Immunohistochemical analysis of cyclin A, cyclin D1 and P53 in mammary tumors, squamous cell carcinomas and basal cell tumors of dogs and cats

The involvement of cyclin A, cyclin D1 and p53 proteins in canine and feline tumorigenesis was analyzed immunohistochemically. In the present study, a total of 176 cases were examined, among which there were 108 canine cases (75 mammary lesions, 16 squamous cell carcinomas and 17 basal cell tumors) and 68 feline cases (43 mammary lesions, 20 squamous cell carcinomas and 5 basal cell tumors). Speckled nuclear staining for cyclin A was observed in 19/38 (50%) canine malignant mammary tumors and 18/37 (48.6%) feline mammary carcinomas, while this was not seen in benign mammary tumors of either dogs or cats. Marked intense nuclear cyclin A staining was seen in 7/16 (43.8%) canine squamous cell carcinomas and 18/20 (90.0%) feline squamous cell carcinomas. Only 3/17 (17.6%) canine basal cell tumors showed slight and scattered staining for cyclin A. Expression of cyclin D1 was very rare in both canine and feline tumors. Nuclear staining of p53 was found in 7/37 (18.9%) feline mammary carcinomas. Intense immunoreactivity for p53 was found in 6/16 (37.5%) canine squamous cell carcinomas and 8/20 (40%) feline squamous cell carcinomas. These results suggest that cyclin A may have a role in the proliferation of canine malignant mammary tumors, feline mammary carcinomas and squamous cell carcinomas of dogs and cats, and p53 may associate with the tumorigenesis of feline mammary carcinomas and squamous cell carcinomas of dogs and cats.     

Coordinate expression of cytokeratins 7 and 20 in feline and canine carcinomas.

Forty-seven feline and 60 canine epithelial tumors were studied to test the coordinate expression of cytokeratin 7 (CK 7) and cytokeratin 20 (CK 20) using commercially available monoclonal antibodies and an avidin-biotin immunoperoxidase staining technique. Previously, the distribution of both cytokeratins was examined in normal tissues from 4 cats and 4 dogs. The pattern of distribution of CK 7 in normal tissues was similar, with minor differences, to that described in humans, whereas the reactivity pattern of CK 20 in cats and dogs was wider than that in humans. The subset of tumors strongly expressing CK 7 and CK 20 included pancreatic adenocarcinomas (100%), transitional cell carcinomas (75%), and endometrial carcinomas (67%) in the cat. None of the canine tumors had this immunophenotype. Feline (50%) and canine (56%) mammary gland carcinomas and canine cholangiocarcinomas (67%) were the only tumors presenting the CK 7 +/CK 20- immunophenotype, whereas the CK 7-/CK 20+ immunophenotype included thyroid carcinomas (100%), intestinal adenocarcinomas (60%), bronchioloalveolar carcinomas (50%), and renal carcinomas (50%) in the cat and intestinal adenocarcinomas (56%), gastric adenocarcinomas (50%), and ovarian carcinomas (50%) in the dog. The CK 7-/CK 20- immunophenotype included the rest of the analyzed tumors. The immunohistochemical evaluation of coordinate expression of both CK 7 and CK 20 in feline and canine carcinomas using monoclonal antibodies provides important information that can help to discriminate among carcinomas from different primary sites and could be particularly helpful in the determination of the primary site of origin of carcinomas presenting as metastatic disease.     

COX‐2, mPGES‐1 and EP2 receptor immunohistochemical expression in canine and feline malignant mammary tumours

Prostaglandin (PG) signalling is involved in human and animal cancer development. PG E₂ (PGE₂) tumour‐promoting activity has been confirmed and its production is controlled by Cyclooxygenase‐2 (COX‐2) and microsomal PGE synthase‐1 (mPGES‐1). Evidence suggests that mPGES‐1 and COX‐2 contribute to carcinogenesis through the EP2 receptor. The aim of our study was to detect by immunohistochemistry COX‐2, mPGES‐1 and EP2 receptor expression in canine (n = 46) and feline (n = 50) mammary tumours and in mammary non‐neoplastic tissues. COX‐2 positivity was observed in 83% canine and 81% feline mammary carcinomas, mPGES‐1 in 75% canine and 66% feline mammary carcinomas and the EP2 receptor expression was observed in 89% canine and 54% feline carcinomas. The frequency of COX‐2, EP2 receptor and mPGES‐1 expression was significantly higher in carcinomas than in non‐neoplastic tissues and adenomas. COX‐2, mPGES‐1 and EP2 receptor expression was strongly associated. These findings support a role of the COX‐2/PGE2 pathway in the pathogenesis of these tumours.     

