High Grade Prostatic Intraepithelial Neoplasia inMilitary Working Dogs With and Without

Prostate Cancer

Joseph W. Aquilina,1 LuAnn McKinney,2 Anna Pacelli,1 Laura K. Richman,3David J. Waters,4 Ian Thompson,5 Walter F. Burghardt Jr.,6 and

David G. Bostwick1*1Mayo Clinic, Rochester, Minnesota

2Armed Forces Institute of Pathology, Washington, DC3Registry of Veterinary Pathology, American Registry of Pathology, Washington, DC

4Cancer Biology Laboratory, Purdue University, West Lafayette, Indiana5Department of Surgery, Brooke Army Medical Center, Fort Sam Houston, Texas

6Department of Defense Military Working Dog Veterinary Services, Lackland Air Force Base, Texas

BACKGROUND. High grade prostatic intraepithelial neoplasia (PIN) is the most likely pre-cursor of human prostate cancer and is commonly found in men undergoing prostatic needlebiopsy for suspected cancer. Recent work has demonstrated that pet dogs, like humans,develop PIN spontaneously and in association with prostate cancer. Pet dogs are the mostdomesticated animal, sharing the habitat and oftentimes the diet of their owners. If PIN andprostate cancer are strongly related to environmental factors, then the prevalence of thesefindings might differ in a population of dogs such as military working dogs which is notexposed to the habitat and diet of humans. In this study, we determined the prevalence of PINin prostates of aged military working dogs with and without prostatic adenocarcinoma.METHODS. Cases were selected from the military working dog slide and tissue archive at theArmed Forces Institute of Pathology, Washington, DC. The most recent 329 necropsies (1991to 1996) were examined histologically by multiple reviewers; of these, 199 dogs (60%) werefound to have evaluable prostatic tissue. In addition, the most recent 50 necropsies (1958 to1996) with the diagnosis of prostatic cancer were examined, of which 25 cases (50%) werefound to have evaluable prostatic adenocarcinoma. In most cases, a single large transversesection of prostatic tissue was available for review. Medical records for each dog were re-viewed independently, and age, clinical history, indications for euthanasia, and other healthproblems were recorded.RESULTS. High grade PIN was identified in 3% of dogs (6 of 199 dogs) without prostatecancer. A total of 50.8% of dogs in this study group (101 of 199 dogs) were known to besexually intact, 26.7% of dogs (53 of 199 dogs) were castrated, and the status of the remaining22.6% of dogs (45 of 199 dogs) was unknown. High grade PIN was present in 18 of 25 dogs(72%) with prostatic adenocarcinoma. Of these cases, 11 dogs (44%) were castrated, 4 dogs(16%) were intact, and the status of 10 dogs (40%) dogs was unknown. Gleason scores rangedfrom 6 to 10, with a mean of 8.4 and a median of 8.CONCLUSIONS. High grade PIN is present in a small but substantial number (3%) ofmilitary working dogs. Of military working dogs with prostatic adenocarcinoma, 72% hadhigh grade PIN. The true prevalence in each of these cohorts is likely to be higher given thesampling variation inherent in evaluating a single random histologic section. Aged male dogs

*Correspondence to: David G. Bostwick, Department of Pathology,Mayo Clinic, 200 First St. S.W., Rochester, MN 55905.Received 2 January 1998; Accepted 18 March 1998

The Prostate 36:189–193 (1998)

© 1998 Wiley-Liss, Inc.

seem to have substantial clinical utility as an animal model for prostatic carcinogenesis. Werecommend that serial sectioning and total embedding of the prostate should be used to morethoroughly characterize premalignant and malignant diseases in aged military working dogs.This method will provide important data to determine whether a model of spontaneous PINin elderly dogs may have clinical utility in developing strategies directed toward preventingand treating prostate. Prostate 36:189–193, 1998. © 1998 Wiley-Liss, Inc.

KEY WORDS: prostatic intraepithelial neoplasia (PIN); prostate; adenocarcinoma

INTRODUCTION

High grade prostatic intraepithelial neoplasia (PIN)is considered to be the precursor of most cases of hu-man prostate cancer [1–3]. This microscopic lesion isfound in 10–16% of men undergoing prostatic needlebiopsy for suspected cancer [1,4]. The presence of highgrade PIN on prostatic biopsy is the most significantrisk factor for subsequent carcinoma [5,6]. Men withhigh grade PIN are at high risk for cancer, and up to57% will be found to have prostate cancer on subse-quent biopsy [1]. Foci of high grade PIN are present in85% of men with prostate cancer [7].

