Evaluation of the Safety Profile of Direct-Acting Antivirals on Patients with Hepatitis C Virus: A Pharmacovigilance Study

Background Hepatitis C virus (HCV) is the primary contributor to chronic hepatic diseases. A rapid change in the situation took place with the advent of oral direct-acting antivirals (DAAs). However, a comprehensive review of the adverse event (AE) profile of the DAAs is lacking. This cross-sectional study aimed to analyze the reported Adverse Drug Reactions (ADRs) with DAA treatment using data from VigiBase, the WHO Individual Case Safety Report (ICSR) database. Methods All ICSRs reported to VigiBase with sofosbuvir (SOF), daclatasvir (DCV), sofosbuvir /ledipasvir (SOF/LDV) and ombitasvir/paritaprevir/ritonavir (OBV/PTV/r) in Egypt were extracted. Descriptive analysis was performed to summarize patients’ and reactions’ characteristics. Information components (ICs) and proportional reporting ratios (PRRs) for all reported ADRs were calculated to identify signals of disproportionate reporting. Logistic regression analysis was performed to identify the DAAs association with serious events of concern while adjusting for age, gender, pre-existing cirrhosis, and ribavirin use. Results Out of 2925 reports, 1131 (38.6%) were serious. The most commonly reported reactions; anaemia (21.3%), HCV relapse (14.5%) and headache (14%). For the disproportionality signals; HCV relapse was reported with SOF/DCV (IC 3.65, 95% CrI 3.47–3.79) and SOF/RBV (IC 3.69, 95% CrI 3.37–3.92), while anaemia (IC 2.85, 95% CrI 2.26–3.27) and renal impairment (IC 2.12, 95% CrI 0.7–3.03) were reported with OBV/PTV/r. Conclusion The highest severity index and seriousness were reported with SOF/RBV regimen. A significant association was found for OBV/PTV/r with renal impairment and anaemia although being the superior regimen in terms of efficacy. The study findings call for further population-based studies for clinical validation.


Introduction:
Hepatitis C virus (HCV) infection is a major cause of morbidity and mortality [1]. About 58 million people worldwide are diagnosed with hepatitis C and nearly 1.5 million infections occur per year [2,3]. Approximately 290,000 people died from hepatitis C in 2019 [2].
Egypt has one of the highest global burdens of hepatitis C virus (HCV) infection [4]. High rates of HCV transmission in Egypt were associated with the nationwide parenteral anti-schistosomal therapy campaigns from the 1950s to the 1980s, due to a series of unsafe intravenous injections [1,5].
However, a rapid change in the situation took place in Egypt [11] after the conduction of multiple national programs to screen and treat HCV. A lower rate of HCV prevalence of 4.6% was reported in a large-scale national study conducted during 2018-2019 [4,7].
Few serious adverse drug reactions (ADRs) associated with DAAs were observed and reported in the clinical trials performed on the DAAs, however, clinical trials are limited in their ability to detect ADRs that occur in a 'real-world' setting and a broad review of the adverse event (AE) profile with DAAs is lacking [14]. Therefore, post-marketing surveillance through pharmacovigilance activities-as spontaneous reporting of ADRs-is essential to identify potential safety and efficacy concerns [13,15].
Although over five million patients were treated with DAAs globally [13] and more than two million patients were treated in Egypt by 2018 [7], there are no large data-based studies with an extended period were conducted in Egypt to detect potential associations between reported adverse events and DAA drugs.
This pharmacovigilance study using VigiBase-the World Health Organization global database of individual case safety reports (ICSRs)-aims to evaluate the safety profile of the direct-acting antivirals (DAAs) regimens (SOF + RBV, SOF + DCV, SOF/LDV, and OBV/PTV/r) on the local population in Egypt through describing the pattern of adverse events, and detecting the signals of disproportionate reporting (SDRs)-statistical associations between medicinal products and adverse events- [16,17].

Study Design and Data Source
This is a cross-sectional pharmacovigilance study conducted using data from VigiBase-developed and maintained by Uppsala Monitoring Centre (UMC)-which contains about 29,699,362 reports received from around 140 countries. Vig-iBase receives reports of ADRs occurred with medicinal products from the national centres for pharmacovigilance in the participating countries [18,19].
The information in VigiBase comes from different sources and the probability that the suspected adverse effect is drug-related is not the same in all cases. Study results and conclusions do not represent the opinion of the UMC or the World Health Organization [20].

