Genetic Mechanisms and Additional Risk Factors Underlying Hip Dysplasia

Overview

Hip Dysplasia, or Developmental dysplasia of the hip (DDH) is a congenital disorder of the hip joint characterized by a shallow, or dysplastic hip socket, with potential risks of developing progressive joint dislocation, early osteoarthritis from young adulthood and serious functional disability. The Hip Cohort Study is the first longitudinal, population-based hip "phenobank" which includes standardized ultrasound examinations of the newborn hip, radiographs at skeletal maturity (around 19 years), as well as clinical data and DNA samples from the participants. The combination of genetic analyses with the rich radiological and clinical data collected at different life stages during the first two decades of life will enable identification of biological pathways (advanced genetic analyses) that are significantly associated with different radiological indices of hip dysplasia. This will allow for early, targeted treatment of the DDH disease and thus reduce the risk of later osteoarthritis.

Study population and DNA samples Clinical and radiological follow up of Hip Cohort Study members born during 1989 (n=4004, response rate 52%) and those revealing sonographically immature or dysplastic hips in the newborn period during 1988 and 1990 (n= 480, response rate 67.7%) was performed during 2007- 2009, thereby enabling characterisation of DDH in 2406 subjects. In 1779 of these, salivary samples were collected with consent, and DNA was extracted and stored in Professor Bill Ollier's laboratory at the University of Manchester. The initial Hip Cohort Study is well described in ClinicalTrials.gov ID: NCT01818934. Genome-wide genotyping, SNP and CNV association analyses: DNA samples will be sent to the Department of Clinical and Molecular Medicine, Norwegian University of Science and Technology (NTNU), Norway, for the genome-wide SNP (single nucleotide polymorphism) genotyping. For this study, we will use the latest version of Illumina Core Exome array, covering known genes, selected exones and promoter regions, as well as a backbone of common genetic variation allowing for imputation of genotypes that will widen the genomic content we will examine. Thus, this array will provide sufficient coverage of the genome for biological pathway analyses. Initial quality control will be performed in Illumina's Genome Studio Software. Subsequent quality control will be performed as described in the paper by Anderson et al in order to avoid common confounding factors such as batch effect (Lam et al 2020, Anderson et al 2010). Coverage of the genome-wide variation will be further improved by imputation of non-genotyped SNPs using directly genotyped SNPs and linkage disequilibrium information from the Haplotype Reference Consortium. Imputation will be performed using Impute2 software, according to recommended guidelines (Howie et al, 2011). In addition to the pathway analysis, an overall association analysis will be performed to uncover SNPs associated with DDH. For these analyses, standard quality control and statistical methods will be used (Purcell et al, 2007). Copy number variations (CNVs) will be determined using PennCNV (Wang, Li, Hadley et al, 2007) and their association with hip dysplasia indices will be examined using regression analyses. Given our current sample size, we have about 50% power to detect a variant in this sample (frequency of 20% of hip dysplasia, and a similar allele frequency and OR of the top candidate in GDF5). However, using multiple phenotypes and genetic analyses, we will be able to explore the genetics of hip dysplasia and hip shape in a unique way. Biological Pathway Analyses: The use of conventional genome-wide association methods does not fully explore the potentially complicated relationships between genetic variants, or between the proteins resulting from the genetic variants. Pathway- and network-based analysis methods have been developed to help address this problem ((e.g. Wang, Li, Bucan et al 2007; Torkamani et al 2008; Baranzini et al 2009; Eleftherohorinou et al 2009; Perry et al 2009; Peng et al 2010). These methods typically combine evidence of multiple genes/SNPs within a particular biological pathway, in order to establish if a pathway may be implicated in the development of a phenotype. Permutation methods are often employed to assess the significance in these analyses. This type of analysis may also help to interpret the biological meaning of association signals. This approach will allow the exploration of the relative contributions of different biological mechanisms possibly underlying DDH, using genetic information to group genes into putative pathways and relate them to different anatomical and clinical features of DDH. For the current study, biological pathways will be obtained from established databases, such as the Gene Ontology and Reactome databases, and other public resources (Ashburner et al, 2000; Croft et al, 2014). Pathways will be analysed in MAGMA, using the genome-wide SNP and/or CNV information to identify underlying biological pathways for an individual radiological index (de Leeuw et al, 2015). Distinctive biological mechanisms related to radiological indices may be identified through pathway analyses. To date, most predictors of severe DDH outcome have been based on clinical and x-ray findings, and on evaluation of the patient's functional impairment. We will, using multivariable prognostic models, construct a more focused, multivariate model for the prediction of improvement, based not only on traditional clinical and radiological risk factors but also on novel genetic information and imaging biomarkers.