Type I collagen ╨Ю┬▒1 Sp1 transcription factor binding site polymorphism is associated with reduced risk of hip osteoarthritis defined by severe joint space narrowing in elderly women.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 52, No. 5, May 2005, pp 1431–1436 DOI 10.1002/art.21011 © 2005, American College of Rheumatology Type I Collagen ␣1 Sp1 Transcription Factor Binding Site Polymorphism Is Associated With Reduced Risk of Hip Osteoarthritis Defined by Severe Joint Space Narrowing in Elderly Women K. Lian,1 J. M. Zmuda,2 M. C. Nevitt,1 L. Lui,1 M. C. Hochberg,3 D. Greene,4 J. Li,4 J. Wang,4 and N. E. Lane1 severe JSN (score >3), and 131 (23%) had moderate or moderate-to-severe femoral osteophytosis (score >2). There was no association of the T/T genotype with either radiographic hip OA or radiographic hip OA characterized by osteophytosis. For radiographic OA of the hip characterized by moderate-to-severe JSN, the odds of disease were significantly reduced among subjects with the T/T compared with the G/G genotype (OR 0.30, 95% CI 0.11–0.81, P ⴝ 0.02) and did not change after adjustment for potential confounders (OR 0.36, 95% CI 0.13–0.99, P ⴝ 0.048). Conclusion. The T/T genotype of the COL1A1 Sp1 polymorphism was associated with a reduced risk of radiographic OA of the hip characterized by JSN. This association should be confirmed in other populations to determine if mechanistic studies are warranted. Objective. A common G/T substitution at an Sp1 binding site in intron 1 of the COL1A1 gene has been reported to be associated with reduced bone mineral density and increased risk of osteoporotic fracture. The purpose of this study was to examine whether there is an association between COL1A1 Sp1 polymorphism and radiographic osteoarthritis (OA) of the hip in elderly women in the Study of Osteoporotic Fractures. Methods. Radiographic hip OA status of subjects was defined by the presence of 1 of the following criteria in either hip: a joint space narrowing (JSN) score of >3, a Croft summary grade of >3, or both definite (score >2) osteophytes and JSN in the same hip. Cases of radiographic OA of the hip were further subdivided into those with JSN score >3 and those with a femoral osteophyte score >2 and JSN score <2. The COL1A1 Sp1 polymorphism was genotyped using allele-specific kinetic polymerase chain reaction in 4,746 women. Multivariate logistic regression was performed to estimate odds ratios (ORs) and 95% confidence intervals (95% CIs). Results. Radiographic OA of the hip was present in 571 women (12%). Of these patients, 325 (57%) had The prevalence of symptomatic osteoarthritis (OA) of the hip is ⬃3% among adults 30 years and older in the US, and accounts for a significant health care burden (1). Multiple risk factors for OA of the hip, including age, obesity, joint injury, occupational factors, sex, and ethnicity, among others, have been well established (1). Furthermore, it is also known that genetic factors play a strong role in the development of hip OA. Several twin and family studies (2,3) have demonstrated a significantly increased risk for OA of the hip among siblings of affected individuals. In addition, genomewide scans, fine-scale mapping, and other association analyses have identified several loci that may be associated with OA of the hip (2,3). Recently, type I collagen ␣1 (COL1A1) has emerged as a candidate gene of interest in OA (3). Type I collagen is a heterotrimer consisting of 2 ␣1 chains and 1 ␣2 chain, and is the major protein in bone (4). The Supported by the NIH (grants 1R01-AG-05407 and 1R01AR-40431), the Doris Duke Clinical Research Fellowship Program for Medical Students (grant 20000684), and the Rosalind Russell Arthritis Research Foundation. 1 K. Lian, MS, M. C. Nevitt, PhD, L. Lui, MA, MS, N. E. Lane, MD: University of California at San Francisco; 2J. M. Zmuda, PhD: University of Pittsburgh, Pittsburgh, Pennsylvania; 3M. C. Hochberg, MD: University of Maryland, Baltimore; 4D. Greene, MD, J. Li, PhD, J. Wang, MS: Roche Molecular Systems, Alameda, California. Address correspondence and reprint requests to N. E. Lane, MD, Division of Rheumatology, Department of Medicine, University of California at Davis and San Francisco, PO Box 589, Burlingame, CA 94011. E-mail: email@example.com. Submitted for publication August 24, 2004; accepted in revised form January 25, 2005. 