Single Penicillin Shot Treats Early Syphilis as Well as Three Weekly Doses

Syphilis rates have surged throughout the 21st century, with congenital syphilis near record highs in the US and disproportionate rates of HIV co-infection among affected individuals. While a single 2.4 million unit (MU) intramuscular injection of benzathine penicillin G (BPG) has been the CDC-recommended standard of care for early syphilis for decades, many clinicians — particularly in HIV care settings — have routinely administered three weekly doses out of concern that immunosuppression might blunt treatment response. No definitive randomized trial had previously resolved this controversy.

Hook and colleagues conducted an open-label multicenter randomized non-inferiority trial at ten US sites between October 2018 and March 2022. Adults aged 18 and older with confirmed early syphilis (primary, secondary, or early latent) were randomized 1:1 to receive either a single BPG injection (2.4 MU) or three weekly BPG injections (7.2 MU total). The primary endpoint was a ≥4-fold decline in Rapid Plasma Reagin (RPR) titer or RPR seroreversion by six months. A non-inferiority margin of 10% was pre-specified. Of 249 enrolled participants, 199 were evaluable at six months. The cohort was predominantly male (97%), Black (62%), and non-Hispanic (95%); 61% were living with HIV, of whom 92% were on antiretroviral therapy and 77% had CD4 counts above 350 cells/mm³.

In the intention-to-treat analysis, serologic response was 76% (95% CI 68–82%) in the single-dose group and 70% (95% CI 61–77%) in the three-dose group. The Farrington-Manning test yielded a difference of −0.06 with a 90% CI upper bound of 0.03 — below the 0.10 non-inferiority margin — confirming that single-dose therapy is non-inferior. Per-protocol results were consistent: 77% vs 71%. Critically, no clinical relapses or treatment failures were observed in either arm. HIV status did not modify the outcome: among people living with HIV, response rates were 76% (one dose) versus 71% (three doses). Even among those with CD4 counts below 200 cells/mm³, 69% achieved a ≥4-fold serologic response.

Several secondary findings add clinical texture. Early latent syphilis had a lower serologic response rate (65%) compared to primary or secondary stages (77%), consistent with prior literature. Participants who experienced a Jarisch-Herxheimer reaction after initial treatment were significantly more likely to respond serologically (85% vs 69%), suggesting this post-treatment inflammatory response may serve as a surrogate marker of robust treatment effect. Adverse events were common but manageable: injection-site pain affected 76% of the single-dose group and 85% of the three-dose group, and Jarisch-Herxheimer reactions occurred in 24% of all participants. Three serious adverse events occurred, none attributed to BPG.

The implications are substantial. Single-dose BPG simplifies treatment, reduces patient burden, supports antimicrobial stewardship, and is critical during the recurring global shortages of BPG that have repeatedly hampered syphilis control programs. The trial was halted early due to COVID-19-related enrollment slowdowns, falling short of its 420-participant target, though futility analyses confirmed that the observed response rates of approximately 79% in both arms still provided 80% power. The study was also predominantly male and conducted in the US, limiting generalizability to women and global populations. Nonetheless, these data provide the strongest evidence to date that three doses of BPG offer no clinical or serologic advantage over one for uncomplicated early syphilis, regardless of HIV status.