Hepatitis C Virus Cure Is the Start of the End for Patients With Advanced Fibrosis/Cirrhosis Free

(See the Brief Report by D�ez et al on pages 2726–9.)

The advent of direct-acting antiviral (DAA) therapy has revolutionized the care for patients with chronic hepatitis C virus (HCV). This is best exemplified by the significant reduction in the number of patients with chronic HCV on United States (US) and European liver transplant waitlists [1]. There is an established reduction in all-cause mortality among patients with DAA-induced sustained virologic response (SVR), specifically in those who do not have decompensated liver disease [2, 3]. Furthermore, the development of de novo hepatocellular carcinoma, a significant cause of morbidity and mortality in chronic hepatitis C, is reduced by DAA-induced SVR [4, 5].

Endpoints evaluated in HCV studies include but are not limited to achievement of SVR, reductions in the hepatic venous pressure gradient (HVPG), improvements in liver stiffness as measured via transient elastography (TE), and changes in surrogate plasma biomarkers. Many believe that improvement in liver stiffness suggests improvement in hepatic fibrosis and its complications. However, this thinking is inherently flawed because reduction in inflammation following SVR may overestimate improvement in fibrosis and provide clinicians with a false sense of security [6].

Measurement of the HVPG is paramount for the determination of the presence of portal hypertension. The HVPG is the difference between the wedged hepatic venous pressure (WHVP) and the free hepatic venous pressure (FHVP). By occluding a catheter proximally in the hepatic vein, the WHVP creates a column of fluid between the hepatic sinusoids (vascular conduits lined with endothelial cells) and the catheter, thereby creating a surrogate for sinusoidal pressure. In sinusoidal causes of liver disease, such as HCV and alcohol, the damaged sinusoids will not allow for equal distribution of the pressure across the liver, thus creating an elevated WHVP. The FHVP is obtained by deflating the balloon more distally in the hepatic vein and is a measure of systemic pressure [7]. Minimal portal hypertension (MPH) exists when the gradient is 6–9 mm Hg. Clinically significant portal hypertension (CSPH) represents a gradient ≥ 10 mm Hg and signifies the critical threshold at which the development of hepatic decompensation including ascites, variceal bleeding, and hepatic encephalopathy may occur [8].

Regarding use of HVPG measurements, there is debate whether a therapy deserves to be discredited if a drug does not elicit transition from CSPH to MPH, whether it is a decrease to MPH that solely matters, or whether changes in HVPG over time also warrant acknowledgement [9]. In this issue of Clinical Infectious Diseases, Diez et al seek to add to the literature by assessing HVPG at a prolonged duration of 48 weeks after DAA-induced SVR and by including patients with human immunodeficiency virus (HIV) coinfection.

All patients in their prospective, multicenter study had an HVPG > 10 mm Hg with a median HVPG of 16.5 mm Hg. Thirty-four patients were included in the final analysis; 17 had decompensated liver disease with a median Model for End-Stage Liver Disease of 12, and 21 patients had concomitant HIV infection (8 decompensated; 13 compensated). Their primary endpoint of reduction of HVPG < 10 mm Hg was not achieved by anyone with decompensated cirrhosis but was met in 6 (18%) patients, all of whom were compensated. In fact, the largest median decrease of 4 mm Hg occurred in HIV/HCV-coinfected patients. Interestingly, 4 patients (11.8%) had a rise in HVPG, only 1 of whom was HIV coinfected and 2 were decompensated; however, potential explanations or further patient characteristics were not provided for these aforementioned findings. However, this finding should serve as a warning to clinicians that worsening liver disease can occur, even after HCV cure.

Their secondary endpoint, which assessed for achievement of decreased pressure measurement among subsets of patients, was met by 44% of patients overall, 52.9% with compensated cirrhosis, and 35.3% with decompensated cirrhosis, but also failed to achieve statistical significance. They report a significant decrease in the median liver stiffness by 7.85 kPa; however, this did not correlate with changes in HVPG. This finding further supports the concept that repeat TE after cure has little value at this time. It was already known that TE is not an accurate estimator of fibrosis after treatment of HCV; therefore, incorporating TE and liver stiffness measurements into the study did not add anything new to the literature. They did note a significant reduction in 4 different plasma biomarkers, but these biomarkers are not used in routine clinical practice. The authors did make a clinically useful note that none of the patients developed worsening liver function or hepatocellular carcinoma during the study period.

Diez and colleagues had a well-designed study that did corroborate recently published literature. Mandorfer et al found that 60% of patients with CSPH had a decrease in HVPG > 10% measured 4 months after non-interferon therapy completion, but overall CSPH still persisted in 76% of patients [10]. While Diez et al state a limitation of a small sample size, Mandorfer had a larger sample size (n = 90) and still showed similar results of 60% of patients vs 52.9% with a 10% HVPG reduction. However, they continued to follow patients for a median of 35 months and at later timepoints demonstrated significant reduction in hepatic decompensation, a clinically meaningful outcome. They also did a subset analysis of HVPG measurements 26 months after therapy completion in 13 (19%) of their patients. Eleven patients (85%) had an HVPG decrease of > 10% to 11.8 mm Hg, but the 2 patients without decrease had experienced hepatic decompensation.

Afdhal et al shared a small, prospective study with paired HVPG measurements measured at baseline, end of treatment, and 48 weeks after sofosbuvir and ribavirin induced SVR [11]. However, a protocol amendment to assess HVPG measurements at 48 weeks after treatment was not incorporated into the protocol by the time most patients completed the study, so only 11 patients had long-term HVPG follow-up at 48 weeks. The mean HVPG at baseline was 17.1 mm Hg, and they found that 9 patients (24%) had a 20% reduction of HVPG during treatment. Of the 11 patients with HVPG follow-up at 48 weeks, 9 patients had a baseline HVPG > 12 mm Hg. Eight patients (89%) had a > 20% reduction in HVPG and 3 patients (33%) had a reduction to < 12 mm Hg. However, further pertinent information as to whether these 3 patients had compensated or decompensated cirrhosis or the exact measurements was not provided [11].

Lens et al had the largest sample size of 226 patients in their 2017 prospective study [12]. They found that despite a 10% reduction in baseline HVPG among 140 patients (62%), 78% of patients overall still had CSPH; however, HVPG was measured just 24 weeks after achieving SVR with oral antiviral regimens. Additionally, they established that changes in TE measurements do not correlate with changes in HVPG [12].

Diez and colleagues’ inclusion of patients with HIV is unique, but the data do not suggest that these patients respond differently than others. Patients with HIV infection may develop portal hypertension from a host of etiologies. Some of these lead to noncirrhotic portal hypertension associated with hepatoportal sclerosis or nodular regenerative hyperplasia. Antiretroviral therapies, particularly zidovudine, didanosine, and stavudine, are established culprits in the development of these entities [13]

In conclusion, Diez et al’s study warns us that although SVR has benefits, cure of the virus does not equal cure of the liver disease. A longer follow-up period may change this view, as true fibrotic regression (not just decreased liver stiffness) could ultimately lead to the ultimate goal, the resolution of liver disease.

Note

Potential conflicts of interest. K. E. S. reports grants from AbbVie, Bristol Myers Squibb, Gilead, and Inovio, and personal fees from MedPace, Watermark (on behalf of Allergan), UniQure, and Inovio, outside the submitted work. A. S. reports no potential conflicts of interest. Both authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.

References