Dogs and man are the only species that developprostate cancer spontaneously. The dog alone hasenough naturally occurring prostate cancer to make itpossibly useful as a model. In humans and dogs, pros-tate cancer is associated with increasing age. Recentwork has demonstrated that dogs, like humans, de-velop PIN spontaneously and in association withprostate cancer [8,9]. Canine and human PIN havesimilar basal cell layer disruption, proliferative index,and microvessel density, and are both associated withcarcinoma. In dogs and humans, there is increasingevidence that high grade PIN is an intermediate pre-invasive lesion in the morphologic continuum of pro-gression from benign epithelium to carcinoma [8,9].

Naturally occurring animal models for human can-cer offer advantages over other models. The dog andhuman prostate are similar morphologically and bothare predisposed to cancer. Canine and human PINseem to be predictive of cancer, so the canine is auseful surrogate model for studying prostate cancer.In addition, this spontaneous animal model of PINoffers promise as a valuable tool in the evaluation ofchemopreventive agents [9] if the prevalence is suffi-ciently high.

We reviewed archived canine necropsy slides fromthe Armed Forces Institute of Pathology (AFIP, Wash-ington DC) to determine the prevalence of PIN withand without adenocarcinoma of the prostate in mili-tary working dogs.

METHODS

Case Selection

Canine necropsies were selected from the militaryworking dog slide and tissue archives at the Armed

Forces Institute of Pathology, Washington, DC. Theclinical records, archived at the Department of De-fense Military Working Dog Veterinary Services inSan Antonio, Texas were available. Military workingdogs receive standard diets and care, regardless ofdeployment status or location. By regulation, each dogreceives a semiannual complete physical examination,including digital rectal examination. All excisional bi-opsies, aspiration cytologies, and other tissue sectionsare evaluated by the Department of Veterinary Pathol-ogy and stored at the AFIP.

Cases were chosen from the most recent 329 unse-lected necropsies (1991 to 1996), and these were com-pared with the most recent 50 necropsies with pros-tatic adenocarcinoma (1958 to 1996). Excluded fromstudy were cases without evaluable prostatic tissueand those with significant tissue autolysis. Further-more, any cases with prostatic involvement by cancerfrom adjacent organs (such as urinary bladder) or me-tastasis were excluded.

Case Evaluation

All prostatic tissue slides from each case were re-viewed by six of the authors (J.W.A., L.M., A.P.,L.K.R., D.J.W., and D.G.B.) without knowledgeof clinical information. In each case, the presenceand extent of high grade prostatic intraepithelialneoplasia was recorded, including the number ofacini involved, and the amount of evaluable prosta-tic tissue. Furthermore, the grade of adenocar-cinoma was evaluated using the Gleason grading sys-tem.

The necropsy record of each dog was reviewed in-dependently, and the following data were recorded:age, breed, neuter status, clinical history, reason foreuthanasia, and other comorbid conditions.

RESULTS

High Grade PIN in Consecutive NecropsiesWithout Cancer

Of 329 cases in the files, we excluded 101 casesbecause no prostatic tissue was available for evalua-tion and 29 cases in which autolysis was substantial,precluding accurate examination; the final study

190 Aquilina et al.

group consisted of 199 cases. In most cases, only asingle large transverse section of prostatic tissue wasavailable for review. Representative military workingdog necropsy tissue sections showing benign prostateand high grade PIN are shown in Figure 1. Dog breedsin the evaluable group included 56.6% Belgian Mali-nois (112 of 199 dogs), 36% German shepherd (71 of199 dogs), and 8.0% other breeds (16 of 199 dogs). Themean age of the study group was 8.9 yr (range, 1 to 14yr; age of 2 dogs unknown), and was 9.3 in dogs withhigh grade PIN (Table I). A total of 50.8% of dogs (101of 199 dogs) were known to be sexually intact, 26.7%of dogs (53 of 199 dogs) were castrated, and the statusof the remaining 22.6% of dogs (45 of 199 dogs) wasunknown.

High grade PIN was identified in 6 of 199 dogs(3%). Four of the dogs (67%) with PIN were sexuallyintact, one dog (17%) was castrated, and one dog(17%) had unknown status. Five of six dogs were Bel-gian Malinois; the other dog was a German shepherd(Table I).

High Grade PIN in Dogs With ProstaticAdenocarcinoma

We also studied 25 dogs with prostatic adenocarci-noma. High grade PIN was present in 18 of 25 dogs(72%) (Table II; Fig. 2). Dog breeds included 14 Ger-man shepherds (56%), 7 Belgian Malinois (28%), and 4dogs of undetermined breed (16%). Of these cases, 11dogs (44%) were castrated, 4 dogs (16%) were sexuallyintact, and the neuter status of 10 dogs (40%) wasunknown. The mean age of dogs that had high gradePIN associated with cancer was 9.6 yr, and for theentire group with prostate cancer was 9.4 yr. In dogprostate cancers, we recognized patterns of growthcorrespondent to those described by Gleason in hu-man prostate cancer. We found enough similarity be-tween human and dog prostate cancer appearance.Although a perfect parallelism was not always evi-dent, there was sufficient correspondence to justify it.Gleason scores ranged from 6 to 10, with a mean of 8.4,and a median of 8. Of 25 cases, 22 cases (88%) had atleast focal Gleason primary pattern 5.

