1472 DECREASE OF THE OKT8 POSITIVE T C E L L SUBSET IN POLYMYALGIA RHEUMATICA Lack of Correlation with Disease Activity CHARIF BENLAHRACHE, PATRICK SEGOND, LOUIS AUQUIER, and JEAN-PIERRE BOUVET Peripheral T cell populations were investigated in 35 patients suffering from polymyalgia rheumatica. The total number of T cells was low compared with those of a control group of similar age (P < 10 -?. This decrease was demonstrated by using both classic E-rosette and monoclonal antibody techniques (OKT3, and OKT4 OKTS) and was shown to be secondary to a selective TS defect (P < There was no correlation between the decrease in T8 (a cytotoxic suppressor T cell subset) and steroid therapy, disease activity, and temporal arteritis, nor between this decrease and the T y percentage and the presence of circulating immune complexes (CIC). The T y cell percentage was low in the patient group (P < lo-’) and correlated with the presence of detectable CIC (P < 0.05). In contrast to the TS and T y defects, concanavalip A-stimulated cells from 5 selected patients were found capable of suppressing in vitro anti-trinitrophenyl response. This suppression was found in both autologous and allogeneic experiments. From these data one can assume that an immune anomaly (T8 defect) could be the origin of CIC and the disease occurrence. + From the Department of Rheumatology. Hbpital Ambroise Pare, Faculte de MCdecine Paris-Ouest. the Department of Haematology, Facultt de Medecine Paris-Sud. and the Laboratoire d’lmmunologie Microbienne, Institut Pasteur, Paris, France. Charif Benlahrache, MD: recipient of a fellowship award from the College de Medecine des Hbpitaux de Paris: Patrick Segond, MD: Assistant Professor in Immunology, Facultk de Medecine Paris-Sud; Louis Auquier, MD: professor of Rheumatology. Head of the Department of Rheumatology. Faculte de MCdecine Paris-Ouest; Jean-Pierre Bouvet, MD. PhD: Charge de Recherches at the Institut National de la Sante et de la Recherche Medicale (INSERM). Address reprint requests to Dr. Jean-Pierre Bouvet. Laboratoire d’Immunologie Microbienne, Institut Pasteur. rue du Docteur Roux, 25, 75724 PARIS Cedex, France. Submitted for publication February 8. 1983; accepted in revised form July 19, 1983. Arthritis and Rheumatism, Vol. 26, No. 12 (December 1983) Polymyalgia rheumatica (PMR) is a relatively common disease in aged patients (1). Although the clinical symptoms and the response to steroid therapy are well documented (2), the etiology and pathogenesis of the disease remain unknown. However, a frequent association with giant cell ajteritis (GCA) (1) points to a vascular origin of the disease. Immunoglobulin deposits are found in the media of affected arteries, suggesting an anti-vascular autoimmunity (3) or an immune complex mediated vasculitis. Hazleman and coworkers reported enhanced in vitro proliferative responses of peripheral blood lymphocytes to homologous artery walls (4). Unfortunately these results were not confirmed by Papaioannou and coworkers ( 5 ) and recent studies have investigated mainly the immune complex hypothesis. Circulating immune complexes (CIC) are often found in patients; the percentages depend partly on the methods of measurement (6-8). The highest percentage is 94, which is found when the Raji cell radioimmunoassay is used (8). However, these data provide only indirect evidence for an immune vasculitis since similar deposits were found in other diseases lacking signs of vasculitis (3), and CIC have also been detected in a wide variety of nonimmune diseases as well as in normal subjects (9,lO). Additional data are therefore required to confirm the immunologic origin of the disease. The recent development of monoclonal antibodies which react with specific antigenic determinants on human T lymphocytes provides a means to readily assess the immune status of individuals (1 1). These antibodies have been widely used in the study of T cell imbalances in autoimmune diseases (12). A decrease in OKT8 positive cells, a cytotoxic/suppressor T cell subset, was found in peripheral blood in DECREASE OF OKT8 IN PMR 1473 Table 1. Clinical findings and T lymphocyte subsets in patients* Patient I 2 3 4 5 6 7 8 9 10 I1 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 Duration from disease onset (years) Steroid therapy (mdday) 3 3 3 5 6 2 0.3 9 4 0.16 2 2 I 2 2 5 9 3 0.16 0.16 0.16 I .4 5 6 0.5 0.3 0.16 0.3 3 3 1.5 0.6 0.1 0.5 0.2 0 0 25 - 19 t 10 17 II 17 15 45 19 16 32 9 33 16 40 in 7 6 I5 130 54 52 