The Risk of Hepatitis C Virus Recurrence in Hepatitis C Virus-infected Patients Treated With Direct-acting Antivirals After Achieving a Sustained Virological Response

A Comprehensive Analysis

Peng Huang; Yan Wang; Ming Yue; Zhijun Ge; Xueshan Xia; Andre J. Jeyarajan; Jacinta A. Holmes; Rongbin Yu; Chuanwu Zhu; Sheng Yang; Wenyu Lin; Raymond T. Chung


Liver International. 2021;41(10):2341-2357. 

In This Article


The flow diagram detailing the database searches and included publications is illustrated in Figure 1. In summary, a total of 2471 references were identified and screened for eligibility. Of these, results were available from 38 studies reporting on recurrence post-SVR. The study characteristics are shown in Table 1.

Figure 1.

Literature screening flow diagram. Low-risk hepatitis C virus (HCV)-mono-infected studies included those examining recurrence among patients with no recognized ongoing risk factors for reinfection (current or former persons who inject drugs, incarceration, men who have sex with men, or active substance abuse). High-risk HCV-mono-infected studies included those studying patients with at least one ongoing identified reinfection risk factor. HIV/HCV coinfected studies included all coinfected participants regardless of risk factors. The total number of studies in the three groups does not equal the total number of studies identified as five studies examined two populations

The demographic and clinical characteristics of the studies are presented in Table S1. A total of 27 061 patients achieved SVR following DAAs therapy, with 15 307 (62.7%) being males, 12 169 (82.9%) Caucasian, 6925 (67.3%) treatment naïve, 18 147 (70.6%) HCV genotype 1, and 5792 (28.4%) with cirrhosis. The median age of the study population was 50 years (range, 34–72).

Five studies evaluated two distinct subgroups of high-risk HCV-mono-infected and HIV/HCV-coinfected patients and were also included in two analysis groups.[14,21,27–29] Ultimately, there were 43 studies included in the analysis. Within the HCV low- and high-risk HCV mono-infection and HIV/HCV coinfection groups, 14/16 (87.5%), 16/19 (84.2%) and 5/8 (62.5%), respectively, were considered high-quality (NOS score ≥ 6). For the low-risk and high-risk HCV mono-infection population groups, funnel plots appeared symmetrical indicating no overt evidence of significant publication bias (Figure S1A,B). However, there was significant publication bias in the HIV/HCV coinfection population group, perhaps due to the small number of studies included. Furthermore, an unexpected shape was observed in the funnel plots and the Egger's test (P < .01; Figure S1C).

Low-risk HCV Mono-infection Population

Sixteen studies were found evaluating the risk of HCV recurrence in low-risk HCV-mono-infected patients.[23,30–44] The recurrence rates per 1000 PYFU and their corresponding 95% CI were calculated for each study and are presented in Table 2. The median follow-up time post-SVR was 21 months (interquartile range, 16–28). Of the included studies, 13 had at least one HCV recurrence. Among those, 10 used genotyping or sequencing to determine recurrence type, one relied on author judgement/terminology, and two did not classify the recurrence.

Sixty-nine HCV recurrences were observed in a total of 16 017 patients. Overall, recurrence rates varied from 0.00/1000 PYFU to 66.93/1000 PYFU among the included studies (Table 2). Consequently, the pooled estimate for the recurrence rate was 0.89/1000 PYFU (95% CI, 0.16–2.03) based on a random effects model and a substantial level of heterogeneity was observed (I 2 = 61%; Table 3).

Of the included 69 HCV recurrences, 36 were assessed as late relapse, and 29 were reinfection, while the recurrence type was not classified in the remaining four patients. The pooled estimate was 0.32/1000 PYFU (95% CI, 0.00–1.40) for late relapse, and 0.06/1000 PYFU (95% CI, 0.00–0.32) for reinfection (Table 3).

High-risk HCV-mono-infected Population

Nineteen studies were identified assessing recurrence in high-risk patients who achieved SVR following DAAs therapy.[14,15,21,27–29,45–57] The median follow-up time post-SVR was 17 months (interquartile range, 9–21). Of those studies, four evaluated the risk in prisoners,[15,49,53,56] one in mono-infected MSM or PWID,[48] and the other primarily in PWID. Among those, 17 studies reported at least one HCV recurrence; nine out of 17 used genotyping or sequencing to determine recurrence type, seven relied on author judgement/terminology/risk factors, and one did not classify the recurrence.