1431 1432 LIAN ET AL gene encoding COL1A1 is an important candidate for the regulation of bone metabolism and bone mass, and mutations in this gene are known to cause osteogenesis imperfecta (4). Grant and colleagues have identified a G-to-T polymorphism in a transcription factor binding site for Sp1 in the first intron of the COL1A1 gene (5). Several studies have examined the relationship of this polymorphism to bone mineral density (BMD), osteoporosis, and fracture (5–8). In addition, higher adjusted BMD increases the risk of OA of the hip (9–11). Studies of the association of the COL1A1 Sp1 polymorphism and OA have produced conflicting results, with one showing a positive association with OA of the knee and hip (12) and two showing a lack of an association with OA of the spine and hip (3,13). However, the studies may have been underpowered to detect significant associations. Since the material properties of the subchondral bone are believed to be associated with both the development and the progression of OA (14), and the COL1A1 Sp1 polymorphism increases the risk of osteoporosis due to changes in both bone composition and bone strength (7), we hypothesized that subjects with the COL1A1 Sp1 polymorphism might have a different risk for OA of the hip. To test this hypothesis, we examined the association between the COL1A1 Sp1 binding site polymorphism and radiographic OA of the hip in a large cohort of older postmenopausal women, in an effort to further elucidate the role of COL1A1 in the pathogenesis of OA of the hip. PATIENTS AND METHODS Study population. All subjects were participants in the Study of Osteoporotic Fractures (SOF), a multicenter cohort study initiated in 1986 to determine risk factors for osteoporotic fractures in elderly women (15). Participants were all age 65 or older at baseline and were recruited from populationbased listings at 4 clinical centers in the US: Baltimore, MD, Minneapolis, MN, Monongahela Valley, PA (near Pittsburgh), and Portland, OR. Exclusion criteria for the parent study (the SOF) included bilateral hip replacement and an inability to walk unassisted; African American women were excluded because of their low risk of hip fracture. The study was approved by the institutional review board at each of the institutions involved, and all subjects provided written informed consent at enrollment and at each clinical examination. Buffy coat specimens were collected from a total of 6,975 participants at either visit 2 (1989–1990) or visit 6 (1997–1998), and genotyping was performed in all women who provided adequate consent for genetic studies (Figure 1). Bilateral anteroposterior pelvic radiographs were obtained at baseline and at visit 5 (mean 8.3 years of followup) (16). The present analysis used only the visit-5 radiograph to evaluate radiographic OA of the hip for this study. Figure 1. Flow chart of the study subjects. SOF ⫽ Study of Osteoporotic Fractures; OA ⫽ osteoarthritis; JSN ⫽ joint space narrowing. COL1A1 genotyping. Genotyping was performed with Axys Sequana on DNA from 3,501 subjects using a TaqMan assay with a standardized protocol, performed on an ABI 7700 (Perkin Elmer, Emeryville, CA). The following primers and probes were utilized (17): forward primer AATCAGCCGCTCCCATTCTCCTA, reverse primer GGAGGGCGAGGGAGGAGAGAA, G2046 probe (FAM fluorophore) TCATCCCGCCCCCATTCCCTG, 2046T probe (TET fluorophore) TCATCCCGCCCACATTCCCTGG. An additional 3,247 samples were genotyped at Roche Molecular Systems for COL1A1G2046T by allele-specific kinetic-method polymerase chain reaction as described by Germer et al (18), with the modification of using a single sample for each reaction. The 2 allele-specific primers and 1 common primer used were as follows: allele-specific G2046 CTGCCCAGGGAATGG, allele-specific 2046T CCTGCCCAGGGAATGT, common primer AAGGGAGGTCCAGCCCTCAT. Amplification reactions were run in GeneAmp 5700 Sequence Detection Systems (SDS) instruments (PE Biosystems, Foster City, CA). The cycle at which the relative fluorescence reached a threshold of 0.5, using the SDS software from PE Biosystems, was defined as Ct. The Ct of each amplification reaction was determined, and the difference between the Ct for allele 1 and that for allele 2 (⌬Ct) was used as the assay result. Samples with ⌬Ct results between ⫺3.0 and 3.0 were considered heterozygous (G/T). Samples with ⌬Ct results below ⫺3.0 were considered homozygous for the major allele (G/G), and samples with ⌬Ct results above 3.0 were considered homozygous for the minor allele (T/T). Since genotyping was performed in 2 different settings HIP OA, COL1A1, AND GENETICS Table 1. 