrophage-colony stimulating factor(GM-CSF) [4] and transforming growth factor-�3 [5]have been reported e�ective in reducing the severity oforal mucositis when administered before chemotherapyand in the mucosal repair phase. Clinical application ofthese agents may be limited however, due to costs asso-ciated with recombinant protein manufacture and pro-blems foreseen with gaining regulatory approval andpatient acceptance.

Bovine milk and colostrum are a rich source ofnatural anti-microbial factors, cytokines and proteingrowth factors with biological activity required forgrowth and development of the newborn and support ofthe gastrointestinal tract [7,8]. Bovine lactoferrin, aminor bioactive milk protein, has been demonstrated asan e�ective anti-microbial factor [9] through its ironbinding capacity, restricting levels of this essentialnutrient for microbial growth. More recently, lacto-ferrin peptides including lactoferricin have been demon-strated with potent antimicrobial activity, unrelated to

ORAL

ONCOLOGY

Oral Oncology 35 (1999) 197±202

1368-8375/99/$Ðsee front matter # 1999 Published by Elsevier Science Ltd. All rights reserved.

PII: S1368-8375(98)00087-6

* Corresponding author. Tel.: 0061 8 8204 7071; fax: 0061 8 8239

0267; e-mail: gregester@medicine.adelaide.edu.au

the iron binding potential of the molecule [10]. Lacto-ferrin has also been demonstrated as an e�ective growthfactor for cultured rat epithelial cells [11] and a med-iator of the in¯ammatory immune response [12]. Fur-thermore, a recent study has identi®ed that oraladministration of bovine lactoferrin improves intract-able stomatitis in cats positive and negative for felineimmunode®ciency virus [13].

These ®ndings suggest that topically administeredlactoferrin, a natural and abundant bovine milk protein,might act to enhance immune competence. It may alsosuppress the growth of opportunistic pathogens in themouth, which is a serious secondary complication oforal mucositis. The present study aims to measure thee�ectiveness of topically administered bovine lactoferrinin reducing the severity of mucosal damage in a hamstermodel of chemotherapy-induced oral mucositis.

2. Materials and methods

Manufacture of milk protein fraction: Bovine lacto-ferrin was purchased from a commercial supplier. Theprotein content of the powder was 85% (w/w) and itslevel of iron saturation was speci®ed at 68% (partiallysaturated). Bovine serum albumin (BSA) in dried form(96% protein w/w) was obtained from Sigma ChemicalCo. St Louis, USA. Samples were kept at 4�C andapplications were pre-weighed into eppendorf tubes.

Animal experimentation: All animal experimentationwas approved by the Animal Ethics Committee of theWomen's and Children's Hospital, Adelaide and fol-lowed the Australian Code of Practice for the Use ofAnimals in Research and Teaching (1990).

Oral mucositis model: Young castrated male Syriangolden hamsters (120±180 g) were purchased from theUniversity of Otago, New Zealand. They were cagedindividually in an animal room maintained at 22�2�Cwith a 10 h:14 h light: dark cycle. Throughout theexperimental period the hamsters had continual access

to water and commercial rodent diet (Ridley Agripro-ducts New Joint Stock Ration).

The animal model of mucositis was based on a mod-i®ed method of Sonis et al. (1990) [6]. An outline of thetrial protocol is shown in Fig. 1. To induce mucosalulceration, hamsters were administered two intraper-itoneal injections of 5-¯uorouracil (5-FU) containing 90and 60 mg/kg on days 0 and 2 (12.5 and 8.33 mg/mlrespectively; Delta West1) and cheek pouches wereeverted and lightly scratched with a small wire brush ondays 1 and 2. Hamsters were anaesthetised with iso-¯urane, and protein treatments applied to the left cheekpouch. These consisted of either lactoferrin or BSA inpowdered form, sprinkled across the ulcer from aninverted eppendorf tube. Applications of 10 mg weremade on days 0, 1 and 2; the amount was increased to15 mg from day 3 to ensure complete wound coverage.Each treatment group initially consisted of 14 animals;two hamsters from each group were either euthanasedor found dead (14% mortality rate).

