The Significance of Antinuclear Factors in Rheumatoid Arthritis By JOHN J. CONDEMI, EUGENE V. BARNETT,EDWARD C. AnvArER, RALPHF. JACOX, EDWARD S. MONGANAND JOHN H. VAUGHAN Sixty-eight patients with definite or classic rheumatoid arthritis were studied to determine the significance of the presence of antinuclear factors. The patients with positive antinuclear factor determinations were those with more severe disease. These patients almost always had subcutaneous nodules, which were not always associated with a high latex titer for rheumatoid factor. Necrotizing vasculitis occurred in both antinuclear factor positive and negative patients. Our data do not support the notion that a positive antinuclear factor test or a positive L.E. cell test should be used to exclude the diagnosis of rheumatoid arthritis. Sexanta-octo patientes con definite o classic arthritis rheumatoide esseva studiate con le objectivo de determinar le signification del presentia de factores antinuclear. Esseva constatate que le patientes in qui le determinationes de factor antinuclear esseva positive esseva etiam le patientes con grados sever del morbo. Iste patientes habeva in quasi omne le casos nodulos subcutanee le quales esseva non semper associate con un alte titro a latex pro factor rheumatoide. Vasculitis necrotisante occurreva in patientes tanto positive como etiam negative pro factor antinuclear. Nostre datos non supporta le these que un positive test pro factor antinuclear o un positive test pro cellulas de L.E. pote esser usate in excluder un diagnose de arthritis rheumatoide. T HE APPLICATION O F Coon's immunofluorescence method' to demonstrate the presence of antinuclear factor^^-^ in the serums of patients with systemic lupus erythematosus (SLE) has made available both a valuable research tool and a useful diagnostic test. It soon became obvious, however, that the technique was revealing these factors to be present in other diseases, such as rheumatoid arthritis (RA), juvenile rheumatoid arthritis, discoid lupus, scleroderma, dermatomyositis, polyarteritis nodosum, liver disease, ulcerative colitis, thyroid disease, myasthenia gravis, and 5 per cent of normal people. Although the incidence of antinuclear factors (ANF) in systemic lupus erythematosus has been reported to be from 97-100 per cent by most workers, From the Departmmt of Medicine, Uniuersity of Rochester School of Medicine and Dentistry, Rochester, N . Y. Supported b y research grants AI 02349 and AM 02443 from the National Institutes of Health and from the National Foundation, Bethesda, Md. JOHN J. CONDEMI, M.D.: Senior Instructor in Mediche, Assistant Physician; EUGENE V. BARNETT,M.D.: Assistant Professor of Medicine, Associate Physician; EDWARD C. ATWATER, M.D.: Assistant Professor of Medicine, Associate Physician; EDWARD S . MONGAN, M.D.: Assistant Professor of Medicine, Associate Fhysician; RALPH F. JACOX, M.D.: Professor of Medicine, Physician; and JOHN H. VAUGHAN,M.D.: Professor of Medicine, Physician, Recipient, Research Career Award from the National Institutes of Health, Bethesda, Md., all at Strong Memorial Hospital, Rochester, N . Y. 1080 .dRlHRITIS AND RHEUi%fAlISM, VOr . 8, NO. 6 (I)kC EhiAEA), I!i)65 SI(;NIE’ICANCE:OF ANTINUCLEAR FACrOHS 1081 there has been a wide variation in the reported incidence of ANF in rheumatoid arthritis-e.g., less than 10 per cent by Friou,5 16 per cent by Weir arld Holbrow,c 16 per cent by Rothfield et al.,‘ 36 per cent by Hall et al.,s 36 per cent by Witebsky,s and 65 per cent by Alexander et a1.I0 This variation has most often been attributed to differences in sensitivity of the systems employed to demonstrate ANF. Another explanation may be patient selection. Alexander, Bremmer, and Duthie,lo who reported the highest incidence, included in their