Apparent feline leukemia virus-induced chronic lymphocytic leukemia and response to treatment

Abstract

Chylothorax secondary to chronic lymphocytic leukemia (CLL) was diagnosed in a feline leukemia virus (FeLV)-positive 8-year-old castrated male domestic shorthair feline. The leukemia resolved following therapy with chlorambucil, prednisone, cyclophosphamide, doxorubicin, and lomustine. To our knowledge, this is the first reported case of CLL in an FeLV-positive cat. Although a causative relationship cannot be proven, patients diagnosed with either disease may benefit from diagnostics to rule out the presence of the other concurrent condition.

An 8-year-old castrated male domestic shorthair feline presented to the Texas A&M University Small Animal Hospital for surgical thoracic duct ligation to correct suspected idiopathic chylothorax. The owner had observed occasional episodes of labored breathing and lethargy approximately 2 weeks before presentation to the local veterinarian and 6 weeks prior to presentation to TAMU.

Upon initial presentation to the local veterinarian, pleural effusion was identified as the cause for dyspnea. A thoracocentesis was performed, yielding 175 ml of opaque, pink fluid with a triglyceride measurement of 382 mg/dl (reference range 10–100 mg/dl, Idexx Laboratories). The chylous effusion had a total protein of 4.5 g/dl and a high cellularity with 68,100 white blood cells per μl. Greater than 80% of these white blood cells were small, mature lymphocytes with no obvious neoplastic characteristics. There was no growth on aerobic or anaerobic cultures. No other abnormalities were present on physical examination and the patient became significantly more comfortable following thoracocentesis. Over the next several days, repeated palliative thoracocenteses were performed.

A complete blood count (CBC) obtained by the local veterinarian identified a leukocytosis of 49,000 per μl (reference range 5500–19,500 per μl) characterized by a moderate lymphocytosis of 7600 cells per μl (reference range 1800–6500 per μl) and a high number of granulocytes (32,900 granulocytes per μl [reference range 2800–13,000 per μl]; in-house CBC). On microscopic evaluation of the blood smear, the machine-read granulocytic population was judged to consist primarily of small lymphocytes. Further diagnostics included a negative feline immunodeficiency virus/feline leukemia virus enzyme-linked immunosorbent assay (ELISA) (Snap FIV/FeLV Test, Idexx Laboratories) and negative heartworm antigen and antibody ELISA tests (Idexx Laboratories).

An abdominal ultrasound was performed, identifying enlarged abdominal lymph nodes (measurements were not obtained). Fine needle aspirates of the enlarged nodes were obtained, identifying reactive lymph nodes with a left shift in the neutrophilic population. No evidence of neoplasia was identified (University of Arizona Diagnostic Laboratory).

Two separate bone marrow aspirates were performed prior to presentation to TAMU. The first bone marrow aspirate identified megakaryocytes present in adequate numbers. Both the erythroid and myeloid series were present and maturing within normal ratios. A mild granulocytic predominance was noted with a mild left shift, but no evidence of either dysplasia or lymphocytic infiltration was noted. The second bone marrow aspirate (obtained 2 weeks later) revealed a left shifted myeloid predominance with normal order and complete maturation. The myeloid:erythroid ratio was 2:1. Approximately 7% of the cells were small mature lymphocytes. An indirect immunofluorescent assay (IFA) of the second bone marrow was positive for FeLV (Idexx Laboratories).

Treatments prescribed by the local veterinarian included one capsule of rutin once every 12 h (Standard Process Cyruta Plus 325 mg capsules), 25 mg marbofloxacin (Zeniquin; Pfizer) once a day, and 5 mg prednisone once every 12 h. The case was referred to TAMU for further evaluation and treatment of the chylothorax and marked leukocytosis.

During the same time period, the owner reported one episode of inappetence that resolved within several days. Following the episode, the cat's appetite was observed to be normal. The cat's diet was changed to Hill's Prescription Diet w/d because the cat was judged to be overweight by the local veterinarian (unknown body condition score). After the diet change, the cat quickly developed diarrhea that did not resolve. The cat's average body weight in health had been around 6.5–6.7 kg; however, the cat's body weight decreased to 5.4 kg during this time.

