Is L5/S1 interbody fusion necessary in long-segment surgery for adult degenerative scoliosis? A systematic review and meta-analysis.

OBJECTIVE There is no consensus regarding the best surgical strategy at the lumbosacral junction (LSJ) in long constructs for adult spinal deformity (ASD). The use of interbody fusion (IF) has been advocated to increase fusion rates, with additional pelvic fixation (PF) typically recommended. The actual benefit of IF even when extending to the pelvis, however, has not been vigorously analyzed. The goal of this work was to better understand the role of IF, specifically with respect to arthrodesis, when extending long constructs to the ilium. METHODS A systematic review of the PubMed and Cochrane databases was performed to identify the relevant studies in English, addressing the management of LSJ in long constructs (defined as ≥ 5 levels) in ASD. The search terms used were as follows: "Lumbosacral Junction," "Long Constructs," "Long Fusion to the Sacrum," "Sacropelvic Fixation," "Interbody Fusion," and "Iliac Screw." The authors excluded technical notes, case reports, literature reviews, and cadaveric studies; pediatric populations; pathologies different from ASD; studies not using conventional techniques; and studies focused only on alignment of different levels. RESULTS The PRISMA protocol was used. The authors found 12 retrospective clinical studies with a total of 1216 patients who were sorted into 3 different categories: group 1, using PF or not (n = 6); group 2, using PF with or without IF (n = 5); and group 3, from 1 study comparing anterior lumbar interbody fusion versus transforaminal lumbar interbody fusion. Five studies in group 1 and 4 in group 2 had pseudarthrosis rate as primary outcome and were selected for a quantitative analysis. Forest plots were used to display the risk ratio, and funnel plots were used to look at the risk of publication bias. The summary risk ratios were 0.36 (0.23-0.57, p < 0.001) and 1.03 (0.54-1.96, p = 0.94) for the PF and IF, respectively; there is a protective effect of overall pseudarthrosis for using PF in long constructs for ASD surgeries, but not for using IF. CONCLUSIONS The long-held contention that L5/S1 IF is always advantageous in long-construct deformity surgery is not supported by the current literature. Based on the findings from this systematic review and meta-analysis, PF with or without additional L5/S1 interbody grafting demonstrates similar overall construct pseudarthrosis rates. The added risk and costs associated with IF, therefore, should be more closely considered on a case-by-case basis.