series only patients with disease of 8 or more years’ duration. The number of patients with Felty’s syndrome in any series may also increase the number of positive reactions obtained.ll With these differences in reported incidences of ANF in patients with rheumatoid arthritis and the uncertainty of the significance of the test in classification, treatment, and prognoses, we decided to study a group of patients with RA from our clinic to determine the incidence in them of ANF as we measure it, and to assess its clinical significance. SELECTIONOF PATIENTS AND METHODOF STUDY During a two-year period (1961-1963), we began utilizing the immuncfluorescence method for detecting ANF and obtained positive results in 18 of 20 patients with SLE, 36 of 132 patients with adult RA, 5 of 15 with juvenile RA, 3 of 75 normal people, and 18 of 143 with a variety of other diseases (Table 1 ) . Among the 316 patients with negative tests in the miscellaneous group were cases of dermatomyositis, Laennec’s cirrhosis, viral hepatitis, Behcet’s syndrome, drug reactions, chronic infections, rheumatic fever, rheumatic heart disease, syphilis, serum sickness, thyroid disease, and solar dermatitis. The patients’ sera were examined because they were sent to 11s by the attending physicians, who considered that autosensitivity might have been playing a role in the patients’ illnesses. The two cases of systemic liipus erythematosus with negative ANF determinations in Table 1 were well-documented cases in which earlier sera had exhibited positive L.E. cell tests, but had been in long remissions when the sera were taken for ANF testing. For the present study, 68 patients from the Arthritis Clinic were investigated. About one-half were on corticosteroids at the time the study was performed. The remainder had never been on corticosteroids or had not received steroids for at least one year. An equal number of the patients were in anatomical Stages I or I1 or in Stages I11 or IV by Steinbrocker’s classification12 ( p. 271 ) . Patients on corticosteroids were equally distributed between the groups with mild disease (Stages I and 11) and the groups with more severe disease (Stages I11 and IV). All patients fulfilled the ARA diagnostic criteria fur definite RA12 (p. 234), except that a positive L.E. cell preparation did not exclude a patient from the study. Each patient was given a questionaire about onset and duration of disease, joints involved, sun sensitivity, nervous system symptoms, leg ulcers, Raynauds phenomenon, and family history. A complete examination was then performed by one of our group. The laboratory studies included a complete CONDEMI ET AL. 1082 Table 1.-Zncidence of Antinuclear Factors in Disease TESTED DIA6NOSIs POSITIVE PER CENT POSITIVE 20 18 90 RHEUMATOID ARTHRITIS I32 36 27 JUVENILE RHEUMATOID ARTHRITIS IDIOPATHIC THROYBOCYTOPENIC WRPURA 15 5 33 8 3 ULCERATIVE COLITIS 8 2 DISCOID LUPUS ERYTHEMATOSUS 6 2 SCLERODERMA 5 2 MYASTHENIA GRAVIS 5 2 SYSTEMIC LUPUS ERYTHEMATOSUS 7 I OSTEOARTHRITIS 31 2 ASTHMA 20 I KIDNEY DISEASE 28 I IDIOPATHIC PERlCARDlTlS II I MYLERAN-TREATED CHRONIC MYOLOGENOUS LEUKEYIA I4 I TOTAL 310 NORMALS 71 POLYARTERITIS NDDOSUY blood count, urine analysis, sedimentation rate, paper electrophoresis, Wassermann test, chest x-ray, electrocardiogram, and biopsy to confirm the diagnosis of vasculitis whenever possible. MATERIALSAND METHODS To demonstrate the presence of antinuclear factors, peripheral smears were made from the bloods of hospitalized patients with elevated peripheral blood leukocyte counts but with no evidence of autosensitivity diseases. The smears were dried for one hour before a fan and fixed in 95 per cent ethanol at 37 C. for 10 minutes. They were dried again at 37 C. for 30 minutes and used immediately or frozen at -20 