Treatment of syphilis: a systematic review.

Abstract

The incidence of syphilis in the United States is increasing; it is estimated that more than 55,000 new infections will occur in 2014. Treatment regimens are controversial, especially in specific populations, and assessing treatment response based on serology remains a challenge. To review evidence regarding penicillin and nonpenicillin regimens, implications of the "serofast state," and treatment of specific populations including those with neurosyphilis or human immunodeficiency virus (HIV) infection and pregnant women. We searched MEDLINE for English-language human treatment studies dating from January 1965 until July 2014. The American Heart Association classification system was used to rate quality of evidence. We included 102 articles in our review, consisting of randomized trials, meta-analyses, and cohort studies. Case reports and small series were excluded unless they were the only studies providing evidence for a specific treatment strategy. We included 11 randomized trials. Evidence regarding penicillin and nonpenicillin regimens was reviewed from studies involving 11,102 patients. Data on the treatment of early syphilis support the use of a single intramuscular injection of 2.4 million U of benzathine penicillin G, with studies reporting 90% to 100% treatment success rates. The value of multiple-dose treatment of early syphilis is uncertain, especially in HIV-infected individuals. Less evidence is available regarding therapy for late and late latent syphilis. Following treatment, nontreponemal serologic titers should decline in a stable pattern, but a significant proportion of patients may remain seropositive (the "serofast state"). Serologic response to treatment should be evident by 6 months in early syphilis but is generally slower (12-24 months) for latent syphilis. Evidence defining treatment for HIV-infected persons and for pregnant women is limited, but available data support penicillin as first-line therapy. The mainstay of syphilis treatment is parenteral penicillin G despite the relatively modest clinical trial data that support its use.