Ulcers were assessed daily between days 5 and 15immediately prior to application of the protein. Assess-ment of the hamster cheek pouch included measuring thelength and width of the ulcer with callipers (mm), tracingthe edge of the wound for area determination, andphotographing the ulcer on the everted pouch. The con-dition of the cheek pouch mucosa surrounding the ulcerwas graded visually on a 0±3 scale for the level of bruis-ing, swelling, redness and scarring; these scores weretotalled to provide the condition score of the cheekpouch mucosa. The ulcer score was also graded visuallyon a 0±10 scale based on the severity and size of the ulcer.

Each animal was weighed daily and its well beingassessed using a score scale of 0±4. The well beingassessment included a grading of acute and chronicweight changes, behaviour, gait, posture and degree ofsubmandibular swelling. Hamsters with an averagescore exceeding 2.5 were culled (data not shown).

Mucositis model ± immunosuppression timecourse: Atimecourse study was undertaken with small groups of

Fig. 1. Schematic representation of the oral mucositis hamster trial procedures: continuous treatment.

198 J. Clarke et al./Oral Oncology 35 (1999) 197±202

hamsters to examine the level and sequence ofimmunosuppression following administration of thechemotherapeutic agent 5-FU. Two hamsters weresacri®ced daily between days 4 and 15 inclusive, andblood samples taken for complete blood counts, plateletcounts and di�erential determinations (VeterinaryPathology Services, Adelaide, South Australia).

3. Results

Immunosuppression time course: Haematologicalassessment showed that the dose of 5-FU administeredresulted in severe leukopenia by day 4 (Table 1). Thehamsters responded with a strong regenerative leukocy-tosis, which peaked at day 10. Many animals had acti-vated lymphocytes, indicating their immune systemswere responsive; several animals were thrombocyto-penic, had reduced pack cell volumes and red blood cellcounts, indicating they had undergone acute haemolytic

crises (data not presented). White cell counts (WCC)had returned to pre-trial levels at the day 14 evaluationtimepoint.

3.1. Mucositis model

Condition score: Fig. 2 provides a photographic com-parison of representative lactoferrin and BSA treatedcheek pouch mucosa on days 7 (peak mucositis) and attotal healing or day 15 at the end of the trial. The ulcersof these two hamsters were selected for comparison asthey were nearly identical on day 5. The lactoferrin

Table 1

White blood cell counts (WCC� 109/l) of hamsters in time course

study (n=2 at each time point)

day 0 day 4 day 8 day 10 day 12 day 14

WCC� 109/l 6.4 1.4 28.8 50.8 40.4 6.6

SE 1.0 0.4 12.9 24.0 27.3 2.7

Fig. 2. Cheek pouch ulcers of hamsters treated with lactoferrin (L7 and L15) or BSA (B7 and B14) on the day of peak mucositis (day 7), and when

healed (day 14 for BSA) or at the end of the trial (day 15 for lactoferrin).

J. Clarke et al./Oral Oncology 35 (1999) 197±202 199

treated ulcer at day 15 was still visible, whereas the BSAtreated lesion had completely healed by day 14.

The levels of bruising, swelling, redness and scarring(scale: 0 normal, to 3: severe for each parameter) of thehamster cheek pouches were assessed subjectively eachday during the timecourse. Results of this assessmentare presented in Fig. 3. Hamsters receiving lactoferrinshowed a signi®cantly higher score (range p<0.05 top<0.001 between days 6 and 9 and days 12 and 13),indicating a more severe oral condition in these animals.The peak di�erence between the groups occurred on day7 ( p<0.001) when the condition score of lactoferrintreated animals was double the control group.

Ulcer score and area: The daily ulcer scores and arearesults are presented in Fig. 4(a) and (b), respectively.The initial mean day 4 ulcer scores (10.0 and 9.7 forlactoferrin and BSA), and areas (83.3 and 86.1 mm2 forlactoferrin and BSA) were similar for both treatmentgroups, but diverged from day 5 onward. Hamstersreceiving lactoferrin showed signi®cantly higher ulcerscores between days 6 and 15 ( p<0.05 to p<0.001)and larger ulcer areas between days 6 and 14 ( p<0.05to p<0.001) compared to hamsters administered BSA.Maximum di�erences occurred on days 11 and 14 forulcer score ( p<0.001), and on day 7 for ulcer area( p<0.001). By day 15 of the trial, BSA treated ulcershad completely healed, while the lactoferrin groupremained above 10% of the day 5 score and area value.The inserted bar charts (Fig. 4(a) and (b)), represent theareas under the mean ulcer score curve and the meanulcer area curve between days 5 and 15 of the trial,showing signi®cantly greater plot areas in the lactoferrintreated group for both parameters ( p<0.005 andp<0.001 for ulcer score and area, respectively).

