Prenatal BPG delivery: 9-day treatment interval linked to higher risk of congenital syphilis

When treating late latent syphilis during pregnancy, extending the interval between benzathine penicillin G (BPG) doses to 9 days may put newborns at risk of congenital infection relative to 6–8-day intervals, according to new research.

In a cohort of mother–infant dyads with prenatal exposure to three doses of BPG, congenital syphilis occurred more frequently when at least one dose was spaced 9 days apart compared to when all dose intervals were 8 days or less (18.2 percent vs 5.6 percent; p=0.01), reported first study author Dr Kelly Johnson from the University of California San Francisco in San Francisco, California, US. [CROI 2025, abstract 161]

A 9-day vs 6–8-day BPG dose interval was associated with a nearly fourfold increase in the odds of congenital syphilis (odds ratio [OR], 3.7, 95 percent confidence interval [CI], 1.2–11.6), Johnson added.

BCG, at a schedule of three weekly intramuscular injections, is the only recommended treatment for syphilis of late latent or unknown duration in pregnancy, she said. “The gold standard is that these injections would be given at strict 7-day intervals. But our national Centers for Disease Control and Prevention STI treatment guidelines imply, based on expert opinion, that intervals up to 9 days may be acceptable without having to restart the full three-dose series.”

The findings of the study contradict the guidelines, suggesting that 9-day dose intervals increase the likelihood of congenital syphilis, according to Johnson.

Study details

For the study, Johnson and colleagues used the California Department of Public Health STI surveillance data. They identified 1,109 mother–infant dyads between 2016 and 2023, where the mother had late latent syphilis during pregnancy with reactive Rapid Plasma Reagin (RPR).

The dyads were grouped into three, as follows: 677 received BCG doses at 6–8-day intervals, 22 received at least one BCG dose at a 9-day interval, and 410 received no or inadequate treatment. Treatment was initiated 30 or more days before childbirth.

Key baseline characteristics were similar between the 6–8 day and 9-day groups. No significant differences were seen in maternal HIV status (positive: 0.5 percent vs 0 percent), maternal RPR ratio of at least 1:32 (45.1 percent vs 40.9 percent), receipt of prenatal care (97.3 percent vs 95.5 percent), or gestational age at syphilis treatment (median 16 vs 21 weeks).

Congenital syphilis occurred in 36.8 percent of infants whose mothers received no or inadequate treatment, yielding an OR of 9.8 (95 percent CI, 6.7–14.4) relative to infants whose mothers received at least one BCG dose at a 9-day interval (p=0.08).

Congenital syphilis was defined by using infant criteria only. These included lab evidence of infection and a clinical sample, report of syphilitic stillbirth, infant’s RPR being fourfold higher than the mother’s at delivery, and a reactive RPR plus any one of the following: either a congenital syphilis finding on an examination or long bone x-rays or either a reactive cerebrospinal fluid (CSF) venereal disease research laboratory test or a CSF white blood cell count or protein elevation.

Err on the side of caution

Johnson acknowledged several study limitations. First was the small numbers of cases, which prevented the performance of adjusted analyses. Others were the possibility of unmeasured confounders, lack of data for all variables of interest (eg, maternal weight), and potentially overly sensitive congenital syphilis surveillance criteria. Additionally, most of the mother–child pairs in the 9-day interval group came from a later period (July 2019-December 2023) than those in the 6–8-day interval group.

“But despite these limitations, data on [the effect of administering BCG doses at longer intervals for treating late latent syphilis during pregnancy] is otherwise lacking. In fact, it’s completely nonexistent, and a randomized control trial on this study is never going to be conducted; it would be unethical,” Johnson noted.

“For those reasons, given the lack of data otherwise and while awaiting larger studies with higher power, we think that clinicians should consider erring on the side of caution and avoiding intervals outside 6–8 days whenever possible,” she said.

In a question-and-answer session, a member of the audience raised the point that the known pharmacokinetics of penicillin would not support a hypothesis that a 9-day interval might be worse than 7 days, given the drug’s slow decline in the system. Johnson, in response, challenged such an assumption, noting that the available pharmacokinetic data on BPG are mostly in nonpregnant people.

“There’s one study I’m aware of in pregnant people,” Johnson said. In that study, participants in their late pregnancy were already below therapeutic penicillin levels at 7 days after a BPG injection, she added. “So, I think that is a theory that is still worth investigating, particularly in pregnancy.”