Cirrhosis, High Age and High Body Mass Index Are Risk Factors for Persisting Advanced Fibrosis After Sustained Virological Response in Chronic Hepatitis C

M. Hedenstierna; A. Nangarhari; A. El-Sabini; O. Weiland; S. Aleman


J Viral Hepat. 2018;25(7):802-810. 

In This Article


Study Population

We identified 402 patients with pretreatment fibrosis stage F3 or F4 and SVR in our treatment records and research databases. Of these, 36 had died, developed HCC or undergone liver transplantation after SVR had been achieved, and they were therefore excluded. Of the 366 remaining patients, 284 accepted to participate in the study, and 269 of these were successfully examined with reliable LSM at a follow–up (Figure 1). Pretreatment fibrosis stage was determined by liver biopsy in 181 patients, by LSM in 81 patients and by a clinical diagnosis of cirrhosis in 7 patients. Of the 269 patients included in the study, 119 (44%) had pretreatment cirrhosis and 158 (59%) were male. Median age at SVR was 53 years (range 18–74). The median follow–up time was 7.7 years, but varied substantially due to the cross–sectional design of the study (range 0–20 years). Follow–up time was <5 years for 115 patients, 5–10 years for 70 patients and >10 years for 84 patients (Table 1). All patients included in this study had been treated with interferon–containing regimens.

Figure 1.

Flow chart for inclusion of patients. Abbreviations: SVR, sustained virological response; HCC, hepatocellular carcinoma; LSM, liver stiffness measurement; FU, follow–up

The study population differed from the patients excluded or lost to follow–up in some respects. Patients not included in the study were to a higher degree male (73% vs 59%) and had higher baseline BMI (median 27 vs 25 kg/m2. The remaining baseline factors, including the proportion with cirrhosis, baseline diabetes mellitus and alcohol abuse, were not significantly different for the two groups.

A history of alcohol abuse was common at baseline (33%), but at follow–up the median reported alcohol consumption was low (18 g/wk), and 110 patients (42%) reported no alcohol consumption during the last year. A minority of patients reported higher alcohol consumption, with 44 patients (17%) and 17 patients (6%) reporting more than 100 or 210 g/wk, respectively. Median BMI at follow–up was 26 kg/m2 (range 17–43), and 45 patients (17%) had a BMI over 30 kg/m2 at the follow–up visit. Diabetes mellitus (DM) had been diagnosed in 35 patients (13%) at follow–up, and seven of these had developed DM after SVR had been achieved.

Fibrosis Improvement After Sustained Virological Response

Median liver stiffness at follow–up was 6.6 kPa (range 2–57 kPa) compared to 13.9 kPa at baseline (range 9.5–74). However, only 81 patients were examined with LSM before treatment, and all of these had a follow–up time of less than 5 years. For patients with longer follow–up time, we only had data on pretreatment METAVIR fibrosis stage. Patients with pretreatment cirrhosis (F4) had significantly higher median liver stiffness at follow–up (8.5 kPa 95% CI 7–9.1) than patients with pretreatment fibrosis stage F3 (6 kPa 95% CI 5.5–6.4). Patients with pretreatment cirrhosis determined by liver biopsy had lower median liver stiffness at follow–up than patients with cirrhosis determined by LSM (7.4 kPa 95% CI 6–8.8 vs 9.4 kPa 95% CI 7.8–11), but this difference was not significant (P = .07). A scatter plot of all individual LSMs by follow–up time and baseline fibrosis stage is presented in Figure 2.

Figure 2.

The distribution of liver stiffness levels at follow–up by baseline fibrosis stage and follow–up time. Regression line over fitted values is shown

A majority (87%) of patients with fibrosis stage F3 at baseline had a liver stiffness below 9.5 kPa at follow–up, and 83% of patients with pretreatment cirrhosis had a liver stiffness below 12.5 kPa, indicating an improved fibrosis stage after achieved SVR. However, 17% of patients with cirrhosis and 13% of patients with fibrosis stage F3 did not improve their fibrosis stage, and 5% had progressed to a more advanced stage at follow–up (Figure 3).

Figure 3.

