Abstract and Introduction
In this prospective study, we evaluated the effectiveness and tolerability of novel therapies against hepatitis C virus (HCV) in a cohort of PWID enrolled at our centre from April 2015 to July 2016. In this analysis, a total of 174 patients were included: eleven (6.3%) were treated with pegylated interferon (PEG-IFN) and ribavirin (RBV) containing regimens, 163 (93.7%) with IFN-free treatments. RBV has been used in 70 patients (40.2%); 59 (33.9%) patients were in opioid substitution therapy (OST) with methadone or buprenorphine. Overall, sustained virological response (SVR) has been observed in 162 subject (93.1%), breakthrough (BT) in three (1.7%), relapse in one (0.6%) and dropout in eight (4.6%). Treatment was interrupted for clinical conditions in seven patients: six (3.4%) had hepatic decompensation and one died for hepatocellular carcinoma (HCC). In multivariate analysis, predictive factors of treatment failure were as follows: albumin level below 3 g/dL (OR=7.190; 95% IC=1.236–41.837; P<.001), MELD score >10 (OR=5.886; 95% IC=1.411–35.994; P<.001) and years of HCV infection >20 (OR=1.286; 95% IC=0.556–9.455; P=.016). In conclusion, treatment with DAAs was effective and well tolerated in PWID; cirrhotic subjects with MELD > 10 and albumin low level showed a higher risk of developing serious adverse events and treatment failure.
Injection drug use (IDU) is the major route of transmission of hepatitis C virus (HCV) in most countries with an estimated global prevalence of 67% among people who inject drugs (PWID). Factors associated with HCV infection in PWID are strictly related to socio-economic conditions and lifestyle of patients such as imprisonment, unemployment, poverty and low education levels.[3,4] The progression of chronic hepatitis C (CHC) to hepatic fibrosis or cirrhosis usually requires many decades, but in some cases and in the presence of other factors, it may be more rapid, especially in PWID who acquired the infection in young age and with moderate/severe alcohol consumption. In these patients, treatment with pegylated interferon alpha (PEG-IFN) and ribavirin (RBV) showed a sustained virological response (SVR) rate comparable to other HCV infected people. Despite the presence of many favourable factors for treatment in PWID, such as the young age, low fibrosis stage or genotype 3, a small fraction of patients has received treatment, for reasons related to social or living conditions, imprisonment, medical and psychiatric comorbidities. In this population, the risk of poor adherence or re-infection discouraged clinicians to start treatment with PEG-IFN and RBV, especially in genotype 1 (GT1) and genotype 4 (GT4), due to long treatment duration and the impact of side effects. The European Association for the Study of the Liver (EASL) guidelines recommend to treat PWID with novel and effective therapies based on direct-acting antiviral agents (DAAs); nevertheless, no data are currently available in a real-life clinical context concerning the treatment of this subpopulation with IFN or IFN-free regimens.
For these reasons, an assessment of effectiveness and tolerability of DAAs in PWID was provided in this prospective study.
End-Points and Study Design
This prospective single-centre study was conducted at our centre of Infectious Diseases “Amedeo di Savoia Hospital” Turin, Italy, from April 2015 to July 2016; the study protocol was approved by our local Ethic Committee and was conducted in compliance with the Declaration of Helsinki and with the local Review Board regulations.
PWID affected by CHC, treatment naïve or experienced, with any genotype or fibrosis condition, eligible for IFN or IFN-free regimens according to EASL guidelines and our Italian policies of treatment access (established by the Italian Agency of Drug, AIFA) have been included in this study.
PWID were defined as patients with at least one episode of intravenous drug use (heroin or cocaine) in the last year and referred by the local public centres for treatment of drug addiction (“SerT”), where these subjects received the first diagnosis, medical care and pharmacological/psychological support, with or without opioid substitution treatment (OST).
Exclusion criteria were decompensated cirrhosis, other risk factors for HCV, pregnancy or coinfection with HIV or HBV.
Primary end-point of this study was the evaluation of the SVR rate in these patients; secondary end-points were the analysis of side effects and the treatment interruption due to toxicity or viral breakthrough.
SVR was defined as undetectable HCV-RNA 12 weeks after the treatment completion.
End-of-treatment response (ETR) was defined as undetectable HCV-RNA at the end of treatment, while relapse (REL) was defined as detectable HCV-RNA after ETR. Clinical and laboratory parameters were documented at the baseline and during the treatment. Liver fibrosis was measured with transient elastography (Fibroscan®) as METAVIR score at baseline and at 12 weeks after end of treatment in patients who gained SVR.
Safety evaluation was performed according to WHO grading scale.
