Active syphilis in HIV infection: a multicentre retrospective survey. The German AIDS Study Group (GASG).

OBJECTIVE To study syphilis in HIV infection focusing on immunocompromised patients with an atypical or aggressive clinical course of syphilis, inappropriate serological reactions or an unreliable response to therapy. STUDY DESIGN A multicentre retrospective chart review using a standardised questionnaire for all patients with active syphilis. SETTINGS Thirteen dermatological and medical centres throughout Germany, all members of the German AIDS Study Group (GASG). PATIENTS Clinical data of 11,368 HIV infected patients have been analysed for cases of active syphilis requiring treatment. Asymptotic patients with reactive serological parameters indicating latent syphilis without a need for treatment were excluded. RESULTS Active syphilis was reported in 151 of 11,368 HIV infected patients (1.33%, range per centre 0.3%-5.1%). Most of the 151 syphilis patients were male (93%) and belonged to the homosexual or bisexual exposure category for HIV infection (79%); another 6% were iv drug users. Among the 151 syphilis patients primary syphilis was diagnosed in 17.2%, maculopapular secondary syphilis in 29.1%, ulcerating secondary syphilis in 7.3%, neurosyphilis in 16.6% and latent seropositive syphilis without clinical symptoms but serological abnormalities indicating active syphilis in 25.2%. A history of prior treatments for syphilis was reported in 50%. At the time of syphilis diagnosis 26.5% of the patients were in CDC stage II, 33.8% in stage III and 24.5% in stage IV of HIV disease (CDC classification 1987). CD4 cell count was lowest in those with ulcerating secondary syphilis (mean 307, SD 140/microliters) and neurosyphilis (351, SD 235/ microliters). The highest CD4 count was found in patients with early primary and early secondary syphilis (444, SD 163/microliters and 470, SD 355/microliters). Inappropriate serological response to syphilis infection was found in 81 of 151 patients (54%). Remarkable findings were false negative VDRL titres (11 patients with non primary syphilis), false negative TPHA (1) or 19S-IgM-FTA-ABS-tests (16), and strongly reactive VDRL (> or = 512, 8) or TPHA titres (> or = 10 240, 47). Treatment failures were reported in at least 6 of 151 cases (4%). CONCLUSIONS Atypical clinical and serological courses of syphilis were observed in HIV infected patients. Ulcerating secondary syphilis with general symptoms ("malignant syphilis") was 60 times more frequent than in historic syphilis series. Neurosyphilis was found in one sixth of those with active syphilis. Therefore lumbar puncture should be considered a routine in coinfections with HIV and syphilis. Treatment efficacy should be monitored carefully.