Background

OnabotulinumtoxinA (Ona A; Botox) powder for intramuscular injection solution is a neuromuscular paralytic agent derived from the fermentation of Clostridium botulinum type A and is one of several immunologically distinct serotypes of botulinum neurotoxin.¹ It is indicated for the prophylaxis of headaches in adults with chronic migraine (CM) who have headaches that occur at least 15 days per month and last four hours a day or longer. Ona A is administered as a minimum of 31 injections and a maximum of 39 injections of five Allergan units (U) per injection to the head and neck (total: 155 U to 195 U per administration). The recommended retreatment schedule is every 12 weeks.² Ona A is available in 50 U, 100 U, and 200 U vial sizes, at a cost of $178.50, $357.00, and $714.00 per vial, respectively ($3.57 per U). CADTH Canadian Drug Expert Committee previously reviewed Ona A in 2014 for CM and provided a do-not-list recommendation due to significant limitations of the underlying clinical evidence.³

The manufacturer submitted a cost-utility analysis for this resubmission that was similar to the initial submission in 2014.⁴ In the current submission, the manufacturer compared Ona A with best supportive care (BSC) for the prophylaxis of headaches in adults with CM over a three-year time horizon. The manufacturer also compared Ona A with topiramate in a pairwise scenario analysis to account for changes in clinical practice since the original submission. The Markov model in the resubmission attempted to address some of the previous limitations identified by CADTH by considering additional health care resource use and costs, and added health states to model headache frequency changes in patients who discontinue treatment. The manufacturer modelled 13 health states: six headache frequency health states (0 to 3, 4 to 9, 10 to 14, 15 to 19, 20 to 23, and 24 to 28 headache days per month [HDPM]) for each treatment status (i.e., on treatment or discontinued treatment due to treatment failure) and a death state. Patients entered the model in the various health states based on headache frequency distribution from a pooled analysis of the PREEMPT trials, and transitioned between states every 12 weeks. Similar to the original submission, efficacy and discontinuation data from the PREEMPT trials were pooled and used to inform transition probability for Ona A and BSC; placebo was used as a proxy for BSC. The manufacturer modelled response-based discontinuation based on current clinical practice, in which patients who do not experience a reduction of at least 50% in headache frequency after the initial 24 weeks discontinue treatment. Patients discontinuing Ona A or BSC were assumed to be treated with BSC (patients treated with BSC effectively did not discontinue).

The manufacturer reported that Ona A was associated with an additional cost of $3,430 and generated an additional 0.10 quality-adjusted life-years (QALYs) compared with BSC, resulting in an incremental cost-utility ratio (ICUR) of $34,407 per QALY. At a willingness-to-pay threshold of $50,000 per QALY, the manufacturer reported that Ona A was associated with a 64% probability of being the optimal intervention compared with BSC.

Summary of Identified Limitations and Key Results

Despite the revisions to the manufacturer’s model to address limitations identified in the 2014 review of Ona A, CADTH identified several key limitations with the manufacturer’s economic evaluation, some which remained from the original submission.

The three key limitations were associated with the model structure and assumptions regarding the clinical pathway, the uncertainty associated with the comparative effectiveness data, and the extrapolation of short-term data over a longer time horizon.

First, it was uncertain whether the manufacturer’s model structure sufficiently captured a spectrum of health states that are meaningful to patients with CM, as the headache frequency–based health states used by the manufacturer do not appropriately account for other important aspects of CM, such as headache severity, and the existing literature on these health states was not informative as to whether they are clinically distinct. Although the manufacturer attempted to consider other aspects of CM using treatment-specific health-state utilities, this does not adhere to best practices for economic modelling. The manufacturer’s assumption that patients who improved from CM to episodic migraine (EM) would continue to receive Ona A does not align with current public insurance coverage in Ontario, which indicates that patients should not continue to receive Ona A after improving to EM.⁵ As practice may differ across jurisdictions in Canada, it is uncertain how EM is managed in Canada in patients who previously had CM.

Second, the comparative clinical evidence for Ona A was associated with uncertainty. The placebo injection arm from the PREEMPT trials was used to approximate BSC for the manufacturer’s base-case analysis. However, the validity of this assumption is uncertain as it is unclear what treatments constitute BSC in Canada. The manufacturer used data from the FORWARD trial for the scenario analysis comparison of Ona A with topiramate. The CADTH Clinical Review highlighted several methodological limitations that could have biased the results in favour of Ona A. As such, the incorporation of these data indicates that the cost-effectiveness estimates for Ona A compared with BSC and topiramate are uncertain. Furthermore, the manufacturer did not support its use of subgroup-based transition probabilities with clear clinical evidence, contributing further uncertainty to the treatment impact in the model.

Third, the use of short-term data to inform change in headache frequency was extrapolated over the time horizon, allowing perpetual improvement or worsening. According to the clinical expert consulted by CADTH, once patients are on treatment beyond a year, those who do not respond would not receive further treatment and that patients were unlikely to improve or worsen, indicating a maintenance of effect over time. The clinical expert indicated that in Canadian practice, patients have received Ona A for CM for up to 10 years.

Additional limitations were identified, including application of treatment-specific baseline patient characteristics, inappropriate sourcing and application of adverse event data, a failure to consider drug wastage, outdated emergency department visit costs, and the use of pairwise comparison when a sequential analysis that assessed Ona A with all relevant comparators (BSC, topiramate, and erenumab) should have been explored.

CADTH undertook reanalyses of the manufacturer’s model to address the above limitations where possible. CADTH could not address limitations associated with the model structure and the quality of the comparative evidence. In the CADTH base case, Ona A is associated with an ICUR of $134,601 per QALY compared with BSC. At a willingness to pay $50,000 per QALY, Ona A was associated with a 9% probability of being the optimal intervention. When comparing Ona A with topiramate, Ona A was associated with an ICUR of $28,968 per QALY; however, concerns with the available comparative efficacy information require that these results be viewed with caution.

Conclusions

CADTH identified several key limitations with the manufacturer’s model. Based on a series of reanalyses, CADTH estimated that the ICUR of Ona A compared with BSC was $134,601 per QALY. At a willingness to pay $50,000 per QALY, Ona A was associated with a 9% probability of being the optimal intervention compared with BSC. A price reduction of more than 75% is required for Ona A to achieve an ICUR of less than $50,000 per QALY compared with BSC.

However, given the limitations with the model structure and comparative effectiveness data that could not be adequately addressed in CADTH reanalyses, the cost-effectiveness of Ona A for the prophylaxis of CM remains uncertain.