The method has been commonly used to identify rare drug-resistant variants of the HIV virus.
A new study showed that analyzing plasma — even with precise, deep sequencing technology — is not enough to detect hepatitis C virus (HCV) variants that are drug resistant.
Though direct acting antivirals (DAAs) are the current standard of care for patients with chronic HCV, strains of HCV resistant to DAAs are emerging. In order to find drug-resistant variants of HCV, clinicians and researchers usually turn to population-based Sanger sequencing.
However, the study researchers — led by V. Stalin Raj of the Department of Viroscience at Erasmus Medical Center in Rotterdam, Netherlands — said that standard Sanger sequencing methods miss minority variants. Deep sequencing (DPS) has been used to identify rare, minority, drug-resistant variants of human immunodeficiency virus (HIV).
“In this study, we developed a DPS approach to obtain insight into HCV NS5B viral quasispecies and the presences of drug-resistance associated NS5B variants in the plasma and liver tissue of treatment naive chronic HCV infected patients,” the researchers said.
In order to carry out their investigation, the researchers obtained liver samples from 18 HCV patients who had not received treatment. They also developed a DPS approach to analyze a genome fragment of amino acid positions from 226 to 337 of the NS5B region, as well as methods to account for potential errors.
“The raw sequence data generated in these experiments contained many errors, not uniformly distributed over the amplified region,” researchers noted.
They added that most errors were introduced by DPS and not by reverse transcription. Furthermore, they used the read cleaning approach, with the goal of purging errors that may be caused by DPS.
This method was tested by simulating 2 sets of reads with 454-specific error profiles, and found that before read cleaning, the average error rate of the stimulated reads were 0.77% and 3.2%, respectively, “thereby exceeding the number of errors effectively encountered in the six control experiments.”
After read cleaning, all errors were removed. However, the researchers found that range of error rates increased per nucleotide or amino acids. Similar results were obtained in the analyses of HIV quasispecies. Evaluating each nucleotide or amino acid position separately is necessary to avoid the increased error rates.
“For the analysis of NS5B quasispecies in liver and plasma, haplotypes were reconstructed and based on the control experiments, a conservative cut-off was placed at a haplotype frequency of 1%,” the authors wrote.
Even with taking errors into consideration and a conservative cut-off, the researchers suggested caution in interpreting the data because a liver biopsy obtained by a needle may not be a representative sample of the whole liver.
Previous studies using Sanger sequencing analyses showed that mutations that allow for drug resistance were present in treatment-naive patients at a rate of 0.19-24%. However, in the present study, those mutations were found in 16 out of 18 patients, and were either dominant (50%) or minority haplotypes (39%).
In addition, there were 2 mutations that allow for resistance in the same genomic strand in 4 of the 18 patients. Among those 4, the majority (3) the double mutant haplotype was also the dominant haplotype, the researchers noted.
The researchers found that analyzing the liver and plasma separately revealed there were slightly more resistant variants in the liver than in the plasma, suggesting that plasma is sufficient to detect the drug-resistant variants of the virus.
“Overall, our data thus provide insight into the HCV Ns5B quasispecies population in liver and plasma in treatment naive patients obtained through state-of-the-art sensitive sequencing technologies,” the researchers wrote.
The study, “Identification of HCV Resistant Variants against Direct Acting Antivirals in Plasma and Liver of Treatment Naïve Patients,” was published online in the journal Scientific Reports.