Effect of HIV-2 infection on HIV-1 disease progression and mortality

Recently, Prince et al.[1] performed a meta-analysis of data from different West African cohorts and concluded that there is no evidence of a difference in mortality rates between HIV-1 single and HIV-1 and HIV-2 dual-infected individuals. Their study raises a number of concerns. Most notably, the meta-analysis was built on data without known HIV-1 infection date and infection order. In addition, most studies recruited hospitalized patients with advanced HIV disease already at enrolment. In an attempt to account for this, the authors performed a subanalysis assuming early disease progression in data from community-based cohorts. However, this assumption was not supported by any clinical or laboratory data. Moreover, they only reported maximum follow-up time of the included studies. It would have been more informative to show the median follow-up (ideally stratified by group). Table 1 shows that all except one of the studies had substantially shorter median follow-up time than the expected survival time in Africa after HIV-1 infection (6–10 years (interquartile range) [2]. One study [3] had longer follow-up (median 10.9 years) but was instead limited by few follow-up visits with long intervals (7–10 years). Also, in many of the studies, dual-infected individuals were older than HIV-1 single-infected individuals at enrolment (up to 16 years). All these potential caveats make the results by Prince et al.[1] difficult to interpret.Table 1: Study characteristics of the six studies used for the comparison between HIV-1 single and HIV-1 and HIV-2 dual-infected individuals in the meta-analysis by Prince et al. [1].We recently demonstrated the importance of infection order by studying a cohort of police officers in Guinea-Bissau [4]. Analysis of seroincident data with long and frequent follow-up showed that dual-infected individuals with an HIV-2 infection preceding the HIV-1 infection on average had a 90% longer progression-time to AIDS than HIV-1 single-infected individuals. Importantly, the results also showed that CD4+ T cells declined at a similar rate, but that the level at the estimated infection date was significantly higher in dual than in single-infected individuals. This indicates that determinants for the difference in progression-time could be related to events during primary HIV-1 infection (PHI), a stage that was not studied in the cohorts included in the meta-analysis by Prince et al.[1]. In these cohorts, the baseline characteristics in terms of disease progression were either similar between the single and dual-infected individuals or not presented. If early events determine differences in disease progression and this is reflected in a higher/lower CD4+ T cell count level after PHI, similar CD4+ T cell count levels and clinics at baseline (often in late-stage disease) would imply a systematic difference between the groups in time since HIV-1 infection. This could explain why Prince et al.[1] did not find any difference in mortality rate. To evaluate the effect of HIV-2 infection on HIV-1 mortality, we analysed a cohort of individuals with known infection date, long follow-up, dense follow-up frequency and known infection order [5]. The survival-time was longer and the mortality rate was lower among dual than among HIV-1 single-infected individuals clearly demonstrating that HIV-2 infection delays the progression of HIV-1 infection. Acknowledgements Conflicts of interest There are no conflicts of interest.