Reply

Authors


  • Potential conflict of interest: Dr. Thomas advises and received grants from Merck. He also received grants from Gilead.

Reply:

We thank Drs. Cobb and Vilchez for their thoughtful comments and for highlighting an important issue related to polymerase chain reaction (PCR)-based testing in the clinical setting. Hepatitis C virus (HCV) RNA testing plays an important role in determining appropriate duration of therapy, both in terms of candidacy for response-guided therapy as well as for implementing treatment futility rules. However, despite the critical importance of HCV RNA quantification, there remains considerable uncertainty as to the most appropriate quantitative/qualitative levels that should trigger these clinical decisions. The Roche Cobas Taqman HCV 2.0 assay with a lower limit of detection of 9.3-10 IU/mL was used in both the boceprevir and telaprevir phase 3 trials. 1, 2 Unlike a qualitative assay, the quantitative assay has a lower limit of quantification (LLOQ) as well as a lower limit of detection (LOD). Thus, there are three possible results reportable with this assay: quantifiable (>25 IU/mL), detected but not quantifiable (<25), and not detected (<10 IU/mL). The important point for the healthcare provider is that a detected but not quantifiable result should not be equated with an undetectable result when making decisions about response-guided therapy. Detectable but below levels of quantification was a frequent observation at critical treatment decision points for response-guided therapy in the registration trials. A recent Food and Drug Administration (FDA) analysis suggested that up to 17% of boceprevir-treated patients at week 8 and up to 28% of telaprevir-treated patients at week 4 fit into this category and that these patients had significantly lower sustained virological response (SVR) than those whose result was not detected (5%-20% lower SVR mainly due to increased relapse rates). 3 Thus, there appears to be a true difference in SVR rates among patients with these different virological test results and the FDA has recommended that undetectability be used for response-guided therapy decision points. We agree with the authors that to avoid confusion between limit of quantification (<25 IU/mL), detected and undetectable, the term HCV RNA not detected (=target not detected) rather than “undetectable” should be used when reporting test results. Given the clinical importance of this issue, a consensus agreement needs to be reached on the terminology used for reporting HCV results. Additional research including distribution of blind panels may be justified to establish the equivalence of various assays run in different laboratories and to determine the optimal cutoff that accurately predicts SVR. Until such time as more data become available, we advise continuing to follow the recommendation in the FDA label to use a sensitive real-time reverse-transcription PCR (RT-PCR) assay for monitoring HCV RNA levels during treatment. The assay should have a lower limit of HCV RNA quantification of equal to or less than 25 IU/mL, and a limit of HCV RNA detection of approximately 10-15 IU/mL.

We thank Dr. Milazzo and colleagues for their interesting comment. The authors propose that the week 4 timepoint of lead-in therapy during boceprevir treatment could be used to determine the need for triple therapy versus combination peginterferon plus ribavirin alone because of the high SVR rates, ranging from 97% in the peginterferon plus ribavirin arm to 89% and 90% in the two boceprevir arms (RGT and fixed duration therapy, respectively) in patients who achieve a rapid virological response. Although this is a valid observation, it cannot be recommended in clinical practice at this time because such an approach was not formally tested and the results were derived from a subanalysis of the main trial. Moreover, only a small number of patients would qualify for this approach, −6% of treatment-naïve genotype 1 patients in the current boceprevir trial; 17% to 26% from other historical reports. 6-8 Interleukin (IL)-28B is unlikely to improve SVR prediction in this setting, as demonstrated in a recent analysis suggesting that a rapid virologic response (RVR) was more predictive of SVR than baseline IL-28B genotype. 9 While we agree this might be an interesting approach to limit exposure to triple therapy, the future availability of potent antiviral agents without the need for interferon (and perhaps ribavirin) will permit even shorter courses of therapy and make the suggested approach redundant.

Marc G. Ghany M.D., M.H.Sc.*, David R. Nelson M.D.†, Doris B. Strader M.D.‡, David L. Thomas M.D.§, Leonard B. Seeff M.D.¶, * National Institutes of Health, Liver Diseases Branch, NIDDK, NIH, Bethesda, MD, † University of Florida, Section of Hepatobiliary Disease, Gainesville, FL, ‡ University of Vermont College of Medicine, Burlington, VT, § Johns Hopkins Medical Institution, Baltimore, MD, ¶ The Hill Group, Bethesda, MD.

Ancillary