The minimally invasive interbody selection algorithm for spinal deformity.

OBJECTIVE Minimally invasive surgery (MIS) for spinal deformity uses interbody techniques for correction, indirect decompression, and arthrodesis. Selection criteria for choosing a particular interbody approach are lacking. The authors created the minimally invasive interbody selection algorithm (MIISA) to provide a framework for rational decision-making in MIS for deformity. METHODS A retrospective data set of circumferential MIS (cMIS) for adult spinal deformity (ASD) collected over a 5-year period was analyzed by level in the lumbar spine to identify surgeon preferences and evaluate segmental lordosis outcomes. These data were used to inform a Delphi session of minimally invasive deformity surgeons from which the algorithm was created. The algorithm leads to 1 of 4 interbody approaches: anterior lumbar interbody fusion (ALIF), anterior column release (ACR), lateral lumbar interbody fusion (LLIF), and transforaminal lumbar interbody fusion (TLIF). Preoperative and 2-year postoperative radiographic parameters and clinical outcomes were compared. RESULTS Eleven surgeons completed 100 cMISs for ASD with 338 interbody devices, with a minimum 2-year follow-up. The type of interbody approach used at each level from L1 to S1 was recorded. The MIISA was then created with substantial agreement. The surgeons generally preferred LLIF for L1-2 (91.7%), L2-3 (85.2%), and L3-4 (80.7%). ACR was most commonly performed at L3-4 (8.4%) and L2-3 (6.2%). At L4-5, LLIF (69.5%), TLIF (15.9%), and ALIF (9.8%) were most commonly utilized. TLIF and ALIF were the most selected approaches at L5-S1 (61.4% and 38.6%, respectively). Segmental lordosis at each level varied based on the approach, with greater increases reported using ALIF, especially at L4-5 (9.2°) and L5-S1 (5.3°). A substantial increase in lordosis was achieved with ACR at L2-3 (10.9°) and L3-4 (10.4°). Lateral interbody arthrodesis without the use of an ACR did not generally result in significant lordosis restoration. There were statistically significant improvements in lumbar lordosis (LL), pelvic incidence-LL mismatch, coronal Cobb angle, and Oswestry Disability Index at the 2-year follow-up. CONCLUSIONS The use of the MIISA provides consistent guidance for surgeons who plan to perform MIS for deformity. For L1-4, the surgeons preferred lateral approaches to TLIF and reserved ACR for patients who needed the greatest increase in segmental lordosis. For L4-5, the surgeons' order of preference was LLIF, TLIF, and ALIF, but TLIF failed to demonstrate any significant lordosis restoration. At L5-S1, the surgical team typically preferred an ALIF when segmental lordosis was desired and preferred a TLIF if preoperative segmental lordosis was adequate.