Journal of Periodontal Research 16: 372-378, 1981

Predictability of probing depthsin ligature-induced periodontal defects

in beagle dogs

JAN JANSEN AND TACO PILOT

Department of Periodontology, Faculty of DentistryUniversity of Groningen, The Netherlands

The aim of the present study was to evaluate the predictability of a method of creatingperiodontal defects in dogs. The dentogingival fibres around the lower premolars of 12beagle dogs were cut to the level of the alveolar bone and closely adapted copper bandswere cemented to the teeth. After 3 weeks the copper bands were replaced by cottonligatures which were left in place for 11 weeks. After this active phase the ligatures wereremoved, and the defects were allowed to heal for varying periods up to 30 weeks. Probingdepths were recorded at different stages of the experiment. Indications for predictabilitywere: 1° the magnitude of the standard deviations of the mean probing depths and 2° thedifferences between results from contralateral areas. Relatively deep defects were obtainedimmediately after creation, but the magnitude of the standard deviation of the meanindicated wide variations in probing depths. Following ligature removal probing depthsdecreased but were of a more uniform depth. These results indicate that variation ofprobing depths in artificial periodontal defects is rather high immediately after creationbut decreases drastically during the healing phase. It was concluded that in young adultbeagle dogs predictable periodontal defects, 3.0-3.5 millimeters in depth, can be producedartificially provided that the initial deeper artificially created defects are allowed to healfor at least 3 months.

(Accepted for publication February 18, 1981)

Introduction

Artificially created periodontal defects inanimals have been used to test therapeuticprocedures for several decades. The ulti-mate aim of animal experiments is the re-production of the human situation. An ar-tificially created periodontal defect shouldtherefore resemble a human pocket asclosely as possible, clinically with regard toprobing depth and/or attachment level andhistologically with regard to epithelization

of the defect, composition of inflammatoryinfiltrate, loss of connective tissue attach-ment, loss of alveolar bone and changes inthe cementum.

Deep periodontal pockets respond dif-ferently to therapeutic and preventive pro-cedures than do shallow pockets or sulci.This has been shown in clinical trials(Knowles et al. 1979) and in animal studies(Lang et al. 1979). If an animal model isused to test periodontal procedures, it isnecessary to create pockets of distinguish-

P R E D I C T A B I L I T Y O F P R O B I N G D E P T H S 373

able depths. A problem in creating deeppockets is, however, that artificial periodon-tal defects tend, to some extent, to healspontaneously, resulting in shallower pock-ets than those required (Hugoson &Schmidt 1978, Jansen et al. 1979, Van Dijk1979).

A further prerequisite when using an ex-perimental model is that it must be re-producible and that the artificial periodon-tal defects should be entirely predictable.Little attention has as yet been paid tothis aspect (Lindhe, Hamp & Loe 1975,Ericsson et al. 1975, Caton & Kowalski1976).

The aim of this experiment, which ispart of a more extensive study on the be-havior of artificially created periodontaldefects in beagle dogs, was to evaluateprobing depths in ligature-induced perio-dontal defects. The results should shed morelight on the predictability of probing depthsand thus ensure more efficient use of theexperimental animal model.

Material and Methods

Twelve beagle dogs, approximately 18months old and weighing ± 15 kg wereused. Throughout the experiment the dogswere fed on a diet which favors gross plaqueaccumulation on the teeth, and no oralhygiene procedures were performed. Perio-dontal defects were created by using a tech-nique adopted from Van Dijk (1979) whomodified the method described by Lindheand Svanberg (1974). The dentogingivalfibres around the necks of the lower rightpremolars ^V, gP and 4P were cut to thelevel of the alveolar bone, and closelyadapted copper bands were cemented to theteeth with the use of copper cement. After3 weeks the copper bands were replaced bycotton ligatures which were forced downas deep as possible. These ligatures wereleft in place for 11 weeks. After this period

the defects were allowed to heal for periodsvarying up to 30 weeks.

Pairs of dogs were scheduled for sacrificeand subsequent histological preparation 0,2, 6, 14 and 30 weeks after removal of thecotton ligatures from the defects. The ex-perimental design is shown in Fig. 1.

Probing depths measurements were per-formed at baseline immediately after theremoval of the ligatures and 2, 6, 14 and 30weeks thereafter.

In this experiment probing depth wasdefined as: the extent of penetration intothe sulcus or pocket of a University ofMichigan 0 probe (terminal diameter 0.35mm) inserted with a non-standardized lighthand pressure until definite resistance wasfelt. All measurements were made to thenearest half millimeter. Only mesial anddistal surfaces of the experimental teethwere investigated.

