The chronic, rheumatoid synovitis in four patients (two children and two adults) with immunodeficiency was studied by means of immunopathological examination of synovial tissues and fluids. Their immunodeficiencies were extensively studied, with various tests for humoral and cell-mediated immunity. The two children were boys with Bruton's disease. One adult, female, had common variable immunodeficiency. One adult male had severe hypogammaglobulinemia. All patients had isolated deposits of complement component C3 in their synovial membranes, usually without traces of immunoglobulins. Depressed levels of complement component C3 were found in the joint fluid, in contrast to high levels in the corresponding serum in one patient. Components of the alternate pathway, properdin and C3A, were found in the tissues, but not components of the classic pathway, C1q and C4. The findings suggest that chronic, rheumatoid arthritis in hypogammaglobulinemic patients may be related to activation of C3 by the alternate pathway.

The presence of a serum factor secreted by the thymus has been demonstrated in the serum of mice and man using a rosette inhibition assay. This factor is a peptide with a molecular weight close to 1,000. Its level is age dependent, being stable until the age of 6 months in mice and until 20 years in man. New Zealand Black mice show a premature cessation of thymic hormone secretion preceeding the onset of other T cell abnormalities and auto-immunity. Before any treatment the serum thymic hormone level is abnormally low in most cases of SLE even in patients under 25, an age where control subjects still show high hormone levels. Conversely, normal or high levels are found in RA, PARTICularly in patients over 40 who showed in 60% of cases a hormone level significantly higher than normal controls of the same age. In keeping with this data T cells evaluated by SRBC spontaneous rosette formation show low figures in some patients with active SLE and normal or high values in most cases of RA.

Anti-lymphocyte antibodies have been demonstrated in patients with SLE and RA by immunofluorescence. They differ from those arising through iso-immunization by being cold reactive and chiefly of the IgM class. Separate anti specificities were shown for determinants on either T or B cells.

Patients with rheumatoid arthritis (RA), 36 cases, and normal subjects, 49 cases were studied by lymphocyte cultures stimulated by phytohemagglutinin (PHA), Concanavalin A (Con A), pokeweed mitogen (PWM) and Con A convalently bound to Sepharose 4 B (Con A-S). The comparisons between the two groups showed a significant difference between the RA lymphocytes and the control lymphocytes stimulated by PHA and Con A. However, no statistical difference was found between the two lymphocyte populations stimulated by PWM and Con A-S. In order to determine the lymphocyte population stimulated by each mitogen, separation of B and T cells from peripheral blood was performed according to the ability for the T cell population to bind the sheep red blood cells (rosette-forming cells). The T cell-rich population was only stimulated by PHA, Con A and PWM. Although the T cell-depleted population showed no response to these mitogens, a response to Con A-S was elicited.

Comparison of virus antibody titers in JRA and matched controls failed to implicate rubella, measles, parainfluenza type 1 or Epstein-Barr viruses. Neither RNA viruses or mycoplasma were detected in rheumatoid synovial cell cultures by 3H-uridine incorporation. Mycoplasma were not isolated from rheumatoid synovial membranes or cell cultures using the large specimen-broth culture procedure. The balance of available evidence, including our own, suggests that mycoplasma are not involved in RA. Viruses, however, are better candidates, in spite of similar generally negative studies to date, because methods for detecting known latent infection in animals and in vitro are still in their infancy. Those that are known have not yet been comprehensively applied to RA. Thus, supported by indirect evidence from animal models and human diseases like polyarteritis, the hypothesis that viruses are involved in the pathogenesis of RA remains attractive, although direct evidence is still lacking.

Knee joint arthritis is induced among rabbits previously immunized by intradermal injection with egg albumin (EA) emulsified in adjuvant containing either M. tuberculosis or M. butyricum, then by intra-articular injection with EA. Arthritis evolution involves two phases, an early one during the first 2 months and a late one from 3 months to 1 year. During the early phase, arthritis intensities are similar no matter which Mycobacterium is used. However, during the late phase, only rabbits immunized with M. tuberculosis develop self-perpetuating arthritis. Among more than 50% of arthritic rabbits, immunological lesions of aortic artery and cardiac valvules are found. Among the rabbits immunized with M. tuberculosis, the humoral anti-EA antibody level remains constant during the whole arthritis evolution; but, among the rabbits immunized with M. butyricum, the arthritis intensity decreases from 3 months of evolution. The correlation between arthritis index (AI) and humoral antibody level is only significant among the rabbits with early arthritis. The intradermally immunized rabbits show a positive skin test with EA and tuberculin. The more intense the cutaneous reactions, the greater the chances of developing self-perpetuating arthritis after the EA intra-articular injection. The fluorescent anit-EA antibodies in the synovia and spleen are found only among the early arthritis. After 2 months of evolution, fluorescent antibodies disappear whatever the immunization may be. Among the immunized rabbits, it is probable that antigenic EA does not persist in the synovia. Indeed, the autologous inflamed synovia transplantation, from the donor-challenged knee joint, does not develop an inflammatory reaction in the non-challenged knee joint. The fluorescent immunoglobulins IgG and IgM in the synovium of arthritic rabbits are found with the same percentages as fluorescent anti-EA antibodies. The lymphocyte response to EA, PHA and PWM are positive whatever the immunization and arthritis evolution may be. There is no correlation between AIs and lymphocyte responses to specific and nonspecific mitogens. It is probable that self-perpetuation depends closely on M. tuberculosis whose adjuvant power is much superior than M. butyricum and not on antigenic EA whose essential role would be to trigger the inflammatory process.

