Antibodies to HTLV-I in sera from patients with connective tissue diseases.

SIR, It has been suggested that HTLV-I infection may be associated with the development of some connective tissue diseases (CTD), including polyand dermatomyositis (PM and DM), systemic lupus erythematosus (SLE), scleroderma, and Sjogrens syndrome (SS).' 2 This was based on eight case reports in which neither the ethnic origin of their patients nor the assays used to determine HTLV-I seropositivity were defined. Six cases of PM in HTLV-I antibody positive Jamaicans have also been reported (Rodgers-Johnson et al, quoted in ref 3). These observations prompted us to examine the HTLV-I antibody status of a large cohort of patients with CFD attending a rheumatology clinic in London. A total of 98 serum samples from 80 patients: 42 with SLE, of whom eight also had SS, and 38 with PMIDM, were screened using an HTLV-I enzyme linked immunosorbent assay (ELISA) kit (Dupont Ltd). Most (55) were Caucasian, 16 were of Indian, Middle, or Far East descent, and nine were West Indians. Positive samples were retested using the same assay system, and repeatedly positive sera were further tested by particle agglutination and immunofluorescence on HTLV-I infected cells. A total of 98 serum samples was tested and the results were as follows: screen positive: 5, repeat positive: 3, confirmed positive: 1. The two serum samples that were positive in the ELISA test on one occasion only were from the same patient, a child with DM; eight other serum samples from this child were negative in the ELISA. The two samples that were consistently positive in the ELISA, but not confirmed positive, were from one patient with PM and one with SLE, both of whom were Caucasian. The only confirmed HTLV-I antibody positive serum was from