The depths measured were divided intoclasses as follows: class I (probing depth< 2.5 millimeter) was chosen because itwas assumed to be within the range ofclinically normal depths after 1-1.5 year ofplaque accumulation (Lindhe, Hamp & Loe1973, 1975, Schroeder & Lindhe 1975).Therefore, pockets with probing depths ex-ceeding 2.5 millimeters were considered tobe deepened. A distinction was made be-tween moderately deep defects: (class n,2.5 < probing depth <C 5.0 millimeter) anddeep defects (class III probing depth > 5.0millimeter), because 5 millimeters is oftenconsidered to be the borderline between aconservative and surgical approach in perio-dontal therapy.

In order to investigate the predictabilityof making defects in the same dog, perio-dontal defects were created in 10 out of theoriginal 12 beagle dogs using the sametechnique at the contralateral (left) pre-molars 14 weeks before sacrifice, thus al-lowing left-right comparison at baseline andat the end of the active phase.

374 J A N S E N A N D P I L O T

Two parameters were used as being indi-cative of the predictability of probingdepths:1°. the magnitude of the standard devia-

tions of the mean probing depths and2°. differences in probing depths from con-

tralateral areas.Means of probing depths were comparedusing Student's t test for each area separatelyin order to avoid the potential problemof dependency of observations (Pilot et al.1980).

Results

The frequency distribution of probing depthsin classes, mean probing depths and stan-dard deviations are presented in Fig. 2for all defects after different periods oftime. Baseline measurements were dom-inated by class I defects. During the ac-tive phase of the experiment class II and IIIdefects predominated and at the end of thisphase deep defects (class III) constituteda large proportion of those measured. Inthe healing phase, however, class II defectsconstituted more than 80 % of the registra-tions at all stages. After 14 weeks defects

that did not fall within the class II rangewere exceptional.

Mean probing depths and standard devia-tions for separate areas are presented inTable 1. After a considerable increase inmean probing depth during the creationphase a rapid decrease was noted duringthe healing phase, especially during the firsttwo weeks after elimination of the ligatures.

At baseline standard deviations of themean were approximately 0.50 millimeterin all areas (Table 1). At the end of the ac-tive phase, deviations in probing depthwere often two to three times as high.,whilst in the following period the standarddeviations returned to baseline levels oreven lower in most instances.

It can be seen from Fig. 2 that thefrequency distribution of classes of probingdepths at baseline and at the end of theactive phase were almost the same in theleft and right quadrants. Two dogs did not,however, follow this general pattern. In dognumber 9 defects were created on the rightside of 7 and 8 millimeters depth whilst onthe contralateral side no such exceptionaldepths were recorded. In dog number 5similar but less pronounced differences

_ active phase of creating defects _ .healing phase.

baseline

weeks14

14weeks

30

AfibrotomyCu-bandsP p p

ACu-bandsreplacedby cottonligatures

eliminationof cotton

ligatures from thedefects

Nn

1272

1272

636

636

424

212

Fig. 1. Design of the experiment for the lower right premoiars. N = number of dogs available for probingdepth registration at different stages, n represents the total number of defects.

Around contralateral teeth of 10 dogs defects were created but no healing phase was included (see text).

P R E D I C T A B I L I T Y O F P R O B I N G D E P T H S 375

x = 2 . 5 7 + 0 . 4 7 x = 4 . 0 5n = 7 2 n = 7 2

96

x = 3 . 8 9 + 0 . 5 7 x = 3 . 3 7 + 0 . 6 6 x = 3 . 3 3 + 0 . 5 0 x = 3 . 2 0 + 0 . 2 9n = 3 b n = 3 5 n = 2 4 n = 1 2

base line end active phase 2 weeks 6 weeks 14weeks 30weeks

x = 2 . 6 1 + 0 . 4 4n = 6 0

22

x = 3 . 8 7 + 1 . 0 9n = 6 0

••class^ M class1 Iclass

1IIIII

P 52.5

P 5

2.5< p<5.0mnn

5.0

Fig. 2. Frequency distribution of classes of probing depths at different stages of the experiment for defects inthe lower right premoiar region (upper bars) and in the lower left premolar region (lower bars).

Mean probing depths (x) and standard deviations are presented at the bottom of each bar for all defects(n) available for registration in that phase of the experiment.

were also found. A comparison betweenthe mean probing depths of separate areasin the left and right jaw in this group ofdogs revealed no significant differences inany area (Table 1).

Discussion

In this study the probing depth of a perio-dontal defect was chosen as the clinicalparameter, because it was felt that thedepth of the defect is more relevant thanthe loss of periodontal attachment when atreatment procedure is tested in a model.

