| |
Effectiveness, safety and clinical outcomes of direct-acting antiviral therapy in HCV genotype 1 infection: Results from a Spanish real-world cohort
|
| |
| |
Download the PDF here
Jnl of Hepatology June 1017 - Jose Luis Calleja1,y, Javier Crespo2,y, Diego Rincon3, Belén Ruiz-Antoran4, Inmaculada Fernandez5, Christie Perello4, Francisco Gea6, Sabela Lens7, Javier Garcia-Samaniego8, Begona Sacristan9, Maria Garcia-Eliz10, Susana Llerena11, Juan Manuel Pascasio12, Juan Turnes13, Xavier Torras14,
Rosa Maria Morillas15, Jordi Llaneras16, Miguel A. Serra17, Moises Diago18, Conrado Fernandez Rodriguez19, Javier Ampuero20, Francisco Jorquera21, Miguel A. Simon22, Juan Arenas23, Carmen Alvarez Navascues24, Rafael Banares3, Raquel Munoz5, Agustin Albillos6, Zoe Marino7, The Spanish Group for the Study of the Use of Direct-acting Drugs Hepatitis C Collaborating Group
The higher response in our study may also reflect differences in patient management. In Spain, HCV patients are managed by experienced clinicians in referral centers rather than in community-based practice, which may result in greater treatment adherence and consequently higher response rates [[25], [26]].
Ours is one of the largest real-world studies in patients treated with the second-generation DAAs OMV/PTV/r + DSV or LDV/SOF. Such studies are critical, since effectiveness and safety are often lower than in clinical trials In particular, safety concerns have been raised in cirrhotic patients [[21], [22]]. Lower SVR rates have also been reported in early real-world studies of combination therapy including second-generation DAAs [[23], [24]].
It is important to reiterate that this study does not allow direct comparisons to be made between the two treatment regimens. Overall, the LDV/SOF subcohort included a significantly higher proportion of patients with traditionally harder to treat characteristics, such as genotype 1a infection and more advanced liver disease. This indicates differences in prescribing tendency for the two regimens. LDV/SOF may be perceived to be a more appropriate choice in patients with more advanced liver disease, as it is indicated in patients with decompensated cirrhosis, a population in which OMV/PTV/r + DSV is contraindicated. Seventy-two percent of patients received treatment in line with guidance from the European Association for the Study of the Liver, with most deviations from this being related to the administration of RBV. Non-adherence to this guidance was not associated with any significant difference in SVR [4].
Although evaluation of HCC was not an objective of this study, recent conflicting reports around the potential association of HCC with DAA-based therapy [[18], [19], [20]] prompted us to evaluate HCC recurrence and incidence in our combined cohort based on a review of patient records. Seventy patients had a prior diagnosis of HCC and had experienced a complete response prior to initiation of DAA therapy. Among these 70 patients HCC recurred in 21 (30%) within 12 months of starting DAA therapy (Table S8). Two patients subsequently died. Incident HCC was confirmed in 30/3,233 patients without a prior diagnosis within 18 months of starting DAA therapy. HCC was more common in patients with cirrhosis (p <0.001) but was not related to achievement of SVR (p = 0.400). At 14 months of follow-up post HCC diagnosis, four incident patients had died as a result of tumor progression, three as a result of liver failure and one from other causes. These observations should be interpreted with caution due to the absence of a specific surveillance protocol being adopted to follow patients in our cohorts.
Background & Aims
Clinical trials evaluating second-generation direct-acting antiviral agents (DAAs) have shown excellent rates of sustained virologic response (SVR) and good safety profiles in patients with chronic hepatitis C virus (HCV) genotype 1 infection. We aimed to investigate the effectiveness and safety of two oral DAA combination regimens, ombitasvir/paritaprevir/ritonavir plus dasabuvir (OMV/PTV/r + DSV) and ledipasvir/sofosbuvir (LDV/SOF), in a real-world clinical practice.
