Efficacy and Contextual (Placebo) Effects of CGRP Antibodies for Migraine: Systematic Review and Meta-analysis

Systematic Review and Meta-analysis

Raeburn B. Forbes, MD(Hons), FRCP, Lond, FRCP, Ed; Mark McCarron, MA, MD, FRCP, MMedED; Chris R. Cardwell, PhD


Headache. 2020;60(8):1542-1557. 

In This Article

Abstract and Introduction


Background: CGRP Antibodies are high-cost newly licensed migraine preventatives.

Objective: To calculate the overall reduction in monthly migraine days and the proportion contextual effect (PCE) using meta-analysis. The PCE is the ratio between the reduction in Monthly Migraine Days in the placebo group and the reduction in Monthly Migraine Days in the CGRP-Ab group after 3 months of treatment.

Methods: Meta-analysis of randomized double-blind placebo-controlled trials of anti-CGRP antibodies in people with episodic migraine (EM) or chronic migraine (CM) in persons aged 18 or over. Non-randomized trials and trials in persons under 18 years excluded. Search of National Clinical Trials Register 2000–2019, MEDLINE to September 2019, Hand search of major headache conference abstract books 2012–2019. Two investigators used standard proforma to reach consensus. Trial quality assessed using Cochrane Collaboration risk of bias tool. PRISMA guidelines followed.

Results: 21 completed trials with 13367 participants (8075 EM, 5292 CM). Compared to placebo, pooled reduction in MMD was 1.50 days in 15 EM trials (95%CI 1.16, 1.84; I2 = 69%, P hetereogeneity < .001) and 2.24 days in 7 CM trials (95%CI 1.82, 2.65, I2 = 15%, P hetereogeneity = .320). In EM trials, pooled PCE was 0.66 (95%CI 0.59,0.75; I2 = 64%, P hetereogeneity = .001). In CM trials the PCE was .68 (95%CI 0.61, 0.75; I2 = 20%, P hetereogeneity = .280). Industry funded every study, but risk of bias was low.

Conclusions: CGRPAbs are effective but sixty-six percent of the benefit is from contextual effects, including placebo effect. Contextual effects merit further scrutiny as a means of improving migraine headache.


Calcitonin gene-related peptide (CGRP) influences nociceptive mechanisms in the trigeminovascular system and plays a pivotal role in the development of migraine pain.[1–3] Treatment with subcutaneous or intravenous monoclonal antibodies against calcitonin gene-related peptide (CGRP) and its receptor demonstrates significant reductions in monthly migraine days (MMD).[4] The journey from basic science to bedside heralds a new era in migraine[5] – one of the world's leading causes of disability.[6] Compared to existing migraine prevention, anti-CGRP monoclonal antibodies (CGRPAbs) are expensive (in the UK approximately £4600 per patient per annum), and use will be restricted until acceptable cost-effectiveness ratios are established.[7] It is inevitable that some of the effectiveness of CGRPAbs is due to placebo effect. Clinicians should be made aware of the extent of any placebo effects and cost-effectiveness evaluations may need to take such effects into consideration.

An overall treatment effect is due to the specific treatment effect (due to the mode of action of the drug – in this case CGRP blockade) and a non-specific treatment effect (due to the context in which the treatment is given). In clinical trials this non-specific effect is called the placebo effect. Before accepting a novel medicine as effective, a clinical trial must demonstrate superiority to placebo in a randomized, controlled trial. In clinical practice, where use of placebo is considered unethical, the non-specific effect is called a "contextual effect." Factors such as mode of administration (injection, tablet, or surgery), the clinical environment, expectations and beliefs of the patient, and attitude or manner of the treating physician all contribute to the "contextual effect."[8] The total treatment effect experienced by the patient, and therefore the real-world benefit of the intervention, is the sum of specific treatment effect and non-specific contextual effects.

In clinical practice contextual effects are important, especially in chronic painful conditions, eg, physicians who do not adopt a reassuring manner are less likely to make their patients feel better.[8] Meta-analysis of clinical trials shows that contextual effects contribute as much as 75% of overall treatment effect in painful disorders such as osteoarthritis[9] or fibromyalgia.[10] Invasive treatments such as joint lavage may offer greater relief due to a greater contextual effect, than less invasive treatments.[9] In migraine, sham acupuncture has been shown to have a higher response rate than oral pharmacological placebo.[11] Different placebo response rates can lead to an "efficacy paradox" where a more powerful intervention with less contextual effect is less likely to reduce symptoms in real life than a less powerful drug with greater contextual effect.[12] The "efficacy paradox" is also observed when a drug with modest effect in a randomized clinical trial seems to be highly effective in clinical practice.[12]

To estimate specific and contextual treatment effects of CGRPAbs in adults with episodic migraine (EM) and chronic migraine (CM) we estimate overall efficacy and the proportion contextual effect (PCE) from a systematic review and meta-analysis of completed randomized double-blind placebo-controlled trials. This information may inform the process of introducing CGRPAbs into practice. To our knowledge this is the first systematic review to estimate PCE for a neurological disorder.