3.3.1. Riociguat Versus Placebo
a. Functional Class Transition Probabilities
Transition probabilities between functional classes (FC II to FC III, FC III to FC II, FC III to FC IV, FC IV to FC III) at different cycles of the model were tested. When the lower and upper 95% CIs of the OR for transition for riociguat in the first 16 weeks are used (assuming CI bounds are 20% above and below mean), the ICUR is between $172,985 to $174,392 per QALY. When ± 20% is used after the first 16 weeks, the ICUR is between $173,204 to $174,093 per QALY. When ± 20% is used for both periods, the ICUR is between $173,652 to $174,294 per QALY. When it is assumed there is no change in transition probabilities after two years (patients transition through FC health states in same manner), the ICUR increases to $209,518 per QALY.
b. Mortality
In the manufacturer’s economic model, mortality varies by FC health state, as well as by treatment. If the risk of mortality is fully reflected by increases in mortality by FC health state (additional mortality benefit is excluded: OR for riociguat versus placebo is set to unity, and same mortality risk applied for each treatment within the same FC), the ICUR increases to $350,519 per QALY. If same mortality is assumed in all FC health states, and the OR for mortality from the ITC is used, the ICUR increases to $187,397 per QALY.
c. Titration Cost
There were four nursing visits for treatment initiation with riociguat in the CCA, but only one visit in the CUA model. If 4 nursing visits are used in the model, the ICUR increases to $173,590 per QALY.
3.3.2. Riociguat Versus Generic Bosentan
FC Transition Probabilities
When ± 20% of the OR for transition for riociguat in the first 16 weeks is explored, the ICUR ranges between $185,841 to $188,335 per QALY. When the ± 20% is used after the first 16 weeks, the ICUR is between $179,939 to $196,626 per QALY. When the ± 20% is used for both periods, the ICUR is between $180,741 to $194,654 per QALY.
a. Mortality
If the mortality risk is fully captured by FC health state (the OR for mortality is set to unity for riociguat versus bosentan, and the same mortality risk applied for each treatment within the same FC), generic bosentan is the dominant strategy (less costly and more effective). If mortality is equal across all FC health states and the OR of mortality from the ITC is used, the ICUR decreases to $134,865 per QALY.
b. Titration Cost
There were four nursing visits for treatment initiation with riociguat in the CCA, but only one visit in the CUA model. If 4 nursing visits are used in the model, the ICUR increases to $187,486 per QALY.
3.3.3. Riociguat Versus Tracleer
a. Discontinuation Rate
It is not clear that discontinuation rates are truly different between riociguat and Tracleer (ITC reported discontinuation due to AE with a credible interval (CrI) overlapping unity: OR riociguat versus Tracleer: 2.53; 95% CrI, 0.23 to 27.73). If set to be equal (3.8%), there is no longer cost savings associated with riociguat, and incremental costs increase to $47,358, resulting an ICUR of $49,099 per QALY.
b. Mortality
If the mortality risk is fully captured by FC health state (the OR for mortality set to unity for riociguat versus Tracleer and same mortality risk applied for each treatment within the same FC), riociguat is less costly, but less effective than Tracleer. The ICUR for Tracleer is $227,861 per QALY when compared with riociguat.
3.3.4. Riociguat Versus Mix of Generic Bosentan and Tracleer
Even if bosentan is available in generics, many provincial drug plans still cover Tracleer under exceptional access program (please refer to the Drug Plan Benefit Listings Table). According to the clinical expert, patients and physicians with CTEPH and PAH are reluctant to switch from Tracleer to generics because they are concerned by the potential variability in bioavailability between generics and lack of patient-support program. A CADTH Common Drug Review (CDR) analysis of utilization data from public plans (except Quebec) showed that, the proportion of claims for bosentan that consisted of Tracleer varied widely across provinces, ranging from 0% to 96% of claims. (PharmaStat data from IMS Health Canada Inc., 2013). However, the PharmaStat data does not differentiate the indication; claims above likely include PAH patients as well as CTEPH patients ().
Proportion of Claims for Tracleer Across Public Plans.
There were no data available for Alberta and Manitoba. According to the clinical expert, approximately 100% in Alberta and 80% of patients in Manitoba are on Tracleer.
The annual drug cost per patient of riociguat versus different mix of generic bosentan (Tracleer) is listed in .
Annual Drug Costs of Riociguat Versus Different Mix of Generic Bosentan (Tracleer).
CDR also estimated the cost per QALY with a different mix of patients on generic versus brand-name drugs based on the manufacturer’s base-case results ().
CADTH Common Drug Review Reanalysis of Incremental Cost-Utility Ratios for Riociguat Versus Bosentan for Varying Utilization of Tracleer.
Although the drug costs of riociguat and Tracleer are similar (daily cost of $128.25 versus $128.36, respectively), there are additional monitoring costs associated with Tracleer (one additional liver-function test per month at an additional cost of $225.84 annually). As such, assuming equal efficacy between the two drugs (as true relative efficacy is unknown), riociguat is slightly less costly than Tracleer.
A summary of key CDR reanalyses is presented in . The CDR reanalyses, which accounted for potential double counting of the benefit of riociguat on mortality, showed that the base-case analysis submitted by the manufacturer likely underestimated the ICUR of riociguat compared with placebo or generic bosentan.
Summary of CADTH Common Drug Review Reanalyses of Incremental Cost-Utility Ratios for Riociguat Versus Placebo and Riociguat Versus Generic Bosentan.
Several price-reduction scenarios were explored using the manufacturer’s base-case analysis for riociguat compared with placebo and generic bosentan, respectively (). Of note, when the CDR reanalysis — which accounted for potential double counting of the benefit of riociguat on mortality — is used as a reference, a price reduction of more than 80% would be needed for the ICUR of riociguat compared with placebo to fall below $50,000 per QALY, and a price reduction of more than 60% would be needed for riociguat to be less costly (but still less effective) than generic bosentan.
CADTH Common Drug Review Analysis of Incremental Cost-Utility Ratios Based on Various Price-Reduction Scenarios.