HCV-RNA was quantified in plasma samples using a commercially available real-time PCR system, the Cobas Ampliprep/Cobas TaqMan system (HCV-RNA v2.0 Roche Molecular System, Branchburg, NJ, USA) with a detection limit of 12 IU/mL. HCV genotyping was performed with a line probe assay (INNO-LiPA HCVTM II, Siemens Healthcare Diagnostics).
The treatment choice for eligible patients was performed according to clinical guidelines, patients history and AIFA criteria for the access to IFN-free regimens in Italy.
In detail, AIFA criteria indicate only patients with severe fibrosis (F3 or F4) or cirrhosis as eligible for IFN-free treatment (according to HCV genotypes and clinical conditions), whereas patients with low stages of hepatic fibrosis can be treated only with IFN-based regimens, such as with simeprevir (SMV) in GT1 or GT4.
On the other hand, patients infected with GT2 and GT3 HCV, without fibrosis, can be treated only with standard dual therapy (RBV plus PEG-IFN).
Only patients with extrahepatic manifestations of CHC, with documented organ involvement, can be treated with an IFN-free regimen without significant fibrosis.
For descriptive statistics, continuous variables were summarized as median values and interquartile ranges (IQR, 25–75 percentile range). Categorical variables were described as frequencies and percentages. All data were assessed for normality using the Shapiro-Wilk test, and differences in continuous variables between groups were evaluated using the Mann-Whitney or Kruskal-Wallis statistical test (for two or more groups respectively). The Spearman correlation test was utilized to investigate continuous data. The associations between categorical variables were assessed through the chi-squared test. To test the predictive role of single variables for treatment failure, univariate logistic regression was performed: only variables with a P value lower than .2 have been tested in the multivariate analysis. P values lower than .05 have been considered statistically significant. All the statistical testing has been performed through SPSS version 22.0.
Baseline Characteristics of the Study Population
Among the 345 patients considered at first, 174 were retained in the analysis. We excluded 145 subjects (non-PWID), 14 (inactive drug users nonfollowed by any public centre) and 12 (without informed consent).
The baseline features of this population are reported in Table 1. The majority of subjects were males (89.7%) and Italian (97.8%), with a median age of 51.5 years; 59 received the OST with methadone or buprenorphine. Eighty-three presented a psychiatric disorder (47.7%), most frequently anxiety and sleep disorders; all patients with major depression, schizophrenia or other behavioural disorders were treatment naïve. Patients with prior imprisonment status were 21 (12%), and 68 were unemployed (39.1%). Fibrosis stage distribution was as follows: twelve patients (7%) with F0-F2, 53 (30.4) with F3, 109 (62.6%) with F4. Cirrhotic patients were Child-Pugh A (102) and B; twelve patients had a MELD score >10.
HCV genotypes distribution was 69 subjects were infected with 1a (39.6%), 32 with 1b (18.4%), three with 2 (1.1%), 50 with 3 (28.7%) and 20 with 4 (11.5%).
The majority of patients were naïve (108, 62%), 66 experienced (37%) with previous failure to PEG-IFN and RBV or triple therapy with telaprevir (TLV) or boceprevir (BOC).
Drug use Characteristics
The majority of patients reported the use of intravenous heroin (93%); the 4% referred the use of cocaine by injection or sniffing; in 3% the combined use of heroin, cocaine and amphetamine were documented. The median duration of drug use was 15.9 years (IQR=8.9–23.5); the median time to follow-up at the public centres for treatment of drug addiction was 9.6 years (IQR=4.6–19.1).
Nine patients with lower fibrosis stage were treated with PEG-IFN, RBV and SMV regimens; two patients with GT3, F3 and naïve were treated with PEG-IFN, RBV and sofosbuvir (SOF) for 12 weeks.
Twenty patients received SOF+SMV with or without RBV for 12 weeks, 54 SOF+ ledipasvir (LDV) for 8, 12 or 24 weeks, 26 ombitasvir (OMB)/paritaprevir (PAR)/ritonavir (r)±dasabuvir (DAS) (“3D” combination) with or without RBV for 12 or 24 weeks, 56 with SOF+daclatasvir (DAC) with or without RBV for 12 or 24 weeks, seven with SOF+RBV for 24 weeks.
Ribavirin was used in all PEG-IFN-based regimens and in IFN-free when indicated according to treatment schedule, patient conditions and clinical evaluation: on this basis, 104 patients received a RBV-free therapy (59.8%), while 70 (40.2%) had a RBV-including regimen.