Upon presentation to TAMU, physical examination parameters were largely unremarkable with the exception of bilateral maxillary canine extrusion and a dry, flaky hair coat. The cat weighed 4.9 kg with a body condition score of 6/9. A CBC identified a marked leukocytosis of 43,700 per μl (reference range 5500–19,500 per μl) in which approximately half of the lymphocytic cell population [absolute number of small lymphocytes equal to 13,547 per μl (reference range 1500–7000 per μl)] was characterized as small-cell lymphocytes with the remaining half characterized as intermediate in size [absolute number of intermediate lymphocytes equal to 18,354 per μl (reference range 0 per μl)]. No other abnormalities were present on the CBC (Fig. 1). Serum biochemistry revealed a blood urea nitrogen (BUN) of 16 mg/dl (5.7 μmol/l) (reference range 19–33 mg/dl [6.8–11.8 μmol/l]), creatinine of 1.4 mg/dl (123.8 μmol/l) (reference range 0.8–1.8 mg/dl [70.7–159.1 μmol/l]), and serum triglyceride equal to 61 mg/dl (0.7 mmol/l) (reference range 15–69 mg/dl [0.2–0.8 mmol/l]). These values were determined to be clinically insignificant. Urinalysis was not performed. A FeLV ELISA (ViraChek/FeLV Lab Pack, Synbiotics Corporation) was performed on whole blood, and a positive result was obtained. Serum folate was 23.5 μg/l (53.3 nmol/l) (reference range 10–22 μug/l [22.7–49.9 nmol/l]) and serum cobalamin was <150 μg/l (<0.11 μmol/l) (reference range 290–1500 μg/l [0.21–1.1 μmol/l]).

Fig 1.

Peripheral blood smear obtained upon presentation to TAMU. Notice the large white blood cells, approximately the size of the neutrophils within the same field. These cells were identified as lymphoblasts. The presence of such cells in large numbers is consistent with the diagnosis of lymphocytic leukemia.

On thoracic ultrasound, anechoic pleural fluid was identified and a therapeutic thoracocentesis was performed. A soft-tissue structure visible cranial to the heart was suspected to be mildly enlarged thymic tissue. On abdominal ultrasound, several enlarged, round, hypoechoic jejunal and mesenteric lymph nodes were identified. A hypoechoic mass was also present in the mesentery extending from the intestines toward the root of the mesentery. The small intestine was diffusely hypoechoic with an appreciable loss of wall layering noted. In addition, the spleen was judged to be enlarged measuring 1.4 cm in thickness. No other abnormalities were identified on abdominal ultrasound. Under ultrasound guidance, a splenic aspirate was obtained and submitted for pathologic evaluation.

The nucleated cell population present in the spleen consisted of predominately lymphocytes, 45% of which were intermediate in size with the remaining 55% small lymphocytes. The types of lymphocytes present in the spleen and their overall morphology were consistent with the populations present in the peripheral blood. A few non-degenerate neutrophils and eosinophils were present throughout the samples in numbers consistent with that found in peripheral blood. Rare plasma cells were seen. No infectious agents were noted.

Whole blood was submitted for polymerase chain reaction assay for antigen receptor rearrangements (PARRs) which identified a clonally expanded T-cell lymphocyte population consistent with a lymphoid malignancy (Department of Microbiology, Immunology, and Pathology, Colorado State University) (Fig. 2).

Fig 2.

