- Email: [email protected]
Canine Lymphomas: a Morphological and Immunohistochemical Study of 55 Cases, with Observations on p53 Immunoexpression
J. Comp. Path. 2004, Vol. 131, 207–213
www.elsevier.com/locate/jcpa
Canine Lymphomas: a Morphological and Immunohistochemical Study of 55 Cases, with Observations on p53 Immunoexpression F. A. R. Sueiro, A. C. Alessi* and J. Vassallo Department of Anatomical Pathology, Faculty of Medical Sciences, State University of Campinas (UNICAMP), Sa˜o Paulo, P. O. Box 6111, 13081-970 Campinas, SP, and *Department of Veterinary Pathology, Faculty of Veterinary Medicine, University of Sa˜o Paulo State, Jaboticabal, Brazil
Summary The purpose was to determine the immunophenotype of canine lymphomas (CLs) classified according to the WHO nomenclature for domestic animals, and to relate these findings to the immunoexpression of p53 protein. Lymphomas were immunophenotyped with antibodies to CD79a, CD3, and p53 protein, suitable for paraffin wax-embedded tissue sections. Of 55 cases, 40 (72.7%) were of the B-cell phenotype, 12 (21.8%) of the T-cell phenotype, and three (5.4%) were non-B-non-T lymphomas. Of the 40 B-cell lymphomas, 31 were of the large B-cell type, six were lymphoplasmacytic, one lymphocytic, one follicular (grade II) and one unclassified low-grade. One of the peripheral T-cell lymphomas was morphologically consistent with anaplastic large cell lymphoma. Immunolabelling for p53 occurred in 24 B-cell and seven T-cell lymphomas. Lymphomas with high grade histology showed a significantly increased frequency of p53 positivity ðP ¼ 0:01Þ: Positivity for p53 (more than 10% positive cells) tended to be associated with the T-cell phenotype ðP ¼ 0:06Þ: Mean patient age was significantly higher in p53-positive cases ðP ¼ 0:02Þ: These data are comparable with findings in human lymphomas. q 2004 Published by Elsevier Ltd. Keywords: canine lymphomas; CD3; CD79a; p53 protein
Introduction Tumours of the haematopoietic system are among the neoplasms most frequently encountered in veterinary medicine (Madewell, 1985). Canine non-Hodgkin’s lymphoma, also known as canine lymphoma (CL), represents the third most common malignant neoplasm in dogs, the annual incidence rate being 13 to 24 per 100 000 dogs (Ba¨ckgren, 1965; Dorn et al., 1967). The mean age of dogs with CL varies from 6.3 to 7.7 years (Van Pelt and Connor, 1968; Madewell, 1985; Teske, 1994), with no obvious gender predilection (Haga et al., 1988). An association between CL and viral infection is controversial and has not been proved Correspondence to: J. Vassallo. 0021–9975/$ - see front matter doi: 10.1016/j.jcpa.2004.04.002
(Van Hoosier et al., 1968; Owen, 1971; Kaplan, 1978; Onions, 1980; Safran et al., 1992). Boxers, Saint Bernards, Scottish terriers, basset hounds, Airedale terriers and bulldogs are reported to be at increased risk, and dachshunds and Pomeranians at reduced risk (Teske, 1994). CLs, like human lymphomas (Jaffe et al., 2001), were included in a World Health Organization (WHO) classification based on morphological and immunophenotypical criteria (Valli et al., 2002), but this classification is as yet too recent to have formed the basis for further studies. CLs have been proposed as animal models for their human counterparts (Owen et al., 1975; Hayes, 1978; MacEwen, 1990). q 2004 Published by Elsevier Ltd.
208
F.A.R. Sueiro et al.
The molecular biology of CL has been studied, but not in great depth. A family of bullmastiffs in Great Britain was described as being at high risk from lymphoma, suggesting that genetic factors play a role in pathogenesis (Onions, 1984). Aberrations of the p53 tumour suppressor gene were demonstrated in CL, and other canine neoplasms (Gamblin et al., 1997; Veldhoen et al., 1998; Setoguchi et al., 2001). In CL, however, no attempt was made to relate p53 to morphological and immunohistochemical findings. The purpose of this study was to examine immunohistochemically CL samples from the routine diagnostic laboratory at the University of Sa˜o Paulo to (1) evaluate the applicability of the WHO nomenclature, and (2) relate the immunophenotypical findings to p53 immunoexpression.
Materials and Methods Samples Formalin-fixed paraffin wax-embedded tissue samples from 65 cases in which a diagnosis of CL had been made during the period 1990– 2000 at the Department of Veterinary Pathology of the University of Sa˜o Paulo State (Jaboticabal) were examined morphologically and immunohistochemically. Haematoxylin and eosin (HE)-stained sections were reviewed and classified according to the WHO International Histological Classification of Hematopoietic Tumors of Domestic Animals (Valli et al., 2002). Immunohistochemistry Serial sections were placed on silanized slides and dewaxed. Antigen retrieval was performed after hydration and endogenous peroxidase blocking with 3% H2O2 in a steamer at 90 8C for 30 min. The slides were then incubated overnight (18 h) at 4 8C with the following primary antibodies: anti-CD79a (clone HM57; Dakopatts, Carpenteria, CA, USA), diluted 1 in 50; anti-CD3 (code A0452, rabbit polyclonal; Dakopatts), diluted 1 in 50; antimyeloperoxidase (rabbit polyclonal; Dakopatts), diluted 1 in 400; anti-lysozyme (rabbit polyclonal; Dakopatts), diluted 1 in 500; and anti-p53 (CM1, rabbit polyclonal; Novocastra Laboratories, Newcastle, UK), diluted 1 in 100. For anti-CD79a, specific binding was detected with the CSA (Catalized Signal Amplification) peroxidase system (K1500; Dakopatts). For all other markers, the EnVision peroxidase system (K1491; Dakopatts) was used. Labelling was “visualized” with
3,30 -diaminobenzidine (D-5637; Sigma, St Louis, MO, USA) added to H2O2 3% in phosphatebuffered saline (pH 7.6). Positive controls consisted of canine reactive lymph nodes with known immunohistochemical properties. In negative controls the primary antibody was replaced by bovine serum albumin (BSA) 1% in phosphatebuffered saline, pH 7.6. Immunolabelling was evaluated microscopically. A tumour was considered to have a B immunophenotype when at least 10% of the neoplastic cells reacted exclusively with anti-CD79a, and a T immunophenotype when at least 10% exclusively expressed CD3. Immunolabelling of neoplastic cells for p53 was scored as follows: 0, none positive; 1, , 10% positive (nuclear labelling); 2, . 10% positive. Statistical Analysis The tests used were the exact Fisher test, Chi-square test or the Mann– Whitney test (Statistical Analysis System: SAS for Windows), P , 0:05 being considered significant.
