Arthritis and Rheumatism o$cial journal of the gmerican Rheumatismassociation VOL. X, NO. 5 OCTOBER, 1967 The Epidemiology of Drug-Induced Systemic Lupus Erythematosus By MORRISSIECEL,STANLEY L. LEE AND NANCYS. PERESS I chemical agents have been implicated in the causation of systemic lupus-like symptoms in patients treated for disorders unrelated to systemic lupus erythematosus, such as essential hypertension and cardiac arrythmia. The relationship of the drug-induced disease to idiopathic systemic lupus erythematosus (SLE) is not clear. The clinical features of drug-induced lupus resemble those of SLE, with the important exception that the symptoms generally disappear upon discontinuation of drug therapy. However, whether the contributory factors in both disorders are also similar is unknown. Because of its biological significance, this question was investigated in the present study with particular reference to host factors and time trend. The variety of drugs implicated and the number of drug-induced lupus cases reported constitute a problem which warrants intensive study for clues to the underlying mechanism in the occurrence of the spon- taneous form of SLE.l In a review of the numerous therapeutic agents incriminated and the related clinical features, it was estimated that drugs were associated with the occurrence of symptoms of lupus in from 3 per cent to 12 percent of 1,193 investigated cases of SLE." The cases related to the use of drugs consisted of two broad categories: 1) the presumably early cases of idiopathic SLE which developed clear-cut symptoms following use of sensitizing drugs; and 2 ) all other cases in which signs and symptoms of lupus were first observed following drug administration, and were apparently dependent on use of the drug. To distinguish between these two groups of cases, comprehensive historical information on drug therapy and clinical findings was essential. Differentiation was particularly difficult in cases with convulsive seizures and petit mal, because such symptoms can occasionally be due to SLE. In these cases, classification by the investigators was large- From the Department of Environmental Medicine and Community Health, State University of New York, College of Medicine at New York City, and the Division of Hematology, Maimonides Hospital, Brooklyn, New York. Aided by Research Grant CD-00001 from the Division of Chronic Diseases, Bureau of State Services, United States Public Health Service, Washington, D . C . MORRIS SIEGEL, M.D., M.P.H.: Professor, Environmental Medicine & Community Health, State University of New York, Downstate Medical Center; Visiting Lecturer on Epidemiology, Hartiard School of Public Health. STANLEYL. LEE, M.D.: Director of Hematology, Maimonides Hospital, Brooklyn; Associate Professor of Medicine, State University of New York, Downstate Medical Center. NANCYS . PERESS,M.D.: Intern in Pathology, Kings County Hospital, Brooklyn, N.Y. N RECENT YEARS, 407 ARTHRITISAND RHEUMATISM,VOL. 10, No. 5 (October 1967) 408 SIEGEL, LEE, PERESS ly based on the time relationship between other manifestations of SLE and the use of anti-convulsant drugs. If manifestations subsequently appeared in other organ systems after prolonged use of such drugs, the case was considered as drug-induced SLE. TREND OF DRUG-INDUCED SLE AND IDIOPATHIC SLE NEW YORK CITY, 1957-1966 a 11M111 “ E l l MOVING AV1lAGEl - - DRUG UlAJS .METHOD The sources of data were the records of cases in voluntary and public hospitals, clinics and laboratories in 3 of the 5 boroughs of New York City, Manhattan, The Bronx and Brooklyn. The hospitals included in the study had a bed capacity of 29,182, or 62.9 per cent of the total hospital beds in the boroughs surveyed? The relevant clinical and epidemiological information was obtained by a public health nurse who annually reviewed the files of hospitals and death certificates for diagnosed cases of SLE, and communicated every 2 weeks with laboratory personnel for newly found cases with LE cells. A detailed summary was then made of past and present clinical data, with particular reference to the temporal sequence of drug use and symptoms. The study group periodically reviewed the clinical information, and the cases of lupus were classified as either SLE or drug-induced SLE according to the predetermined criteria. The criteria for a diagnosis of SLE were: involvement of at least three organ systems; laboratory confirmation by repeated findings of LE cells, or typical splenic or renal microscopic lesions; and exclusion of other diseases? The criteria for