Treatment schedules were as follows: twelve weeks of triple therapy with PEG-IFNα, RBV and SMV, followed by 12 weeks of dual therapy, for treatment-naïve F0-F2 patients infected with GT1 or GT4 (9, 5.2%); 8 weeks with SOF+LDV (2, 1.1%) for naïve F3 patients with GT1; 12 weeks with SOF+SMV±RBV for F3-F4 patients with GT1 or GT4; 12 weeks with OMB/PAR/r±DAS±RBV for F3-F4 patients with GT1b or for F3 patients with GT1a-GT4; finally, a treatment duration of 24 weeks was scheduled for cirrhotic patients treated with SOF+LDV, SOF+DCV±RBV or OMB/PAR/r+DAS+RBV (GT1a, F4).
A summary of treatment outcomes is reported in Table 2. SVR was achieved in 162 patients (93.1%), virological BT in three (1.7%), REL in one (0.6%); in eight patients, treatment was stopped before the expected treatment duration; in six patients, the interruption was due to hepatic decompensation, with need for hospitalization; one patient died due to the onset of multinodular hepatocellular carcinoma (HCC); one patient, in treatment with PEG-IFNα, RBV and SMV, was lost after the first 3 months of therapy.
Cases of BT were due to the selection of resistance-associated variants (RAVs): one in treatment with PEG-IFN, RBV and SMV (D168V at NS3 locus), another one in treatment with SOF+RBV (S282T at NS5B locus) and, finally, one in a patient treated with SOF+DCV+RBV (L28V at NS5A+S282T at NS5B loci).
Frequently observed side effects were as follows: pruritus (9.2%), grade 1 anaemia in patients receiving RBV (5.7%), flu-like syndrome in IFN-based regimens (3.4%), headache (12.6%) and gastrointestinal symptoms (8%); hepatic decompensation was observed during the treatment in 15 subjects with MELD score >9; among these, nine continued therapy with supportive treatment (diuretics, potassium and rest).
Treatment Outcomes According to Type of Treatment and Genotypes
In Figure 1A, treatment outcomes are shown according to treatment; for PEG-IFN regimens, SVR was 82%, BT 9%, dropout 9%; in SOF+SMV±RBV, SVR was 95%, relapse 5%; in 3D±RBV, SVR was 100%; in SOF+DCV±RBV, SVR was 91.1%, BT 1.8%, dropout 7.1%; in SOF+RBV, SVR was 85.7 and BT 14.3%.
Treatment outcomes according to regimens (A) and viral genotypes (B)
According to HCV genotypes (Figure 1B), the SVR rates were as follows: 97.1% in GT1a, 93.8% in GT1b, 100% in GT2, 88% in GT3 and 90% in GT4.
Treatment Outcomes According to Clinical Characteristics
In Figure 2, the rate of SVR is shown by patient characteristics: no differences were observed between RBV-free and RBV-including regimens (92.2% vs 93.6%), naïve or experienced patients (93.6% vs 93.9%), whereas a significant higher SVR rate was present in patients with a baseline MELD score <10 (96.9% vs 58.3%, P<.001).
SVR12 rates according different clinical and therapeutic characteristics
Liver Stiffness Variation After Treatment in Patients With SVR
Liver stiffness measurement, performed by transient elastography, was repeated at 12 weeks after the end of treatment in patients with SVR. As follows, the available data regarding 90 cirrhotic patients and 50 with F3 are reported: the median stiffness reduction was 5.2 kPa (IQR: 1.9–8.7) in F4 patients and 3.6 kPa (IQR: 1.8–7.7) in F3.
In detail, among cirrhotic patients, liver stiffness decreased in 84 cases, increased in two cases and remained unchanged in four cases. All the patients with F3, conversely, evidenced a decrease in liver stiffness.
Based on these data, we described the METAVIR change after SVR (Figure 3): considering cirrhotic patients, 71 (79%) remained in F4 stage, 16 changed to F3 (18%) and three to F0-F2 (3%); conversely, among F3 patients, 32 changed to F1 stage (64%), 12 to F2 (24%) and six remained F3 (12%).
Fibrosis stage (Metavir score) at baseline and follow-up in patients with SVR12
Predictive Factors for Treatment Failure
Receiver operating characteristic (ROC) curve has been used in order to determine albumin, MELD score, liver stiffness, years of infection and drug abuse cutoff values to predict treatment failure. By univariate analysis, the following factors were related to treatment failure: albumin level ❤ g/dL at baseline, MELD score >10, liver stiffness >30 kPa, years of HCV infection >20, HCV GT3, insulin resistance and years of drug abuse>10. By multivariate analysis, the most important predictive factor for treatment failure was the albumin level ❤ g/dL (OR=7.190; 95% IC=1.236–41.837; P<.001) followed by the MELD score >10 (OR=5.886; 95% IC=1.411–35994; P<.001) and the time of HCV infection >20 years (OR=1.286; 95% IC=0.556–9.455; P<.016)(Table 3).