Most authors producing experimentalperiodontal breakdown in dogs recordshallow pockets (Ericsson et al. 1975,Schroeder & Lindhe 1975, Lindhe & Erics-son 1978, Hugoson & Schmidt 1978). Liga-tures, in particular placed coronally to thegingival margin, often produce considerablerecession, leading to extensive loss of at-tachment but extremely shallow pockets(Lindhe & Ericsson 1978). The fact thatthe cotton ligatures were situated subgin-givally in this experiment, immediately afterplacement as well as at the end of the activephase of creating the defects, probably ex-

376 J A N S E N A N D PI L O T

plains why relatively deep pockets were ob-tained in a period of 14 weeks.

At baseline standard deviations were ap-proximately 0,5 millimeters thus demon-strating very little variation between ani-mals. At the end of the active phase stan-dard deviations varied from 0,8 to 1,6 milli-meters. This means that the degree of varia-tion was high at this stage. The results of3 dogs with exceptional reactions did, how-ever, contribute to these high values. In onedog it was impossible to create defects ofany considerable depth because of the gin-gival recession, and in two dogs very deepdefects of 7 and 8 millimeters were mea-sured on only one side whilst the contra-lateral side reacted "normally,"

In spite of these individual variations, nostatistically significant differences (P<;0,05)were found in any area between left and

right, which provides some support forpredictability at this stage, (Table 1).

During the healing phase predictabilitywas estimated through the values of thestandard deviations, since the present ex-perimental design did not allow for left-right comparison during this phase. Follow-ing ligature removal standard deviations ofmean probing depths in separate areas de-creased gradually to baseline levels or evenlower, indicating that aberrant high valuesregressed to the mean.

The observation that, following ligatureremoval, mean probing depths decreasedsuggests a gain in periodontal attachment,but can also be explained by probing er-rors due to differences in the inflammatorystate of the tissues (Armitage et al, 1977),Recently Jansen et al, (1981), in evaluatingprobing errors in experimental periodontal

Table 1

Mean probing depths (x) for separate areas and standard deviations (s,d.) of the defectsat different stages of the experiment,

right side of the jaw

phase ofthe experiment

baseline

end active phase

2 weeks of healing

6 weeks of healing

14 weeks of healing

30 weeks of healing

left side of the jaw

base line

end active phase

num-berof

dogs

12

12

6

6

4

2

10

10

disi

X +

2,92 +

4 ,75 +

4 ,25 +

3.6b +

3.37 +

3,25_+

dist

2,85 +

4,45 +

4

s . d .

0,51

0,92

0,42

0.40

0.48

0.35

0,47

0.79

Pmes.

X +

2,50 +

3,66 +

3,50 +

3.33 +

3,12 +

3,00 +

s

0,

0,

0,

0 .

0 .

0 .

mes.

2,80 +

4 ,20 +

0

1

d .

64

86

43

60

25

00

,42

,13

3dist.

X +

2,45 +

3.92 +

3,25 +

3,50 +

3.62 +

3.50 +

dist2.50-t

3 . 5 5 ^

s . d .

0 ,50

1,60

0,41

0,70

0,47

0.00

0,47

0,98

Pmes.

X

2,50

4,04

3,00

3,25

3.37

3.25

+

+

+

+

+

+

+

s . d .

0,37

1 .27

0.63

0 ,41

0,48

0,35

mes.

2,7C

3.65

+

+

0,48

- 0 , 8 5

dist.

X + S

2,50 + 0,

3,62 + 1

2.83 + 0

3 . 0 0 + 0

3,25 + 0

3 , 2 5 + 1 ,

dist.

2.40 + 0

3,45 + 0

2

, d .

37

00

75

70

86

06

,4b

,85

P

2 ,

4 ,

3,

3,

3,

3,

2

3

mes.

X + S

5 0 + 0

33 + 1

58 + 0

50 + 1

25 + 0

00 + 0,

mes..55 + 0

.95 + 1

, d .

i l

11

80

IB

50

00

,28

,01

P R E D I C T A B I L I T Y O F P R O B I N G D E P T H S 377

defects in beagle dogs, demonstrated thatin defects at the end of the creation phase,the probe went past the apical terminationof the junctional epithelium; in the healingphase the probe tip on average indicated theconnective tissue attachment level correctly.This would suggest that probing depthmeasurements obtained at different timepoints during the present experiment shouldbe interpreted with care.

The ultimate values for mean probingdepths varied from 3.0 to 3.5 millimetersafter 30 weeks of healing (Table 1). Lindheet al. (1975) demonstrated in beagle dogsthat accumulation of plaque for a period of48 months resulted in pockets with the samemean probing depth (3.4 millimeter). How-ever, they reported that the magnitude ofthe standard deviation of the mean showedthat "the rate of periodontal tissue break-down varies considerably from one dog tothe other and from one experimental periodto the next." Their standard deviation forprobing depth after 48 months of plaqueaccumulation was comparable to our stan-dard deviations in the present study at theend of the active phase but was muchhigher than the standard deviations after14 and 30 weeks of healing.

