Efficacy and safety of sofosbuvir in the treatment of hep C among patients on hemodialysis: a systematic review and meta-analysis

Hepatitis C virus (HCV) infection among maintenance hemodialysis patients is implicated in increased morbidity and mortality compared to uninfected patients. Sofosbuvir (SOF)-based regimens may not be optimal among patients requiring hemodialysis. Several studies, however, provide evidence that use of SOF among HCV-positive patients with renal impairment, is effective and safe. We searched Pubmed and Embase to identify studies reporting the efficacy and safety of SOF-based regimens for the treatment of HCV-positive patients on maintenance hemodialysis and performed a random effects meta-analysis. The overall pooled estimate of the efficacy of SOF-based therapy was 95% (95% CI 91–98%). The efficacy of the SOF-based regimen was 92% (95% CI 80–99%), 98% (95% CI 96–100%), and 100% (95% CI 95–100%) for the following doses: 400 mg on alternate days, 400 mg daily, and 200 mg daily, respectively. The most frequent adverse event was fatigue with a pooled prevalence of 16% (95% CI 5–29%), followed by anemia 15% (95% CI 3–31%), and nausea or vomiting 14% (95% CI 4–27%). Anemia was more prevalent in treatment regimens containing ribavirin (46%, 95% CI 33–59%) compared to ribavirin-free regimens (3%, 95% CI 0–9%). This study suggests that SOF-based regimens in the treatment of HCV infection among hemodialysis patients are both effective and safe.


Results
Our systematic search yielded 777 studies of which 60 were eligible for full text review. A total of 20 studies were in accordance with our eligibility criteria and were included in the meta-analysis ( Table 1). The selection process is detailed in Fig. 1.

Discussion
A systematic review and meta-analysis was conducted to determine the safety and efficacy of SOF-based regimens in the treatment of HCV infection among patients requiring maintenance hemodialysis. Among the 934 studies that were identified in our database search, 20 studies met our inclusion criteria. Overall, the pooled estimate of efficacy, based on SVR12, was 95%, ranging from 73 to 100%. Using meta-regression analyses, no statistically significant differences were identified for patient demographics, use of different types of DAA in addition to SOF, or HCV genotypes. Among the studies that reported the dose of SOF, no statistically significant differences in SVR12 were noted with pooled estimates of 92%, 98%, and 100% for SOF doses of 400 mg on alternate days, 400 mg daily and 200 mg daily, respectively. Among the studies that reported adverse events, anemia was the most common with a pooled prevalence of 15%. A total of 10 (50%) studies reported anemia as an adverse event. Among these, 6 (60%) included ribavirin in the SOF-based regimens. Since ribavirin frequently causes anemia, a subset analysis was performed to compare rates of anemia among ribavirin and ribavirin-free regimens 30 . The prevalence of anemia was significantly higher in regimens containing ribavirin (46%) compared to ribavirin-free regimens (3%). Other adverse events identified in this meta-analysis, included fatigue, rash or itching, headaches, insomnia and nausea or vomiting with pooled estimates of 16%, 7%, 7%,3% and 14%, respectively.
Several large-scale studies evaluated the efficacy and safety among patients with HCV infection and severe renal impairment, treated with SOF-based regimens 8,9 . A longitudinal prospective study by Saxena et al. compared outcomes of HCV infection between 1716 pts with estimated glomerular filtration rate (eGFR) > 45 ml/ min/1.73 m 2 to 73 patients with an eGFR ≤ 45, of which five were on hemodialysis. Patients were treated with SOF 400 mg daily and 60% also received ribavirin. Among patients with eGFR ≤ 45, SVR12 was achieved among 83% of patients. Reported rates of adverse events, in this subset of patients, were similar to our study, with anemia, fatigue and headaches occurring in 30%, 30% and 14% of patients, respectively. Rash or itching was not . Their outcome of efficacy combined SVR at 12 and 24 weeks. Twenty-one studies were included, with a total of 717 HCV infected patients, of whom 306 (58.4%) were on either hemodialysis or peritoneal dialysis. Doses of SOF varied and a subset also received ribavirin. Among patients requiring dialysis, the pooled SVR12/24 was 95%. Rates of anemia, fatigue, headache, rash or itching were not reported 8 . Lastly, in November 2019, the U.S. Food and Drug Administration (FDA) updated the prescribing of several SOF-based regimens, which now states that no dosage adjustment is required for patients with any degree of renal impairment. In the updated labeling report for sofosbuvir/velpatasvir, for example, the FDA reported an overall SVR of 95% among 59 patients requiring hemodialysis and peritoneal dialysis. Adverse events were not provided 31 . Our findings support the notion that the use of Sofosbuvir based regimens in patients receiving hemodialysis is both safe and effective. In our analysis, hemodialysis did not www.nature.com/scientificreports/ seem to be an unfavorable factor, in terms of safety and efficacy, when compared to other meta-analyses which included Hepatitis C patients without renal disease 32,33 .
There are several limitations that warrant discussion. First, substantial heterogeneity was found among studies, which we addressed by using a random effects model instead of a fixed effects model. Furthermore, metaregression analysis of specific variables, including patient demographics, HCV genotype and SOF dose did not identify any statistically significant differences. Second, although some publications reported lower SVR rates based on different SOF-based regimens, a subset analyses of types of SOF-based regimens and a comparison of NS5A and NS3 protease inhibitor regimens could not be performed due to the small number of studies in each group. Future studies will need to compare differences in efficacy and safety among SOF-based regimens using different DAA. Third, efficacy by specific HCV genotype could not be determined, again due to the small number of patients and that data pertaining to specific genotypes and outcomes were not extractable.
In conclusion, our findings support that the use of SOF-based regimens in the treatment of HCV infection among patients requiring hemodialysis is both safe and effective. Future randomized controlled studies are needed to confirm these results and to evaluate the optimal SOF-based treatment regimen.

