Cost-Effectiveness of Sofosbuvir-Based Treatments for HCV
Cost-Effectiveness of Sofosbuvir-Based Treatments for HCV
Treatments aimed at the same group of patients (genotype, existence of cirrhosis) are compared. We consider results categorized by whether cirrhosis exists. The results for patients without cirrhosis are shown in Table 8 , and the results for patients with cirrhosis are shown in Table 9 . Note that the treatments are sorted according to cost in ascending order. In both tables, the incremental cost-effectiveness ratio (ICER) is calculated as the additional cost divided by additional effectiveness between each treatment and the benchmark treatment. Specifically, all treatments are first compared to the standard of care treatment (e.g., standard of care treatment for genotype 1 is Telaprevir + Peginterferon + Ribavirin), and then compared to the adjacent efficient treatment. For example, in Table 8 for genotype 2, the Adjacent ICER between two-phase treatment with 24 + 12 versus Peginterferon + Riavirin is 4233. The Adjacent ICER between two-phase treatment with 24 + 16 versus two-phase treatment with 24 + 12 is 44,458. The Adjacent ICER between Sofobuvir + Ribavirin versus two-phase treatment with 24 + 16 is 1,805,952.
For genotype 1 patients without cirrhosis, compared to the acknowledged efficient benchmark treatment (peginterferon + ribavirin + telaprevir), all new treatments are cost-effective with ICER less than the threshold of $50,000/QALY. In particular, treatments Harvoni and Viekira Pak both achieve higher QALYs with reduced costs, which makes the benchmark treatment no longer efficient, whereas treatments olysio + sovaldi and sofosbuvir + peginterferon + ribavirin both achieve higher QALYs but with increased costs compared to benchmark treatment. However, if compared to Viekira + Pak, Harvoni, olysio + sovaldi and sofosbuvir + peginterferon + ribavirin are no longer efficient characterized with higher costs and lower QALYs. Thus, we conclude that all four new regimens are alternatives of the current standard care of treatment (peginterferon + ribavirin + telaprevir), but Viekira Pak is the most cost-effective for genotype 1 patients without cirrhosis, whereas the other three sofosbuvir-based treatments are featured with higher costs and lower QALYs. For genotype 2 treatments, compared to standard care of treatment (peginterferon + ribavirin), all three sofosbuvir-based treatments are cost-effective. However, the comparative ICER of two-phase treatment with a 16-week follow-up versus 12-week single treatment of sofosbuvir + ribavirin is $1,805,952/QALY, which is far beyond the threshold. Thus, for genotype 2 the two-phase treatments with peginterferon + ribavirin as initial and 12 and 16 weeks of sofosbuvir as follow-up are cost-effective, whereas single treatment with sofosbuvir + ribavirin is not. For genotype 3, similarly, the two-phase sofosbuvir-based treatments are cost-effective compared to the standard care of treatment; note that ICERs between adjacent treatments are also below the threshold. The 24-week single treatment with sofosbuvir + ribavirin is not cost-effective compared to standard care of treatment with an ICER of $88,833/QALY. Further, it is not cost-effective compared to two-phase treatment with a 16-week follow-up with an ICER of $374,594/QALY. Overall, single-phase treatment with sofosbuvir for both genotype 2 and 3 patients dominates two-phase treatments using peginterferon + ribavirin regimen as initial and sofosbuvir-based new treatment as follow-up. The reason is attributable to the fact that the traditional regimen (peginterferon + ribavirin) has a relatively high SVR for genotypes 2 and 3, and the new treatment's additional SVR increase does not justify its much higher cost.
When patients with cirrhosis are considered, for genotype 1 all new treatments are efficient compared to the standard care of treatment, whereas adjacent ICERs show that sofosbuvir + peginterferon + ribavrin and Viekira Pak are both inefficient compared to Harvoni with higher costs and lower QALYs. Although olysio + sovaldi has higher cost as well as higher QALYs compared to Harvoni, the ICER is $1,305,213/QALY, which is far beyond the threshold and hence not efficient. Therefore, for genotype 2 patients with cirrhosis, Harvoni is the most cost-effective treatment. For genotype 2, both two-phase treatments with 12-week and 16-week follow-ups are cost-effective, whereas 12-week single treatment with sofosbuvir + ribavirin is not cost-effective. For genotype 3 treatments, a 24-week single treatment with sofosbuvir + ribavirin is cost-effective, as well as the two-phase treatment with a 16-week follow-up.
