Amgen Response to ICER’s Draft Scoping Document on CGRP Preventive Therapy for Migraine

SUMMARY OVERVIEW

Amgen and Novartis appreciate the opportunity to comment on ICER’s draft scoping document for Calcitonin Gene-Related Peptide (CGRP) Inhibitors as preventive therapy in migraine.  Migraine is a serious debilitating disease which is overdue for innovation. A fair assessment of new migraine preventive treatments is incredibly important to patients, and needs to be framed carefully based on the unique patient experience, disease burden, and impact to society. Based on our review of ICER’s draft scope, our recommendations are as follows:

1) Employ a patient-centered approach – the base-case should be the patient and employer perspective that factors in the missed work days and reduced productivity of migraine patients, as opposed to a payer perspective.

2) Assess erenumab only in treatment-experienced patients – those who have tried and failed topiramate or propranolol as opposed to treatment-naïve patients.

3) Factor in treatment discontinuation rates with currently available preventives in the analysis and in the impact on efficacy.

4) Incorporate responder rates i.e., a fifty percent responder rate assesses the proportion of patients who saw their migraine days reduced by half or more.

5) Apply full continuous modelling techniques rather than a fixed health state model to more accurately capture the complexities of migraine burden and treatment benefit.

A more comprehensive discussion of these recommendations is below.

KEY RECOMMENDATIONS

ICER’s healthcare system perspective should not just reflect the payer perspective but be inclusive of all those who are incurring costs due to migraine, most notably the patient and employer.  The adoption of a payer perspective, as opposed to a broader societal perspective, is at odds with established accepted methodologies in the economic evaluation of new treatments.^1,2,3,4 The gold standard for health economic assessment methodology, the Second Panel on Cost-Effectiveness in Health and Medicine, recommends that all cost-effective analyses capture both the healthcare payer and the societal perspective (in this case, societal is defined as all costs incurred by society due to migraine, including the often overlooked costs to patients).^5,6   These include, for example, patient co-payments for treatment, employer costs, and providers costs of services and transactions not covered by insurance. Moreover, there are higher overall total costs associated with this debilitating disease that impact the healthcare system including the  impact on patient physical, psychological and social lives which are not captured in an assessment that is only payer centric. Patients with migraine have much higher average healthcare expenditures averaging 1.5 times matched controls.⁷  While these are high costs to the healthcare system, they are unlikely to be captured in ICER’s base-case. ICER’s current base-case leaves out one of the largest healthcare payers of migraine costs, the self-insured employer. Self-insured plans cover 83% of all employees at private sector companies with >200 employees.⁸ Each employee with frequent migraine costs employers as much as $13,000, a hidden cost representing nearly three times what individuals pay in insurance premiums.^9,10,11 ICER should define the healthcare perspective as greater than the payer to be inclusive of all stakeholders (e.g., patients and employers): this more comprehensive approach will more accurately capture the full value of CGRP’s to patients.

Failure to include a broader perspective may inadvertently lead to migraine being deprioritized over conditions that incur significant payer costs. Health economists have recommended this broader perspective for over 20 years¹² to ensure that untreated conditions that incur very little comparative cost to payers are not deprioritized. Migraine qualifies as one of these as it is 1) extremely debilitating if left untreated 2) incurs little directly identifiable cost to the payer if left untreated and 3) may be more expensive to payers when effectively treated.  Economists recognize that health as a merit good has benefits that go beyond the healthcare system that methodologically should be reflected in any economic evaluation, including migraine.¹³

The use of a payer perspective as the base-case is biased against a fair evaluation of therapies to prevent migraines (such as CGRPs) as it fails to capture more than half of the value to patients. As emphasized in the Amgen open comment period response, on top of the non-payer incurred direct medical costs, there are also substantial costs that lie outside of the healthcare system. Highly debilitating, migraine causes sufferers to frequently miss work and be less productive while at work, costing employers, patients and society billions of dollars per year.^14,15 Absenteeism accounts for 81%, short term disability 13%, and workers compensation 6% of costs from lost productivity due to migraine.^{16,17,18,19,20} ICER should use a broader healthcare system perspective, inclusive of the patient and employer, as their base-case. Additionally, we urge ICER to make all inputs, assumptions and data available and fully transparent –especially in the event that ICER chooses to run this as a separate scenario instead of a base-case.

