Volume 151, Issue 3, Pages 457–471.e5
Effectiveness of Sofosbuvir, Ledipasvir/Sofosbuvir, or Paritaprevir/Ritonavir/Ombitasvir and Dasabuvir Regimens for Treatment of Patients With Hepatitis C in the Veterans Affairs National Health Care System
DOI: http://dx.doi.org/10.1053/j.gastro.2016.05.049
Background & Aims
We investigated the real-world effectiveness of sofosbuvir, ledipasvir/sofosbuvir, and paritaprevir/ritonavir/ombitasvir and dasabuvir (PrOD) in treatment of different subgroups of patients infected with hepatitis C virus (HCV) genotypes 1, 2, 3, or 4.Methods
Results
Conclusions
Discussion Only
Download Full Text Article
LDV/SOF, PrOD, and SOF-based antiviral regimens resulted in remarkably high SVR rates in the VA national health care system, approaching the rates reported in clinical trials. This is in contrast to previous interferon-based regimens, which consistently resulted in much lower SVR rates in real-world clinical practice than in clinical trials.12, 13, 14, 15, 16, 17 SVR rates were higher in genotype 1– (SVR 92.8%) and genotype 4–infected patients (SVR 89.6%) than genotype 2– (SVR 86.2%) or 3–infected patients (SVR 74.8%), and these differences were even greater among cirrhotic and treatment-experienced patients. Among genotype 1–infected patients, there was no significant difference in SVR rates between LDV/SOF and PrOD regimens in either unadjusted or multivariable, propensity-score-adjusted analyses, and SVR rates >90% were achieved even in subgroups such as treatment-experienced or cirrhotic patients. The short, 8-week LDV/SOF monotherapy regimen resulted in excellent SVR rate (94.3%), but was used in only 48.6% of genotype 1–infected patients eligible for 8-week therapy (ie, treatment-naïve patients with viral load <6 million IU/mL, without cirrhosis). Long, 24-week regimens did not result in higher SVR rates and were rarely used, despite being FDA-approved and American Association for the Study of Liver Diseases/Infectious Diseases Society of America–recommended3232 for certain genotype 1 patients with cirrhosis.
Among a total of 17,487 patients, our study included 5250 patients with a diagnosis of cirrhosis and 5960 with a FIB-4 score >3.25 (which is highly suggestive of cirrhosis), who achieved surprisingly high overall SVR rates of 86.8% and 87.4%, respectively. These high SVR rates were driven by genotype 1–infected cirrhotic patients who had much higher SVR (90.6%; 95% CI, 89.7%–91.5%) than genotype 2 (77.3%; 95% CI 73.3%–80.9%) or genotype 3 (65.7%; 95% CI, 61.2%–69.8%). To our knowledge, this is the largest study of DAAs in cirrhotic patients and the SVR rates in genotype 1–infected patients are the highest reported in real-world clinical practice. LDV/SOF, PrOD, and SOF regimens have allowed patients with cirrhosis to be cured of HCV in substantial numbers and proportions for the first time ever. Longer follow-up of these patients is necessary to determine whether patients with cirrhosis who achieve SVR by DAAs are protected from developing progressive liver dysfunction, liver failure, or HCC and, whether they are capable of liver remodeling and regression of cirrhosis.
American Association for the Study of Liver Diseases/Infectious Diseases Society of America guidelines during the time period of our study,3232 as well as the LDV/SOF package insert,3333 recommend that LDV/SOF regimens should extend for 12 or 24 weeks, with the single exception of a short, 8-week LDV/SOF monotherapy regimen that “can be considered,”3333 “with caution and at the discretion of the practitioner”3232 in treatment-naïve, genotype 1–infected patients without cirrhosis with an HCV viral load <6 million IU/mL. This is based on a post-hoc analysis of the ION-3 clinical trial showing higher relapse rates in those treated for 8 weeks who had a viral load ≥6 million (9 of 92 [10%]) compared with those with a viral load <6 million IU/mL (2 of 123 [2%]).33 VA treatment guidelines explicitly recommended 8 weeks of treatment for this subgroup of patients.3434 Indeed, our study confirmed that 8 weeks of LDV/SOF monotherapy had similarly high SVR rates (94.8%) as 12 weeks (95.3%) in this favorable subgroup. However, our results also showed that treatment was unnecessarily extended beyond 8 weeks in 1833 of 4066 patients in this subgroup, dramatically increasing the cost of treatment without increasing SVR. Our results should offer reassurance to treatment providers that 8 weeks of LDV/SOF monotherapy is sufficient duration in this subgroup.
