Lima VD, Bangsberg DR, Harringan PR, et al. Risk of viral failure declines with duration of suppression of highly active antiretroviral therapy irrespective of adherence level. JAIDS. 2010 Dec 1;55(4):460-5.
Highly active antiretroviral therapy (HAART) is effective at reducing viral load (VL) to undetectable levels for extended periods of time, which improves clinical outcomes.(1) To achieve optimal virologic suppression requires greater than 95% adherence during the first year of treatment.(1, 2) Maintaining such high adherence levels over a lifetime may be difficult to achieve. As the total body burden of virus declines with treatment, the level of HAART required to maintain suppression may be lower than that required to achieve the initial response.(3)
To model the effect of adherence and duration of viral suppression on the risk of viral rebound.
The British Columbia (BC) Centre for Excellence in HIV/AIDS registry; the only distribution source of HAART in BC; HAART-naïve patients who were subsequently treated.
HAART-naïve adult patients (≥18 years) who were subsequently started on two nucleoside/nucleotide analogs plus a non-nucleoside reverse transcriptase inhibitor (NNRTI) or a protease inhibitor boosted with up to 400 mg/day of ritonavir (boosted PI), or a non-boosted PI were included, provided that they had at least one baseline CD4 cell count and VL measurement within six months of initiating treatment. Patients seen between January 1, 2000 and June 30, 2006 were included and followed until June 30, 2007. Consistent with an intention-to-treat approach, changes to the regimens were not included in the analysis. Patients had CD4 and VL monitoring at baseline, four weeks after HAART initiation, and quarterly thereafter. Patients who 1) had scheduled treatment interruptions, 2) moved, 3) died from any cause, 4) were lost to follow-up, or 5) enrolled in a blinded clinical trial that included a placebo, were censored at the last contact prior to any of these events. HIV drug resistance genotyping was done on samples of 250 copies/mL or more collected at baseline and after starting HAART.
Covariates included adherence level, follow-up time with viral suppression, and presence of drug resistance. Adherence levels for each period were defined as the number of days of HAART dispensed, divided by the number of days of follow-up. Adherence was categorized as >40%, 40%-79%, 80%-94%, and 95% and above. Independent predictors of rebound were determined using Generalized Estimating Equations (GEE) assuming a binomial distribution.
A total of 1,305 subjects were included. Eighty-four percent were men, the median age was 42 years, the median CD4 cell count was 130 cells/µl, and the median VL was 5.0 log10 copies/mL. Twenty-two percent of subjects had a history of injection drug use (IDU).
Of 1,305 subjects, 274 (21%) had virologic rebound. The median time of suppression prior to the rebound was two years. Rebound was independently associated with being female, having a history of IDU, initiating treatment in 2000 to 2001 (compared to later years), a treatment without an NNRTI or boosted PI, younger age, and resistance to any class of HAART. Viral rebound was less likely among persons with longer duration of suppression (OR=0.37, 95% CI 0.32-0.42 per year suppressed). Among those with adherence of 95% or greater, the probability of rebound was 0.10 after being suppressed for 12 months and 0.04 after 72 months. Among subjects with adherence between 80 and 95%, this probability was 0.85 after 12 months and 0.8 after 72 months. For persons with adherence less than 40% the probability was 0.68 after 12 months and 0.05 after 72 months. After adjustment for the duration of suppression, persons with less than 95% adherence were 11% more likely to rebound and this risk was higher for persons with less than 40% adherence (OR 1.16, 95% CI 1.12, 1.21). Subjects with drug resistant strains prior to initiating HAART were at increased risk for rebound (OR 2.78, 95% CI 1.68, 4.95).
The risk of viral rebound is reduced with greater duration of viral suppression, within differing levels of adherence.
This was a high quality cohort study. The subjects are likely to be representative of the population to which findings are generalized, the records abstracted were secure and unbiased, the outcome was clearly not present at the start and was measured accurately at follow-up, follow-up time was adequate although studies of longer duration would be useful, and the loss to follow-up was accounted for. However, as noted by the authors, the findings may be due, at least in part, to survivor bias.
This study demonstrates that while adherence is still an important factor in maintaining viral suppression, the duration of viral suppression is an independent factor in sustained viral suppression. This finding is consistent with a cohort study of homeless persons(4) and with induction maintenance studies.(5) Given that the risk of rebound decreases over time, efforts should be made to ensure near perfect adherence, particularly during the early years of treatment. Additional studies that can better measure the degree to which treatment can, if at all, be reduced following sustained viral suppression are warranted.
- Lima VD, Harrigan R, Murray M, et al. Differential impact of adherence on long-term treatment response among naive HIV-infected individuals. AIDS. 2008;22:2371-2380.
- Low-Beer S, Yip B, O'Shaughnessy MV, et al. Adherence to triple therapy and viral load response. J Acquir Immune Defic Syndr. 2000;23:360-361.
- Furtado MR, Callaway DS, Phair JP, et al. Persistence of HIV-1 transcription in peripheral-blood mononuclear cells in patients receiving potent antiretroviral therapy. N Engl J Med. 1999;340:1614-1622.
- Rosenblum M, Deeks SG, van der Laan M, et al. The risk of virologic failure decreases with duration of continuous viral suppression at greater than 50% adherence. PLoS One. 2009;4:e7196.
- Gulick RM, Ribaudo HJ, Shikuma CM, et al. Triple-nucleoside regimens versus efavirenz-containing regimens for the initial treatment of HIV-1 infection. N Engl J Med. 2004;350:1850-1861.