Resolution of clinically significant portal hypertension after sustained virologic response to interferon-free regimens prevents hepatic decompensation

Background and aims:

Sustained virologic response (SVR) to interferon (IFN)-free therapies ameliorates portal hypertension. However, the impact of the hemodynamic response on hepatic decompensation has yet to be investigated in this setting.

Methods:

Seventy-seven patients with portal hypertension (HVPG≥6 mmHg) who underwent hepatic venous pressure gradient (HVPG) and liver stiffness (LS) measurement before (baseline [BL]) and after (follow-up [FU]) IFN-free therapy were retrospectively studied.

(Further) hepatic decompensation was defined by variceal (re-)bleeding, incident ascites/worsening of ascites (requirement of paracentesis), incident hepatic encephalopathy (HE)/worsening of HE (admission for grade 3/4 HE).

Results:

Patient characteristics at BL: Child-Pugh A: 81%/B: 19%; MELD: 8 (interquartile range [IQR]: 2) points; HVPG: 13 (IQR: 8)mmHg.

In the subgroup of patients with a BL-HVPG of 6 – 9 mmHg (n = 22), no patient progressed to clinically significant portal hypertension (CSPH; HVPG ≥10 mmHg). Among patients with BL-CSPH, a HVPG-decrease≥10% was observed in 65%(36/55), while 24%(13/55) had a FU-HVPG< 10 mmHg (i.e., resolved CSPH).

During a median FU of 24.9 (IQR: 15.1) months after the end of HCV treatment, 10 patients developed (further) hepatic decompensation, with variceal bleeding (n = 1), ascites (n = 4), or HE (n = 5) being the first events. Two patients underwent liver transplantation (both had further decompensation) and one patient died (non-liver-related).

In patients with BL-CSPH, a HVPG-decrease≥10% tended to decrease the risk of (further) hepatic decompensation at 2 years (10% vs. 27%; P= 0.072). Overall, the absence of FU-CSPH was fully protective of the development of (further) hepatic decompensation (0% vs. 21%; P= 0.002). This was confirmed in the subgroup of patients with BL-CSPH (P= 0.004).

The area under the receiver operating characteristic curve (AUROC) of FU-LS for diagnosing FU-CSPH/predicting (further) hepatic decompensation was 0.923/0.887. None of the patients with FU-LS values < 12kPa/< 18.9kPa had FU-CSPH/(further) hepatic decompensation, respectively.

Conclusions:

The resolution of CSPH after SVR to IFN-free regimens is associated with a negligible risk of (further) hepatic decompensation, while a HVPG-decrease≥10% is not fully protective. Moreover, FU-LS seems to provide prognostic information.