One hundred and seventy-four HCV recurrences were observed in a total of 7932 patients. In total, the recurrence rate varied from 0.00/1000 PYFU to 403.85/1000 PYFU among the studies. The pooled recurrence rate estimate was 29.37/1000 PYFU (95% CI, 15.54–46.91; I 2 = 90%; Table 3). After adjustment for study follow-up duration, meta-regression analysis indicated that this population experienced a 10.79% increased risk of HCV recurrence (relative risk [RR] 1.1079, 95% CI, 1.0324–1.1890; P = .0044) compared to the low-risk HCV-mono-infected population (Table S2).

Of the included 174 HCV recurrences, 156 were reinfection, and none were late relapse; the recurrence type was not classified in the remaining 18 patients. In regard to recurrence type, it appears HCV recurrence was driven by reinfection (25.82/1000 PYFU, 95% CI, 13.08–42.13) rather than late relapse (Table 3).

HIV/HCV Co-infection Population

Eight articles assessed HCV recurrence among HIV/HCV-coinfected patients.[13,16,21,27–29,58] Of those, one was carried out predominantly in MSM,[16] one in targeted PWID,[58] and the remaining studies reported mixed populations. Four studies were retrospective cohorts, while the remainder were prospective cohorts. The median follow-up time post-SVR was 16 months (interquartile range, 11–21). Of the included studies, six studies reported at least one recurrence.

In summary, 46 recurrences were observed in a total of 3112 patients. The pooled recurrence rate was 23.25/1000 PYFU (95% CI, 4.24–53.39; Table 3) and the recurrence rates in individual studies ranged from 0.00/1000 PYFU to 145.45/1000 PYFU, although there was a considerable level of heterogeneity. As in the high-risk HCV mono-infection group, HIV/HCV co-infection was associated with an increased risk of HCV recurrence compared to the low-risk HCV mono-infection population (RR 1.0960, 95% CI, 1.0001–1.2010, P = .0498; Table S2). However, no significant difference was observed between the high-risk HCV-mono-infected population and HIV/HCV-coinfected population (RR 0.9893, 95% CI, 0.9056–1.0807, P = .8106).

By recurrence type, the pooled estimates for late relapse and reinfection were 0.00/1000 PYFU (95% CI, 0.00–0.00) and 23.25/1000 PYFU (95% CI, 4.24–53.39), respectively (Table 3). Furthermore, when combining the HIV/HCV-coinfected patients with high-risk HCV-mono-infected patients, the estimated reinfection risk among PWID following DAA-induced SVR was 18.31/1000 PYFU (95% CI, 9.70–29.16; I 2 = 89%; Figure S3).

HCV Recurrence Following DAAs Therapy Versus IFN Therapy

A previous study reported HCV recurrence following SVR after treatment with IFN-based or DAAs-based therapies.[11] Of the included studies, there were 58 IFN-based and three DAAs-based studies. Using these data, we then demonstrated that recurrence rates following IFN-based therapy were 2.00/1000 PYFU (95% CI, 0.73–3.69), 22.09/1000 PYFU (95% CI, 12.90–33.23) and 30.89/1000 PYFU (95% CI, 0.00–123.23) in low-risk HCV-mono-infected, high-risk HCV-mono-infected and HIV/HCV-coinfected populations, respectively (Figure S4A-C).

In a meta-regression analysis after adjusting for age and study follow-up covariates, DAAs therapy was not associated with lower HCV recurrence in the low-risk HCV-mono-infected population (RR 0.9893, 95% CI 0.9520–1.0280, P = .5813), high-risk HCV-mono-infected population (RR 0.9924, 95% CI 0.9205–1.0700, P = .8439) and HIV/HCV-coinfected population (RR 0.8784, 95% CI, 0.7354–1.0493, P = .1528) when compared to IFN therapy (Table 4). After adjusting for age, study follow-up was negatively correlated with HCV recurrence following HCV 'cure' and the recurrence rate was lower with longer follow-up. The partial correlation coefficients were −0.296, −0.449 and −0.822 in the low-risk HCV-mono-infected, high-risk HCV-mono-infected and HIV/HCV-coinfected populations, respectively (P < .05; Figure 2A-C).

Figure 2.

Hepatitis C virus (HCV) recurrence incidence rates per 1000 person-years. HCV recurrence rates by average follow-up duration in (A) low-risk HCV-mono-infected population; (B) high-risk HCV-mono-infected population and (C) HIV/HCV coinfected populations. HCV recurrence rates were significantly associated with average follow-up duration in three populations of differing risk levels (A, B, C). Partial r and P values were obtained after adjustment for age