Image of the PARR assay gel obtained by the Department of Microbiology, Immunology, and Pathology at Colorado State University. The PARR test is run similarly to a Western blot – a patient's serum sample is run through a gel impregnated with specific receptor antigens. Unlike other PCR tests, however, the PARR test is not simply a positive or negative test. Rather, one is looking for a single band within a test lane versus multiple or blurred lines within a lane. In this patient's test, the first lane to the far left is a positive control lane used to indicate the presence of DNA within the sample, and therefore a reliable test. Lanes two and three contain unique immunoglobulin antigen receptors that will bind with any B-cell immunoglobulins present. The two far right lanes, lanes five and six, contain T-cell receptor antigens that will bind with any T-cell antibodies present. Multiple lines or blurred areas within a lane indicate variable B- or T-cell populations. The presence of a single line in a lane indicates a clonal T- or B-cell population. The middle lane, lane four, was not used due to a defect within the gel. Lane two contains multiple lines, indicating a variable B-cell population. Contrast this with lane six in which there is a clear, bright line present at the base of the lane. The lymphocyte population present in this patient can, therefore, be characterized as a clonal T-cell population, making this a diagnosis of T-cell lymphoma. No significant data is present in lane three or five.

Based on diagnostic test results, a diagnosis of chronic lymphocytic (small-cell) leukemia (CLL) was made. In addition, the patient was FeLV-positive confirmed with both ELISA and IFA. Although intestinal biopsies were not obtained, the loss of layering detail on abdominal ultrasound along with the markedly decreased cobalamin levels were consistent with distal small intestinal disease, including infiltrative disease processes (eg, lymphoma). The patient was discharged with chlorambucil (Leukeran; GlaxoSmithKline) at a dosage of 0.42 mg/kg given by mouth three times a week and prednisone given orally at 2 mg/kg q24h. Recommendations for subcutaneous cobalamin supplementation at 250 μg weekly for 6 weeks, then once every 14 days for 6 weeks, then once every 3–4 weeks as necessary to maintain cobalamin within the normal range were sent to the referring veterinarian.

Further diagnostics that were not pursued included aspiration of the thoracic mass and the enlarged mesenteric lymph nodes as well as endoscopic intestinal biopsies. These diagnostics were not pursued due to concerns related to the patient's stability under anesthesia, which would have been required to obtain the samples.

During the first 5 months of treatment, a marked but variable decrease in both small and intermediate sized circulating lymphocytes (8064–10,640 per μl [reference range 1200–8000 per μl]) was appreciated. The cat no longer needed therapeutic thoracocentesis, was gaining weight, and doing well clinically. Six months after initial therapy, the peripheral lymphocyte count had climbed to 17,404 per μl (1200–8000 per μl Antech Diagnostics). Clorambucil was discontinued and cyclophosphamide (Cytoxan; Mead Johnson Oncology) added to the protocol at 15 mg orally once a day for 4 days. One week following cyclophosphamide therapy, the white blood cell count had increased to 28,500 per μl (reference range 3500–16,000 per μl Antech Diagnostics). A 6 ml dose of doxorubicin (Adriamycin PFS; Pfizer) was given intravenously. The white cells remained elevated near 30,000 per μl for the next 2 weeks. A single dose 10 mg dose of lomustine (CeeNu; Bristol Labs Oncology) was given orally; however, leukocytosis of 36,000 per μl (reference range 3500–16,000 per μl Antech Diagnostics) was still present.

Several weeks after completing lomustine therapy, the cat became acutely lethargic, anorexic, and poorly responsive. A severe anemia (packed cell volume of 10%) was identified by the local veterinarian. Although initially declined by the owner, the cat was given a whole-blood transfusion several days later (donor and recipient type A, post-transfusion packed cell volume of 20%). Following the transfusion, the cat began to show marked improvement clinically and was maintained on 10 mg prednisone daily.

By 10 months after initiating therapy, the cat had gained 0.9 kg and was reported to be doing quite well at home. A recheck CBC revealed a total white cell count of 5400 per μl (reference range 3500–16,000 per μl Antech Diagnostics) with no clinically significant abnormalities appreciated in the serum biochemistry.