40 16 2n 10 20 I19 21 20 94 37 40 60 + + + + + + + + + + + + + + + 112 + 0 0 6 20 0 0 n 15 15 6 7 6 0 0 15 10 10 15 n 0 0 10 10 15 15 8 6 17 15 30 20 0 ESRt no CICf Temporal vasculitis - + + - + + + - + + 96 % E-RFCt T3 65 71 72 61 36 59 72 54 68 55 66 45 75 4n 75 65 40 63 48 72 58 47 82 7n 59 60 53 59 65 60 47 65 78 50 59 71 64 74 65 40 56 7n 51 75 52 68 42 76 46 71 68 4n 60 58 69 56 50 79 75 54 63 55 5n 76 63 49 * - = absence of; + = presence of; ND = not determined. t ESR = erythrocyte sedimentation rate; ClC = circulating immune complexes; E-RFC rheumatoid arthritis (13,14), multiple sclerosis (15,16), Sjogren’s syndrome (17), and other diseases (12). In the majority of cases the decreases seemed t o be correlated with the disease activity. I n the present study we have investigated lymphocyte subsets in polymyalgia rheumatica patients for t h e presence of similar abnormalities. PATIENTS AND METHODS Patients. Thirty-five patients (7 men and 28 women) with polymyalgia rheumatica were selected for the study. The diagnosis was based on the ptesence of 3 or more criteria proposed by Bird and coworkers (2). The mean age was 72.6 years 2 7.8 SD. The presence of a temporal arteritis was found in 7 patients. The patients were not investigated at the same stage of the disease, thus allowing for the study of the influence of steroid therapy and disease 61 71 53 54 % T4 +8 75 65 68 sn 42 52 67 56 72 52 65 3n 73 41 69 63 46 59 45 66 52 46 80 73 51 58 52 50 71 55 54 63 69 5n 52 % % % T4 TS TY 50 48 53 49 36 49 55 46 61 48 51 34 50 32 56 53 42 22 23 28 26 I0 13 10 7 53 50 55 46 44 53 56 49 61 48 50 61 49 45 54 52 46 43 23 12 n 10 n 6 6 5 4 2 I1 3 25 2 2 3 2 11 35 n 25 16 9 16 11 24 12 11 24 31 16 n 14 13 14 16 7 16 31 in 15 10 6 9 7 12 11 13 13 ND ND ND ND ND ND ND ND ND ND ND ND ND = E-rosette forming cells activity (Table 1). The median duration from the onset of symptoms was 2 years (mean: 2.41 2 2.4). Twenty-four patients were undergoing steroid therapy (mean dosage: 8.48 ? 7.42 mg/day). Disease activity was estimated by the erythrocyte sedimentation rate (ESR) and by serum levels of haptoglobin and orosomucoid: mean ESR was 36.9 2 32.3 mm. haptoglobin was 2.41 % 1.12 mg/ml (normal value: 0.971, and orosomucoid was 1.08 2 0.37 mg/ml (normal value: 0.52). The results of these 3 tests for biologic inflammation were found to vary according to a linear positive correlation: ESR-haptoglobin (P < 2 x 10-7, ESR-orosomucoid (P < orosomucoid-haptoglobin (P< 10-61. Control subjects. Thirty-five subjects (7 men and 28 women) of similar age (77.2 % 8.9 years) without any infectious, inflammatory, or neoplastic disease, were selected as controls. This control group included patients hospitalized for hemiplegia or osteoporosis. All controls had normal values for ESR, haptoglobin (0.59 to 1.44 mg/ml), and orosomucoid (0.40 to 0.68 mg/ml). BENLAHRACHE ET AL 1474 Assay for immune complexes. The procedure for the detection o f circulating immune complexes was carried out on fresh sera, hsing a simple qualitative technique: the PEGC l q immunodiffusion test (18). CIC were precipitated with 5% polyethylene glycol (PEG), washed with 2.5% PEG, and after dissolution of the pellet with saline, investigated for the presence of IgG and complement by immunodiffusion ass.ay. Positive sera showed precipitating lines in the presence of anti-Fcy and anti-CIq sera .(Betiring). Mononuclear cell isolation. Purified mononuclear cell (MNC) suspensions were obtained from hepatinized blood by standard Ficoll-Hypaque density cedtrifugation and kept frozen in liquid nitrogen after being treated with dimethylsulfoxide.. E-rosettes. Rosetting of mononuclear cells was performed using sheep red blood cells treated with neuraminidase. and the percentage of rosette forming cells (E-RFC) was established from 200 cells. E-RFC were purified by centrifugation on Ficoll-Hypaque followed by lysis of erythrocytes with Tris-ammonium chloride (19). Purified E-RFC were used for T y cell percentage estimation as described in the following section. Ty cells. T cells that bear Fc receptor specific for IgG were identified with the erythrocyte-antibody rosette technique performed on E-RFC preparations. These cells (Ty) formed rosettes with ox erythrocytes sensitized with IgG from rabbits hyperimmunized with ox erythrocytes (19). The CYR C FUR C hyperimmune serum was a gift from Dr. N. Chiorazzi of the Rockefeller University (NY). B cells. Membrane immunoglobulin positive cells (B cells) were