孕激素诱导的生长激素与犬的乳腺肿瘤

乳腺肿瘤是犬的最常见的肿瘤，约占母犬肿瘤总数的一半；乳腺肿瘤约有一半是恶性肿瘤，发生转移的比率很高。犬乳腺肿瘤的发病原因还不很明确，它的形成都受哪些因子调节还未完全确定。孕激素能够诱导乳腺组织生成并蓄积生长激素，使乳腺增生，最终导致犬乳腺肿瘤的形成。在此过程中，雌激素受体、孕激素受体、生长激素结合蛋白和生长激素受体在乳腺癌的形成、发展中起关键作用。从内分泌的角度对犬乳腺肿瘤发病机理作了简单综述，以期为激素治疗乳腺肿瘤提供一些理论依据。     

An immunohistochemical study of cyclooxygenase-2 expression in various feline neoplasms

Cyclooxygenase (COX) enzymes catalyze the synthesis of prostaglandins and exist as two isoforms, COX-1 and COX-2. COX-2 is a potent inducible mediator of inflammation. COX-2 is also upregulated in several human tumors and in canine squamous cell, renal cell, and transitional cell carcinomas, prostatic adenocarcinoma, and intestinal neoplasia. The purpose of this study was to determine whether COX-2 is expressed in various feline tumors. Results of this study may help determine whether COX-2 is a potential target for therapeutic and preventive strategies in cats. Immunohistochemical studies were performed on paraffin-embedded tissues using the amplified streptavidin-biotin-horseradish peroxidase system. COX-2 was found in 7 of 19 (37%) feline transitional cell carcinomas and in 2 of 21 (9%) feline oral squamous cell carcinomas. No COX-2 immunoreactivity was detected in cutaneous squamous cell carcinomas (6), adenocarcinomas (nine mammary, eight pulmonary, seven intestinal), lymphomas (six nasal, six intestinal), or 10 vaccine-associated sarcomas. The widespread absence of COX-2 expression in most feline neoplasms might suggest that COX-2 inhibitors would have a low potential as anticancer agents.     

p63: a novel myoepithelial cell marker in canine mammary tissues

Several immunohistochemical markers have been used to demonstrate the presence of myoepithelial cells in order to determine their role in the histogenesis of mammary tumors. p63, a recently characterized p53 homologue, is consistently expressed in myoepithelial cells of the human breast; however, no assessment of its immunoreactivity has been reported so far in canine mammary tissues. We investigated p63 immunohistochemical expression, as a novel myoepithelial cell nuclear marker, in 81 samples of normal (n = 2), hyperplastic (n = 11), and neoplastic (n = 68) canine mammary tissues. Myoepithelial phenotype was confirmed by using complementary monoclonal antibodies: alpha-smooth muscle actin, cytokeratin 14, cytokeratin AE1/AE3, and vimentin. p63 expression was observed in 91.4% (74/81) of the samples evaluated. Normal mammary glands, mammary hyperplasias, and benign tumors showed 100% immunoreactivity, with p63 expression restricted to myoepithelial cell nuclei. In general, benign mixed tumors showed a basal cell compartment immunoreactive to p63, with a gradual decrease of its expression during myoepithelial transformation. p63 expression was found in 72% of malignant tumors, allowing myoepithelial or basal cell identification in spindle-cell carcinomas (2/2), tubulopapillary carcinomas (8/9), solid carcinomas (7/10), and carcinosarcomas (1/3). The osteosarcoma analyzed was p63 negative. In our series, stromal components were consistently nonreactive to p63. In conclusion, the present study reveals p63 as a sensitive and highly specific marker of myoepithelial cells in canine mammary tissues, and the authors suggest p63 as an additional marker for defining myoepithelial histogenesis.     