Methods
This systematic review and meta-analysis was performed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Statement (PRISMA) 34 and has been registered in PROSPERO (CRD42018116632, https ://www.crd.york.ac.uk/PROSP ERO).

Search strategy. A systematic search of Pubmed and Embase databases was performed to identify studies
reporting the efficacy and safety of SOF-based regimens for the treatment of HCV positive patients on maintenance hemodialysis. The following search term was used: (end stage renal disease OR renal replacement therapy OR chronic kidney failure OR severe renal impairment OR chronic kidney disease OR hemodialysis) AND (sofosbuvir). Results were limited to the English language and publication dates until May 22, 2020. Additionally, references of systematic reviews focusing on HCV treatment were reviewed manually.

Eligibility criteria and study selection. Two reviewers (KB and EMCD) independently screened and
evaluated the studies that were potentially eligible for inclusion with discrepancies were resolved either between the two reviewers or a third party (FS), if consensus could not be reached. Studies were included if they reported extractable data on the number of maintenance hemodialysis patients that completed a SOF-based regimen and had values for a sustained virologic response 12 weeks after the completion of treatment (SVR12). Our primary outcome of interest was the efficacy of SOF-based therapy in HCV positive patients on maintenance hemodialysis. Regimen efficacy was defined as the proportion of patients that had SVR12. As a secondary outcome of interest, we evaluated the safety of SOF-based regimens, by quantifying the proportion of patients with reported adverse events.
Studies that reported only aggregate data on patients, regardless of their dependency on hemodialysis, were excluded as were opinion pieces and conference abstracts. Studies that focused on patients < 18 years old, HIV positive patients or transplant recipients were also excluded. SOF-based regimens that included pegylated interferon were also excluded. We assessed the quality of the eligible studies using the Newcastle-Ottawa Scale 35 . A maximum of six stars could be assigned to each study, as the parameters "selection of the non-exposed cohort" and "comparability between cohorts" were not applicable to our analysis. High quality studies were defined as studies that received ≥ 5 stars (Supplementary Table 1).
Data synthesis and analysis. We performed a random effects meta-analysis to estimate the pooled efficacy of SOF-based therapy in HCV positive patients on dialysis, using the DerSimonian and Laird approach 36 . The Freeman Tukey double arcsine transformation was used to stabilize the variances 37 . We selected a random www.nature.com/scientificreports/ effects model, because we assumed that the effects are heterogeneous due to differences in the study design, SOF dosage and combination drugs of each study. Studies were grouped by regimen dosage where possible, and a random effects meta-analysis was carried out in these sub-groups. We assessed the heterogeneity among studies and subgroups using the I 2 statistic 38 . The Egger's test was used to explore publication bias and small study effects 39 .
To evaluate the safety of the SOF-based regimens, we performed a random-effects meta-analysis for each adverse event reported in 3 or more studies, and estimated the pooled proportion of patients that reported each adverse event. A meta-regression analysis was also conducted to investigate the extent to which the differences in study characteristics were correlated with the between study heterogeneity 40 . For our review, we used an intention to treat analysis. Patients lost to follow-up were considered as treatment failures.
Stata v15 (Stata Corporation, College Station, TX, USA) was used to perform the statistical analysis. The statistical significance threshold was set at 0.05.

Data availability
All data generated or analyzed during this study are included in this published article (and its Supplementary Information files).

Figure 5.
Forest plot of included studies that reported anemia as an adverse event. Individual and combined estimates of prevalence of each adverse event with 95% confidence intervals. ES effect size (prevalence).