COMPLEMENT-FIXING HIDDEN RHEUMATOID FACTOR IN CHILDREN WITH BENIGN RHEUMATOID

Transcription

1 93 COMPLEMENT-FIXING HIDDEN RHEUMATOID FACTOR IN CHILDREN WITH BENIGN RHEUMATOID * NODULES TERRY L. MOORE, ROBERT W. DORNER. and JACK ZUCKNER Rheumatoid nodules have been described in many children without evidence of clinical disease. These have been referred to as benign rheumatoid nodules (BRN). Except for one report, no serological abnormalities have been demonstrated. Four children with BRN have been studied. By means of a hemolytic assay, high titers of hidden rheumatoid factor (RF) have been found in the blood, i.e., 19s IgM RF detected after acid separation of IgM-containing fraction from the serum. The median hidden RF titer of the children with BRN was 1:362 and in healthy and disease controls was I :7. The difference was significant at P <.1. The results indicate that patients with BRN have active complement-fixing 19s IgM RF in their serum. Thus, the association of hidden RF with BRN raises the possibility that the deposition of immune complexes ( 19s IgM-7S IgG) leads to small vessel vasculitis and nodular formation. Subcutaneous nodules that histologically re- From the Section of Rheumatology, Department of Internal Medicine, St. Louis University School of Medicine, 142 S. Grand Avenue, St. Louis, Missouri Supported in part by USPHS grant No. 5 S7RR5388. Terry L. Moore, M.D.: Assistant Professor of Internal Medicine; Robert W. Dorner. Ph.D.: Associate Professor of Internal Medicine and Biochemistry; Jack Zuckner, M.D.: Clinical Professor of Internal Medicine, Director, Section of Rheumatology. Address reprint requests to Terry L. Moore. M.D., St. Louis University School of Medicine, Section of Rheumatology, 142 S. Grand Avenue, St. Louis, Missouri Submitted for publication April 12, 1978; accepted June 7, semble rheumatoid nodules have been described in many children (1-19) in whom there was no evidence of juvenile rheumatoid arthritis (JRA) or other connective tissue disease. These have been referred to as benign rheumatoid nodules (BRN). The finding of these nodules in otherwise asymptomatic children has evoked concern about the future development of JRA; in one followup study, no children subsequently developed any type-of systemic rheumatic disease (15). The nodules have usually been located subcutaneously, predominantly in the scalp and pretibial regions. They have varied in size, have been observed to regress and recur spontaneously, and have usually been painless. Sera of patients with BRN have been reported not to contain rheumatoid factors (RF) or antinuclear antibody (ANA) (1-19) except in one study (2). Other laboratory tests, such as erythrocyte sedimentation rate (ESR), complete blood count (CBC), and blood chemistries have not shown abnormalities (1-19). The possibility that hidden RF might be present in these patients was entertained because of recent studies in our laboratory that revealed the presence of hidden RF in high titer when a hemolytic assay was performed on separated IgM fractions of patients with seronegative JRA (21). Hidden RF is 19s IgM RF that probably has been complexed in the peripheral blood with 7s IgG and therefore is not detected by standard laboratory tests. In the present study this hemolytic assay has been extended to patients with BRN. By using this procedure, Arthritis and Rheumatism, Vol. 21, No. 8 (November-December 1978)

2 BENIGN RHEUMATOID NODULES 93 1 complement-fixing RF was demonstrated in the separated 19s IgM fractions from sera of 4 children with BRN. Case 1 CASE REPORTS A 6.5-year-old white boy (M.S.) was first admitted at the age of I.5 years for evaluation of a nodule over his occiput. Subsequently, other nodules developed over the occiput and, also, over the anterior tibial area bilaterally. The history was remarkable in that the patient also has had multiple occurrences of swelling of lymph nodes in his neck, axilla, and inguinal area. Biopsies of these nodes showed reactive lymphoid hyperplasia. His physical examination revealed