Malignant lymphoma is the most common neoplasm in

Transcription

1 J Vet Intern Med 2003;17: Feline Epitheliotropic Intestinal Malignant Lymphoma: Cases ( ) Janet K. Carreras, Micheal Goldschmidt, Martin Lamb, Robert C. McLear, Kenneth J. Drobatz, and Karin U. Sørenmo The clinical, histopathologic, and immunohistochemical features of cats with epitheliotropic intestinal malignant lymphoma (EIL) are described. Intestinal biopsy samples were reviewed by 3 pathologists to confirm the diagnosis of EIL. These samples (n ) were compared to the intestinal biopsies of normal cats (n 11), cats with inflammatory bowel disease (IBD; n 7), and cats with non-eil (n 9) for quantification and immunophenotyping of intraepithelial lymphocytes. Immunophenotypic studies were performed with CD3 and CD79a antibody stains to assess for T- and B-cell immunoreactivity, respectively. EIL biopsies had markedly more intraepithelial lymphocytes than normal intestine (NRL) and samples from cats with IBD. However, no marked difference was observed in the number of intraepithelial lymphocytes in cats with non-eil compared to cats with EIL. Regardless of the histologic diagnosis, the intraepithelial lymphocytes in all cats were small- to intermediate-sized T cells. Clinical findings and imaging studies in the cats identified minimal or nonspecific findings in affected cats. Most cats fit the typical profile of cats with IBD or alimentary malignant lymphoma. Nine of cats with EIL were treated with prednisone with or without additional chemotherapy. Four cats were refractory to chemotherapy and were euthanized within 3. months. The remaining cats had longterm survival times of 11 months or greater. The median survival time was 11 months. Additional studies are warranted to better characterize EIL and its relationship to IBD in cats and non-eil and to identify optimal treatment strategies for this disease. Key words: Alimentary; Cancer; Cats; Gastrointestinal; Lymphosarcoma. Malignant lymphoma is the most common neoplasm in the cat, and the gastrointestinal tract is the most commonly affected primary site. 1 8 The term epitheliotropism in alimentary lymphosarcoma refers to the characteristic homing of neoplastic T cells to the mucosal epithelium of the intestinal tract. This form of malignant lymphoma has been reported previously in the dog, cat, and cow but has been most extensively studied in humans, where it is classified by the Revised European and American Lymphosarcoma classification system as intestinal T- cell malignant lymphoma, with or without enteropathy (ITCL) ITCL is uncommon in humans. It has been documented to occur as a sequela to, or in conjunction with, celiac disease. However, the correlation between celiac disease and the development of ITCL remains controversial because not all cases of ITCL are associated with proven celiac disease.,11 The severity of clinical signs and histologic findings varies widely. Very mild cases may be limited to purely intramucosal infiltration of small T cells into the epithelial layer of the intestine. At the other end of the spectrum, extensive remodeling of normal intestinal architecture by pleomorphic large or anaplastic T lymphocytes may be seen. 11 Approximately 90% of malignant lymphomas in cats are classified as intermediate or high grade on histologic as- From the Department of Clinical Studies-Philadelphia, School of Veterinary Medicine, University of Pennsylvania, Veterinary Teaching Hospital (Carreras, McLear, Drobatz, Sørenmo), and the Laboratory of Pathology at the Veterinary School of the University of Pennsylvania (Goldschmidt), Philadelphia, PA; and Hoffman-LaRoche, Preclinical Research and Development, Nutley, NJ (Lamb). Reprint requests: Janet K. Carreras, VMD, Gulf Coast Veterinary Oncology, Suite 10, 1111 W Loop South, Houston, TX 77027; drcarreras@gcvs.com. Submitted June 13, 2002; Revised August 12, 2002; Accepted December 6, Copyright 2003 by the American College of Veterinary Internal Medicine /03/ /$3.00/0 sessment. 4,17,18 A previous study examined 67 cats with feline gastrointestinal malignant lymphoma. In that report, 0 of the 67 cats had lymphocytic versus lymphoblastic malignant lymphoma. Although immunophenotypic studies were not performed, the authors reported that the majority of cats with lymphocytic malignant lymphoma showed histologic characteristics of epitheliotropism. 