Submitted by: Patricia Shea, DVM
Vet Sci 2022; 9:477
Kehl A, Törner, K, et al.
Chronic enteropathies are encountered almost daily in general feline practice. At present, achieving a detailed diagnosis of these diseases requires histopathology and immunohistochemistry, but these methodologies require invasive, and for many clients, prohibitively expensive, sampling techniques. Serum biomarkers (“liquid biopsy”) for various types of canine neoplasms are under development, and these tests are available for some human cancers. These biomarkers can be ions, proteins, nucleic acids, or circulating cells. The ideal tumor biomarker for any species should be noninvasive, inexpensive, facilitate early diagnosis of a specific neoplasm, serve as a prognostic indicator, and predict tumor behavior, allowing the clinician to select the best therapeutic options. Neoplasia biomarkers also should aid in the monitoring of the disease over time, including its response to therapy.
The goals of this retrospective study of pathology reports of 860 feline gastrointestinal masses, including 679 alimentary lymphomas, 122 carcinomas, 29 spindle cell tumors, 7 mast cell tumors (MCTs), and 23 polyps, were to describe the frequency of occurrence, histological characteristics, and immunohistochemistry of feline gastrointestinal tumors, and to report the results of the authors’ pilot study on two potential biomarkers of feline enteric neoplasms, the microRNAs miR-20b and miR-192. MicroRNAs (miRNAs) are small, highly conserved, noncoding RNAs that are involved in the post-transcriptional regulation of gene expression. They may be active in either suppressing the activity of a gene or enhancing it, and in the case of neoplasia, may have tumor-suppressive or oncogenic effects. Expression of miRNAs in neoplastic disease can vary markedly between different types of cancer. In dogs with intestinal T-cell lymphoma, a number of miRNAs, including miR-192, are down-regulated, and miRs 20, 18b, and 363 are up-regulated.
As has been reported in this study as well as a number of others, lymphomas are the most common feline intestinal neoplasm, and adenocarcinomas are the second most common; MCTs and sarcomas also can occur in the feline gastrointestinal tract. Polyps, which are benign tumor-like lesions, were also included in this study because they are a differential diagnosis for patients with gastrointestinal masses. Only histopathological samples were evaluated by the authors; cytological and necropsy samples were excluded. The samples included in the study were all from cats of known breed, age, and sex, and were all clearly identifiable as of gastric or intestinal origin, with a definitive diagnosis available. The biopsies were from mucosal, transmucosal, or surgically resected samples.
Of the 860 cases that fulfilled all these criteria, 307 also had tissue submitted from regional lymph nodes. Almost all (97.3%) of the tumor tissue samples in the 860 patients demonstrated malignancy; there were 23 polyps (2.7%), and one leiomyoma. Cats with alimentary lymphoma were significantly younger (median age 10 years; 25% quartile, 8 years; 75% quartile 13 years) than those with carcinomas (median age 11 years; 25% quartile, 10 years; 75% quartile, 14 years). Lymphomas mainly occurred in the stomach (n = 118/679) and small intestine (n = 310/679), MCTs only in the small intestine (7/7), and carcinomas were more frequent in the large intestine (n = 71/122). The 29 spindle cell tumors were found in a variety of sites in the gastrointestinal tract, equally common in the small and large intestine, and tended to occur in cats > 10 years, although there was a wide range of ages (2-18 years). Polyps also tended to occur in cats 10 or more years old, with a median age of 12 years, and were mostly found in the intestines (n = 20/23); 3/23 were in the stomach.
No breed predispositions to gastrointestinal cancers were identified in the present study. About 80% of the animals were Domestic Shorthairs (DSH) or other mixed breeds. Other studies, principally from the USA, found that Siamese cats were overrepresented in cases of feline alimentary neoplasia, but in this study, which utilized data from cats in Germany and other European countries, cases in Siamese cats were rare (10/860). No data regarding the clinical history, treatment, or survival time of any patients were available in the present study.
The very high percentage of samples demonstrating gastrointestinal lymphoma (79.0%) was considered by the authors to be surprising because older studies reported that alimentary lymphoma prevalence in samples of feline gastrointestinal tumors was in the 41-52.3% range. Improved diagnostic techniques as well as an increasing willingness on the part of cat owners to choose advanced diagnostics such as immunohistochemical subtyping, may have, at least in part, contributed to a higher proportion of tissue samples reading out as lymphoma. The regional lymph node was affected in 69.7% (170/258) biopsied lymphoma cases, regardless of tumor site or tumor cell morphology.
Lymphoma cell size and an average number of mitoses per high-powered microscope field (hpf) were evaluated in 472 and 452 of lymphoma cases, respectively. Medium-sized and large-cell lymphomas were the most common in the data set available. There was a significant difference in the frequency of occurrence of small-cell lymphomas based on anatomical site. The small intestine was the most common site of small-cell lymphomas; small-cell lymphomas were rare in the stomach, and none were found in the large intestinal samples evaluated in the study. Mitotic figures varied from 0-30/hpf, and mitotic count was significantly higher in the intermediate- and large-cell lymphomas than in small-cell lymphomas of the small intestine. However, the number of mitotic figures varied a lot; 42 small cell lymphomas demonstrated > 2 mitotic figures/hpf, and 11 of these had > 5 mitotic figures/hpf. On the other hand, 45.5% of the intermediate-cell and 33.7% of the large-cell lymphomas had < 6 mitotic figures/hpf.
