L plasm, reportedly accounting for approximately 90%

Transcription

1 Feline Spinal Lymphosarcoma: A Retrospective Evaluation of 23 Cats Stephen B. Lane, Joe N. Kornegay, J. Robert Duncan, and John E. Oliver Jr. A retrospective study of pathologically confirmed cases of feline spinal lymphosarcoma (FSL) admitted to the Collages of Veterinary Medicine at the University of Georgia and North Carolina State University from 1973 to 1988 was conducted. Two hundred fourteen cases of feline lymphosarcoma were diagnosed histopathologically; involvement of the central nervous system (CNS) was identified in 26 (12.1%). Twenty-three of these tumors involved the spinal cord, and 22 of the 23 were solitary. A predilection for the thoracic and lumbar vertebral canal was noted. Most cats with spinal disease were young, with mean and median ages of 43 and 24 months, respectively; 67 cats ymphosarcoma (LSA) is the most common feline neo- L plasm, reportedly accounting for approximately 90% of hematopoietic and 22% to 30% of all tumors of cats.'-5 The exogenous feline leukemia virus (FeLV) has been implicated as the most important factor associated with hematopoietic neoplasia in with FeLV viremia occumng in approximately 30% of cats with neoplasia.* Most published reports of central nervous system (CNS) LSA have included single cases or small series of affected cats.'-" Based on this material, the vertebral canal appears to be preferentially involved.9-' However, the anatomic distribution and other clinicopathologic features have not been established. Data in this study pertains principally to the spinal form of feline LSA (FSL) and includes material collected retrospectively from cases of LSA evaluated at the Colleges of Veterinary Medicine at North Carolina State University and the University of Georgia. Clinical and diagnostic features, treatment, prognosis, and pathological findings are presented. Data from 10 of the 14 cats from the University of Georgia have been discussed briefly previously.'' Results from the 23 cats published here largely parallel those obtained in a recently published series of 2 1 cats with spinal lymphosarcoma. l3 Materials and Methods Case Selection The records of 2 14 cats with pathologically confirmed LSA admitted to the Colleges of Veterinary Medicine at North Carolina State University and the University of Georgia between 1973 and 1988 were reviewed. Special attention was paid to 26 cases in which there was CNS involvement. Historical, clinical, diagnostic, and postmortem data were obtained from the university records. Systematic pathological evaluation of the CNS was not performed in all cats. Thus, the data presented here may underrepresent the frequency of neurological involvement in this neoplasm. Assay Systems Enzyme-linked immunosorbent assay (ELISA) (Leukassay HI ELIZA, Pittman-Moore, Inc, Washington Crossing, NJ) was used to detect the presence of circulating FeLV p27 antigen in 17 cats; an indirect fluorescent antibody (IFA) (Assays, Nashville, TN) technique was used to evaluate the bone marrow of 1 cat for the presence ofcell-associated FeLV p27. The concentration immunoassay technology (CITE) feline immunodeficiency virus (FIV) antibody test kit (Feline Immunodeficiency Virus Antibody Test Kit, IDEXX were 36 months of age or younger. In most cases, affected cats had acute neurological deterioration after an initial insidious course. Extraneural abnormalities were not consistently present. Neoplastic lymphocytes diagnostic of FSL were identified on cerebrospinal fluid (CSF) analysis in 6 of 17 cats evaluated. Sixteen of 17 cats evaluated had serologically positive test results for feline leukemia virus (FeLV) p27 antigen, and all cats tested for feline immunodeficiency virus (FIV) antibodies had negative test results. J Vet Intern Med 1994;8: Copyright by the American College of Veterinary Internal Medicine. Corp, Portland, ME) was used to identify the presence of antibody to FIV in serum in 5 cats and cerebrospinal fluid (CSF) in 1 cat. Classification Lymphoid neoplasms were classified and staged according to their cytological and histological appearance'4s15 and anatomical distribution.16 Neoplastic tissue from 14 ofthe 23 cats was reviewed retrospectively by one of us (JRD) to establish a histopathologic classification of the tumor. Cerebrospinal Fluid Analysis Cerebrospinal fluid was obtained from the cerebellomedullary (n = 12) or lumbar (n = 3) cisterns, or both (n = 2) in 17 cats. Cerebrospinal fluid protein was quantified using a commercial dye-binding