The efficacy of minimally invasive discectomy compared with open discectomy: a meta-analysis of prospective randomized controlled trials.

Object: Advocates of minimally invasive discectomy (MID) have promoted this operation as an alternative to open discectomy (OD), arguing that there may be less injury to the paraspinal muscles, decreased postoperative pain, and a faster recovery time. However, a recently published large randomized controlled trial (RCT) comparing these approaches reported inferior relief of leg pain in patients undergoing MID. The authors conducted a meta-analysis to evaluate complications and improvement in leg pain in patients with radiculopathy enrolled in RCTs comparing OD to MID.

Methods: The authors performed a literature search using Medline and EMBASE of studies indexed between January 1990 and January 2011. Predetermined RCT eligibility included the usage of tubular retractors during MID, a minimum follow-up duration of 1 year, and quantification of pain with the visual analog scale (VAS). Trials that only evaluated patients with recurrent disc herniation were excluded. Data on operative parameters, complications, and VAS scores of leg pain were extracted by 2 investigators. A meta-analysis was performed assuming random effects to determine the difference in mean change for continuous outcomes and the risk ratio for binary outcomes.

Results: Six trials comprising 837 patients (of whom 388 were randomized to MID and 449 were randomized to OD) were included. The mean operative time was 49 minutes during MID and 44 minutes during OD; this difference was not statistically significant. Incidental durotomies occurred significantly more frequently during MID (5.67% compared with 2.90% for OD; RR 2.05, 95% CI 1.05-3.98). Intraoperative complications (incidental durotomies and nerve root injuries) were also significantly more common in patients undergoing MID (RR 2.01, 95% CI 1.07-3.77). The mean preoperative VAS score for leg pain was 6.9 in patients randomized to MID and 7.2 in those randomized to OD. With long-term follow-up (1-2 years postoperatively), the mean VAS score improved to 1.6 in both the MID and OD cohorts. There was no significant difference in relief of leg pain between the 2 approaches with either short-term follow-up (2-3 months postoperatively, 0.81 points on the VAS, 95% CI -4.71 to 6.32) or long-term follow-up (2.64 on the VAS, 95% CI -2.15 to 7.43). Reoperation for recurrent herniation was more common in patients randomized to the MID group (8.50% compared with 5.35% in patients randomized to the OD group), but this difference was not statistically significant (RR 1.56, 95% CI 0.92-2.66). Total complications did not differ significantly between the operations (RR 1.50, 95% CI 0.97-2.33).

Conclusions: The current evidence suggests that both OD and MID lead to a substantial and equivalent long-term improvement in leg pain. Adequate decompression, regardless of the operative approach used, may be the primary determinant of pain relief-the major complaint of many patients with radiculopathy. Incidental durotomies occurred significantly more frequently during MID, but total complications did not differ between the techniques.