identified on MNC preparations after 1 hour incubation at 37°C. Membrane immunoglobulins were revealed by immunofluorescence examination using fluoresceinconjugated F(ab’Iz goat anti-human Ig (Kallestad). MOUOCIOMI antibodies. Membrane markers specific for total peripheral T cells (T3) or for T cell subsets (T4 and T8) were identified by indirect immunofluorescence studies with commercial monoclonal antibodies (Ortho Pharmaceutical Corporation). In the peripheral blood OKT3 identifies all T cells (T3), OKT4 is specific for T helper cells (T4). and OKT8 for suppressorkytotoxic cells (T8) (1 1). Stock solutions of antibodies were made in phosphate buffered saline (0.5 mg/ml) and diluted in appropriate volumes of RPMI medium containing 10% fetal calf serum (Gibco, Grand Island, NY) and 0.02% sodium azide: these dilutions were 1:30 (OKT3). 1 : 5 0 (OKT4), and 1:100 (OKT8), aliquoted and stored at -80°C. Fluorescence studies. Suspensions of MNC (5 x lo4 cells per well) were washed 3 times in Linbro plates with an RPMI-fetal calf serum solution. They were incubated for 30 minutes at 4°C with 2 p1 of diluted monoclonal antibody, and after 3 additional washes, positive cells were revealed by indirect immunofluorescence using 1 :20 diluted fluoresceinated goat F(ab’)z anti-mouse IgG antibody (Cappel Labora- C CYR ma F -p.f+ :f i- 4 .. .. : 8 ... I 1 t -& E- RFC .. L 144 14 1s Figure 1. Petcentages of T lymphocytes in polymyalgia rheumatica patients (PMR) and controls (C): significant decreases are apparent between the patients and the controls when measured by 4 different methods. Dots represent individual values, horizontal lines correspond to means (long tines) and standard deviations (short lines). E-RFC = cells forming rosettes with sheep erythrocytes. T3 = cells positive with OKT3 antibody. T4+8 = cells positive with OKT4 andor OKTE (simultaneous labeling experiments).T4+TH = sum of cells positive with OKT4 and OKTH (distinct labeling experiments). 1475 DECREASE OF OKT8 IN PMR tory, Cochranville, PA). Washed fluorescent cells were suspended in a buffered solution, set on slides, and examined with a Leitz Orthoplan microscope equipped for epifluorescence. Percentages were established on 200 cells. Assay of suppressor cell function. The capacity of freshly isolated lymphocytes to inhibit the in vitro antibody response was investigated in 5 patients and compared with that of 4 aged and 15 young controls. The patients were selected for their high T4:T8 ratios: 7.62 (patient 9), 3.43 (patient 27), 4.36 (patient 29), and 6.43 (patient 31), or for their high disease activity (patient 34). The suppressive function of M N C was measured by their ability to inhibit the in vitro response to trinitrophenyl (TNP) antigen after having been stimulated by concanavalin A (Con A) (20). Suppressor cell macrophages were eliminated by nylon adherence as previously described (21). Suppression was induced either by direct addition of Con A to the autologous TNP-polyacrylamide (TNP-PAA) stimulated culture or by transfer of patient cells, incubated with Con A. to a TNP-PAA stimulated 2-day culture from a normal young subject. Assays for anti-TNP response were performed at day 7 , using a conventional agarose plaque technique (20) and the percentage of suppression was calculated by comparing cells incubated with and without Con A. Statistical analysis. Statistical methods used were Student's t-test, linear correlation, and partial correlation tests. When data did not show a Gaussian distribution, nonparametric tests (chi-square and Mann-Whitney) were used. RESULTS Presence of circulating immune complexes. CIC were detected in 24 of 35 patients (68.5%) and in 4 of 32 controls (12.5%). The difference between the 2 groups was clearly significant ( P < lop4). No correlation was found between CIC and the age of patients, disease duration, presence of steroid therapy, presence of temporal arteritis, nor with results of a n y of the 3 tests for biologic inflammation (Table 1). E-RFC and Ty. The percentage of E-RFC was found to be clearly lower in patients (60.86 2 11.17%) than in the controls (68.83 5 6.39%) (Figure 1 ) . The difference was significant ( P < lop3). The percentages of Ty were established from purified E-RFC from 22 patients and 20 controls. Patients were found to have a lower percentage of Ty cells (median: 6%) than the controls (median: 15%); the difference was highly significant ( P < lo-')). Similar results were found when the percentages of T y were calculated from total MNC by multiplying individual Ty percentages with corresponding E-RFC percentages. E-RFC and Ty did not show any statistical correlation with clinical data or with biologic tests for inflammation. E-RFC were present in the % aa 8 PMR C PMR C .. PMR C 7 7c 6c ... 