assay (Bio-Rad Protein Assay, Bio-Rad Chemical Division, Richmond, CA). Leukocyte and erythrocyte cell counts were performed on undiluted CSF using a standard hemacytometer. Cytological evaluation was done in all cases using a cytocentrifuge (Cytospin 2, Shandon Southern Instruments, Inc, Sewickley, PA) and modified Wright-Giemsa staining (Leukostat, Fisher Scientific, Orangeburg, NY). Results Prevalence Baseline hospital data were available for the time periods of 1973 to 1988 (University of Georgia) and 1985 to 1988 (North Carolina State University). During this time, 1,897 cats were hospitalized; LSA was diagnosed in 2 14 (1 1.3%). Twenty-six of these 2 14 cats (12.1%) had CNS involvement, From the Department of Companion Animal and Special Species Medicine, College of Veterinary Medicine, North Carolina State University, Raleigh, NC (Lane, Kornegay), and the Departments of Pathology (Duncan) and Small Animal Medicine (Oliver), College of Veterinary Medicine, University of Georgia, Athens. Accepted December 7, Dr Lane's current address is Veterinary Referral Clinic ofmississauga, 305 Matheson Boulevard East, Mississauga, Ontario NI E 2R4, CANADA. Reprint requests: Dr Joe Kornegay, College of Veterinary Medicine, North Carolina State University, 4700 Hillsborough St, Raleigh, NC Copyright by the American College of Veterinary Internal Medicine /94/ $3.00/0 Joumal of Veterinav Internal Medicine, Vol8. No 2 (March-April), 1994: pp

2 100 LANE ET AL Age (years) Fig 1. Age distribution of cats with spinal lymphosarcoma at the onset of neurologic dysfunction. and 23 of these 26 (88.5%) cats had tumors in the vertebral canal. Results given below refer to the 23 cats with FSL, unless otherwise indicated. The mean and median ages of the 23 cats were 43 and 24 months, respectively (range: 6 months to 14 years); 16 (69.6%) were 36 months of age or younger (Fig 1). There was no preferential breed or sex distribution, with 12 female and I I male cats included in the study. Clinical Findings Extraneural clinical abnormalities were noted infrequently. Weight loss, renomegaly, and peripheral lymph node enlargement were noted in 4, 2, and I of the cats, respectively. Only 7 of the 23 cats had hematologic or serum biochemical abnormalities. There was lymphoblastic leukemia in 3 cats (2 ofthe 3 also had intracranial leptomeningeal lymphoblastic infiltration); 3 had macrocytic, normochromic nonregenerative anemia; and Haemobartonella felis was identified in 2 cats. Neurological Findings Asymmetric paraparesis, focal hyperesthesia, and rapidly progressive ataxia were the most common clinical signs, occumng in 20 (86.9%), 19 (82.6%), and 11 (47.8%) cats, respectively. These signs generally progressed over a period of 7 days or less. Cats with cervical spinal cord or nerve root involvement generally had peracute tetraparesis, lower motor neuron dysfunction, and diminished sensation in the thoracic limbs. Clinical signs typical of root lesions (root signature), such as lameness and hyperesthesia upon shoulder extension, were noted in only 1 of 3 cats with cervical root involvement. Virological Testing Sixteen of 17 cats tested for circulating FeLV p27 antigen were positive at the time of diagnosis; the cat with seronega- Fig 2. Ventral aspect of the cervical roots and associated spinal cord segments from an 8-year-old male domestic cat with acutely progressive left thoracic limb monoparesis and eventual tetraplegia. A large neoplastic mass extends to the cervical roots at multiple levels, resulting in root enlargement. tive results had IFA-positive results on bone marrow evaluation. Tests done in 6 cats for the presence of either serum (n = 5) or CSF (n = 1) FIV antibodies had negative results. Tumor Topography Neoplastic tissue was found in an extradural location in 20 of the 23 (87.0%) cats with spinal LSA. In the other 3 cats, the cervical roots of the brachial plexus were affected, and there was extension to the subarachnoid space (Fig 2). All but 1 of the 23 cats had solitary tumors in the CNS; this cat had two separate tumors. Fourteen of the total of 24 (58.3%) tumors occurred between the second thoracic and the fourth lumbar vertebrae (Fig 3). Cerebrospinal Fluid Analysis All but one CSF sample were clear and colorless. A mixed pleocytosis (mean of 140 cells/pl; range of 0 to 1625 cells/ C1-4 C5-Tl T2-5 T6-9 T10-13 L1-3 L4-7 Vertebrae Fig 3. Distribution by vertebral body level of cats with spinal lymphosarcoma. The total number of cats adds to 25 because of multifocal involvement in 1 cat.