nontender, nonmovable nodules over the occiput and anterior tibial area bilaterally. In addition, multiple soft benign nodes that were movable and non-tender were noted in the epitrochlear, axillary, anterior triangle of the neck, and inguinal areas. The rest of the physical exam was unremarkable. Biopsy of the anterior tibial lesions on two occasions and the occipital lesion showed foci of fibrinoid necrosis adjacent to which were histiocytes and chronic inflammatory cells. The findings of each biopsy were compatible with those of rheumatoid nodules. Case 2 A 4-year-old white boy (M.G.) was admitted for evaluation of multiple nodules on the skull. Nine months prior to admission the mother noticed a nodule on the left side of his scalp. In the ensuing months the nodule enlarged and others developed. His history was completely normal. His physical examination was unremarkable except for the presence of 6 painless subcutaneous nodules, cm in diameter, on both sides of the skull. Biopsy of one of the nodules showed dense connective tissue and nodular areas of degenerative connective tissue surrounded by palisading histiocytes. These findings were compatible with those of a rheumatoid nodule. Case 3 A 5-year-old white girl (C.K.) was first admitted at age 2 for multiple nodules on the right posterior aspect of the skull. Over the next 3 years, nodules developed in other regions of the skull, in the occipital area, on the forehead, over the anterior tibial area bilaterally, and on both ankles. The history was normal except for frequent upper respiratory infections up to the age of 3. Her physical examination was unremarkable except for multiple non-tender subcutaneous nodules in the above mentioned areas. Biopsy of one of the nodules showed fibrous tissue containing stellate shaped cells with a central core of necrotic collagen surrounded by radially oriented epithelioid cells. The tissue findings were compatible with those of a rheumatoid. nodule. Case 4 A 3-year-old white boy (L.P.) was admitted for evaluation of subcutaneous nodules on the scalp, back, elbows, knees, and anterior tibial area. The nodules were nontender. The rest of his history was unremarkable. Results of his physical examination were normal except for firm, non-tender nodules cm in diameter over all the described areas. Biopsy of one of the nodules showed a central area of fibrinoid necrosis surrounding proliferating fibroblasts and a palisading mononuclear layer, findings compatible with a rheumatoid nodule. Examination of the biopsy by immunofluorescence showed deposits of IgM, IgG, IgA, and C3 in the small vessels. MATERIALS AND METHODS Four children with BRN were studied, along with controls consisting of 12 healthy children and 1 children with various connective tissue diseases as previously described (2 1 ).,These children were attending the Pediatric Arthritis Clinic of Cardinal Glennon Memorial Hospital for Children at St. Louis University or were private outpatients. Nodules were biopsied and the histology appeared compatible with a rheumatoid subcutaneous nodule in each case. The children were evaluated by a variety of laboratory tests including CBC, urine analysis, latex fixation (LFT) and sheep cell agglutination tests (SCAT) for RF, ESR, ANA, and when available, complement levels (C3 and C4). Blood specimens for hidden RF were drawn in clot tubes. Sera were obtained by centrifugation and stored at - 2" C until tested. Sera were subjected to gel filtration on a Sephadex G-2 column (2.5 X 85 cm) essentially by the method of Allen and Kunkel (22), as previously described in detail (23). Sera from patients with JRA and controls were handled in the same manner. The hemolytic assay was performed on whole serum and on the IgM-containing fraction after heat inactivation at 56" C for 3 minutes and absorption with a 1:s dilution of washed packed sheep erythrocytes (SRBC) at 37" C for 9 minutes to remove all natural antibody to SRBC. The target cell-srbc were sensitized with a 1 : 1 dilution of rabbit anti- SRBC IgG hemolysin (Colorado