syphilis treatment in HIV positive patients

a recent article assessed retrospectively the utility of a single shot of benzathine penicillin 2.4 M units (BPG) for the treatment of early syphilis in hiv positive patients (see hiv early syphilis treatment CID 2015 in dropbox, or DOI: 10.1093/cid/ciu888​), as is recommended by the CDC.

baseline:

--US Military HIV Natural History Study cohort of 350 people with 478 cases of early syphilis diagnosed between 2002-2009 (median year 2006) 

--median age 28, 22% white/63% African American, 99% male, and median CD4 count at HIV diagnosis 506 (348-649) and at first syphilis case of 483 (376-636), 59% on HAART at syphilis diagnosis

--30% were treated with one dose of BPG and 52% with >= 2 doses

--early syphilis was defined as converting to a positive nontreponemal test, confirmed by treponemal test, within 365 days of a negative test

--response to treatment at 13 months was defined as a >= 4-fold decline in nontreponemal titers

results:

--91% responded to treatment by 13 months, 97% by 24 months

--median time to treatment response was 5 months

--no difference in response to treatment by numbers of BPG shots

--multivariate analysis: older age was associated with delayed response; higher pretreatment nontreponemal titer and higher CD4 at time of syphilis diagnosis was associated with shorter time to treatment response. HAART use and HIV viral load were not related to response

the authors note that the CDC recommendations are based on pretty poor data, and that there are no good randomized controlled trials on the question of the best treatment regimen.

so, I remain unconvinced that a single BPG injection should be standard therapy, for the following reasons.

--this is a retrospective analysis. were there clinical differences between the patients that led the treating provider to choose 1 vs 2+ injections of BPG?? were those given only one shot fundamentally different/healthier than those given more?

--this group had largely preserved immunologic function, with >50% having CD4 counts of >500, the lowest CD4 count being 348, and had a low viral load (median log of 2.4 copies/ml). and even with that, the study did find that those with lower CD4 counts in this pretty healthy group took longer to respond to treatment. what if their CD4 count were 58?? would one shot work???

--​there were several case reports awhile ago (probably 20+ years and, as i recollect, prior to HAART) of very rapid progression from syphilis infection to clinical neurosyphilis within 1-2 years in patients treated with only 1-2 shots of BPG (I did find one reference for rapid development of neurosyphilis: Ann Intern Med. 1990;113(11):872). as a result of these early reports, i have been using 3 shots (the same as required for late syphilis) for the few cases i have seen​

--the reason I bring this up is, first, that there has been a significant increase in syphilis cases over the last several years, and, second, that there seems to be a lot of chatter on the internet about how we can now use one one shot of BPG to treat early syphilis in HIV patients, based on the above article. and, it may be true that one shot is sufficient, especially in the era of HAART, but i do not think this article proves it.....