Multiple-Case Leukemia Cluster Serological Studies ...

Serological Studies of Normal and Leukemic Cats in a Multiple-Case Leukemia Cluster Susan M. Cotter, M. Essex and William D. Hardy, Jr. Cancer Res 1974;34:1061-1069. Published online May 1, 1974.

Updated Version

Citing Articles

E-mail alerts Reprints and Subscriptions Permissions

Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/34/5/1061

This article has been cited by 1 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/34/5/1061#related-urls

Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected]. To request permission to re-use all or part of this article, contact the AACR Publications Department at [email protected].

Downloaded from cancerres.aacrjournals.org on July 11, 2011 Copyright © 1974 American Association for Cancer Research

[CANCER RESEARCH 34, 1061 1069. May 1974]

Serological Studies of Normal and Leukemic Cats in a Multiple-Case Leukemia Cluster1 Susan M. Cotter. M. Essex,2 and William D. Hardy, Jr.3 Angeli Memorial Animal Hospital [S. M. C.], and Department of Microbiology, Harvard University School of Public Health, Boston, Massachusetts 02115 [M. E.\, and Sloan-Kettering Institute for Cancer Research, New York, New York 10021 [W. D. H.}

SUMMARY Eleven cases of lymphoblastic leukemia and four cases of feline infectious peritonitis occurred in a private household population of 35 predominantly unrelated cats over a 39month period. The study conducted on this population con sisted of periodic physical examinations, hemograms, serological testing for feline leukemia virus group-specific antigens (gs) and for antibody to feline oncornavirus-associated cell membrane antigen (FOCMA). Serological values were compared to those for healthy and leukemic control cats from the same geographical area. Leukemic household cats had a higher mean age and had been in the household longer than those that remained healthy. Six of 9 leukemic cats were positive for FeLV (as were 9 of 10 leukemic cats living in other environments). Both the control and cluster household leukemic cats also had low geometric mean antibody titers to the FOCMA antigen (0.72 and 0.76, respectively). This differed from healthy controls from outside the house that were all nega tive for gs antigens and had a mean antibody titer of 1.66. For healthy controls within the house, the mean antibody titer was 4.65 and 5 of 17 were positive for gs antigens. This difference is apparently due to more frequent or more recent exposure to virus in the cluster household. Relatives of leu kemic cats were no more likely to develop leukemia or be come gs positive than relatives of healthy cats. Cats de veloping leukemia appear to be unable to produce or maintain a high FOCMA antibody titer. It was concluded that these data support the theory of horizontal transmis sion of feline leukemia virus. INTRODUCTION Leukemia and sarcoma of cats are caused by C-type oncornaviruses that are biochemically and biophysically simi lar to, and antigenically related to, the murine oncornaviruses (7-9, 15, 24). The virus-induced cat tumors are of 1This study was supported by grants from the National Cancer Insti tute (CA-08748), the New York Cancer Research Institute, the Oliver S. and Jennie R. Donaldson Charitable Trust, the Massachusetts Branch of the American Cancer Society, the Jane Coffin Childs Fund for Medical Research, and the Anna Fuller Fund. 2Scholar of the Leukemia Society of America. 3Special Fellow of the Leukemia Society of America. Received September 7, 1973; accepted February 6, 1974.

particular interest, however, because they represent na turally occurring cancers of noninbred mammals (8, 9, 16, 24, 25, 27, 39). While some of the strains of avian and mu rine oncornaviruses that have received extensive experimen tal attention might eventually be proven to cause naturally occurring tumors in outbred jungle fowl and feral rodents, evidence for this is lacking at the present time. Similarly, although vertical transmission appears largely responsible for distribution of the C-type oncornaviruses in inbred lab oratory mice and commercial chickens (21, 22), earlier speculation that this would also be true for cats (22, 37) has not materialized. Horizontal transmission had proven to be efficient in cats and now appears to be the most important route for distribution of the feline agents under natural con ditions (24, 8, 9, 16-18, 25-27, 32). Clustering of feline leukemia has been described by sev eral investigators in the past (2, 3, 17, 18, 24, 38). Such clustering of cases presents an opportunity to study spread of disease or resistance to it within the members of the clus ter. One of us (S. M. C.) recently gave a clinical description of a naturally occurring cluster where 10 cases of lympho blastic leukemia occurred within a 30-month period among a population of 35 predominantly unrelated outbred cats (4). One additional case of leukemia has occurred in this house within the last 9 months. The purpose of this paper is to present Serological data obtained on further study of this household. In addition to the greatly increased incidence of leukemia over that to be expected (5, 19, 23, 24), this popu lation was particularly interesting in that all cases had similar clinical signs and lesions, which were characterized by primary involvement of the hematopoietic system and an absence of gross tumor masses in any organ. Even though some of these cats had mild histolÃ³gica!infiltration of liver, spleen, and some lymph nodes with poorly differentiated lymphoid cells, the disease has been classified as leukemia rather than lymphosarcoma because of the major involve ment of blood and bone marrow. Most cases of spontaneous or induced leukemia in cats involve the infiltration of either thoracic or abdominal organs with malignant lymphocytes while the lymphoblastic leukemia type is less frequent (19). This suggested that the disease occurring in the above clus ter, might be due to 1 particular strain of FeLV.4 We recently described 2 Serological procedures that 'The abbreviations used are: FeLV, feline leukemia virus: FOCMA, feline oncornavirus-associated cell membrane antigen: FIP. feline infectious peritonitis; PCV, packed cell volume.