1433 Characteristics of the study subjects* Characteristic Radiographic hip OA cases (n ⫽ 571) Controls (n ⫽ 4,175) P 79.6 ⫾ 5.0 157.6 ⫾ 6.3 66.3 ⫾ 12.8 26.7 ⫾ 5.0 4.4 462.2 ⫾ 548.3 48.4 ⫾ 5.4 17.9 48.0 46.1 73.7 78.4 ⫾ 4.6 157.8 ⫾ 6.2 66.2 ⫾ 12.9 26.6 ⫾ 4.8 4.8 570.2 ⫾ 636.9 48.3 ⫾ 5.6 19.0 47.9 50.0 81.0 ⬍0.01 0.51 0.88 0.47 0.66 ⬍0.01 0.68 0.51 0.96 0.08 ⬍0.01 0.74 ⫾ 0.13 0.67 ⫾ 0.14 0.37 ⫾ 0.10 0.73 ⫾ 0.14 0.62 ⫾ 0.11 0.37 ⫾ 0.09 0.61 ⬍0.01 0.14 Age, years Height, cm Weight, kg BMI, kg/m2 Current smoker, % Physical activity (walking), kcal/week Age at menopause, years Current estrogen use, % Current vitamin D use, % Calcium supplement use, % Health status, good/excellent, % Areal BMD by DXA, gm/cm2 Total hip Femoral neck Calcaneal Genotype frequency, % G/G G/T T/T 66.8 30.0 3.2 65.3 30.9 3.8 * Except where indicated otherwise, values are the mean ⫾ SD. P values were determined by chi-square test for dichotomous variables and by Student’s t-test for continuous variables. OA ⫽ osteoarthritis; BMI ⫽ body mass index; BMD ⫽ bone mineral density; DXA ⫽ dual x-ray absorptiometry. with different protocols, 40 subjects were genotyped by both methods and sequenced, with 100% agreement. Radiographic assessments. Details of radiographic scoring methods for hip OA have been described previously (9,16). Subjects were considered to have radiographic hip OA if they satisfied at least 1 of 3 criteria: joint space narrowing (JSN) score ⱖ3, Croft summary grade ⱖ3 (ⱖ3 individual radiographic features), or both definite (score ⱖ2) osteophytes and JSN in the same hip. Controls (no radiographic hip OA) were subjects who had no findings of hip OA on the visit-5 radiograph. For the present study, subjects with a diagnosis of rheumatoid arthritis, Paget’s disease of bone, or bilateral hip fractures were excluded from this analysis (14). In this study, 32 subjects had unilateral total hip replacements (THRs). For these subjects, we used the contralateral hip (non-THR hip) to define radiographic hip OA from the visit-5 radiograph. In addition, we further characterized patients with radiographic hip OA into those with JSN score ⱖ3 (n ⫽ 325) and those with a femoral osteophyte score of ⱖ2 and a JSN score ⱕ2 (n ⫽ 131), to try to further assess radiographic OA of the hip by phenotype (Figure 1). Reproducibility was good, with kappa scores for interrater reliability of 0.66 for definite JSN (score ⱖ2), 0.71 for definite osteophytes, and 0.65 for summary grades ⱖ2. The intraclass correlation coefficient was 0.85 for the continuous measurement of minimal joint space (9,16). Covariate assessment. All study participants completed a self-administered questionnaire at the baseline and followup visits that included age, self-reported health status, current medication use (including multivitamins and other supplements), daily physical activity, and smoking history. Demographic data presented are from visit 5. Height was measured using a wall-mounted Harpenden stadiometer (Hol- tain, Dyfed, UK), and weight was measured with a balance beam scale; methods have been previously described (9,15,16). Protocols for BMD measurements of the hip were performed using dual x-ray absorptiometry (QDR 1000; Hologic, Waltham, MA), and calcaneus was measured using single x-ray absorptiometry (OsteoAnalyzers; Siemens-Osteon, Wahiawa, HI), as previously described (9). Statistical analysis. Differences in demographic and BMD variables between radiographic hip OA cases and controls without radiographic hip OA were assessed by Student’s t-test for continuous variables and by chi-square test for dichotomous variables. Hardy-Weinberg equilibrium was assessed in both groups by chi-square test, and allele frequencies were calculated using the gene counting method. Logistic regression was used to estimate