Feline Intestinal T-Cell Lymphoma: Assessment of Morphologic and Kinetic Features in 30 Cases

Abstract

In this study, 30 feline intestinal T-cell lymphomas (ITCLs) from 77 cats with gastrointestinal lymphoma were evaluated. Neoplastic lesions were composed predominantly of small (n = 21) or medium to large (n = 9) anaplastic cells. Different patterns of tumor growth were observed. A starry-sky pattern was evident in 7 cases (23.3%), fibrosis in 18 cases (60%), and neovascularization in 19 cases (63.3%). The cell proliferation index (assessed by MIB1 [mindbomb homolog 1] immunohistochemistry) ranged from 0.21% to 66.91%. Mean MIB1 index was 3.49% within the first 33rd percentile, 18.31% within the second 33rd percentile, and 40.16% within the third 33rd percentile (P < 0.000001). Microscopic and kinetic features provided evidence that ITCL in cats exhibits a spectrum in cytological composition and growth patterns that could putatively reflect differences in biologic behavior.

Lymphoma is the most common hematopoietic neoplasm of the cat 8 and represents one third of all neoplasms in this species. 1 In 2002, the World Health Organization (WHO) proposed a new classification system for domestic animal lymphomas, 7 which was adapted from the human REAL (Revised European American Lymphoma) classification 5 in addition to the consolidated Kiel classification 2 and the Working Formulation (WF) of the National Cancer Institute. 3 With respect to previous systems, the WHO-REAL classification provides no separation between low, medium, or high malignancy grade tumors, except for follicular neoplasms. The identification of B- or T-cell lineage is critical, 6 followed by further classification of B-cell lymphomas on the basis of their morphological characteristics. In contrast, T-cell lymphomas are classified on the basis of their anatomical location. In this study, primitive lymphomas of the gastroenteric tract, in particular those deriving from T cells, were evaluated.

The alimentary form of lymphoma is the most common in cats and represents 31% of all feline lymphomas. 8 The current WHO-REAL classification 7 tends to group all intestinal lymphomas derived from T lymphocytes in a single class (intestinal T-cell lymphoma [ITCL]), suggesting that all of these neoplasms have similar clinical characteristics and pathological patterns. This retrospective study aimed to verify the homogeneity of the ITCL class in 30 archived pathological tissue samples collected from cats diagnosed with intestinal T-cell lymphoma.

The alimentary lymphoma samples were collected and archived during 20 years of diagnostic service in the Pathology Division of the Department of Veterinary Public Health and Animal Pathology of the University of Bologna, Italy. Specimens were collected from the intestine and mesenteric lymph nodes of affected cats during necropsy or surgical biopsy between 1984 and 2004. These specimens were fixed in 10% phosphate buffered formalin (pH 7.2), processed routinely, embedded in paraffin, and sectioned at 5 μm. Sections were stained with hematoxylin and eosin (HE). Immunohistochemistry was performed on replicate tissue sections for cluster of differentiation 3 (CD3) and CD79 expression to confirm the diagnosis of T-cell lymphoma and to select appropriate cases for study.

On HE-stained sections, the size of lymphocytes was estimated relative to measurement of red blood cells (RBCs). Specifically, RBC size was used as a unit of measurement compared with the nuclear size of the lymphocytes (small lymphocyte nucleus, <2 times the RBC diameter; medium to large lymphocyte nucleus, ≥2 times the RBC diameter). Other assessed parameters included the presence of neovascularization, fibrosis, and macrophages containing lymphocytic debris (starry sky pattern). Immunohistochemistry was performed on replicate tissue sections with the use of antibodies to CD3 a and CD79a b antigens to assess the cellular immunophenotype. Furthermore, sections were stained with MIB1 [mindbomb homolog 1] antibody, c which recognizes the nuclear protein Ki67 antigen that is expressed only in proliferating cells. Endogenous peroxidase activity was blocked by treatment with 3% hydrogen peroxide for 30 min. Sections were then immersed in citrate buffer (pH 6.0) and irradiated twice for 5 min in a microwave oven at 750 W to retrieve antigenicity. The sections were allowed to cool to room temperature for 20 min, and incubated overnight at 4°C with primary antibodies. A streptavidin–biotin complex kit d was used according to the manufacturer's instructions to identify sites of primary antibody binding. Diaminobenzidine was used as the chromogen. Sections were then counterstained with Papanicolaou's hematoxylin, coverslipped, e and examined microscopically.

Cases were classified as ITCL if a monomorphic infiltration of lymphocytes was present that stained positively for CD3 (>50% of cells) and negatively for CD79a. The percentage of neoplastic lymphocytes with positive MIB1 nuclear staining (proliferating cells) were assessed by image analysis and expressed as the MIB1 index. The image analysis system consisted of a light microscope f and a digital camera g connected to a personal computer from which images were acquired then analyzed and measured by dedicated software. h

After confirmation of the diagnosis of alimentary lymphoma, only those cases of neoplasia with appropriate cellular morphology and minimal artifacts were considered for further investigation. Seventy-seven cases diagnosed as feline alimentary lymphoma were initially selected for review, but only 30 cases met the criteria for acceptable cellular morphology and CD3 immunoreactivity for classification as ITCL.