Treatment of HCV infection in PWID represented an interesting challenge in medicine and public health; in the past, treatment with PEG-IFN and RBV in this setting was debated considering the problems of adherence, difficult management of side effects and the higher prevalence of psychiatric adverse events. However, the effectiveness and tolerability of this therapy in PWID were similar to than observed in other patients,[6,13–15] and the global observed adherence was not different between PWID and non-PWID patients. Despite these encouraging data, the access to treatment in PWID was lower compared to other HCV infected subjects;[7,8] another debatable issue was the risk of re-infection in patients who gained SVR as a possible contraindication for treatment in PWID, but available data reported a low risk (approximately 1%-5%) even in presence of risk behaviours. Major barriers for the access to treatment in this population were primarily the socio-economic conditions, the low educational level and the presence of other behavioural disorders, which limited the patients admittance at local health centres.
Nowadays, the introduction of new DAAs against HCV and the availability of IFN-free regimens have rapidly changed the clinical approach to CHC; however, no data are currently available about the real-life effectiveness and tolerability of these novel therapies in PWID.
In this study, we reported an encouraging overall SVR rate of 93%, with a low rate of dropout (4%); however, many aspects deserve to be better clarified. First, despite the prevalence of fibrosis stage F3 or F4 in our population (93.1%), 102 (62%) patients were treatment naïve: the main reason of this finding is related to the high prevalence of psychiatric disorders (45%) among these subjects, who were previously excluded from the standard treatment with PEG-IFN and RBV.
Other problems causing previous treatment refusal were active alcoholism, unemployed status, social or life troubles, detention or living alone condition. The majority of other naïve patients, with low fibrosis stage and eligible for IFN-based regimens, refused the PEG-IFN plus RBV and SMV therapy due to fear of side effects and to the theoretical availability in the near future, in Italy, of IFN-free regimens also for F0-F2 patients; therefore, a low number of naïve patients without fibrosis were enrolled in this study and treated with IFN-based therapies.
Although IFN-based regimens evidenced a lower SVR rate than IFN-free regimens (Figure 2), this data should be considered with caution: first, the number of cases is widely different between these two groups (11 vs 163) and, second, the rate of SVR in IFN-based group is an intention-to-treat (ITT) analysis, conditioned by one dropout patient; considering a per-protocol (PP) analysis, in this group, the SVR rate was 90.9%. Moreover, the IFN-based treatments were well tolerated in these patients, and only mild dermatological symptoms (pruritus or mild erythema) were observed in patients treated with SMV. For these reasons, treatment with PEG-IFN, RBV and SMV should be considered and proposed in naïve patients with GT1 or GT4 and low fibrosis stage, especially in the presence of strong motivation.
On the other hand, the majority of treated PWID received an IFN-free regimen with good SVR rate and tolerability. Virological relapse was observed in two patients treated with SOF+LDV and SOF+DCV+RBV, whereas the only BT was reported in a suboptimal treatment with SOF+RBV in a GT3 cirrhotic patient. However, a significant different rate of SVR was evidenced according to baseline MELD score: 58.3% in MELD >10 vs 96.9% in MELD <10 (P<.001). This aspect was confirmed in the multivariate analysis and related to treatment interruption in patients with advanced liver cirrhosis, who evidenced a hepatic decompensation with need of hospital admission. Furthermore, a low albumin level was observed as a significant predictor of treatment failure, strongly related to advanced hepatic disease, as recently demonstrated in cirrhotic patients treated with SOF+SMV. Interestingly, age was not different between patients with MELD scores under and over 10 (48.5 vs 49.1, respectively, P=.482), but a significant difference was found considering the years of HCV infection (19.5 vs 13.5, respectively, P<.001) and years of drug abuse (20.5 vs 12, respectively, P<.001). The duration of infection in addition to drug abuse (and probably other factors) led to serious liver damage with consequent higher risk of serious adverse events during the treatment. This could be an essential issue for the future approach of treatment in PWID, considering that an early treatment might lead not only to viral eradication, but also to prevention of fibrosis, cirrhosis and end-stage liver disease. An interesting strategy of viral eradication in PWID was proposed by de Vos et al. with a first treatment in low-risk PWID and, in a second time, the treatment of HCV infection in high-risk PWID; other policies of care should be considered according to HCV prevalence and availability of novel therapies in different countries.
In this real-life study, we demonstrated for the first time the high effectiveness and safety of IFN-free regimens in PWID with advanced fibrosis or cirrhosis and the feasibility of these treatments due to low number of pills, good tolerability and need of less frequent blood test than IFN regimens, with an excellent global adherence. An important open issue remains the chance of SVR in patients with advanced liver cirrhosis and the risk of serious adverse events and treatment interruption. Finally, we suggest that the choice of IFN-based regimens for eligible subjects, despite requiring more efforts by both patients and clinicians, can be still a viable option.