Study design:

Systematic review and meta-analysis.

Objective:

To compare complication rates and clinical and radiologic outcome between the mini-open prepsoas and mini-open transpsoas approaches for lateral lumbar interbody fusion.

Summary of background data:

Both approaches are believed to be safe with similar complication rates. Previous studies suggest that the rate of neurological injury might be higher in the transpsoas group, whereas visceral or vascular injury might be more frequent in the prepsoas group.

Methods:

A systematic review of the literature was performed. Data were extracted from original publications up until December 26, 2018. Evidence was extracted from well-designed case-control or cohort studies and sorted in 2 groups, the prepsoas and transpsoas approaches. A meta-analysis was performed using a random-effects model (I statistic >50% for all analyses).

Results:

A total of 115 studies included data of 13,260 patients, 2450 in the prepsoas group and 10,810 in the transpsoas group. Demographics for prepsoas versus transpsoas group were (N-weighted means): age 61.9 versus 60.9 years; %female sex 53% versus 63%, levels fused 1.4 versus 2.6, blood loss 52.4 versus 122.3 mL, and operating time 125.1 versus 200.7 min. The following statistically significant differences in complication rates between prepsoas and transpsoas approaches were found: transient psoas weakness or thigh/groin numbness 4% versus 26% [95% confidence interval (CI): 11%-17%], motor neural injury 0.4% versus 1.3% (95% CI: 16%-62.3%); no statistically significant differences were found for: major vascular injury 2% versus 1% (95% CI: 1.04%-2.31%), kidney or ureter injury 0.04% versus 0.08% (95% CI: 0.057%-5.2%), injury pleural/peritoneal structures 0.6% versus 0.2% (95% CI: 0.89%-6.58%), cage subsidence 5% versus 4% (95% CI: 0.9%-1.97%), surgical site infection 1% versus 1% (95% CI: 0.57%-1.66%), abdominal wall pseudohernia 1% versus 1% (95% CI: 0.07%-21.22%), sympathetic chain injury 5% versus 0% (95% CI: 0.34%-97.86%), and directly procedure-related death 0.04% versus 0% (95% CI: 0.127%-76.8%). Pooled mean perioperative changes between prepsoas and transpsoas approaches were: segmental sagittal Cobb angle 3.07 versus 1.99 degrees; foraminal height 2 versus 6.96 mm.

Conclusions:

The prepsoas had fewer complications than the transpsoas approach. Furthermore, the prepsoas approach showed superior restoration of segmental lordosis, whereas foraminal height restoration was superior with the transpsoas approach. This could be explained by the differences in location of the interbody device placement in relation to the center of rotation of the spine between the 2 surgical techniques.