The manufacturer submitted a cost-utility analysis comparing asfotase alfa (Strensiq) to best supportive care (BSC, defined as the need for surgical interventions, hospitalizations, intensive care unit services, respiratory assistance, outpatient visits, consultations, and pain medication) in patients diagnosed with pediatric-onset hypophosphatasia (HPP).² The analysis was conducted under the Canadian publicly funded health care system perspective.²

The model consists of six health states, four of which were defined by the severity level of disease, based on the observed over the predicted six-minute walk test (6MWT) score; death by HPP; and death by other causes. Invasive ventilator is also included as a toll state (i.e., a temporary health state associated with disutility and additional costs). In the manufacturer’s base case, patients entered the model at an age of 5.8 years at any of the severity health states, and transitioned between these states every 12 weeks over a lifetime time horizon (101 years).² The health states and the baseline distribution were defined as follows (based on 28 patients for whom there were 6MWT data available, from studies ENB-006-09/ENB-008-10 and ENB-009-10):^4,5

• Severity level I: (observed 6MWT/predicted 6MWT) > 82.2% (22.2 % of patients)
• Severity level II: 82.2% ≥ (observed 6MWT/predicted 6MWT) > 64.4% (33.3% of patients)
• Severity level III: 64.4% ≥ (observed 6MWT/predicted 6MWT) > 46.6% (29.6 % of patients)
• Severity level IV: (observed 6MWT/predicted 6MWT) > 46.6% or those who could not complete the 6MWT (14.8% of patients)
• HPP-related death (0% of patients)
• Background death, not related to HPP (0% of patients)
• Invasive ventilator toll state (0% of patients).

In the model, patients progress between severity health states and can transition to death or the invasive ventilator toll state at any time point. As there were no severity categories based on the 6MWT outcome in HPP, the manufacturer defined the severity levels using an approach previously used in patients with Duchenne muscular dystrophy,^2,9 although there is no clear evidence to support this. In the manufacturer’s analysis, the severity of HPP was considered to be age-dependent. The probability of transitioning between severity states was based on 6MWT data from 28 patients who received asfotase alfa or BSC. These data were obtained from the following open-label, phase II clinical trials assessing asfotase alfa with a historical or no-treatment concurrent control (where 6MWT was an exploratory or secondary outcome): ENB-006-09 and its extension (ENB-008-10), which assessed patients aged five to 12 years; and ENB-009-10, which assessed patients aged 13 to 66 years.^4,5 The manufacturer estimated an ordered probit regression model that allowed for the calculation of age-specific transition probabilities, based on 6MWT data from the above-mentioned trials, allowing for predictions for patients younger than five years and older than 65 years for whom 6MWT data are not available. Additionally, when a patient is in the invasive ventilator toll state, the manufacturer assumed a health utility decrement and additional direct medical costs. Furthermore, all patients who require an invasive ventilator were assumed to transition to the most severe health state afterward (severity level IV). The transition probabilities for HPP-related death and the invasive ventilator health state were taken from the following studies assessing patients younger than five years: asfotase alfa studies ENB-002-08 and its extension (ENB-003-08), and ENB-010-10; and a natural history study, ENB-011-10.^3,6,7 In the manufacturer’s base case (where patients started at 5.8 years old), the probability of transitioning into the HPP-related death or the invasive ventilator health state was 0% after the age of five years, as both of these outcomes were not observed past the age of four in the trials.

The manufacturer conducted a utility elicitation study to determine utility weights for each of the health states. In summary, case histories were created based on two natural history studies (ENB-011-10 and ALX-HPP-502).^7,8 These case histories were then summarized into descriptions of the health states, with input from clinical experts. Following this, the clinical experts were asked to rank each description in order of severity, as related to the 6MWT, and then rate each of the descriptions using the EuroQol 5-Dimension 5 Levels Questionnaire (EQ-5D-5L).

Resource utilization was determined by estimating the frequency of clinical events expected in addition to the background care level needed for each of the health states, with input from a clinical expert. Costs information was primarily obtained from the Ontario Schedule of Benefits for Physician Services (2015),¹⁰ among other sources.² The manufacturer assumed that the cost of wastage associated with partially used vials of asfotase alfa would not be incurred by the publicly funded health care system, although nothing was stated regarding who would incur this cost. Additionally, the manufacturer assumed that 10 years from the start of the model, loss of data exclusivity will lead to a 30% decrease in the list price of asfotase alfa.²