13 To our knowledge, only 1 report of histologically confirmed feline epitheliotropic gastrointestinal tract malignant lymphoma (EIL) has been published. 9 The population of epitheliotropic neoplastic lymphocytes in that case report was intermediate-sized T cells. The cat was treated with a modified Madison-Wisconsin protocol but was euthanized after 3 weeks because of progressive disease. There are few reports assessing the immunophenotype of the normal population of intraepithelial lymphocytes in the small intestine of cats. Two recent studies investigated the immunophenotype and distribution of leukocyte subsets in healthy, specific pathogen-free (SPF) cats with no history of gastrointestinal disease. 19,20 Both studies found that the majority of the intraepithelial lymphocytes were CD3. However, it remains unknown whether or not the intraepithelial lymphocyte population in the intestines of these SPF cats reflects the population of intraepithelial lymphocytes in the intestines of normal cats. Furthermore, we are unaware of studies that have attempted to assess the number and immunophenotype of intestinal intraepithelial lymphocytes in cats with intestinal disease. The goals of this study were as follows: to identify multiple cases of EIL in cats; to describe the clinical, clinicopathologic, and imaging findings in these cats; to assess the response and outcome to chemotherapy; and to determine the immunophenotype of the neoplastic lymphocyte population. We also wanted to compare the immunophenotype and number of intraepithelial lymphocytes in cats with EIL to those of cats with each of the following histologic diagnoses: normal intestine (NRL), lymphocytic-plasmacytic inflammatory bowel disease (IBD), and intestinal malignant lymphoma without histologic evidence of epitheliotropism (LSA).

2 Feline Epitheliotropic Intestinal Malignant Lymphoma 327 Materials and Methods The surgical pathology database of the Laboratory of Pathology at the Veterinary School of the University of Pennsylvania was searched for cats with a histologic diagnosis of alimentary malignant lymphoma from 1997 to The years 1997 to 2000 were selected in the belief that there was an increased awareness by pathologists of the entity of EIL due to recently published reports in the veterinary literature. 9,14 16 Seventeen cats were selected for histologic review on the basis of a description in the original biopsy report suggesting the possible homing of lymphocytes to the epithelium of the gastrointestinal mucosa. Biopsy sections stained with hematoxylin and eosin (HE) were reviewed for histologic evidence of epitheliotropism. EIL showed extensive infiltration of the surface epithelium (enterocytes) by the neoplastic lymphoid cells often forming small aggregates (microabscesses). Slides were reviewed simultaneously by 3 pathologists (LC, MG, ML), and a consensus was required for a diagnosis of epitheliotropism to be confirmed. If a consensus was not reached, the cat was removed from the study. For comparison of the immunophenotype and number of intraepithelial lymphocytes between cats with EIL and other cats with normal and diseased intestines, intestinal biopsies of cats with each of the following diagnoses were selected at random from the biopsy database: group 1, NRL; group 2, IBD; and group 3, LSA. The biopsy database was searched for examples of each diagnosis reported over the past years, and approximately 1 biopsy sample per year was selected for review. All histologic diagnoses from the comparison groups were confirmed by 1 pathologist (ML). NRLs of cats were characterized by small numbers of lymphocytes and plasma cells within the lamina propria and variable numbers of lymphocytes showing infiltration into the overlying mucosa, whereas cats with IBD had increased numbers of lymphocytes and plasma cells within the lamina propria with a few interspersed eosinophils. Malignant lymphoma was characterized by a monomorphous infiltrate of lymphoid cells within the lamina propria and (when excisional biopsies were available) infiltration through the muscularis mucosa into the submucosa and, in some cats, the outer muscle layers. Medical records of the cats with EIL included in the study were reviewed for signalment, history, and clinical signs at the time of initial examination; results of diagnostic testing, including laboratory values before institution of therapy; thoracic and abdominal radiographs; abdominal ultrasonography; bone marrow cytology; surgical or endoscopic procedures; and type of therapy with outcome. Thoracic and abdominal radiographs and abdominal sonograms were reviewed in a blinded fashion by 1 board-certified radiologist (RCM) who was unaware of the cats clinical signs and histologic diagnoses. To evaluate the immunophenotype of the lymphocytes, -m-thick serial sections were prepared from formalin-fixed, paraffin-embedded blocks of intestinal biopsies from all groups (NRL, IBD, LSA, and EIL). The Envision PlusR system a was used, employing a rabbit polyclonal antibody against CD3 b to identify T cells and a murine monoclonal antibody against CD79a c with a CSAR amplification system d to identify B cells. The lymph nodes of cats were used as positive controls for both CD3 and CD79a staining, whereas the smooth muscles of cats were used as negative controls. The number of intraepithelial lymphocytes in a representative region of the villous and crypt epithelium encompassed by approximately 00 enterocytes was assessed (ML). The morphology of the neoplastic lymphocytes and their overall distribution in the mucosa and wall of the intestine also were determined (MG). The number of intraepithelial lymphocytes per the number of enterocytes was converted to a percentage for use in statistical analysis. The data were not normally distributed, and lymphocyte percentages for each of the groups were expressed as a median and range. The Kruskal-Wallis test was performed to identify marked differences in the lymphocyte populations in the different groups of cats (NRL, IBD, LSA, and EIL). If these differences were dramatic (P.0), pairwise comparisons were performed by the Wilcoxon rank sum test to determine which groups were significantly different. A P value.0 was considered significant. Results The original biopsy reports from 17 cats indicated possible epitheliotropism. After review by the 3 pathologists, 7 of these were excluded because a consensus regarding a diagnosis of EIL was not achieved, leaving to be included in the final analysis. None of the cats excluded from the EIL group were reassigned to the control population. All cats were domestic shorthairs, and the median age was 12. years (range, 14 years). Three were spayed females, and 7 were castrated males. The median weight of the cats at the time of presentation was 4.2 kg (range, kg), with 4 of the cats categorized as being underweight or having evidence of muscle wasting. Physical examinations were normal in 3 cats, and in 1 cat, the only abnormal finding was dehydration. Thickening of the intestines was palpated in 3 cats, 2 of which had discrete masses. One cat presented in shock due to intestinal perforation. Historic complaints at the time of presentation included diarrhea (n 9), vomiting (n 7), weight loss (n ), anorexia (n 3), decreased appetite (n 2), polyphagia (n 2), polydipsia (n 2), lethargy (n 2), and increased respiratory rate (n 1). Vomiting, diarrhea, and weight loss were present for 1 month in the majority of cats, and clinical signs tended to be progressive. All 9 cats tested were negative for feline leukemia virus (FeLV) and feline immunodeficiency virus on the basis of serologic tests. The results of clinicopathologic tests are shown in Table 1. Monocytosis, neutrophilia, and anemia were the most common abnormalities found on CBC, and abnormally high concentrations of liver enzymes were the most common abnormality found on serum biochemistry. Resting blood ammonia concentration was measured in 3 cats and was found to be abnormally high in 1 cat with a concurrent portosystemic shunt. Trypsinlike immunoreactivity was measured in 4 cats and was abnormally high in 2 cats. Serum cobalamin and folate concentrations were measured in 2 cats, and serum folate concentration was low in 1 cat. Zinc sulfate fecal flotation tests were negative in all 7 cats tested. Gastric biopsy samples were submitted for rapid urease testing in 3 cats, and all were negative. Fecal cultures were performed in 4 cats, and 1 sample produced growth of Campylobacter sp. Survey thoracic radiographs were available for 6 of the cats, and abdominal radiographs were available for 2. Thoracic radiographs showed no clinically relevant findings in any of the 6 cats. Abdominal radiographs were normal in 1 cat, and the other cat had a nodule visible in the tip of the spleen. Ultrasonographic studies of the abdomen were available for all cats. The intestine was considered normal in cats. Abnormalities seen in the remaining cats included visible masses (n 2), a thickened intestinal wall (n 1), intussusception (n 1), loss of wall layer definition (n 1), increased echogenicity in the mucosal layer (n 1), and fluid distension (n 1). Intra-abdominal lymph nodes were considered normal in 6 cats and enlarged

3 328 Carreras et al Table 1. Clinicopathologic findings in cats with epitheliotropic intestinal malignant lymphoma. Parameter Hematocrit (%) White blood cells ( 3 /L) Neutrophils ( 3 /L) Monocytes ( 3 /L) Lymphocytes ( 3 /L) Glucose (mg/dl) Blood urea nitrogen (mg/dl) Creatinine (mg/dl) Alanine transferase (U/L) Aspartate aminotransferase (U/L) Alkaline phosphatase (U/L) Albumin (g/dl) Globulin (g/dl) -Glutamyl transpeptidase (U/L) Total bilirubin (mg/dl) Mean Standard Deviation Reference Range in 4. Liver parenchyma was normal in cats and diffusely hyperechoic 4 cats. The liver of the remaining cat was diffusely and mildly hyperechoic, with small hyperechoic islands within the parenchyma. Four cats had a small amount of peritoneal effusion. One cat had a mottled spleen, and another cat had a nodule present in the tip of the spleen. Enlarged, hyperechoic kidneys were found in one cat. Of the intestinal biopsies examined, were obtained via exploratory laparotomy and via endoscopy. In the cats that had additional biopsies taken, malignant lymphoma was found in 2 of 4 liver biopsies, 2 of 2 mesenteric lymph node biopsies, 0 of 1 splenic biopsy, and 0 of 1 renal biopsy. Cytologic analyses of aspirates from the liver (n 1) and bone marrow (n 4) were negative for malignant lymphoma. Of the 4 cats with enlarged intra-abdominal lymph nodes on abdominal ultrasound, 1 had a diagnosis of malignant lymphoma confirmed on histopathology. Cytology from an intra-abdominal lymph node aspirate in the same cat was considered to show reactive changes. Three additional cats also had fine-needle aspirates of intra-abdominal lymph nodes. Two had enlarged lymph nodes on abdominal ultrasonography, and 1 had normal lymph nodes. The cytologic interpretation was reactive in the 2 cats with enlarged lymph nodes and positive for malignant lymphoma in the cat with normal-appearing intra-abdominal lymph nodes. The low number of intra-abdominal lymph node cytology results positive for malignant lymphoma (2 positive of 4 tested) may be because the neoplastic lymphocyte population in the cats with EIL consisted of small- to intermediatesized lymphocytes, which would not be considered unusual in a lymph node. In the intestinal biopsies, EIL was found in the duodenum (n 7), jejunum (n 3), ileum (n 1), and small intestine (region unspecified, n 1). Concurrent abnormalities in affected regions included lymphoid microabscesses in the mucosal epithelium (n ), villous blunting and fusion (n 3), increased plasma cell infiltrate within the lamina propria (n 3), and eosinophil infiltrate into the lamina propria (n 1). The overall distribution of the Table 2. Intestinal layers infiltrated by malignant lymphocytes in cats with epitheliotropic intestinal malignant lymphoma. Layer a No. Affected No. Assessed Epithelium Lamina propria Submucosa Muscularis externa Mesenteric adipose tissue 2 4 a Deeper tissue layers for submucosa, muscularis externa, and mesenteric adipose tissue were evaluated only for patients undergoing fullthickness excisional biopsy. neoplastic lymphocytes and their extent of invasion into the deeper tissue layers of the intestinal wall are shown in Table 2. The size of the neoplastic lymphocytes was small in 8 cats and intermediate in 2 cats. One cat had evidence of a secondary, neoplastic population of large histiocytoid cells separate and distinct from the epitheliotropic neoplastic cells. These histiocytoid cells did not stain positively for either CD3 or CD79a and were not epitheliotropic, being located entirely within the lamina propria. Follow-up information was available for 9 cats with EIL, and the median overall survival time for the 9 cats was 11 months. Standard chemotherapy protocols for feline malignant lymphoma were administered to 7 cats. Six cats received alternating single-agent chemotherapy with L-asparaginase, vincristine, methotrexate, and cyclophosphamide in addition to prednisone, and 1 cat received L-asparaginase, cyclophosphamide, vincristine, and prednisone at a tapering dosage. Four cats were refractory to chemotherapy and were euthanized 3. months after diagnosis of EIL. Three of these 4 cats had concurrent diseases including histiocytic intestinal lymphosarcoma, iris melanoma, aural mast cell tumor, lymphangiectasia, cholangiohepatitis, pancreatitis, and meningioma causing seizures. Two cats treated with chemotherapy were still alive at the time of the last follow-up, 28 and 29 months after initial diagnosis. The th cat was euthanized 23 months after initiation of chemotherapy. Only lung parenchyma was submitted for postmortem examination in this cat, and the diagnosis was metastatic carcinoma. No clinical evidence of malignant lymphoma was found at the time of euthanasia. The owners of 2 cats declined chemotherapy but approved single-agent therapy with prednisone. One cat survived 11 months. The other cat was started on a novel protein diet in addition to the prednisone and was alive 28 months after initial diagnosis. The results of immunophenotypic analysis with antibodies to CD3 and CD79a showed that the cats with EIL and the 7 cats with LSA had T-cell malignant lymphoma (ie, stained positive for CD3) (see Fig 1). The intraepithelial lymphocytes in all cats (NRL, IBD, LSA, and EIL) were CD3 (T-cell immunophenotype). The neoplastic lymphocytes in the 7 randomly selected cats with nonepitheliotropic intestinal LSA also were T cells. The neoplastic lymphocytes were small in size in 4 of 7, intermediate in size in 2 of 7, and blastic in 1 of 7 cats with LSA. The results of immunophenotyping and the magnitude of intraepithelial

4 Feline Epitheliotropic Intestinal Malignant Lymphoma 329 Table 3. Comparison of magnitude and immunophenotype of intraepithelial lymphocytes from intestinal biopsies of cats. Group N a Mean b Median b Range b CD3 CD79a NRL IBD LSA EIL NRL, normal feline intestine; IBD, inflammatory bowel disease; LSA, malignant lymphoma; EIL, epitheliotropic intestinal malignant lymphoma. a N, number of intestinal biopsies examined in each group. b Numbers presented as percentages calculated from number of intraepithelial lymphocytes per enterocytes. lymphocyte invasion in all 4 groups of cats (NRL, IBD, LSA, and EIL) are shown in Table 3. There were markedly more intraepithelial lymphocytes in cats with EIL than in NRL cats (P.0002) and IBD cats (P.01). However, the increase in numbers of intraepithelial lymphocytes in the group considered epitheliotropic proved not to be statistically significant when compared to nonepitheliotropic LSA. Fig 1. (a) Photomicrograph of a section of small intestine from a cat with epitheliotropic intestinal malignant lymphoma. There is marked infiltration of the malignant lymphocyte population within the mucosal epithelium and lamina propria (hematoxylin and eosin, 12; bar 28 m). (b) Photomicrograph of a section of small intestine from the same cat as pictured in (a), but this time, stained for CD3 immunoreactivity. The malignant lymphocytes are CD3 (stain brown) and therefore T cells (CD3 A42 [Dako, Carpenteria, CA], 62.; bar 6 m). (c) Photomicrograph of a section of small intestine from the same cat as in (a) and (b). The section has been stained for CD79a immunoreactivity and is negative (lymphocytes do not stain brown). (CD79a M701 [Dako], 62.; bar 6 m). Discussion Previous reports in the veterinary literature have suggested a potential relationship between lymphocytic-plasmacytic IBD and the development of malignant lymphoma. 