1061

MAY 1974


Susan M. Cotter, M. Essex, and William D. Hardy, Jr. proved valuable for studying the natural history of the feline oncornaviruses and for detecting horizontal transmission of FeLV under laboratory conditions. The 1st, indirect mem brane immunofluorescence with living cells, involves detec tion of antibody to FOCMA, which is induced by FeLV and feline sarcoma virus in vitro and in vivo (10 13). Using this test, rapid transmission of virus has been demonstrated to occur from cats inoculated with FeLV (8, 9, 27) to uninoculated contact controls of various ages. In the case of FeLV, a significant number of contact controls also developed leu kemia (27). The 2nd test, indirect immunofluorescence with fixed cells, involves the detection of feline oncornavirus gs (group-specific) antigens in circulating leukocytes and platelets (17, 18). This test was shown to be an effective index of virus transmission when contact control cats were exposed to cats inoculated with FeLV under field conditions (18). The presence of gs antigens detected in this manner was shown to be closely correlated with, and indicative of viremia (18). The purpose of this study is to use the 2 proce dures together in normal as well as leukemic cats in a cluster environment and to see whether serological data in normal cats reflect increased exposure to FeLV. This was measured both by the presence of viremia and antibody to FOCMA. MATERIALS

AND METHODS

Cats. Thirty-five cats lived in the leukemia cluster house hold between January 1970 and April 1973, while the study took place (Table 1). The cats were kept in a 3-room Boston, Mass., apartment with a total of approximately 600 sq ft and were never allowed out after first entering. In addition to the cats the area was inhabited by 3 dogs, 1 quinea pig, and 1 person. As Table 1 illustrates, 22 cats were present in the house when the study began, and 13 were brought in between January 1970 and January 1972. All cats were brought into the environment as strays except cats 8,11, and 21, which were born in the house. All 35 cats were 1 year of age or less at the time of entry into the house. Eighteen of the 35 cats were unrelated to any other cats in the house; all were domestics of mixed breeding except Cat 22, a Persian. Eighteen were males and 17 were females. The ages are listed in Table 1. For helathy controls from outside the cluster house, cats from the same geographical area that were brought to the Angeli Memorial Animal Hospital for such things as vacci nation were used. They were randomly selected during the period from July 1972 to April 1973 and matched by age and sex to the cats from the leukemia house. None had any known exposure to cats with leukemia. For leukemic con trols, 10 cats from the Boston area with lymphoblastic leuke mia similar in type to that seen in cats from this household were randomly selected. Case records are on file at the An geli Memorial Animal Hospital for all sick cats which lived in this house over the last 16 years. These were reviewed, and no case of leukemia was known to occur there prior to Jan uary 1970. A prospective study of the group has been in progress since January 1971, after 3 cats had developed leukemia within 7 months. All healthy cats in this household have had regular clinical examinations and hemograms at 1062