Ethics Approval
The data used in the study were anonymized and there was no direct interaction with human participants. Approval to conduct this study and extract data from VigiBase was obtained from the Egyptian Drug Authority (EDA). The study protocol and methodology were revised and approved by the Committee of Ethics, Faculty of Pharmacy, Ain Shams University (approval number: 236 dated 13-03-2019).

Data Extraction and Analysis
For this study, all ICSRs with designated direct-acting antivirals (DAAs) reported in Egypt since the introduction of DAA regimens in 2013 till the 7th of September 2021 (data extraction date) were selected. All reports with the following suspected drugs were included: sofosbuvir (SOF), daclatasvir (DCV), sofobuvir/ledipasvir (SOF/ LDV), and ombitasvir/paritaprevir/ritonavir (OBV/PTV/r) for the treatment of Hepatitis C Virus (HCV).
The data sets were extracted from VigiBase to Microsoft Excel files using the standard terminology in VigiBase as follows; Anatomical Therapeutic Chemical (ATC) classification of medications, the active substance of the medicines, according to WHO Drug dictionary for medicinal products and the Preferred Term (PT) level in the Medical Dictionary for Regulatory Activities (MedDRA) for coding the ADRs.
The extracted data were then reviewed to exclude reports with concomitant interferon (IFN), and reports that included SOF alone or DCV alone as they are not approved as monotherapies. Moreover, a case-by-case review was performed to obtain additional data pertaining to the severity assessment of the ADR and the logistic regression analysis.
The reports were then analysed based on age, gender, causality, System Organ Class (SOC) and Preferred Term (PT) levels of the ADR, ADR seriousness and type of seriousness, and ADR severity.
Severity assessment of the ADR outcome was performed and classified utilizing Hartwig's severity assessment scale, which is divided into seven levels [21][22][23][24].
Any ADR that was associated with death, life-threatening condition, hospital admission or prolongation of hospitalization, persistent disability, congenital anomaly and any medically important event or condition that was considered clinically critical by the reporting health care professional, is considered a serious reaction according to the ICH definition [25].
Causality was defined as probability that the drug caused the adverse event [26]; which differs between case reports [20], so causality was assessed and categorized by the reporter according to WHO-UMC criteria [22,27].

Statistical Analysis
The statistical analysis was performed using R software (v. 4.1.1). P-values less than 0.05 were considered to represent statistically significant results [28].

Descriptive Analysis
Descriptive statistics for patients' characteristics, system affected (SOC), reaction PT, seriousness and type of seriousness, causality, and outcome severity all were presented as frequencies and percentages. Distributions of the aforementioned data were compared among regimens using the χ 2 or Fisher's exact tests.

Disproportionality Analysis
Disproportionality analysis -which is a tool for signal identification by describing the extent to which the reported adverse drug reaction is related to the suspected drug compared with the other drugs in the database was performed [29].
Information components (ICs) and Proportional Reporting Ratios (PRRs) -which are indicator quantities used to describe the disproportionality in reporting-were calculated for each ADR per treatment regimen. PRRs greater than 2, ICs greater than 1, IC 95% credibility interval (95% CrI) lower-bound greater than 0, and PRR 95% confidence interval (95% CI) lower-bound greater than 1 were all considered as signals of disproportionate reporting (SDR). In addition, the minimum number of reports required to detect a signal was 3 [29][30][31].

