Richard Malik, BVSc, DipVetAn, MVetClinStud, PhD, FACVSc, MASM; Leslie J. Gabor, BVSc, PhD; Paul J. Canfield, BVSc, PhD, FACVSc, FRCPath
Post Graduate Foundation in Veterinary Science, University Veterinary Centre, Sydney and Diagnostic Services Laboratory, The University of Sydney
Most feline physicians would agree that lymphoma (synonym: lymphosarcoma; malignant lymphoma) [LSA] is the most common malignancy of cats. Indeed the discovery in the 1960s that some types of feline LSA were associated with a retrovirus (feline leukaemia virus [FeLV]) resulted in an explosion of knowledge concerning LSA, cats in general and virus-induced cancer.
A substantial literature has been generated in relation to this important feline malignancy, the majority of reports arising from the United Kingdom and North America, but with important contributions also from Europe. Although a very large number of papers have covered a variety of aspects of this malignancy, a comprehensive review of the literature provides an insight that LSA may take on different guises in different feline populations, whether they are separated geographically, or in time. In other words, the pattern and epidemiology of LSA appears distinct in different countries or even regions within the same country, and at different points in time. Thus, some series of feline LSA cases have had a preponderance of young cats with mediastinal involvement associated also with leukaemia, while others have been comprised mainly of older cats typically with involvement of the gastrointestinal tract, mediastinal lymph nodes and kidneys.
A retrospective analysis of feline LSA cases seen in our hospital suggested that the clinical and epidemiologic features of LSA in eastern Australia differed markedly from that which had been reported in the past for the UK and North America. Accordingly we embarked on a large prospective study of feline LSA in order to obtain a more rigorous data set, with a view to precisely defining the features of this important malignancy in Australian cats as at 1995.
This project involved myself (a feline physician with an interest in chemotherapy and infectious disease), Paul Canfield (a veterinary pathologist with an interest in comparative oncology) and an enterprising PhD student (Les Gabor). This combination of personnel was very successful in recruiting cases, and we were able to obtain data from 118 cats over an 18-month period. These cases were studies using standard clinical methods (haematology, serum biochemistry, radiology, ultrasonography and necropsy). Immunohistology was used to determine the immunophenotype of the tumours. Daria Love developed a sensitive Western blot test to detect antibodies against FIV (adapting protocols developed originally in Glasgow), while Marion Jackson and colleagues at Saskatoon used their PCR to determine whether FeLV provirus was present in tumours from affected cats. The study had two broad aims: (1) to define risk factors for the development of LSA in our cohort of Australian cats, and (2) to determine patient or tumour factors predictive of a successful outcome using a single multi-agent chemotherapy protocol.
Analysis of the data from our cohort of 118 cats was initially hindered by our inability to classify cases according to accepted anatomic schemes or staging protocols. This problem arose because many cases had either unusual combinations of tissues affected (e.g., gut, nearby mesenteric lymph nodes and one peripheral lymph node) or the involvement of atypical structures (e.g., the nasal cavity). For this reason we developed our own classification system, including separate categories for cases with mixed and atypical tissue involvement.
In our study cohort, the age distribution of cases was biphasic with a small group of young cats (less than 4-years-of-age) and a much larger group of cats about 10 years-old (overall range 5 to 212 months; median age 120 months). The subgroup of young cats tended to have mediastinal involvement, with or without involvement of nearby lymph nodes. Siamese cats and Siamese crosses were over-represented (22 cases) across-the-board in comparison to domestic crossbred cats, and were especially conspicuous amongst the group of young cats with mediastinal involvement. Male cats were over- represented, accounting for 72/118 cats.
Compared to series from the UK or North America, mixed cases and atypical cases were much more common. 70% of lymphomas were B cell tumours, 26% were T cell tumours while 4% failed to stain with any T or B cell markers. Medium grade (66%) and high grade (24%) tumours predominated, while low-grade tumours (10%) were quite rare. There was no detectable association between histological grade and immunophenotype.
