Greater occipital nerve blocks as migraine therapy: A discussion of two recent randomized,double-blinded,placebo-controlled clinical trials

Dear Editor:

Although greater occipital nerve block (GONB) had been advocated for decades as migraine therapy, the procedure had not been scientifically studied. Recently two randomized double-blinded placebo-controlled trials (RCTs) for prophylactic treatment of migraine were published (1,2). The RCTs came to opposite conclusions. The RCT by Dilli et al. concluded that GONB was no better than placebo while the RCT by Inan et al. found that GONB was statistically better than placebo. There were two major differences between the studies. In the Dilli et al. RCT, the anesthetic (bupivacaine) was injected in the conventional one-time administration while in the study by Inan et al. the injections were administered weekly for one month.

The placebos were defective in both RCTs. In the Dilli et al. study, the placebo was 1% lidocaine diluted in 2.75 ml of saline. This was an attempt to mask the placebo by evoking sensory changes similar to bupivacaine, but the participants were not tested to confirm this assumption. It was also assumed that this small dose of lidocaine would have no effect on the headache. The fact that both placebo and bupivacaine (with methylprednisolone) evoked a 30% positive response suggests that the placebo was successfully masked. The alternative explanation is less tenable; i.e. 1% lidocaine (diluted to 0.08%) was as active as 0.5% bupivacaine and 20 mg methylprednisolone.

Participants in the Inan et al. study were falsely told that the side effects of injections (such as local numbness) would be unrelated to whether the injection was placebo or verum. Surely some individuals who did not experience scalp numbness recognized or suspected that they had received placebo. They would have no expectation of improvement (the nocebo effect), adding to the nonresponsive group and skewing the negative to positive ratio toward the latter.

Of the two RCTs, the Dilli et al. study would appear to be the most valid. But additional RTCs will be needed to confirm these findings. Future protocols should use better placebo masking, e.g. application of topical anesthetic to all groups, and placebo verification by examination and subject questioning.