odds ratios (ORs) and 95% confidence intervals (95% CIs) for each genotype (G/G, G/T, and T/T) and the risk of any radiographic hip OA, radiographic hip OA defined by JSN score ⱖ3, or radiographic hip OA defined by femoral osteophytosis of score ⱖ2. All analyses for the association of COL1A1 Sp1 genotypes and radiographic hip OA categories were performed both with and without adjustments for known and potential confounding variables including age, height, weight, estrogen use, and femoral neck BMD. Statistical analysis was performed using the statistical software program SAS, version 8.2 (SAS Institute, Cary, NC). RESULTS Characteristics of the study subjects and genotype frequencies. Among the 4,746 study subjects, 571 satisfied the criteria for radiographic hip OA, of whom 1434 LIAN ET AL Table 2. Association between COL1A1 and radiographic hip OA status* Genotype Cases/total (%) Unadjusted OR (95% CI)† Adjusted OR (95% CI)‡ Radiographic hip OA G/G G/T T/T Osteophytosis G/G G/T T/T Joint space narrowing G/G G/T T/T 571/4,746 (12.0) 382/3,108 (12.3) 171/1,462 (11.7) 18/176 (10.2) 131/4,746 (2.8) 79/3,108 (2.5) 46/1,462 (3.2) 6/176 (3.4) 325/4,746 (6.8) 224/3,108 (7.2) 97/1,462 (6.6) 4/176 (2.3) – 1.00 (referent) 0.95 (0.78–1.15) 0.81 (0.49–1.34) – 1.00 (referent) 1.23 (0.85–1.80) 1.31 (0.56–3.05) – 1.00 (referent) 0.92 (0.72–1.17) 0.30 (0.11–0.81) – 1.00 (referent) 0.90 (0.79–1.19) 0.83 (0.49–1.40) – 1.00 (referent) 1.32 (0.91–1.93) 1.22 (0.48–3.11) – 1.00 (referent) 0.95 (0.73–1.23) 0.36 (0.13–0.99) * OA ⫽ osteoarthritis; OR ⫽ odds ratio; 95% CI ⫽ 95% confidence interval. † Unadjusted logistic regression analysis. ‡ Adjusted for age, height, weight, estrogen use, and femoral neck bone mineral density. 325 subjects had JSN score ⱖ3 and 131 had moderateto-severe osteophytes (Figure 1). Compared with the 4,175 control subjects who did not satisfy the radiographic hip OA criteria, the patients identified as having radiographic OA of the hip were older, reported less walking activity and poorer overall health status, and had higher femoral neck BMD (Table 1). No other demographic or BMD variables were significantly different between the 2 groups. The frequency of the T allele in the cohort was 19.1% and was similar to frequencies observed in similar populations (6,12). Genotype frequencies did not deviate from Hardy-Weinberg equilibrium. Odds ratios for radiographic OA of the hip. There was no significant association between the T/T genotypes and radiographic hip OA case status in unadjusted models or in models adjusted for age, height, weight, estrogen use, and femoral neck BMD (Table 2). Similarly, there was no association between either of the genotypes and the presence of radiographic OA of the hip as characterized by femoral osteophyte score ⱖ2, with or without similar adjustment for covariates. However, there was a 70% reduction in the odds of radiographic hip OA characterized by severe JSN in subjects with the T/T genotype, compared with the G/G genotype (OR 0.30, 95% CI 0.11–0.81, P ⫽ 0.02). This reduction in radiographic hip OA risk remained significant after adjustment for age, height, weight, estrogen use, and femoral neck BMD (P ⫽ 0.048) (Table 2). DISCUSSION The present study addressed the association between the COL1A1 Sp1 binding site polymorphism and the risk of radiographic OA of the hip. While there was no association between the Sp1 polymorphism and radiographic hip OA defined by any of the 3 different criteria or by osteophytosis as the predominant radiographic feature, there was a significant reduction in the odds of radiographic hip OA as characterized by severe JSN among subjects with the T/T genotype. Because joint space narrowing observed by hip radiography is a surrogate measure of articular cartilage loss, we speculate that this effect may reflect the role of the COL1A1 genotype on articular cartilage or subchondral bone metabolism in the development of hip OA. The relationship between the COL1A1 Sp1 binding site polymorphism and osteoporosis, as well as fracture risk, has been well established (5–8). Moreover, haplotype analysis has demonstrated that the relationship with fracture risk is specific for the Sp1 site, rather than other polymorphisms (8). The Sp1 transcription factor