Methods
Data from HCV genotype 1 patients treated with either OMV/PTV/r + DSV ± ribavirin (RBV) (n = 1567) or LDV/SOF ± RBV (n = 1758) in 35 centers across Spain between April 1, 2015 and February 28, 2016 were recorded in a large national database. Demographic, clinical and virological data were analyzed. Details of serious adverse events (SAEs) were recorded.
Results
The two cohorts were not matched with respect to baseline characteristics and could not be compared directly. The SVR12 rate was 96.8% with OMV/PTVr/DSV ± RBV and 95.8% with LDV/SOF ± RBV. No significant differences were observed in SVR according to HCV subgenotype (p = 0.321 [OMV/PTV/r + DSV ± RBV] and p = 0.174 [LDV/SOF]) or degree of fibrosis (c0.548 [OMV/PTV/r/DSV ± RBV] and p = 0.085 [LDV/SOF]). Only baseline albumin level was significantly associated with failure to achieve SVR (p <0.05) on multivariate analysis. Rates of SAEs and SAE-associated treatment discontinuation were 5.4% and 1.7%, in the OMV/PTV/r + DSV subcohort and 5.5% and 1.5% in the LDV/SOF subcohort, respectively. Hepatocellular carcinoma (HCC) recurred in 30% of patients with a complete response to therapy for previous HCC. Incident HCC was reported in 0.93%.
Renal function
ledipasvir/sofosbuvir ± ribavirin - Both baseline and week 12 post-treatment eGFR data were available for 713 patients, including 36 (5.0%) with abnormal baseline renal function. For patients with normal baseline renal function, mean (SD) change in eGFR was −1.3 (11.3) ml/min/1.73 m2. Of the 36 patients with abnormal baseline renal function, 22/36 (61.1%) showed an improvement in eGFR at week 12 post-treatment (mean [SD] change + 10.1 [19.6] ml/min/1.73 m2), including 13 (36.1%) who showed >10 ml/min/1.73 m2 improvement.
Renal function
ombitasvir/paritaprevir/ritonavir and dasabuvir ± ribavirin - Both baseline and week 12 post-treatment glomerular filtration rate (eGFR) data were available for 659 patients, including 38 (5.8%) with abnormal baseline renal function (defined as eGFR <60 ml/min/1.73 m2). For patients with normal baseline renal function, mean (SD) change in eGFR was −1.6 (12.4) ml/min/1.73 m2. Among patients with abnormal baseline renal function, 19 (50%) showed an improvement in eGFR at week 12 post-treatment (mean [SD] change +3.19 [13.1] ml/min/1.73 m2), including seven (18.4%) who showed >10 ml/min/1.73 m2 improvement.
Conclusions
In this large cohort of patients managed in the real-world setting in Spain, OMV/PTV/r + DSV and LDV/SOF achieved high rates of SVR12, comparable to those observed in randomized controlled trials, with similarly good safety profiles.
Lay summary
In clinical trials, second-generation direct-acting antiviral agents (DAAs) have been shown to cure over 90% of patients chronically infected with the genotype 1 hepatitis C virus and have been better tolerated than previous treatment regimens. However, patients enrolled in clinical trials do not reflect the real patient population encountered in routine practice. The current study, which includes almost 4,000 patients, demonstrates comparable rates of cure with two increasingly used DAA combinations as those observed in the clinical trial environment, confirming that clinical trial findings with DAAs translate into the real-world setting, where patient populations are more diverse and complex.
Discussion
Ours is one of the largest real-world studies in patients treated with the second-generation DAAs OMV/PTV/r + DSV or LDV/SOF. Such studies are critical, since effectiveness and safety are often lower than in clinical trials In particular, safety concerns have been raised in cirrhotic patients [[21], [22]]. Lower SVR rates have also been reported in early real-world studies of combination therapy including second-generation DAAs [[23], [24]].