Fig. 1. Military working dog prostate sections showing benignacinus (A), high grade PIN (B and C). Hematoxylin and eosin stain.

Canine PIN 191

DISCUSSION

Our results indicate that high grade PIN is presentin a small but substantial number of aged militaryworking dogs without clinical evidence of prostaticcancer. The necropsy prevalence rate was 3% in thiscohort, but is likely underestimated given the sam-pling variation inherent in reviewing a single randomhistologic section. No large scale, systematic necropsystudies on aged dogs have evaluated totally embed-ded prostates for foci of prostate cancer and PIN, andwe recommend that future studies include serial sec-tioning and total embedding of the prostate. We alsofound that 72% of dogs with prostatic adenocarcinomahave high grade PIN, a rate similar to that observed inmen [1]. These results probably also represent an un-derestimate due to sampling variation.

The histologic appearance and prevalence of PIN inmilitary working dogs was similar to that reportedpreviously in pet dogs (66% of 29 cases) [8], despite

differences in breed, lifestyle, and diet. This findingsuggests that high grade PIN is a common pathwayfor cancer in both cohorts, and that there is eithershared genetic disposition or similar environmentalinfluences that cause cancer or PIN. This remarkablesimilarity provides a fruitful area of epidemiologic in-vestigation.

It is interesting to note that the dog and human arethe only populations that develop spontaneous pros-tatic adenocarcinoma; these results and previous stud-ies [8,9], indicate that they also share a predispositionfor spontaneous high grade PIN, unlike rats, mice, andother species.

Military working dogs tend to be large dogs; themost common breeds are the Belgian Malinois andGerman shepherd. In recent years, there has been agradual shift in the military working dog populationfrom the German shepherd to the Belgian Malinois.The prevalence of high grade PIN and prostate cancer

TABLE I. Age, Breeds, Neuter Status, and Prevalence of High Grade PIN inMilitary Working Dogs Without Prostate Cancer*

Total number of dogs(intact/castrated/

unknown)

Mean actuarial age (yrs)(intact/castrated/

unknown)

Breeds without PIN(n = 193) 193 (97/52/44) 8.9 (9.1/9.4/8.0)German shepherd 70 (27/21/22) 9.2 (9.3/9.9/8.5)Belgian Malinois 107 (63/24/20) 8.7 (9.0/8.8/7.6)Other/unknown 16 (7/7/2) 9.2 (9.3/9.9/6.5)

Breeds with PIN(n = 6) 6 (4/1/1) 9.3 (9.5/11/7)German shepherd 1 (1/0/0) 13 (13/—/—)Belgian Malinois 5 (3/1/1) 8.6 (8.3/11/7)

*Consecutive necropsies 1991–1996 (study group, n = 199).

TABLE II. Age, Breeds, Neuter Status, and Prevalence of High Grade PINin Military Working Dogs With Prostate Cancer*

Total number of dogs(intact/castrated/

unknown)

Mean actuarial age (yrs)(intact/castrated/

unknown)

Dogs without PIN(n = 7) 7 (1/3/3) 9 (13/7.3/9.3)German shepherd 3 (0/1/2) 9.3 (—/7/10.5)Belgian Malinois 3 (0/2/1) 7.3 (—/7.5/7)Other/unknown 1 (1/0/0) 13 (13/—/—)

Dogs with PIN(n = 18) 18 (3/8/7) 9.6 (9.7/9.6/9.6)German shepherd 11 (2/4/5) 10.3 (11.5/9.5/10.4)Belgian Malinois 4 (0/2/2) 8.5 (—/9.5/7.5)Other/unknown 3 (1/2/0) 8.7 (6/10/—)

*Unselected consecutive necropsies with prostate cancer (study group, n = 25).

192 Aquilina et al.

may vary between breeds; however, our series did nothave sufficient numbers to determine whether thereare breed-specific differences in prevalence of highgrade PIN.

Androgens are a necessary factor for developmentof human prostate cancer. Although the mechanismsthat govern this process remain uncertain, preopera-tive androgen-deprivation therapy in men caused asignificant decrease in the prevalence and extent ofhigh grade PIN in radical prostatectomy patients com-pared with untreated matched controls [10]. In a re-cent prospective randomized series, Vaillancourt et al.[11] demonstrated a significant reduction in the preva-lence of high-grade PIN from 67 to 6% after 3 monthsof combination of neoadjuvant endocrine therapy. Theavailability of drugs capable of causing androgen de-privation by a variety of mechanisms should aid infuture studies [12–15]. These agents may be useful indesigning appropriate studies to determine the hor-monal influences on the canine prostate.