Immunological aspects of mammary tumors in dogs and cats: a survey including own studies and pertinent literature

Naturally occurring cancer in companion animals parallels cancer in man more closely than does experimentally induced cancer in inbred laboratory animals. In dogs and cats, as in man, a role for immune responses is indicated in the development of tumors. A survey is presented based on the literature and our own studies concerning the immunological and immunotherapeutic aspects of canine and feline mammary neoplasia. In dogs bearing mammary neoplasms, circulating immune complexes appear to play a negative role in the generation of effective antitumor immune responses. The functional role of peripheral blood lymphocytes and tumor-infiltrating lymphocytes in dogs and cats with mammary tumors is not yet fully established. No tumor antigen responsible for humoral or cellular responses has yet been identified. Extracorporeal perfusion of serum of dogs with mammary tumors and subcutaneous administration of mitomycin- and neuraminidase-treated autologous tumor cells are associated with improved prognosis. The opposite was true for i.v. treatment with BCG or Corynebacterium parvum vaccine in our study, in contrast to a previous report. A number of other treatment modalities in cats and dogs with mammary carcinomas failed to induce tumor regression. Canine and feline mammary carcinomas are good candidates for modern immunotherapeutic approaches.     

Expression of TopBP1 in canine mammary neoplasia in relation to histological type, Ki67, ERalpha and p53

TopBP1 is aberrantly expressed in human and feline mammary carcinomas, but expression of this BRCA1-related protein has not been investigated in canine mammary carcinomas. In this study, 132 canine mammary tumours (46 benign, 86 carcinomas) were examined immunohistochemically for expression of TopBP1, oestrogen receptor α (ERα), Ki67 and p53. Positive staining for TopBP1 was evident in all canine mammary lesions, although five samples had <20% positive cells. The number of samples with high levels of staining increased in different categories from benign mixed tumour to adenoma to carcinoma. Most TopBP1 staining was nuclear, but both nuclear and cytoplasmic staining were observed as the degree of malignancy increased, similar to human and feline mammary carcinomas. Benign mixed tumours, however, had more cytoplasmic staining than adenomas. Expression of p53 and the proliferation marker Ki67 increased from benign mixed tumour to adenoma to carcinoma, but the differences between benign and malignant tumours were more distinct than for TopBP1 expression. ERα expression decreased from malignant to benign tumours, although over half of the benign mixed tumours were negative. TopBP1 was expressed in canine mammary tumours at higher levels than has been reported previously for cats, although the shift in cellular localisation with malignancy was similar.     

Aglepristone decreases proliferation in progesterone receptor-positive canine mammary carcinomas

Background: Progesterone receptor (PR) antagonist aglepristone (RU534) has been used successfully for pregnancy termination and therapy of pyometra, vaginal tumors, and mammary hyperplasia in bitches and queens. All of these conditions share with canine mammary carcinomas the expression of PR. Objectives: To study the effect of RU534 on proliferation and apoptosis in canine mammary carcinomas in relation to PR expression. Animals: Twenty‐seven nonspayed bitches with mammary carcinomas were treated with either 2 doses of 20 mg/kg RU534 (n = 22, RU534‐treated group) or oil placebo (n = 5, control group) on days 1 and 8. Methods: Tumor samples were collected before (day 1) and after (day 15) treatment for immunohistochemistry. PR expression, proliferation index (PI), and apoptotic index (AI) were determined using antibodies against PR, Ki67, and cleaved lamin A/C antigens, respectively. The effect of treatment on these parameters was analyzed. Results: Differential expression of PR between day 1 (59.1% PR‐positive tumors) and day 15 (36.4% PR‐positive tumors) was observed in RU534‐treated tumors exclusively. After RU534 treatment, mean PI was significantly decreased in PR‐positive but unchanged in PR‐negative RU534‐treated tumors. A reduction of ≥20% in PI was found in 61.5% of RU534‐treated tumors with PR expression. Conversely, no effect on AI was observed after RU534 treatment. Conclusions and Clinical Importance: Neoadjuvant RU534 treatment had PR expression‐related inhibiting effects on proliferation of canine mammary carcinoma cells.     