In the current study, both OMV/PTV/r + DSV and LDV/SOF resulted in high rates of SVR12, similar to those achieved in clinical trials, despite the inclusion of a high proportion of patients with cirrhosis and prior treatment failure. The SVR rates in our study were somewhat higher than those reported in a recent US study of 4,365 genotype 1 treatment-naïve veterans treated with LDV/SOF ± RBV [25]. This is likely to be related to differences in baseline characteristics; the US cohort included a high proportion of African-American patients, in whom SVR rates were significantly lower than in Caucasian patients (89.8% vs. 92.8%; p = 0.003) [25]. The higher response in our study may also reflect differences in patient management. In Spain, HCV patients are managed by experienced clinicians in referral centers rather than in community-based practice, which may result in greater treatment adherence and consequently higher response rates [[25], [26]].
The week 4 response with LDV/SOF in our study was lower than reported in clinical trials [[6], [7], [8]]. However, only up to 20% of patients in these studies were cirrhotic compared with almost 60% in our LDV/SOF cohort. Although the presence of cirrhosis had no impact on SVR12 in clinical trials, on-treatment data are not available to evaluate the effect of cirrhosis on early responses. In our study, patients with cirrhosis treated with LDV/SOF showed significantly lower on-treatment week 4 responses, compared with patients without cirrhosis. This was not the case in patients treated with OMV/PTV/r + DSV, possibly due to a higher proportion of patients with less advanced disease. Lower rates of week 4 response to LDV/SOF have been reported in other studies in advanced liver disease (83% in Child-Pugh B patients) [27]. Lack of response at week 4 was found to be related to significantly lower rates of SVR in the real-world study by Backus et al. [25]. However, lack of response at week 4 was not predictive of failure to achieve SVR in our cohort. In addition, despite a lower initial on-treatment response, the EOT and SVR12 rates in our patients treated with LDV/SOF were similar to those reported in pivotal clinical trials, and not significantly lower in cirrhotic patients. The exception to this was SVR12 following 8 weeks of LDV/SOF in 14 cirrhotic patients, a regimen which is not recommended in the product label. However, it is important to mention that failure to achieve SVR12 in these patients was mostly associated with treatment withdrawal and not with virologic failure.
Treatment for 8 vs. 12 weeks was associated with a significantly lower rate of SVR in the recent study by Backus and colleagues, including those who were eligible for this regimen according to treatment recommendations. However, although significant, the numerical difference in SVR rates was small (93.2% vs. 96.6%, respectively; p = 0.001). In the current study, patients who were eligible and treated with LDV/SOF for 8 weeks achieved an SVR12 rate of 95.7%, although they accounted for only 23% of those eligible for this regimen. Similarly, a substantial proportion of patients in the US Veteran study who would have been eligible for this regimen received 12 weeks instead [25], suggesting some reluctance to use the shorter regimen. However, data from our study should encourage confidence in this regimen in appropriate patients.
In patients with cirrhosis and a genotype 1b infection, treatment with OMV/PTV/r + DSV + RBV is recommended for 12 weeks [28], and was the most common regimen in our study. However, 98 patients with cirrhosis and a genotype 1b infection were treated with OMV/PTV/r + DSV without RBV for 12 weeks; of these, 93% achieved SVR12, a rate not significantly different to those receiving RBV (97%). These findings support data from the recently published TURQUOISE-III study (ClinicalTrials.gov identifier: NCT02219503) which showed an SVR in 100% with OMV/PTV/r + DSV without RBV for 12 weeks in 60 similar patients [16]. Taken together, these data confirm the effectiveness of OMV/PTV/r + DSV without RBV in cirrhotic patients with genotype 1b infection.
Overall in our cohort cirrhotic and non-cirrhotic patients achieved similarly high rates of SVR12 as did Child-Pugh A and B patients. Neither Child-Pugh B nor the presence of portal hypertension (esophageal varices) was associated with failure to achieve SVR. These results are particularly encouraging given the favorable safety profile in cirrhotic patients discussed below.