the category of “drug-induced SLE” were symptoms consistent with systemic lupus and confirmatory laboratory evidence (LE cells) following use of drugs for 15 days or longer in subjects with disorders apparently unrelated to the connective tissue disease. Excluded from the drug-induced category were cases with symptoms of connective tissue diseases antedating the use of drugs known to be implicated in the occurrence of drug-induced lupus; other disorders associated with auto-immune serological manifestation, such as myasthenia gravis; cases with inadequate information on drug therapy; and cases with obscure systemic manifestations antedal ing drug therapy. The present review is limited to cases of druginduced SLE following the prolonged use of 4 commonly implicated drugs, diphenylhydantoin, isoniazid, hydralazine and procainamide. The cases occurred among residents of New York City from 1957 to 1966, inclusive. During this 10-year period, a total of 54 cases was identified and 0 1957 0 l9b Id59 19W 19U I& 1963 YEAR OF DIAGNOSIS 1964 1965 (966 Fig. 1.-Trend of drug-induced SLE and idiopathic SLE, N e w York City, 1957-1966 (three-year moving average). systematically investigated for associated host and environmental factors. A summary of the results follows, including a comparison with data on 101 cases of SLE observed during the same study period among residents in a defined area of New York City with a population of 662,000 in 1960.1 Moreover, the host factor data for drug-induced lupus are compared with the expected values. Because the population at risk to drug therapy is unknown, expected values are based on the prevalence of the disorders for which the abovementioned drugs are commonly used: epilepsy, tuberculosis, hypertension, and cardiac arrythmia when complicating myocardial infarction.”’.’” Adequate prevalence data by race were not obtainable for epilepsy, and so an expectancy based on racial distribution in the population for the commonly affected age groups was assumed. The only specific prevalence data available for Puerto Ricans were for tuberculosis. In estimating the expectancy of other diseases in Puerto Ricans, the rates for whites were applied. RESULTS The time trend for the occurrence of cases of drug-induced lupus and comparative data on the incidence of SLE are shown in Fig. 1. The annual number of recognized drug-lupus cases rose somewhat at first and then stayed at an even level for several years before increasing markedly in 1965 and 1966, the last 2 years of the study period. By contrast, data on the incidence of idiopathic SLE showed no significant rise in rate during the entire period. 409 DRUG-INDUCED SYSTEMIC LUPUS ERYTHEMATOSUS Table 1.-Drug-Induced Year of Diagnosis Total 1957 1958 1959 2 2 3 1960 6 4 1961 1962 1963 1964 SLE by Year of Diagnosis and Drug Therapy, 1957-1966 Number of Cases by Implicated Drug Diphenylhydantoin Isoniazid Hydralazine 1 1 1 3 3 Procainamide 1 3 2 1 1 1 1 1 2 2 1965 1966 2 5 1 13 16 3 2 2 2 1 7 3 9 Tofaat: 54 16 9 11 18 1 The relationship of the number of druginduced lupus cases diagnosed each year to the specific drugs used is shown in Table 1. The cases noted in the early years of the study period were attributable to diphenylhydantoin, hydralazine and isoniazid. The abrupt rise in 1965 was due to procainamide hydrochloride. Over the entire period, most of the cases were related to the use of procainamide, followed in decreasing order by diphenylhydantoin, hydralazine and isoniazid. The many cases associated with procainamide in 1965 and 1966 were unexpected in view of the few cases noted in previous years. The abrupt increase in 1965 may be due in part to greater awareness of the complication since 1962, when it was first reported. The increase may also be associated with the more extensive use of procainamide, as evident in a two-fold rise in the volume of sales of this drug in 1965 and 1966.8 Analysis of the data by host factors, such as age, sex and race, is shown in Tables 2 to 5. The distribution of cases varied directly with age, with 5.6 per cent under 15 years of age and 55.6 per cent over 44 years of age (Table 2 ) . The average age at time of diagnosis for all cases was 47.4 years, as compared with 35.8 years for the 101 cases of SLE observed during the same period. The age distribution of cases reflected the age of the population at risk to the drugs used. Thus, there were only 2 (12.5 per cent) of 16 epileptic patients treated with diphenylhydantoin over 44 years of age as against 25 (86.2 per cent) of 29 subjects on hydralazine or procainamide for hypertensive and cardiac disorders, respectively. The average age at onset of drug-induced lupus in the different groups was 33.4 years for diphenylhydantoin, 40.3 for isoniazid, 53.5 for hydralazine and 60.7 for procainamide. The distribution of drug-induced cases by sex is given in Table 3. There was a preponderance of female cases, 38 (70.4 per cent) as against 16 (29.6 per cent) males, a sex ratio of 2.4. These values were significantly different from those observed for idiopathic SLE.