Erenumab should not be assessed in populations who are naïve to migraine preventive treatment. Patients that can gain disease control from currently available preventive treatments and who can persist on them, represents maximum value to both the patients and the healthcare system. Multiple clinical and insurer sources suggest that in clinical practice, erenumab will be used after failure of topiramate or propranolol, addressing the high unmet need of migraine patients who have experienced a lack of efficacy or tolerability from prior preventives.^{21,22,23,24,25} Further, in the U.S. ICER CTAF 2014 report, Botox^® is reserved for CM patients who have failed topiramate or propranolol.²⁶ In the U.S., this reinforces the place of CGRPs in preventive treatment-experienced patients. Moreover, while ICER recommends topiramate and propranolol, the evidence on these interventions is not directly comparable to the evidence on erenumab. Studies are not comparable in terms of population, definition of endpoints and other study design elements, which weakens the comparative efficacy estimation.^27,28,29,30

Erenumab has been shown to benefit treatment experienced patient populations including those who have failed prior treatment (including topiramate and propranolol). Erenumab is an investigational migraine preventive that has demonstrated sustained efficacy specifically in these patients in clinical trials.^31,32 It is of paramount importance that ICER’s assessment address the unmet needs of patients who have not benefited from currently available preventive treatments. If ICER were to assess erenumab in all patients, this is not aligned with how these treatments are expected to be used in real world practice.  It may inadvertently restrict access in patients who will benefit the most and undermine the potential value of these innovative treatments.

There are several ways to model the economics of migraine prevention with CGRPS. In treatment experienced patients, one way is to model all patients with > 4 migraines days per month.  In this case, the appropriate comparator for erenumab in the patient populations who have tried and failed topiramate or propranolol is ‘no preventive treatment’.  Based on ICER’s prior report on Chronic Migraine, ICER may choose to model EM and CM separately, in which case based on ICER’s prior analysis, Botox in CM may be the comparator for the CM population.  What is known, is that there is currently no defined standard of care for patients with > 4 migraines per month who have tried and failed topiramate or propranolol. Additionally, there are no clinical trials or observational cohort data that are available or published with propranolol³³ or topiramate ^{34,35,36,37,38,39} in patients who have tried and failed these treatments. As such, neither topiramate nor propranolol are appropriate comparators in preventive treatment experienced patients with 4 or more migraine days per month.

An adequate therapeutic trial with preventive therapy is approximately 2-3 months based on treatment guidelines for prevention.^40,41 Requiring failure of multiple classes of therapies for 18-24 months is overly burdensome to patients and is not supported by available evidence.^42,43,44 Treatment guidelines, published sources, real-world and clinical practice data do not to support this approach.⁴⁵ The time a patient tries a treatment can be short, reflected in persistence rates.  Annual persistence rates with current preventive therapies is low: 50% of patients discontinue treatment at one month, 70-75% by month 6.⁴⁶ In another study, similar trends were observed at month 12.⁴⁷ Given that there are no data in a robust controlled setting or in current clinical practice that support a treatment paradigm of patients requiring failure of multiple classes, ICER should evaluate erenumab in migraine patients after failure of topiramate or propranolol with an adequate therapeutic trial.

Discontinuation rates and poor adherence must be accurately captured since patients can only derive continued benefit when they persist with therapy. Adherence is a fundamental driver of value.^48,49 Despite discontinuation of preventive migraine medications, acute medication use persists.⁵⁰ While acute medication use is observed to decrease in the 30 days after discontinuation of the index preventive, acute medication use rebounds to pre-preventive usage in the 31-90 days after discontinuation.⁵¹ ICER should ensure that they accurately account for and measure discontinuation rates in all treatments as they have a significant impact on efficacy in clinical practice. Failure to capture this accurately would significantly underestimate the value of CGRPs.

Per guidelines, RR is a key measure of preventive treatment success.⁵² Odds ratio (OR) and confidence interval (CI) relative to placebo, in all comers for erenumab 140-mg in EM was 2.81 (2.01 to 3.94) ⁵³ and in CM 2.3 (1.6, 3.5). ⁵⁴ In treatment experienced patients with >1 prior treatment failures, OR relative to placebo, in EM was 3.06 (1.70, 5.52) ⁵⁵ and in CM was 3.30 (1.98, 5.51)⁵⁶, nominal p<0.001. ICER should reflect RR as an important outcome measure since it accurately captures a with-in patient response.

In a study comparing the modeling of migraine preventives using a health state approach versus a continuous approach, parametric continuous distributions provide more accurate approximations of migraine day frequency.⁵⁷ ICER should use a continuous distribution approach rather than a “health state” approach to more accurately capture migraine day frequency.

CONCLUSION

The migraine prevention space has had no real innovation in two decades and patients are in need of new treatment options. As ICER has accurately reflected, migraine is one of the most debilitating diseases in the world.  Its profound effects, which impact every facet of patients’ lives, are frequently underestimated. Current preventive treatment options are suboptimal for many patients. We recommend that ICER focus on areas where CGRPs are more likely to be used in real world clinical practice: those migraine patients that do not have available treatments that work.  Moreover, ICER should ensure that its assessment of value is inclusive of all stakeholders, not just healthcare payers, by incorporating a more comprehensive healthcare system and societal perspective in order to capture migraine costs that CGRPs could alleviate.  At this time, we are also providing a supplemental data appendix with further data to support our comments here.

References: