Author Affiliations: Vanderbilt Institute for Global Health, Departments of Pediatrics and Medicine, Vanderbilt University School of Medicine, Nashville, Tennessee.
In the July 23, 2012, online and September 10, 2012, print versions of Archives, an impressive French clinical surveillance system has been used to assess the impact of syphilis infection on the viral load (VL) and CD4+ cell counts of human immunodeficiency virus (HIV)–infected men.1 In France, the national health system facilitates high continuity in primary care, reduces fiscal barriers to care, and enables clinical data to be harvested readily for outcomes research. Because this system facilitates the management of sexually transmitted infections (STIs) in the context of HIV primary care, it is markedly easier to examine the impact of coinfection than in systems in which patients are evaluated for STIs in settings outside of their primary care practice.
This study—ANRS CO4—reports data from a well-known hospital cohort of HIV-infected patients, the French Hospital Database on HIV (FHDH). There are 70 hospitals participating in the FHDH, but only the 24 hospitals in the Paris region contributed here. Of the almost 29 000 men followed in the 24 Paris-region HIV clinics, 628 had a first syphilis diagnosis between 1998 and 2006. The investigators conducted a nested case-control study from the cohort of HIV-infected men followed up for all or some of the years 1998 through 2006. All the eligible patients had been followed up for at least 4 years, 2 years before and 2 years after the index period of syphilis diagnosis. In addition, eligible men had at least 2 CD4+ cell counts before, 1 during, and 2 after the index period. Not all the men were taking combination antiretroviral therapy (cART) in the 9-year period.
The 282 study-eligible men diagnosed as having incident primary or secondary syphilis were compared with 1233 matched men who had not had a syphilis diagnosis. Up to 5 controls were selected per case based on the following characteristics: age ±5 years, sexual orientation (men who have sex with men [MSM] vs other), hospital, length of follow-up (±6 months), and virologic/immunologic status just before the date of the syphilis diagnosis, termed the index date. The principal outcome was an increase in HIV RNA VL of at least 0.5 log copies/mL plasma or a rise to greater than 500 copies/mL in patients with previously fully suppressed VL in the 6 months after the index date. The research design was a clever and valid way to study meaningful VL changes in fully virally suppressed men and other HIV-infected men due to syphilis coinfection.
The FHDH–ANRS CO4 protocol team found that 77 men with syphilis (27.3%) and 205 syphilis-free men (16.6%) met the VL increase criteria in the 6 months after the index date (adjusted odds ratio [aOR], 1.87; 95% CI, 1.40-2.49). It was notable that in men with a VL of less than 500 copies/mL before an event who were receiving cART, VL increase was still more common in syphilis coinfected men: 20% vs 14% (aOR, 1.52; 95% CI, 1.02-2.26). A secondary end point was that of CD4+ cell count changes. The CD4+ cell count decreased during the syphilis event period significantly more than in the nonsyphilis group (mean difference of −28/μL, P = .001); the CD4+ cell count rebounded after syphilis therapy (mean difference of 25/μL) such that the cases and controls did not differ after therapy. The FHDH–ANRS CO4 protocol team's documentation that syphilis is associated with a significantly higher likelihood of increased HIV VL is not the first time the observation has been made,2- 5 but it is the largest study by far of the question, and it sends an important message about public health implications of coinfection leading to higher VL and presumed infectiousness in men with HIV infection.
There is evidence that MSM typically reduce their high-risk behavior when they learn of their HIV-seropositive status. The FHDH–ANRS CO4 study did not report a person-time incidence rate, but the 628 cases in 28 956 men during 9 years are likely to calculate to a higher incidence than a recent 27-state US estimate of syphilis incidence trends in MSM, for example.6 It is likely that men with HIV (100% of the men in the FHDH) are at higher risk than are MSM in the US surveillance group, representing most who are not HIV infected, but it would still be worrisome to have higher syphilis rates in men with HIV than in MSM in general.1,6
The STI incidence could be a surrogate measure for comparing success for risk reduction programs across programs and populations in HIV-seropositive persons. The information-motivation-behavioral skills model7 has shown promise for helping HIV-seropositive persons adhere to cART and to reduce risky behaviors that harm themselves (STI acquisition) and others (HIV transmission), but this is not standard practice in most HIV clinical practices. Every case of an STI, such as syphilis, in an HIV-infected person confirms that high-risk sexual practices are occurring and that STIs are being spread, making this a highly specific surrogate for assessing positive prevention program functioning; it is not sensitive, however, and might underestimate failure with a smaller program, that is a program with too few persons to justify the use of STIs as a suitable surrogate of ongoing sexual risk.
This FHDH–ANRS CO4 study confirms the findings of many previous small studies2- 5,8- 10 that show that syphilis coinfection increases VL beyond baseline variation for persons undergoing suppressive cART and for those not receiving such therapy. A 0.5 log copies/mL VL rise may not sound like much in this era of cART, but such a rise, nevertheless, increases infection risk. Note that VL rises were more likely with syphilis infection even when men had been fully suppressed with cART in the 6 months before the syphilis diagnosis. The study documented the rise of VL to above detectable levels in some men with and without syphilis (albeit more often in those with syphilis), suggesting the wisdom of HIV-infected men reducing sexual risk and using condoms even when VL is undetectable; VL may rise unpredictably with coinfections or for other reasons, even when persons are receiving cART.
On a population level, coinfections, such as syphilis, can increase transmission markedly when in many persons with infection, they increase VL and infectious risk.8,9 This FHDH–ANRS CO4 study reminds us that coinfection prevention, detection/treatment, and public health control are valuable tools for decreasing HIV disease progression or transmission, even in the cART era. The STI control campaigns based on revitalized case finding, contact tracing, and primary prevention, for example, with information-motivation-behavioral skills–based counseling, could help and have been successful with MSM populations elsewhere. In fact, aggressive screening, treatment, and contact tracing for many coinfections, such as tuberculosis, STIs, and helminths, along with primary prevention, could help lower HIV VL.10 Reducing VL through coinfection prevention and control, coupled with optimized primary care for millions of infected persons globally, must be considered a viable and vital component for HIV prevention and treatment.11,12
Correspondence: Dr Vermund, Vanderbilt Institute for Global Health, Vanderbilt University School of Medicine, 2525 West End Ave, Ste 750, Nashville, TN 37203 (firstname.lastname@example.org).
Published Online: July 23, 2012. doi:10.1001/archinternmed.2012.3770
Financial Disclosure: None reported.
Vermund SH. Prevention, Detection, and Treatment of Coinfections: A Priority in Human Immunodeficiency Virus–Infected Persons to Reduce Viral Load and Consequent Disease Progression and Transmission Comment on “Effect of Early Syphilis Infection on Plasma Viral Load and CD4 Cell Count in Human Immunodeficiency Virus–Infected Men”. Arch Intern Med. 2012;172(16):1243-1244. doi:10.1001/archinternmed.2012.3770