Results After immunolabelling, 55 cases were confirmed as non-Hodgkins lymphoma (NHL). The other 10 cases were classified as myeloid/hystiocytic neoplasia (exclusive reactivity for myeloperoxidase or lysozyme, or both). Dogs with NHL The animals in which NHL was diagnosed included the following breeds: nine boxers (16.3%); five cocker spaniels (9.1%); six German shepherd dogs (10.9%); four Rottweilers (7.2%); two Brazilian filas (3.6%); three poodles (5.4%); two minipinschers (3.6%); six (10.9%) dogs of other breeds (teckel, basset hound, Dobermann, Brazilian terrier, Great Dane, Tenerife); and 18 crossbred animals (39.7%). There were 34 males (61.8%) and 21 females (38.2%), the mean age being 6.58 years (standard deviation [SD]: 3.2). Lymphoma was multicentric in 41 dogs (74.5%), cutaneous in seven (12.7%), splenic in three (5.4%), gastrointestinal in two (3.6%), hepatic in one (1.8%) and in an isolated cervical lymph node in one (1.8%). Immunophenotype of NHLs Of the 55 cases, 40 (72.7%) were of a B immunophenotype and 12 (21.8%) a T immunophenotype. In three cases (5.4%), all with a high
209
Canine Lymphomas Table 1 Details of 55 canine lymphomas
Fig. 1. Large B-cell lymphoma. HE. £ 1000. Inset: Immunolabelling of CD79a. £ 100.
histological grade, no reaction occurred for either B- or T-cell markers, and the phenotype therefore remained undetermined. The B-cell lymphomas were low grade in nine cases ð9=55 ¼ 16:3%Þ (one lymphocytic, six lymphoplasmacytic, one follicular and one unclassified). High histological grade was found in 31 cases ð31=55 ¼ 56:3%Þ (Fig. 1). Most low grade B-cell lymphomas showed plasmacytic differentiation ð6=9 ¼ 66:6%Þ; while plasmablastic differentiation was seen in 12 of 31 large B-cell lymphomas (38.7%). The only case diagnosed as follicular lymphoma was graded II (8– 15 blasts per £ 400 field). Of the 12 T-cell lymphomas, one ð1=55 ¼ 1:8%Þ had a low histological grade and 11 ð11=55 ¼ 20%Þ a high grade (Fig. 2). One of these had the morphology of anaplastic large cell lymphoma. The seven cutaneous presentations were of B-cell phenotype in three cases and T-cell
Fig. 2. Peripheral T-cell lymphoma. HE. £ 1000. Inset: Immunolabelling of CD3. £ 100.
Immunophenotype
Number of cases (and % of 55)
Number (and %) of cases þ ve for p53
Number of cases with stated morphology
B cell
40 (72.7)
24 (60)
Low grade histologyp—9 High grade histology†—31
T cell
12 (21.8)
7 (58.3)
Low grade histology—1 High grade histology‡—11
Non-Bnon-T
3 (5.4)
ND
Large cell lymphomas—3
ND, not done. p Types: lymphoplasmacytic (n ¼ 6); lymphocytic (n ¼ 1); follicular, grade II (n ¼ 1); unclassified ðn ¼ 1Þ. † Diffuse large B-cell lymphomas ðn ¼ 31Þ. ‡ Large T-cell lymphomas ðn ¼ 11Þ.
phenotype in four. There was no statistically significant age effect in relation to low and high grade histological patterns ðP ¼ 0:73Þ: These data are summarized in Table 1. Expression of p53 in NHLs Immunolabelling for p53 protein was attempted only in the 52 cases in which the immunophenotype could be determined. Reactivity occurred in 24/40 (60%) B-cell lymphomas: in 14/40 (35%) labelling was scored 2 and in 10 (25%) labelling was scored 1 (Fig. 3). Of the T-cell lymphomas, all seven cases positive for p53 protein (58.3%) were scored 2. There was no significant difference in
Fig. 3. Nuclear immunolabelling of p53 in a large B-cell lymphoma. £ 1000.
210
F.A.R. Sueiro et al.
p53 expression between B- and T-cell lymphomas, either when negative vs positive groups were tested ðP ¼ 1:0Þ; or when scores of 1 vs 2 were tested ðP ¼ 0:066Þ: A significant difference in p53 immunoexpression was noted in respect of low and high grade histological patterns ðP ¼ 0:01Þ: The mean age of animals with score 0 was 5.38 years (SD: 2.50), and of animals with score 1 or 2 was 7.48 years (SD: 3.35). This difference was statistically significant ðP ¼ 0:02Þ: Eight of nine boxers and three of four Rottweillers with NHL showed positive expression of p53 protein. This high frequency was not observed in the other pure breeds, occurring in only two out of six German shepherd dogs, two out of five cocker spaniels and one out of three poodles.
Discussion The use of the WHO International Histological Classification of Hematopoietic Tumors of Domestic Animals, which is based on the recently published human nomenclature, may facilitate communication between medical and veterinary pathologists, both for diagnostic and research purposes. The predominance of the multicentric presentation of lymphoma (41 dogs; 74.5%) accords with the observation of Teske (1994). The present study showed that 72.7% of CLs represented B-cell neoplasia and 21.8% T-cell neoplasia; the results thus resembled those of a cooperative study of 1378 human NHLs, in which the corresponding percentages were 88 and 12 (Non-Hodgkin’s Lymphoma Classification Project, 1997). A predominance of B immunophenotype in CL was also reported by others. Thus, Teske et al. (1994b) found 58.9% B- and 37.9% T-cell lymphomas in 95 cases. Caniatti et al. (1996) found 14 Band three T-cell lymphomas, and one non-B-non-T lymphoma in a series of 18 cases. Fournel-Fleury et al. (1997) studied 47 frozen tissue samples of CL, from which 29 (61.7%) were B and 18 (38.3%) T; they also studied 92 paraffin wax-embedded tissue samples of CL, in which 68 (73.9%) proved to be B and 24 (26.1%) T. Four of the seven cutaneous lymphomas in the present study had a T-cell phenotype and three a B-cell phenotype. In a study of seven cutaneous lymphoid neoplasms, Day (1995) found a T-cell phenotype (CD3 positive) in three cases, the remainder being negative for CD3 and for a panel of B-cell markers; however, anti-CD79a was not used. The B-cell marker CD79a and the T-cell marker CD3 were used on all tissues (paraffin-wax embedded) in the present study. The high
