Baseline data were available for 19 subjects (43 ± 10 years; 84% female), who had failed an average of 11.7 ± 3.2 preventive treatments at the time of recruitment. The average number of headache days at baseline was 23.3 ± 5.2 days of which 21.6 ± 6.6 days were migraine days. The average migraine-specific quality of life (MSQ) score at baseline was 32.0 ± 15.7. Compared to baseline, the average reduction in headache days at 12 weeks was 3.9 days, which increased to 4.9 days at week 24. The average reduction in migraine days at 12 weeks was 6.2 days/month, which increased to 7.2 days at weeks 24. Thirty-eight percent of the subjects at week 12 and 43% at week 24 obtained a 30% reduction in headache days; 44% of subjects at week 12 and 50% of subjects at week 24 reported a reduction of more than 30% in migraine days compared to baseline. In 50% of the subjects, the CM pattern reverted to an episodic pattern (<15 headache days/month) at week 24. We observed an improvement in the MSQ score of an average of 17.2 points at 24 weeks (50.8 ± 23.4). At week 24, five subjects reported pain at the site of implant, four reported musculoskeletal pain in the cervical and shoulder area and one experienced slight lead movement. No subjects required any further surgical procedure.

This preliminary evidence suggests that high cervical SCS, contrary to the first study in CCH, is a safe and well-tolerated treatment at least in experienced hands. In the only large-scale, prospective, controlled study evaluating ONS for CM with 1-year follow-up to date, the safety profile of the therapy was questioned by the authors themselves, as 183 device- or procedure-related AEs occurred, 8.6% of which required hospitalization and almost 41% required additional surgery [30]. In the first HF10 SCS study, four subjects (24%) required an additional surgical procedure, whereas none of the 20 patients implanted in the more recent trial required further surgery at 6 months f/u.

In terms of efficacy, these initial evidence may suggest that HF10 SCS is beneficial in rCM. One of the strengths of these studies is the fact that the migraine population included in these two trials was very refractory to medical treatments. Not many trials in ONS included a population who failed that many treatments and Botox. The ONS RCT in CM included patients who failed at least two preventive treatments; however, it was unclear the mean numbers of treatments failed at enrolment. Furthermore, none of the patients tried Botox according to the PREEMPT trial in those trials. Botox is currently considered the standard of care for CM treatment [31, 32]; hence, its failure has been included in the most recent international guidelines for the definition of rCM [33]. A recent prospective open-label study, which studied the efficacy of a group of 35 rCM patients who failed an average of 9.36 (±2.61, range 4–19) preventative treatments, is somehow comparable to the HF10 SCS study populations, although there was no mention of the proportion of patients who failed Botox before implant [34]. At baseline, the patient group had a mean of 26.83 (±6.74) moderate to severe headache days. At a median f/u of 42 months, there was an average reduction of 5.8 headache days/month and 34.3% of rCM patients obtained >30% in monthly headache days. Although it is not appropriate to compare these trials, it is possible that HF10 SCS is as effective or even more effective than ONS in rCM. Similarly to chronic pain, paraesthesia may not be needed to obtain relief in headache conditions.

A retrospective analysis of seven patients with chronic refractory primary headache disorders followed up for a mean of 28 months also reported for the first time the long-term safety, tolerability and efficacy of HF10 cervical SCS not only in rCM but also in other primary headaches, namely CCH and SUNA (Fig. 7.2). Results demonstrated an improvement of at least 50% in headache frequency and/or intensity in all rCM subjects, improvement in attacks frequency and duration in one SUNA and headache freedom in the other SUNA patient. The CCH patient reported a reduction in attack duration but not frequency [35].