* Of the 101 cases of SLE under consideration, 89 (88.1 percent) were female and 12 (11.9 per cent) male, a sex ratio of 7.4. As shown in Table 3, the distribution of drug-induced lupus cases by sex varied with the drug used. Thus, the preponderance of females was marked for diphenylhydantoin, while the cases were more equally distributed between males and females for procainamide. However, since these differences depend on the number of male and female subjects treated, they * Value of P = 0.05 or less. 410 SIEGEL, LEE, PERESS Table 2.-Age Distribution of Drug-Induced SLE by Drug Therapy Total NO. Under 15 15-24 25-34 3544 45-54 55-64 65-74 75 & over Total: Mean age: % 3 5.6 5 9.2 9.2 20.4 18.5 16.7 20 4 0.0 5 11 10 9 I1 0 54 47.4 100.0 Number of Cases by Implicated Drug Diphenylhydaiitoin Isoniazid Hydralazine 3 3 3 5 2 16 33.4 were compared with expected values based on prevalence As shown in Table 3, the observed ratio of female to male cases exceeded the expected ratio in each drug category. Furthermore, the extent of the difference between observed and expected ratios was two-fold or more, an indication that the female may be considerably more susceptible than the male to drug-induced lupus, unless we are to assume that females were treated twice as much. Specific rates by sex could not be determined because of lack of adequate information on the population under treatment. The foregoing data on age and sex were compared with case reports of drug-induced lupus published by others. The cases considered were limited to those with a positive LE cell test. The cumulated data on 47 cases reported by others are shown in Table 4 by the specific drugs under cons i d e r a t i ~ n . ~ , The ~ J ~ current study is broadIy in agreement with the resuIts of others. They too showed a preponderance of cases among females, which appeared to be most marked for hydralazine and diphenylhydantoin. And the age at onset was lowest for diphenylhydantoin, followed in increasing progression by hydralazine and procainarnide. The association between drug-induced lupus and other host factors, such as race, Pracainamde 2 2 2 1 3 3 2 2 J" 4 7 6 9 40.3 11 53.5 18 60.7 1 CASES OF DRUG INDUCED SLE BY RACE NEW YORK CITY, 1957-1966 501 r a Observed 0Expected WHITE PUERTO RI CAN Number Number NEGRO Fig. 2.-Cases of drug-induced SLE by race, New York City, 1957-1966. was difficult to determine because of the lack of adequate information on the number of subjects treated in each racial group. Some indication of the relationship to race may be evident by comparing the observed number of cases in each racial group with the expected number as estimated from prevalence data or from the population distribution, as described above.5,6,? The results are summarized in Fig. 2, which shows the observed and expected number of cases for Negro, Puerto Rican, and other white groups. The observed and the expected values were 411 DRUG-INDUCED SYSTEMIC LUPUS ERYTHEMATOSUS Table 3.-Sex Distribution of Drug-Induced SLE by Implicated Drug - __ Sex Total Number of Cases by Implicated Drug Dlphenylhydantoiri Isoniazid Hydralazine Procainamide XO. To Male Female 16 38 29.6 70.4 2 14 3 6 3 8 10 Total: 54 100.0 16 9 11 18 Observed Ratio* Expected Ratio 2.4 0.7 7.0 1.o 2.0 0.5 2.7 1.3 0.5 Observed Ratio Expected Ratio 3.4 7.0 4.0 2.1 2.6 * Itatlo of 8 1.3 female to male cases comparable in each racial group, an indication that there was no apparent racial difference in the occurrence of druginduced lupus. The failure to demonstrate a racial difference for drug-induced lupus was at variance with the results observed for idiopathic SLE. As reported previously, the prevalence of SLE was found to be highest for non-whites, followed in descending order by the rate for Puerto Ricans and then other whites. The morbidity rates for SLE were 2 to 3 times greater among Negroes than whites.ll The foregoing difference in racial characteristics between drug-induced and idiopathic lupus seems to hold for each of the drugs under consideration (Table 5 ) . As summarized in Table 5, the