Logistic Regression Analysis
Logistic regression analyses were performed to calculate the reporting odds ratio (ROR) and investigate the association between the exposure to a DAA and the pattern of four specific serious events, while adjusting for the effect of covariates that possibly having an impact on the pattern of these ADRs including, age, gender, ribavirin use, and pre-existing cirrhosis. The four selected events were HCV/ HBV reactivation, anaemia, renal impairment (including AKI, renal insufficiency, and elevated blood creatinine), and hepatic complications (composite endpoint including hepatocellular carcinoma and hepatic encephalopathy). The aforementioned events were selected as being serious and co-reported with the four DAAs of interest.
Preceding this analysis, case review was performed to include only reports with the aforementioned events and covariates. Reports with SOF only and SOF/IFN regimens were not excluded from the case review and logistic regression analysis and were merged with SOF/RBV reports under a reference group named 'Legacy treatment', denoting the HCV treatment regimens used before the release of newer DAAs including DCV and LDV, which were thereafter used in combination with SOF [2,12]. The study flowchart is shown in Fig. 1.
From the results of descriptive and disproportionality analyses, estimates of safety and effectiveness -per IC, severity, fatality and seriousness indices-for the four DAA regimens are summarized in (Table 3).
Regarding the gender, male patients were at a lower risk of anaemia (OR 0.26, 95% CI 0.2-0.33). However, male patients were at higher risk of HCV/HBV reactivation (OR    ADR adverse drug reaction, DAA direct-acting antiviral, HCV hepatitis C virus, IC information component, PRR proportional reporting ratio, CI confidence interval, CrI credibility interval, DCV/SOF daclatasvir/sofosbuvir, OBV/PTV/r ombitasvir/paritaprevir/ritonavir, SOF/RBV sofosbuvir/ribavirin, SOF/LDV sofosbuvir/ledipasvir, RBV ribavirin, MPGN membranoproliferative glomerulonephritis. a N observed : the actual number of case reports for the drug-reaction combination [1]. b N expected : the number of case reports expected for the drug-reaction combination [1]. c N reaction : the number of case reports for the ADR irrespective of drug [1].

Discussion
In this study, we found unusual ADR signals associated with certain DAAs, particularly the signals of renal impairment and anaemia with OBV/PTV/r that was independent of ribavirin use as revealed by the logistic regression analysis, and the disproportionate reporting signal of HCC with SOF/DCV. The reported signals were exceedingly high, clinically relevant, and unexpected, given the normal use of DAAs and the natural progression of HCV infection. These findings mandate further research to confirm the exposure-effect relationship and to further validate the clinical relevance.
Comparing the results of descriptive and disproportionality analyses among the four treatment regimens revealed that OBV/PTV/r regimen was superior to the other DAAs concerning treatment effectiveness and compensated disease. Besides, OBV/PTV/r had the least reported serious, severe and fatal outcomes, indicating a favourable riskbenefit, despite the reported high signals of anaemia and renal impairment which mandate close monitoring of renal and hematologic parameters.
On the other hand, the highest percentage of severe outcomes, the highest signal of HCV relapse and associated hepatic complications and positive signals of anaemia and HBV reactivation were reported with SOF/RBV. Our conclusions regarding the risk-benefit evaluation are valid for the time of this cross-sectional study and may require an update in subsequent analysis.
This is the first pharmacovigilance study that highlighted the safety of DAAs in HCV-infected Egyptian patients using data from the WHO VigiBase, which includes a considerable number of reports. We performed a disproportionality analysis using Bayesian and frequentist signal indicators (ICs and PRRs), and multivariate logistic regression analysis to adjust for clinically relevant covariates. Our approach differs from Cazacu et al. (2020) and Hayes et al. (2021), who were only concerned with descriptives to evaluate DAAs' safety using VigiBase [13,32].
Our results regarding the most commonly reported ADRs were not much different from those reported in clinical trials on DAA regimens in concern; Kowdley et al. reported fatigue [35], headache and anaemia, Zeuzem [39].
Concerning OBV/PTV/r, the disproportionality analysis showed that the highest signal for OBV/PTV/r was for drug-drug interactions (DDIs), supported by the work of Cazacu et al. (2020), who highlighted potential DDIs in 1.9% and major DDIs in 20.4% of all ICSRs included in their analysis; Amlodipine was the medicine most frequently involved in the ICSRs with reported major DDIs, followed by furosemide, metformin, omeprazole and digoxin [32].
OBV/PTV/r, however, was also associated with serious and potentially fatal ADRs including anaemia and renal impairment. The observed number of reports of renal  [40].
The logistic regression analysis also showed that OBV/ PTV/r was independently associated with nearly fourfold risk of renal impairment. In addition, it showed that advancing age was associated with increased risk of renal impairment (ROR 1.05, 95% CI 1.02-1.08). These previously mentioned findings are in agreement with Cazacu et al. (2020), who reported an age-dependent risk of renal impairment with OBV/PTV/r [32].
Concerning anaemia, Gheorgh et al. (2017) reported anaemia as the most frequent ADR observed with OBV/ PTV/r (33.5%) [34]. In the present work, OBV/PTV/r was associated with nearly ninefold risk on anaemia, which was independent of ribavirin use (ROR 8.97, 95% CI 4.91-17.29). The strikingly high ROR for OBV/PTV/r induced anaemia calls for further studies to confirm a clinical correlation, if any. Despite the heightened risk of anaemia and renal impairment, OBV/PTV/r was superior to all other studied DAA combinations in terms of lower risk of HCV/HBV reactivation and progression to hepatic encephalopathy and HCC and being the regimen reported with the lowest severity index (0.53%) and fatal outcomes (0%).
A SDR of hepatocellular carcinoma (HCC) was observed with SOF/DCV (IC 3.46, 95% CrI 2.62-4.04), which was comparable with Lashen et al. (2018) findings, who reported discontinuation in 8 patients (1.3%) due to HCC development while on treatment with SOF/DCV [33]. HCC was also reported in two previous studies on patients treated with SOF/DCV; Hassany et al. (2020) reported HCC development in 22 (6.7%) patients who achieved SVR and 5 (23.8%) patients who failed to achieve SVR concluding that the occurrence of HCC is significantly higher in patients who failed to achieve SVR and that treatment response decreases the risk of HCC occurrence [6]. However, Aziz et al. (2019) reported that 10 (3.33%) of their study patients developed HCC, from which 5 patients achieved SVR, suggesting that the frequency of HCC following SOF/DCV regimen is independent of achieving SVR [43].
On contrary, other studies negates the correlation between the DAA treatment and HCC; Lin et al. (2020) concluded from their study that DAA therapy did not only increase the probability of HCC recurrence in patients who received curative treatment for HCC but it also improved the survival outcome of those patients [44]. Moreover, in a large cohort of North American patients by Singal et al. (2019), it was found that there was no significant difference in the HCC recurrence patterns between the DAA-treated patients and the untreated ones supporting that DAA therapy was not associated with the increased HCC recurrence [45].
Regarding the impact of gender on the risk of HCC, the logistic regression analysis showed that male patients were at higher risk of hepatic complications including HCC (OR 5.35, 95% CI 4.18-6.89), which supports the work of Cazacu et al. (2020), who highlighted the relationship between gender and neoplasms reporting higher percentage in male patients [32].
Our study had certain limitations. First, the total number of patients included in the analysis is the total number of patients who reported ADRs only, not the actual number of patients who received DAAs in Egypt, this is due to the issue of underreporting [18]. Second, the study did not include the role of drug interactions in the incidence of the reported ADRs. Third, Among the four studied DAA combinations, there were disparities in the frequency of reports which may had affected our comparisons; the highest proportion of 2032 reports were with SOF/DCV, and the lowest proportion of only 51 reports was with SOF/LDV. Finally, the findings resulting from the use of pharmacovigilance data and signal detection methods have to be explored for clinical validation as ADR reporting is a passive process that cannot provide a measurement of the true risk associated with the drug [13,18,46].