Hull, Soames and Davies (1974), investi-gating naturally occurring periodontal dis-ease in a beagle dog colony aged between1 and 8 years, found that "the amount ofbone loss in beagle dogs increased with age,but to a degree which varied considerably,despite the fact that all dogs were main-tained under identical conditions and werefed identical diets throughout their lives."Recently Sorensen et al. (1980) reported onnaturally occurring periodontal disease in across sectional study on seventy-four beagledogs from one to twelve years of age. Theyfound only a slight to moderate increase inpocket depth with age but even in theoldest age group mean pocket depth didnot exceed 3.5 millimeters. This increase in

probing depth was not found in all areas.These studies indicate that the accumula-tion of plaque alone without interventionof e.g. ligature placement is neither a quicknor a reliable method of obtaining pocketsof a predictable depth in dogs. The experi-mental periodontal defects in the presentstudy were, after a period of healing, atleast as deep as plaque induced periodontaldefects after 4 years but showed consider-ably less variation in depth.

From the results of the present experi-ment in young adult beagle dogs, it can beconcluded that predictable periodontal de-fects, 3-3.5 millimeters in depth, can beproduced provided that the initial artificial-ly created defects are allowed to heal forat least 3 months.

References

Armitage, G. C, Svanberg, G. K. & Loe, H.1977. Microscopic evaluation of clinicalmeasurement of connective tissue attachmentlevel. Journal of Clinical Periodontology 4:173-190.

Caton, J. G. & Kowalski, C. J. 1976. Primatemodel for testing periodontal treatment pro-cedures. II Production of contralaterally sim-ilar lesions. Journal of Periodontology 47:506-510.

van Dijk, L. J. 1979. Surgery and oral hygieneon artificial periodontal defects in beagledogs. University of Groningen, Thesis.

Ericsson, I., Lindhe, J., Rylander, H. & Oka-moto, H. 1975. Experimental periodontalbreakdown in the dog. Scandinavian Journalof Dental Research 83: 189-192.

Hugoson, A. & Schmidt, G. 1978. Influence ofplaque control on the healing of experimen-tally induced bone defects in the dog. Journalof Periodontology 49: 135-141.

Hull, P. S., Soames, J. V. & Davies, R. M.1974. Periodontal disease in a beagle dogcolony. Journal of Comparative Pathologyand Therapeutics 84: 143-150.

Jansen, J., Corba, N. H. C , Pilot, T. & Jan-sen, H. W. B. 1979. Spontaneous regenera-tion of artificially created periodontal defectsin beagle dogs. Journal of Dental Research58: Special Issue A 591.

378 J A N S E N A N D PI L O T

Jansen, J., Pilot, T. & Corba, N. H. C. 1981.Histologic evaluation of probe penetrationduring clinical assessment of periodontal at-tachment levels. Journal of Clinical Perio-dontology In press.

Knowles, J. W., Burgett, F. G., Nissle, R. R.,Shick, R. A., Morrison, E. C & Ramfjord,S. P. 1979. Results of periodontal treatmentrelated to pocket depth and attachment level.Eight years. Journal of Periodontology 50:225-233.

Lang, N. P., Morrison, E. C, Loe, H. & Ram-fjord, S. P. 1979. Longitudinal therapeuticeffects on the periodontal attachment leveland pocket depth in beagle dogs. Journal ofPeriodontal Research 14: 418^27.

Lindhe, J. & Ericsson, L 1978. Effect of liga-ture placement and dental plaque on perio-dontal tissue breakdown in the dog. Journalof Periodontology 49: 343-350.

Lindhe, J., Hamp, S. E. & Loe, H. 1973. Ex-perimental periodontitis in the beagle dog.Journal of Periodontal Research 8: 1-10.

Lindhe, J. & Svanberg, G. 1974. Influence oftrauma from occlusion on progression of

experimental periodontitis in the beagle dog.Journal of Clinical Periodontology 1: 3-14.

Lindhe, J., Hamp, S. E. & Loe, H. 1975.Plaque induced periodontal disease in beagledogs. Journal of Periodontal Research 10:243-255.

Pilot, T., Van der Weele, L. T., van Dijk, L. J.& Jansen, J. 1980. Variations in artificialperiodontal defects in beagle dogs. Journalof Dental Research 59: Si>ecial Issue B 373.

Schroeder, H. E. & Lindhe, J. 1975. Conver-sion of stable established gingivitis in the doginto destructive periodontitis. Archives ofOral Biology 20: 775-782.

Sorensen, W. P., Loe, H. & Ramfjord, P. 1980.Periodontal disease in the beagle dog. Journalof Periodontal Research 15: 380-389.

Address:

Department of PeriodontologyUniversity of GroningenAnt. Deusinglaan 19713 A V GroningenThe Netherlands