Preventive treatment of refractory chronic cluster headache: systematic review and meta-analysis

Preventive treatment for refractory chronic cluster headache (rCCH) is challenging and many therapies have been tried.

To study what could be considered the therapy of choice in rCCH through a systematic review and meta-analysis.

This review was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. The protocol was registered in PROSPERO (ID CRD42021290983). A systematic search was performed in MEDLINE, Embase, Cochrane, clinicaltrials.gov, and the WHO's-International-Clinical-Trials-Registry-Platform. Studies on the preventive treatment for rCCH as defined by the European Headache Federation consensus statement were included. A meta-analysis of the pooled response rate was conducted for the different therapies.

Of 336 results, 45 were eligible for inclusion. Most articles studied the effect of neuromodulation as a preventive treatment for rCCH. The most studied neuromodulation technique was occipital nerve stimulation (ONS), with a pooled response rate in the meta-analysis of 57.3% (95% CI 0.481-0.665). Deep brain stimulation (DBS) was the second most studied treatment with a pooled response rate of 77.0% (95% CI 0.594-0.957). DBS results were more heterogeneous than ONS, which could be related to the different stimulation targets in DBS studies, and reported more serious adverse events than in ONS studies. The remaining therapies (anti-CGRP pathway drugs, warfarin, ketamine-magnesium infusions, serial occipital nerve blocks, clomiphene, onabotulinum toxin A, ketogenic diet, sphenopalatine ganglion radiofrequency or stimulation, vagus nerve stimulation, percutaneous bioelectric current stimulation, upper cervical cord stimulation, and vidian neurectomy) present weaker results or have less quality of evidence.

The results of this systematic review and meta-analysis suggest that ONS could be the first therapeutic strategy for patients with rCCH based on the current evidence.