C. for future use. Prior to testing, the slides were thawed before a fan and rinsed in buffered saline. The test serum was applied for 30 minutes, with care taken to prevent drying. The excess serum was then Iremoved by washing for 20 minutes with two changes of phosphate buffered saline. Fluorescein-labeled rabbit antihuman gamma globulin was then applied for 30 minutes. This was also washed with two changes of saline and then covered with a 90 per cent glycerin solution and overlayed with a cover slip. The rabbit antihuman gamma globulin for labeling was prepared by immunization with alum-precipitated Cohn Fr. 11.13 The antisera reacted only with y G globulin on immnnoelectrophoresis against normal serum. By Ouchterlony analysis, there were cross reactions with gamma A and gamma M globulins.* *According to the new convention, the terms yG. yA, and yM are used in preference to the prior terms y2 ( y s s ) , ylA ( B2A), and ylM (B,M) respectively.14 1083 SIGNIFICANCE OF ANTINUCLEAR FACTORS Table 2.-Relationship X. FEMALES 2 4 ANF to Stage* of Disease n. et 13 3 8 4 6 5 5 13 30 33 5 J m. m. TOTAL of J 5 13 J 5 4 27 5 5 5 54 *ANATOMICAL STAGE AS MFINEO BY STEINBROCKER. 12 fONE OF THESE EXHIBITED A POSITIVE ANF OETERYINA1,TK)N ONLY WITH ALTERED CELLS. The modification of Marshall et al.,15 was used in coupling fluorescein isothyocyanate to globulin. Nonspecific staining was removed by absorption with acetone-dried rabbit liver powder.16 A Zeiss microscope with an OSRAM HBO 200 W. mercury arc light source was used. In viewing the slides, a BG 12 exciting filter and 500 blocking filter were used. The sera tested were used iininediately or stored for short periods of time at -20 C. Titers were determined by twofold dilutions of serum and are expressed as reciprocals of the highest dilution giving a pozitive reaction. The latex fixation test was a modification of the Singer and Plotz method as described by Atwater.17 The L.E. cell preparations were generally performed on the same bleeding as the ANF test and the defibrinationls method was used. In one patient with a positive L.E. cell preparation the ANF were demonstrated in the serum only when altered nuclear material was used19 Titers were not performed on this sera. RESULTS Of the 68 patients studied there were 24 who were ANF positive and 44 who were ANF negative (Table 2 ) . Eleven of 25 males and 13 of 43 females were ANF positive. The frequency of males and females in this series of patients with rheumatoid arthritis is within expected limits. The greater frequency of occurrence of ANF in males over females is statistically significant only at the 10-20 per cent level. Of 34 patients in Stages I11 and IV, 18 were ANF positive, while only 6 positive tests were seen in 34 patients who were in Stages I and 11. This difference of distribution is significant at the 1 per cent level. The highest ANF titers (32 and above) occurred in patients with disease in Stage I11 or IV. As can be seen, however, there were also patients in Stages I11 and IV who had low titers of ANF. Table 3 shows that there was also an increased incidence of positive ANF tests in patients in functional Class 111 or IV. Thirty-eight patients were in Class I11 or IV and 18 were ANF positive, while of 30 in Class I or I1 only 6 were ANF positive. This difference is significant at the 1 per cent level. CONDEMI ET AL. 1084 Table 3.-Relationship of ANF to Class" of Disease MALES FEMALES I ~ T A 1L 1 1 3 3 0 1 I I I I I I I I ~~ *FUYCTIOWAL CLASS AS DEFINED BY STEINBRDCKER. ' 2 tDNE OF THESE EXHIBITED A POSITIVE ANF DETERMIMATIOW ONLY WITH ALTERED CELLS. Table 4.-Relationship of (o ANF to Age of Patient MA1-ES FEMALES ONE OF THESE EXHIBITED A POSITIVE ANF DETERYINATION ONLY WITH ALTERED CELLS. Is 1085 SIGNIFICANCE OF ANTINUCLEAR FACTORS Table 5.