number of observed and expected cases by race and specific drug was comparable in every instance. In Negro subjects, there was no increase of drug-lupus for any of the drugs incriminated thus far. This was most evident for the procainamide group, where none of the 18 cases was a Negro. However, the expectancy among Negroes in this drug group was only 2.0 cases, and consequently a larger experience would be required to determine the biological significance of these results. The study on drug-induced lupus led to a search for evidence of a predisposition to the disorder as manifested by the presence of serological abnormalities before the onset of therapy or before the appearance of drug reactions in treated subjects. The sera were examined for LE cells, antinuclear antibodies, direct Coombs’ reaction, rheumatoid factor, biologic false-positive reaction for syphilis, and for protein fractions by paper electrophoresis by methods previously described.l The expected frequency of serological abnormalities was explored in a pilot study of 80 subjects under therapy. Half of the group had active pulmonary tuberculosis, and the remainder were equally divided between epilepsy and essential hypertension. The serological findings were within normal limits except for antinuclear antibodies in 4 (10 per cent) tuberculous patients, and elevated gamma globulin levels in 13 (34 per cent) tuberculous patients, 3 (15 per cent) hypertensives and 1 ( 5 per cent) epileptic. The average gamma globulin level was significantly higher in the tuberculous patients than in the others. Following the pilot study, a prospective, longitudinal investigation was made of serological abnormalities before and after the onset of drug treatment for hypertension, epilepsy and tuberculosis. Serial bleedings were obtained shortly before or less than 1 week after onset of treatment, 1 month later and then every 6 months thereafter. A review of the first 50 cases in the program of which 29 were taking isoniazid for tuberculosis, 10 taking a variety of drugs for 412 SIEGEL, LEE, PERESS Table 4.-Neported Cases of Drug-Induced SLE by Age, Sex, and h D w Male Total Under 30 60 & Total Under 30 60 Bi 30 to 59 Over to Diphenylhydantoin Isoniazid Hydralazine Procainamide Total: * Ratio of 6 2 4 4 16 30 59 5 1 5 g Therapy Female Over 1 1 4 2 8 14 0 10 2 7 3 31 13 1 1 8 1 3 14 4 12 5 Sex Ratio> 2.3 012 2.5 1.8 1.9 female to male cases. hypertension and 11taking diphenylhydantoin or trimethadione for epilepsy, revealed no increase in serological abnormalities attributable to drug therapy. On the contrary, the elevated gamma globulin level in tuberculous subjects fell progressively with treatment and clinical improvement. Nor could the frequency of antinuclear antibodies be correlated with duration of therapy. The period af follow-up varied from less than a year in 32 (64 per cent) cases to 1 to 3 years in 18 (36 per cent). From these preliminary results, the serological abnormalities observed originally in the pilot study appear to be largely a manifestation of the disorder treated and not a consequence of the drugs used. It is also apparent that the frequency of the other serological abnormalities, which might be attributable to drug therapy, was so low that a very large series of patients would be required for an adequate investigation. DISCUSSION In the current study, comparative epidemiological data were presented on druginduced SLE and idiopathic SLE occurring during the past decade in New York City. The results reveal significant similarities and differences in respect to contributory factors. Thus, the preponderance of females, so typical of SLE, was also observed for drug-lupus, although the difference was less marked. Distinctive differences between drug-inducrd lupus and idiopathic SLE were ob- served for such characteristics as race, age and time trend. These may reflect significant differences in causation. Unlike SLE, there was no evidence of an increase in Negroes. Similar results have been noted by Holley in Birmingham, Alabama.12 The occurrence of drug-induced lupus fluctuated over time, varying with the drugs in popular use, and the dominant conditions and age group of the subjects treated. On the other hand, in the past 10 years at least, the incidence of SLE has been fairly constant and comparable in such widely separated areas as New York and Malmo, Sweden.13 In addition, the disorder has been largely concentrated in females of child-bearing age regardless of time or place. The recent increase of recognized cases of drug-induced lupus need not cause undue alarm, because it partly reflects greater awareness of the complication and thus earlier and more complete casefinding. Early recognition of the complication is particularly important in elderly individuals with impaired functional capacities. The complication should be suspected in both male and female subjects with chronic disorders under prolonged drug therapy. The selective action of the drugs in affecting only certain individuals under prolonged treatment reflects the presence of predisposing host factors which might be identified in prospective studies of drugtreated cases. In this connection, our sero- 413 DRUG-INDUCED SYSTEMIC LUPUS ERYTHEMATOSUS Table 5.