Reply

Abstract

We thank Rodriguez-Frias et al. for their interest in our work and their careful reading of our article. We understand that the authors mistook our criticisms of the prior literature on the topic to be specific to their work, but want to clarify this point. Our sentence “These works however suffered from important methodological flaws, including lack of sensitivity, no consideration of the error rate of the method to establish reliable cutoffs and ensure specificity, too short genomic region analyzed, and/or no linkage studies” refers to three distinct articles, including the one by Rodriguez-Frias et al.1 In this sentence, “and/or” means that the three studies did not all suffer from the same flaws. We found the analysis of ultra-deep pyrosequencing (UDPS) data by Rodriguez-Frias et al. of good quality, and their sensitivity appropriate for this type of study. However, the hepatitis B virus (HBV) reverse transcriptase fragment they analyzed, which spanned positions rt148 to rt208, was far too short to be appropriate for HBV resistance studies, especially in the context of adefovir (or even tenofovir) therapy, which selects variants with substitutions at position rt236 or neighboring positions, as shown in our study. In this context, linkage studies are of no use because mutations that play a key role in the fitness of the variants are not taken into consideration. Additionally, the authors studied only four patients, with a single baseline assessment in three of them. Several timepoints were analyzed in one patient only, who received sequential nucleos(t)ide analog-based therapies, with samples taken at treatment changes only. Overall, the study by Rodriguez-Frias et al. is a valuable pioneer study describing a new, well-handled technical approach, but it suffers from the methodological limitations of a too-short genomic region analyzed that resulted in poorly contributive linkage analyses, and provides limited novel information on HBV resistance to antiviral drugs due to limitations in the number and type of patient samples analyzed. We were a bit confused by the comments on the sensitivity of our analysis, which is in fact similar to that in the PLoS One article (0.03%). The authors question our coverage and sensitivity, and suggest that a variant present as a single viral particle (or as a single sequence?) cannot be reliably detected by our approach. They give the example of a patient with a variant representing 0.34% of 2.1 log IU/mL. First, 0.34% of 4,853 sequences means that 17 identical sequences of very good quality were generated from this single sequence, a high enough number to rely on its presence. Second, this calculation is purely theoretical if one remembers that the HBV DNA “international unit” is an arbitrary unit that in fact represents a much larger number of viral particles. That said, we completely agree with Rodriguez-Frias et al. that UDPS standardization is needed, at both the experimental and analytical levels. We did not discuss these aspects in detail in the article, as our research focus was on using the new technology and our in-house software tools to provide novel information on HBV resistance to nucleotide analogs. We agree that classification of patients as treatment-naïve or treatment-experienced is always challenging, in clinical trials and even more in real life. Primary drug resistance transmitted at the time of infection is also a possibility that would require further investigation. Nevertheless, speculations must be avoided. The patient with 9.57% of resistant variants at baseline was a treatment-naïve pediatric patient included in a clinical trial with adefovir in the United States. Prior exposure is very unlikely, but primary transmitted resistance is possible. The second patient with 11.29% of resistant variants at baseline was a French blood donor who declared no risk factor and no prior therapy for hepatitis B. Prior therapy or primary resistance are possible in this case. Finally, we would like to thank Rodriguez-Frias et al. for their comments and for giving us the opportunity for this interesting discussion. We look forward to reading more from their group on this topic, and to working with them and others to improve standardization of next-generation sequencing strategies in viral hepatitis studies. Christophe Rodriguez, Pharm.D., Ph.D. Stéphane Chevaliez, Pharm.D., Ph.D. Jean-Michel Pawlotsky, M.D., Ph.D. National Reference Center for Viral Hepatitis B, C and Delta, Department of Virology, Hôpital Henri Mondor, Université Paris-Est, Créteil, France; INSERM U955, Créteil, France