reliability of these markers for the study of CL in routinely processed tissues was corroborated by our experience and that of others (Milner et al., 1996). Immunophenotyping is of major importance, since a large proportion of T-cell lymphomas may have B-cell morphology in HE-stained sections of formalin-fixed tissue (Teske et al., 1994b). In three (5.4%) of our cases no immunolabelling was detected; this may have been due to fixation problems or to true lack of cell phenotype. A small proportion of non-B-non-T cases was also reported by the above-mentioned authors. However, unlike Thomas et al. (2001), we found no cases in which co-expression of CD79a and CD3 occurred. The value of myeloid/histiocytic markers in studying undifferentiated neoplasms, such as lymphomas, was reported by Vassallo et al. (1993). In the present study these markers detected 10 cases of myeloid/hystiocytic neoplasia incorrectly classified as lymphomas by histological methods. The large B-cell lymphoma was the type most frequently encountered (31/55; 56.3%), as in human patients (30.6%; Non-Hodgkin’s Lymphoma Classification Project, 1997). In a study of 72 CL, Weller et al. (1980) classified 40 cases (55.5%) as “histiocytic”-type, according to the Rappaport classification (Rappaport, 1966). In 176 CLs, Greenlee et al. (1990) found that 83 (47.2%) were of the centroblastic type, and 45 (25.6%) of the immunoblastic type, according to the Kiel classification (Lennert, 1978). Teske et al. (1994b), using the Kiel classification in an examination of 95 cases, found that 54% were of the centroblastic ðn ¼ 47Þ or immunoblastic ðn ¼ 5Þ B-cell immunophenotype. Among 92 immunophenotyped CLs, Fournel-Fleury et al. (1997) classified 38 cases as centroblastic (41.3%) and 13 as immunoblastic (14.1%). The comparatively high frequency of lymphoplasmacytic lymphoma among our B-cell lymphomas ð6=55 ¼ 10:9%Þ contrasts with the much lower reported frequency in human NHLs (1.2%; Non-Hodgkin’s Lymphoma Classification Project, 1997). On the other hand, we found only one case of follicular lymphoma (1.8%), in contrast to 22.1% reported in human NHLs (Non-Hodgkin’s Lymphoma Classification Project, 1997). Weller et al. (1980) reported that seven of 72 CLs (9.7%) were of the “nodular” type, according to the Rappaport (1966) classification. Greenlee et al. (1990), using the Kiel classification in studying 176 CLs, found that six were lymphoplasmacytic or lymphoplasmacytoid (3.4%) and that 18 were centroblasticcentrocytic (10.2%). Teske et al. (1994b) reported
211
Canine Lymphomas
one immunocytoma (1.1%) and one centroblasticcentrocytic (1.1%) CL among 92 CLs. In the study by Fournel-Fleury et al. (1997), 2/92 cases were classified as lymphoplasmacytic and one as centroblastic-centrocytic. These differences may have been due to geographical variations or to the use of different diagnostic criteria. The CM-1 polyclonal antibody used in the study marks both mutant and wild type p53. Polyclonal antibodies are generally less specific than monoclonal antibodies, but our experience with p53 monoclonal antibodies used in human pathology was that they did not react with canine tissues. As the half-life of mutant p53 is long (up to 12 h), while that of the wild type is comparatively short (30 min), it may be assumed that the amount of wild type p53 labelled by the polyclonal antibody was negligible (Batsakis and El-Naggar, 1995). Other authors, using CM-1, showed overexpression of p53 in canine tumours, including CL (Gamblin et al., 1997; Jaffe et al., 2000). Our survey identified nuclear labelling for p53 (scores 1 and 2) in 31/52 cases (59.6%). This proportion is greater than that of 40% found by Gamblin et al. (1997), but in that study the number of CLs was only 15, and no immunophenotyping was performed. Veldhoen et al. (1998), who analysed DNA sequences of the p53 gene in CL, found mutations of this gene in three of eight cases (37.5%). Reactivity for p53 was present in 60% of our B-cell lymphomas. In human high-grade B-cell lymphomas, p53 overexpression was present in 24/ 93 cases (26%) in one study and in 9/32 (28%) in another (Piris et al.; 1994; Pagnano et al., 2002). Overexpression of p53 in human NHL was associated with decreased survival (Piris et al., 1994; Pagnano et al., 2001, 2002). In CL, however, it is difficult to relate p53 to survival because only limited numbers of cases are available and treatment varies. Although not statistically significant, there was a tendency in the present study for higher degrees of p53 immunolabelling to be shown by T-cell lymphomas than by B-cell lymphomas. As Tcell lymphomas were shown to differ from B-cell lymphomas in being characterized by shorter survival periods and higher mortality rates (Greenlee et al., 1990; Teske et al., 1994a) it might be speculated that overexpression of p53 plays a role in unfavourable outcome. Overexpression of p53 in CLs was significantly associated with CLs of high histological grade (score 2; P ¼ 0:01), which accords with their aggressive clinical behaviour. In the present study and others, the significantly increased mean age of cases showing p53 positivity (scores 1 and 2) supports the notion
(Warner, 2002) that genetic mutations tend to accumulate with increasing age, while DNA repair becomes less efficient, thus enhancing susceptibility to tumours. The importance of genetic influence in lymphomagenesis is supported by the comparatively high proportion of boxers and Rottweilers showing p53 overexpression, as also mentioned in earlier reports (Priester, 1967; Teske, 1994). Our data on CLs are in many aspects (morphology, immunophenotype and p53 overexpression) comparable with findings in human lymphomas. These data emphasize the potential value of CL as a model for the pathogenesis and treatment of human lymphomas.
Acknowledgments The authors thank the Statistics Service, Research Committee of the Faculty of Medical Sciences, State University of Campinas (UNICAMP), for help in the analysis of data. They also thank Prof. Dr L. S. Queiroz for reviewing the manuscript.
References Ba¨ckgren, A. W. (1965). Lymphatic leucosis in dogs. Acta Veterinaria Scandinavica, 6, 16S–26S. Batsakis, J. and El-Naggar, A. (1995). p53: Fifteen years after discovery. Advances in Anatomic Pathology, 2, 71–88. Caniatti, M., Roccaabianca, P., Scanziani, E., Paltrinieri, S. and Moore, P. F. (1996). Canine lymphoma: immunocytochemical analysis of fine needle aspiration biopsy. Veterinary Pathology, 33, 204–212. Day, M. J. (1995). Immunophenotypic characterization of cutaneous lymphoid neoplasia in the dog and cat. Journal of Comparative Pathology, 112, 79 –96. Dorn, R. C., Taylor, D. O. N. and Hibbard, H. H. (1967). Epizootiological characteristics of canine and feline leukemia and lymphoma. American Journal of Veterinary Research, 28, 933 –1001. Fournel-Fleury, C., Magnol, J. P., Bricaire, P., Marchal, T., Chabanne, L., Delverdier, P. A., Bryon, P. A. and Felman, P. (1997). Cytohistological and immunological classification of canine malignant lymphomas: comparison with human non-Hodgkin’s lymphomas. Journal of Comparative Pathology, 117, 35 –59. Gamblin, R. M., Sagartz, J. E. and Couto, C. G. (1997). Overexpression of p53 tumor suppressor protein in spontaneously arising neoplasms of dogs. American Journal of Veterinary Research, 58, 857– 863. Greenlee, P. G., Filippa, D. A., Quimby, F. W., Patnaik, A. K., Calvano, S. E., Matus, R. E., Kimmel, M., Hurvitz, A. I. and Lieberman, P. H. (1990). Lymphoma in dogs: a morphologic, immunologic, and clinical study. Cancer, 66, 480–490.
212
F.A.R. Sueiro et al.