3- to 5-month intervals since the serological studies have been in progress. Autopsies were done on all cats that died after 1970. Since the start of the general study (January 1970) 18 cats have died; 11 from leukemia, 4 from FIP, 1 from aplastic anemia, and 2 from other causes (Table 1). All cats with leukemia were anemic and had similar clinical signs, which were vague initially, often limited to fever, lethargy, and anorexia. Various infections such as stomatitis, s.c. abscesses, anterior uveitis, or respiratory infections were noted in 10 of the 11 cats that eventually developed leuke mia. These infections occurred prior to the onset of anemia and the appearance of malignant cells in the marrow or blood. The anemia was hypoplastic and gradually progres sive. Normoblasts were consistently present in the blood of 9 of the II leukemic cats. The 1st apparent change in the blood in addition to anemia was the appearance of a few atypical lymphocytes. Disease progression was reflected in the gradual increase in prolymphocytes until they became the predominant cell. Lymphoblasts were consistently seen in the blood in later stages of the disease. In 2 cats (Cats 5 and 9), increasing numbers of reticuloendothelial cells were present in the blood in conjunction with abnormal lympho cytes and in 1(Cat 3) a proliferation of abnormal cells of the granulocytic series and early cells of the erythrocytic series as seen. Cells of the bone marrow generally reflected those of the blood but changes in the marrow occurred earlier. A diagno sis of leukemia was not made in any cat unless large numbers of immature lymphocytes were consistently present in blood or marrow. No gross tumor masses were present in any cat, although most had slight to moderate enlargement of the liver and spleen. At necropsy, the bone marrow, liver, spleen, and some lymph nodes were mildly infiltrated with poorly differen tiated cells mainly of the lymphocytic series (reticulum cells). Four cats developed FIP between January and March 1972. It was thought that this disease was introduced in November 1971 by two 2-month-old stray kittens (Cats 12 and 13) taken into the house. The following January, both kittens developed FIP as did 2 other cats (Nos. 14 and 15) in the ensuing 2 months. A detailed description of the clinical signs and autopsy findings in each cat with leukemia or FIP was published previously (4). Hematology and Serology. Starting in December 1971, blood samples were taken on all cats with signs of illness. In addition to routine blood counts (PCV, total and differential white blood cell counts), blood smears were prepared for detection of gs antigens. Testing was repeated at varying intervals in 6 of the leukemic cats. Similar samples were collected from all healthy cats in the house at 3- to 5-month intervals from March 1972 until April 1973. Starting in July 1972, serum was tested at 3- to 5-month intervals for the presence of FOCMA antibody. Blood samples from ageand sex-matched healthy controls as well as from 10 cats with lymphoblastic leukemia were also examined for gs anti gens and FOCMA antibody. To detect gs antigens, acetone-fixed blood smears were examined by indirect immunofluorescence. First the smears were reacted with rabbit antiserum to FeLV gs antigens that CANCER


RESEARCH

VOL. 34

Feline Leukemia Cluster was absorbed in vivo in a healthy kitten to remove nonspe cific activity against cat cells. The slide was then incubated at 37Â°for 60 min. Following 0.9% NaCl solution washes, the smears were stained with goat anti-rabbit IgG conjugated with fluorescein isothiocyanate (Hyland Laboratories, Costa Mesa, Calif.). The smears were counterstained with 0.03% Evans blue and then examined on a fluorescence microscope with a fluorescein isothiocyanate filter. This procedure re sults in specific punctate or granular staining of FeLV gs antigens in the cytoplasm of infected leukocytes and plate lets. Additional details about the performance of this test were published elsewhere (17, 18). FOCMA antibody titers were detected by reacting 2-fold dilutions of the serum in question with washed cultured leukemic lymphoblasts. The lymphoblast culture was de rived from a cat with a FeLV-induced lymphoma by Theilen et al. (34). The target cell line is harvested from culture within 24 hr of when saturation density is reached, since 95 to 100% of the cells then express the FOCMA in bands or complete rings when reacted with highly positive serum. Since the target cell can be maintained as a positive control in this manner, an estimation of the degree of antibody ac tivity for unknown samples can be determined by choosing the highest serum dilution that still gives small dots of fluorescence. All of the 3 known subgroups of FeLV are produced by the target cell line (36). After incubating the cells with serum at 37Â°for 30 min, the cells were washed twice with cold 0.9% NaCl solution and reacted with fluorescein-labeled rabbit serum to cat 7-globulin (Sylvana, Millburn, N. J.). The cells were incubated at 37Â°for 30 min, rinsed twice again with cold 0.9% NaCl solution, and resuspended in cold 50% glycerol. They were then examined on a Zeiss fluorescence microscope with a BG12 filter. For each serum sample a series of three 2-fold dilutions were checked at least 3 times under blind code. The highest dilution at which at least 50% of the cells are specifically stained is taken as the end point titer. Positive and nega tive control sera are also included in each test and read under code along with the unknown samples. RESULTS

