Implications of HCV RNA level at week 4 of direct antiviral treatments for hepatitis C

We aimed to determine whether the HCV viral load after four weeks of treatment (W4VL) with direct‐acting antiviral agents (DAAs) predicts sustained virologic response (SVR) in a real‐world clinical setting. We identified 21 095 patients who initiated DAA‐based antiviral treatment in the national Veterans Affairs (VA) healthcare system from 01/01/2014 to 06/30/2015. Week 4 viral load was categorized as undetectable, detectable below quantification (DBQ), detectable above quantification (DAQ) with viral load ≤42 IU/mL and DAQ with viral load >42 IU/mL. Week 4 viral load was undetectable in 36.1%, detectable below quantification in 45.6%, DAQ ≤42 in 9.3%, DAQ >42 in 9.1%. Detectable above quantification was much more common and undetectable week 4 viral load much less common when tested with the Abbott RealTime HCV assay vs the Roche COBAS AmpliPrep/COBAS TaqMan Version 2 assay. Compared to patients with undetectable week 4 viral load (SVR=93.5%), those with detectable below quantification (SVR=91.8%, adjusted odds ratio [AOR] 0.79, P‐value=.001), DAQ ≤42 (SVR=90.0%, AOR 0.63, P‐value<.001) and DAQ >42 (SVR=86.2%, AOR 0.52, P‐value<.001) had progressively lower likelihood of achieving SVR after adjusting for baseline characteristics and treatment duration. Among genotype 1‐infected patients who were potentially eligible for 8‐week sofosbuvir/ledipasvir monotherapy, we did not find evidence that treatment for 12 weeks instead of 8 weeks was associated with higher SVR, even among those with detectable above quantification. In summary, DBQ and DAQ W4VL are very common in real‐world practice, contrary to what was reported in clinical trials, and strongly predict reduced SVR across genotypes and clinically relevant patient subgroups. Whether and how week 4 viral load results should influence treatment decisions requires further study.