The distribution of liver fibrosis stages at follow–up by fibrosis stage at baseline. Categories are based on pretreatment cut–offs for METAVIR fibrosis stages

The distribution of liver stiffness values remained similar irrespective of follow–up time for patients with pretreatment fibrosis stage F3. Significant improvement with diminishing liver stiffness with longer follow–up time was, however, seen in patients with pretreatment cirrhosis (Figure 4). For patients with pretreatment cirrhosis, the proportion with a LSM > 9.5 kPa at follow–up thus diminished from 48% for patients with <5 years of follow–up to 36% for patients with 5–10 years of follow–up, and 21% when follow–up time was >10 years (P = .02). In the small subset of patients with follow–up >15 years (n = 11), the median LSM at follow–up was 5.2 (2.7–15.5) kPa and only 9% had a LSM > 9.5 kPa, indicating persisting advanced fibrosis.

Figure 4.

The distribution of liver fibrosis stages at follow–up by fibrosis stage at baseline and 5–y follow–up period. Categories are based on pretreatment cut–offs for METAVIR fibrosis stages. The Kruskal–Wallis rank test is used to test equality of populations

Risk Factors Associated With Persisting Advanced Fibrosis at Follow–up

Overall, 64 of the included patients (24%) had mean LSM levels of ≥9.5 kPa at follow–up, indicating persisting advanced fibrosis. Of these, 19 patients had pretreatment fibrosis stage F3, and 45 patients had pretreatment cirrhosis (F4). Among those with LSM levels ≥ 12.5 kPa at follow–up (n = 27), seven patients had pretreatment F3 fibrosis, and 20 had pretreatment cirrhosis (F4). Median APRI score at follow–up was 0.38 (range 0.19–3.21) for patients with LSM values ≥ 9.5 kPa, and 0.26 (range 0.07–1.01) for patients with LSM values < 9.5 kPa. This difference was significant (P < .001), but the median for both groups was below standard cut–offs used to predict cirrhosis. None of the patients with a LSM below 9.5 kPa at follow–up had an APRI score indicating cirrhosis (≥1.5).

We estimated the risk for persisting advanced fibrosis (defined as ≥9.5 kPa) at follow–up, by calculating odds ratios with logistic regression for patients with different risk factors. Age, alcohol use and BMI were tested both as continuous and dichotomized variables with similar significance levels. Results are presented for the dichotomized variables to make the ORs easier to interpret (Table 2).

In the univariate analysis, pretreatment cirrhosis, age at SVR, BMI and diabetes mellitus at follow–up were significantly associated with a follow–up liver stiffness of ≥9.5 kPa. Alcohol use was not found to be associated with LSM levels ≥ 9.5 kPa at follow–up. However, median alcohol consumption was only 18 g/wk in this study, and 42% of the included patients did not use alcohol at all. Consequently, the influence of alcohol use on liver stiffness was difficult to evaluate in this study.

In the multivariate analysis, pretreatment cirrhosis, age ≥ 55 years at SVR and BMI ≥ 25 kg/m2 at follow–up remained risk factors indicating persisting advanced fibrosis with ORs of 3.9 (95% CI 2.0–7.2), 2.3 (95% CI 1.2–4.3) and 2.3 (95% CI 1.1–4.6), respectively.

No significant interaction was found between risk factors and follow–up time after SVR, suggesting that the different risk factors had the same impact on liver stiffness regardless of follow–up time. Because of this, no separate analysis was performed for the different 5–year follow–up periods.

A small subset of patients with pretreatment F3 fibrosis (n = 7) had a LSM level of ≥12.5 kPa at follow–up, indicating progression to cirrhosis after successful treatment. These seven patients were much more likely to have diabetes (3/7 compared to 7/141, P = .002) than F3 patients that did not progress to cirrhosis after SVR. They also had higher BMI (29 (95% CI 25–33) vs 25 (95% CI 25–26), P = .08) and drank more alcohol (65 g/wk (95% CI 11–119) vs 18 g/wk (95% CI 7–29), P = .09). These differences were not significant, but are still interesting considering the small number of patients. Age at SVR was not significantly different between the two groups: 49 years (95% CI 38–60) vs 52 years (95% CI 50–54), P = .6. Six patients were diagnosed with hepatocellular carcinoma (HCC) after inclusion in this study. Three of these had a LSM value < 9.5 kPa at follow–up. The characteristics of these patients are presented in Table 3.