3 SPINAL LYMPHOSARCOMA 101 Table 1. Cerebrospinal Fluid Analysis Protein Tumor location Cat No. RBC/gL WBC/pL (mg/dl) (vertebrae) 1' C6-T T L3-4 4t , ML 126 multifocal L L1-3 7' T T L4 9' NA ML NA L1-4 llt , ML 103 C6-T L4-S3 12 QNS ML QNS L3 14' QNS ML QNS L L7-S T v ,ML 404 C7 Note: All samples except one were clear and colorless. Abbreviations: QNS, insufficient quantity; NA, not available; ML, malignant lymphocytes identified on cytological examination. Lumbar cistern centesis. t lntramedullary tumor involvement. 10). Although adjacent vertebrae were not evaluated histologically in all cases, none of the cats reviewed had bony involvement. Fourteen tumors were classified histologically as lymphoblastic (n = 6), small noncleaved (n = 4), immunoblastic (n = 3), and small lymphocytic (n = I). Microscopically, all tumors had medium to moderate mitotic activity (3 to 6 mitoses per 40X power field), compatible with the medium to high grade categories of the National Cancer Institute Working Formulation, which suggests a poor survival rate.l4.l5 Treatment and Outcome Four cats with extradural spinal tumors were treated. Chemotherapy (L-asparaginase, vincristine, and prednisone) was given after local spinal radiation (n = 3) or surgical cytoreduction (n = I). Improvement was noted in 3 cats with cranial lumbar neoplasms, even though they received different treatments. Neurological deterioration occurred in 1 cat with lower motor neuron disease and absence of pain sensation associated with a caudal lumbar lesion. Longterm results were poor, even in those cats that initially improved. One cat is alive 13 months after presentation. The other 3 cats died or were euthanatized within 5 months of treatment due to systemic relapse. pl) with elevated protein content (mean of mg/dl; range of mg/dl) was identified in most cats (Table I). Neoplastic lymphocytes were seen in three lumbar and three cerebellomedullary cistern samples. Radiography Osseous lesions were not noted on survey radiographs. Myelograms were done in 16 cats, revealing an extradural pattern of compression in 11 of these cats; the lesion was asymmetric in 9 and symmetric in 2 of the affected cats. Thinning, deviation, and splitting of the contrast columns were noted (Fig 4). The myelograms in the other 6 cats had either an intramedullary pattern (n = 3) or were judged to be normal (n = 2). The cats with the intramedullary myelographic pattern corresponded to those in which there was neoplastic infiltration from spinal nerve roots across the subarachnoid space in the cervical area. Pathological Findings Solitary, epidural lesions were identified in I8 of the 23 (78.3%) cats with spinal tumors. Cervical root lesions extended to the subarachnoid space in 3 cats; I cat had multifocal epidural involvement in the thoracolumbar area. Some cats with solitary tumors had additional distant meningeal infiltration by anaplastic lymphocytes. Subarachnoid and parenchymal extension occurred in the 3 cats with cervical root lesions. Extensive intratumoral and peritumoral intramedullary hemorrhage was noted in adjacent cord segments in these cats. Other organ systems were affected in 10 of 23 (43.5%) cats with FSL. The kidneys were most commonly affected (n = Fig 4. Ventrodorsal myelogram from a 6-month-old female domestic cat with progressive paraplegia and a cranial mediastinal mass. The contrast medium is displaced from right to left over the bodies of L2-L4. The cat tested positive for FeLV, and multifocal lymphosarcoma was suspected. Cytology of an aspirate of the cranial mediastinal mass was compatible with lymphosarcoma. Combined chemotherapy and radiation was administered, and there was clinical improvement. However, the cat died 5 months later. A necropsy was not done. Reprinted with permission from Kornegay JN. Feline neurology. Probl Vet Med 1991;3:367.