MIB-1 labeling index in feline dysplastic and neoplastic mammary lesions and its relationship with postsurgical prognosis

Samples from feline normal, dysplastic, and neoplastic mammary tissues were used to investigate the usefulness of MIB-1 labeling index (MIB-1 I) as a prognostic indicator. Forty-eight queens bearing invasive carcinomas were included in a 2-year follow-up study. Mammary lesions were classified according to the World Health Organization system, and invasive carcinomas were further graded on the basis of the degree of tubule formation, the degree of nuclear and cellular pleomorphism, and mitotic count. Additional sections were immunostained using MIB-1 antibody, and MIB-1 I was expressed as a percentage of positive nuclei. In normal mammary gland tissues, the mean MIB-1 I was <1%. A low proliferation rate was found in all mammary adenosis and in situ carcinomas, and the highest rates were observed in feline mammary hypertrophy and invasive carcinomas. Twenty-one (43.7%) of the queens bearing invasive carcinomas were still alive at the end of the trial, and 27 (56.2%) had died. The MIB-1 I was not significantly correlated with clinical outcome, age, histologic type, or grading of the tumors, but a borderline correlation was observed with invasion of lymphatic vessels. Univariate analysis showed that high MIB-1 I was also not associated with decreased overall survival, whereas the grading system of the tumors had high predictive value (P = 0.0040) for postsurgery survival. The lack of correlation between MIB-1 I and postsurgery survival suggests that this marker alone is not sufficient to determine a correct prognosis in feline mammary carcinomas, even if it is a useful proliferation marker.     

Comparison of histology and clinical variables to DNA ploidy in canine mammary tumors

Flow cytometric DNA analysis was done on 132 canine mammary tumors from 99 dogs to evaluate the relation to histology and to clinical staging. Seventy-one tumors (54%) were histologically malignant; 38 (54%) of these were aneuploid and 33 (46%) were diploid. Fifty-two (39%) tumors were histologically benign, of which 45 (87%) were diploid and seven (13%) aneuploid. There were nine dysplastic mammae (7%); two were aneuploid and the rest diploid. DNA indices varied from 0.72 to 2.35. Of 58 mammary carcinomas, 25 (43%) were diploid and 33 (57%) were aneuploid (of the latter, 16 showed hypodiploidy and 17 hyperdiploidy with a predominance between DNA index 1.10 and 1.50). Three tumors (two carcinomas and one malignant mixed tumor) were multiploid with two aneuploid cell populations. The histological type varied within eight tumors, and in four of these the DNA index also varied. DNA indices varied within three tumors with uniform morphology. No correlation was found between DNA index and age of the dogs, nor between DNA index and tumor size. No significant differences were found between DNA index and histology, tumor growth pattern, or tumor location. Benign tumors were smaller than carcinomas, which were smaller than malignant mesenchymal tumors. Tumors growing adherent to the skin were larger than those not adherent to the skin. The regional lymph nodes were examined in 33 cases. No significant difference between the mean DNA index and presence of lymph node metastasis was found. These results show the possibility of using flow cytometry for DNA analysis in canine mammary tumors.(ABSTRACT TRUNCATED AT 250 WORDS)     

p53 gene mutations occurring in spontaneous benign and malignant mammary tumors of the dog

Sixty-three cases of benign and malignant canine mammary tumors were analyzed to define the alteration of exons 5-8 for the p53 tumor suppressor gene using polymerase chain reaction direct sequence analysis with paraffin-embedded tissues. Four missense mutations were found in 38 benign mammary tumors (11%), and five missense (one tumor had two missense mutations) and one nonsense mutations were found in 25 mammary carcinomas (20%). These data suggest that the p53 gene alterations might be initiated at an early stage of canine mammary carcinogenesis and p53 mutations might be associated with malignancy. However, there was no evidence of any relationship between the p53 alterations and the histologic types of tumors or breeds of dogs.     

Dysregulation of beta-catenin is common in canine sporadic colorectal tumors.

Human colorectal tumorigenesis is often initiated by APC (adenomatous polyposis coli) or beta-catenin (CTNNB1) mutations, which result in dysregulation of beta-catenin expression, followed by alterations in E-cadherin and/or p53. We examined 32 canine intestinal tumors for expression and intracellular distribution of beta-catenin, E-cadherin, and p53 using immunohistochemistry. beta-Catenin in normal mucosal epithelial cells was restricted to lateral cell membranes, but 13/13 (100%) colorectal adenomas had intense cytoplasmic and/or nuclear reactivity. Three of six (50%) colorectal carcinomas, 2/13 (15%) small intestinal carcinomas, and dysplastic cells in 1/2 focal hyperplastic lesions in the small intestine had a similar pattern of staining; remaining tumors had normal membranous beta-catenin reactivity. There was a correlation (P = 0.007) between abnormal beta-catenin and E-cadherin staining with 11/13 (85%) colorectal adenomas, 3/6 (50%) colorectal carcinomas, and 3/13 (23%) small intestinal carcinomas showing decreased membranous reactivity compared with normal mucosal epithelium. E-cadherin staining was reduced more often in adenomas than in carcinomas (P = 0.04). There were two patterns of nuclear p53 staining: > 60% of nuclei in 2/26 (8%) carcinomas (one colorectal, one small intestinal) were strongly labeled, whereas three colorectal adenomas and one small intestinal carcinoma had fainter staining in 10-20% of cells. Dysregulation of beta-catenin appears to be as important in canine colorectal tumorigenesis as it is in the human disease and could be due to analogous mutations. Malignant progression in canine intestinal tumors does not appear to be dependent on loss of E-cadherin or beta-catenin expression or strongly associated with overexpression of nuclear CMI antibody-reactivity p53.     

