You don't need to be signed in to read BMJ Blogs, but you can register here to receive updates about other BMJ products and services via our site.

Primary Care Corner with Geoffrey Modest MD: 1 shot of penicillin for early syphilis in HIV patients??

13 Apr, 17 | by gmodest

by Dr Geoffrey Modest

2 articles of note just came out on syphilis.

MMWR presented data on rates of primary and secondary syphilis in the US in 2015 (see ). The overall case rate was 7.5/100K population, nearly 4 times the previous lowest documented rate of 2.1/100K in the year 2000, a nadir after which it has increased each year. The rate increased 22% during 2011-13.


— in 2015, there were 23,872 reported primary and secondary syphilis cases in the United States.

— 81.7% of male primary and secondary syphilis cases were among gay, bisexual, and other men who have sex with men (MSM)

— among the 44 states reporting information on the sex of sex partners for > 70% of male cases, the rates were:

— overall for men over 18 years old: 17.5/100K

— men who have sex only with women: 2.9/100K

— MSM: 309.0/100K, which translates to:

106.0 times the rate among men who have sex with women only, varying by states from 39.2-342.1 times

— 167.5 times the rate among women

— the highest rates of primary and secondary syphilis among MSM were in the South and West, the top 5 being in North Carolina (peaking at 748/100K), Mississippi, Louisiana, South Carolina, and New Mexico

—  the highest rates of primary and secondary syphilis overall were in Louisiana, California, North Carolina, Nevada, Florida, Arizona, Oregon, Maryland, Illinois, and Mississippi.

— As a point of historical reference, the lowest state specific MSM primary and secondary syphilis rate in 2015 was 73.1/100K in Alaska, surpassing the highest overall US primary and secondary syphilis rates in 1946, at 70.9/100K

— this analysis was limited by a few issues: only 44 states had the sex of sex partners reported for >70% of male cases; the number of MSM in each state was estimated based on surveys and there may be significant underestimation; and the incidence of syphilis infections may be underreported.



Another article looked at 1-dose versus 3-dose regimens of intramuscular benzathine penicillin for early syphilis in patients with HIV (See DOI: 10.1093/cid/ciw862).


— 64 patients were randomized to 2.4 million versus 7.2 million units of intramuscular benzathine penicillin for early syphilis (2.4 million units every week for 3 weeks). The study was from 2009-2013.

–mean age 35, 95% male, 84% MSM, 58% African-American/31% Hispanic/11% white, 6% primary syphilis/61% secondary/33% early latent, 59% had had syphilis before, mean CD4=388/64% on HAART, 49% of those on HAART had undetectable viral loads

— primary syphilis was defined as having compatible genital, anal, or oropharyngeal ulcers; secondary syphilis if they had skin rash or mucosal lesions.  Those with positive serologies (all had positive RPR as well as the more specific TP-PA, T pallidum particle agglutination) were classified as early latent syphilis if they had a documented negative result followed by a positive within 12 months, or at least 4-fold increase in RPR titer.

— median RPR at baseline was 1:128

— RPR and symptoms were monitored every 3 months, and treatment success was defined as at least a fourfold (2 dilution) decrease in RPR during 12 month follow-up.


— only 9 of the 64 patients had seroconversion (negative RPR after treatment), 4 in the 1-shot and 5 in the 3-shot groups

— intention to treat analysis: treatment success rate was 80% in single-dose versus 93% in 3-dose regimens, absolute difference 13%, but not statistically significant.

— Per protocol analysis: success rates were 93% with single-dose and 100% with 3-dose regimens, also not statistically significant.

— no difference by CD4 counts (< 350 vs >350), HIV viral load, use of HAART at baseline, RPR at baseline (<32 vs >=32), or syphilis stage

— only 1 of 20 (5%) patients with undetectable HIV viral load did not achieve treatment success ; whereas 8 of 44 with detectable HIV did not achieve treatment success, a non-significant difference, but there were 6 in the 1-shot group and 2 in the 3-shot group.

— no severe reactions (eg Jarisch-Herxheimer), and none developed neurologic symptoms during the follow-up period.

— They conclude that the current CDC recommendations for a single-dose of benzathine penicillin is reasonable for HIV-infected patients with early syphilis


— the historic concern here was that:

–Treponema pallidum was and is quite susceptible to penicillin

–essentially all adults with early syphilis have their RPR revert to normal after treatment with 2.4 million units of benzocaine penicillin, i.e. a single dose

— but, in those with HIV infection, about one third do not serorevert.

–there seemed to be a higher rate of abnormal CSF findings as well as clinical neurosyphilis in HIV-positive people at earlier stages of syphilis infection (these data predated  HAART therapies).

— However, clinical failure after 1-shot treatment was actually quite rare, though I saw a report of at least one HIV-infected patient who had progression to neurosyphilis after 2.4 M units of benzathine penicillin for early syphilis infection.

–so, many of us, myself included, automatically prescribed 3 weekly doses, or 7.2 million units. However subsequent data on seroreversion were not much better with this higher dose.


–is this study generalizable? And should we just follow the CDC guidelines (ie 1-shot of 2.4 million units of benzathine penicillin)?  There are a few issues:

–really small numbers of patients overall, so above study was underpowered to detect clinically significant differences

–overall a pretty healthy group from CD4 perspective. They did comment that of the 17 patients with CD4 <200 at the time of the syphilis diagnosis, all 11 with 1-shot and 5 of 6 with 3 shots had appropriate serologic response

–but the questions remain: does this apply to those who are more immunocompromized (eg a patient with CD4=50)? or those with detectable viral loads (there was difference noted above, though not statistically significant in this small study)? Does short-term followup of the surrogate marker of RPR titers necessarily correspond to clinical efficacy in the longer-term? Could some of these patients (perhaps with lower CD4 counts) still infect others while their RPR more slowly responds? Should we look at CSF findings (perhaps a better marker of neurosyphilis) and potential long-term neurologic outcomes instead of RPR?


So, my conclusions from this study:

— the big conclusion is that syphilis is increasing in MSM around the country, meaning that people seem to be less protected from the spread of other sexually-transmitted infections as well (eg HIV). And the syphilis/HIV coinfection rates are quite high: reported rates of 15.8% in Los Angeles and up to 47.4% in Philadelphia. Sounds like a potential public health (and individual) disaster waiting to happen…

— in terms of the appropriate treatment for syphilis, my guess is that the CDC guidelines are reasonable. But, given the dearth of clear data, my inclination would still be to use 7.2 million units (3 shots) in those more severely immunocompromised (eg CD4 under 200 or so ????, even more so if nonsuppressed HIV viral load). Would be great to have a larger study with more varied patients (different CD4 counts, viral loads, longer term followup including clinical outcomes, etc)

By submitting your comment you agree to adhere to these terms and conditions
You can follow any responses to this entry through the RSS 2.0 feed.
EBM blog homepage

Evidence-Based Medicine blog

Analysis and discussion of developments in Evidence-Based Medicine Visit site

Creative Comms logo

Latest from Evidence-Based Medicine

Latest from EBM