In the present study, immunohistochemical evaluation was performed on 91 lymphomas, which is the largest number of cases in any report to date. Twelve of the 91 samples were eliminated from the analysis because they represented lymphomas from the stomach and intestine in the same patient. The majority of small intestinal lymphomas (64%) were T-cell lymphomas, but slightly more of these small intestinal T-cell lymphomas (n = 19/53) were large- or medium-sized cells rather than small-cell (n =15/53). A significant number of the small intestinal lymphomas (n = 18/53) were large- or medium-size B-cell tumors; only one small intestinal tumor composed of small B-cells was identified. A preponderance of B-cell lymphomas occurred in the stomach and large intestine (n = 20/26) rather than in the small intestine. Most of the medium-cell and large-cell gastrointestinal lymphomas (68.4%) were B-cell tumors.
In evaluating the many lymphoma cases, the authors found some cases that did not fit into the current, common diagnostic system comprising two types: indolent, small-cell, low mitotic count (<2/hpf) T-cell lymphomas, and intermediate or aggressive medium- to large-cell, higher mitotic count (>2/hpf) B-cell lymphomas. Three cases of small-cell B-cell lymphoma were found, one in the small intestine, and two in the stomach. More medium- to large-cell T-cell lymphomas were found than small-cell T-cell lymphomas in this study, principally in the small intestine (19/53 lymphoma cases) and a few in the large intestine (3/9 lymphoma cases). No lymphoma cases negative for both T- and B-cell markers, or cases of T-cell rich B-cell lymphoma, were identified in the present study. There is still no prognostically relevant grading system for feline gastrointestinal lymphomas. Criteria such as tumor cell size, mitotic count, and immunological phenotype are needed to facilitate the workup and treatment of each patient.
Regarding gastrointestinal carcinomas (n = 122), most (58.1%) were found in the large intestine; 32.0% were present in the small intestine, only 3 (2.5%) in the stomach, and in 9 of the biopsies, the precise intestinal location was either not available or not reported. Nearly 90% of these tumors demonstrated transmural infiltrative growth into the muscularis mucosa, submucosa, muscularis, serosa, and adjacent mesenteric fat. Surgical margins could not be evaluated in 31 of the samples as only pieces of the tumor were submitted. In the remaining 91 biopsies, tumor margins were disease-free in only 14 cases; in the remaining 77, surgical margins were affected. Metastasis of the carcinoma to a regional lymph node was identified in 25/41 (61%) of cases in which a lymph node was submitted. The rate of metastasis to regional lymph nodes was not significantly different between small intestinal carcinomas and carcinomas of the large intestine.
Of the 29 spindle cell tumors included in the study, most were found in the small (n = 6) and large intestines (n =10), and 5 were present in the stomach. Eight of these tumors came from an intestinal site that could not be more specifically defined anatomically. Tumor types included leiomyosarcoma, hemangiosarcoma, leiomyoma, neurofibrosarcoma, spindle cell sarcoma without further differentiation, and pleomorphic sarcoma. Most of these tumors (n = 21) were incompletely excised, and all sarcomas (28/29) were moderately inflamed with necrotic areas. No gastrointestinal stromal tumors were found in the study group.
Over half of the polyp cases (12/23) were found in the large intestine, whereas in previously published literature, polyps were principally identified in the stomach and small intestine. None of the polyps evaluated in the present study had any obvious signs of transformation into malignant carcinoma. The age range of those cats with large intestinal polyps and large intestinal carcinoma was similar. Mast cell tumors were present in only 7/860 cats, confirming the suggestion in current literature that these are rare. Special stains and immunohistochemistry were required to identify these tumors. All the MCTs in the study cohort were found in the small intestine, although large intestinal MCTs in cats have been reported elsewhere. The second portion of the study focused on the expression of two miRNAs, miR-20b, likely pro-oncogenic, and miR-192, probably tumor-suppressive, that could serve as serum biomarkers of neoplasia in feline patients. Expression of these two miRNAs was evaluated in 5 randomly selected carcinomas, 6 large-cell B-cell lymphomas (3-14 mitoses/hpf), and 5 small cell T-cell lymphomas (1-4 mitoses/hpf), all from the small intestine, and these findings were compared to miR-20b and miR-192 levels in 5 controls not exhibiting inflammation or a tumor. All the controls had a slightly lower expression of miR-20b than that found in any of the tumor samples. In each of the tumor groups, one cat had very high miR-20b levels: an 8-year-old Russian Blue with a tubular carcinoma, a B-cell lymphoma with 5 mitotic figures/hpf in an 11-year-old DSH, and a T-cell lymphoma with 1 mitosis/hpf from a 7-year-old DSH. The expression of miR-192 was also lower in the controls than in the carcinomas and B-cell lymphomas. However, there was no difference in miR-192 expression between the T-cell lymphoma group and the controls. The cats with the highest expression of miR-192 in the carcinoma and B-cell lymphoma groups were the same ones that had the highest miR-20b expression as well. The cat with the highest miR-20b expression in the T-cell lymphoma group had a very low level of miR-192.
Future studies of these and other potentially relevant miRNAs in feline patients will need to involve more cases and more miRNAs, including miRNAs other than miR-20 and miR-192 identified as misregulated in canine lymphomas. Serum samples from a much larger population of cats will also need analysis to determine profiles of these biomarkers in health and disease, and to find out if these biomarkers have diagnostic utility in cases of feline gastrointestinal neoplasia.