Background: Epidural hemostasis needs to use small, adapted material in minimally invasive surgery, including bilateral decompression via a unilateral approach for lumbar spinal stenosis. Most surgeons avoid external material for hemostasis because of possible neural tissue damage or complications. We compared epidural hemostasis in minimally invasive surgery by fat graft versus gelatin sponge.

Methods: The design was a prospective randomized controlled in-vivo human experimental study. The 24 levels operated on for lumbar spinal stenosis were evaluated in two groups: Group A (control group: gelatin sponge) and Group B (experimental group: fat graft). International Normalized Ratio and Prothrombin Time were assessed preoperatively. Number of cotton hemostats and systolic and diastolic blood pressure were assessed intraoperatively. Epidural hemorrhage area, spinal cord size and ratio of epidural hemorrhage area to spinal cord size were evaluated on early postoperative lumbar MRI.

Results: Mean epidural hemorrhage area in groups A and B was respectively 1.3±0.5 and 1.2±0.6 cm², and mean spinal cord size 1.2±0.6 and 1.8±0.6 cm² on early postoperative axial lumbar MRI. The two groups did not significantly differ in ratio of epidural hemorrhage/spinal cord size or number of intraoperative hemostats (p=0.36, and p=0.71).

Conclusions: The autologous fat graft ensured sufficient and safe epidural hemostasis without serious adverse events in minimally invasive spinal surgery, and is preferable as autologous tissue is easily and quickly harvested. The surgeon feels safe with this technique and does not need external hemostatic agents.