4 102 LANE ET AL Discussion Primary CNS lymphoma in human beings is rare. Less than 0.7% of all malignant lymphomas, and 1.6% of all extranodal malignant lymphomas, are considered primary CNS neoplasms." Spinal cord compression more commonly occurs as a late complication of generalized lymph~ma.'~~'' Lymphomatous involvement ofthe CNS in cats is also thought to be uncommon, accounting for 0.2% and 5.3% of feline malignancies in two comprehensive studie~.'',~' In our series, 12.1% of cats with LSA had CNS involvement; this represented 0.1 % of all feline accessions. The high prevalence of spinal lesions in cats with CNS LSA is consistent with the anatomic distribution of previously published cases?-' 1913,21-24 Neoplasms involving the spinal cord have traditionally been classified as primary or secondary, depending upon the origin of the predominant cell type and the presence or absence of systemic involvement. It is not clear if LSA can arise de novo in the epidural space, and therefore, be classified as a primary spinal neoplasm. Attempts to locate lymphatic tissue in the vertebral canal of human beings have been unsuc~essful.~~*~~ Extramedullary hematopoietic tissue has been cited as the possible anlage for the development of primary LSA of the feline spine.22 As such, spinal LSA has traditionally been classified as either a secondary or primary neural tumor based on whether neoplastic lymphocytes were or were not identified in other organ systems, respectively. Our data and that of others indicate that feline spinal LSA is a disease of young cats. The mean and median ages of the cats were 43 and 24 months, respectively, and 16 cats were under 36 months of age. In contrast, the mean age of dogs with spinal tumors, regardless of histological type, is 60 to 84 months.23s27 This age difference is likely related to the infectious etiology of most feline lymphoid tumors, because the FeLV appears to be associated with FSL.I3 All cats tested for FeLV in our series were positive. In another series of cats with FSL 16 of 19 (84%) affected cats were also FeLV- ~0sitive.I~ Tumor location and animal age have been shown to influence results of serological testing for the presence of circulating p27 antigen (viremia).'6,28 It is not known whether serum neutralizing and feline oncornavirus cell membrane-associated (FOCMA) antibodies confer protection to the CNS. The presence of FOCMA, which reportedly is expressed on the cell membranes of FeLV-induced neoplastic lymphocyte^?^ was not assessed in our study or in previously reported cats with FSL.8,'3928*29 Although FeLV seems to be causally associated with FSL, lymphatic tumors of the CNS also may occur spontaneously. Primary CNS lymphomas have been associated with a variety of primary and acquired immunodeficiency disorders in human beings.30 The prevalence of primary CNS lymphoma in patients infected with the human immunodeficiency virus (HIV/HTLV-111) has increased, ranking as the second leading cause of intracranial mass lesions in this Little is known about chronic immunosuppression and CNS disease in the cat. Feline leukemia virus-induced immunosuppression and resultant secondary diseases (FeLV-induced feline acquired immune deficiency syndrome [FAIDS]) lead to more deaths than FeLV-induced neoplasms.7v33 The isolation and early charactenzation of a feline lentivirus, now termed FIV, has led investigators to associate nonspecific behavioral changes in FIVpositive cats with those now recognized to be part of the acquired immunodeficiency syndrome (AIDS) dementia complex.34 Whether this trend will be verified in FIV-infected cats studied in the future is not known. Serum and CSF from 6 FeLV-positive cats with FSL in our study did not contain antibody to FIV. This is in accord with serological testing of FeLV-positive cats, in which evidence of concomitant RV infection is unu~ua1.