The 2018 Otto Aufranc Award: How Does Genome-wide Variation Affect Osteolysis Risk After THA?

Background Periprosthetic osteolysis resulting in aseptic loosening is a leading cause of THA revision. Individuals vary in their susceptibility to osteolysis and heritable factors may contribute to this variation. However, the overall contribution that such variation makes to osteolysis risk is unknown.

Questions/purposes We conducted two genome-wide association studies to (1) identify genetic risk loci associated with susceptibility to osteolysis; and (2) identify genetic risk loci associated with time to prosthesis revision for osteolysis.

Methods The Norway cohort comprised 2624 patients after THA recruited from the Norwegian Arthroplasty Registry, of whom 779 had undergone revision surgery for osteolysis. The UK cohort included 890 patients previously recruited from hospitals in the north of England, 317 who either had radiographic evidence of and/or had undergone revision surgery for osteolysis. All participants had received a fully cemented or hybrid THA using a small-diameter metal or ceramic-on-conventional polyethylene bearing. Osteolysis susceptibility case-control analyses and quantitative trait analyses for time to prosthesis revision (a proxy measure of the speed of osteolysis onset) in those patients with osteolysis were undertaken in each cohort separately after genome-wide genotyping. Finally, a meta-analysis of the two independent cohort association analysis results was undertaken.

Results Genome-wide association analysis identified four independent suggestive genetic signals for osteolysis case-control status in the Norwegian cohort and 11 in the UK cohort (p ≤ 5 x 10^-6). After meta-analysis, five independent genetic signals showed a suggestive association with osteolysis case-control status at p ≤ 5 x 10^-6with the strongest comprising 18 correlated variants on chromosome 7 (lead signal rs850092, p = 1.13 x 10^-6). Genome-wide quantitative trait analysis in cases only showed a total of five and nine independent genetic signals for time to revision at p ≤ 5 x 10^-6, respectively. After meta-analysis, 11 independent genetic signals showed suggestive evidence of an association with time to revision at p ≤ 5 x 10^-6 with the largest association block comprising 174 correlated variants in chromosome 15 (lead signal rs10507055, p = 1.40 x 10^-7).

Conclusions We explored the heritable biology of osteolysis at the whole genome level and identify several genetic loci that associate with susceptibility to osteolysis or with premature revision surgery. However, further studies are required to determine a causal association between the identified signals and osteolysis and their functional role in the disease.

Clinical Relevance The identification of novel genetic risk loci for osteolysis enables new investigative avenues for clinical biomarker discovery and therapeutic intervention in this disease.