One-way sensitivity analyses are performed on prices, transition probabilities, SVR rates, utility weights and costs of care. Only when the sofosbuvir price is reduced by at least 30 %, olysio + sovaldi and sofosbuvir + peginterferon + ribavirin can achieve cost-effectiveness compared to Harvoni or Viekira Pak for genotype 1 patients. However, even when the sofosbuvir price is halved, using sofosbuvir as initial treatment is still not cost-effective compared to two-phase treatments that use sofosbuvir as a follow-up treatment for genotypes 2 and 3. Further, increasing SVR for sofosbuvir-based treatments by 10 % does make them cost-effective compared to Harvoni and Viekira Pak for genotype 1 patients, and reducing SVR for Harvoni and Viekira Pak by 10 % also makes sofosbuvir-based treatments cost-effective.
Changing SVR rates of sofosbuvir-based treatments for genotypes 2 and 3 does not change the effectiveness of the treatments. Two-phase treatments of 24 + 12 weeks are always cost-effective for both genotypes 2 and 3, compared to which, a change in the values of cost of care or utility weights can push 24 + 16 weeks treatments' ICER beyond or below $50,000/QALY when benchmarked with the base case. However, it does not change the conclusion that single phase treatment of sofosbuvir + ribavirin is not cost-effective for both genotypes 2 and 3. Thus, for genotype 1 patients, a reduction in sofosbuvir price can improve the cost-effectiveness of sofosbuvir-based treatments compared to the alternatives of Harvoni and Viekira Pak. For genotypes 2 and 3, however, sofosbuvir-based new treatments serve better as follow-up treatments rather than initial treatments.
Probabilistic analysis results are shown via the cost-effectiveness acceptability curves (Fig. 1), which are interpreted as the probability that the data are consistent within a true cost-effectiveness ratio falling below that value between two treatments. The figure that the ICER of olysio + sovaldi versus Harvoni falls either below 0 or above $50,000/QALY with approximate probability 0.95, implying that compared to Harvoni, olysio + sovaldi either achieves lower QALYs with higher costs or achieves higher QALYs with higher cost but with ICER exceeding the threshold $50,000/QALY. Thus, with 95 % confidence we conclude that olysio + sovaldi is not cost-effective. For Viekira Pak and sofosbuvir + peginterferon + ribavirin, compared to Harvoni, they both have roughly a 0.50 probability of being effective according to figure. For both genotypes 2 and 3, the ICER of two-phase treatment with peginterferon + ribavirin as initial and 12-week sofosbuvir as follow-up versus the standard of treatment is below $10,000/QALY with 0.95 probability according to the figure, whereas two-phase treatment with peginterferon + ribavirin as initial, a 16-week sofosbuvir as follow-up and single treatment with sofosbuvir + ribavirin both have equal probability of being effective and not effective for genotypes 2 and 3.
(Enlarge Image)
Figure 1.
