Osteoarthritis (OA) is the most common form of arthritis and one of the ten most disabling diseases in developed countries. Total joint replacement (TJR) is considered by far as the most effective treatment for end-stage OA patients. The majority of patients achieve symptomatic improvement following TJR. However, about 22% of the TJR patients either do not improve or deteriorate after surgery. Several potential non-genetic predictors for the TJR outcome have been investigated. However, the results were either inconclusive or had very limited predictive power. The aim of this study was to identify genetic variants for the poor outcome of TJR in primary OA patients by a genome-wide association study (GWAS). Study participants were total knee or hip replacement patients due to primary OA who were recruited to the Newfoundland Osteoarthritis Study (NFOAS) before 2017. The Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) was used to assess pain and functional impairment pre- and 3.99±1.38 years post-surgery. Two non-responder classification criteria were used in our study. One was defined by an absolute WOMAC change score. Participants with a change score less than 7/20 points for pain were considered as pain non-responders; and those with less than 22/68 points for function were classified as function non-responders. The second one was the Outcome Measures in Arthritis Clinical Trials and the Osteoarthritis Research Society International (OMERACT-OARSI) criteria. Blood DNA samples were genotyped using the Illumina GWAS microarrays genotyping platform. The quality control (QC) filtering was performed on GWAS data before the association of the genetic variants with non-responders to TJR was tested using the GenABEL package in R with adjustment for the relatedness of the study population and using the commonly accepted GWAS significance threshold p < 5*10. −8. to control multiple testing. In total, 316 knee and 122 hip OA patients (mean age 65.45±7.62 years, and 58% females) passed the QC check. These study participants included 368 responders and 56 non-responders to pain, and 364 responders and 68 non-responders to function based on the absolute WOMAC point score change classification. While 377 responders and 56 non-responders to pain, and 366 responders and 71 non-responders to function were identified by the OMERACT-OARSI classification criteria. Interestingly, the same results were obtained by both classification methods, and we found that the G allele of rs4797006 was significantly associated with pain non-responders with odds ratio (OR) of 5.12 (p<7.27×10. -10. ). This SNP is in intron one of the melanocortin receptor 5 (MC5R) gene on chr18. This gene plays central roles in immune response, pain sensitivity, and negative regulation of inflammatory response to antigenic stimulus. The A allele of rs200752023 was associated with function non-responders with OR of 4.41 (p<3.29×10. -8. ). The SNP is located in intron three of the RNA Binding Fox-1 Homolog 3 (RBFOX3) gene on chr17 which has been associated with numerous neurological disorders. Our data suggested that two chromosomal regions are associated with TJR poor outcomes and could be the novel targets for developing strategies to improve the outcome of the TJR

While total joint replacement (TJR) is considered as an effective intervention to relieve pain and restore joint function for end-stage osteoarthritis (OA) patients, a significant proportion of the patients are dissatisfied with their surgery outcomes. The aim of this study was to identify genetic factors that can predict patients who do or do not benefit from these surgical procedures by a genome-wide association study (GWAS). Study participants were derived from the Newfoundland Osteoarthritis Study (NFOAS) which consisted of 1086 TJR patients. Non-responders to TJR was defined as patients who did not reach the minimum clinically important difference (MCID) based on the self administered Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) in terms of pain reduction or function improvment. DNA was extracted from the blood samples of the study participants and genotyped by Illumina GWAS genotyping platform. Over two million single nucleotide polymorphisms (SNPs) across the genome were genotyped and tested for assocition with non-responders. 39 non-responders and 44 age, sex, and BMI matched responders were included in this study. Four chromosome regions on chromosomes 5, 7, 8, and 12 were suggested to be associated with non-responders with p < 1 0–5. The most promising one was on chromosome 5 with the lead SNP rs17118094 (p=1.7×10–6) which can classify 72% of non-responders accurately. The discriminatory power of this SNP alone is very promising as indicated by an area under the curve (AUC) of 0.72 with 95% confidence interval of 0.63 to 0.81, which is much better than any previously studied predictors mentioned above. All the patients who carry two copies of the G allele (minor allele) of rs17118094 were non-responders and 75% of those who carry one copy of the G allele were non-responders. The discriminatory ability of the lead SNPs on chromosomes 7 and 12 were comparable to the one on chromosome 5 with an AUC of 0.74, and 88% of patients who carry two copies of the A allele of rs10244798 on chromosome 7 were non-responders. Similarly, 88% of patients who carry two copies of the C allele of rs10773476 on chromosome 12 were non-responders. While the discriminatory ability of rs9643244 on chromosome 8 was poor with an AUC of 0.26, its strong association with non-responders warrants a further investigation in the region. The study identified four genomic regions harboring genetic factors for non-responders to TJR. The lead SNPs in those regions have great discriminatory ability to predict non-responders and could be used to create a genetic prediction model for clinical unitilty and application