Results
Long-term Health Outcomes
Table 3 presents the projected long-term health outcomes in terms of HCV-related deaths and advanced liver disease. In treatment-naïve patients, LDV/SOF regimens were associated with the lowest incidence of liver-disease progression, including new cases of DCC, HCC, liver transplants and HCV-related deaths, with the exception of SOF + SMV for 12/24 weeks. For example, the incidence of HCV-related deaths in treatment-naïve patients was 337 per 10 000 with LDV/SOF, but ranged from 330 to 4447 per 10 000 across the comparator regimens. LDV/SOF decreased liver-disease progression by 0–93% compared with current regimens or no treatment. In treatment-experienced patients for whom previously therapy with PR or PI + PR had failed, treatment with LDV/SOF resulted in the lowest incidence of advanced liver disease complications in most of the cases analysed, including patients with and without cirrhosis, with the exception of SOF + SMV. In PI + PR-experienced patients LDV/SOF decreased liver-disease progression by 74% when compared with SOF + PR for 12 weeks, and by 76% when compared with SOF + RBV for 24 weeks.
The NNT outcomes are presented in Table 4. In the treatment-naïve patient sub-cohort, LDV/SOF was associated with the lowest NNT to avoid liver-disease progression against all comparators, with the exception of SOF + SMV 24 weeks in patients with cirrhosis (which performed on a par with LDV/SOF). For example, the NNT to avoid a liver transplant in treatment-naïve patients was 25 for LDV/SOF but ranged from 25 to 43 for current regimens. In treatment-experienced patients for whom previous PR or PI + PR therapy had failed (both with and without cirrhosis), LDV/SOF was generally associated with the lowest NNT in order to avoid liver-disease progression against all treatment comparators.
Short-term Health Economic Outcomes
The 1-year cost per SVR for each treatment regimen is presented in Figure 2. LDV/SOF was associated with the lowest 1-year cost per SVR in all the patient sub-cohorts. For treatment-naïve patients, LDV/SOF resulted in cost savings per SVR ranging from 35% to 227% and 29% to 268% in patients without and with cirrhosis, respectively. In difficult-to-treat PR- or PI + PR-treatment-experienced patients, LDV/SOF resulted in a reduction in cost per SVR of 6–89% and 8–92%, respectively.
(Enlarge Image)
Figure 2.
Short-term 1-year total costs per sustained virologic response for (a) treatment-naïve patients without cirrhosis, (b) treatment-naïve patients with cirrhosis, (c) PI + PR treatment-experienced patients, and (d) PR treatment-experienced patients. BOC, boceprevir; F0–F4, METAVIR liver fibrosis score F0–F4; LVD, ledipasvir; LDV/SOF; ledipasvir/sofosbuvir; PR, pegylated interferon + ribavirin; RBV, ribavirin; SMV, simeprevir; SOF, sofosbuvir; SVR, sustained virologic response. LDV/SOF regimens included both LDV/SOF for 8 weeks (60% of patients) and LDV/SOF for 12 weeks (40% of patients) in treatment-naïve patients, and LDV/SOF for 12 weeks (F0–F3 patients) and LDV/SOF for 24 weeks (F4 patients only; 17% of patients) in treatment-experienced patients.
Long-term Health Economic Outcomes
Long-term health economic outcomes are presented in Table 5. Treatment-naïve patients who received LDV/SOF lived 0.31, 0.70, 1.16 and 1.11 years longer than those treated with SOF + PR for 12 weeks, SMV for 12 weeks + PR for 24 weeks, SOF + RBV for 24 weeks, and BOC + PR (variable duration), respectively. Patients treated with SOF + SMV for 12 or 24 weeks lived 0.03 years longer than those treated with LDV/SOF. Patients who had previously been treated with PI + PR lived 0.57 and 0.66 years longer when treated with LDV/SOF than with SOF + PR for 12 weeks and with SOF + RBV for 24 weeks, respectively.
LDV/SOF dominated (i.e., was less costly and more effective than) comparator regimens in most of the cases analysed for treatment-naïve and treatment-experienced patients. An exception to this was the comparison with SOF + SMV for 12 or 24 weeks, in which LDV/SOF was less costly and less effective (incremental QALY −0.01 and incremental cost −$101 504 in treatment-naïve patients; incremental QALY −0.02 and incremental cost −$63 609 in treatment-experienced patients). Although LDV/SOF did not dominate SOF + SMV, it is a cost-effective option, with ICERs greater than $50 000 per QALY gained. When comparing patients by disease severity at the initiation of treatment (i.e., patients without vs. those with cirrhosis) for the same regimen, total costs were 25–49% lower among patients without cirrhosis than among those who had cirrhosis.
Impact of Treatment Initiation by Fibrosis Stage in Treatment-naïve Patients
One-year costs per SVR, long-term health economic outcomes and long-term health outcomes at different stages of liver fibrosis are presented in Tables S6–S8 http://onlinelibrary.wiley.com/store/10.1111/apt.13081/asset/supinfo/apt13081-sup-0001-TableS1-S9.docx?v=1&s=2d42b5c1823a07fa2699dec9a9a267d755ea8c40, respectively. Overall, treating patients at an earlier stage of liver fibrosis resulted in lower costs per SVR (S6). Total costs per SVR of LDV/SOF in patients with METAVIR fibrosis score F0–F1, F2 and F3–F4 were $82 152, $82 399 and $90 878, respectively. Treating patients with METAVIR fibrosis score F0–F1 and F2 with LDV/SOF resulted in a reduction in cost of $8479 and $8726 per SVR, respectively, compared with treating those with score F3–F4 (cost of $90 878 per SVR), a cost reduction of between 9% and 10% per SVR.