~~~~~ That the thoracolumbar spine and epidural space were selectively involved in cats with FSL in this series is consistent with prior report^.",'^,^^,^',^^ The epidural space provides a low resistance chamber for gradual tumor expansion along or around the spinal cord and its nerve roots. Extension over contiguous vertebrae occurred frequently in cats we studied. Dural invasion and subsequent parenchymal extension was noted only in cats with cervical nerve root involvement. Vertebral lesions were not identified in those cats evaluated histologically. However, paraspinal muscle and dural and vertebral invasion by neoplastic cells has been reported.'-' 1~'3,21,22 Although LSA also frequently affects the thoracolumbar spine of dogs, cervical and intracranial involvement is relatively common as ell.^^,^' Reasons for the apparent predilection for the thoracic, and to a lesser extent, lumbar spine in cats with LSA are not known. Mechanisms leading to extension of FSL from nerve roots to the spinal cord preferentially in the cervical area in cats of our study also are unclear. Interestingly, however, in one study of spinal tumors of various histological types in dogs, cervical tumors often originated or extended intradurally, whereas those in the thoracolumbar area generally remained extrad~ral.~~ The absence of detectable extraneural involvement in almost half of the cats in this series contrasts with prior reports in which systemic involvement was present in most case^.',^',^^ Spinal LSA in dogs and human beings seems to occur more commonly as a complication of generalized disease; primary or focal hematogenously derived epidural me tastases occur infreq~ently.~~"' The likelihood of metastasis of FSL to other areas of the CNS appears low; only 1 of the 23 cats had multiple mass lesions. The clinical signs and progression of spinal neoplasms depend on the location of the tumor within the spinal axis and relative to the dural sheath. Focal hyperesthesia, a finding which is present in advance of clinical dysfunction in hu. man beings with spinal lymphoma,4'was present in most of the cats discussed here. That neurological dysfunction was often insidious in cats of this study is compatible with the typical course of slowly expanding extradural lesions. However, rapid clinical deterioration was noted in almost half of the cats. Cerebrospinal fluid analysis is supportive, but infrequently diagnostic, of CNS neoplasia. Nonspecific increases in CSF white blood cell count and protein concentration were noted in most cats of this series. The character of the pleocytosis varied. Neutrophils predominated in some cases. Although neutrophilic pleocytosis is commonly asso-

5 SPINAL LYMPHOSARCOMA 103 bated with infectious CNS disease, increased numbers have also been noted secondary to tumors associated with the CSF pathway, particularly rneningioma~.~' In our series, the highest neutrophil counts were noted in cats with cervical root tumors that had extended intradurally. In all of these cases, hemorrhage and necrosis of the infiltrating tumor and adjacent spinal cord was noted at necropsy. Neoplastic lymphocytes were seen on CSF analysis in only 6 of 17 cats studied in this series and have been noted infrequently, if at all, in other reported casesl3 Analysis of CSF appears to be diagnostic of LSA more commonly in affected dogs.3733s This species difference may relate to the fact that unlike cats, most dogs with CNS LSA have multifocal leptomeningeal infiltration. Site of CSF collection may be an additional factor. Considering the predominant caudal flow of CSF, samples taken from the cerebellomedullary cistern are more likely to be helpful when the underlying lesion affects the cervical spine or brain.42 As has been stated already, whereas LSA of the CNS in dogs seems to affect the brain and cervical cord relatively commonly, cats have preferential thoracolumbar spinal involvement. Modalities that alone or in combination may be beneficial in treating cats with LSA include surgical cytoreduction, focal