Percutaneous retrogasserian glycerol rhizolysis for treatment of chronic intractable cluster headaches: long-term results.

OBJECTIVE

To analyze the long-term effectiveness and safety of percutaneous retrogasserian glycerol rhizolysis (PRGR) in the treatment of medically refractive chronic cluster headache (CH). The current mainstay of surgical intervention for these patients is percutaneous radiofrequency retrogasserian rhizotomy (PRFR). However, when performed for V1 distribution pathology, PRFR can lead to corneal anesthesia, which places the patient at risk for future visual loss. It also increases the risk of facial dysesthesia.

METHODS

In a prospective, consecutive series, 18 patients with intractable CH were followed for a mean of 5.2 years (range, 40-78 mo) after they had undergone PRGR, performed using a standard technique. The significance of this technique as an alternative to PRFR is that it should result in a lower rate of both corneal and facial anesthesia and provide an acceptable degree of pain relief.

RESULTS

Fifteen patients (83%) obtained immediate pain relief after one or two injections; the majority of them experienced relief after the first injection. CH recurred in seven patients (39%) over the course of the study. Two of these patients received a second injection, and both met with equal success. Two other patients underwent PRFR. Excluding those who underwent PRFR, the overall daily headache frequency decreased from 3.5 +/- 0.3 attacks per day preoperatively to 0.6 +/- 0.2 attacks per day at last follow-up. The severity of these headaches, as assessed by verbal pain scales, also decreased from 10 preoperatively to 4.4 +/- 1.4 at follow-up. None of the patients, including those who required a second procedure, experienced corneal anesthesia or facial dysesthesia.

CONCLUSIONS

This study provides the first long-term evaluation of PRGR for the treatment of medically refractive chronic CH and lends support to both the safety and long-term efficacy of this procedure. Further investigations are needed to compare directly the relative efficacy and safety of PRGR and PRFR.