Taking between 60% and 90% of prescribed antiretrovirals is the surest way to acquire resistance, according to the results of a large clinical cohort study from the British Columbia Centre for Excellence in HIV/AIDS, which were reported during a late breaker session at the recent IAS Conference on HIV Pathogenesis and Treatment in Paris.
90% adherence is the equivalent of missing no more than five doses of a twice daily regimen in a four week period, or three doses of a once daily drug in a month.
However, although adherence to therapy was strongly associated with the acquisition of new resistant HIV strains, it was not the only factor.
High baseline viral load (a Cox proportional hazard ratio (HR) of 1.5) and, to a lesser extent, low baseline CD4 cell count (HR = 1.1), were also found to have a bearing on the emergence of at least one resistant strain, even adjusting for adherence, suggesting that antiretroviral therapy itself – as least as prescribed in Vancouver between 1996 and 1999 – was not sufficient to control viral replication and therefore to avoid the emergence of some resistance.
Additionally, a significant minority of those who adhered to antiretroviral therapy more than 95% of the time (as measured by prescription refill) were found to have sub-optimal levels of drugs in one (150/1219), or both (70/1219) routine blood samples taken after therapy began. In fact, only 44% of the entire cohort had adequate levels of nonnucleoside reverse transcriptase inhibitors (NNRTIs) or protease inhibitors (PIs), when measured during their first two routine blood tests. Twenty-eight percent did not have adequate drug levels in both tests, and a further 28% only had adequate drug levels in either their first or second test. Around 45% of the people in the latter two groups developed resistance to at least one drug during the study. This suggests that drug interactions and/or absorption issues play an important part in the emergence of resistance.
HIV drug resistance was assessed in every plasma sample with a viral load measurement of over 1000 copies/mL taken from a cohort of 1219 antiretroviral-naïve adults who began triple therapy between August 1996 and September 1999. This resulted in over 3000 separate genotypic tests: a median of two per patient (range 0-13).
During the first 30 months of follow-up, about 30% of the cohort developed some evidence of genotypic resistance, the most common being the M184V mutation, classically associated with lamivudine. Very few people acquired resistance to more than one class of drugs and when it did occur it was mostly amongst people on an NNRTI whose HIV became resistant to both their NNRTI (usually nevirapine) and lamivudine (3TC). However, neither the initial use of an NNRTI versus a PI, nor a history of injection drug use was found to be a statistically significant factor associated with the emergence of resistance in this cohort.
Those who took none or up to 20% of their prescribed drugs had a HR of 1 of acquiring resistance. Those who took 95% or more of their prescribed drugs had a HR of two. Those who took 80-90% of their prescribed drugs had the highest risk, with a HR of 4.
In summary, in multivariate analysis, the strongest associations with HIV drug resistance in this cohort were:
- 60-90% prescription refills
- inconsistent drug levels in those who filled 95% or more of their prescriptions
- high baseline viral load
Harrigan PR et al. The Association between Drug Resistance and Adherence Determined by Two Independent Methods in a Large Cohort of Drug Naïve Indidividuals Starting Triple Therapy. Programme and abstracts of the 2nd IAS Conference on HIV Pathogenesis and Treatment; July 13-16, 2003; Paris, France. Abstract LB12.