Abstract
Background
Estimates of treatment failure, change and interruption are lacking for individuals treated in early HIV infection.
Methods
Using CASCADE data, we compared the effect of treatment in early infection (within 12 months of seroconversion) with that seen in chronic infection on risk of virological failure, change and interruption. Failure was defined as two subsequent measures of HIV RNA>1,000 copies/ml following suppression (<500 copies/ml), or >500 copies/ml 6 months following initiation. Treatment change and interruption were defined as modification or interruption lasting >1 week. In multivariable competing risks proportional subdistribution hazards models, we adjusted for combination antiretroviral therapy (cART) class, sex, risk group, age, CD4+ T-cell count, HIV RNA and calendar period at treatment initiation.
Results
Of 1,627 individuals initiating cART early (median 1.8 months from seroconversion), 159, 395 and 692 failed, changed and interrupted therapy, respectively. For 2,710 individuals initiating cART in chronic infection (median 35.9 months from seroconversion), the corresponding values were 266, 569 and 597. Adjusted hazard ratios (HRs; 95% CIs) for treatment failure and change were similar between the two treatment groups (0.93 [0.72, 1.20] and 1.06 [0.91, 1.24], respectively). There was an increasing trend in rates of interruption over calendar time for those treated early, and a decreasing trend for those starting treatment in chronic infection. Consequently, compared with chronic infection, treatment interruption was similar for early starters in the early cART period, but the relative hazard increased over calendar time (1.54 [1.33, 1.79] in 2000).
Conclusions
Individuals initiating treatment in early HIV infection are more likely to interrupt treatment than those initiating later. However, rates of failure and treatment change were similar between the two groups.