A prospective cohort study of microvascular decompression and Gamma Knife surgery in patients with trigeminal neuralgia.

Object: The aim of this study was to analyze 1 surgeon's 4-year experience with microvascular decompression ([MVD], 36 patients) and Gamma Knife surgery ([GKS], 44 patients) in 80 consecutive patients with trigeminal neuralgia (TN).

Methods: The authors conducted a prospective cohort study from March 1999 to December 2003 with an independent clinical assessment of the results and serial patient satisfaction surveys. All patients completed a 2004 patient satisfaction survey (0.5-5 years postoperative), and 70% of surviving patients completed the same survey in 2007 (3.9-8.5 years postoperative). Follow-up was undertaken in 100% of the patients (mean 3.4 +/- 2.14 years, range 0.17-8.5 years).

Results: Respective initial and latest follow-up raw pain-free rates were 100 and 80.6% for MVD and 77.3 and 45.5% for GKS. The median time to the maximal benefit after GKS was 4 weeks (range 1 week-6 months). Respective initial, 2-, and 5-year actuarial pain-free rates were 100, 88, and 80% for MVD and 78, 50, and 33% for GKS (p = 0.0002). The relative risk of losing a pain-free status by 5 years posttreatment was 3.35 for patients in the GKS group compared with the MVD group. Initial and 5-year actuarial rates for >or= 50% pain relief after GKS were 100 and 80%, respectively. The respective rates of permanent mild and severe sensory loss were 5.6 and 0% for patients in the MVD group, as opposed to 6.8 and 2.3% for patients in the GKS group. Anesthesia dolorosa did not occur during the study. Both procedures enjoyed a high degree of early patient satisfaction (95-100%). Microvascular decompression maintained the same rate of patient satisfaction, but satisfaction with GKS decreased to 75% as pain control waned. Twenty-three patients (29%) died of causes unrelated to the TN or the surgical intervention during the follow-up, and their pain status was known at the time of death. Statistically significant intergroup differences for the MVD versus GKS cohorts were age (median 54 years, range 36-70 years vs median 74 years, range 48-92 years, respectively), preoperative symptom duration (median 2.58 years, range 0.33-15 years vs median 7.5, range 0.6-40 years, respectively), and the presence of major comorbidities (2.8 vs 58.3%, respectively).

Conclusions: In this nonrandomized prospective cohort trial of selected patients with potentially relevant intergroup differences, MVD was significantly superior to GKS in achieving and maintaining a pain-free status in those with TN and provided similar early and superior longer-term patient satisfaction rates compared with those for GKS. The complications of wound cerebrospinal fluid leakage, hearing loss, and persistent diplopia (1 case each in the MVD group) were not seen after GKS.