Antigen-induced arthritis was established in the mouse by immunization with methylated bovine serum albumin (mBSA) in complete Freund's adjuvant with B pertussis vaccine. The knee joint was injected after 21 days with mBSA in saline. The arthritis was chronic, antigen-specific, and T-cell dependent in hypothymic nu/nu mice. C57BL and BALB/c mice were susceptible, whereas CBA mice were relatively resistant. Susceptibility was dominant: one gene was loosely linked to the "b" allele of the H-2 complex of C57BL mice. and macrophages in the rheumatoid synovium. In animals various experimental diseases of articular and periarticular tissues have been induced by infective, chemical, endocrine, physical, and immunologic procedures. The models that most closely resemble rheumatoid arthritis depend upon either infectious agents (1,2) or immunologic responses (3-5). Antigen-induced arthritis was induced initially by intraarticular injection of antigen in immunized rabbits, and may be the closest experimental analogue of human rheumatoid arthritis ( 5 ) . This model was therefore established in mice to investigate humoral and cell-mediated immunity in the induction and persistence of disease, and to study genetic aspects of susceptibility and resistance. MATERIALS AND METHODSMice. Ten-week-old male and female mice were used.The strains were BALB/c, CBA/CaH WEHI, C3H/HeJ and the closely related C3H.SW, and C57BL/6J WEHI and the closely related BIO.BR. F, and F, hybrids and backcross mice were selectively bred. The mice were reared in a specific pathogen-free (SPF) facility and maintained conventionally after 6 weeks of age. Hypothymic "nude" (nu/nu) mice of CBA and BALB/c strains were used at 6 weeks of age. The BALB/c. nu/nu and CBA. nu/nu mice were derived from the eighth backcross to BALB/c and CBA respectively. These mice, which are homozygous for the recessive mutation "nude," so called because they are hairless, suffer from thymic hypoplasia and are highly deficient in T cells. The heterozygotes (nu/+) are normal. Antigens. The antigens included bovine serum albumin (BSA, Cohn fraction V ) from Armour Pharmaceutical Com-

Human granulocyte elastase (EC 3.4.21.37) is released from granulocytes in large amounts in chronic inflammatory joint diseases and is therefore of special pathogenic and diagnostic importance. In order to examine the diagnostic significance of this enzyme as a clinico-chemical parameter, we determined the concentration of granulocyte elastase in complex with alpha 1-proteinase inhibitor by an enzyme immunoassay in synovial fluids and plasma of patients with chronic joint diseases. In inflammatory synovial fluids the concentration of complexed elastase correlates well with the granulocyte number and may increase to an extremely high level. In 90% of patients with manifest rheumatoid arthritis increased elastase levels are also observed in the plasma, probably due to the large gradient between the synovial fluid and plasma concentration, whereas in osteoarthrosis normal plasma concentrations were observed. Thus, these results indicate that normal plasma concentrations in patients with chronic joint diseases exclude the diagnosis of rheumatoid arthritis with high probability. The simultaneous determination of complexed elastase in plasma and synovial fluid improves the nosological differentiation of chronic joint diseases. Elastase activity on a specific chromogenic substrate, which was found in many inflammatory synovial fluids, is mainly attributed to elastase alpha 2-macroglobulin complexes. In some purulent synovial fluids, however, we were able to detect free elastase, which has been shown to play an important role in the destruction of articular cartilage.

In a multicenter placebo-controlled double-blind randomized clinical study, 91 patients with rheumatoid arthritis were given 28 days' treatment with recombinant interferon-gamma (50 micrograms daily for 20 days, then 50 micrograms each second day up to day 28, given by subcutaneous injection). The aim of the study was to provide a methodologically clear demonstration of the efficacy of treatment with interferon-gamma, using criteria that could be handled by statistical tests. Evaluatable documentation was available for 79 patients, of whom 40 were treated with the active compound. The principal criterion for the statistical evaluation of the therapeutic success was improvement of the Ritchie "joint pain index" or Lansbury "joint pain index" by at least 30% within 28 days. The chi-square test showed superiority of the interferon arm over the placebo arm with an error probability of alpha less than 1%. In addition, efficacy of interferon-gamma was demonstrated in respect of practically all parameters investigated. The frequency of side-effects, including febrile reactions, was the same for the active compound and the placebo. During interferon treatment the daily maximum body temperature was raised by 0.3 degrees C on average, but was below 37.2 degrees C at all times.

Cross‐reactivity between type II collagen (CII) and C1q, the collagen‐like subunit of the first component of complement, has been demonstrated in synovial fluid (SF) from rheumatoid arthritis (RA) patients. Many authors have studied autoimmunity to CII in RA, but little work has been done on autoimmunity to C1q in RA. In the data presented here, we have been able to show that in addition to native C1q, an altered form of C1q is present in SF from RA patients. Furthermore, a low molecular weight form of C1q is present in RA SF, although its role, if any, in the pathogenesis of RA is unclear. The presence in these RA SF of C1q‐specific antibodies (IgG and IgM) has been studied and we have partially characterized the antibody moieties involved. As well as binding to C1q and fragments representing the collagen‐tails from C1q, 7 S IgG autoantibodies against C1q also bind to a C1q molecule altered in vitro by incubation with reactive oxygen species and to the non‐apeptide KGEQGEPGA (representing residues 26–34 from the C1q A‐chain), which has previously been shown to suppress the onset of CII‐induced arthritis in an animal model.

Summary The half‐life of125I labelled CINH was determined in patients with HAE and normal controls. There was no significant difference between these two groups. The half‐life in the HAE patients was 67.7 hr ± 4.9 hr (s.d.) and in the normals 64 hr± 1.4 hr (s.d.). This finding is consistent with a defect in synthesis as an explanation of the low serum C1NH levels in HAE patients.