HIV-1 genotypic and phenotypic susceptibility testing (GPT) optimizes antiretroviral selection, but its effect on survival is unknown, write Frank J. Palella Jr, MD, from Northwestern University Feinberg School of Medicine at Chicago in Illinois, and colleagues from the HIV Outpatient Study (HOPS) Investigators. In our analyses, we sought to evaluate the effect of GPT on survival among a diverse sample of HIV-infected patients with varied antiretroviral treatment histories and therapeutic responses who have participated in the HOPS.
At 10 US HIV clinics, from 1999 to 2005, 2699 HIV-infected patients with plasma HIV RNA levels higher than 1000 copies/mL were eligible for GPT who were followed-up for a median of 3.3 years. Recorded data consisted of clinical factors and demographic, use of GPT, all-cause mortality, and crude and adjusted hazard ratios (HRs) for the association of GPT with survival.
915 patients (34%) of 2699 patients had GPT. Patients who did have GPT had lower mortality rates (2.0 vs 2.7 deaths per 100 person-years) when compared to patients who did not have GPT. After adjusting for demographic factors, HIV RNA level, CD4+ cell count and intensity of clinical follow-up, GPT was associated with improved survival in standard Cox models (adjusted HR, 0.69; 95% confidence interval [CI], 0.51 – 0.94; P = .017).
According to the results of subgroup analyses (2.2 vs 3.2 deaths per 100 person-years for patients who had GPT vs those who did not; adjusted HR, 0.60; 95% CI, 0.43 – 0.82; P = .002), GPT was associated with improved survival among HAART-experienced patients (n = 2107). GPT was also linked to improved survival among triple antiretroviral class–experienced patients (n = 921; 2.1 vs 3.1 deaths per 100 person-years; adjusted HR, 0.61; 95% CI, 0.40 – 0.93; P = .022).
GPT appeared to be linked to improved survival in the overall cohort (adjusted HR, 0.54; P = .001) and in the HAART-experienced group (adjusted HR, 0.56; P = .003), based on marginal structural models.
Use of GPT was independently associated with improved survival among HAART-experienced patients, the study authors write. We also found reduced mortality among patients who had genotypic testing alone or phenotypic testing alone.
The study limitations include its observational design, inconsistent measurement of clinical study variables, lack of randomization for use of GPT, missing data and possible residual confounding.
We believe that we have provided an important first step in elucidating a possible causal relation between GPT and better survival, the study authors conclude. Future work is needed to replicate our findings, elucidate further on patient subgroups for whom GPT has a particularly beneficial effect on survival, and determine the pathways through which GPT’s posited effect is mediated.