Similarly, treating patients at an earlier stage of liver fibrosis resulted in an improvement in life-years and QALYs gained (Table S7 http://onlinelibrary.wiley.com/store/10.1111/apt.13081/asset/supinfo/apt13081-sup-0001-TableS1-S9.docx?v=1&s=2d42b5c1823a07fa2699dec9a9a267d755ea8c40). Treating patients with METAVIR fibrosis score F0–F1 or F2 with LDV/SOF resulted in a gain of 0.61 and 0.53 life-years, as well as a gain of 0.71 and 0.64 QALYs, respectively, compared with treating those with score F3–F4. Treatment with LDV/SOF in patients with METAVIR fibrosis score F0–F1 and F2 vs. score F3–F4 resulted in a reduction of lifetime costs per patient of $17 980 and $16 524, a decrease of 18% and 16% per patient, respectively. In patients with METAVIR fibrosis score F0–F1, the total lifetime costs were $83 440 with LDV/SOF and $116 354 with no treatment (28% higher with no treatment than with LDV/SOF). In patients with score F2 these costs were $84 896 and $139 991 (39% higher with no treatment than with LDV/SOF), and in those with score F3–F4 they were $101 420 and $170 984 (41% higher with no treatment than with LDV/SOF), respectively (Table S7 http://onlinelibrary.wiley.com/store/10.1111/apt.13081/asset/supinfo/apt13081-sup-0001-TableS1-S9.docx?v=1&s=2d42b5c1823a07fa2699dec9a9a267d755ea8c40). Regardless of the stage at which treatment was initiated, treatment with LDV/SOF dominated no treatment from a cost-effectiveness perspective. For all treatment-naïve patients, the total lifetime costs associated with LDV/SOF were $90 127 and with no treatment they were $141 856 (Table 5).
Overall, treatment initiation at an earlier stage of liver fibrosis resulted in improved downstream outcomes (Table S8 http://onlinelibrary.wiley.com/store/10.1111/apt.13081/asset/supinfo/apt13081-sup-0001-TableS1-S9.docx?v=1&s=2d42b5c1823a07fa2699dec9a9a267d755ea8c40). Treating patients with LDV/SOF resulted in a 66% reduction of overall liver-disease-related complications when patients were treated METAVIR fibrosis score F2 compared with score F3–F4, and by 82% when patients were treated at score F0–F1 compared with score F3–F4. LDV/SOF resulted in the most significant reductions in liver-disease-related progression across all strategies for treatment initiation regardless of fibrosis stage at which treatment commenced (i.e., treatment initiated at F0–F1, F2 or F3–F4).
Sensitivity and Scenario Analyses
One-way deterministic sensitivity analysis is presented in Figure 3. This shows the top 20 inputs that most heavily influenced the ICER values for LDV/SOF compared with SOF + PR for 12 weeks among treatment-naïve patients. ICERs remained under the $50 000 per QALY gained WTP threshold, indicating that LDV/SOF remains a cost-effective strategy even when clinical, cost and natural history parameters are varied across wide yet plausible ranges. These results are similar to those obtained when LDV/SOF was compared with no treatment. Similar trends were observed for other sub-cohorts of patients across fibrosis distributions, levels of treatment experience, and treatment comparisons. Results from the probabilistic sensitivity analysis for treatment-naïve patients are presented in Figure 4. At a WTP threshold of $50 000, LDV/SOF for 12 weeks in treatment- naïve patients was 98% likely to be cost-effective compared with SOF + PR for 12 weeks and 100% likely to be cost-effective compared with no treatment (Figure 4).
(Enlarge Image)
Figure 3.
Deterministic sensitivity analysis for treatment-naïve patients: lifetime incremental cost per QALY gained for LDV/SOF for 12 weeks compared with SOF + PR for 12 weeks. Inputs varied and associated distributions are detailed in Table S4. CC, compensated cirrhosis; DCC, decompensated cirrhosis; EM, early mortality; F0–F4, METAVIR liver fibrosis score F0–F4; HCC, hepatocellular carcinoma; LDV/SOF; ledipasvir/sofosbuvir; NC, non-cirrhotic; PR, pegylated interferon + ribavirin; QALY, quality-adjusted life-year; SOF; sofosbuvir; SVR, sustained virologic response; TP, transition probabilities.
(Enlarge Image)
Figure 4.
Probabilistic sensitivity analysis for treatment-naïve patients. Lifetime incremental cost per QALY for LDV/SOF for 12 weeks compared with SOF + PR for 12 weeks: (a) cost-effectiveness acceptability curve and (b) incremental cost-effectiveness plane. Lifetime incremental cost per QALY for LDV/SOF for 12 weeks compared with no treatment: (c) cost-effectiveness acceptability curve and (d) incremental cost-effectiveness plane. LDV/SOF; ledipasvir/sofosbuvir; PR, pegylated interferon + ribavirin; QALY, quality-adjusted life-year; SOF, sofosbuvir; WTP, willingness-to-pay.
Scenario analyses for treatment-naïve patients treated with SOF + SMV based on real-world data and treatment-experienced patients with cirrhosis treated with LDV/SOF + RBV for 12 weeks are presented in Table S9 http://onlinelibrary.wiley.com/store/10.1111/apt.13081/asset/supinfo/apt13081-sup-0001-TableS1-S9.docx?v=1&s=2d42b5c1823a07fa2699dec9a9a267d755ea8c40. Compared with the real world data for SOF + SMV (which included lower SVRs relative to clinical trial data [COSMOS]) LDV/SOF dominated SOF+SMV. Compared with LDV/SOF 24 week and SOF + SMV 24 week regimens, treatment with LDV/SOF + RBV 12 weeks was slightly less effective but also less costly, and therefore the more cost-effective option.