Haga, T., Yokomori, K., Nakayama, H., Hayashi, T., Goto, N., Takahashi, R. and Fujiwara, K. (1988). Canine and feline lymphoid and myeloid tumors encountered in Tokyo. Japanese Journal of Veterinary Science, 50, 809 –813. Hayes, H. M. (1978). The comparative epidemiology of selected neoplasms between dogs, cats, and humans. A review. European Journal of Cancer, 14, 1299 –1308. Jaffe, M. H., Hosgood, G., Taylor, H. W., Kerwin, S. C., Hedlund, C. S., Lopez, M. K., Davidson, J. R., Miller, D. M. and Paranjpe, M. (2000). Immunohistochemical and clinical evaluation of p53 in canine cutaneous mast cell tumors. Veterinary Pathology, 37, 40–46. Jaffe, E. S., Harris, N. L., Stein, H. and Vardiman, J. W. (2001). World Health Organization Classification of Tumors. Pathology and Genetics. Tumors of Haematopoietic and Lymphoid Tissues, International Agency for Research on Cancer Press, Lyon. Kaplan, H. S. (1978). Etiology of lymphoma and leukemias: role of C-type RNA viruses. Leukemia Research, 2, 253 –271. Lennert, K. (1978). Malignant Lymphomas other than Hodgkin’s Disease, Springer, New York. MacEwen, E. G. (1990). Spontaneous tumors in dogs and cats: models for the study of cancer biology and treatment. Cancer Metastasis Reviews, 9, 125 –136. Madewell, B. R. (1985). Canine lymphoma. Veterinary Clinics of North America: Small Animal Practice, 15, 709 –723. Milner, R. J., Pearson, J., Nesbit, J. W. and Close, P. (1996). Immunophenotypic classification of canine malignant lymphoma on formalin-fixed paraffin waxembedded tissue by means of CD3 and CD79a cell markers. Onderstepoort Journal of Veterinary Research, 63, 309 –313. Non-Hodgkin’s Lymphoma Classification Project (1997). A clinical evaluation of the international lymphoma study group classification of non-Hodgkin’s lymphoma. Blood, 89, 3909 –3918. Onions, D. E. (1980). RNA-dependent DNA polymerase activity in canine lymphosarcoma. European Journal of Cancer, 16, 345–350. Onions, D. E. (1984). A prospective survey of familial canine lymphosarcoma. Journal of the National Cancer Institute, 72, 909 –912. Owen, L. N. (1971). Serial transplantation of canine lymphocytic leukemia. European Journal of Cancer, 7, 525 –528. Owen, L. N., Bostock, D. E., Betton, G. R., Onions, D. E., Holmes, J., Yoxall, A. and Gorman, N. (1975). The role of spontaneous canine tumours in the evaluation of the aetiology and therapy of human cancer. Journal of Small Animal Practice, 16, 155 –162. Pagnano, K. B. B., Vassallo, J., Lorand-Metze, I., Costa, F. F. and Saad, S. T. O. (2001). p53 Mdm2, and c-myc overexpression is associated with a poor prognosis in aggressive non-Hodgkin’s lymphomas. American Journal of Hematology, 67, 84– 92.
Pagnano, K. B. B., Silva, M. D., Vassallo, J., Aranha, F. J. P. and Saad, S. T. O. (2002). Apoptosis-regulating proteins and prognosis in diffuse large B-cell nonHodgkin’s lymphomas. Acta Haematologica, 107, 29–34. Piris, M. A., Pezella, F., Martinez-Montero, J. C., Orradre, J. L., Villuendas, R., Sanchez-Geato, M., Cuena, R., Cruz, M. A., Martinez, B., Garrido, M. C., Gatter, K., Aiello, A., Delia, D., Giardini, R. and Rilke, F. (1994). p53 and bcl-2 Expression in high-grade B-cell lymphomas: correlation with survival time. British Journal of Cancer, 69, 337–341. Priester, W. A. (1967). Canine lymphoma: relative risk in the boxer breed. Journal of the National Cancer Institute, 39, 833 –845. Rappaport, H. (1966). Tumors of the hematopoietic system. In: Atlas of Tumor Pathology, section 3, Armed Forces Institute of Pathology, Washington, DC, pp. 8–97. Safran, N., Perk, K. and Eyal, O. (1992). Isolation and preliminary characterisation of a novel retrovirus isolated from a leukemic dog. Research in Veterinary Science, 52, 250– 255. Setoguchi, A., Sakai, T., Okuda, M., Minehata, K., Yazawa, M., Ishizaka, T., Watari, E., Nishimura, R., Sasaki, N., Hasegawa, A. and Tsujimoto, H. (2001). Aberrations of the p53 tumor suppressor gene in various tumors in dogs. American Journal of Veterinary Research, 62, 433 –439. Teske, E. (1994). Canine malignant lymphoma: a review and comparison with human non-Hodgkin’s lymphoma. Veterinary Quarterly, 16, 209–219. Teske, E., van Heerde, P., Rutteman, G. R., Kurzman, I. D., Moore, P. F. and MacEwen, E. G. (1994a). Prognostic factors for treatment of malignant lymphoma in dogs. Journal of the American Veterinary Medical Association, 205, 1722 –1728. Teske, E., Wisman, P., Moore, P. F. and van Heerd, P. (1994b). Histologic classification and immunophenotyping of canine non-Hodgkin’s lymphomas: unexpected high frequency of T cell lymphomas with B cell morphology. Experimental Hematology, 22, 1179 –1187. Thomas, R., Smith, K. C., Gould, R., Gower, S. M. and Binns, M. M. (2001). Molecular cytogenetic analysis of a novel high-grade canine T-lymphoblastic lymphoma demonstrating co-expression of CD3 and CD79a cell markers. Chromosome Research, 9, 649–657. Valli, V. E., Jacobs, R. M., Parodi, A. L., Vernau, W. and Moore, P. F. (2002). Histological Classification of Hematopoietic Tumors of Domestic Animals (World Health Organization International Histological Classification of Tumors of Domestic Animals), Armed Forces Institute of Pathology, Washington, DC. Van Hoosier, F. L., Stenback, W. A., Mumford, D. M., Hill, W. A., Dunn, S. C., MacDonald, E. J., MacDonald, M. C., Taylor, H. G. and Trentin, J. J. (1968). Epidemiological findings and electron microscopic observations in human leukemia and canine contacts. International Journal of Cancer, 3, 7 –16.
213
Canine Lymphomas
Van Pelt, R. W. and Connor, G. H. (1968). Clinicopathological survey of malignant lymphoma in the dog. Journal of the American Veterinary Medical Association, 152, 976 –989. Vassallo, J., Altemani, A. M., Cardinalli, I. A., Crespo, A. N., Lima, C. S. P., Eid, K. A. and Souza, C. A. (1993). Granulocytic sarcoma of the larynx preceding chronic myeloid leukemia. Pathology Research and Practice, 189, 1084 –1086. Veldhoen, N., Stewart, J., Brown, R. and Milner, J. (1998). Mutations of the p53 gene in canine lymphoma and evidence for germ line p53 mutations in the dog. Oncogene, 16, 249–255.
Warner, H. R. (2002). Recent progress in understanding the relationship among aging, replicative senescence, cell turnover and cancer. In Vivo, 16, 393–396. Weller, R. E., Holmberg, C. A., Theilen, G. H. and Madewell, B. R. (1980). Histologic classification as a prognostic criterion for canine lymphosarcoma. American Journal of Veterinary Research, 41, 1310 –1314.