The initially discordant FeLV test results obtained in this patient may be accounted for by a simple product failure of the Snap ELISA test. Alternately, prior investigations into similar FeLV test result patterns have raised the possibility that the p27 antigen tested for using conventional Snap ELISA tests may not be present in sufficient quantities in peripheral blood to cause a positive test. The antigen is, however, present in sufficient quantities intracellularly and can therefore be identified by IFA. ¹

Chronic lymphocytic (small-cell) leukemia (CLL) is rare but reported in felines. ^2–4 Anecdotal evidence suggests that T-cell derivative of CLL is the most common presentation ³ with clinical signs varying in severity from mild lethargy and gastrointestinal signs to more severe illness as the bone marrow and other internal organs become involved. Despite the lack of information available in veterinary literature, CLL in cats seems to be a slowly progressive disease in which patients can be expected to survive for up to 2 years or longer when treated with chlorambucil and/or prednisone. ^2,3,5 The cause for the development of CLL in humans, dogs, or cats is unknown.

The diagnosis of CLL can present a unique challenge as differentiating non-neoplastic small lymphocytes from a truly neoplastic population may be exceptionally difficult. In addition to immunophenotyping of lymphocytic cell populations using flow cytometry and analysis of antigenic expression, the polymerase chain reaction for antigen receptor rearrangements (PARRs) test can be used to determine the clonality of a lymphocytic cell population. ⁶ Lymphocytes contain unique hypervariable regions within their DNA. These areas are unique to each lymphocyte unless the cells are all derivative of a single original lymphocytic. These clonal populations share identical hypervariable regions. The PARR test identifies clonal hypervariable regions via polymerase chain reaction and can further distinguish between T- and B-cell populations. In the dog, the sensitivity of the PARR assay performed at Colorado State University is 80% with a specificity of 92%. ⁷ Although initially of limited use prognostically, ⁸ more recent canine data suggests that the use of the PARR assay to distinguish between CD8+ T-cells and CD21+ B-cells carries important prognostic significance when evaluated along with cell size. ⁷

In the cat, the sensitivity of the PARR assay is reported to be 65% with a specificity of 90% (unpublished data from Colorado State University). Although limited, recent data suggest that the PARR assay in felines may also provide some prognostic information for the clinician. T-cell populations with large granular lymphocytic (LGL) morphology may have a poorer prognosis than non-LGL T-cell or CD21+ B-cell populations, likely because these LGL populations are associated with intestinal lymphoma. ⁷ No prognostic information is available for felines with non-LGL T-cell or CD21+ B-cell populations. ⁷

To the authors' knowledge, an FeLV-induced CLL has not been reported to date. FeLV-positive cats have been demonstrated to be at increased risk for development of acute monoblastic leukemia, erythroleukemia, and acute lymphoblastic lymphoma (both aleukemic and leukemic). Additionally, these cats are increasingly susceptible to a variety of secondary diseases associated with generalized immunosuppression. ^9,10 These secondary illnesses are often fatal.

Current research of the pathophysiological mechanisms by which the FeLV retrovirus causes neoplasia specifically is centered on identifying specific tumor suppressor genes that are disrupted and/or proto-oncogenes that may be activated when pro-virus DNA from the FeLV virus is inserted into the subject's own DNA. ¹¹ This research has not yet identified a causative relationship between the DNA damage caused by the FeLV retrovirus and the development of specific neoplastic diseases. The link between the two is currently confined to the disruption of what may be thought of as generalized genetic controls of unregulated cellular replication that may arise in a number of different cell populations, leukemias being a single categorical subset. Further research into the specific mechanisms by which specific tumor populations arise is, of course, ongoing in a variety of veterinary and human medical fields.

If, in fact, FeLV does activate generic proto-oncogenes and/or render tumor-suppression genes ineffective, or both, it may be reasonable to hypothesize that there exists a causative relationship between the FeLV-positive status of this patient with the development of CLL. Despite the rarity of feline CLL and the increasingly rare occurrence of FeLV in domesticated cats in the United States, patients diagnosed with either disease may benefit from diagnostics to rule out the presence of the other concurrent condition.

One cannot ignore the possibility that the FeLV status of this particular cat had no causative relationship with the development of the concurrent CLL. The paucity of peer-reviewed literature examining CLL in feline patients, especially as is related to the pathogenesis of the development of the disease, limits what conclusions can be reasonably drawn from single patient data. Further research into CLL is necessary to determine what role FeLV infection may play in the development of this neoplastic process.