6 w 5 5c U 3 ( n ... 4 3 8 * .. W Figure 2. T cell subsets in polymyalgia rheumatica patients (PMR) and controls (C): decreases of T8 and Ty percentages and high T4R8 ratio in patients. The percentages of T4 (cells positive with OKT4) and T8 (cells positive with OKT8) were established from total mononuclear cell preparations. The percentages of Ty(T cell bearing Fcy receptor) were established from purified T cell preparations. Dots enclqsed in boxes correspond to 25% and 75% of individual values. The horizontal lines in the boxes correspond to the medians. BENLAHRACHE ET A L 1476 same percentages in patients with and without CIC. In contrast, among the patients investigated for Ty percentage, the 5 subjects without CIC had higher levels of Ty (median: 10.00%) than those with detected CIC (median: 6.17%) (P < 0.05) and this was also noticed when percentages were calculated on total MNC ( P < 0.05). B lymphocytes. Similar percentages of B cells were found in patients (5.77 2 1.69%) and controls (6.03 ? 2.31%). The difference was not significant. T cell percentage as defined by monoclonal antibodies. Using the OKT3 monoclonal antibody, we found a lower percentage of T cells (T3) in patients (61.57 4 10.60%) than in controls (66.89 +- 6.45%). The difference was significant ( P < 0.01) (Figure 1). Comparable results were found when the total T cell percentage was established by simultaneous labeling with OKT4 and OKT8 antibodies. In this case the positive cells (T4t-8) corresponded to the sum of T4-8+, and T4+8+ subsets. They were lower in patients (58.74 10.60%) than in controls (66.17 ? 6.03%) (P < 0.001) (Figure 1). The sums of individual * % T3 percentages of OKT4 (T4) and OKT8 (T8) positive cells (T4+T8) also showed lower levels in patients (66.63 2 11.8%) than in controls (74.03 t 6.67%) (P< 0.001). The increased levels of T4+T8, when compared with T3 and T4+8, were because of a double count of T4+8+ subset in the sum T4+T8. A close linear positive correlation was found among the 4 different methods for measuring T cell percentages (P < lop9). As found with E-RFC, T3, T4+8, and T4+T8 did not correlate with clinical data nor with CIC, biologic inflammation, or Ty. T cell subsets. Percentages of T4 were not found to be statistically different in patients (49.54 2 6.73%) and controls (47.17 ? 4.55%) (Figure 2). In contrast, percentages of T8 were found to be dramatically lower in patients (17.26 -+ 7.70%) when compared with controls (26.60 3.67%) ( P < lop9). The T4:T8 ratio was similarly modified; it was clearly higher in patients (3.415 -+ 1.506) than in controls (1.801 2 0.264) (P < Both the T8 and T4:T8 ratio were not found to correlate with clinical data, biologic inflammation, CIC, or Ty. On the contrary, * % E RFC 00- O0l 80 * 0 i” me 10 me* 0. 60. * * e. A 50. *e 40. m 10 . M 30 %T* %- 0 10 20 30 lG. Figure 3. A, Positive correlation between percentages of T3 and T8. 9, Lack of correlation between percentages of E-rosette forming cells (E-RFC) and TG (Ty). = individual values of circulating immune complex (CIC) positive polymyalgia rheurnatica patients. = values of CIC negative patients. The solid bar is the regression line. + DECREASE OF OKT8 IN PMR 1477 Table 2. Con A-induced suppression of the in vitro anti-TNP response* Number of anti-TNP-PFC per lob cells Patient Culture with Con A Culture without Con A 9 27 31 34 2 0 9 0 15 young 21.7 -t 16.2 allogeneic cells from normal young patients. Using this method, the suppressive capacity of patient cells, except for 1 case (patient 27), was found to be normal and similar to that of aged controls (Table 3). % suppression DISCUSSION 8 62 99 14 75 100 89 299 t 161 90.8 t 9.8 This study was undertaken to determine whether there could be lymphocyte abnormalities which would support the hypothesis of an immune mechanism as the origin of polymyalgia rheumatica. The results indicate a decrease in detected T y cells (P < lop6)(Figure 2) which was found to correlate with the presence of CIC, and mainly, a decrease in peripheral (Figure 1) which correlated with a T cells (P < major decrease of T8 subset ( P < lop9)(Figure 2). No statistical relationship was found between these two groups of