Is the Lateral Transpsoas Approach to the Lumbar Spine Safe at L4–L5?
Popularity of the lateral transpsoas approach (or lateral lumbar interbody fusion, LLIF, or extreme lateral interbody fusion, XLIF) for interbody fusion in the lumbar spine continues to grow. As it is performed today, the technique was originally described by Ozgur et al1 as a minimally invasive modification of the anterior retroperitoneal approach. The advantages of the lateral approach generally extolled in the literature include a minimally invasive incision, access to the anterior disk space without the need for an approach surgeon, and a theoretical decrease in complications associated with open anterior abdominal surgery.1 However, there are known limitations to the lateral approach, particularly the inaccessibility of the L5–S1 interspace, the frequency of postoperative psoas weakness and thigh numbness, and controversy with regard to the potential risks of serious neurological and vascular injury with accessing the L4–L5 disk space.2,3
It is generally considered common knowledge that moving in the caudal direction down the lumbar spine, the lumbar plexus migrates from a dorsal to ventral location on the lateral aspect of the disk. Thus, when accessing the disk space from a lateral approach, the plexus is at highest risk of injury at the L4–L5 level. Several studies have attempted to quantify a safe working space lateral to the L4–L5 disk: Uribe et al4 performed a cadaveric study describing the course of the genitofemoral nerve and found that it is frequently located along the anterior medial fourth of the L4 and L5 vertebral bodies, which the authors state puts it at risk of direct injury during the lateral approach. In an anatomical study based on magnetic resonance imaging (MRI), Guerin et al5 found that the window for safe access at L4–L5 was only 37.8% of the lower endplate of the vertebral body sagittal diameter, which was statistically significantly smaller than at L1–L2. Banagan et al6 performed a cadaveric study in which the authors recreated a standard clinical minimally invasive transpsoas approach with guide wires placed into the disk space under fluoroscopic guidance. The authors found that at the L4–L5 disk space, there the guidewire actually caused direct nerve injury.6 In fact, the authors conclude that “there is no zone of absolute safety when using the direct lateral transpsoas approach at the L4–L5 level.”6 Anatomically, the neurological risks inherent in traversing the psoas at this level cannot be overstated.
A retrospective cohort study published in 2012 found that injury to the femoral nerve occurred in 1.7% of 118 total patients, but the incidence was 4.8% for LLIF performed at the L4–L5 disk space; no femoral nerve injuries occurred at any of the other levels.3 The authors therefore recommend only “judicious use” of LLIF at L4–L5 because of the risk of severe nerve injury.3 Another retrospective study of 235 patients found a similar trend for lumbar plexus-related motor deficits in patients undergoing L4–L5 LLIF, though the authors were unable to show a statistically significant difference (OR=3.79; P=0.145).7 A recent systematic review of 24 studies examining postoperative thigh pain and neurological dysfunction after LLIF found that thigh pain, weakness, and paresthesias were the most commonly reported postoperative symptoms, with the rate ranging as high as 60.7%.8 The authors found that a longer operative time, and specifically including the L4–L5 disk space, were independent risk factors for the development of postoperative thigh symptoms.8 Further corroborating these results, Lykissas et al9 evaluated motor deficits after LLIF in over 900 patients; with a multivariate regression model, the authors demonstrated that LLIF involving the L4–L5 level significantly increased the risk of immediate postoperative sensory deficit (OR, 1.774; P=0.45).