Differentiation of acute from chronic hepatitis C virus infection by nonstructural 5B deep sequencing: A population‐level tool for incidence estimation

Abstract

The ability to classify acute versus chronic hepatitis C virus (HCV) infections at the time of diagnosis is desirable to improve the quality of surveillance information. The aim of this study was to differentiate acute from chronic HCV infections utilizing deep sequencing. HCV nonstructural 5B (NS5B) amplicons (n 5 94) were generated from 77 individuals (13 acute and 64 chronic HCV infections) in British Columbia, Canada, with documented seroconversion time frames. Amplicons were deep sequenced and HCV genomic diversity was measured by Shannon entropy (SE) and a single nucleotide variant (SNV) analysis. The relationship between each diversity measure and the estimated days since infection was assessed using linear mixed models, and the ability of each diversity measure to differentiate acute from chronic infections was assessed using generalized estimating equations. Both SE and the SNV diversity measures were significantly different for acute versus chronic infections (P < 0.009). NS5B nucleotide diversity continued to increase for at least 3 years postinfection. Among individuals with the least uncertainty with regard to duration of infection (n 5 39), the area under the receiver operating characteristic curve (AUROC) was high (0.96 for SE; 0.98 for SNV). Although the AUROCs were lower (0.86 for SE; 0.80 for SNV) when data for all individuals were included, they remain sufficiently high for epidemiological purposes. Synonymous mutations were the primary discriminatory variable accounting for over 78% of the measured genetic diversity. Conclusions: NS5B sequence diversity assessed by deep sequencing can differentiate acute from chronic HCV infections and, with further validation, could become a powerful population-level surveillance tool for incidence estimation. (HEPATOLOGY 2015;61:1842-1850 A pproximately 184 million people worldwide have been infected with hepatitis C virus (HCV), 1 including an estimated 138,600-221,300 people in Canada.2 Approximately 75% develop chronic infection and the remaining 25% clear infection spontaneously.3 Without treatment, 15%-25% of those chronically infected will develop progressive liver disease over several decades. Fortunately, HCV can be virologically cured, which reduces allcause and liver-related mortality, 4 and newer, welltolerated direct acting antiviral treatments hold the promise of curing >95% of infections.

Genomic variability of within‐host hepatitis C variants in acute infection

“…With the advent of NGS, the contribution of synonymous and nonsynonymous mutations to viral evolution became more observable but there is very limited data for HCV in this regard. A study that compared the Shannon entropy of the NS5B region between acute (n = 16) and chronic infections (n = 78) with NGS concluded that genomic variability was significantly higher in the latter group and progressively increased over time . Interestingly, synonymous mutations accounted for over 78% of the genetic diversity observed in the total sample with the distinction (synonymous vs nonsynonymous) being more visible for chronic than acute phase samples.…”

“…have been limited to subsegments of the genome-probably due to the technical challenges in generating full-length HCV amplicons in a high throughput and cost-effective manner. 15 Also, obtaining viraemic samples from acute infection time points is logistically difficult as the infection is largely asymptomatic. Finally, most studies todate have not further characterized the observed genomic diversity as synonymous and nonsynonymous mutations.…”

“…Previous studies utilizing NGS to examine viral variant diversity have been limited to subsegments of the genome—probably due to the technical challenges in generating full‐length HCV amplicons in a high throughput and cost‐effective manner . Also, obtaining viraemic samples from acute infection time points is logistically difficult as the infection is largely asymptomatic.…”

Analysis of Heterogeneous Genomic Samples Using Image Normalization and Machine Learning

“…It is usually done by identifying the numerical features that are relevant to the problem under consideration. They can include various diversity measures [24], population genetics parameters [4], physico-chemical properties [21] and other parameters specifically tailored to particular problems. These features are generally identified in consultation with domain experts, and selection of the most relevant features is daunting and resource-consuming task.…”

Evaluation of a recombinant multiepitope antigen for diagnosis of hepatitis C virus: A lower cost alternative for antigen production

“…most of the patients remain undiagnosed until they develop severe liver damage or submitted for serological screening. 3 About 184 million people worldwide are HCV-infected, 4 and early diagnosis and treatment allow virological cure in most of the cases. 5 Because HCV infection is usually asymptomatic, with insidious evolution and epidemic, 6 the screening for this infection can be considered strongly recommended for individuals who have some risk factors, who is blood donor or pregnant.…”