Botulinum toxin type A for the prevention of headaches in adults with chronic migraine

4 Consideration of the evidence

4.1 The Appraisal Committee reviewed the data available on the clinical and cost effectiveness of botulinum toxin type A, having considered evidence on the nature of chronic migraine and the value placed on the benefits of botulinum toxin type A by people with the condition, those who represent them, and clinical specialists. It also took into account the effective use of NHS resources.

4.2 The Committee considered the impact of chronic migraine on the everyday life of people with the condition. It heard from the patient experts that chronic migraine is accompanied by severe pain, which impacts greatly on people's quality of life, affecting their ability to work and participate in social activities. The patient experts also noted that people with chronic migraine often experience anxiety and depression related to their condition. The Committee considered chronic migraine to be a debilitating condition which significantly affects health-related quality of life.

4.3 The Committee considered current clinical practice for the treatment of chronic migraine. The Committee heard from the clinical specialists that it is important for people first presenting with chronic migraine to try a range of oral preventive treatment options before considering treatment with botulinum toxin type A. The clinical specialists stated that there is no one measure of response to treatment and that response is multifaceted. The Committee heard from the clinical specialists that a good response to treatment is typically considered to be a 30–50% reduction in the frequency of headache days or headache episodes. The Committee also noted that when a person's response to treatment is assessed, clinicians also take account of any improvement in the person's quality of life, and that improvements may not always be accompanied by substantial reductions in the number of headache days. It heard from clinical specialists that people considered for treatment with botulinum toxin type A are assessed for medication overuse before treatment starts, and that this is monitored during treatment. The Committee also heard from the patient experts that many people with chronic migraine would consider any degree of response to treatment to be valuable and that further treatment options were important.

4.4 The Committee considered the administration of botulinum toxin type A and monitoring of the potential benefit associated with treatment. It heard from clinical specialists that an initial consultation typically lasts between 45 minutes and 1 hour and includes administration of botulinum toxin type A, which can take between 15 and 30 minutes depending on the experience of the person administering the injection. After treatment patients are asked to keep a headache diary. The clinical specialists stated that the re-treatment interval with botulinum toxin type A varies in practice. Some clinical practices have routine follow-up about every 3 months. This will be either a telephone consultation with a consultant or headache specialist nurse, or a 30 minute clinic appointment with the consultant during which repeat injections may be administered. In other clinical practices there is no routine follow-up and patients must request an appointment to be considered for further treatment with botulinum toxin type A. The Committee concluded that the re-treatment interval with botulinum toxin type A is variable in practice, being 3 months in some patients, but at much longer intervals in others.

4.5 The Committee considered the evidence on the clinical effectiveness of botulinum toxin type A for the prevention of headaches in adults with chronic migraine from the PREEMPT trials, which compared botulinum toxin type A with placebo. The Committee noted that the pooled results for the intention-to-treat population indicated a statistically significant reduction in frequency of headache days per month, migraine days per month and cumulative headache hours with botulinum toxin type A compared with placebo, but considered the absolute numerical benefit of botulinum toxin type A over placebo to be modest. It also noted the statistically significant reduction in the frequency of headache episodes, migraine episodes and frequency of acute headache medication days with botulinum toxin type A compared with placebo, but that the absolute numerical differences between the treatments were small. The Committee observed that there was no statistically significant difference in how often acute pain medication was taken. Further, the Committee noted the large placebo effect seen in the trials. It was also aware that the incremental effect of botulinum toxin type A compared with placebo may have been increased by people in the active treatment arm realising that they were receiving botulinum toxin type A because of its side effects. The Committee noted that the duration of follow-up in the PREEMPT trials was short (1 year) and the study design, which meant that all patients received the active drug from week 24, could not exclude natural improvements in people's condition over time. The Committee heard from the clinical specialists that at baseline, patients in the clinical trials had fewer headache days per month (a mean of 19) than people with chronic migraine in secondary care in the UK (a mean of 25–26 headache days per month). The clinical specialists said that this might explain why the benefit observed in the trials was less than what they would expect to see in clinical practice. The Committee discussed the duration of the therapeutic effect of botulinum toxin type A and heard from the clinical specialists and patient experts that when a person's condition responds to botulinum toxin type A, the re-treatment interval varies, and that many patients need treatment for longer than 2 years. The Committee was aware of the difficulties of conducting clinical trials in people with chronic migraine, particularly with regard to the known significant placebo effect observed in such studies. It noted that the placebo comparator in the trial would not be given in clinical practice in the manner used in the trial. Although the absolute magnitude of benefit with botulinum toxin type A was modest, evidence from clinical specialists and patient experts suggested that the effect was clinically meaningful in people whose chronic migraine had failed to respond to three prior preventive treatments and whose condition responds (with at least a 30% reduction in the number of headache days per 28 days) to botulinum toxin type A treatment. The Committee therefore concluded that botulinum toxin type A was clinically effective in people with chronic migraine whose condition had not responded to three prior preventive treatments.

4.6 The Committee considered the clinical trial evidence in light of the views of the patient experts and clinical specialists. The Committee noted the improvements in quality of life for patients whose condition responded to botulinum toxin type A. The Committee then considered the use of botulinum toxin type A in clinical practice in the UK. The Committee heard from clinical specialists that if a person's chronic migraine was responding to botulinum toxin type A, the treatment would be continued until the number of headache days was reduced to fewer than 15 headache days per month (episodic migraine). The clinical specialists estimated that people would have at least two treatment cycles. They said that after this approximately 50% of people would continue on treatment, and of those 30% would need five cycles of treatment before their condition was reclassified as episodic migraine. The remaining patients would continue to receive treatment for longer than 2 years. Alternatively, if chronic migraine does not respond adequately to botulinum toxin type A after at least two cycles of treatment, and there is little benefit to the patient, treatment is discontinued and patients would receive standard care. The Committee noted that the clinical trial evidence on the effectiveness of botulinum toxin type A was for 1 year. The Committee was aware of the abstract by Rothrock et al. (2011) and the update by Hanlon et al. (2011), which estimated duration of response to treatment with botulinum toxin type A, with response to treatment defined as a 50% reduction in the number of headache days per month. The Committee noted that only 24% of the patients who responded to treatment were able to stop treatment and maintain a good response for at least 6 months. The Committee concluded that because there are no long-term clinical trial data, the estimates of Rothrock et al. and Hanlon et al. are likely to be the best current estimates of the duration of benefit for people who respond to botulinum toxin type A in clinical practice in the UK.

4.7 The Committee discussed the adverse reactions of treatment with botulinum toxin type A for chronic migraine. It noted the adverse reactions reported in the trials with botulinum toxin type A (see section 3.8). It heard from the patient experts that there is often pain around an injection site lasting a few days after treatment with botulinum toxin type A. However, people would be willing to tolerate the adverse reactions with botulinum toxin type A treatment to reduce the frequency or severity of their chronic migraine. The Committee concluded that botulinum toxin type A is generally well tolerated; a conclusion supported by the patient experts and clinical specialists.

4.8 The Committee considered the cost effectiveness of botulinum toxin type A compared with standard care, based on the manufacturer's economic model and the critique by the ERG, in the context of the decision problem. The Committee was aware that the scope specified that the population should include adults with chronic migraine whose condition has failed to respond to at least three prior pharmacological prophylaxis therapies, and whose medication overuse has been appropriately managed. The Committee was aware that the manufacturer had focused on this population in a sensitivity analysis and that its main submission compared botulinum toxin type A with placebo in people whose condition has failed to respond to at least one prior preventive medication. The Committee noted comments from the clinical specialists that people first presenting with chronic migraine will be prescribed a range of oral preventive medication options before treatment with botulinum toxin type A is considered. The Committee concluded that the main decision problem proposed by the manufacturer, comparing botulinum toxin type A with placebo in people whose condition has failed to respond to at least one prior preventive medication, did not reflect clinical practice. It concluded that it was only relevant for the Committee and the NHS to consider the clinical and cost effectiveness of botulinum toxin type A in people whose chronic migraine has failed to respond to at least three prior preventive medications and that the trial data from this subgroup should be used to calculate the transition probability matrices used in the economic model.

4.9 The Committee discussed the revised economic model submitted by the manufacturer as part of the response to consultation. The Committee noted that the revised model included their preferred assumptions and inputs. It was aware that the revised model included a negative stopping rule of a two health state reduction in two cycles of treatment. The Committee noted that the manufacturer's cumulative analyses on the impact of each of the Committee's preferred assumptions resulted in a revised deterministic ICER for botulinum toxin type A compared with placebo of £15,300 per QALY gained when different utilities are applied to each arm, and £24,500 per QALY gained when the same utilities are applied to each arm. The Committee noted that the revised probabilistic ICERs for botulinum toxin type A compared with placebo were similar to the deterministic ICERs. The Committee concluded that the large differences between the deterministic and probabilistic ICER in the original model were no longer apparent in the revised economic model.

4.10 The Committee considered the costs and resource use in the revised economic model. The Committee was satisfied that its request for the placebo follow-up cost to be the same as the administration cost of botulinum toxin type A had been resolved because the revised economic model had included a standard NHS neurological outpatient follow-up reference cost of £140 for the placebo follow-up cost. It noted that the routine care costs used in the placebo arm were applied to people who discontinue botulinum toxin type A because of lack of efficacy in the model, and an average accident and emergency cost of £77.33 was used by the manufacturer, as requested in the Committee's preliminary recommendations. The Committee considered the additional clarification on resource use supplied by the manufacturer. It noted the ERG's concern that the resource use data for people from the UK with chronic migraine in the IBMS presented by the manufacturer included a small number of people (n=57), and a significant number of outlying results (for patients whose resource use contributed to most of the total resource use observed). The Committee concluded the Blumenfeld et al. resource use data to be most appropriate given the small number of people from the UK with chronic migraine in the IBMS.

4.11 The Committee noted that its preliminary recommendations requested any revised economic model to incorporate a range of negative stopping rules based on the reduction in the number of headache days per 28 days after two cycles of treatment for people with chronic migraine. The Committee noted from comments received during consultation that there was agreement between the manufacturer and the clinical community that a 50% response rate is considered to be too high, and that a 30% response rate recommended by the British Association for the Study of Headache is used in clinical practice. The Committee noted that the manufacturer's revised base-case deterministic ICER (based on a negative stopping rule of less than 30% reduction in headache days per month) for botulinum toxin type A compared with placebo was £15,000 per QALY gained when applying different utilities, and £24,900 per QALY gained when applying the same utilities. The Committee concluded that a 30% response rate (that is, a 30% reduction in the number of headache days per month after two cycles of treatment) was the most clinically relevant and reasonable negative stopping rule on which to base its decision.

4.12 The Committee discussed the appropriate positive stopping rule. It noted the consultation comments that a positive stopping rule, in which patients stop treatment if their migraine has changed to episodic migraine and remained stable in episodic migraine for at least 3 months, is the most clinically appropriate. The Committee also noted that the marketing authorisation for botulinum toxin type A does not include use in people with episodic migraine. It therefore concluded that a positive stopping rule in which patients stop treatment if their condition has changed to episodic migraine (that is, fewer than 15 headache days per month) for three consecutive months is the most clinically relevant. However, the Committee recognised that according to the only published longer term follow-up of patients who responded to treatment with botulinum toxin type A, only 24–25% were able to stop treatment with botulinum toxin type A and maintain a good response for at least 6 months (Rothrock et al. and Hanlon et al.) The Committee concluded that these publications provided the most plausible estimate for the likely implementation of the positive stopping rule in clinical practice in England and Wales, with 24–25% being the most appropriate figure on which to base a positive stopping rule in the economic model.

4.13 The Committee considered the use of the same or different utilities within each health state in the botulinum toxin type A and placebo arms. The Committee noted comments from consultees and commentators that treatment with botulinum toxin type A is associated with a range of clinical and non-clinical benefits, which are not included in the reduction in the number of headache days per month. It considered data on botulinum toxin type A on improving secondary outcomes, including the number of headache episodes, the number of migraine or probable migraine days, the number of moderate or severe headache days, and the number of acute medication days from the pooled PREEMPT trial data and evidence given by the patient experts in writing and at the Committee meeting. The Committee was also aware of an online survey of 60 people with chronic migraine recently conducted by Migraine Action, which also provided supportive data. However, the Committee was aware of the absence of robust data supporting the size of the difference in utility values in each arm. The Committee concluded that although using different utility values within each health state in the botulinum toxin type A and the placebo arm was plausible and better than applying the same utility values within each health state to calculate the most appropriate ICER for considering cost effectiveness, there was still considerable uncertainty around the degree to which differential utilities existed within each health state.

4.14 The Committee was aware that its preliminary recommendations stated that any revised economic model should explore removing the non-monotonicity in the original model, that is, for the botulinum toxin type A arm, the utility in health state 5 (20–23 headache days per month) is lower than the utility in health state 6 (24–28 headache days per month). It heard from the manufacturer that the non-monotonicity in the utility values in the original model was caused by the small number of people in the three or more prior preventive treatment group. The Committee further noted the ERG's explanation that the non-monotonicity was not surprising given that health states were defined in terms of the number of headache days per 28 days, while the MSQ utility mapping functions were constructed so as to pick up other elements of the condition (such as severity or intensity of headaches). The Committee considered that this explanation was plausible and that removing the non-monotonicity had little impact on the ICERs. It concluded that there was no need to remove the non-monotonicity in the updated economic model.

4.15 The Committee considered the explanation from the manufacturer on the way that utilities had been calculated, noting the way in which utilities had been mapped from the MSQ. The Committee noted the ERG's concern that the non-MSQ parameter values were different in the botulinum toxin type A and placebo utility mapping functions. It agreed that the manufacturer had not provided a clear rationale as to why these parameter values should be different. The Committee noted that when the ERG equalised the non-MSQ parameter values, less non-monotonicity was observed, and the deterministic ICER was £18,900 per QALY gained when applying different utility values to each arm. The Committee concluded that this was the most plausible ICER because it incorporated the Committee's preferred inputs and assumptions including a 30% negative stopping rule, applied different utilities to treatment arms (within the Committee's reservations stated in section 4.13), and equalised the non-MSQ parameter values in the utility mapping functions.

4.16 The Committee discussed whether botulinum toxin type A was an appropriate use of NHS resources for the prevention of headaches in adults with chronic migraine. It noted that botulinum toxin type A is the only technology with a UK marketing authorisation for the prevention of headaches in adults with chronic migraine, and that the method of administration of botulinum toxin type A is novel compared with other preventative treatments used in the management of chronic migraine. The Committee accepted the plausibility of using different utility values in the botulinum toxin type A and placebo arms (within the reservations expressed in section 4.13), and considered that the utility values in the economic model encompass the major health-related quality of life benefits associated with treatment with botulinum toxin type A, including duration and intensity of migraine, reduction in symptoms, need for rescue treatment, and lower dose of acute medication. It therefore considered that all of the significant or substantial health benefits of botulinum toxin type A treatment had been included in the model. The Committee concluded that because the most plausible ICER presented was less than £20,000 per QALY gained, botulinum toxin type A could be considered an appropriate use of NHS resources for the prevention of headaches in adults with chronic migraine that has not responded to at least three prior pharmacological prophylaxis therapies and whose condition is appropriately managed for medication overuse.

4.17 The Committee discussed whether NICE's duties under the equalities legislation required it to alter or add to its preliminary recommendations in any way. The Committee was aware that during scoping, consultees and commentators suggested that there is inequity in access to diagnosis and treatment of migraine for people whose first language is not English. It also noted that comments suggested that there is unequal access to treatment for chronic migraine for people with mental health issues and that greater recognition of chronic migraine as a significant clinical problem will help eliminate discrimination in the workplace. It heard from the patient expert and from consultation comments that chronic migraine is more prevalent in women than men. The Committee was aware that consultation comments also suggested that there is inequity in access to treatment with botulinum toxin type A for people on low income. The Committee did not consider access to treatment for people whose first language is not English to be relevant because the recommendations do not specify a particular English language based test for the diagnosis of chronic migraine. Further, because the recommendations do not differentiate between any groups of people, the Committee concluded that its recommendations did not limit access to the technology for any specific group compared with other groups.