Cindy Boer

Hand Osteoarthritis Phenotypes | 105 2.2 * ** * * * ** *** **** ** * * ** * * # *** * *** *** *** * *** * * ** ** ** *** * * * * # *** * * * * *** *** *** * * ** * * *** *** * * * *** Thumb KLsum Finger Klsum Hand KLsum rs116882138 SNV Chr 9 27313557 A 0.02 1.34 5.1*10 Hospital diagnosedhip andor knee /Kneeª MOB3B/EQTN rs13283416 9 119301607 G 0.42 1.10 5.3*10 - HipOA ASTN2 rs4836732 9 119266695 C 0.47 1.20 6.1*10 -10 Hip THR ASTN2 rs2480930 9 117842307 A 0.49 1.09 6.6*10 -12 HipOA ASTN2 rs1330349 9 117840742 C 0.60 1.08 4.1*10 -11 HipOA ASTN2 rs919642 9 116911147 T 0.27 1.05 8.5*10 -15 OA COL27a1 rs1078301 9 116909146 T 0.27 1.07 1.4*10 -10 KneeOA COL27a1 rs4730250 7 107207695 G 0.17 1.17 9.2*10 -09 Knee COG5 rs3815148 7 106938420 C 0.23 1.14 8.0*10 -08 Knee and/orhand COG5 rs10948172 6 44777691 G 0.29 1.14 7.9*10 -08 Hip and/or knee RUNX2 rs12154055 6 44449697 G 0.34 1.03 2.7*10 -08 OA RUNX2 rs3884606 5 170871074 G 0.47 1.04 8.3*10 -09 Knee andHipOA FGF18 rs1913707 4 13039440 A 0.59 1.08 3.0*10 -11 HipOA RAB28 rs143383 20 34025983 T 0.64 1.16 8.3*10 -09 Knee,Hip GDF5 rs143384 20 34025756 A 0.61 1.10 1.4*10 -19 HipOA GDF5 rs3771501 2 70717653 G 0.53 0.94 1.7*10 -08 Self-reported / anysite TGFa rs4252548 19 55879672 T 0.01 1.32 2.0*10 -12 HipOA IL11 rs12982744 19 2177193 C 0.61 1.17 7.8*10 -09 Hip DOT1L rs1560707 19 10750738 T 0.40 1.04 1.4*10 -13 OA SLC44A2 rs10502437 18 20970706 G 0.42 1.03 2.5*10 -08 OA TMEM241 rs6499244 16 69735271 A 0.52 1.06 3.9*10 -11 KneeOA WWP2 rs12901071 15 67370389 A 0.37 1.08 3.1*10 -10 Hip and /or knee SMAD3 rs3204689 15 58246802 C 0.52 1.46 1.1*10 -11 SevereHand ALDH1A2 rs4764133 12 15064363 T 0.39 0.83 b 1.8*10 -15 HandKLsum MGP rs1060105 12 123806219 C 0.77 1.07 1.9*10 -08 KneeOA SBNO1 rs11059094 12 122606837 T 0.45 1.08 7.4*10 -11 HipOA BCL7A rs17659798 11 28874997 A 0.69 1.06 2.1*10 -10 Knee andHipOA LMX1B rs11583641 1 183906245 C 0.68 1.08 5.6*10 -10 HipOA COLGALT2 rs4338381 1 103572927 A 0.62 1.10 4.4*10 -15 HipOA COL11A1 Pos EA EAF OR P-value Phenotype Locus Hand OA Phenotypes −2 −1 0 1 2 Beta (scaled) Legend P-value & Beta * < 0.05 ** < 0.005 *** < 0.0005 **** < 5*10 -06 # < 5*10 -08 Thumb KLsum Finger Klsum Hand KLsum SNV Chr EA EAF Locus # # # # # # * rs10916199 1 A 0.81 [ZNF678] rs2070852 11 C 0.69 [F2] rs4764133 12 C 0.62 MGP-[]-ERP27 rs12049916 12 A 0.77 [CCDC91] a. b. ▲ Figure 4: Heatmap depicting the effect of OA associated SNVs in in each hand OA phenotype. a. The found associated lead SNVs of the hand OA phenotypes and their beta and p-value in the other phenotypes. b. Depicted are all known OA SNVs which had a nominal significant effect in at least one stratified hand OA phenotype. All beta’s were calculated for the reported effect allele and scaled. Co- lours represent the scaled beta of the effect allele, which is here the minor allele. P-values are repre- sented by * in the box. Gene: reported gene from the GWAs study. Chr, chromosome; Pos: base pair position on the chromosome Hg19; EA, effect allele; EAF, effect allele frequency; OR, odds ratio; KL, kellgren-Lawrence score. were previously found to be associated with knee and/or hip OA, they may reflect common mechanisms across all joints in OA. Discussion We identified four genome-wide significant loci associated with hand OA phenotypes, of which two were novel and specific for thumb OA. Integration of multiple lines of data provided cumulative evidence that WNT9A may be a causal gene for thumb OA. We first conducted a cluster analysis of hand joints to identify less heterogeneous clusters of joints that served as the basis of the hand OA phenotypes assessed in this GWAS. With this approach of using radiographically defined biologically relevant OA phenotypes to reduce phenotype heterogeneity and increase statistical power, we were able to robust-ly identify known and novel OA loci despite our modest sample