Serum, Denver, Colorado), as previously described (24). Briefly, the IgG fraction of hemolysin was obtained by DEAE column chromatography and checked for purity by immunodiffusion against antiserum to whole rabbit serum and to rabbit Cohn Fraction 11. The IgG hemolysin was reduced and alkylated to abolish its ability to fix complement as previously described (24.25) and subjected to acid treatment with.5 M glycine-hci, ph 2.8, at room temperature for 9 minutes to enhance hybridization, as previously described (26). This preparation was then used to sensitize the SRBC. One-tenth milliliter of serum or IgM fraction was diluted serially in.15 M veronal-buffered saline containing.1% gelatin and optimal concentrations of Cat+ and Mg++ (GVB). To each dilution,.1 ml of a suspension of I% sensitized SRBC was added. Following incubation at 37" C for 6 minutes and overnight at 4" C, the cells were washed with I ml of cold GVB and the supernatant discarded. Next, 1 ml of GVB and.4 ml of 1:4 guinea pig serum previously absorbed

3 932 MOORE ET AL Table 1. Laboratory and Clinical Data of Patients with Benign Rheumatoid Nodules RF Titers Duration IgM Fraction Serum CBC of Complement Age Nodules Hemolytic Hemolytic Hemo- WBC Patient (yrs) Sex (yrs) Assay LFT SCAT Assay LFT SCAT globin Count ESR ANA C3 C4 M.S. 6.5 M 5. I: I24 NR* NR I:2 NR NR , M.G. 4 M.75 1: NR 1.2 NR NR ,2 I2 Neg C.K. 5 F 3. 1:64 1:16 NR 1.4 NR NR ,4 17 Neg - - L.P. 3 M.25 1:32 1:16 NR l:2 NR NR Neg * NR = nonreactive. with SRBC as a source of complement were added to the target cells. Following incubation for 6 minutes at 37" C with vortexing every 12 minutes, the reaction was stopped with 1.5 ml cold phosphate buffered saline. The tubes were centrifuged for 5 minutes at 2 RPM at 4" C. Hemoglobin released into the supernatant was assayed spectrophotometrically at 41 2 nm. Zero and 1% controls were included in all assays. Titers > 1 : 16 were considered positive. The LFT was performed on the whole serum and on the IgM-containing fraction after heat inactivation for 3 minutes at 56" C, essentially according to Singer and Plotz (27). with a commercial test kit (Difco Laboratories, Detroit, Michigan). Titers > I : 2 were considered positive. The SCAT was also performed on the whole serum and IgM-containing fraction. All sera and fractions were heat inactivated and absorbed as in the hemolytic assay. The SCAT was performed essentially according to Ball (28) using settling patterns in microtiter plates (Cooke Engineering Co., Alexandria. Virginia) to determine the end point. Titers > 1 : 16 were considered positive. ESR was determined by the Westergren method. ANA were determined by an indirect immunofluorescent staining method using mouse kidney and fluorescein-labeled rabbit anti-human IgG or IgM antibody. Complement levels (C3 and C4) were measured by commercial imrnunodiffusion kits (Meloy Laboratories, Springfield, Virginia). RESULTS Laboratory and clinical data on the patients are listed in Table 1. Marked hemolytic RF activity was noted in the IgM-containing fractions (1 : 124, 1 : 124, 1 : 64, 1 : 32). There was also some activity noted by LFT in 3 patients (1 : 16, 1 : 16, 1 :4), but none by SCAT. The median hemolytic titer of the IgM-fractions of the patients with BRN was 1 : 362: the median titer of the disease and healthy controls was 1 : 7. The difference was significant at P <.1 according to the Mann-Whitney analysis (2). Only 1 of 1 disease controls showed hemolytic activity in the IgM-fraction (1 : 32). and no activity was demonstrated in the 12 normal controls (Figure I). In the serum, LFT and SCAT for RF were negative and the hemolytic assay for RF activity showed 1:24 1:512 1: :128 - I- 3 A 1:64 a z - 1:32 l- a E 1:16 I s - m 1:8 1 :4 1:2 vv Neg. Benign Normal Disease Rheumatoid Children Controls Nodules Figure 1. Scattergraph of hemolytic titers of IgM-containing Jractions front children with benign rheumatoid nodules. normal children. and disease controls. Horizontal dotted line represents 2 SD from median liter oj'normal controls.