-Relationship of ANF to Duration of Disease I MALES FEMALES IN ONE OF THESE, A POSITIVE ANF MTERHINATION OCCURRED ONLY WITH ALTERED CELLS. 19 The correlation between positive A N F determinations and more advanced stages and classes of disease was more marked in males. There was no correlation between titer of A N F and functional class of rheumatoid arthritis. The patients’ ages, recorded in Table 4, are the ages attained at the time of the study. As can be seen, the men were all over 40, and the incidence of positive ANF tests increased with age. This changing incidence, however, is not statistically significant. The women had disease at earlier ages, and no significant correlation was noted between presence of ANF and age. However, it may be noted that in the combined sexes, 9/19 (47 per cent) of patients over 60 years were ANF positive, while only 15/49 (30 per cent) who were under 60 years were A N F positive. In Table 5 the duration of disease is compared to the presence of ANF. In neither sex could the presence of A N F be correlated with greater duration of disease at a high probability, but in each sex the highest incidence occurred in those with disease of longest duration. The presence or absence of ANF is correlated in Table 6 with several clinical signs considered to reflect the vasculitis of rheumatoid arthritis. As can be seen, nodules21 occurred in 20 of 24 patients with positive tests but in only 15 of 44 with negative tests. This difference is significant at the 1 per cent level. Symptoms or signs of peripheral neuropathy, including paresthesias, decreased vibratory or pinprick sensation, or absent reflexes occurred without a significant difference in ANF-positive and ANF-negative groups. Nail fold thrombi22 (Fig. 1) occurred in 2 of 24 A N F positive patients and 1086 CONDEMI ET AL. Table 6.-Euidence of Vasculitis in 4 of 44 of the ANF-negative patients. Two patients (R. C . ) and (L. R.) had foot drop and positive biopsies for vasculitis (Figs. 3 and 5 ) . Both had gangrenous changes in the lower extremities (Figs. 2 and 4 ) . Patient L. R. was ANF negative; patient R. C. was ANF positive. The ANF-positive patients were also compared to the ANF negative patients for the incidence of infection, Raynauds phenomenon, drug sensitivity, sun sensitivity, family history of rheumatic disease, gamma globulin level, sedimentation rate, hematocrit, blood count, urine protein and sediment, Wassermann test, and steroid treatment. No significant differences were Fig. 1.-Nail fold thrombi. SIGNIFICANCE OF ANTINUCLEAR FACTORS Fig. 2.-A 1087 gangrenous leg ulcer in a patient (R. C.) who was ANF positive. noted. There were 14 positive ANF tests in 35 patients taking corticosteroids and 10 positive tests in 33 patients not on steroids. In Table 7 the ANF titers of patients with RA are compared to those of patients with SLE. In general, it can be seen that patients with SLE had higher ANF titers than did patients with RA. When the RA patients with positive L.E. cell tests were studied by the three layer immunofluorescent test for specific immune globulins, nearly all had rG ANF.20 The 7 G ANF titers were significantly lower in the RA patients than in those with SLE. The two groups did not differ significantly in their r M ANF titers. Only 3 of the 7 patients with RA and positive L.E. cell tests had evidence of vasculitis. All 3 had leg ulcers and peripheral neuropathy. One of the 3 also had nail fold thrombi. Table 8 demonstrates the relationships existing among F I1 latex titers, ANF titers, nodules, and L.E. cell tests. It is obvious from the spread of the points that there is no correlation between titers of ANF and of F I1 latex flocculation. In the 44 patients with negative ANF tests (left column) ncdules occurred more frequently in those with higher latex titers (640 and 1088 CONDEMI ET AL. Fig. 3.