radiotherapy, and systemic chemotherapy. Regardless of the therapeutic regimen, FSL should be treated as an emergency. Although laminectomy allows spinal cord and nerve root decompression, as well as accurate histological identification prior to therapeutic intervention, irradiation is the mainstay of treatment in humans with spinal LSA and is given irrespective of surgical intervention. Large fractional doses (400 cgy) totaling >50 Gy may be necessary for local tumor control.43 Rapid reduction in tumor mass may occur within hours of delivering a single large dose of radiation (500 to 1000 cgy) in cats with LSA.44 The effectiveness of radiotherapy as the sole means of therapy for FSL is not known. The use of chemotherapy alone, or as an adjuvant therapy, is critical for control of tumor tissue in other organ systems. Chemotherapy protocols for feline hematopoietic tumors vary depending on the stage of disease and personal preference. Multiple agents are generally used in combination. Duration of remission and survival times vary among Chemotherapy combined with local spinal radiation or surgery resulted in improvement in 3 cats in our series, but only 1 improved long-term. Six cats treated with vincristine, cyclophophamide, and prednisone in another series achieved either a complete or partial remission.13 The median duration of complete remission was 14 weeks. Many chemotherapeutic agents poorly penetrate the bloodbrain (spinal cord) barrier; thus, cats with parenchymal involvement may respond more poorly to chemotherapy. References I. Schmidt RE, Langham RF. Survey of feline neoplasms. J Am VetMed Assoc 1967; 151: Whitehead JE. Neoplasiain the cat. Vet Med Small Anim Clin 1967;62: Dorn CR, Taylor D, Schneider R, et al. Survey of animal neoplasms in Alameda and Contra Costa Counties, California. 11. Cancer morbidity in dogs and cats from Alameda Co. J Natl Cancer Inst 1968;40: Engle GC, Brodey RS. A retrospective study of 395 feline neoplasms. J Am Anim Hosp Assoc 1969; 5: Dorn CR. Epidemiology of canine and feline tumors. J Am Anim Hosp Assoc 1976; 12: Jarrett WFH, Crawford EM, Martin WB, et al. A virus-like particle associated with leukemia (lymphosarcoma). Nature 1964;202: Hardy WD Jr. The virology, immunology, and epidemiology ofthe feline leukemia virus. In: Hardy WD Jr, Essex M, McClelland AJ, eds. Feline Leukemia Virus. North Holland: Elsevier: 1980: Cotter SM. Feline viral neoplasia. In: Greene CE, ed. Clinical Microbiology and Infectious Diseases of the Dog and Cat. Philadelphia, PA: WB Saunders; 1984: Zaki FA, Hurvitz AI. Spontaneous neoplasms of the central nervous system of the cat. J Small Anim Pract 1976; 17: Schappert HR, Geib LW. Reticuloendothelial neoplasms involving the spinal canal of cats. J Am Vet Med Assoc 1967; 150: I 1. Northington JW, Juliana MM. Extradural lymphosarcoma in six cats. J Small Anim Pract 1978; 19: Kornegay JN. Feline neurology. Compend Contin Educ Pract Vet 1981;3: Spodnick GJ, Berg J, Moore FM, et al. Spinal lymphoma in cats: 21 cases ( ). J Am Vet Med Assoc 1992;200: Carter RF, Valli VEO, Lumsden JH. The cytology, histology, and prevalence of cell types in canine lymphoma classified according to the National Cancer Institute Working Formulation. Can J Vet Res 1986;50: Non-Hodgkin's lymphoma pathologic classification project. National Cancer Institute-sponsored study of classifications of non-hodgkin's lymphomas. Cancer 1982;49: Hardy WD Jr. Hematopoietic tumors of cats. J Am Anim HOSPASSOC 1981; 17: Gregory MC, Hughes JT. Intracranial reticulum cell sarcoma associated with IGA deficiency. J Neurol Neurosurg Psychiatry 1973;36: Freeman C, Berg JW, Cutter SJ. Occurrence and prognosis of extranodal lymphomas. Cancer 1972;29: Meincke JE, Hobbie WV Jr, Hardy WD Jr. Lymphoreticular malignancies in the cat: clinical findings. J Am Vet Med Assoc 1972; 160: Hardy WD Jr. Epidemiology of primary neoplasms of lymphoid tissues in animals. In: Twomey JJ, Good RA, eds. The