Cost-effectiveness Acceptability Curves between Treatments. Olysio + Sovaldi versus Harvoni for Genotype 1 (Subfigure 1); Viekira Pak versus Harvoni for Genotype 1 (Subfigure 2); Sofosbuvir + Peginterferon + Ribavirin versus Harvoni for Genotype 1 (Subfigure 3); Two-phase of 24 + 12 versus Peginterferon + Ribavirin for Genotype 2 (Subfigure 4); Two-phase of 24 + 16 versus Two-phase of 24 + 12 for Genotype 2 (Subfigure 5); Sofosbuvir + Ribavirin versus Two-phase of 24 + 16 for Genotype 2 (Subfigure 6); Two-phase of 24 + 12 versus Peginterferon + Ribavirin for Genotype 3 (Subfigure 7); Two-phase of 24 + 16 versus Two-phase of 24 + 12 for Genotype 3 (Subfigure 8); Sofosbuvir + Ribavirin versus Two-phase of 24 + 16 for Genotype 3 (Subfigure 9)
Results
Base Case Analysis
Treatments aimed at the same group of patients (genotype, existence of cirrhosis) are compared. We consider results categorized by whether cirrhosis exists. The results for patients without cirrhosis are shown in Table 8 , and the results for patients with cirrhosis are shown in Table 9 . Note that the treatments are sorted according to cost in ascending order. In both tables, the incremental cost-effectiveness ratio (ICER) is calculated as the additional cost divided by additional effectiveness between each treatment and the benchmark treatment. Specifically, all treatments are first compared to the standard of care treatment (e.g., standard of care treatment for genotype 1 is Telaprevir + Peginterferon + Ribavirin), and then compared to the adjacent efficient treatment. For example, in Table 8 for genotype 2, the Adjacent ICER between two-phase treatment with 24 + 12 versus Peginterferon + Riavirin is 4233. The Adjacent ICER between two-phase treatment with 24 + 16 versus two-phase treatment with 24 + 12 is 44,458. The Adjacent ICER between Sofobuvir + Ribavirin versus two-phase treatment with 24 + 16 is 1,805,952.
For genotype 1 patients without cirrhosis, compared to the acknowledged efficient benchmark treatment (peginterferon + ribavirin + telaprevir), all new treatments are cost-effective with ICER less than the threshold of $50,000/QALY. In particular, treatments Harvoni and Viekira Pak both achieve higher QALYs with reduced costs, which makes the benchmark treatment no longer efficient, whereas treatments olysio + sovaldi and sofosbuvir + peginterferon + ribavirin both achieve higher QALYs but with increased costs compared to benchmark treatment. However, if compared to Viekira + Pak, Harvoni, olysio + sovaldi and sofosbuvir + peginterferon + ribavirin are no longer efficient characterized with higher costs and lower QALYs. Thus, we conclude that all four new regimens are alternatives of the current standard care of treatment (peginterferon + ribavirin + telaprevir), but Viekira Pak is the most cost-effective for genotype 1 patients without cirrhosis, whereas the other three sofosbuvir-based treatments are featured with higher costs and lower QALYs. For genotype 2 treatments, compared to standard care of treatment (peginterferon + ribavirin), all three sofosbuvir-based treatments are cost-effective. However, the comparative ICER of two-phase treatment with a 16-week follow-up versus 12-week single treatment of sofosbuvir + ribavirin is $1,805,952/QALY, which is far beyond the threshold. Thus, for genotype 2 the two-phase treatments with peginterferon + ribavirin as initial and 12 and 16 weeks of sofosbuvir as follow-up are cost-effective, whereas single treatment with sofosbuvir + ribavirin is not. For genotype 3, similarly, the two-phase sofosbuvir-based treatments are cost-effective compared to the standard care of treatment; note that ICERs between adjacent treatments are also below the threshold. The 24-week single treatment with sofosbuvir + ribavirin is not cost-effective compared to standard care of treatment with an ICER of $88,833/QALY. Further, it is not cost-effective compared to two-phase treatment with a 16-week follow-up with an ICER of $374,594/QALY. Overall, single-phase treatment with sofosbuvir for both genotype 2 and 3 patients dominates two-phase treatments using peginterferon + ribavirin regimen as initial and sofosbuvir-based new treatment as follow-up. The reason is attributable to the fact that the traditional regimen (peginterferon + ribavirin) has a relatively high SVR for genotypes 2 and 3, and the new treatment's additional SVR increase does not justify its much higher cost.
When patients with cirrhosis are considered, for genotype 1 all new treatments are efficient compared to the standard care of treatment, whereas adjacent ICERs show that sofosbuvir + peginterferon + ribavrin and Viekira Pak are both inefficient compared to Harvoni with higher costs and lower QALYs. Although olysio + sovaldi has higher cost as well as higher QALYs compared to Harvoni, the ICER is $1,305,213/QALY, which is far beyond the threshold and hence not efficient. Therefore, for genotype 2 patients with cirrhosis, Harvoni is the most cost-effective treatment. For genotype 2, both two-phase treatments with 12-week and 16-week follow-ups are cost-effective, whereas 12-week single treatment with sofosbuvir + ribavirin is not cost-effective. For genotype 3 treatments, a 24-week single treatment with sofosbuvir + ribavirin is cost-effective, as well as the two-phase treatment with a 16-week follow-up.