Bodyweights: The mean body weights for each groupthroughout the trial are presented in Fig. 5. There was

no signi®cant weight di�erence between treatmentgroups across days -1 to 15. A peak bodyweight loss of9% was recorded at day 8 for both treatment groups.The bodyweight of the hamsters was approximately95% of their starting ®gure by day 15.

4. Discussion

Lactoferrin is an iron binding glycoprotein ofapproximately 80 Kda with a wide range of biologicalfunctions ascribed to it, including the control of ironavailability and immune modulation [14]. The protein ispresent in milk and other body ¯uids of mammals suchas plasma, saliva, tears and secretions that bathe muco-sal surfaces, and is considered to be an important factorin the hosts defence of mucosal surfaces [15]. Lacto-ferrin is also synthesised by maturing granulocytes, andis stored in the secondary granule system of neutrophils[12] where it probably plays a role in host defence.

Fig. 3. Condition score of hamster cheek pouch mucosa from the

lactoferrin and control protein treated groups assessed between days

2±15 of the trial. Grading was between 0 (normal) and 5 (severe) con-

dition. Levels of signi®cance: *p<0.05; ^p<0.01; #p<0.005;

+p<0.001.

Fig. 4. Ulcer score (a) and area (b) values for lactoferrin and control

protein treated hamsters after chemotherapy (n=12 for both groups).

Bar charts (inserts) represent area under the curve between days 5 and

15 for each group. Levels of signi®cance: *p<0.05; ^p<0.01;

#p<0.005; +p<0.001.

200 J. Clarke et al./Oral Oncology 35 (1999) 197±202

Several recent studies have pointed to biologicalactivity from lactoferrin consistent with aiding repair ofmucosal epithelial injury [11] and in mediating thein¯ammatory response [12,16]. Relevant to our study,oral administration of bovine lactoferrin has beenshown to improve intractable stomatitis and con-currently improve the immune defence system in catswith and without feline immunode®ciency virus [13].This activity was partly attributed to lactoferrin medi-ated activation of neutrophil phagocytosis.

Other studies have reported related biological activ-ities for lactoferrin. The iron-saturated form of lacto-ferrin and lactoferrin peptide hydrolysates, promote thein vitro growth of lymphocytic [17] and intestinal epi-thelial cells [11]. The mechanism of this proliferativeaction is thought to be via active sites in the proteinmolecule which bind cell surface receptors and up-reg-ulate endogenous growth factors. In support of this, lac-toferrin binding sites at the surface of HT29 cells havebeen identi®ed, as has its ability to stimulate prolifera-tion of this cell line [18]. Further bene®cial activity fromtopically applied lactoferrin might be expected fromcontrolling cellular release of in¯ammatory cytokines.Indeed, antiin¯ammatory activity of lactoferrin has beenreported in recent studies which demonstrate its capacityto inhibit the release of interleukin-6, a pro-in¯amma-tory cytokine from HT-29 and U-937 cell lines [19].

The reported antimicrobial and immune regulatoryfunctions of bovine lactoferrin suggests potential bene-®ts from the molecule in reducing oral mucositis wheremucosal infection and associated leukopenia prevail.The link between low immune competence and mucosi-tis severity was supported in this study with oral condi-tion score coinciding with peak immunosuppressionbetween days 6 and 10 of the trial. However, we identi-®ed that lactoferrin signi®cantly decreased the rate ofulcer closure and worsened related measures of mucosal

condition and well being in the oral cavity of immune-compromised hamsters. This activity apparently con-tradicted the bene®cial e�ects reported from bovinelactoferrin in the feline model of oral stomatitis [13].