-Racial Race Distribution of Drug-Induced SLE by Drug Therapy . , Total Obs. Exp. Dinhenvlhydantoin Obs. Exp. Isoniazid Obs. Exp. Hydralazine Exp. Obs. Procainamide Obs. Exp. White Puerto Rican Negro 41 6 39.0 12 2 11.7 1.7 4 2 4.5 1.3 8 4.5 1 7.6 0.7 17 1 15.2 0.8 7 10.5 2 2.6 3 3.2 2 2.7 0 2.0 Total: 54 54.0 16 16.0 9 9.0 11 11.0 18 18.0 Obs. = Observed number of cases; Exp. = Expected number of cases. logical studies of subjects observed before and after treatment with isoniazid, anticonvulsants and anti-hypertensive drugs failed to reveal evidence of such a predisposition, nor evidence of an increase in serological abnormalities following prolonged drug therapy. These results are inconclusive because of small sample size. They differ somewhat from a report by others that the frequency of antinuclear factors is higher in tuberculous patients treated with isoniazid for one year or more than in untreated patients14 In other studies bearing on the underlying mechanism, a relationship to hyperglobulinemia and to familial aggregation of connective tissue abnormalities has been reported by various investigators in an occasional drug-lupus case.15J6 The information available is only suggestive of an association with an overactive immune mechanism or an hereditary tendency. A relationship to such specific processes as the formation of antigen-antibody complexes and the release of lysosomes has not yet been demonstrated in subjects with drug-induced SLE.17 Whether the underlying mechanism responsible for drug induced SLE is similar to that postulated for idiopathic SLE is still not clear.18 Nevertheless, the fact that different ‘chemical substances can induce symptoms of lupus which disappear upon withdrawal of the offending drugs makes it reasonable to suppose that chemical or biological agents might also be implicated in the development of symptoms of SLE, and that clinical remission in SLE might follow elimination of the offending agents or neutralization of their ill effects. This concept is consistent with the character of the disorder, viz, its episodic course and protean manifestations. It emphasizes the need for renewed efforts in search of the specific agents involved and of methods for counteracting their effects. In this connection, a filterable viruslike agent has been observed in a susceptible crossbred strain of NZB/BL mice that develop renal disease and other manifestations resembling SLE.19v20Neonatal Swiss mice also appeared to be susceptible to the filterable agent, developing “features of renal and hemolytic disease” and positive Coombs’ tests following intraperitoneal inoculation.20This finding provides new evidence for the viral etiology of auto-immune disorders, and promises to make possible the investigation of contributory factors and underlying mechanism in SLE and related connective tissue diseases under laboratory conditions. SUMMARYAND CONCLUSIONS An increase in the incidence of druginduced SLE was observed among New York City residents from 1957 to 1966 in an epidemiological study of the disorder following prolonged use of diphenylhydantoin, hydralazine, isoniazid and procainamide. The increase was partly related to greater awareness of the complication and 414 SIECXL, LEE, PEHESS to more widespread use of procainamide. IIost factors associated with drug-induccd hipus varied with the d r u g used a n d the population treated as follows: 1 ) Fernales seemed to be m o r e susceptible than males. The extent of the difference appeared to be a b o u t two-fold. It was found to be somewhat greater for diphenylhydantoin t h a n for other drugs. 2 ) The m e a n age for the occurrence of druglupus varied from 33.4 years for epileptic patients treated with diphenylhydantoin to 60.7 years for cardiac subjects treated with procainamide. 