4 BENIGN RHEUMATOID NODULES 933 values of I I :4. CBC, ESR, ANA, C3, and C4 values revealed no abnormalities. DISCUSSION The cause of subcutaneous nodules in children without associated disease is not understood. Defining pathogenic mechanisms has not been advanced by the failure to find serologic abnormalities in all previous reports (1-1 9), with one exception (2). In the present study of 4 additional children with BRN, avid complement-fixing IgM RF has been found in the separated 19s IgM fractions of the patients sera. Hemolytic titers of the IgM fractions were comparable to those observed in patients with JRA and were significantly greater than those of normal children and other children with different connective tissue diseases who acted as controls (P <.1) (21). The association of hidden RF and BRN has raised the possibility that the production of nodules was secondary to the deposition of immune complexes (19s IgM-7S IgG) leading to small vessel vasculitis and subsequent nodule formation. Hidden RF might indeed have been complexed in the serum as a rather stable macromolecular complex of 19s IgM and 7s IgG since no IgM RF was demonstrated in the patients sera without acid gel filtration. Such complexes, if deposited in small vessels, could then fix complement causing an inflammatory process and an arteritis. As a consequence, proliferation of capillaries and fibroblasts, a chronic inflammatory cell infiltrate, and secondary necrosis could occur leading to the central fibrinoid necrotic zone of the mature nodule, thus resulting in findings similar to those of the subcutaneous nodule associated with rheumatoid arthritis (3). It has been demonstrated that deposits of RF and complement have been found in large amounts in nodules from rheumatoid patients (31). Also, IgG and IgM deposits have been demonstrated in the central zone of the nodule as well as in some small vessel walls (32). This was noted in the biopsy of patient L.P. (Case 4) which suggests the possibility that stable immunoglobulin complexes of hidden RF and IgG could have precipitated out in small vessels and initiated the arteries responsible for the formation of the rheumatoid nodule. Why such hidden RF-IgG complexes have not been deposited in all small vascular beds to initiate further disease activity and possibly even rheumatoid arthritis is not clear. It is interesting to speculate on prognosis in these children. The presence of hidden RF is not well under- stood at this stage of our knowledge. It may have a potential role in prognosticating disease or in initiating further pathology. However, a negative followup of previously reported patients with BRN would seem to contradict this possibility (19, but it is not known that the latter patients had hidden RF. The 4 patients in this report will be followed periodically to determine whether hidden RF titers correlate with regression and recurrence of the nodules. This will help to determine if the production of hidden RF is constant or varies. It may also provide information on whether hidden RF is resulting from a chronic stimulus such as in subacute bacterial endocarditis and is self-limited or, perhaps, an ongoing production. IgM levels in these patients have not been elevated. However, there is the possibility that in the acid dissociable IgM fraction, not only is hidden RF present, but other antibody titers are also elevated. However, IgM cold agglutinins have shown no elevation. The authors are also presently attempting to characterize the hidden RF in the patients with BRN and compare its specificity and avidity to the hidden RF in JRA and the R F of adult rheumatoid arthritis. This work may show that the RF has different characteristics and may help in determining why the children do not go on to develop arthritis. ACKNOWLEDGMENTS We wish to thank Drs. Jesse Ternberg and Dennis