631
Viral Replication Capacity under Varying Adherence-adjusted Cmin Drug Levels Explains Differing NNRTI and PI Adherence/Resistance Relationships
David Bangsberg*1, E Acosta2, R Gupta3, D Guzman1, E Riley1, R Harrigan4, N Parkin5, and S Deeks6
1San Francisco Gen Hosp, Univ of California, US; 2Univ of Alabama at Birmingham, US; 3Harvard Med Sch, Boston , MA, USA; 4BC Ctr for Excellence in HIV/AIDS, Vancouver, Canada; 5Monogram Biosci, South San Francisco, CA, US; and 6San Francisco Gen Hosp, CA, US
Background:
Non-nucleoside reverse transcriptase
inhibitor (NNRTI) resistance is more common than protease inhibitor (PI)
resistance at low levels of adherence. We
hypothesized that differences in PI and NNRTI
resistance at variable drug concentrations is explained by the relative
advantage of resistant and wild type virus to replicate at relevant in vivo drug concentrations.
Methods: Participants were selected from the Research on Access to Care in the Homeless (REACH) cohort
and were included if they were receiving a regimen containing efavirenz (EFV), nevirapine (NVP),
or nelfinavir (NFV). The replicative
capacity of drug-resistant and drug-susceptible variants was determined using a
single-cycle recombinant phenotypic susceptibility assay. The ability of the
patient-derived resistant variant to replicate was compared to the ability of
the drug-susceptible reference to replicate at varying drug levels, and
expressed as a ratio. Drug concentrations used in vitro were adjusted for protein binding to approximate in vivo concentrations. The estimated in vivo level of drug exposure was
estimated using protein-adjusted population-averaged Cmin
(a fixed value for each drug) and multiplied by participant-specific levels of
adherence (0 to 100%).
Results: Of 108 individuals, 54 were on a PI-based
regimen and 54 were on a NNRTI-based regimen. The prevalence of PI resistance
was less common than NNRTI resistance in the lowest quartile of adherence (67% vs 23%, p = 0.05).
Average resistant : wild type replication curves
indicate that the minimum level of adherence required to select for resistance
was 2% for NVP and efavirenz whereas it was 85% for NFV.
Individuals who harbored drug-resistant variants (dark
circles in the figure) were more likely to have levels of drug-exposure where
the resistant variant was more fit than the drug-susceptible variant in vitro. Those who harbored
drug-susceptible virus (light squares in the figure) were more likely to have
levels of drug exposure where the drug-susceptible virus was more fit than the
drug-resistant variant
(p = 0.005). EFV resistant : wild type replication curves (not shown) were
similar to NVP curves.
.
 
Conclusions:
Among individuals with low adherence,
resistance mutations are more common in those treated with NNRTI than in those
treated with PI. These class-specific adherence/resistance relationships appear
to be predicted by the relative replicative capacity of drug-resistant vs wild type
variants to replicate in the presence of clinically relevant drug levels.
|