-The vasculitis noted in patient (R. C.) at postmortem examination. There is a perivascular mononuclear inflammatory reaction with necrosis and hyaline degeneration of the media. above). In the ANF-positive group, the presence of nodules was not related to high F I1 latex titers. In 3 patients with negative ANF tests and 3 patients with positive ANF tests, nodules were noted in the past but were not found at the time of this study. The patients with nodules had either positive F I1 latex tests or ANF tests or both. Positive L.E. cell tests were limited to patients with positive ANF determinations, but they occurred more frequently if the patient also exhibited a strongly positive latex test. Only 1 of 7 patients with positive L.E. cell tests failed to exhibit nodules. In our series we did not observe positive F I1 latex tests in patients with active systemic lupus erythematosus, although others have reported positive tests in up to 25 per cent of SLE patients7 DISCUSSION The presence of ANF in patients with RA has been correlated in other studies with severity of disease, duration of disease, vasculitis and infection. SIGNIFICANCE O F ANTINUCLEAR FACTORS Fig. 4.-A 1089 gangrenous toe in a patient (L. K . ) who was ANF negative. This study confirms the impression of Alexander et al.'" and Hall et a1.,8 that ANF is more apt to be present in the serums of RA patients with severe disease, but we could not correlate it with the presence of vasculitis or infection. were able to find a positive correlation between the Alexander et presence of ANF and duration of disease. Hall et a1.8 stated that their patients with RA and positive ANF tests had a mean duration of disease of 8.4 years and those with negative ANF tests had a mean duration of 7.4 years. Although we found the highest percentage of positive ANF tests in patients with disease of longer duration, many positive tests were also seen in patients with disease of shorter duration. It is to be noted, however, that 17 of the 34 patients in our group in Stage I11 or IV had disease of less than 4 years duration. There were 7 positive ANF tests in this group of 17 patients. Because Alexander and Duthie selected for their study patients with disease of 8 or more years' duration, patients with early rapidly progressing active disease were excluded. This emphasizes to us that severity of disease is probably a more important correlate of positive ANF tests than duration. It is generally agreed that rheumatoid nodules are more commonly seen with more severe disease. Thus, the higher incidence of rheumatoid nodules among our patients with positive ANF tests also emphasizes the relationship 1090 Fig. 5.-The patient (L. R.) . CONDEMI ET AL. perivascular inflammatory response found on muscle biopsy in between severity of disease and the positive tests. Wilkens and Deckerz3 also noted that patients with positive ANF tests had severe joint deformities and rheumatoid nodules. We did not, however, find a positive correlation between presence of A N F and presence of other evidences of vasculitis or infection as they23 and Ward et aLZ4have reported. Our experience is like that of Halls Alexander,lo and V a l k e n b ~ r gin~ ~this regard. In Ward's study the ANF test employed was somewhat less sensitive than ours, and so their positive sera may have come from patients with somewhat more A N F than ours. In the study by Wilkens and Decker, the antinuclear tests included indirect L.E. cell preparations, nucleoprotein-coated latex particle agglutination, and DNA- and nucleoprotein-coated red blood cell agglutination in addition to the direct L.E. cell test and the immunofluorescence method for the detection of ANF. Wilkens and Decker obtained positive tests by immunofluorescence in only 7 of 14 patients. If this figure alone is considered, then positive tests also occurred independently of vasculitis in their study. In our laboratory we have been unable to find instances in which DNA- or 1091 SIGNIFICANCE OF ANTINUCLEAR FACTORS Table 7.