Sensitivity Analysis
One-way sensitivity analyses are performed on prices, transition probabilities, SVR rates, utility weights and costs of care. Only when the sofosbuvir price is reduced by at least 30 %, olysio + sovaldi and sofosbuvir + peginterferon + ribavirin can achieve cost-effectiveness compared to Harvoni or Viekira Pak for genotype 1 patients. However, even when the sofosbuvir price is halved, using sofosbuvir as initial treatment is still not cost-effective compared to two-phase treatments that use sofosbuvir as a follow-up treatment for genotypes 2 and 3. Further, increasing SVR for sofosbuvir-based treatments by 10 % does make them cost-effective compared to Harvoni and Viekira Pak for genotype 1 patients, and reducing SVR for Harvoni and Viekira Pak by 10 % also makes sofosbuvir-based treatments cost-effective.
Changing SVR rates of sofosbuvir-based treatments for genotypes 2 and 3 does not change the effectiveness of the treatments. Two-phase treatments of 24 + 12 weeks are always cost-effective for both genotypes 2 and 3, compared to which, a change in the values of cost of care or utility weights can push 24 + 16 weeks treatments' ICER beyond or below $50,000/QALY when benchmarked with the base case. However, it does not change the conclusion that single phase treatment of sofosbuvir + ribavirin is not cost-effective for both genotypes 2 and 3. Thus, for genotype 1 patients, a reduction in sofosbuvir price can improve the cost-effectiveness of sofosbuvir-based treatments compared to the alternatives of Harvoni and Viekira Pak. For genotypes 2 and 3, however, sofosbuvir-based new treatments serve better as follow-up treatments rather than initial treatments.
Probabilistic analysis results are shown via the cost-effectiveness acceptability curves (Fig. 1), which are interpreted as the probability that the data are consistent within a true cost-effectiveness ratio falling below that value between two treatments. The figure that the ICER of olysio + sovaldi versus Harvoni falls either below 0 or above $50,000/QALY with approximate probability 0.95, implying that compared to Harvoni, olysio + sovaldi either achieves lower QALYs with higher costs or achieves higher QALYs with higher cost but with ICER exceeding the threshold $50,000/QALY. Thus, with 95 % confidence we conclude that olysio + sovaldi is not cost-effective. For Viekira Pak and sofosbuvir + peginterferon + ribavirin, compared to Harvoni, they both have roughly a 0.50 probability of being effective according to figure. For both genotypes 2 and 3, the ICER of two-phase treatment with peginterferon + ribavirin as initial and 12-week sofosbuvir as follow-up versus the standard of treatment is below $10,000/QALY with 0.95 probability according to the figure, whereas two-phase treatment with peginterferon + ribavirin as initial, a 16-week sofosbuvir as follow-up and single treatment with sofosbuvir + ribavirin both have equal probability of being effective and not effective for genotypes 2 and 3.
(Enlarge Image)
Figure 1.
Cost-effectiveness Acceptability Curves between Treatments. Olysio + Sovaldi versus Harvoni for Genotype 1 (Subfigure 1); Viekira Pak versus Harvoni for Genotype 1 (Subfigure 2); Sofosbuvir + Peginterferon + Ribavirin versus Harvoni for Genotype 1 (Subfigure 3); Two-phase of 24 + 12 versus Peginterferon + Ribavirin for Genotype 2 (Subfigure 4); Two-phase of 24 + 16 versus Two-phase of 24 + 12 for Genotype 2 (Subfigure 5); Sofosbuvir + Ribavirin versus Two-phase of 24 + 16 for Genotype 2 (Subfigure 6); Two-phase of 24 + 12 versus Peginterferon + Ribavirin for Genotype 3 (Subfigure 7); Two-phase of 24 + 16 versus Two-phase of 24 + 12 for Genotype 3 (Subfigure 8); Sofosbuvir + Ribavirin versus Two-phase of 24 + 16 for Genotype 3 (Subfigure 9)
Source...