The age of the cats was known for 20 ofthe 30 individuals and ranged from 2 to 15 years of age. The average age was 8.8 years. Most of the cats were domestic shorthairs; other breeds were Persian (2 cases) and Siamese (1 case). Ten cats were males or neutered males, 11 were females or spayed females, and the sex was undetermined for 9 cats.

All intestinal specimens contained a dense lymphoid infiltrate that partially or completely invaded all the layers of the intestinal wall (muscularis mucosa, submucosa, and muscularis). Lymphocytes varied from small mature cells with condensed chromatin to large immature cells with fine chromatin, irregular nuclear profiles, and prominent nucleoli. Monomorphism of the lymphoid infiltrates was further evidenced by diffuse CD3 expression.

The nuclear size of lymphocytes varied from 1 to 5 times the diameter of the RBCs, with a prevalence of small-cell lymphoma (21 of 30 cases). A starry sky appearance, because of macrophages containing debris of dead lymphocytes, was observed in 7 of 30 cases (23.3%). Fibrosis was present in 18 of 30 cases (60%), and neovascularization was present in 19 of 30 cases (63.3%).

The MIB1 index ranged from 0.21% to 66.91%. Cases within the first 33rd percentile had a mean MIB1 index of 3.49% (range: 0.21–9.26%;standard deviation [SD]:3.18%); cases within the second 33rd percentile had a mean MIB1 index of 18.31% (range: 13.4–19.74%; SD: 2.10%); and cases within the third 33rd percentile had a mean MIB1 index of 40.16% (range: 22.75–66.91%; SD: 15.73%). The statistical difference among the 3 percentile classes was highly significant (analysis of variance [ANOVA], P < 0.000001).

The nuclear size of the neoplastic lymphocytes correlated with proliferation activity. Large-cell lymphomas were more frequently represented in the high-proliferation category (first 33rd percentile, 6 small-cell and 4 large-cell lymphomas; second 33rd percentile, 4 small-cell and 6 large-cell lymphomas; third 33rd percentile, 3 small-cell and 7 large-cell lymphomas). Fibrosis occurred at a greater frequency in highly proliferative lymphoma (first 33rd percentile, evidence of fibrosis in 5 cases; second 33rd percentile, evidence of fibrosis in 6 cases; third 33rd percentile, evidence of fibrosis in 7 cases). Conversely, other morphologic parameters (starry sky pattern and neovascularization) did not vary significantly according to the proliferation percentile.

The WHO-REAL classification of ITCL is primarily based on anatomic characteristics. After immunophenotyping to separate B-cell from T-cell lymphomas, the latter are divided on the basis of anatomic location without any further evaluation of malignancy. All ITCLs are grouped in the same class, suggesting that all of these neoplasms have the same morphologic characteristics and behavior.

Nevertheless, evaluation of simple morphological parameters in this study has shown some inconsistencies within the ITCL class, even in a small series of cases. The nuclear size of the neoplastic lymphocytes ranged from 1 to 5 times the diameter of the RBCs, and the starry sky pattern was present only in 23.3% of cases. Conversely, fibrosis was evident in 60% of the cases, whereas neovascularization was observed in 63.3% of the cases. These observations suggest that these parameters could be important diagnostic features of ITCL, although these criteria are not considered in the current WHO classification scheme. 6 A similar variability in cellular morphology has also been observed in a smaller number of lymphomas involving the feline alimentary tract 4 : 6 ITCLs out of 23 alimentary lymphomas consisted of 4 small-cell lymphomas and 2 large-cell lymphomas, whereas every B-cell lymphoma was of the large-cell type.

The proliferation index (MIB1 index) revealed wide variability among intestinal lymphomas, even if they belonged to the same class. On the basis of the percentage of cycling cells, 3 subclasses were identified in this study that could putatively reflect a different biologic behavior. If cases were stratified according to the MIB1 index, the first 33rd percentile group had a low proliferation index (<10%), the second 33rd percentile had an intermediate proliferation index (10–20%), and the third 33rd percentile had a very high proliferation index (22–70%). Among the various morphologic parameters considered, only nuclear size, and to a lesser degree fibrosis, seemed to correspond with proliferative activity.

A strong variability in morphologic characteristics and proliferative activity of ITCL was observed in this study. This lack of homogeneity contrasts with the idea that a single classification of neoplastic lymphocytes should have a similar appearance and biological behavior. To confirm or refute the premise that feline ITCL varies in its biological behavior, further studies are needed that evaluate a large series of neoplasms and assesses the clinical course of disease.