Rates of virologic failure were low, and similar to those reported in clinical trials. Resistance testing is not routinely performed in Spain and therefore data on the presence of resistance mutations at relapse to any of the DAAs used are not available. Relapse seemed to occur later in the OMV/PTV/r + DSV sub-cohort. Time of relapse is not available in phase 3 trials of OMV/PTV/r + DSV. However, in the phase 2b AVIATOR study (ClinicalTrials.gov identifier: NCT01464827) with OMV/PTV/r + DSV all relapses occurred before post-treatment week 4 [29]. The reason for, and relevance of, the later relapses with OMV/PTV/r + DSV seen in our cohort are unclear and long-term follow-up would be required to evaluate this. With LDV/SOF, the majority of patients who relapsed in the phase 3 ION-2 study did so between EOT and week 4 post-treatment [7], as in our cohort. In a recent analysis of 159 patients treated in National Institutes of Health studies followed for a period ranging from 1 to 116 weeks after achieving SVR12 with LDV/SOF, all patients maintained a virologic response with no evidence of late relapse [30].
It is important to reiterate that this study does not allow direct comparisons to be made between the two treatment regimens. Overall, the LDV/SOF subcohort included a significantly higher proportion of patients with traditionally harder to treat characteristics, such as genotype 1a infection and more advanced liver disease. This indicates differences in prescribing tendency for the two regimens. LDV/SOF may be perceived to be a more appropriate choice in patients with more advanced liver disease, as it is indicated in patients with decompensated cirrhosis, a population in which OMV/PTV/r + DSV is contraindicated. Seventy-two percent of patients received treatment in line with guidance from the European Association for the Study of the Liver, with most deviations from this being related to the administration of RBV. Non-adherence to this guidance was not associated with any significant difference in SVR [4].
Safety and tolerability, including renal safety, with both regimens was good. However, despite the real-world nature of our cohort, few patients with markedly impaired renal function were included, thus precluding firm conclusions. Overall, reported rates of SAEs (5.5%) were only slightly higher or similar to those reported in the pivotal clinical trials (up to 5.5% for OMV/PTV/r + DSV and up to 3.8% for LDV/SOF) [[6], [7], [8], [9], [10], [11], [12], [13], [14], [15]] although the rate of SAEs in cirrhotic patients was slightly higher in both subcohorts (OMV/PTV/r + DSV 8.1%; LDV/SOF 8.3%) than in clinical trials (5.5% with OMV/PTV/r + DSV; [12] 3.0% for LDV/SOF [31]). Liver-related SAEs were rare (<1.5%) and decompensation rates were low (<1%). Decompensation and liver failure have been reported in cirrhotic patients treated with OMV/PTV/r + DSV, which led to its contraindication in Child-Pugh B as well as Child-Pugh C patients in Spain [32] and recommendations for additional monitoring [33]. In the current study, only eight out of 732 patients with cirrhosis (0.5%) who were treated with OMV/PTV/r + DSV experienced decompensation. Although not significantly different, this rate was lower than observed with LDV/SOF (0.9%). Overall, the probability of incident decompensation was significantly higher in patients with a history of decompensation than in those without, which is consistent with recently published data [17]. Factors associated with poorer safety on univariate analysis were similar with both treatments and, as expected, were generally related to more advanced liver disease and dysfunction. As the primary baseline predictors of SVR12 by univariate analysis were also related to liver function, these data support treating patients earlier during disease. Post hoc analyses of recurrence and incidence of HCC in our combined cohort agreed with previous studies [[18], [19], [20]], but must be interpreted with caution given the lack of a routine surveillance monitoring protocol.
This study has the usual limitations related to its observational, real-world design and to electronic data collection, including potential physician prescribing bias, incomplete patient records, local practice discrepancies, and data entry errors. Nevertheless, the large number of patients included gives an important insight into the effectiveness and safety of two increasingly used treatment regimens in the diverse patient population managed in routine practice.
In summary, both OMV/PTV/r + DSV and LDV/SOF yielded similar or higher rates of SVR12 in the real-world setting compared with randomized clinical trials, with similarly good safety profiles. It is the authors' collective opinion that these results provide definitive evidence of the effectiveness of these regimens in the management of patients with chronic HCV genotype 1 infection in routine practice.
|
|
| |
| |
|
|
|