VA treatment guidelines during the study period designated PrOD as the preferred regimen in genotype 1–infected patients except for prior null responders, those previously treated with protease inhibitors, and patients with Child’s B or C cirrhosis (in whom the preferred regimen was LDV/SOF and ribavirin for 12 weeks) and except for treatment naïve, non-cirrhotics with a viral load <6 million (in whom 8-week LDV/SOF and 12-week PrOD regimens were equally preferred). This was due to the lower cost of 12 weeks of PrOD ($22,850) compared with 12 weeks of LDV/SOF ($37,157) in the VA system during the study period and the absence of evidence that one is more effective than the other in the subgroups for which PrOD was preferred. Our data support the VA treatment recommendations because we found no difference in SVR between PrOD and LDV/SOF regimens in either adjusted or unadjusted analyses, in the entire population or in clinically relevant subgroups (cirrhosis or not, treatment-experienced or naïve). Despite these recommendations and the higher cost, LDV/SOF regimens constituted 77% and PrOD only 23% of regimens in genotype-1–infected patients. This could be due to higher prevalence of drug–drug interactions, higher pill burden, and more frequent requirement for co-prescription of ribavirin in PrOD regimens compared with LDV/SOF regimens.
After the end of the study period, and after FDA approval of elbasvir/grazoprevir as an additional regimen for genotype 1 HCV on January 28, 2016, regimen costs in the VA were further reduced dramatically, to approximately $17,000 per 12-week course for LDV/SOF, PrOD, and elbasvir/grazoprevir and treatment recommendations changed to “equally recommend” all 3 agents as of March 2016.
Genotype 3–infected patients had the lowest SVR rates in our study, just as in clinical trials. We found that the non-FDA–approved regimen of LDV/SOF and ribavirin had a higher SVR rate (77.9%; 95% CI, 73.1%–82.0%) than the longer and more expensive FDA-approved regimen of SOF and ribavirin for 24 weeks (70.6%; 95% CI, 66.9%–74.1%). However, the highest SVR rate in genotype 3–infected patients was observed in the regimen that included PEG together with SOF and ribavirin (87.0%; 95% CI, 80.0%–91.8%), the only interferon-containing regimen that is still recommended.3232
Few large real-world studies of interferon-free regimens are currently available for comparison with ours. The HCV-TARGET, a prospective cohort study of patients undergoing HCV treatment in routine clinical care in academic centers, reported SVR rate to SOF and simeprevir in genotype 1–infected patients of 88% among 151 transplant recipients and 84% among 836 non-transplant recipients.35, 36 This regimen has been superseded by LDV/SOF and PrOD-based regimens. Among 487 patients with decompensated cirrhosis treated in the United Kingdom under an Expanded Access Programme with SOF, LDV/SOF, or daclatasvir, SVR was achieved in 90.5% of genotype 1– and 68.8% of genotype 3–infected patients3737—very similar to our findings. A smaller VA study looked at only treatment-naïve, genotype 1–infected patients treated with LDV/SOF and reported SVR rates almost identical to ours among this subgroup.3838 The TRIO Network, which compiles data from participating “real-world” academic and community HCV treatment clinics in the United States, reported in an abstract an SVR rate of 94% among 1521 genotype 1–infected patients treated with LDV/SOF monotherapy.3939
The main limitation of our study was that SVR data were unavailable in 9% of patients, which can lead to overestimated SVR rates among those with available SVR data. We think this is unlikely for 2 reasons. First, patients with missing SVR data were similar to those with available SVR data (Supplementary Table 2 Supplementary Table 2). Although early discontinuation of treatment in <8 weeks was more common in patients with missing SVR data (25% vs 4.4%), the majority of patients with missing SVR completed 8 or more weeks of treatment, demonstrating that patients with missing SVR data were not patients who “dropped out” of treatment or were “lost to follow-up,” but rather patients (or physicians) who were simply delinquent in getting their SVR viral load measured after the end of their treatment—not an uncommon phenomenon outside of clinical trials. Second, we used comprehensive multiple imputation models that included duration of treatment in addition to baseline, pretreatment characteristics to impute the missing SVR data and found only a non-substantial reduction in SVR after imputation (Table 5Table 5), suggesting that it is unlikely that our results of observed SVR are biased toward overestimation due to the missing SVR data. Important strengths of the study include the complete ascertainment of filled pharmacy prescriptions and the utilization of complete electronic medical records since 1999 from a national health care system that treats the greatest number of HCV-infected patients in the United States.
Our results demonstrate that LDV/SOF, PrOD, and SOF regimens can achieve remarkably high SVR rates in real-world clinical practice, especially in genotype 1–infected patients. The main obstacle to curing HCV infection in the maximum possible number of patients is currently the cost of HCV antiviral regimens. It is expected that cost will decline dramatically as more antiviral regimens become FDA-approved, resulting in competition between manufacturers. In fact, costs decreased dramatically within the VA after the completion of our study and after the FDA approval of elbasvir/grazeprevir in January 2016. The VA health care system has budgeted $1.5 billion nationally for antiviral medications for fiscal year 2016, while every health care organization in the United States is faced with similar budgetary constraints due to the cost of antiviral medications. We hope that our results will be used to determine the most cost-effective ways to treat HCV-infected patients and to reassure patients, clinicians, and health care systems that current treatments for HCV, though costly, appear to be effective in the real-world setting.