A canine model, using high grade PIN as a surro-gate for prostate cancer, may have utility for prevent-ing and treating prostate cancer. Similar to its humancounterpart, advanced prostate cancer in dogs may beindependent of testicular androgens. Of the 25 dogs inour study with prostate cancer, 4 dogs were intact, 11dogs were castrated, and the status of 10 dogs wasunknown. The adrenal and pituitary glands may havea significant influence [16]; therefore, an effort to un-derstand the mechanisms of canine prostatic cellularhormonal response is warranted.

CONCLUSIONS

The dog is the only species other than man in whichspontaneous high grade PIN and prostate cancer com-

monly occur. We found a 3% prevalence of high gradePIN in single random histologic sections in aged mili-tary working dogs, indicating that this may be an ap-propriate model for the study of human prostatic car-cinogenesis and may provide a useful tool in the de-velopment of future therapeutic strategies forchemoprevention of prostate cancer in dogs and man.

REFERENCES

1. Bostwick DG: High grade prostatic intraepithelial neoplasia:The most likely precursor of prostate cancer. Cancer 1995;75:1823–1836.

2. Nagle RB, Petein M, Brawer M, Bowden GT, Cress AE: Newrelationships between prostatic intraepithelial neoplasia andprostatic carcinoma. J Cell Biochem Suppl 1992;16H:26–29.

3. Bostwick DG, Brawer MK: Prostatic intra-epithelial neoplasiaand early invasion in prostate cancer. Cancer 1987;59:788–794.

4. Bostwick DG, Qian J, Frankel K: The incidence of high gradeprostatic intraepithelial neoplasia in needle biopsies. J Urol1995;154:1791–1794.

5. Brawer MK, Bigler SA, Sohlberg OE, Nagle RB, Lange PH: Sig-nificance of prostatic intraepithelial neoplasia on prostateneedle biopsy. Urology 1991;38:103–107.

6. Davidson D, Bostwick DG, Qian J, Wollan P, Oesterling JE, Rud-ders RA, Siroky M, Stilmant M: Prostatic intraepithelial neopla-sia is a risk factor for adenocarcinoma: Predictive accuracy inneedle biopsies. J Urol 1995;154:1295–1299.

7. McNeal JE, Bostwick DG: Intraductal dysplasia: A premalignantlesion of the prostate. Hum Pathol 1986;17:64–71.

8. Waters DJ, Hayden DW, Bell FW, Klausner JS, Qian J, BostwickDG: Prostatic intraepithelial neoplasia in dogs with spontane-ous prostate cancer. Prostate 1997;30:92–97.

9. Waters DJ, Bostwick DG: Prostatic intraepithelial neoplasia oc-curs spontaneously in the canine prostate. J Urol 1997;157:713–716.

10. Ferguson J, Zincke H, Ellison E, Bergstrahl E, Bostwick DG:Decrease of prostatic intraepithelial neoplasia following andro-gen deprivation therapy in patients with stage T3 carcinomatreated by radical prostatectomy. Urology 1994;44:91–95.

11. Vaillancourt L, Tetu B, Fradet Y, Dupont A, Gomez J, Cusan L,Suburu ER, Diamond P, Candas B, Labrie F: Effect of neoadju-vant endocrine therapy (combined androgen blockade) on nor-mal prostate and prostate carcinoma: A randomized study. AmJ Surg Pathol 1996;20;86–93.

12. Aquilina JW, Lipsky JJ, Bostwick DG: Androgen deprivation asa strategy for prostate cancer chemoprevention. J Natl CancerInst 1997;89:689–696.

13. Bostwick DG, Aquilina JW: Prostatic intraepithelial neoplasia(PIN) and other prostatic lesions as risk factors for cancer che-moprevention trials. J Cell Biochem 1996;25:156–164.

14. Kelloff GJ, Boone CW, Crowell JA, Steele VE, Lubet R, DoodyLA: Surrogate endpoint biomarkers for phase II cancer chemo-prevention trials. J Cell Biochem 1994;19(suppl):1–9.

15. Bostwick DG: Target populations and strategies for chemopre-vention trials of prostate cancer. J Cell Biochem 1994;19(suppl):191–196.

16. Obradovich J, Walshaw R, Goullaud E: The influence of castra-tion on the development of prostatic carcinoma in the dog, 43cases (1978–1985). J Vet Intern Med 1987;1:183–187.

Fig. 2. Military working dog prostate section showing high gradePIN (top) in association with adenocarcinoma of the prostate.Hematoxylin and eosin stain.

Canine PIN 193