2 The procedure
2.1.1 Migraine is a severe headache, often accompanied by sensitivity to light and sound. It may be preceded by an aura, consisting of perception of an unusual light, an unpleasant smell or, occasionally, confusing thoughts or experiences. The International Classification of Headache Disorders (International Headache Society 2004) provides a classification of migraine types.
2.1.2 Current treatment for patients with migraine aims to prevent or stop episodes with drugs such as painkillers, anti‑emetics and triptans (as recommended in NICE clinical guideline 150). If these fail, invasive treatments such as nerve blocks, botulinum toxin type A (NICE technology appraisal guidance 260) or acupuncture are sometimes used.
2.2.1 ONS for intractable chronic migraine is usually done in 2 stages, although a single‑stage procedure is sometimes used. In the first, trial stage, using local anaesthesia and usually with fluoroscopic guidance, electrodes are passed through a subcutaneous tunnel and placed over the occipital nerve(s) around the level of C1. Correct placement of electrodes is verified by intraoperative stimulation and patient feedback before they are sutured to subcutaneous tissue. A lead is tunnelled under the skin from the electrode to an exit site in the posterior cervical region, where it is connected by an external extension lead to a hand‑held neurostimulator.
2.2.2 The second stage is carried out if the trial is successful. With the patient under general anaesthesia, an implantable neurostimulator is secured in a subcutaneous pocket, usually in the infraclavicular region or the abdominal wall. A lead is tunnelled from the electrode to the implantable neurostimulator. The patient uses a remote control to stimulate the occipital nerves when needed.
2.3.1 A randomised controlled trial (RCT) of 157 patients compared ONS (active stimulation, n=105) against sham stimulation (n=52). It reported a statistically significantly greater decrease in the 'Migraine disability assessment score' (MIDAS), which takes into account headache days and their impact on the patient's life (maximum score 200), at 12‑week follow‑up for the ONS group than for the sham stimulation group (64.6 and 20.4 reduction respectively, p=0.001).
2.3.2 An RCT of 67 patients comparing ONS (n=33) against sham stimulation (n=17) or medical management (n=17) reported a responder rate (defined as a reduction in headache days per month of 50% or more, or a 3‑point or greater reduction in average overall pain intensity compared with baseline at 3 months) of 39% (11/28) in the ONS group, 6% (1/16) in the sham stimulation group and 0% (0/17) in the medication group (p value not reported).
2.3.3 A case series of 25 patients reported that headache frequency per 90 days reduced from 75.56 (standard deviation [SD] 26.81) before implantation to 37.45 (SD 7.49) over a mean follow‑up of 18 months (p<0.001).
2.3.4 The RCT of 157 patients reported no significant difference between the groups in the proportion of patients whose pain reduced by 50% or more (measured on a visual analogue scale) (17% for ONS and 14% for sham stimulation, p=0.55) at 12‑week follow‑up.
2.3.5 The case series of 25 patients reported a significant reduction in headache severity (0–10 scale) from a baseline of 9.32 (SD 1.28) to 5.72 (SD 3.31) over a mean follow‑up of 18 months (p<0.001).
2.3.6 The Specialist Advisers listed key efficacy outcomes as a reduction in migraine or headache days, headache severity, frequency and duration, disability score (measured by MIDAS), medication use and improvements in quality of life (SF‑36).
2.4.1 Infections at the implant site were reported in 14% (7/51) of patients in the RCT of 67 patients at 3‑month follow‑up. Infection was reported in 4% (4/105) of patients in the ONS group and 6% (3/52) of patients in the sham stimulation group in the RCT of 157 patients at 12‑week follow‑up (no further details available).
2.4.2 Skin erosion was reported in 4% (4/105) of patients in the ONS group and 4% (2/52) of patients in the sham stimulation group in the RCT of 157 patients at 12‑week follow‑up.
2.4.3 Lead migration or dislodgement was reported in 10% (5/52) of patients in the sham stimulation group and 14% (15/105) of patients in the ONS group in the RCT of 157 patients after 3 months; and in 24% (12/51) of patients in the RCT of 67 patients at 3‑month follow‑up. Lead migration was reported in 36% (9/25) of patients in the case series of 25 patients at mean 18‑month follow‑up.
2.4.4 Problems with ineffective device programming and ineffective leads were reported in 12% (6/51) and 4% (2/51) of patients respectively, in the RCT of 67 patients at 3‑month follow‑up.
2.4.5 Persistent pain or numbness at the implant site was reported in 13% (14/105) of patients in the ONS group and 17% (9/52) of patients in the sham stimulation group in the RCT of 157 patients at 12‑week follow‑up. Loss of motor or musculoskeletal control was reported in 1% (1/105) of patients in the ONS group in the same RCT (timing not reported).
2.4.6 Unintended stimulation effect (no further details available) was reported in 6% (6/105) of patients in the ONS group and 2% (1/52) of patients in the sham stimulation group in the RCT of 157 patients.
2.4.7 In addition to the above, the Specialist Advisers listed haemorrhage, nerve damage and lead fracture as theoretical adverse events.
2.5.1 The Committee recognised that patients being considered for ONS for intractable chronic migraine commonly have very distressing and long‑term symptoms that other methods of treatment have failed to control.
2.5.2 The Committee recognised that research in this area is difficult because there is uncertainty about the percentage level of relief that should be considered significant and it is difficult to achieve blinding in trials, and because of the complex and heterogeneous nature of chronic migraine. Currently, there are not enough good‑quality comparative studies to be able confidently to evaluate the procedure's efficacy. This underpins the recommendations in section 1.
2.5.3 The Committee recognised that techniques and technology are evolving and, implicitly, this may produce better results.