O Connor for contributing patients to the study REFERENCES Altman RS, Caffrey PR: Isolated subcutaneous rheumatic nodules. Pediatrics 34: , 1964 Beatty EC, Jr: Rheumatic-like nodules occurring in nonrheumatic children. Arch Pathol 68: , 1959 Burrington JD: Pseudorheumatoid nodules in children: Report of ten cases. Pediatrics 45: , 197 Calabro JJ: Discussion of arthritis rounds: Rheumatoidlike nodules presenting as pump bumps in a patient without rheumatoid arthritis. Arthritis Rheum 13: , 197 Calabro JJ: Editorial comment. J Pediatr 8 l: I972 Caughey DE, Calabro JJ, Cracchiolo A, Goldberg LB, Pearson CM: Benign rheumatoid (pseudorheumatoid) nodules in children. Arthritis Rheum 12:285, 1969 Depowski M, Miezyneski W Pseudorheumatic changes in the subcutaneous tissue of children. Patol Pol 18: , I967 Draheim JH. Johnson LC, Helwig EB: A clinico-patho-

5 934 MOORE ET AL logic analysis of rheumatoid nodules occurring in 54 children. Am J Pathol 35678, Kosmin M, Vail JT, Sturman MJ, Luke DR: Streptococcal infection and isolated rheumatic-like subcutaneous nodules. Med Bull US Army (Europe) 21:28-31, Mesara BW, Brody GL, Oberman HA: Pseudorheumatoid subcutaneous nodules. Am J Clin Pathol 45: , Miller EH: Isolated, subcutaneous, rheumatic-like nodules. JAMA 215: , Pournaras J, Gibson AAM: Pseudorheumatoid nodules in children. J Bone Joint Surg 53: , Sokoloff L: Discussion of arthritis rounds: Rheumatoidlike nodules presenting as pump bumps in a patient without rheumatoid arthritis. Arthritis Rheum 13: , Schaller JG: Benign rheumatoid nodules. Arthritis Rheum 2:277, Simons FER, Schaller JG: Benign rheumatoid nodules. Pediatrics 56:29-33, Sturgill BC, Allan JH: Arthritis rounds: Rheumatoid-like nodules presenting as pump bumps in a patient without rheumatoid arthritis. Arthritis Rheum 13: , Taranta A: Occurrence of rheumatic-like subcutaneous.nodules without evidence of joint or heart disease. N Engl J Med 266:13-16, Ziegler E Rheumatismus nodosus als einzige Manifestation der rheumatischen Krankheit. Arch Kind 122:l-6, Zuckner J, Baldassare A: The nonspecific subcutaneous nodule: Its presence in fibrositis and scleroderma. Am J Med Sci 27 1 :69-75, Berardinelli JL, Hyman CJ, Campbell EE, Fireman P: Presence of rheumatoid factor in ten children with isolated rheumatoid-like nodules. J Pediatr 81: , Moore TL, Dorner RW, Zucker J: Complement-fixing hidden rheumatoid factor in juvenile rheumatoid arthritis. Arthritis Rheum 21: 22. Allen JC, Kunkel HG: Hidden rheumatoid factors with specificity for native gamma globulins. Arthritis Rheum 9: , Moore TL, Dorner RW, Zuckner J: Hidden rheumatoid factor in seronegative rheumatoid arthritis. Ann Rheum Dis 33: , Tanimoto K, Cooper NR, Johnson JS, Vaughan JH: Complement fixation by rheumatoid factor. J Clin Invest 55: , Robbins DL, Moore TL, Carson DA, Vaughan JH: Relative reactivities of rheumatoid factors in serum and cells: evidence for selection deficiency in serum rheumatoid factor. Arthritis Rheum , Moore TL, Robbins DL, Rose JE, Vaughan JH: Drugs affecting the release of rheumatoid factor in a plaque forming cell assay. Arthritis Rheum 21: , Singer JM, Plotz CM: The latex fixation test. I. Application to the serologic diagnosis of rheumatoid arthritis. Am J Med 21: , Ball J: Serum factor in rheumatoid arthritis agglutinating sensitized sheep red cells. Lancet , Sokal RR, Rohlf FJ: Introduction to Biostatistics. First edition. San Francisco, W. Freeman, 1973, pp Moore CP, Wilkens RF: The subcutaneous nodule: Its significance in the diagnosis of rheumatic disease. Sem Arthritis Rheum 7:63-79, Munthe E: Immunofluorescent studies on rheumatoid tissue. Symposium on Immunology, Stockholm, March Acta Pathol Microbiol Scand 77: , Nowoslawski A, Brzosko WJ: Immunopathology of rheumatoid arthritis. 11. The rheumatoid nodule (the rheumatoid granuloma). Pathol Eur 2:32-321, 1967