6,9,13,14 In this study, 1 of the cats diagnosed with EIL had concurrent lymphocytic-plasmacytic gastritis, 1 cat had concurrent gastric lymphangiectasia, and 2 cats had concurrent colitis. Furthermore, 0% of the cats had chronic clinical signs of vomiting, diarrhea, decreased appetite, or weight loss persisting for 6 months or longer. These signs are consistent with previously reported clinical signs of lymphocytic-plasmacytic IBD. In cats with IBD, however, clinical signs tend to be intermittent, whereas in this study, clinical signs were progressive in almost all affected cats. 21 There is speculation that certain forms of human celiac disease and feline lymphocytic-plasmacytic IBD actually represent low-grade intraepithelial T-cell malignant lymphoma and not a reactive T-cell proliferation.,13,22 Ninety percent of malignant lymphomas in cats are classified as intermediate or high grade on histologic assessment. 4,17,18 Of the cats with EIL in this study, 80% had a neoplastic population of small lymphocytes, and 20% had lymphocytes that were intermediate in size. Thus, on the basis of cell morphology alone and without the ability to perform clonal analysis, it may be extremely difficult to distinguish IBD from EIL. Additional studies quantifying numbers of intraepithelial lymphocytes in the intestinal tissues of cats may help better distinguish between NRL, IBD, LSA, and EIL. One cat with IBD in this report had a very high number of intraepithelial lymphocytes (lymphocyte : enterocyte ratio, 9%). For statistical purposes, this cat was considered an outlier. Humans with celiac disease also may have high numbers of intraepithelial lymphocytes. 23 Of the 9 cats for which follow-up was available, 1 cat survived 11 months, and 4 cats survived 23 months.

5 330 Carreras et al These findings are similar to an earlier study in which 29 of the 0 cats with lymphocytic malignant lymphoma were treated with a less aggressive chemotherapy protocol consisting of prednisone and chlorambucil. 13 In that study, the median survival time of 20 cats achieving complete remission was 22.8 months. 13 In another previous report of 14 cats with malignant lymphoma at various anatomic locations, those cats with CD3 tumors that were not associated with FeLV viremia had the longest survival times. 4 Two cats in this study were treated with prednisone (with or without novel protein diet), and their survival times were 11 and 28 months. Prolonged responses to prednisone monotherapy also have been described in human patients with small ITCLs confined to the mucosa that appear to have developed as a complication of celiac disease. 11 Two recent studies investigating the normal resident population of intraepithelial lymphocytes in the healthy intestines of cats were performed with SPF cats. 19,20 In these studies, the majority of intraepithelial lymphocytes were CD3. Waly et al 19 also found markedly more intraepithelial lymphocytes in the villous epithelium than in the crypt epithelium in all regions of the small intestines of cats. We also found that the distribution of intraepithelial lymphocytes was subjectively greater in the villous epithelium than in the crypt epithelium, not only in NRL cats, but also in cats with IBD, LSA, and EIL. 19 In our study, the population of feline intestinal intraepithelial lymphocytes was CD3 immunoreactive regardless of the histologic diagnosis (NRL, IBD, LSA, or EIL). The only dramatic differences in numbers of intraepithelial lymphocytes were found between NRL cats and cats with EIL and between IBD cats and cats with EIL. Immunohistochemical staining of feline intestinal tissues from more cats with NRL, IBD, LSA, and EIL would be necessary to draw additional conclusions regarding the significance of epitheliotropism. All cats with EIL had histologic evidence of neoplastic lymphocytes in the lamina propria. Five of the cats in this study had biopsies obtained endoscopically; therefore, the extent of infiltration of the neoplastic lymphocytes into the submucosal and deeper tissue layers could not be assessed because of the nature of the tissue submitted. Despite this limitation, endoscopic biopsy may allow for early detection and treatment of EIL and thus offer the opportunity for a favorable outcome with the appropriate treatment. It has been proposed that grade be considered when treating feline gastrointestinal malignant lymphoma. 