Received; November 21st; 2003 Accepted; April 5th; 2004
www.elsevier.com/locate/jcpa
Canine Lymphomas: a Morphological and Immunohistochemical Study of 55 Cases, with Observations on p53 Immunoexpression F. A. R. Sueiro, A. C. Alessi* and J. Vassallo Department of Anatomical Pathology, Faculty of Medical Sciences, State University of Campinas (UNICAMP), Sa˜o Paulo, P. O. Box 6111, 13081-970 Campinas, SP, and *Department of Veterinary Pathology, Faculty of Veterinary Medicine, University of Sa˜o Paulo State, Jaboticabal, Brazil
Summary The purpose was to determine the immunophenotype of canine lymphomas (CLs) classified according to the WHO nomenclature for domestic animals, and to relate these findings to the immunoexpression of p53 protein. Lymphomas were immunophenotyped with antibodies to CD79a, CD3, and p53 protein, suitable for paraffin wax-embedded tissue sections. Of 55 cases, 40 (72.7%) were of the B-cell phenotype, 12 (21.8%) of the T-cell phenotype, and three (5.4%) were non-B-non-T lymphomas. Of the 40 B-cell lymphomas, 31 were of the large B-cell type, six were lymphoplasmacytic, one lymphocytic, one follicular (grade II) and one unclassified low-grade. One of the peripheral T-cell lymphomas was morphologically consistent with anaplastic large cell lymphoma. Immunolabelling for p53 occurred in 24 B-cell and seven T-cell lymphomas. Lymphomas with high grade histology showed a significantly increased frequency of p53 positivity ðP ¼ 0:01Þ: Positivity for p53 (more than 10% positive cells) tended to be associated with the T-cell phenotype ðP ¼ 0:06Þ: Mean patient age was significantly higher in p53-positive cases ðP ¼ 0:02Þ: These data are comparable with findings in human lymphomas. q 2004 Published by Elsevier Ltd. Keywords: canine lymphomas; CD3; CD79a; p53 protein
Introduction Tumours of the haematopoietic system are among the neoplasms most frequently encountered in veterinary medicine (Madewell, 1985). Canine non-Hodgkin’s lymphoma, also known as canine lymphoma (CL), represents the third most common malignant neoplasm in dogs, the annual incidence rate being 13 to 24 per 100 000 dogs (Ba¨ckgren, 1965; Dorn et al., 1967). The mean age of dogs with CL varies from 6.3 to 7.7 years (Van Pelt and Connor, 1968; Madewell, 1985; Teske, 1994), with no obvious gender predilection (Haga et al., 1988). An association between CL and viral infection is controversial and has not been proved Correspondence to: J. Vassallo. 0021–9975/$ - see front matter doi: 10.1016/j.jcpa.2004.04.002
(Van Hoosier et al., 1968; Owen, 1971; Kaplan, 1978; Onions, 1980; Safran et al., 1992). Boxers, Saint Bernards, Scottish terriers, basset hounds, Airedale terriers and bulldogs are reported to be at increased risk, and dachshunds and Pomeranians at reduced risk (Teske, 1994). CLs, like human lymphomas (Jaffe et al., 2001), were included in a World Health Organization (WHO) classification based on morphological and immunophenotypical criteria (Valli et al., 2002), but this classification is as yet too recent to have formed the basis for further studies. CLs have been proposed as animal models for their human counterparts (Owen et al., 1975; Hayes, 1978; MacEwen, 1990). q 2004 Published by Elsevier Ltd.
208
F.A.R. Sueiro et al.
The molecular biology of CL has been studied, but not in great depth. A family of bullmastiffs in Great Britain was described as being at high risk from lymphoma, suggesting that genetic factors play a role in pathogenesis (Onions, 1984). Aberrations of the p53 tumour suppressor gene were demonstrated in CL, and other canine neoplasms (Gamblin et al., 1997; Veldhoen et al., 1998; Setoguchi et al., 2001). In CL, however, no attempt was made to relate p53 to morphological and immunohistochemical findings. The purpose of this study was to examine immunohistochemically CL samples from the routine diagnostic laboratory at the University of Sa˜o Paulo to (1) evaluate the applicability of the WHO nomenclature, and (2) relate the immunophenotypical findings to p53 immunoexpression.
Materials and Methods Samples Formalin-fixed paraffin wax-embedded tissue samples from 65 cases in which a diagnosis of CL had been made during the period 1990– 2000 at the Department of Veterinary Pathology of the University of Sa˜o Paulo State (Jaboticabal) were examined morphologically and immunohistochemically. Haematoxylin and eosin (HE)-stained sections were reviewed and classified according to the WHO International Histological Classification of Hematopoietic Tumors of Domestic Animals (Valli et al., 2002). Immunohistochemistry Serial sections were placed on silanized slides and dewaxed. Antigen retrieval was performed after hydration and endogenous peroxidase blocking with 3% H2O2 in a steamer at 90 8C for 30 min. The slides were then incubated overnight (18 h) at 4 8C with the following primary antibodies: anti-CD79a (clone HM57; Dakopatts, Carpenteria, CA, USA), diluted 1 in 50; anti-CD3 (code A0452, rabbit polyclonal; Dakopatts), diluted 1 in 50; antimyeloperoxidase (rabbit polyclonal; Dakopatts), diluted 1 in 400; anti-lysozyme (rabbit polyclonal; Dakopatts), diluted 1 in 500; and anti-p53 (CM1, rabbit polyclonal; Novocastra Laboratories, Newcastle, UK), diluted 1 in 100. For anti-CD79a, specific binding was detected with the CSA (Catalized Signal Amplification) peroxidase system (K1500; Dakopatts). For all other markers, the EnVision peroxidase system (K1491; Dakopatts) was used. Labelling was “visualized” with
3,30 -diaminobenzidine (D-5637; Sigma, St Louis, MO, USA) added to H2O2 3% in phosphatebuffered saline (pH 7.6). Positive controls consisted of canine reactive lymph nodes with known immunohistochemical properties. In negative controls the primary antibody was replaced by bovine serum albumin (BSA) 1% in phosphatebuffered saline, pH 7.6. Immunolabelling was evaluated microscopically. A tumour was considered to have a B immunophenotype when at least 10% of the neoplastic cells reacted exclusively with anti-CD79a, and a T immunophenotype when at least 10% exclusively expressed CD3. Immunolabelling of neoplastic cells for p53 was scored as follows: 0, none positive; 1, , 10% positive (nuclear labelling); 2, . 10% positive. Statistical Analysis The tests used were the exact Fisher test, Chi-square test or the Mann– Whitney test (Statistical Analysis System: SAS for Windows), P , 0:05 being considered significant.
Results After immunolabelling, 55 cases were confirmed as non-Hodgkins lymphoma (NHL). The other 10 cases were classified as myeloid/hystiocytic neoplasia (exclusive reactivity for myeloperoxidase or lysozyme, or both). Dogs with NHL The animals in which NHL was diagnosed included the following breeds: nine boxers (16.3%); five cocker spaniels (9.1%); six German shepherd dogs (10.9%); four Rottweilers (7.2%); two Brazilian filas (3.6%); three poodles (5.4%); two minipinschers (3.6%); six (10.9%) dogs of other breeds (teckel, basset hound, Dobermann, Brazilian terrier, Great Dane, Tenerife); and 18 crossbred animals (39.7%). There were 34 males (61.8%) and 21 females (38.2%), the mean age being 6.58 years (standard deviation [SD]: 3.2). Lymphoma was multicentric in 41 dogs (74.5%), cutaneous in seven (12.7%), splenic in three (5.4%), gastrointestinal in two (3.6%), hepatic in one (1.8%) and in an isolated cervical lymph node in one (1.8%). Immunophenotype of NHLs Of the 55 cases, 40 (72.7%) were of a B immunophenotype and 12 (21.8%) a T immunophenotype. In three cases (5.4%), all with a high
209
Canine Lymphomas Table 1 Details of 55 canine lymphomas
Fig. 1. Large B-cell lymphoma. HE. £ 1000. Inset: Immunolabelling of CD79a. £ 100.
histological grade, no reaction occurred for either B- or T-cell markers, and the phenotype therefore remained undetermined. The B-cell lymphomas were low grade in nine cases ð9=55 ¼ 16:3%Þ (one lymphocytic, six lymphoplasmacytic, one follicular and one unclassified). High histological grade was found in 31 cases ð31=55 ¼ 56:3%Þ (Fig. 1). Most low grade B-cell lymphomas showed plasmacytic differentiation ð6=9 ¼ 66:6%Þ; while plasmablastic differentiation was seen in 12 of 31 large B-cell lymphomas (38.7%). The only case diagnosed as follicular lymphoma was graded II (8– 15 blasts per £ 400 field). Of the 12 T-cell lymphomas, one ð1=55 ¼ 1:8%Þ had a low histological grade and 11 ð11=55 ¼ 20%Þ a high grade (Fig. 2). One of these had the morphology of anaplastic large cell lymphoma. The seven cutaneous presentations were of B-cell phenotype in three cases and T-cell
Fig. 2. Peripheral T-cell lymphoma. HE. £ 1000. Inset: Immunolabelling of CD3. £ 100.
Immunophenotype
Number of cases (and % of 55)
Number (and %) of cases þ ve for p53
Number of cases with stated morphology
B cell
40 (72.7)
24 (60)
Low grade histologyp—9 High grade histology†—31
T cell
12 (21.8)
7 (58.3)
Low grade histology—1 High grade histology‡—11
Non-Bnon-T
3 (5.4)
ND
Large cell lymphomas—3
ND, not done. p Types: lymphoplasmacytic (n ¼ 6); lymphocytic (n ¼ 1); follicular, grade II (n ¼ 1); unclassified ðn ¼ 1Þ. † Diffuse large B-cell lymphomas ðn ¼ 31Þ. ‡ Large T-cell lymphomas ðn ¼ 11Þ.
phenotype in four. There was no statistically significant age effect in relation to low and high grade histological patterns ðP ¼ 0:73Þ: These data are summarized in Table 1. Expression of p53 in NHLs Immunolabelling for p53 protein was attempted only in the 52 cases in which the immunophenotype could be determined. Reactivity occurred in 24/40 (60%) B-cell lymphomas: in 14/40 (35%) labelling was scored 2 and in 10 (25%) labelling was scored 1 (Fig. 3). Of the T-cell lymphomas, all seven cases positive for p53 protein (58.3%) were scored 2. There was no significant difference in
Fig. 3. Nuclear immunolabelling of p53 in a large B-cell lymphoma. £ 1000.
210
F.A.R. Sueiro et al.
p53 expression between B- and T-cell lymphomas, either when negative vs positive groups were tested ðP ¼ 1:0Þ; or when scores of 1 vs 2 were tested ðP ¼ 0:066Þ: A significant difference in p53 immunoexpression was noted in respect of low and high grade histological patterns ðP ¼ 0:01Þ: The mean age of animals with score 0 was 5.38 years (SD: 2.50), and of animals with score 1 or 2 was 7.48 years (SD: 3.35). This difference was statistically significant ðP ¼ 0:02Þ: Eight of nine boxers and three of four Rottweillers with NHL showed positive expression of p53 protein. This high frequency was not observed in the other pure breeds, occurring in only two out of six German shepherd dogs, two out of five cocker spaniels and one out of three poodles.
Discussion The use of the WHO International Histological Classification of Hematopoietic Tumors of Domestic Animals, which is based on the recently published human nomenclature, may facilitate communication between medical and veterinary pathologists, both for diagnostic and research purposes. The predominance of the multicentric presentation of lymphoma (41 dogs; 74.5%) accords with the observation of Teske (1994). The present study showed that 72.7% of CLs represented B-cell neoplasia and 21.8% T-cell neoplasia; the results thus resembled those of a cooperative study of 1378 human NHLs, in which the corresponding percentages were 88 and 12 (Non-Hodgkin’s Lymphoma Classification Project, 1997). A predominance of B immunophenotype in CL was also reported by others. Thus, Teske et al. (1994b) found 58.9% B- and 37.9% T-cell lymphomas in 95 cases. Caniatti et al. (1996) found 14 Band three T-cell lymphomas, and one non-B-non-T lymphoma in a series of 18 cases. Fournel-Fleury et al. (1997) studied 47 frozen tissue samples of CL, from which 29 (61.7%) were B and 18 (38.3%) T; they also studied 92 paraffin wax-embedded tissue samples of CL, in which 68 (73.9%) proved to be B and 24 (26.1%) T. Four of the seven cutaneous lymphomas in the present study had a T-cell phenotype and three a B-cell phenotype. In a study of seven cutaneous lymphoid neoplasms, Day (1995) found a T-cell phenotype (CD3 positive) in three cases, the remainder being negative for CD3 and for a panel of B-cell markers; however, anti-CD79a was not used. The B-cell marker CD79a and the T-cell marker CD3 were used on all tissues (paraffin-wax embedded) in the present study. The high
reliability of these markers for the study of CL in routinely processed tissues was corroborated by our experience and that of others (Milner et al., 1996). Immunophenotyping is of major importance, since a large proportion of T-cell lymphomas may have B-cell morphology in HE-stained sections of formalin-fixed tissue (Teske et al., 1994b). In three (5.4%) of our cases no immunolabelling was detected; this may have been due to fixation problems or to true lack of cell phenotype. A small proportion of non-B-non-T cases was also reported by the above-mentioned authors. However, unlike Thomas et al. (2001), we found no cases in which co-expression of CD79a and CD3 occurred. The value of myeloid/histiocytic markers in studying undifferentiated neoplasms, such as lymphomas, was reported by Vassallo et al. (1993). In the present study these markers detected 10 cases of myeloid/hystiocytic neoplasia incorrectly classified as lymphomas by histological methods. The large B-cell lymphoma was the type most frequently encountered (31/55; 56.3%), as in human patients (30.6%; Non-Hodgkin’s Lymphoma Classification Project, 1997). In a study of 72 CL, Weller et al. (1980) classified 40 cases (55.5%) as “histiocytic”-type, according to the Rappaport classification (Rappaport, 1966). In 176 CLs, Greenlee et al. (1990) found that 83 (47.2%) were of the centroblastic type, and 45 (25.6%) of the immunoblastic type, according to the Kiel classification (Lennert, 1978). Teske et al. (1994b), using the Kiel classification in an examination of 95 cases, found that 54% were of the centroblastic ðn ¼ 47Þ or immunoblastic ðn ¼ 5Þ B-cell immunophenotype. Among 92 immunophenotyped CLs, Fournel-Fleury et al. (1997) classified 38 cases as centroblastic (41.3%) and 13 as immunoblastic (14.1%). The comparatively high frequency of lymphoplasmacytic lymphoma among our B-cell lymphomas ð6=55 ¼ 10:9%Þ contrasts with the much lower reported frequency in human NHLs (1.2%; Non-Hodgkin’s Lymphoma Classification Project, 1997). On the other hand, we found only one case of follicular lymphoma (1.8%), in contrast to 22.1% reported in human NHLs (Non-Hodgkin’s Lymphoma Classification Project, 1997). Weller et al. (1980) reported that seven of 72 CLs (9.7%) were of the “nodular” type, according to the Rappaport (1966) classification. Greenlee et al. (1990), using the Kiel classification in studying 176 CLs, found that six were lymphoplasmacytic or lymphoplasmacytoid (3.4%) and that 18 were centroblasticcentrocytic (10.2%). Teske et al. (1994b) reported
211
Canine Lymphomas
one immunocytoma (1.1%) and one centroblasticcentrocytic (1.1%) CL among 92 CLs. In the study by Fournel-Fleury et al. (1997), 2/92 cases were classified as lymphoplasmacytic and one as centroblastic-centrocytic. These differences may have been due to geographical variations or to the use of different diagnostic criteria. The CM-1 polyclonal antibody used in the study marks both mutant and wild type p53. Polyclonal antibodies are generally less specific than monoclonal antibodies, but our experience with p53 monoclonal antibodies used in human pathology was that they did not react with canine tissues. As the half-life of mutant p53 is long (up to 12 h), while that of the wild type is comparatively short (30 min), it may be assumed that the amount of wild type p53 labelled by the polyclonal antibody was negligible (Batsakis and El-Naggar, 1995). Other authors, using CM-1, showed overexpression of p53 in canine tumours, including CL (Gamblin et al., 1997; Jaffe et al., 2000). Our survey identified nuclear labelling for p53 (scores 1 and 2) in 31/52 cases (59.6%). This proportion is greater than that of 40% found by Gamblin et al. (1997), but in that study the number of CLs was only 15, and no immunophenotyping was performed. Veldhoen et al. (1998), who analysed DNA sequences of the p53 gene in CL, found mutations of this gene in three of eight cases (37.5%). Reactivity for p53 was present in 60% of our B-cell lymphomas. In human high-grade B-cell lymphomas, p53 overexpression was present in 24/ 93 cases (26%) in one study and in 9/32 (28%) in another (Piris et al.; 1994; Pagnano et al., 2002). Overexpression of p53 in human NHL was associated with decreased survival (Piris et al., 1994; Pagnano et al., 2001, 2002). In CL, however, it is difficult to relate p53 to survival because only limited numbers of cases are available and treatment varies. Although not statistically significant, there was a tendency in the present study for higher degrees of p53 immunolabelling to be shown by T-cell lymphomas than by B-cell lymphomas. As Tcell lymphomas were shown to differ from B-cell lymphomas in being characterized by shorter survival periods and higher mortality rates (Greenlee et al., 1990; Teske et al., 1994a) it might be speculated that overexpression of p53 plays a role in unfavourable outcome. Overexpression of p53 in CLs was significantly associated with CLs of high histological grade (score 2; P ¼ 0:01), which accords with their aggressive clinical behaviour. In the present study and others, the significantly increased mean age of cases showing p53 positivity (scores 1 and 2) supports the notion
(Warner, 2002) that genetic mutations tend to accumulate with increasing age, while DNA repair becomes less efficient, thus enhancing susceptibility to tumours. The importance of genetic influence in lymphomagenesis is supported by the comparatively high proportion of boxers and Rottweilers showing p53 overexpression, as also mentioned in earlier reports (Priester, 1967; Teske, 1994). Our data on CLs are in many aspects (morphology, immunophenotype and p53 overexpression) comparable with findings in human lymphomas. These data emphasize the potential value of CL as a model for the pathogenesis and treatment of human lymphomas.
Acknowledgments The authors thank the Statistics Service, Research Committee of the Faculty of Medical Sciences, State University of Campinas (UNICAMP), for help in the analysis of data. They also thank Prof. Dr L. S. Queiroz for reviewing the manuscript.
References Ba¨ckgren, A. W. (1965). Lymphatic leucosis in dogs. Acta Veterinaria Scandinavica, 6, 16S–26S. Batsakis, J. and El-Naggar, A. (1995). p53: Fifteen years after discovery. Advances in Anatomic Pathology, 2, 71–88. Caniatti, M., Roccaabianca, P., Scanziani, E., Paltrinieri, S. and Moore, P. F. (1996). Canine lymphoma: immunocytochemical analysis of fine needle aspiration biopsy. Veterinary Pathology, 33, 204–212. Day, M. J. (1995). Immunophenotypic characterization of cutaneous lymphoid neoplasia in the dog and cat. Journal of Comparative Pathology, 112, 79 –96. Dorn, R. C., Taylor, D. O. N. and Hibbard, H. H. (1967). Epizootiological characteristics of canine and feline leukemia and lymphoma. American Journal of Veterinary Research, 28, 933 –1001. Fournel-Fleury, C., Magnol, J. P., Bricaire, P., Marchal, T., Chabanne, L., Delverdier, P. A., Bryon, P. A. and Felman, P. (1997). Cytohistological and immunological classification of canine malignant lymphomas: comparison with human non-Hodgkin’s lymphomas. Journal of Comparative Pathology, 117, 35 –59. Gamblin, R. M., Sagartz, J. E. and Couto, C. G. (1997). Overexpression of p53 tumor suppressor protein in spontaneously arising neoplasms of dogs. American Journal of Veterinary Research, 58, 857– 863. Greenlee, P. G., Filippa, D. A., Quimby, F. W., Patnaik, A. K., Calvano, S. E., Matus, R. E., Kimmel, M., Hurvitz, A. I. and Lieberman, P. H. (1990). Lymphoma in dogs: a morphologic, immunologic, and clinical study. Cancer, 66, 480–490.
212
F.A.R. Sueiro et al.
