Hepatitis C Virus Eradication With Direct-acting Antiviral Improves Insulin Resistance

Francesco Paolo Russo; Alberto Zanetto; Martina Gambato; Ilaria Bortoluzzi; Ramona Al Zoairy; Enrica Franceschet; Federica De Marchi; Luca Marzi; Erica Nicola Lynch; Annarosa Floreani; Fabio Farinati; Benedikt Schaefer; Patrizia Burra; Heinz Zoller; Andrea Mega


J Viral Hepat. 2020;27(2):188-194. 

In This Article


Study Cohort

Out of 189 patients who were eligible for the study, 138 were enrolled. Patients were excluded for the following reasons: incomplete data regarding glucose metabolism (18), presence of T2DM (23), decompensated T2DM at the time of evaluation (2), concurrent HBV infection (4) and significant alcohol consumption (4). Detailed clinical characteristics of the study cohort are described in Table 1.

Out of 138 patients and according to liver stiffness measurements obtained with the Fibroscan, 95 (68.8%) were cirrhotic, 29 (21%) had severe fibrosis (F3), 9 (6.6%) had moderate fibrosis (F2) and 5 (3.6%) had mild fibrosis (F1). Mean liver stiffness was 19 ± 12 kPa. Mean 'Model for End-Stage Liver Disease' score was 8 (±2.5). Out of 95 patients with cirrhosis, 87 (91.6%) of them were classified as Child A and 8 (8.4%) as Child B. No Child C patients were recruited. Eleven patients with cirrhosis reported oesophageal bleeding in their past medical history, while 4 patients had mild ascites which was controlled by diuretic therapy. Eight patients with cirrhosis (8.4%) had received previous treatment for HCC but were in complete oncological remission when they started DAA therapy.

Hepatitis C virus genotype 1 was the most prevalent (87/138, 63%), and the therapeutic regimen that was most commonly used was the 3D (Ombitasvir, Paritaprevir/Ritonavir and Dasabuvir) combination (55/138, 40%). Thirty per cent (41/138) of patients had previously received IFN-based antiviral therapy (Table 1).

The mean BMI was 25 ± 7 with 2.2% (3/118) of the patients that were obese. Overall, fasting glucose levels were 94 ± 10 mg/dL and insulinemia was 16.4 ± 8.3 mU/mL (HOMA-IR: 3). Sixty-eight per cent of the patients (94/138) were insulin-resistant (IR group), while 32% (44/138) were not (non-IR group). No significant correlation was found between baseline IR and virus-related factors (HCV genotype 3, IL28 polymorphism, HCV-RNA and previous history of antiviral treatment).

IR patients had significantly higher liver stiffness than non-IR patients (23 ± 12 kPa vs 15 ± 8 kPa, respectively; P < .001). Furthermore, the number of patients with cirrhosis was significantly higher in the IR group than in the non-IR group (77 vs 18, P = .003).

The mean BMI was not significantly different in the IR group than in the non-IR group (26 ± 3 vs 25 ± 4, respectively; P = .2).

Changes in Biochemical Results, HCV-RNA and Liver Stiffness During Follow-up

Figure 1 shows patients' follow-up after antiviral therapy. One hundred and twenty-four out of 138 (90%) patients reached the last follow-up (48W after EOT). Seven patients were lost to follow-up before 24W (3 patients changed residence, 2 patients withdrew their informed consent, 1 patient died of sepsis and 1 patient had missing data regarding IR) and 4 patients before 48W (2 withdrew their informed consent, 1 patient had missing data and 1 patient died for pancreatic cancer).

Figure 1.

Flow chart of patients' follow-up. W, weeks after the end of antiviral therapy; pts: patients; IR: insulin resistance; SVR: sustained virological response

Transaminase and bilirubin levels significantly decreased at EOT and throughout the follow-up compared with baseline. Serum albumin significantly increased at 12W and throughout the follow-up compared with baseline (Table 2). HCV-RNA was undetectable in all patients at EOT. SVR12 was achieved in 135 (98%) patients. Patients who did not achieve SVR12 (3) were subsequently excluded from the analysis. At 24W after EOT, TE values significantly decreased compared with baseline (19 ± 12 kPa vs 13 ± 3.5 kPa; P < .001) and remained stable at 48W after EOT (12.7 ± 4 kPa; P = .4) (Table 2).

Changes in Insulin Resistance and Metabolic Parameters During Follow-up

Table 3 shows changes in glucose homeostasis and metabolic parameters over the follow-up, during which no patient changed lifestyle or increased physical activity. Twelve weeks (25 ± 11), twenty-four weeks (26.5 ± 12) and forty-eight weeks (25 ± 13) BMI were not significantly different from baseline (P = .2, P = .4, P = .2, respectively).

At 12W after EOT, 86 (64%) patients and 49 (36%) patients were insulin-resistant and noninsulin-resistant, respectively (Figure 2). The number of patients with IR was lower than baseline, although the difference was not statistically significant (86 [64%] vs 94 [68%]; P = .6). There was a significant reduction of insulin level, which became lower than baseline level (16.4 ± 8.3 vs 13.3 ± 8.7; P < .05). On the contrary, fasting glucose level remained unchanged with no significant difference from baseline (94 ± 10 vs 99 ± 22; P > .05). Therefore, although HOMA-IR decreased (2.4 vs 3), the difference was not statistically significant (P = .08).

Figure 2.

Insulin resistance during follow-up. White: noninsulin resistance group; Grey: insulin resistance group; W: weeks

At 24W after EOT, 72 (56%) patients and 56 (44%) patients were IR and non-IR, respectively (Figure 2). The number of patients with IR was significantly lower than baseline (72 [56%] vs 94 [68%]; P = .04). Insulin level continued to decrease, being the difference between 24W and baseline (12.6 ± 6.9 vs 16.4 ± 8.3; P < .0001) even more pronounced than the one at the previous time point. We did not observe a reduction of blood fasting glucose though, which was even slightly higher than baseline (101 ± 26 vs 94 ± 10; P > .05). However, because of the profound reduction in insulin level, HOMA-IR significantly dropped and became lower than baseline (1.9 vs 3; P < .0001). No significant difference was found between 12W and 24W levels of both fasting glucose and insulin (P = .3 and 0.7, respectively).

At 48W after EOT, 61 (49%) patients and 63 (51%) patients were insulin-resistant and noninsulin-resistant, respectively (Figure 2). The number of patients with IR significantly decreased with respect to baseline (61 [49%] vs 94 [68%]; P = .01) and 24W after EOT (61 [49%] vs 72 [56%]; P = .05). The decreasing trend of insulin was confirmed, with a statistically significant difference with baseline level (11.7 ± 6.3 vs 16.4 ± 8.3; P < .001). No difference was found between 48W and 24W. Fasting glucose blood was 90 ± 29, being not significantly different than baseline. Because of the decreased level of insulin, HOMA-IR was significantly lower than baseline (1.8 vs. 3; P < .001), while no difference was found between 48 and 24W.

Viral- and Host-related Factors Associated With IR Trends During Follow-up

Because of their potential influence on IR, the following factors were analysed: presence of cirrhosis at baseline, genotype 3 and obesity (BMI >30). No difference was found between patients with and without cirrhosis at 12W (P = .7), 24W (P = 1) and 48W (P = 1) after EOT. No difference was found between patients with genotype 3 of HCV vs those with nongenotype 3 infection at 12W (P = .7), 24W (P = .1) and 48W (P = .1) after EOT. Conversely, obesity was associated with a significantly lower probability of achieving a non-IR status at 24W (P = .05) and 48W (P = .04).