13 All published cases of EIL or lymphocytic malignant lymphoma with characteristic epitheliotropism in cats have involved small- to intermediate-sized lymphocytes. 9,13 It is not known whether the grade of the malignant lymphoma will dictate what type of therapy will be indicated in cats with EIL or affect response to therapy. Dogs with T-cell malignant lymphomas have a poorer prognosis than those with B-cell malignant lymphomas. 24 A similar situation may not occur in cats with EIL because of 9 cats in this study did well in the long term, suggesting that a subset of cats with EIL with a good long-term prognosis may exist. Ours was a retrospective evaluation, the number of cats was small, and no attempt was made to correlate the severity of histologic findings with the clinical outcome. Most studies that have investigated the immunophenotype of alimentary LSA in cats have found a preponderance of B-cell tumors. 18,2,26 However, one recent study found a predominance of T-cell tumors. 8 The authors found no marked difference in survival times on the basis of immunophenotype. 8 In the current study, the neoplastic lymphocytes in the 7 randomly selected cats with nonepitheliotropic intestinal LSA were T cells. Similar to the cats with EIL, the neoplastic lymphocytes were small to intermediate in size in 6 of 7 cats with LSA. Furthermore, no statistically significant increase in the numbers of intraepithelial lymphocytes in the cats with EIL compared to those with LSA was found, and one explanation is that there is a continuum of disease from this form of T-cell LSA to EIL. Further characterization of the population of intraepithelial lymphocytes for clonality to determine whether the intraepithelial lymphocytes were a polyclonal inflammatory population or a monoclonal neoplastic infiltrate would be useful. Another possibility is that we have inadvertently selected a skewed sample of LSA cats from our database. During this study, the authors noted a discrepancy between the perceived lower number of intraepithelial lymphocytes based on HE stains and the actual higher number of intraepithelial lymphocytes as seen after immunohistochemical staining for CD3 (compare Fig 1a to b). Waly et al 19 described a similar difficulty in differentiating intraepithelial lymphocytes from enterocytes in their study investigating leukocyte subsets within the small intestine of SPF cats. They reported that CD3 cell numbers were marginally higher than [intraepithelial lymphocytes] counted in HE-stained sections...possibly because CD3-labelled cells were much easier to identify and count than HEstained lymphocytes. The consistent misinterpretation of feline intestinal intraepithelial lymphocyte populations based on HE staining underscores the importance of immunohistochemical staining of intestinal samples from cats with lymphocytic disease processes. The limitations of this study include a small sample size and retrospective design. However, our findings provide a basis for further prospective, randomized, controlled studies to establish the significance of EIL, to better elucidate the relationship between feline lymphocytic-plasmacytic IBD, LSA, and EIL, and to develop optimal interventional strategies. In addition, as our understanding of this disease broadens, cats with EIL may prove to be an excellent model for humans with ITCL. Footnotes a Envision PlusR system, Dako, Carpinteria, CA b A42 rabbit polyclonal antibody against CD3, Dako, Carpinteria, CA c M701 murine monoclonal antibody against CD79a, Dako, Carpinteria, CA d CSAR amplification system, Dako, Carpinteria, CA Acknowledgments The authors would like to acknowledge Linden E. Craig, DVM, PhD, DACVP, Department of Pathology, Veterinary Hospital of the University of Tennessee, for her contribu-

6 Feline Epitheliotropic Intestinal Malignant Lymphoma 331 tion to this research. This work was performed at the University of Pennsylvania, Veterinary Teaching Hospital, and the Laboratory of Pathology at the Veterinary School of the University of Pennsylvania, Philadelphia, PA. This work was funded by a grant from the Oncology Service Fund at the University of Pennsylvania. Portions of this work were presented at the Veterinary Cancer Society Meeting, 2001, in Baton Rouge, LA. References 1. Hardy WD. Hematopoietic tumors of cats. J Am Anim Hosp Assoc 1981;17: Dorn CR, Taylor DON, Hibbard HH. Epizootiologic characteristics of canine and feline leukemia and lymphoma. Am J Vet Res 1967;28: Meincke JE, Hobbie WV, Hardy WD. Lymphoreticular malignancies in the cat: Clinical findings. J Am Vet Med Assoc 1972;160: Vail DM, Moore AS, Ogilvie GK, et al. Feline lymphoma (14 cases): Proliferation indices, cluster of differentiation 3 immunoreactivity, and their association with prognosis in 90 cats. J Vet Intern Med 1998;12: Mauldin GE, Mooney SC, Meleo KA, et al. Chemotherapy in 132 cats with lymphoma: Vet Cancer Soc Proc 199;1: Mahony OM, Moore AS, Cotter SM, et al. Alimentary lymphoma in cats: 28 cases ( ). J Am Vet Med Assoc 199;207: Moore AS, Cotter SM, Frimberger AE, et al. A comparison of doxorubicin and COP for maintenance of remission in cats with lymphoma. J Vet Intern Med 1996;: Zwahlen CH, Lucroy MD, Kraegel SA, et al. Results of chemotherapy for cats with alimentary malignant lymphoma: 21 cases ( ). J Am Vet Med Assoc 1998;213: Krecic MR, Black SS. Epitheliotropic T-cell gastrointestinal tract lymphosarcoma with metastasis to lung and skeletal muscle in a cat. J Am Vet Med Assoc 2000;216: Chott A, Dragosics B, Radaszkiewicz T. Peripheral T-cell lymphomas of the intestine. Am J Pathol 1992;141: Schmitt-Gräff A, Hummel M, Zemlin M, et al. Intestinal T-cell lymphoma: A reassessment of cytomorphological and phenotypic features in relation to patterns of small bowel remodelling. Virchows Arch 1996;429: Murray A, Cuevas EC, Jones DB, et al. Study of the immunohistochemistry and T cell clonality of enteropathy-associated T cell lymphoma. Am J Pathol 199;146: Fondacaro JV, Richter KP, Carpenter JL, et al. Feline gastrointestinal lymphoma: 67 cases ( ). Eur J Comp Gastroenterol 1999;4: French RA, Seitz SE, Valli VEO. Primary epitheliotropic alimentary T-cell lymphoma with hepatic involvement in a dog. Vet Pathol 1996;33: Kadota K, Wada Y, Ishikawa Y, et al. T-cell lymphoma with tropism for various types of epithelium in a cow. J Comp Pathol 2001; 124: Steinberg H, Dubielzig RR, Thomson J, et al. Primary gastrointestinal lymphosarcoma with epitheliotropism in three Shar-Pei and one Boxer dog. Vet Pathol 199;32: Valli VE, Jacobs RM, Norris A, et al. The histologic classification of 602 cases of feline lymphoproliferative disease using the National Cancer Institute working formulation. J Vet Diagn Invest 2000;2: Gabor LJ, Canfield PJ, Malik R. Immunophenotypic and histological characterisation of 9 cases of feline lymphosarcoma. Aust Vet J 1999;77: Waly N, Gruffydd-Jones TJ, Stokes CR, Day MJ. The distribution of leucocyte subsets in the small intestine of healthy cats. J Comp Pathol 2001;124: Roccabianca P, Woo JC, Moore PF. Characterization of the diffuse mucosal associated lymphoid tissue of feline small intestine. Vet Immunol Immunopathol 2000;7: Baez JL, Hendrick MJ, Walker LM, et al. Radiographic, ultrasonographic, and endoscopic findings in cats with inflammatory bowel disease of the stomach and small intestine: 33 cases ( ). J Am Vet Med Assoc 1999;21: Dieter RS, Duque K. Enteropathy associated T-cell lymphoma: A case report and literature review. Wis Med J 2000;99: Massachusetts General Hospital. Case records of the Massachusetts General Hospital (case ). N Engl J Med 2001;4: Greenlee PG, Filippa DA, Quimby FW. Lymphomas in dogs: A morphologic, immunologic, and clinical study. Cancer 1990;66: Jackson ML, Wood SL, Misra V, et al. Immunohistochemical identification of B and T lymphocytes in formalin-fixed, paraffin-embedded feline lymphosarcomas: Relation to feline leukemia virus status, tumor site, and patient age. Can J Vet Res 1996;60: Holmberg CA, Manning JS, Osburn BI. Feline malignant lymphomas: Comparison of morphologic and immunologic characteristics. Am J Vet Res 1976;37: