There is extensive literature reporting discordance between the presence and severity of symptoms and the degree of radio-graphic structural osteoarthritis (OA).1–5 Genetic differences may account for some of this discordance. Indeed, certain genetic variants implicated in pain sensitivity have been shown to be significantly different between asymptomatic radiographic cases of OA and symptomatic cases.6–9
The catechol-O-methyltransferase, encoded by the COMT gene, is a major degrading enzyme in the metabolic pathways of catecholaminergic neurotransmitters.10 Genetic variation at the COMT gene has been shown to result in differential pain sensitivity.10–12 Carriers of the Val158Met COMT variant have been reported to have a higher risk (OR=2.9, 95% CI 1.2 to 6.1) of hip pain as compared with carriers of the Val/Val genotype among those with hip OA.9 This result has not been replicated in independent cohorts, nor for OA in other joints.
We assessed whether the Met allele in the COMT gene is involved in increased risk of symptomatic knee OA in seven cohorts: five cohorts from the UK, one from Australia and two from the USA were included (table 1). Assembly of the cohorts was approved by the local research ethics committees and all study participants gave fully informed consent to participate in genetic studies.
Table 1.
Descriptive characteristics of study subjects
| Cohort name | Genetics of osteoarthritis and lifestyle | Hertfordshire cohort study | Nottingham case-control study | TwinsUK study | Chingford study | Tasmanian older adult cohort | Health ABC study | Framingham osteoarthritis study |
|---|---|---|---|---|---|---|---|---|
| Cohort acronym | GOAL | HCS | Nott | TwinsUK | Chingford | TASOAC | HABC | FOA |
| Reference for cohort details | 8 | 8 | 8 | 8 | 8 | 8 | 13 | 14 |
| Country of origin | UK | UK | UK | UK | UK | Australia | USA | USA |
| Knee OA | ||||||||
| Symptomatic* | n=1199 | n=86 | n=1703 | n=45 | n=127 | n=86 | n=361 | n=327 |
| Age years mean (SD) | 68.4 (7.2) | 65.3 (2.8) | 68.7 (8.9) | 57.5 (7.2) | 65.7 (5.8) | 65.0 (7.2) | 75.1 (3.0) | 73.1 (8.5) |
| F gender % | 46.88% | 44.3% | 55.3% | 100% | 100% | 58.50% | 62.6% | 62.1% |
| BMI kg/m2 mean (SD) | 31.2 (5.4) | 29.7 (5.4) | 29.7 (5.4) | 27.9 (4.5) | 26.7 (3.4) | 31.3 (6.3) | 30.4 (5.3) | 30.3 (6.2) |
| K/L≥3% | 89.7% | 38.6% | 80.8% | 29.0% | 44.1% | 42.6% | 65.9% | 56.3% |
| TKR % | 94.8% | 9.3% | 90.3% | 26.7% | 23.5% | 0% | 16.9% | 13.8% |
| COMT Met+%† | 77.6% | 67.2% | 75.9% | 74.4% | 81.9% | 72.7% | 64.0% | 78.6% |
| Asymptomatic* | n=376 | n=44 | n=0 | n=235 | n=58 | n=136 | n=88 | n=222 |
| Age years mean (SD) | 69.0 (6.7) | 65.2 (6.4) | N/A | 58.1 (7.8) | 75.1 (2.9) | 75.5 (7.6) | 75.1 (2.9) | 65.8 (7.6) |
| F gender % | 43.9% | 31.8% | N/A | 100% | 100% | 43.80% | 61.4% | 55.4% |
| BMI kg/m2 mean (SD) | 28.9 (4.9) | 28.4 (4.0) | N/A | 26.7 (4.9) | 26.2 (4.4) | 28.1 (4.3) | 30.1 (5.0) | 27.9 (5.1) |
| K/L≥3% | 49.5% | 15.9% | N/A | 34.0% | 26.4% | 24.8% | 55.7% | 37.8% |
| TKR % | 0% | 0% | N/A | 0% | 0% | 0% | 0% | 0% |
| COMT Met+% | 76.4% | 73.8% | N/A | 72.6% | 70.7% | 77.9% | 55.7% | 75.2% |
| Unaffected | n=743 | n=542 | n=728 | n=978 | n=401 | n=329 | n=631 | n=1270 |
| Age years mean (SD) | 62.6 (8.4) | 64.8 (2.7) | 66.3 (7.4) | 52.9 (7.2) | 62.9 (5.7) | 61.8 (7.5) | 74.4 (2.8) | 67.9 (9.9) |
| F gender % | 49.9% | 50.1 % | 57.4% | 100% | 100% | 47.9% | 58.9% | 58.2% |
| BMI kg/m2 mean (SD) | 27.1 (4.4) | 26.2 (3.8) | 26.6 (3.9) | 24.5 (4.0) | 24.7 (3.6) | 27.4 (4.8) | 24.7 (3.6) | 27.4 (4.8) |
| COMT Met+% | 74.4% | 77.5% | 74.1% | 73.7% | 77.0% | 76.9% | 67.0% | 76.4% |
Subjects were considered to have symptomatic knee OA if they had at least one knee with radiographic knee OA (as defined by Kellgren and Lawrence (KL) grade≥2) and if they had pain in the same knee. Symptoms were considered present in a knee if pain (or in some cohorts, pain, aching or stiffness) was present in the knee for most days of the month during the prior year, or most days of the prior month. Those with a total knee replacement were classified as symptomatic regardless of current pain status. Because the Nottingham Case-Control Study did not have any asymptomatic subjects and the Hertfordshire Cohort Study had very few symptomatic subjects, and these two UK study cohorts comprised participants of similar ages, data from these two cohorts were pooled.
Genomic DNA was extracted from peripheral blood. For GOAL, Chingford, Hertfordshire Cohort Study (HCS), TASOAC and Nottingham samples, the rs4680 single nucleotide polymorphism (SNP) genotyping was carried out by Kbioscience Ltd, Hertfordshire UK. The TwinsUK samples were genotyped with the Infinium HumanHap 300 and 610 assays (Illumina, San Diego, USA). FOA Study samples were genotyped using the Affymetrix GeneChip Human Mapping 500 K array set (Santa Clara, California). Health ABC samples were genotyped on the Illumina Human1M-Duo BeadChip array. For both TwinsUK and HABC, rs4680 was directly typed. The SNP was imputed in FOA and had MACH_RS=0.963006, average MACH-QUAL=0.981084. The polymorphism was in Hardy–Weinberg equilibrium in controls for all cohorts (p>0.05). A dominant effect model was used.
KL, Kellgren–Lawrence; OA, osteoarthritis; TKR, total knee replacement.
COMT Met+; frequency of carriage of a Met (either Met-Met or Met-Val) at codon position 158 of the COMT gene.
The association between the Met+ genotype at COMT position 158 and knee OA was evaluated using logistic regression adjusting for age, sex and body mass index (BMI). The ORs were meta-analysed using the methods described in reference.8 For the symptomatic versus asymptomatic knee OA comparison, we also adjusted for Kellgren and Lawrence grade. We also evaluated sex-specific associations. Robust variance estimation was used for the TwinsUK and Framingham Osteoarthritis Study (FOA) studies with regard to relatedness. Given the sample size available, the study is powered to detect with p<0.05 and 80% power an OR of 1.245 for the 158COMTMet+ genotype comparing symptomatic with asymptomatic knee OA and an OR of 1.14 for symptomatic knee OAversus controls.
The association between this genotype and the presence of knee OA, irrespective of symptoms, was close to the null for most cohorts, as was the association with asymptomatic knee OA versus controls (table 2). The effect estimate was suggestive of a 10% increased prevalence of symptoms in those with knee OA compared with those without the polymorphism (adjusted OR 1.10 (95% CI 0.95 to 1.27), p=0.2), but did not reach statistical significance and is much lower than the original report on symptomatic hip OA.9 Similar results were seen in sex-stratified analyses (table 2).
Table 2.
Association between the 158COMTMet+ genotype and symptomatic or asymptomatic knee OA in seven independent cohorts adjusted for covariates
| Cohort study | Adjusted OR (95% CI), p value |
|||
|---|---|---|---|---|
| Any knee OA versus controls* | Symptomatic knee OA versus controls* | Asymptomatic knee OA versus controls* | Symptomatic knee OA versus asymptomatic knee OA† | |
| Chingford (only F) | F=1.09 (0.70, 1.69) | F=1.38 (0.82, 2.34) | F=0.66 (0.35, 1.26) | F=1.38 (0.82, 2.34) |
| Framingham | B=1.13 (0.87, 1.47) | B=1.22 (0.89, 1.67) | B=1.02 (0.72, 1.45) | B=1.10 (0.72, 1.68) |
| M=1.33 (0.89, 1.99) | M=1.65 (0.97, 2.80) | M=1.04 (0.62, 1.74) | M=1.63 (0.81, 3.27) | |
| F=1.00 (0.72, 1.34) | F=1.02 (0.69, 1.50) | F=1.00 (0.62, 1.60) | F=0.88 (0.50, 1.52) | |
| GOAL | B=1.06 (0.84, 1.34) | B=1.11 (0.87, 1.43) | B=0.96 (0.71, 1.3) | B=1.03 (0.78, 1.36) |
| M=0.89 (0.65, 1.21) | M=0.97 (0.69, 1.35) | M=0.94 (0.72, 1.23) | M=1.31 (0.86, 1.99) | |
| F=1.33 (0.92, 1.93) | F=1.38 (0.92, 2.07) | F=1.01 (0.73, 1.39) | F=0.70 (0.43, 1.14) | |
| Health ABC | B=0.95 (0.72, 1.25) | B=1.03 (0.76, 1.40) | B=0.69 (0.43, 1.11) | B=1.43 (0.88, 2.32) |
| M=1.25 (0.81, 1.92) | M=1.27 (0.80, 1.01) | M=1.26 (0.57, 2.77) | M=0.95 (0.42, 2.17) | |
| F=0.80 (0.56, 1.14) | F=0.89 (0.61, 1.31) | F=0.48 (0.26, 0.88) | F=1.84 (0.99, 3.40) | |
| HCSNott | B=0.99 (0.88, 1.11) | B=1.00 (0.89, 1.13) | B=0.74 (0.48, 1.15) | B=1.31 (0.85, 2.02) |
| M=0.90 (0.68, 1.18) | M=0.93 (0.71, 1.23) | M=1.12 (0.44, 2.81) | M=0.80 (0.31, 2.04) | |
| F=1.01 (0.77, 1.33) | F=1.00 (0.76, 1.32) | F=0.61 (0.20, 1.84) | F=1.55 (0.51, 4.74) | |
| TASOAC | B=0.88 (0.70, 1.09) | B=0.92 (0.70, 1.22) | B=1.00 (0.81, 1.24) | B=0.94 (0.72, 1.24) |
| M=1.17 (0.66, 2.05) | M=0.74 (0.48, 1.15) | M=1.09 (0.70, 1.69) | M=0.47 (0.18, 1.19) | |
| F=0.48 (0.26, 0.91) | F=1.14 (0.72, 1.8) | F=1.41 (0.89, 2.24) | F=0.52 (0.20, 1.37) | |
| TwinsUK (only F) | F=0.98 (0.71, 1.35) | F=1.03 (0.51, 2.11) | F=0.96 (0.68, 1.36) | F=1.11 (0.53, 2.37) |
| Summary fixed effects by gender | M=1.02 (0.87, 1.20) | M=1.00 (0.85, 1.19) | M=1.01 (0.83, 1.22) | M=1.12 (0.84, 1.51) |
| F=0.98 (0.86, 1.12) | F=1.08 (0.92, 1.25) | F=0.93 (0.78, 1.1) | F=1.04 (0.82, 1.33) | |
| Overall summary fixed effects (95% CI) | 0.99 (0.92, 1.08) p=0.88 | 1.04 (0.95, 1.13) p=0.42 | 0.93 (0.82, 1.06) p=0.28 | 1.10 (0.95, 1.27) p=0.21 |
| Met+ carriers | ||||
| Inter-study heterogeneity I2 (95% CI) | 0% (0%, 36%) | 0% (0%, 48.6%) | 0% (0%, 62.4%) | 0% (0%, 55.5%) |
| Sample size | 5093 OA, 5622 controls | 3934 Sx OA, 5622 controls | 1159 nonSx OA, 5622 controls | 3934 SxOA, 1159 nonSx OA |
Adjusted for age, sex and BMI.
Adjusted for age, sex, BMI and KL grade.
OR are shown for B, both genders; F, females only; M, males only; BMI, body mass index; KL, Kellgren–Lawrence; OA, osteoarthritis.
Despite biologic plausibility and demonstration of an association in other musculoskeletal pain conditions, we did not find an association between the 158COMTMet+ genotype and the knee OA phenotypes studied. The COMT variant may contribute differentially in knee and hip OA pain with the effect being smaller in knee OA. Additionally, while the original finding in hip OA was driven primarily by an association among women, we found no evidence of a stronger effect in women in our data. There are some study limitations: some of the included cohorts contributed only a small number of cases and the definition of asymptomatic OA used may further reduce the power to detect a genetic association. Other variants of this gene or a haplotype approach11 may provide additional insight for the importance of COMT in the pain experience of OA. Other factors that contribute to the pain experience (eg, catastro-phising, affect) may need to be accounted for.
As with the structural disease, the experience of pain in knee OA is multifactorial. Identification of genetic variants in large, well-phenotyped cohorts may provide much-needed rational therapeutic targets for the substantial unmet clinical need of pain management in OA.
Acknowledgments
Funding Supported by EC-FP7 programme grant 200 800 TREAT-OA and by a EULAR project grant to AMV (grant 108239). Dr. Neogi was supported by NIH AR055127. AstraZeneca UK funded the GOAL study sample and data collection. The TwinsUK cohort is supported by the Wellcome Trust. The HCS is supported by the Medical Research Council (UK) and the Oxford NIHR Musculoskeletal Biomedical Research Unit. The Framingham Osteoarthritis Study was funded by NIH AR47785 and AG18393 and the NHLBI N01-HC-25195. The Health, Aging and Body Composition Study was supported by NIA contracts N01AG62101, N01AG62103 and N01AG62106. The genome-wide association study was funded by NIA grant 1R01AG032098-01A1 to Wake Forest University Health Sciences and genotyping services were provided by the Centre for Inherited Disease Research (CIDR). CIDR is fully funded through a federal contract from the National Institutes of Health to The Johns Hopkins University, contract number HHSN268200782096C. The Tasmanian Older Adult Cohort collection was funded by the National Health and Medical Research Council of Australia, Arthritis Foundation of Australia; Tasmanian Community Fund and a University of Tasmania Grant-Institutional Research Scheme.
Footnotes
Contributors All authors contributed to the study design, data interpretation and the final manuscript. In addition, AMV and TN analysed and interpreted the data and prepared manuscript. AMV and TN supervised the study.
Competing interests None.
Ethics approval The appropriate ethics committee approved the study protocol of each of the contributing cohort studies included.
Provenance and peer review Not commissioned; externally peer reviewed.
REFERENCES
- 1.Cobb S, Merchant WR, Rubin T. The relation of symptoms to osteoarthritis. J Chronic Dis. 1957;5:197–204. doi: 10.1016/0021-9681(57)90135-2. [DOI] [PubMed] [Google Scholar]
- 2.Hannan MT, Felson DT, Pincus T. Analysis of the discordance between radiographic changes and knee pain in osteoarthritis of the knee. J. Rheumatol. 2000;27:1513–17. [PubMed] [Google Scholar]
- 3.Dieppe PA, Cushnaghan J, Shepstone L. The Bristol `OA500' study: progression of osteoarthritis (OA) over 3 years and the relationship between clinical and radiographic changes at the knee joint. Osteoarthritis Cartilage. 1997;5:87–97. doi: 10.1016/s1063-4584(97)80002-7. [DOI] [PubMed] [Google Scholar]
- 4.Hochberg MC, Lawrence RC, Everett DF, et al. Epidemiologic associations of pain in osteoarthritis of the knee: data from the National Health and Nutrition Examination Survey and the National Health and Nutrition Examination-I Epidemiologic Follow-up Survey. Semin. Arthritis Rheum. 1989;18(4 Suppl 2):4–9. doi: 10.1016/0049-0172(89)90008-5. [DOI] [PubMed] [Google Scholar]
- 5.Lawrence JS, Bremner JM, Bier F. Osteo-arthrosis. Prevalence in the population and relationship between symptoms and x-ray changes. Ann Rheum Dis. 1966;25:1–24. [PMC free article] [PubMed] [Google Scholar]
- 6.Malfait AM, Seymour AB, Gao F, et al. A role for PACE4 in osteoarthritis pain: evidence from human genetic association and null mutant phenotype. Ann Rheum Dis. 2012;71:1042–8. doi: 10.1136/annrheumdis-2011-200300. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Reimann F, Cox JJ, Belfer I, et al. Pain perception is altered by a nucleotide polymorphism in SCN9A. Proc Natl Acad Sci USA. 2010;107:5148–53. doi: 10.1073/pnas.0913181107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Valdes AM, De Wilde G, Doherty SA, et al. The Ile585Val TRPV1 variant is involved in risk of painful knee osteoarthritis. Ann Rheum Dis. 2011;70:1556–61. doi: 10.1136/ard.2010.148122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.van Meurs JB, Uitterlinden AG, Stolk L, et al. A functional polymorphism in the catechol-O-methyltransferase gene is associated with osteoarthritis-related pain. Arthritis Rheum. 2009;60:628–9. doi: 10.1002/art.24175. [DOI] [PubMed] [Google Scholar]
- 10.Nackley AG, Shabalina SA, Tchivileva IE, et al. Human catechol-O-methyltransferase haplotypes modulate protein expression by altering mRNA secondary structure. Science. 2006;314:1930–3. doi: 10.1126/science.1131262. [DOI] [PubMed] [Google Scholar]
- 11.Diatchenko L, Slade GD, Nackley AG, et al. Genetic basis for individual variations in pain perception and the development of a chronic pain condition. Hum Mol Genet. 2005;14:135–43. doi: 10.1093/hmg/ddi013. [DOI] [PubMed] [Google Scholar]
- 12.Tammimaki A, Mannisto PT. Catechol-O-methyltransferase gene polymorphism and chronic human pain: a systematic review and meta-analysis. Pharmacogenet Genomics. 2012;22:673–91. doi: 10.1097/FPC.0b013e3283560c46. [DOI] [PubMed] [Google Scholar]
- 13.Conroy MB, Kwoh CK, Krishnan E, et al. Muscle strength, mass, and quality in older men and women with knee osteoarthritis. Arthritis Care Res (Hoboken) 2012;64:15–21. doi: 10.1002/acr.20588. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Felson DT, Zhang Y, Hannan MT, et al. The incidence and natural history of knee osteoarthritis in the elderly. The Framingham Osteoarthritis Study. Arthritis Rheum. 1995;38:1500–5. doi: 10.1002/art.1780381017. [DOI] [PubMed] [Google Scholar]