abnormalities, nor with clinical data, biologic inflammation, or steroid therapy. The results obtained in the control group did not show the T cell imbalance usually found in aged persons (22,23). This is because of the low ma1e:female ratio (1:5) in our control group; the T cell abnormality occurs mainly in aged males. The comparison with such a sex-matched control group avoided bias differences due to the age of the patients, cells freezing, or any unexpected technical artifact. The presence of elevated levels of CIC in 8 of 10 patients with polymyalgia rheumatica alone was found by Papaioannou and coworkers (6). In these patients the CIC levels did not seem to correlate with the disease activity, contrasting with the results obtained in patients with giant cell arteritis in whom CIC and disease activity seemed clearly correlated (6). Using another method of CIC detection, Park and coworkers 100 controls * Con A = concanavalin A; anti-TNP-PFC plaque forming cells. = anti-trinitrophenyl they correlated in the patient group with E-RFC ( P < lop9)and T3 ( P < (Figure 3). Indirect estimation of the T cell subset positive for both OKT4 and OKT8 antigens (T4+8+) was made by subtracting the individual values of T4+8 from the corresponding sum T4+T8. No significant difference was found in patients (8.4 +- 4.4%) compared with controls (7.5 -+ 3.9%). Suppressor cell function. To correlate T8 defect with a functional T cell status, we investigated the capacity of T cells from 5 selected patients to inhibit autologous or allogeneic in vitro anti-TNP response. In autologous experiments, nylon-passed mononuclear cells from 4 patients exerted, after Con A stimulation, a significant inhibition of the specific plaque forming cell response similar to that of 15 different young controls (Table 2). Since some control responses were poor (patients 9 and 34), the suppressive capacity of patient Con A incubated cells was also investigated by inhibiting TNP-stimulated cultures of Table 3. Suppression of the in vitro anti-TNP response of cells cultured from young controls by Con A incubated cells from patients and aged controls* Number of TNP-PFC per 10' cells Subjects Con A incubated cells Cells without Con A suppression Patient 9 Experiment I ( Aged control Patient 27 Experiment 2 Aged control Patient 31 Experiment 3 Aged control (Patient 29 Experiment 4 Patient 34 [Aged control I1 3 27 27 10.8 15 160 129 I18 73 34 31 170 99.6 66 777 39 1 359 85 91 13 84 89 77 79 67 67 i % * Except for patient 27, all patients demonstrated normal suppression when Con A incubated cells were added to TNP-stimulated cells from normal young controls. TNP-PFC = anti-trinitrophenyl plaque forming cells; Con A = concanavalin A. 1478 (7) investigated 2 groups of polymyalgia rheumatica and/or giant cell arteritis patients and found 44% positive cases in patients with active disease and only 23% in the inactive group. However, in this study the polymyalgia rheumatica patients were not investigated separately and the relatively low percentage of positivity could be due to a lack of sensitivity of the technique (no positivity was found in controls). In contrast, Espinoza and coworkers (8) found 94% positive cases by the use of the same method as Papaioannou (6) and showed an obvious correlation with the presence of CIC and disease activity. This discrepancy in results of CIC investigations, which has also been noted in other diseases (24,25). casts doubt on the reliability of different techniques. It is indeed possible that a particular method could, by chance, detect mainly the pathogenic molecules among the CIC molecule population. However the well-known interferences between aggregated IgG and CIC detection could also explain conflicting results, particularly in these patients who usually have increased levels of IgG during periods of disease activity. In the present study, CIC were detected in 67.7% of patients; this was statistically different from The presence in vivo of the control group ( P < such complexes was suggested by their correlation with the decrease in Ty cells (P < 0.05). These cells are defined and characterized by the presence of a membrane receptor specific for the Fc part of IgG. This Fc receptor reacts with CIC, and this could therefore inhibit the detection of Ty rosettes. In addition, the interaction with CIC could also trigger cellular modifications of Ty, leading to the loss of their Fcy receptor (26). Thus, the apparent decrease of Ty cells could