All 11 cats that developed leukemia and 89% of those that developed any disease were present in the house when the study began (Tables 1 and 2), while only 2 of 13 (15.4%) cats entering since 1970 have died. Cats developing leukemia had spent an average of 4.5 years in the house at the time of death while cats that are still healthy had spent an average of 2.7 years in the house at the time this study was com pleted. The mean age at which cats developed leukemia was 4.6 years as compared to 3.0 years as the latest age reached by the healthy cats (Table 2). Thirteen of 17healthy cats are still below the mean age for leukemia development. Seventeen of the 35 cats (48.6%) had relatives in the house. Leukemic cats were no more likely to be related to other cats in the house than were healthy cats (Table 2). Cats whose relatives died with leukemia were no more likely to develop leukemia themselves than cats who had no relatives in this environment or those whose relatives in this environ MAY

ment remained healthy. Littermates in the cluster household showed no higher susceptibility to leukemia than unrelated cats as can be seen in Table 5. Hematological studies on healthy cats at 3- to 5-month intervals from March 1972 until April 1973 showed no sig nificant difference between PCV values, total or differential white blood cell counts, or lymphocyte morphology of healthy gi-positive cats and healthy g.v-negative cats. Of 28 cats tested for gs antigens at various times, 14 were positive and 14 were negative (Table 3). One cat (Cat 8), negative in December 1971, became positive in March 1972 and de veloped leukemia 2 months later. Three additional leukemic cats were known to be positive for 5 months (Cat 5), 8 months (Cat 10), and 9 months (Cat 11) before clinical signs of leukemia became evident. For Cat 5, physical examina tions and blood counts were conducted monthly from Jan uary 1971, when the cat was first found to be virus positive, until May 1971, when the diagnosis of leukemia was made. The cat showed no sign of illness during this time and no abnormalities were found on physical examination. In Jan uary 1971, however, mild anemia was present (PCV 27%). Lymphocytes were normal in size and morphology. Over the next 4 months, the PCV remained relatively constant but occasional atypical and immature lymphocytes were seen even though the total number of lymphocytes remained normal. In May 1971, when a diagnosis of leukemia was made, the PCV was 21%, and 95% of 18,700 white blood cells were lymphocytes, the majority being prolymphocytes. Of 14 cats that have been positive for gs antigens, 6 de veloped leukemia, 3 developed FIP, and 5 are still healthy. These 5 cats have a mean age of 2.1 years and all but 1 are 3 years of age or under. Unlike the leukemic cats, only 1 of the 5 gi-positive healthy cats was present in the house in 1970. Of the 14 cats negative forgi antigens, 2 developed leukemia and 12 are still healthy. Five cats were not tested either be cause their illnesses predated the serological study (3 cats with leukemia, 1 with anemia, and 2 that died of other causes) or because they were not in the household at the time of illness (1 kitten with FIP left the house shortly before it became ill). Six of 8 leukemic cats in the household tested for g.vanti gens were positive (Table 4). The 2 that were negative were consistently negative on more than 1 occasion (Table 3). Of 10 cats with lymphocytic leukemia from the Boston area but not from this household, 9 were positive forgi antigens. Four cats in this household had FIP. Three were tested for gÃ¬antigens and all 3 were positive. Of 10 control cats in the Boston area with FI P, 7 were positive for gs antigens. All members of the household were tested forgi antigens on 4 occasions between March 1972 and April 1973. Five of 17 healthy cats were positive on each test (Table 3). No change in g.v status was noted in any healthy cat over the year's time. Of 17 healthy cats of comparable age and breed from the Boston area not known to be exposed to leukemia, all were negative on a single test for gs antigens. Cats with gj-positive or leukemic relatives in the house were no more likely to be gs positive themselves than those cats having only healthy household relatives or no relatives living within the household (Table 4). The mean age of healthy gs-positive cats (2.5 years) was considerably lower than the mean 1063

1974


Susan M. Cotter, M. Essex, and William D. Hardy, Jr. Biographical

Table 1 data on cats in cluster household

entering

status"Died householdJuly Cat1234567891011121314151617181920212223242526272S29303132333435Age"2yr3yr3yr4 1968November 14NoneNoneNoneFather of 1967November 1967April yr8yr4yr4yr6yr6yr5yr6 1967October 1963April 21NoneDaughter of 1968April 1968November 1NoneSisterof 19; sister of 1 1966November 1966March 17Daughter of 1969November 8Sister of 19; sister of 1966November yr4 13Brother of 1971November mo.4 12Brotherof 1971November mo.4 1967April 2NoneNoneBrother of yr4 1968October yr5yr4 1968March 10NoneMother of yr5yr8yr6yr3yr3yr3yr2yr11969Julv 1968June 1Mother of 8 and 1 1966May 21Daughter of 1966November 1969January 6NoneNoneNoneBrother of 20 and 1971June 1971September 1971November 26Sister of 1971November yr1 yr2yr2yr2yrlyrlyrlyr4yr9 1971November 25NoneNoneNoneSister of 1971December 1971February 1972May 32Brother of 21 and 1972May 32Sister of 30 and 1972May 31NoneNoneNoneHealth of 30 and 1972August 19661961February yr1 1970RelationshipsNoneSister yrSexFFMFMMMFFFFFMMMMMMFFFMMMMFFMMFMFMFFDate