Haga, T., Yokomori, K., Nakayama, H., Hayashi, T., Goto, N., Takahashi, R. and Fujiwara, K. (1988). Canine and feline lymphoid and myeloid tumors encountered in Tokyo. Japanese Journal of Veterinary Science, 50, 809 –813. Hayes, H. M. (1978). The comparative epidemiology of selected neoplasms between dogs, cats, and humans. A review. European Journal of Cancer, 14, 1299 –1308. Jaffe, M. H., Hosgood, G., Taylor, H. W., Kerwin, S. C., Hedlund, C. S., Lopez, M. K., Davidson, J. R., Miller, D. M. and Paranjpe, M. (2000). Immunohistochemical and clinical evaluation of p53 in canine cutaneous mast cell tumors. Veterinary Pathology, 37, 40–46. Jaffe, E. S., Harris, N. L., Stein, H. and Vardiman, J. W. (2001). World Health Organization Classification of Tumors. Pathology and Genetics. Tumors of Haematopoietic and Lymphoid Tissues, International Agency for Research on Cancer Press, Lyon. Kaplan, H. S. (1978). Etiology of lymphoma and leukemias: role of C-type RNA viruses. Leukemia Research, 2, 253 –271. Lennert, K. (1978). Malignant Lymphomas other than Hodgkin’s Disease, Springer, New York. MacEwen, E. G. (1990). Spontaneous tumors in dogs and cats: models for the study of cancer biology and treatment. Cancer Metastasis Reviews, 9, 125 –136. Madewell, B. R. (1985). Canine lymphoma. Veterinary Clinics of North America: Small Animal Practice, 15, 709 –723. Milner, R. J., Pearson, J., Nesbit, J. W. and Close, P. (1996). Immunophenotypic classification of canine malignant lymphoma on formalin-fixed paraffin waxembedded tissue by means of CD3 and CD79a cell markers. Onderstepoort Journal of Veterinary Research, 63, 309 –313. Non-Hodgkin’s Lymphoma Classification Project (1997). A clinical evaluation of the international lymphoma study group classification of non-Hodgkin’s lymphoma. Blood, 89, 3909 –3918. Onions, D. E. (1980). RNA-dependent DNA polymerase activity in canine lymphosarcoma. European Journal of Cancer, 16, 345–350. Onions, D. E. (1984). A prospective survey of familial canine lymphosarcoma. Journal of the National Cancer Institute, 72, 909 –912. Owen, L. N. (1971). Serial transplantation of canine lymphocytic leukemia. European Journal of Cancer, 7, 525 –528. Owen, L. N., Bostock, D. E., Betton, G. R., Onions, D. E., Holmes, J., Yoxall, A. and Gorman, N. (1975). The role of spontaneous canine tumours in the evaluation of the aetiology and therapy of human cancer. Journal of Small Animal Practice, 16, 155 –162. Pagnano, K. B. B., Vassallo, J., Lorand-Metze, I., Costa, F. F. and Saad, S. T. O. (2001). p53 Mdm2, and c-myc overexpression is associated with a poor prognosis in aggressive non-Hodgkin’s lymphomas. American Journal of Hematology, 67, 84– 92.
Pagnano, K. B. B., Silva, M. D., Vassallo, J., Aranha, F. J. P. and Saad, S. T. O. (2002). Apoptosis-regulating proteins and prognosis in diffuse large B-cell nonHodgkin’s lymphomas. Acta Haematologica, 107, 29–34. Piris, M. A., Pezella, F., Martinez-Montero, J. C., Orradre, J. L., Villuendas, R., Sanchez-Geato, M., Cuena, R., Cruz, M. A., Martinez, B., Garrido, M. C., Gatter, K., Aiello, A., Delia, D., Giardini, R. and Rilke, F. (1994). p53 and bcl-2 Expression in high-grade B-cell lymphomas: correlation with survival time. British Journal of Cancer, 69, 337–341. Priester, W. A. (1967). Canine lymphoma: relative risk in the boxer breed. Journal of the National Cancer Institute, 39, 833 –845. Rappaport, H. (1966). Tumors of the hematopoietic system. In: Atlas of Tumor Pathology, section 3, Armed Forces Institute of Pathology, Washington, DC, pp. 8–97. Safran, N., Perk, K. and Eyal, O. (1992). Isolation and preliminary characterisation of a novel retrovirus isolated from a leukemic dog. Research in Veterinary Science, 52, 250– 255. Setoguchi, A., Sakai, T., Okuda, M., Minehata, K., Yazawa, M., Ishizaka, T., Watari, E., Nishimura, R., Sasaki, N., Hasegawa, A. and Tsujimoto, H. (2001). Aberrations of the p53 tumor suppressor gene in various tumors in dogs. American Journal of Veterinary Research, 62, 433 –439. Teske, E. (1994). Canine malignant lymphoma: a review and comparison with human non-Hodgkin’s lymphoma. Veterinary Quarterly, 16, 209–219. Teske, E., van Heerde, P., Rutteman, G. R., Kurzman, I. D., Moore, P. F. and MacEwen, E. G. (1994a). Prognostic factors for treatment of malignant lymphoma in dogs. Journal of the American Veterinary Medical Association, 205, 1722 –1728. Teske, E., Wisman, P., Moore, P. F. and van Heerd, P. (1994b). Histologic classification and immunophenotyping of canine non-Hodgkin’s lymphomas: unexpected high frequency of T cell lymphomas with B cell morphology. Experimental Hematology, 22, 1179 –1187. Thomas, R., Smith, K. C., Gould, R., Gower, S. M. and Binns, M. M. (2001). Molecular cytogenetic analysis of a novel high-grade canine T-lymphoblastic lymphoma demonstrating co-expression of CD3 and CD79a cell markers. Chromosome Research, 9, 649–657. Valli, V. E., Jacobs, R. M., Parodi, A. L., Vernau, W. and Moore, P. F. (2002). Histological Classification of Hematopoietic Tumors of Domestic Animals (World Health Organization International Histological Classification of Tumors of Domestic Animals), Armed Forces Institute of Pathology, Washington, DC. Van Hoosier, F. L., Stenback, W. A., Mumford, D. M., Hill, W. A., Dunn, S. C., MacDonald, E. J., MacDonald, M. C., Taylor, H. G. and Trentin, J. J. (1968). Epidemiological findings and electron microscopic observations in human leukemia and canine contacts. International Journal of Cancer, 3, 7 –16.
213
Canine Lymphomas
Van Pelt, R. W. and Connor, G. H. (1968). Clinicopathological survey of malignant lymphoma in the dog. Journal of the American Veterinary Medical Association, 152, 976 –989. Vassallo, J., Altemani, A. M., Cardinalli, I. A., Crespo, A. N., Lima, C. S. P., Eid, K. A. and Souza, C. A. (1993). Granulocytic sarcoma of the larynx preceding chronic myeloid leukemia. Pathology Research and Practice, 189, 1084 –1086. Veldhoen, N., Stewart, J., Brown, R. and Milner, J. (1998). Mutations of the p53 gene in canine lymphoma and evidence for germ line p53 mutations in the dog. Oncogene, 16, 249–255.
Warner, H. R. (2002). Recent progress in understanding the relationship among aging, replicative senescence, cell turnover and cancer. In Vivo, 16, 393–396. Weller, R. E., Holmberg, C. A., Theilen, G. H. and Madewell, B. R. (1980). Histologic classification as a prognostic criterion for canine lymphosarcoma. American Journal of Veterinary Research, 41, 1310 –1314.
Received; November 21st; 2003 Accepted; April 5th; 2004