be due to the presence in vivo of CIC acting by two mechanisms: masking effect on the Fc receptor and triggering a transformation in another T cell subset. Thus, the total number of E-RFC might not be influenced by such a decrease of Ty, as was shown by the lack of correlation between the Ty percentage and the E-RFC. The decrease of Ty cells, even in patients with no CIC, might be due to another mechanism or to a low level of CIC, undetectable by our method. Although this method of detection is not quantitative, the lack of correlation with disease activity, also observed for the T y percentage, did not favor the hypothesis of an immune complex mechanism as the origin of polymyalgia rheumatica. Nevertheless this hypothesis is not ruled out since pathogenic CIC BENLAHRACHE ET AL molecules are probably not correlated to the total level of CIC and might be present even in CIC negative patients. The decrease in the peripheral total T cells was shown by 4 different methods: E-RFC, T3, T4+8, and the sum T4+T8. The 4 measurements provided similar results in each patient and were found to be highly correlated ( P < This decrease was relatively low (5 to 8% of MNC) and highly correlated to a selective defect in T8 subset (P< lop9). The normal percentages of T4, T4+T8, and B cell populations and the high negative correlation with T4:T8 ratio also suggest that T cell decrease was not due to variations in the non-T population but was secondary to the major T8 defect. The T8 subpopulation was found clearly lower in patients when compared with controls (P < This decrease was correlated to the total T cell decrease and was of similar size. No correlation was found with the other clinical or biologic data, nor with the decrease of Ty. Though both T y and T8 subsets include cells with suppressor activity, this lack of correlation was not inconsistent since T y detection was probably partially inhibited by CIC, and since the two populations have been shown to be clearly different (27). T8 marker defines a T cell population which mainly includes T3 positive cells with suppressor or cytotoxic activities (1 1). In contrast, the precise nature of T y cells is controversial. Reinherz and coworkers first established the presence of a myeloid marker (OKMI) together with the absence of the T3 marker at the surface of these cells (27) and concluded that T y cells may be non-T in origin. Further studies have modified these conclusions by showing the existence of T3 positive cells in a monoclonal T y clone (28), and also the presence of other markers specific for T cells in the T3 negative-OKM1 positive Ty population (29). Recent reports suggest the existence of different Ty subsets belonging to the T population and having OKMl plus T4 or T8 markers; these cells should be related to natural killer cells (30). The decrease of T8 cells in polymyalgia rheumatica patients was comparable to those observed in other diseases, though their mechanisms might be different. In rheumatoid arthritis patients, the T8 defect seems to be moderate in the peripheral blood and to contrast with a relatively high percentage of T8 in the synovial fluid and synovial tissue cell suspensions (13,14,31). A possible explanation of these data is that cytotoxic T8 would migrate to the synovial lesions and therefore be decreased in peripheral blood during 1479 DECREASE OF OKT8 IN PMR periods of disease activity (14). However, this hypothesis needs additional data since opposite results were reported showing a low T4:T8 ratio in lymphocytes from the rheumatoid synovial membrane when examined in situ (32). In multiple sclerosis, a similar T8 decrease was found and it also clearly correlated with disease activity ( 1 5 ~ 6 )In . leprosy, a decrease in T8 subset has been found only in patients with associated nodosum leprosum eruption (33), a vascular skin lesion possibly due to an Arthus phenomenon (34) with local deposits of immune complexes (35). Such a relationship between T8 percentage and this particular form of the disease has suggested a primitive T8 defect as the origin of an abnormal response t o the pathogen (33). In polymyalgia rheumatica, the drop in T8 proportion did not correlate with disease activity nor with the presence of CIC. This suggests that T8 are not responsible for the local lesions nor is this subset decrease a simple consequence of the disease. The T8 defect might preexist and favor the occurrence of the disease. Our data showing a normal suppressor cell