Risk of Injury to Retroperitoneal Structures in Prone and Lateral Decubitus Transpsoas Approaches to Lumbar Interbody...

Introduction The retroperitoneal approach for lateral lumbar interbody fusion (LLIF) originally described an initial posterolateral fascial incision enabling finger dissection from behind the peritoneum and guidance of instruments through a second direct-lateral fascial incision. It has since become common for single direct-lateral incisional access to the retroperitoneum. This study attempted to quantify the distance of the peritoneum from posterior landmarks in the space, assess the risk of peritoneal violation in each access trajectory (i.e., posterolateral versus direct lateral retroperitoneal dissection), and determine whether there are differences based on patient position (prone versus lateral decubitus). Methods In three prone cadaveric torsos, Steinman pins were percutaneously placed mid-disc at each level L2-5 bilaterally (for a total of 18 prone approaches). Open dissections exposed the retroperitoneum including the quadratus lumborum and psoas muscles, maintaining the natural reflection of the peritoneum. Visual assessment qualified whether any pin violated any retroperitoneal structure. Distance from the anterior border of the quadratus lumborum to the posterior-most reflection of the peritoneum was measured. For comparison, three additional torsos were positioned in lateral decubitus, and the above steps were repeated, only unilaterally (for a total of nine lateral decubitus approaches). Results In prone, no pin violated the peritoneum; three (3/18 total approaches) violated the kidney, all at L2-3 (3/6 approaches at L2-3). In lateral decubitus, all three L2-3 pins violated the kidney (3/3 approaches at L2-3); five of the six remaining pins from L3-5 violated the peritoneum (totaling eight violations in the nine total approaches). The incidence of any violation was significantly greater in lateral decubitus vs. prone (8/9 vs. 3/18, p=0.0006). The structure at risk (kidney vs. peritoneum) was significantly associated with disc level (p=0.0041): all kidney violations occurred at L2-3 and all peritoneal violations occurred at L3-4 or L4-5. Distance from the quadratus lumborum to the posterior-most reflection of the peritoneum averaged 8.7 cm (range: 6-10) in prone, and 2.9 cm (range: 2.5-3.2) in lateral decubitus (p=0.0129). Conclusion A cadaveric study of retroperitoneal anatomy demonstrates that there is an increased distance from the quadratus lumborum to the peritoneum in prone versus lateral decubitus and that the trajectory of approach to the lumbar discs risks violation of the peritoneum more frequently when accessing directly laterally versus posterolaterally. In either approach, care should be taken to identify and release the peritoneal reflection to create a safe passage to the lumbar discs.