Adolescent idiopathic scoliosis affects up to 3% of children from all ethnicities. African Americans present with larger curves, report greater pain, and have more surgical complications. Healthcare disparities also increase the risk for curve progression and serious, life-threatening neurologic and cardiac complications, many of which can be prevented with early diagnosis. In the initial funding period, we identified several important risk factors for severe scoliosis, including rare variants in fibrillin-1 (FBN1), musculoskeletal collagen genes, and distal chromosome 16p11.2 duplications that confer >10-fold increased risk. However, it is not yet known whether risk factors are generalizable to diverse patient populations. Even when genes are known, precisely determining the pathogenicity of individual human genetic variants remains a bottleneck because of the high frequency of variants of uncertain significance, meaning that there is insufficient clinical or functional data to assign them as either pathogenic or benign. In an era of Precision Medicine, there is an unmet need to change the clinical practice paradigm for scoliosis, however, our ability to interpret genetic data in underserved populations remains limited. Our central hypothesis is that genetic data, when combined with functional analysis, improves the diagnostic precision of variant interpretation for scoliosis and related conditions. To accomplish these goals, we will study a cohort of 1000 African American scoliosis patients recruited during the initial funding period in order to determine whether known risk factors are generalizable to African Americans, and to identify risk variants that can only be identified by studying individuals of African ancestry. Second, we will identify phenotypes associated with scoliosis risk variants in an older population using a gene-first approach that leverages the Geisinger DiscovEHR dataset consisting of >175,000 participants with linked electronic health record and exome sequence data. Finally, to test the hypothesis that knowledge of functional effects improves genetic variant classification, we will utilize high-throughput assays developed in our laboratory for deep mutational scanning. The effects of every possible coding variants in three genes associated with scoliosis and life-threatening aortic aneurysm (COL3A1, SMAD3, and FBN1) will be quantified. Computational classifiers to predict pathogenicity of variant alleles will be built based on our functional data. Variants will be validated in zebrafish models of scoliosis. Renewal of this multicenter study of scoliosis will speed up the pace of gene discovery and its clinical application for patients of all ages and ethnicities. By comprehensively and quantitatively determining the effects of genetic variants on protein function, as well as their impact on diverse individuals across the lifespan, we move closer to the goal of precision medicine for scoliosis.
Adolescent idiopathic scoliosis affects up to 3% of all children with >20,000 children undergoing major spinal fusion surgery annually at a cost of $5 billion. Identification of patients at risk for scoliosis curve progression or serious comorbidities is needed to implement prevention strategies but is hindered by lack of data from underrepresented minorities, including African Americans. In this project, we are setting a new research direction for our competitive renewal by proposing to study scoliosis in diverse populations and to improve genetic variant interpretation for all patients using high-throughput functional genomics.
