1207 Profiling the Genetic Landscape of Spine Surgery Utilization and Failure

INTRODUCTION: Current evidence suggests heritability in spinal pathologies including adult spinal deformity, intervertebral disc degeneration, and spondylotic myelopathy. Furthermore, heritability exists in both their sequela (i.e.: low back pain) and subsequent spine surgery utilization (SSU). The genes and biological processes governing the utilization and revision of spine surgery are unknown but represent a potential approach for improved patient selection. METHODS: A Genome Wide Association Study (GWAS) was conducted using the UK Biobank (N=502,364). Separate GWAS were conducted for SSU (N=11,096) and the revision of posterior spinal decompression (PSD) (No revision: N=5,260; Revision: N=414). The All of Us Research Program was utilized to externally validate the identified SNPs across the entire cohort and historically underrepresented racial groups. Machine learning models were developed to predict revision outcome from the identified SNPs. RESULTS: Baseline participant factors including demographic, environmental, radiographic, and previously developed polygenic risk scores (BMI, osteoporosis, rheumatoid arthritis) did not separate the SSU cohort from those who did not undergo spine surgery. Significant (p<10-5) SNPs were identified (SSU: 34; PSD revision: 9). SNP mapping demonstrated that coding sequence was often a functional consequence (SSU: 13/34; PSD revision: 4/9). Only few of the corresponding genes were implicated in literature (SSU: 5/34; PSD revision: 0/9). The 9 revision SNPs predicted revision outcome (F1 score: 0.22, Sensitivity: 0.24, Specificity: 0.92, PPV: 0.20, NPV: 0.93) where 77.8% (7/9) had odds ratio =4. CONCLUSIONS: SSU heritability and PSD revision are evidenced to be not solely mediated through collagen and bone related genes, but also those of exoskeletal stability/modeling (CNN1, COL6A6, PIGL, U2AF2, COBLL1, TTN), immune response (IFNA5, IFNA13, TRPV2), and neuronal excitability (KCNH8, PCLO, DPYSL2). The identified SNPs strongly predict revision surgery and may be developed towards improved patient selection.