function in 5 patients with abnormal T4:T8 ratios indicate that such a defect in T8 does not affect the suppressor cell subpopulation. Although these results need additional functional data, they point to the problem of correlations between functional experiments and numerations of subsets. In normal subjects, Chudwin and coworkers (36) found a good correlation between T lymphocyte subset enumerations and phytohemagglutinin o r mixed lymphocyte culture responses, though poor correlations were found in immunodeficient patients. In patients with acute polyradiculoneuritis (37) or lupus (38), normal T suppressor cell functions were found, contrasting with abnormal proportions of T cell subsets. T h e possibility of a cytotoxic precursor cell defect was suggested t o explain these unexpected results; however, a comparison with results obtained on human T cell lines shows that T8 includes cells with different immune functions whose selective defects are actually explored by using the same surface marker (39). In the absence of a better delineation of T cell subsets, the precise nature of the T8 defect remains obscure in polymyalgia rheumatica patients. It is likely that this defect corresponds t o a n abnormal immune status that might favor the occurrence of the disease. Participation of pathogenic immune complexes secondary t o this abnormal status might be the cause of local lesions. AKNOWLEDGMENTS We wish thank Dr. M. A. Bach and Df. S. Iscaki for their advice and critical reading of the manuscript, A. Portier for his excellent technical assistance, and Mole. C. Gauvin for secretarial support. REFERENCES 1. Bengtsson BA, Malmvall BE: The epidemiology of giant cell arteritis including temporal arteritis and polymyalgia rheumatica: incidences of different clinical presentations and eye complications. Arthritis Rheum 24:899-904. 1981 2. Bird HA, Esselinckx W, Dixon ASJ, Mowat AG, Wood PHN: An evaluation of criteria for polymyalgia rheumatics. Ann Rheum Dis 38:434-439. 1979 3. Liang GC, Simkin PA, Mannik M: Immunoglobulins in temporal arteries: an immunofluorescent study. Ann Intern Med 81:19-24, 1974 4. Hazleman BL, MacLennan ICM, Esiri MM: Lymphocyte proliferation to artery antigen as a positive diagnostic test in polymyalgia rheumatica. Ann Rheum Dis 34: 122-127. 1975 5. Papaioannou CC, Hunder GG, McDuffie FC: Cellular immunity in polymyalgia rheumatica and giant cell arteritis: lack of response to muscle or artery homogenates. Arthritis Rheum 22:740-745, 1979 6. Papaioannou CC, Gupta RC. Hunder GG, McDuffie FC: Circulating immune complexes in giant cell arteritis and polymyalgia rheumatica. Arthritis Rheum 23: 1021-1025, 1980 7. Park JR, Jones JG, Harkiss GD. Hazleman BL: Circulating immune complexes in polymyalgia rheumatica and giant cell arteritis. Ann Rheum Dis 40:360-365, 1981 8. Espinoza LR, Bridgeford P, Lowenstein M, Bocanegra TS, Vasey FB, Germain BF: Polymyalgia rheumatica and giant arteritis: circulating immune complexes. J Rheumatol 9556-560, 1982 9. Levinsky RJ: The measurement of immune complexes: is it clinically useful? Immunology Today 2:94-97, 1981 10. Plotz PH: Studies of immune complexes. Arthritis Rheum 25:1151-1155, 1982 11. Reinherz EL, Schlossman SF: The differentiation and function of human T lymphocytes. Cell 19:821-827, 1980 12. Bach MA, Bach JF: The use of monoclonal anti-T cell antibodies to study T cell imbalances in human diseases. Clin Exp Immunol 45:449-456, 1981 13. Veys EM, Hermanns P. Schindler J, Kung PC, Goldstein G, Symoens J , Van Wauwe J: Evaluation of T cell subsets with monoclonal antibodies in patients with rheumatoid arthritis. J Rheumatol 9:25-29, 1982 14. Fox RI, Fong S, Sabharwal N, Carstens SA, Kung PC, Vaughan JH: Synovial fluid lymphocytes differ from peripheral blood lymphocytes in patients with rheumatoid arthritis. J Immunol 128:351-354, 1982 IS. Reinherz E, Weiner HL, Hauser SL, Cohen JA, Distaso 1480 JA, Schlossman SF: Loss of suppressor T cells in active multiple sclerosis: analysis with monoclonal antibodies. N Engl J Med 303:125-129, 1980 16. Bach MA, Phan-Dinh-Tuy F, Tournier E, Chatenoud L, Bach JF, Martin C, Degos JD: Deficit of suppressor T cells in active multiple sclerosis. Lancet ii: 1221-1223, 1980 17. Fox RI, Carstens SA, Fong S, Robinson CA, Howell F, Vaughan JH: Use of monoclonal antibodies to analyze peripheral blood and salivary gland lymphocyte subsets in Sjogren’s syndrome. Arthritis Rheum 25:419-426, 1982 18. Grangeot-Keros L, Segond P, Capel F, Iscaki S, Pillot J: Detection of immune complexes: a simple assay based on characterization of the in vivo bound Clq (PEG-Clq immunodiffusion test). J Immunol Methods 23:349-362. 