Early Detection of Hepatitis E Virus Ribavirin Resistance Using Next-Generation Sequencing

LETTER

Treating chronic hepatitis E virus (HEV) infections with ribavirin (RBV) is a recommended option (1).

With great interest, we read the article of Lhomme and coworkers, who described the influence of the HEV variant G1634R to RBV treatment in immunocompromised patients. The presence of G1634R in the individuals assessed by Lhomme et al. did not lead to absolute RBV resistance and did not compromise the response to a second RBV treatment (2). The role of G1634R and other variants for clinical HEV resistance against RBV is under discussion (3–6). Here, we report a longitudinal observation of clinical RBV-induced mutations associated with increasing selection of the variant G1634R in an immunocompromised patient.

Our 62-year-old male patient had undergone allogeneic stem cell transplantation (aSCT) in 2012 for primary myelofibrosis. During 4 years, he received a combination of prednisolone, cyclosporine, and mycophenolic acid for suspected chronic graft-versus-host disease. An HEV infection with genotype 3c and HEV RNA level of 5,570,000 IU/ml was first diagnosed in May 2016.

Oral treatment with RBV (1,200 mg/day) was conducted for 3 months. Subsequently, alanine transaminase (ALT) levels normalized, and HEV RNA decreased rapidly to undetectable levels within 4 weeks (Fig. 1). RBV treatment was stopped after 3 months, as previously proposed (7). Unexpectedly, shortly thereafter, HEV RNA reappeared, and ALT levels reincreased. Therefore, we decided to reinitiate RBV treatment.

• Open in new tab
• Download powerpoint

FIG 1

Courses of HEV RNA in blood serum and stool, HEV variants detected in serum, and ALT levels before and during ribavirin treatment in an immunosuppressed patient. Amino acids are displayed in green for the wild type and in red for resistance-associated variants and are given in ratios for next-generation sequencing (NGS) results. The percentage refers to the mutation. HEV RNA, hepatitis E virus RNA (continued line, lower limit of detection 250 IU/ml in serum; dots, lower limit of detection 250 IU/ml in stool); RBV, ribavirin; ALT, alanine transaminase; dashed line, upper limit of normal of alanine transaminase.

Consecutively, HEV RNA became again undetectable, and ALT levels normalized (Fig. 1). However, as HEV reappeared during treatment, we decided to discontinue RBV to prevent a possible further selection of resistant variants. Instead, we reduced the immunosuppression, which was followed by a durable decrease in HEV RNA levels. ALT levels temporarily increased after resolution of the HEV infection, which we interpreted as a sign of immune reconstitution.

Using Sanger sequencing of the RNA-dependent RNA polymerase (RdRp) region of the HEV genome (open reading frame 1 [ORF1]), the HEV wild type was found at baseline. However, after the first rebound, the RBV-induced mutations K1383N, G1634R Y1320H, D1384G, K1398R, V1479I, and Y1587 could be detected (Fig. 1). After the second rebound and thereafter, RBV-induced mutations showed increasing frequency. To investigate a possible presence of RBV-induced mutations before exposure to RBV, we sequenced the HEV genome from the baseline sample on the MiSeq platform (Illumina, San Diego, CA, USA). Intriguingly, by this approach, we could detect the K1383N and G1634R variants even before the initiation of RBV treatment (Fig. 1; for methods, see the supplemental material).

In accordance with Lhomme and colleagues, we feel that the significance of the presence of RBV-induced mutations before treatment is uncertain. However, we could show that treatment monitoring may reveal a selection of those variants and allow the early prediction of clinical resistance. It needs to be further investigated if there is a threshold of RBV-resistant variants that make a viral breakthrough more likely, and if those patients may have stronger benefit from other treatment options, e.g., modification of the immune suppression or sofosbuvir treatment, although this treatment is to date not evidence based.