-ANF Titers in Systemic Lupus Erythematosus and R.A. ANF TITER II m-POSITIVE SYSTEMIC LUPUS ERYTHEMATOSUS LE PREP. RHEUMATOID ARTHRITIS I I024 512 m 256 ] X p J i 128 axx 64 TIJX xx 32 xxxx I lex 4 xx 2 axx N DXXXX xx 0 ~ *o;ONE + LE PATIENT NOT INGLUDED HAD I9 ALTERED CELLS. PREP AND POSITIVE ANF DETERMINATION ONLY WITH nucleoprotein-coated latex particles gave positive tests in the absence of a positive test by immunofluorescence. This has also been the experience of others.26-28In our hands, hemagglutination tests for antinuclear factors have been uniformly unsuccessful, even with sera of high ANF titer. We recognize that comparison of ANF data derived from various laboratories is diffxult because of differences in patient selection, sensitivity of detection systems, and sources of antigens and antibodies used as reagentsz9 Combinations of these differences may expIain the variations in r e s ~ l t s ~ J ~ J ~ - ~ ~ appearing in the current literature. The finding of patients with clinical RA and positive L,.E. cell tests have in recent years led to difficulties in classification. These patients have been classified as SLE, as RA with positive L.E. cell tests, or as an intermediate syndrome. Among the 24 patients with positive ANF tests we noted 7 with positive L.E. cell tests. Six of these had classic RA with subcutaneous nodules, high latex titers, and chronic symmetrical joint disease in Stage I1 or IV. The seventh was classifiable as definite RA by the ARA criteria. This group of 7 did not have a greater incidence of vasculitis or signs and symptoms suggestive of SLE than did the ANF-positive group. We have stated previously that the numerous indices that might distinguish SLE from RA were no different in the ANF-positive and the ANF-negative groups. Therefore, we are unable to justify the use of the presence of a positive ANF test or of a positive L.E. cell test as tools of diagnostic significance in separating 1092 CONDEMI ET AL. Table 8.-Relationship Between ANF and Latex Titers in R.A. @ * PITIENT WITH SUBCUTANEOUS NODULES PRESENT ANF TtTER POSITIVE LE CELL TEST AND POSITIVE ANF TEST ONLY WITH ALTERED CELLS. Is RA patients. Early in the disease, however, high titers of A N F and more specifically high titers of rG A N F may be of value in distinguishing SLE from RA.20330These data indicate that the presence of A N F in patients with RA at present should be considered simply a manifestation of an immunological response that either SLE or RA patients may exhibit in the course of their d'isease. REFERENCES 1. Coons, A., and Kaplan, M.: Localization of antigen in tissue cells. II. Improvements in a method for the detection of antigen by means of fluorescent antibody. J. Exp. Med. 9:1, 1950. 2. Friou, G., Finch, S., and Detre, K.: Nuclear localization of a factor from disseminated lupus serum, Fed. Proc. 16:413, 1957. 3. Holbrow, E. J., Weir, D. M., and Johnson, G. D.: A serum factor in lupus erythematosus with affinity for tissue nuclei. Brit. Med. J. 2:732, 1957. 4. Holman, H. R., and Kunkel, H. G.: Affinity between the lupus erythemato- sus serum factor and cell nuclei and nucleoprotein. Science 126:162, 1957. 5. Friou, G . L.: The significance of the lupus globulin nucleoprotein reacaction. Ann. Intern. Med. 49:866. 6. Wier, D. M., and Holbrow, E. J.: A clinical study of serum antinuclear factor. Brit. Med. J. 1:953, 1961, 7. Rothfield, N., Phythyon, J. M., McEwen, C., and Miescher, P.: The Role of Antinuclear Reactions in the Diagnosis of SLE: A Study of 53 Cases. Arthritis Rheum. 