1978 19. Hallberg T , Gurner BW, Coombs RA: Opsonic adherence of sensitized ox red blood cells to human lymphocytes as measured by rosette formation. Int Arch Allergy Appl Immunol 44:500-513, 1973 20. Galanaud P, Crevon MC, Delfraissy JF. Segond P, Wallon C, Dormont J: Regulation of the primary in vitro antibody response in human peripheral blood lymphocytes: different effects of mitogen-induced and spontaneous T suppressor cells. Clin Exp Immunol 38: 106-1 IS, 1979 21. Segond P, Salliere D. Galanaud P, Desmottes RM. Massias P. Fiessenger JN: Impaired primary in vitro antibody response in progressive systemic sclerosis patients: role of suppressor monocytes. Clin Exp Immunol 47: 147-154, 1982 22. Nagel JE. Chrest FJ, Adler WH: Enumeration of T lymphocyte subsets by monoclonal antibodies in young and aged humans. J Immunol 127:2086-2088, 1981 23. Mascart-Lemone F, Delespesse G, Servais G, Kunstler M: Characterization of immunoregulatory T lymphocytes during ageing by monoclonal antibodies. Clin Exp Immunol 48:148-154, 1982 24. Fiist G, Fekete B, Angyal I. Jakab A, Pal A, Meretey K , Falus A, Torok K, Szegedi G, Kavai M, Puskas E, Csecsi-Nagy M, Szabo I , Lenkey A, Misz M: Evaluation of different methods for detecting circulating immune complexes: studies in patients with lung cancer. J Immunol Methods 46:259-276, 1981 25. McDougal JS, Hubbard M, McDuffie FC, Strobel PL, Smith SJ. Bass N , Goldman JA, Hartman S, Myerson G , Miller S, Morales R, Wilson CH Jr: Comparison of five assays for immune complexes in the rheumatic diseases: an assessment of their validity for rheumatoid arthritis. Arthritis Rheum 2S:llS6-1166, 1982 26. Moretta L, Mingari MC, Romanzi CA: Loss of FC receptors for IgG from human T lymphocytes exposed to IgG immune complexes. Nature 272:618-620, 1978 BENLAHRACHE ET A L 27. Reinherz EL, Moretta L, Roper M, Breard JM, Mingari MC, Cooper MD, Schlossman SF: Human T lymphocyte subpopulations defined by Fc receptors and monoclonal antibodies: a comparison. J Exp Med 151:969974, 1980 28. Strong DM, Pandolfi F, Slease RB, Budd JE, Woody JN: Antigenic characterization of a T-CLL with heteroantisera and monoclonal antibodies: evidence for the T cell lineage of an Ia-positive, Fc-IgG-positive, suppressor-cell subpopulation. J Immunol 126:22OS-2208, 1981 29. Fox RI, Thompson LF, Huddlestone JR: T y cells express T lymphocyte-associated antigens. J Immunol 126:2062-2063, 1981 30. Van de Griend RJ, Ten Berge I, Tanke HJ, Roos D, Schellekens PTA, Melief CJM, Zeijlemaker WP, Astaldi A: Characterization of two subsets of human T y cells. J Immunol 128:1979-1985, 1982 31. Burmester GD, Yu DTY, Irani AM, Kunkel HG, Winchester RJ: Ia+ T cells in synovial fluid and tissues of patients with rheumatoid arthritis. Arthritis Rheum 24:1370-1376, 1981 32. Meijer CJLM, de Graaf-Reitsma CB, Labefer GJM. Cats A: In situ localization of lymphocyte subsets in synovial membranes of patients with rheumatoid arthritis with monoclonal antibodies. J Rheumatol 9:359-365, 1982 33. Bach MA, Chatenoud L, Wallach D, PhanDinh Tuy F, Cottenot F: Studies on T cell subsets and functions in leprosy. Clin Exp Immunol 44:491-500, 1981 34. Wemembu SNC, Turk JL, Waters MFR, Rees RJW: Erythema nodosum: a clinical manifestation of the Arthus phenomenon. Lancet ii:933-935, 1969 35. Bjorvatn B, Barnetson RS, Kronvall G, Zubler RH, Lambert PH: Immune complexes and complement hypercatabolism in patients with leprosy. Clin Exp Immunol 26:388-396, 1976 36. Chudwin DS, Cowan MJ, Wara DW, Ammann AJ: Patients with abnormal proportions of T-lymphocyte subsets have reduced in vitro cellular immunity. Clin Immunol Immunopathol 26: 126-136, 1983 37. Hughes RAC, Aslan S, Gray IA: Lymphocyte subpopulations and suppressor cell activity in acute polyradiculoneuritis (Guillain-Barre syndrome). Clin Exp Immunol 5 1:448-454, 1983 38. Tsokos GC, Balow JE: Phenotypes of T lymphocytes in systemic lupus erythematosus: decreased cytotoxicl suppressor subpopulation is associated with deficient allogeneic cytotoxic responses rather than with concanavalin A-induced suppressor cells. Clin Immunol Immunopathol 26:267-276, 1983 39. Mingari MC, Moretta A, Pantaleo G, Moretta L: Surface markers of resting and activated human T cells: functional implications and experimental limits. Semin Immunopathol S:477-488, 1982
1/--страниц