Haematology and serum biochemistry were generally unhelpful in the diagnostic investigation of these cases, although as expected, cats with renal involvement tended to be azotaemic and cats with alimentary LSA tended to have low albumin concentrations. Although 13 cats had detectable neoplastic lymphoid cells in buffy coat preparations of peripheral blood, only 5 had distinct lymphocytosis.
2/107 (2%) of cases for which serum was available were positive for FeLV p27 antigen; both of these were young cats with T cell tumours of mixed cell type, and both were leukaemic. Considering this, it was surprising that 25/97 (26%) of cases for which tissue was available for PCR testing had FeLV provirus in tumour tissue. Unfortunately we did not have the foresight to collect age-matched lymphoid tissue from cats without LSA to use as 'controls'.
50/101 (50%) cats were FIV-positive on the basis of Western blot analysis. These cats were more likely to be male, domestic crossbreds and have atypical LSA. Only 39/107 (36%) were positive on the basis of one manufacturer's in house ELISA, and using Western blot analysis as the gold standard this corresponded to 5 false-positives and 16 false-negatives. The 5 false-positives were negative using an immunomigration FIV test, and of the 16 sera that were ELISA-negative but Western-positive, 9 were positive using immunomigration.
We concluded from these virological studies that persistent FeLV antigenaemia was very rare in Australian cats with LSA, but there was insufficient data upon which to assess the importance of FeLV infection in relation to lymphomagenesis. Overall, FeLV appeared to be much less important as a cause of LSA than in studies emanating from UK and North America in the 1970s and 1980s. In contrast, FIV infection was very prevalent in Australian cats with LSA and circumstantial evidence points to Australian FIV strains playing a key role in the development of LSA. Although a number of st8udies had previously described a link between FIV and LSA, particularly in small groups of experimentally infected cats, this was the first large study to emphasise this potential association. The influence of FIV as a predisposing cause goes some way to explaining a number of the epidemiologic features of LSA in our study cohort, such as the preponderance of males, the commonness of B cell tumours and the large number of atypical and mixed forms of LSA. Mediastinal LSA in young cats, especially in Siamese/Oriental cats, is usually linked with neither FeLV or FIV in Australian cats, and is likely a genetic disease based on recent data from North America.
An important component of this study was to correlate patient, tumour, retroviral status and other measurable factors with clinical outcomes in response to a single multi-agent chemotherapy protocol. We used a modified Wisconsin protocol, using slightly different doses of drugs, given in a slightly different order and sometimes using different routes of administration, such that owners were only required to attend the veterinary clinic every second week for the first 16 weeks of therapy, and monthly thereafter. Doxorubicin was given at a dose of 1 mg/kg, L-asparaginase was given more frequently than is customary and via the intramuscular route at a dose of 450 U/kg, cyclophosphamide and methotrexate were given orally (rather than intravenously), routine haematological monitoring was not attempted and intravenous chemotherapy was typically given into he jugular vein under sedation or sevoflurane anaesthesia.
61 cats were treated, 60 with chemotherapy (of which 7 had surgery also), while one was treated with surgery alone. The complete remission rate for chemotherapy was 80%. The median survival time of 116 days was disappointing, perhaps reflecting the fact that cats were treated irrespective of their disease severity at presentation. Considering only cats that went into complete remission, the median survival was 187 days. Importantly, 20 cats were long-term survivors, and indeed perhaps 25% to 30% of cases were cured.
Despite a rigorous analysis of variables, only three factors had a discernable impact on prognosis. The predictors of a favourable outcome were (1) attainment of complete remission (2) a T cell phenotype (3) being less than 4-years-of age. Surprisingly, FIV status had no impact on survival. Interestingly, one cats with solitary nodal (mandibular) LSA was cured by surgical excision alone. Two cats died of recrudescent toxoplasmosis during therapy.
As one of the best predictors of a successful outcome was the attainment of complete remission, we strongly recommend that cats with LSA be given induction therapy with vincristine and L-asparaginase before making a definitive decision about embarking on a full course of chemotherapy. FeLV-negative young cats (less than 4-years-old) are good candidates for chemotherapy (especially if they have mediastinal involvement) and we believe a majority of these cases can be cured. Recent work from the Netherlands suggests that these tumours may respond equally well to simpler protocols without doxorubicin or L-asparaginase, and if this is proven by subsequent works a doxorubicin-free protocol would be probably be preferable.