Acknowledgments
Among a total of 17,487 patients, our study included 5250 patients with a diagnosis of cirrhosis and 5960 with a FIB-4 score >3.25 (which is highly suggestive of cirrhosis), who achieved surprisingly high overall SVR rates of 86.8% and 87.4%, respectively. These high SVR rates were driven by genotype 1–infected cirrhotic patients who had much higher SVR (90.6%; 95% CI, 89.7%–91.5%) than genotype 2 (77.3%; 95% CI 73.3%–80.9%) or genotype 3 (65.7%; 95% CI, 61.2%–69.8%). To our knowledge, this is the largest study of DAAs in cirrhotic patients and the SVR rates in genotype 1–infected patients are the highest reported in real-world clinical practice. LDV/SOF, PrOD, and SOF regimens have allowed patients with cirrhosis to be cured of HCV in substantial numbers and proportions for the first time ever. Longer follow-up of these patients is necessary to determine whether patients with cirrhosis who achieve SVR by DAAs are protected from developing progressive liver dysfunction, liver failure, or HCC and, whether they are capable of liver remodeling and regression of cirrhosis.
American Association for the Study of Liver Diseases/Infectious Diseases Society of America guidelines during the time period of our study,3232 as well as the LDV/SOF package insert,3333 recommend that LDV/SOF regimens should extend for 12 or 24 weeks, with the single exception of a short, 8-week LDV/SOF monotherapy regimen that “can be considered,”3333 “with caution and at the discretion of the practitioner”3232 in treatment-naïve, genotype 1–infected patients without cirrhosis with an HCV viral load <6 million IU/mL. This is based on a post-hoc analysis of the ION-3 clinical trial showing higher relapse rates in those treated for 8 weeks who had a viral load ≥6 million (9 of 92 [10%]) compared with those with a viral load <6 million IU/mL (2 of 123 [2%]).33 VA treatment guidelines explicitly recommended 8 weeks of treatment for this subgroup of patients.3434 Indeed, our study confirmed that 8 weeks of LDV/SOF monotherapy had similarly high SVR rates (94.8%) as 12 weeks (95.3%) in this favorable subgroup. However, our results also showed that treatment was unnecessarily extended beyond 8 weeks in 1833 of 4066 patients in this subgroup, dramatically increasing the cost of treatment without increasing SVR. Our results should offer reassurance to treatment providers that 8 weeks of LDV/SOF monotherapy is sufficient duration in this subgroup.
VA treatment guidelines during the study period designated PrOD as the preferred regimen in genotype 1–infected patients except for prior null responders, those previously treated with protease inhibitors, and patients with Child’s B or C cirrhosis (in whom the preferred regimen was LDV/SOF and ribavirin for 12 weeks) and except for treatment naïve, non-cirrhotics with a viral load <6 million (in whom 8-week LDV/SOF and 12-week PrOD regimens were equally preferred). This was due to the lower cost of 12 weeks of PrOD ($22,850) compared with 12 weeks of LDV/SOF ($37,157) in the VA system during the study period and the absence of evidence that one is more effective than the other in the subgroups for which PrOD was preferred. Our data support the VA treatment recommendations because we found no difference in SVR between PrOD and LDV/SOF regimens in either adjusted or unadjusted analyses, in the entire population or in clinically relevant subgroups (cirrhosis or not, treatment-experienced or naïve). Despite these recommendations and the higher cost, LDV/SOF regimens constituted 77% and PrOD only 23% of regimens in genotype-1–infected patients. This could be due to higher prevalence of drug–drug interactions, higher pill burden, and more frequent requirement for co-prescription of ribavirin in PrOD regimens compared with LDV/SOF regimens.
After the end of the study period, and after FDA approval of elbasvir/grazoprevir as an additional regimen for genotype 1 HCV on January 28, 2016, regimen costs in the VA were further reduced dramatically, to approximately $17,000 per 12-week course for LDV/SOF, PrOD, and elbasvir/grazoprevir and treatment recommendations changed to “equally recommend” all 3 agents as of March 2016.