Mammary mucinous carcinoma in the cat

Mucinous carcinoma of the mammary gland is a rare tumor characterized by excessive mucin production. In human and canine pathology, the diagnosis of mucinous carcinoma is based on the demonstration of an epithelial phenotype of mucus-producing cells and periodic acid-Schiff (PAS)-diastase positivity of the mucin. The histologic and immunohistologic characteristics of feline mucinous mammary carcinoma were examined. Of 656 cases of feline mammary neoplasms and dysplasias, 3.2% were found to be mucin-producing tumors. Cytokeratin 19 (16 cases positive, 4 heterogenous, and 1 negative) and vimentin (15 cases positive, 2 heterogenous, and 4 negative) expression were examined, and the mucin produced was alcian blue positive. PAS-diastase staining was variable (38.1%). Based on these findings, mucinous mammary carcinoma in the cat varies significantly from the human and canine varieties and alcian blue is the prominent stain in the diagnosis of feline mucinous carcinoma.     

The expression of sialyl Lewis X in canine and feline mammary gland tumors

The expression of sialyl Lewis X (sLe(x)) in 93 canine and 15 feline mammary gland tumors (MGT) obtained by surgical resection at Veterinary Medical Center, the University of Tokyo was examined by immunohistochemistry. Their clinicopathological features and prognosis were also reviewed. Approximately 60% of MGT tissues showed sLe(x) positive expressions, while all normal mammary gland tissues were negative. However, its expression was not correlated with clinicopathological features and prognosis significantly. This study suggests that sLe(x) may be a tumor-associated antigen in canine and feline MGTs.     

Loss of p27 expression in canine mammary tumors and their metastases

The p27 gene is a member of the cyclin-dependent kinase inhibitors, which arrest G1- to S-phase transition of the cell cycle. We have previously shown a significant reduction of p27 mRNA expression level in laser-microdissected mammary carcinomas and their lymph node metastases when compared to non-neoplastic mammary gland of the same dog. Here, p27 expression was analyzed on the protein level in non-neoplastic mammary gland, primary mammary carcinomas, their lymph node metastases and intravascular tumor cells of 49 dogs, adenomas of 49 dogs and non-neoplastic mammary gland of 98 dogs by immunohistochemistry. A significantly (p ? 0.05) decreased percentage of p27 positive tissue samples was found when normal gland was compared with adenomas, carcinomas and lymph node metastases. Specifically, 91% of normal gland epithelium displayed nuclear p27 expression. In contrast, only 22% of the adenomas, 20% of carcinomas, 12% of lymph node metastases and 32% of intravascular tumor cells had p27 reactivity. Cell cycle control by p27 is therefore lost in the majority of canine mammary tumors. The lack of significant differences between benign and malignant mammary tumors indicates that decreased p27 expression is an early step in carcinogenesis of canine mammary tumors and hinders the use of p27 as a marker of malignancy for this tumor type.     

Computerized morphometry of mean nuclear diameter and nuclear roundness in canine mammary gland tumors on cytologic smears

BACKGROUND: The use of computer-based image analysis systems in veterinary oncology has increased. Computerized morphometry is a part of image analysis that describes geometric figures of cellular structures in any dimension. Most investigators have performed morphometric analysis on histologic specimens. Computer-assisted nuclear cytomorphometry can provide important preoperative information on neoplastic lesions in animals. OBJECTIVES: The aim of this study was to define whether the morphometric parameters of mean nuclear diameter and nuclear roundness could be used to differentiate benign from malignant canine mammary gland tumors on cytologic specimens. METHODS: Mean nuclear diameter and nuclear roundness were determined by computer-assisted morphometry of epithelial cells in Hemacolor-stained cytologic smears from normal canine mammary gland (n = 7) and from canine mammary adenomas (n = 8), tubulopapillary carcinomas (n = 9), and solid carcinomas (n = 6). Data were analyzed by the Mann-Whitney U test. RESULTS: Significant differences (P <.001) were found in mean nuclear diameter and nuclear roundness among all tumor types and in comparison with normal canine mammary gland epithelial cells (except for nuclear roundness between tubulopapillary carcinomas and solid carcinomas). CONCLUSIONS: The morphometric parameters of mean nuclear diameter and nuclear roundness can be used in the preoperative differentiation of benign from malignant canine mammary gland tumors.     