Conclusion
OBV/PTV/r was superior to other regimens in terms of lower risk of HCV/HBV reactivation and hepatic complications and being the regimen with the lowest severity index (0.53%), seriousness (9.5%) and fatal outcomes (0%). However, OBV/PTV/r correlated with fourfold higher risk of renal impairment and ninefold higher risk of anaemia than SOF based regimens, which calls for further populationbased studies for clinical validation.
The DAA combination SOF/RBV reported the highest severity index (4.02%) and seriousness (85.7%). besides, the highest signal of HCV relapse among the four regimens was reported with SOF/RBV which is possibly related to the efficacy of regimen.

Author Contributions
All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by MGE-M. The first draft of the manuscript was written by MGE-M, NAS and MHS commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding
Open access funding provided by The Science, Technology & Innovation Funding Authority (STDF) in cooperation with The Egyptian Knowledge Bank (EKB). No funding has been received for the conduct of this study.

Conflict of interest
Mai G, Mohamed H and Nagwa A declare that they have no conflict of interest.

Ethical approval
The data used in the study were anonymized and there was no direct interaction with human participants. Approval to conduct this study and extract data from VigiBase was obtained from the Egyptian Drug Authority (EDA). The study protocol was revised and approved by the Committee of Ethics, Faculty of Pharmacy, Ain Shams University (approval number: 236 dated 13-03-2019).

Open Access
This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.