Genotype 3–infected patients had the lowest SVR rates in our study, just as in clinical trials. We found that the non-FDA–approved regimen of LDV/SOF and ribavirin had a higher SVR rate (77.9%; 95% CI, 73.1%–82.0%) than the longer and more expensive FDA-approved regimen of SOF and ribavirin for 24 weeks (70.6%; 95% CI, 66.9%–74.1%). However, the highest SVR rate in genotype 3–infected patients was observed in the regimen that included PEG together with SOF and ribavirin (87.0%; 95% CI, 80.0%–91.8%), the only interferon-containing regimen that is still recommended.3232
Few large real-world studies of interferon-free regimens are currently available for comparison with ours. The HCV-TARGET, a prospective cohort study of patients undergoing HCV treatment in routine clinical care in academic centers, reported SVR rate to SOF and simeprevir in genotype 1–infected patients of 88% among 151 transplant recipients and 84% among 836 non-transplant recipients.35, 36 This regimen has been superseded by LDV/SOF and PrOD-based regimens. Among 487 patients with decompensated cirrhosis treated in the United Kingdom under an Expanded Access Programme with SOF, LDV/SOF, or daclatasvir, SVR was achieved in 90.5% of genotype 1– and 68.8% of genotype 3–infected patients3737—very similar to our findings. A smaller VA study looked at only treatment-naïve, genotype 1–infected patients treated with LDV/SOF and reported SVR rates almost identical to ours among this subgroup.3838 The TRIO Network, which compiles data from participating “real-world” academic and community HCV treatment clinics in the United States, reported in an abstract an SVR rate of 94% among 1521 genotype 1–infected patients treated with LDV/SOF monotherapy.3939
The main limitation of our study was that SVR data were unavailable in 9% of patients, which can lead to overestimated SVR rates among those with available SVR data. We think this is unlikely for 2 reasons. First, patients with missing SVR data were similar to those with available SVR data (Supplementary Table 2 Supplementary Table 2). Although early discontinuation of treatment in <8 weeks was more common in patients with missing SVR data (25% vs 4.4%), the majority of patients with missing SVR completed 8 or more weeks of treatment, demonstrating that patients with missing SVR data were not patients who “dropped out” of treatment or were “lost to follow-up,” but rather patients (or physicians) who were simply delinquent in getting their SVR viral load measured after the end of their treatment—not an uncommon phenomenon outside of clinical trials. Second, we used comprehensive multiple imputation models that included duration of treatment in addition to baseline, pretreatment characteristics to impute the missing SVR data and found only a non-substantial reduction in SVR after imputation (Table 5Table 5), suggesting that it is unlikely that our results of observed SVR are biased toward overestimation due to the missing SVR data. Important strengths of the study include the complete ascertainment of filled pharmacy prescriptions and the utilization of complete electronic medical records since 1999 from a national health care system that treats the greatest number of HCV-infected patients in the United States.
Our results demonstrate that LDV/SOF, PrOD, and SOF regimens can achieve remarkably high SVR rates in real-world clinical practice, especially in genotype 1–infected patients. The main obstacle to curing HCV infection in the maximum possible number of patients is currently the cost of HCV antiviral regimens. It is expected that cost will decline dramatically as more antiviral regimens become FDA-approved, resulting in competition between manufacturers. In fact, costs decreased dramatically within the VA after the completion of our study and after the FDA approval of elbasvir/grazeprevir in January 2016. The VA health care system has budgeted $1.5 billion nationally for antiviral medications for fiscal year 2016, while every health care organization in the United States is faced with similar budgetary constraints due to the cost of antiviral medications. We hope that our results will be used to determine the most cost-effective ways to treat HCV-infected patients and to reassure patients, clinicians, and health care systems that current treatments for HCV, though costly, appear to be effective in the real-world setting.
Acknowledgments
Author contributions: George Ioannou: Study concept and design, acquisition of data, statistical analysis and interpretation of data, drafting of the manuscript, critical revision of the manuscript, obtained funding. Pamela Green: Analysis of data. Elliott Lowy: Analysis of data. Kristin Berry: Study design and analysis of data. Feng Su: Study design and critical revision of the manuscript. Michael F. Chang: Study design and critical revision of the manuscript. Judith Tsui: Study design and critical revision of the manuscript. Lauren Beste: Study design and critical revision of the manuscript. George Ioannou is the guarantor of this paper. All authors approved the final version of the manuscript.
Article Outline
Article Outline
- Methods
- Results
- Treatment Regimens by Genotype
- Patient Characteristics
- Early Discontinuation of Treatment
- Overall Sustained Virologic Response Rates by Genotype
- Utilization and Sustained Virologic Response Rates of 8-Week Ledipasvir/Sofosbuvir Regimens
- Utilization and Sustained Virologic Response Rates of 24-Week Ledipasvir/Sofosbuvir and Paritaprevir/Ritonavir/Ombitasvir Regimens
- Sustained Virologic Response Rates in Patients With Cirrhosis
- Sustained Virologic Response Rates in Treatment-Experienced vs Treatment-Naïve Patients
- Independent Predictors of Sustained Virologic Response
- Impact of Missing Sustained Virologic Response Data and Imputation for Missing Sustained Virologic Response
- Discussion
- Supplementary Material
- References