immunopathology of lymphoreticular neoplasms. New York: Plenum Publishing Corp: 1978: Holzworth J. Leukemia and related neoplasms in the cat. J Am Vet Med Assoc 1960; 136: Averill DR. Feline neoplasia. In: Holzworth J, ed. Diseases of the Cat. Philadelphia: WB Saunders; Luttgen PJ, Braund WR, Vandevelde M. A retrospective study of twenty-nine spinal tumours in the dog and cat. J Small Anim Pract 1980;21: Mooney SC, Hayes AA, Matus RE, et al. Renal lymphoma in cats: 28 cases ( ). J Am Vet Med Assoc 1987;191: Browder J, De Veer JA. Lymphomatoid disease involving the spinal epidural space. A pathologic and therapeutic consideration. Arch Neurol Psychiatry 1939;41: Rubenstein LJ. Atlas of Tumor Pathology. Tumors of the central nervous system. Fascicle 6, Tumors of the lymphoreticular system. Armed Forces Institute of Pathology: Washington DC; 1972: Wright JA, Bell DA, Clayton-Jones DG. The clinical and ra-

6 104 LANE ET A1 diological features associated with spinal tumours in thirty dogs. J Small Anim Pract I979; Francis DP, Cotter SM, Hardy WD Jr, et al. Comparison of virus positive and virus negative cases of feline leukemia and lymphoma. Cancer Res 1979;39: Hardy WD Jr, McClelland AJ, Zuckerman EE, et al. Development of virus non-producer lymphosarcomas in pet cats exposed to FeLV. Nature 1980;288: Pattengale PK, Taylor CR, Parke T, et al. Selective immunodeficiency and malignant lymphoma of the central nervous system. Acta Neuropathol (Berl) 1979;48: Snider WD, Nielsen S, Gold JWM, et al. Neurological complications of acquired immunodeficiency syndrome: analysis of 50 patients. Ann Neurol 1983; 14: Levy RM, Bredesen DE, Rosenblum ML. Neurological manifestations of the acquired immunodeficiency syndrome (AIDS): experience at UCSF and review of the literature. J Neurosurg 1985;62: Hardy WD Jr. Feline leukemia virus non-neoplastic diseases. J Am Anim Hosp Assoc 1981; 17: Pedersen NC, Ho EW, Brown ML, et al. Isolation of a T- lymphotropic virus from domestic cats with immunodeficiencylike syndrome. Science 1987;235: Shelton GH, McKim KD, Cooley PL, et al. Feline leukemia virus and feline immunodeficiency virus infections in a cat with lymphoma. J Am Vet Med Assoc 1989; 194: Yamamoto JK, Hansen H, Ho EW, et al. Epidemiology and clinical aspects of FIV infection in cats from the continental United States and Canada and possible mode of transmission. J Am Vet Med Assoc 1989; I94: Rosin A. Neurologic disease associated with lymphosarcoma in 10 dogs. J Am Vet Med Assoc 1982; 181: Couto CG, Cullen J, Pedroia V, et al. Central nervoussystem lymphosarcoma in the dog. J Am Vet Med Assoc 1984; 184: Mead GM, Kennedy P, Smith JL, et al. Involvement ofthe central nervous system by non-hodgkin s lymphoma in adults. A review of 36 cases. Q J Med 1986;60: Aabo K, Walbom-Jorgensen S. Central nervous system complications by malignant lymphomas: radiation schedule and treatment results. J Rad Oncol Biol Phys 1986; 12: Bailey CS, Higgins RJ. Characteristics of cisternal cerebrospinal fluid associated with primary brain tumors in the dog: A retrospective study. J Am Vet Med Assoc 1986; 188: Thomson CE, Kornegay JN, Stevens JB. Cerebrospinal fluid from the cerebellomedullary and lumbar cisterns ofdogs with focal neurologic disease: 145 cases ( ). J Am Vet Med Assoe 1990; 196: Berry MP, Simpson WJ. Radiation therapy in the management ofprimary malignant lymphoma ofthe brain. Int J Rad Oncol Biol Phys 1981;7: MacEwen EG, Mooney S, Brown NO, et al. Management of feline neoplasms. In: Holzworth J, ed. Diseases of the cat. Philadelphia, PA; WB Saunders; Mooney SC, Hayes AA, MacEwen G, et al. Treatment and prognostic factors in lymphoma in cats: 103 cases ( I). J Am Vet Med Assoc 1989; 194: Cotter SM. Treatment of lymphoma and leukemia with cyclophosphamide, vincristine, and prednisone. 11. Treatment of cats. J Am Anim Hosp Assoc 1983; 19: Jeglum KA, Whereat A, Young K. Chemotherapy of lymphoma in 75 cats. J Am Vet Med Assoc 1987;