leukemiaDied April 1970, leukemiaDied August 1970, leukemiaDied October 1970, leukemiaDied February 1971, leukemiaDied August 1971, leukemiaDied October 1971, leukemiaDied January 1972, leukemiaDied November 1972, leukemiaDied October 1972,

leukemiaDied May 1973,

leukemiaDied March 1973, F1PDied January 1972, FIPDied January 1972, FIPDied February 1972, FIPHealthyHealthyHealthyHealthyHealthyHealthyHealthyHealthyHealthyH March 1972,

anemiaDied June 1970, pancreatitisDied October 1970, June 1970, panleukopenia

" At time of death or April 1973. whichever came first. * As of April 1973, unless otherwise indicated.

age reached by leukemic cats (5.4 years), whether gs positive or gs negative no correlation was seen between gs positivity and negativity for cats within the same litter (Table 5). Four of the 11 cats with leukemia were tested for FOCMA antibody. The geometric mean titer in this group was 0.78 at the time of death (Table 8) and 1.0 at the time of 1st diagnosis. No cat in the household that died of leukemia had a titer over 2.0 when the diagnosis was made (Table 6). The geometric mean titer of 10 control leukemic cats from outside this household was also low at 0.72 (Table 8). FOCMA antibody titers were measured on 3 occasions from July 1972 to April 1973 for the 17 healthy cats in this household and compared with a single test on an equal num ber of matched controls from outside the household (Table 7). The mean titer of healthy cats in the household in July 1972 was 4.65; that of the controls was 1.66. Of the 6 healthy cats obtained prior to 1970, only 1 cat has a titer greater than 8 while 4 of 11 cats (36.4%) obtained after that date have titers greater than 8. In cats with leukemia, titers re mained low during treatment even though many showed transient clinical improvement. The mean titers of healthy cats declined gradually over the period of time from July

1064

1972 until April 1972. Only 1 cat (Cat 27) showed a sharp rise in titer during the period of study, rising from 1 to 128 from July to December. Healthy cats that were gs positive had a mean titer of 4.20 while virus-negative healthy cats had a mean titer of 4.81. The highest titers in healthy cats were found in younger cats even though younger cats did not differ greatly from older cats in gs status. Healthy cats 1 to 2 years of age had a mean titer of 7.80 while cats over 2 years had a mean titer of 2.42. Healthy cats entering the house after 1970 had a higher mean titer (5.85) than those entering before 1970 (2.96). This finding correlates with the ages noted above. No significant difference in titer was noted between cats with relatives de veloping leukemia and cats with healthy relatives. DISCUSSION The incidence of lymphoid neoplasia in the general feline population was determined to be about 40 cases per 100,000 population at risk per year (5). In a study at the Angeli Memorial Animal Hospital, the leukemic form was reported

CANCER RESEARCH


VOL. 34

Feline Leukemia Cluster Comparison

Table 2 faciors lo health status of cats in cluster household

of genelic and environmental

Mean age"

Time in house'

No. entering house before

No. with relatives

No. with relatives developing

No. with healthy

(yr)

(yr)

1970

in house

leukemia

relatives

Category

(35)Cats All cats with leu (11)Cats kemia alldiseases with (18)Healthy cats(17)4.14.64.13.03.34.53.92.722(62.8)11

(48.6)5(45.5)8

(31.4)4(27.1)4(28.7)7(35.4)

(100)16(89.0)6(35.3)17 (44.4)9(53.0)6(17.2)2(18.4)3(16.7)3(17.6)11

1At time of death or April 1973, whichever came first. ' From 1st entry until death or April 1973, whichever came first.

Prevalence

Table 3 of gs antigens in cats in cluster household 1971

1970

Cat

since December

1970a 1973

1972

g >Â»ee Ã® L Â«i > lllÂ£Ã¤lfS|Â£

Diagnosis

u

Â¿ Â¿ S