The role of acute inflammation and of specific antibody in the retention of antigen in joint collagenous tissues of immunized rabbits was examined. The role of the acute synovitis occurring immediately after IA antigen injection in the retention of immune complexes was investigated by the concomitant IA injection of trace amounts of 1258-BSA and mono-sodium urate crystals and by the production of an acute Arthus reaction using an unrelated antigen in doubly immunized rabbits. In neither case was more 125I-BSA retained in the inflamed joint tissues compared to the contralateral noninflamed joints 7 days after antigen injection. In addition, immunized rabbits depleted of ciruclating polymorphonuclear leukocytes by previous treatment with nitrogen mustard retained 2.7 times more antigen than control, nonimmune animals. In another experiment, antigen retention in joint collagenous tissues was greatest 30 min after IA injection, before the appearance of acute inflammatory synovitis. These findings suggest that acute inflammation is not a major factor in the retention of antigen in collagenous tissues. To investigate the role of antibody in the retention of antigen, nonimmune rabbits were injected i.v. with purified anti-BSA antibody three days prior to the intraarticular injection of BSA. Over 20 times more antigen was retained irreversibly in collagenous tissues obtained from the injected joints of passively immunized animals as compared to similar tissues of control rabbits. When rabbits were injected i.v. with purified anti-BSA antibody and either killed 20 min or 3 days later, in vitro binding of antigen by joint collagenous tissues was seen only in animals where antibody was allowed to equilibrate with the extravascular space for 3 days. These findings indicate that retention of antigen depends on the presence of extravascular antibody. It is concluded that the trapping of immune complexes in collagenous joint tissues of immunized animals depends on: (1) the presence of antibody in the extravascular space; (2) the diffusion of antigen or soluble complexes into this space; (3) the interaction of antigen or soluble complexes with extravascular antibody with subsequent formation of larger and more insoluble complexes, and (4) the trapping of these complexes in the collagen fiber meshwork.

The knee joint in rabbits has been used as an arthritis model for the study of acute immunologic injury. Histologic examination of synovium plus a standardized procedure for irrigating inflammatory exudate from the joint allowed study of damage due to immune complexes, urate crystals, and Arthus type vascular lesions by multiple parameters. With this model Arthus lesions could be abolished with neutrophil depletion and reconstituted with intra-articular injection of neutrophil suspension. The rate and degree of injury could also be diminished by depletion of complement components.

Peripheral blood lymphocytes from some rheumatoid arthritis (RA) patients show a reduced response to PHA when stimulated directly after isolation from blood. Cells which have been preincubated in medium with normal serum for 22 h before exposure to PHA show a relatively enhanced response, which approximates the normal. These findings could be compatible either with the blocking of surface PHA receptors of circulating RA lymphocytes, or with progressive maturation of these cells in vitro. Preincubation does not enhance the relatively reduced response of RA lymphocytes to allogeneic RA lymphocytes in one-way mixed lymphocyte cultures. An increased reduction in response to PHA following irradiation of RA lymphocytes suggests an increased radiosensitivity of the PHA-responsive subpopulation. This abnormally reduced response was not seen if the cells were preincubated in vitro prior to irradiation.

An anaphylactic mechanism involving IgE-sensitized basophils is necessary for immune complex deposition in rabbits. Basophils under stimulation by the IgE-bound antigen actively release a platelet-activating factor. Increased vascular permeability results from action of vasoactive amines from basophils and from aggregated platelets and possibly direct action of PAF on vascular walls. This allows trapping of circulating complexes along vascular membranes. Once deposited complexes produce tissue injury by neutrophil-complement-dependent mechanisms. However, complex deposition seems pathogenic in itself for the glomerular basement membrane. PAF is a 1,100-dalton MW lipid, with hydrophobic properties and positive charge; it exists in human basophils. Patients suffering from SLE or undergoing anaphylactic shocks exhibit marked reduction of basophil counts and PAF level. We propose that anaphylactic increase of vasopermeability may be a common feature at early stages of many immune diseases associated with severe structural injury.