Calponin expression and myoepithelial cell differentiation in canine, feline and human mammary simple carcinomas

Calponin is a 34‐kDa smooth muscle‐specific protein that has been shown to be a highly sensitive marker of myoepithelial cells in canine, feline and human mammary tissue and tumours. The expression of calponin was studied in 15 canine, 32 feline and 28 human simple mammary carcinomas using a monoclonal mouse antihuman calponin antibody and the avidin–biotin peroxidase complex (ABC) immunohistochemical technique. Calponin expression was compared with the expression of cytokeratin 14, a marker of normal mammary myoepithelial cells in the three species. Four different types of calponin‐positive cells were identified: (1) Type 1: cytokeratin‐14‐positive pre‐existing myoepithelial cells forming a continuous layer with images of focal disruptions; (2) Type 2: cytokeratin‐14‐positive isolated nests of fusiform, polygonal or round cells without atypia; (3) Type 3: cytokeratin‐14‐positive atypical cells indistinguishable from non‐reactive atypical cells, which should have never been detected in haematoxylin and eosin‐stained sections and (4) Type 4: cytokeratin‐14‐negative stromal fusiform cells around the neoplastic growth or cell nests, identified as myofibroblasts. Calponin‐negative and cytokeratin‐14‐positive atypical neoplastic cells were observed in three canine, 28 feline and two human carcinomas. The latter were indicative of altered expression of high‐molecular‐weight cytokeratins in luminal epithelial‐type simple carcinomas. Our findings show that calponin is a good marker of myoepithelial cell differentiation in feline, human and, particularly, canine simple carcinomas. The high number (six out of 15) of canine tumours with type 3 cells points to the need of both introducing calponin examination in the routine diagnostic schedule and performing further studies on its prognostic significance.     

Sensitivity and specificity of cytologic evaluation in the diagnosis of neoplasia in body fluids from dogs and cats

Sensitivity and specificity were determined for the cytologic detection of malignant tumors in canine and feline body cavity effusions. In a prospective study, 424 body cavity effusions from dogs and cats were collected and evaluated, including 70 pleural and 163 peritoneal effusions from dogs, and 77 pleural and 114 peritoneal effusions from cats. Final diagnoses were confirmed in 339 of the 424 cases by clinical follow-up, necropsy, and in the case of malignant tumors, Histopathology. Malignant tumors were found in 18% of canine and 25% of feline body cavity effusions. Approximately one-half of tumors in both dogs and cats were carcinomas. Discrete cell tumors accounted for 56% of feline neoplastic effusions. The sensitivity of cytologic evaluation for the detection of malignant tumors in body cavity effusions was 64% for dogs and 61% for cats. Specificity was 99% for canine and 100% for feline effusions. Sensitivity and specificity were comparable to those obtained with cytologic evaluation of human samples.     

Cyclooxygenase-2 expression in canine mammary tumors

Mammary tumors are the most common neoplasms in female dogs. Induction of cyclooxygenase-2 (COX-2) has been implicated in various cancers in humans. However, expression of COX-2 has not been investigated in canine mammary tumors. Normal mammary gland (n = 4), simple or complex adenomas (n = 63), and simple or complex adenocarcinomas (n = 84) were studied by immunohistochemistry. Results showed that COX-2 was not expressed in the normal gland but was detected in 24% of adenomas and in 56% of adenocarcinomas (P < 0.001). The incidence of COX-2 expression and the intensity of the COX-2 signal were higher in adenocarcinomas than in adenomas (P < 0.001). These results demonstrate for the first time that COX-2 is induced in a proportion of canine mammary tumors and that COX-2 expression is more frequent and more intense in malignant than in benign tumors, suggesting a potential role for COX-2 in canine mammary tumorigenesis.