3 ) There appeared to be no significant racial difference in susceptibil- ity to drug-induced lupus. A comparison of epidemiological data o n drug-induced SLE and idiopathic SLE ob- served during the same period of study revealed differences in time trend, a n d host factors which may be of etiological significance. The possibility of a drug-induced disorder following prolonged d r u g therapy should be considered in cases diagnosed as SLE. Drug-induced serological abnormalities characteristic of SLE were not detected in small groups of patients tested before a n d after treatment for epilepsy, tuberculosis and hypertension in a preliminary prospective study. The feasibility of a definitive study under present conditions is questioned because of the large sample that would be required for significant results. SUMMARIO IN INTERLINGUA Comparative datos epidemiologic relative a pharmacogene lupus e a idiopathic disseminate lupus erythematose, occurrente inter 1957 e 1966, ha revelate siinilitudes e differentias de signification biologic. Un abrupte augment0 in lupus inducite per pharaccompaniate de un simile disveloppamento inacos occurreva in 1965 e 1966. Isto-non con respecto a disseminate lupus erythematose-esseva attribuibile a un plus extense LISO de procainamida. Le characteristicas del lupus phaimacogene con respecto a1 distribution per etate e sexo del subjectos afficite dependeva in parte del popillation tractate. Per contrast0 con disseminate lupus erythematose, lupus pharmacogene revelava nulle evidentia de differentias racial con respecto a1 susceptibilitate. lp REFERENCES 1. Siegel, M., Lee, S. L., Widelock, D., Reilly, 2. 3. 4. 5. 6. E. B., Wise, G. J., Zingale, S. B., and Fnerst, H. T.: The epidemiology of systcmic lupus erythematosus. Preliminary results in New York City. J. Chron. Dis. 15:131-140, 1962. Lee, S. L., Rivero., I., and Siegel, M.: Activation of systemic lupus erythematosus by drugs. Arch. Int. Med. 117:620-626, 1966. Hospitals and related facilities in southern New York. Jannary 1, 1965, Hospital Review and Planning Council of Southern New York, New York City. Lennox, W. G.: Epilepsy and related disorders. New Ylork, Little, Brown & Co., 1960, Vol. 1, p,, 502. Lowell, A. M.: 'l'uberculosis in New York City, 1961. New York Tuberculosis Association, New 'York, 1962, pp. 31-36. National Center for Health Statistics, Division of Health Examinations Statistics: Blood Pressure of Adults by Age and Sex, 7. 5. 9. 10. 11. United States, 1960-G2. U. S. Dept. Health, Education & Welfare, Public Health Service Publication No. 1000-, Series 11-No. 4, June 1964, p. 9. National Center for Health Statistics, Division of Health Examinations Statistics: Coronary Heart Disease in Adults, United States, 1960-62. U. S. Dept. Health, Education & Welfare, Public Health Service Publication No. 1000-, Series 11-No. 10, Sept. 1965, pp. 17-19. Lavy, N. W.: Personal communication. Kaplan, J. M., Wachtel, H. L., Czarnecki, S. W., and Sampson, J. J.: Lupus-like illness precipitated by procainamide hydrochloride, J.A.M.A. 192:444-447, 1965. Paine, R.: Procainamide hydrochloride and lupus erythematosus. J.A.M.A. 194:23-26, 1965. Siegel, M., Reilly, E. B., Lee, S. L., Fuerst, H. T., and Seelenfreund, M.: Epidemiology of systemic lupus erythematosus: time DRUG-INDUCED SYSTEMIC LUPUS ERYTHEMATOSUS 12. 13. 14. 15. 16. trend and racial differences. Amer. J. Public Health 54:3343, 1964. Holley, H. L.: Personal communication. Leonhardt, T.: Family studies in systemic lupus erythematosus. Acta Med. Scand. 176 ( SUPPI. 416) :37-38, 1964. Cannat, A., and Seligmann, M.: Possible induction of antinuclear antibodies by isoniazid. Lancet 1: 185-187, 1966. Shulman, L. E., and Harvey, A. M.: The nature of drug-induced systemic lupus erythematosus. Arthritis Rheum. 3:464, 1960. Holley, H. L.: Evidence for a predisposition to rheumatic diseases in families of patients developing drug-induced systemic lupus erythematosus. Arthritis Rheum. 7:684686, 1964. 415 17. Weissmann, G.: Lysosomes, New Eng. J. Med. 273:1084-1090 and 1143-1149, 1965. 18. Conference on the immunologic aspects of rheumatoid arthritis and systemic lupus erythematosus. Chap. XII, Shulman, L. E. (Chairman) : Inducing agents and relationship to other diseases. Arthritis Rheum. 6: Suppl. 558-571, 1963. 19. Helyer, B. J., and Howie, J. B.: Renal disease associated with positive lupus erythematosus tests in a cross-bred strain of mice. Nature 197: 197, 1963. 20. Mellors, R. C., and IIuang, C. Y . : Immunopathology of NZB/BL mice. V. Viruslike (filterable) agent separable from lymphoma cells and identifiable by electron microscopy. J. Exp. Med. 124:1031-1038, 1966.