Reasons for the negative response to lactoferrin treat-ment in our mucositis model are largely speculative.Inconsistency with reports of bene®cial activity in thefeline model might relate to species di�erences in bacteriabetween the feline oral cavity and the hamster cheekpouch. Indeed, a di�erent micro¯ora is likely as thehamster cheek pouch is used for the transport of food andothermaterials. It is also known that the ironwithholdingprotective e�ects from lactoferrin are minimised againstsomemicrobes, which secrete iron chelators (siderphores)that compete for iron or express speci®c lactoferrinreceptors [16]. A further negative impact may have beenthe inability of lactoferrin administered to sequester freeiron; iron saturation levels of the topically applied proteinbeing outside a protective concentration range. In thisscenario, instead of restricting iron availability for bac-teria propagation, lactoferrin applied to the mucosalsurface with a high iron loading may provide a nutrientsource for pathogenic opportunistic bacteria throughlocalised iron release. Other workers have reported adelicate iron balance as critical to physiological functionsfrom lactoferrin. For example, lactoferrinmediated T cellproliferation was reported inhibitory at low ironconcentrations and stimulatory when excess iron wasavailable [16]. The ability of lactoferrin to protectlymphocyte function at the site of in¯ammation bysequestering potentially toxic iron arising from oral tissuedestruction might also be negated with insu�cient pro-tein in the apo-form. Inhibitory activity from lactoferrinhas also been demonstrated in cell culture experimentsmeasuring blastogenesis [21] and mitogen-induced pro-liferation of bovine mononuclear cells [20], and in thegrowth of mammary epithelial cells [22].

Alternatively, the di�erent response may have beendue to fundamental di�erences in the type of woundfound in the cats selected for the study [13], and theinduced wound in the hamsters. The cats had intract-able, chronic stomatitis, whereas the hamsters had rela-tively acute wounds. Acute wounds are characterised bya predominance of neutrophils, whereas macrophagesderived from monocytes form the basis of the chronicin¯ammatory reaction [20]. When stimulated by anti-gen, neutrophils release a range of bioactive productsincluding superoxide anion (O2ÿ) and lactoferrin. Lac-toferrin may enhance free radical production and rela-ted tissue damage when iron loaded, or protect againstfree radical damage when iron-free [23]. The additionof iron-loaded lactoferrin into an acute wound mayincrease the degree of tissue damage caused by endo-genous lactoferrin.

In conclusion, we report that topical application ofbovine lactoferrin to cheek pouch mucosal wounds in

Fig. 5. The body weights of hamsters treated with lactoferrin and

control protein.

J. Clarke et al./Oral Oncology 35 (1999) 197±202 201

hamsters increases lesion severity and reduces the over-all rate of wound healing in this experimental animalmodel. Further studies using the hamster model arewarranted, and could be directed toward investigatingthe activity of lactoferrin peptide hydrolysates identi®edwith non-iron related bacteriocidal activity, or alter-natively testing the apo (iron-free) form of lactoferrin asa more e�cient nutrient (iron) binding agent.

References

[1] Mueller BA, Millheim ET, Farrington EA, Brusko C, Wiser TH.

Mucositis management practices for hospitalized patients:

National survey results. J Pain and Symptom Management 1995;

10:510±18.

[2] Mueller BA, Millheim ET. Pharmaceutical aspects of mucositis

mouthwash mixtures. Amer J Health System Pharmacy 1995;

52:2596±7.

[3] Epstein JB, Stevenson-Moore P, Jackson S, Mohamed JH, Spi-

nelli JJ. Prevention of oral mucositis in radiation therapy: a con-

trolled study with benzydamine hydrochloride rinse. Internat J

Radiation Oncology Biology and Physics 1989;16:1571±5.

[4] Kannan V, Bapsy PP, Anantha N et al. E�cacy and safety of

granulocyte macrophage-colony stimulating factor (GM-CSF) on

the frequency and severity of radiation mucositis in patients with

head and neck carcinoma. Internat J Radiation Oncology Biol-

ogy and Physics 1997;37(5):1005±10.

[5] Sonis ST, Van Vugt AG, Brien JPO, Muska AD, Bruskin AM,

Rose A, Haley JD. Transforming growth factor-�3 mediated

modulation of cell cycling and attenuation of 5-¯uorouracil

induced oral mucositis. Oral Oncology 1977;33(1):47±54.

[6] Sonis ST, Tracey C, Shklar G, Jenson J, Florine D. An animal

model for mucositis induced by cancer chemotherapy. Oral Sur-

gery, Oral Medicine, Oral Pathology 1990;69:437±43.

