New-generation therapies for hepatitis C virus (HCV) have been wildly successful – so much so that they've caused a conundrum at market leader Gilead Sciences Inc., whose drugs Harvoni (ledipasvir 90 mg/sofosbuvir 400 mg), Sovaldi (sofosbuvir 400 mg) and Epclusa (sofosbuvir 400 mg/velpatasvir 100 mg) have produced high cure rates in patients with chronic HCV, leading to slumping sales. (See BioWorld Today, May 3, 2017.)

But about 5 percent of HCV patients with decompensated cirrhosis fail to improve or worsen after achieving sustained virologic response (SVR), according to Winston Dunn, associate professor at the University of Kansas Medical Center. Identifying those who likely will be cured by direct-acting antiviral (DAA) agents and those who should be referred directly for liver transplantation could improve patient care while reducing health care costs.

For now, clinicians have no reliable method to separate those two patient populations, given equal severity in baseline disease. But a study by Dunn and colleagues presented at Digestive Disease Week (DDW) 2017 in Chicago suggested that, using genotyping, an answer may be at hand.

Dunn and his team hypothesized that differences in clinical recovery from HCV were associated with genetic variability. They looked at the patatin-like phospholipase 3 (PNPLA3) gene – the most important genetic risk factor for alcoholic liver disease and nonalcoholic fatty liver disease, two of the most severe sequelae of HCV – focusing on the Rs738409 single-nucleotide polymorphism, a variation in a single base pair of DNA in the PNPLA3 gene. Patients possess one of three genotypes – CC, CG or GG – at the PNPLA3 locus. The G allele is associated with a higher risk of worsening disease.

The team followed 32 patients with decompensated cirrhosis who initially achieved SVR using interferon-free DAAs such as Harvoni and the combination of Sovaldi with Olysio (simeprevir, Johnson & Johnson).

Twelve, 24 and 48 weeks following SVR, researchers measured changes in Model for End-Stage Liver Disease (MELD) and Child-Pugh (CPT) scores to assess the severity of chronic liver disease. Five of the 16 patients with the CG or GG genotypes had MELD or CPT scores that worsened by at least one point. In contrast, only one patient with the CC genotype had scores that worsened by at least one point and the remainder of patients in the CC genotype group improved, Dunn reported.

The findings suggest providers could move toward a precision medicine approach to HCV treatment by using a person's genetic makeup to identify those most likely to benefit from HCV treatment, even late in the game, Dunn reported.

"These findings suggest screening for the Rs738409 CG and GG genotypes in hepatitis C patients with decompensated cirrhosis can help to identify individuals who are less likely to recover after achieving a so-called cure of their hepatitis C," he said.

The study was one of the first presented on the opening day of DDW and was featured in a news conference at the meeting.

Genotype offered much greater predictability in MELD score in the study than other variables, such as body mass index, which had no correlation with poor outcomes, according to Dunn. But the findings aren't yet ready for prime time, he cautioned, noting that some patients with the CG or GG genotypes also had improved MELD or CPT scores following HCV treatment.

"The PNPLA3 gene has limited prognostic value when used in isolation," Dunn told BioWorld Today. "It is about the same as having a 1.67 difference in CPT score or 1.70 difference in MELD score at 24 weeks compared to baseline. Therefore, it's not enough to predict clinical recovery by PNPLA3 alone."

The scores took into account the correlation of repeated measurement and adjusting for confounding.

The researchers are seeking to develop other genetic markers that can be used in conjunction with PNPLA3 to develop a genetic signature that could eventually be used in clinical practice to identify which patients are more likely to recover following treatment with DAAs. "We are currently very far from this stage," Dunn acknowledged.

To that end, Dunn and his team next plan to study underlying mechanisms that may explain why the presence of the CG and GG genotypes lead to worsened health outcomes. They'll examine how fatty liver and insulin resistance affect recovery after treatment with DAAs to see if the findings correspond to Rs738409.

Eventually, determining the optimum treatment plan for HCV patients with decompensated cirrhosis could be as simple as collecting genetic material from cheek swabs and sending them for analysis, Dunn said.

DDW continues through Tuesday.