It may be inappropriate to extrapolate our results for chemotherapy in Australian cats to other countries in which the clinical features and epidemiology of LSA is different. Comparing outcomes of LSA therapy in different places may thus be like comparing apples with oranges! For example, the outstanding outcomes obtained in the recent study of LSA in Dutch cats is probably in part a reflection of the high proportion of young Siamese cats with mediastinal involvement in the study cohort.
The pattern of disease observed in our Australian study is similar in some respects for what has been reported more recently for cats in the USA, in contradistinction to what is commonly recorded in veterinary texts. This may reflect a lesser impact of FeLV more recently, as a result of widespread vaccination in North America, or more likely other factors. It is thus probably wise to consider LSA to be an evolving disease, reflecting both changes in the genetic constitution in cats over time and changes in the prevalence and virulence of FeLV and FIV over time.
More work is required to further refine guidelines for chemotherapy of LSA in cats. It is the impression of the authors that cats almost never require more than one-year of chemotherapy and anecdotally, that perhaps 6 months of intensive chemotherapy is sufficient to effect a cure in many cases. There is no doubt in our mind that prednisolone contributes to complications, but we are not convinced it makes an additional contribution to efficacy, based on good success in some cases where prednisolone was not included in the protocol because of contraindications based on pre-existing disease conditions.
Finally, since the completion of this study the broad brushstrokes of the first 60 cases have been confirmed. Additionally, we have seen 3 cases of gastric LSA that were cured using radical surgical excision and limited follow-up chemotherapy. Nasal cases continue to be seen, including a substantial proportion that have a prominent nasopharyngeal component. Many of these do well with cytoreduction (by massage/flushing) and follow-up chemotherapy. Also, many cats, especially older cats, cope much better with the chemotherapy regimen if the drugs are administered at longer intervals, typically with intravenous chemotherapy given every 3 weeks rather than every 2 weeks.
1. Gabor, L. J. Malik, R. Canfield, P. J. Clinical and anatomical features of lymphosarcoma in 118 cats. Australian Veterinary Journal 1998. 76: 11, 725-732. 36 ref.
2. Gabor, L. J. Canfield, P. J. Malik, R. Immunophenotypic and histological characterisation of 109 cases of feline lymphosarcoma. Australian Veterinary Journal 1999. 77: 7, 436-441.
3. Gabor, L. J. Canfield, P. J. Malik, R. Haematological and biochemical findings in cats in Australia with lymphosarcoma. Australian Veterinary Journal 2000. 78: 7, 456-461
4. Gabor, L. J. Jackson, M. L. Trask, B. Malik, R. Canfield, P. J. Feline leukaemia virus status of Australian cats with lymphosarcoma. Australian Veterinary Journal 2001. 79: 7, 476-481
5. Gabor, L. J. Love, D. N. Malik, R. Canfield, P. J. Feline immunodeficiency virus status of Australian cats with lymphosarcoma. Australian Veterinary Journal. 2001.
6. 79: 8, 540-545
7. Malik, R. Gabor, L. J. Foster, S. F. McCorkell, B. E. Canfield, P. J. Therapy for Australian cats with lymphosarcoma. Australian Veterinary Journal 2001. 79: 12, 808-817
8. Teske, E. Straten, G. van. Noort, R. van. Rutteman, G. R. Chemotherapy with cyclophosphamide, vincristine, and prednisolone (COP) in cats with malignant lymphoma: new results with an old protocol. Journal of Veterinary Internal Medicine 2002. 16: 2, 179-186.
9. Vail, D. M. Moore, A. S. Ogilvie, G. K. Volk, L. M. Feline lymphoma (145 cases): proliferation indices, cluster of differentiation 3 immunoreactivity, and their association with prognosis in 90 cats. Journal of Veterinary Internal Medicine 1998. 12: 5, 349-354.