[7] Donovan SM, Odle J. Growth factors in milk as mediators of

infant development. Ann Rev Nutrition 1994;14:147±67.

[8] Belford DA, Rogers M, Regester GO, Smithers GW, Liepe IJ,

Priebe IK, Ballard FJ. Milk-derived growth factors as serum

supplements for the growth of ®broblast and epithelium cells. In

Vitro Cellular and Developmental Biology 1995:31:752±60.

[9] Dionysius DA, Grieve PA, Milne JM. Forms of lactoferrin: their

antibacterial e�ect on enterotoxigenic Escherichia coli. J Dairy

Sci 1993;76:2597±606.

[10] WakabayashiH,BellamyW,TakaseM,Tomita,M. Inactivation of

Listeria monocytogenes by lactoferrin, a potent antimicrobial pep-

tide derived from cow's milk. J Food Protection 1992; 55:238±40.

[11] Hagiwara T, Shinoda I, Fukuwatari Y, Shimamura. E�ects of

lactoferrin and its peptides on proliferation of rat intestinal

epithelial cell line, IEC-18, in the presence of epidermal growth

factor. Bioscience Biotechnology Biochemistry 1995;59(10):1875±

81.

[12] Baynes RD, Bezwoda WR. Lactoferrin and the in¯ammatory

response. In: Hutchen TW et al., editor. Lactoferrin: Structure

and Function, vol. 357. Advances in Experimental Medicine and

Biology, New York: Plenum Press, 1994:133±41.

[13] Sato R, Inanami O, Tanaka Y, Takase M, Naito Y. Oral

administration of bovine lactoferrin for treatment of intractable

stomatitis in feline immunode®ciency virus (FIV)-positive

and FIV-negative cats. Amer J Veterinary Res 1996;57(10):1443±

6.

[14] Levay PF, Viljoen M. Lactoferrin: a general review. Haematolo-

gica 1995;80(3):252±67.

[15] Hutchens WT, Rumball SV, Lonnerdal B. Lactoferrin: Structure

and Function. In: Hutchen TW et al., editor. Lactoferrin: Struc-

ture and Function, vol.357. Advances in Experimental Medicine

and Biology, New York: Plenum Press, 1994.

[16] Brock, J. Lactoferrin: a multifunctional immunoregulatory pro-

tein? Immunology Today 1995;16(9):417±9.

[17] Hashizume S, Kuroda K, Murakami H. Identi®cation of lacto-

ferrin as an essential growth factor for human lymphocytic cell

lines in serum-free medium. Biochemica et Biophysica Acta

1983;763:377±82.

[18] Roiron D, Amouric M, Marvaldi J, Figarella C. Lactoferrin-

binding sites at the surface of HT29-D4 cells, comparison with

transferrin. European J Biochemistry 1989;186:367±73.

[19] Hanson LA, Mattsby-Baltzer I, Engberg I, Roseanu A,

Elverfors J, Motas C. Anti-in¯ammatory capacities of human

milk: lactoferrin and secretory IgA inhibit endotoxin-induced

cytokine release. In: Mestecky J et al., editors. Advances in

Mucosal Immunology. New York: Plenum Press, 1995:669±72.

[20] Torre PM, Oliver SP. Inhibition of bovine peripheral blastogen-

esis by fractional mammary secretion. Comparative Biochemistry

and Physiology B 1989;92:157±65.

[21] Rejman JJ, Payne KD, Lewis MJ, Torre PM, Muenchen RA,

Oliver SP. Identi®cation of whey proteins that in¯uence bovine

blood mononuclear cell proliferation. J Dairy Sci 1992;75 Suppl

308.

[22] Rejman JJ, Oliver SP, Muenchen RA, Turner JD. Proliferation

of the MAC-T bovine mammary epithelial cell line in the pre-

sence of mammary secretion whey proteins. Cell Biology Internat

Rep 1992;16:993±1001.

[23] Britigan BE, Edeker BL. Pseudomonas and neutrophil products

modify transferrin and lactoferrin to create conditions that favor

hydroxyl radical formation. J Clinical Investigation 88:1092±

102.

202 J. Clarke et al./Oral Oncology 35 (1999) 197±202