4:223, 1961. 8. Hall, A. P., Bardawil, W. A., Bayles, T. B., Mednes, A. D., and Gallins, N.: The relations between the anti- SIGNIFICANCE OF ANTINUCLEAR FACTORS nuclear rheumatoid and LE cell factors in the systemic rheumatic diseases. New Eng. J. Med. 263:769, 1960. 9. WitebsLy, E.: Paper read at the 19th Annual Meeting of the American Academy of Allergy, Montreal, Canada, 1962. 10. Alexander, W. R. M., Bremmer, J. M., and Duthie, J. J. R.: Incidence of the antinuclear factor in human sera. Ann. Rheum. Dis. 19:338, 1960. 11. Calabresi, P., Edwards, E. A., and Schilling, R. F.: Fluorescent antiglobulin Studies in Leucopenic and Related Disorders. J. Clin. Invest. 38: 2091, 1959. 12. Hollander, J. L., et al.: Arthritis. Philadelphia, Lea and Febiger, 1960. 13. Kabat, E. A., and Mayer, M. M.: Experimental Immunochemistry. Springfield, Ill., Charles C Thomas, 1961. 14. Nomenclature for Human Immunoglobulin. Bull. WHO. 30:447, 1964. 15. Marshall, J. D., Eveland, W. C., and Smith, C. W.: Superiority of fluorescein isothiocyanate for fluorescent antibody technique with a modification of its application. Proc. Soc. Exp. Biol. Med. 98:898, 1958. 16. Coons, A. H., Leduc, E. H., and Connolly, J. M.: Stuclies on antibody production. I. A method for the histochemical demonstration of specific antibody and its application to a study of the hyperimmune rabbit. J. Exp. Med. 102:49, 1955. 17. Atwater, E. C., and Jacox, R. F.: The latex fixation test in patients with liver disease. Ann. Intern. Med. 58: 419, 1963. 18. Zinkham, W. H., and Conley, C. L.: Some factors influencing the formation of LE cells. Bull. Johns Hopkins Hosp. 98:102, 1956. 19. Svec, K. H., and Kaplan, M. H.: A varient LE cell factor reactive only with altered nuclear material. Arthritis Rheum. 6:11, 1963. 20. Barnett, E. V., Condemi, J. J., Leddy, J. P., and Vaughan, J. H.: Gamma,, gammalA, gammalM antinuclear factors in human sera. J. Clin. Invest. 43:1104, 1964. 1093 21. Sokoloff, L., McCluskey, R. T., and Bunim, J. J.: Vascularity of the early subcutaneous nodule of rheumatoid arthritis. Arch. Path. 55:475, 1953. 22. Scott, J. T., Hourchane, D. O., Doyle, F. H., Steiner, R. E., Laws, J. W., Dixon, A. ST. J., and Bywaters, E. G. L.: Digital arthritis in rheumatic disease. Ann. Rheum. Dis. 20:224, 1961. 23. Wilkens, R. F., and Decker, J. L.: Rheumatoid arthritis with serological evidence suggesting systemic lupus erythematosus, clinical, serological, and chromatographic studies. Arthritis Rheum. 6:720, 1963. 24. Ward, D. J., Johnson, G. D., and Holbrow, E. J.: Antinuclear factor in rheumatoid arthritis. Ann. Rheum. Dis. 23:306, 1964. 25. Valkenburg, H. A.: Conference on the immunologic aspects of rheumatoid arthritis and systemic lupus erythematosus. Conference Series, No. 10, p. 567, 1963. 26. Dubois, E. L., Drexler, E., and Arteberberry, J. D.: A latex nucleoprotein test for diagnosis of systemic lupus erythematosus. J. A. M. A. 177:141, 1961. 27. Holman, H. R., Deicher, M. D., and Robbins, W. C.: Antinuclear antibodies in lupus erythematosus. Henry Ford Hospital, Symposium Mechanisms of Hypersensitivity, p. 361, 1959. 25. Seligmann, M.: DNA antibodies. Coiiference on Immunological Aspects of Rheumatoid Arthritis and Systemic Lupus Erythematosus, Conference Series No. 10, Chapter XI, p. 542. 29. Hijmans, W., Scherit, H. R. E., Mandema, E., Nicnhuis, T. E. W., Holbrow, E. J., and Johnson, G. D.: Comparative study for the detection of antinuclear factors with the fluorescent antibody technique. Ann. Rheum. Dis. 23:73, 1964. 30. Barnett, E. V., North, A. F., Condemi, J. J., Jacox, R., and Vaughan, J. H.: Antinuclear factors in systemic lupus erythematosus and rheumatoid arthritis. Submitted to Ann. Intern. Med.