binding site is located in the first intron of the COL1A1 gene, a region that has been identified as important for the regulation of collagen transcription (7). In a series of experiments, Mann et al demonstrated that the G/T genotype was associated with altered production of the collagen protein, with an increased ratio of ␣1(I) to ␣2(I) protein chains (7). This result was accompanied by an increase in expression of COL1A1 messenger RNA (mRNA) relative to COL1A2 mRNA and reduced biomechanic strength among those with the G/T genotype versus those with the G/G genotype. In the current study, we found that women with 2 copies of the variant T allele had decreased hip OA risk, but this will have to be evaluated in another study population. At present, it is unclear how altered collagen HIP OA, COL1A1, AND GENETICS production and reduced bone strength as a result of the presence of the T allele contributes to the development of hip OA characterized by severe joint space narrowing. These are recognized subsets of radiographic hip OA, and we don’t know about differences in pathogenesis. Higher adjusted levels of BMD are associated with increased odds of radiographic hip OA with osteophytes (9–11) but not radiographic hip OA characterized by JSN alone. Since the T/T genotype is associated with lower BMD and increased risk of fractures, we speculate that it is also associated with radiographic OA of the hip characterized by severe JSN. This association was present, however, even after adjustment for BMD. This suggests that bone strength may be important. While Radin et al found that changes in the subchondral bone strength preceded articular cartilage loss in a rabbit model of OA (13), Burr has shown that increased bone turnover and decreased subchondral bone strength are both important in the development of OA with cartilage loss (19). Since the pathogenesis of hip OA may differ in these 2 subtypes of OA, it is possible that altered collagen in subjects with this COL1A1 polymorphism may result in weaker bone that slows the development of OA. Furthermore, this latter effect appears to be mediated by increased joint space narrowing. Additional work is now required to determine how altered bone composition adjacent to the hip joint may reduce the risk of hip OA. Two previous studies have investigated the role of COL1A1 in OA. Aerssens et al found no association of the COL1A1 Sp1 polymorphism with hip OA in a case–control study comparing postmenopausal women with THR for OA of the hip and those with no hip OA (14). Given the small sample of only 75 cases, it is possible that the lack of association was due to insufficient power to detect a difference between the groups. The study by Aerssens and colleagues focused on elderly women with THR, while our study focused on women with radiographic OA of the hip and excluded subjects with hip replacement. These differences may also explain our divergent results. In a separate proband– spouse case–control study, Loughlin et al found a weak association of the COL1A1 Sp1 polymorphism with hip or knee OA (12). Compared with female controls, elderly women with OA requiring hip or knee replacement had a significantly different genotype distribution, with a relative increase in the G allele (OR 1.58, 95% CI 1.09–2.29). However, this effect did not remain significant after stratifying by hip or knee replacement (11). It is possible that these results are consistent with ours, in that the G allele may be associated with the develop- 1435 ment of hip OA, while the T allele may be protective against hip OA. Also, Loughlin et al defined cases by joint replacement, so their results may not be comparable with ours. This study offers several strengths, including its large community-based cohort of elderly white women and a validated radiographic scoring system making phenotype-specific definitions to fully characterize the extent of radiographic hip OA. However, there are also several potential limitations. Our definitions of radiographic OA of the hip may not be comparable with other grading schemes. Also, despite the large number of subjects in our cohort, we only had 571 subjects with radiographic hip OA, and the numbers of participants were smaller when we subdivided by either genotype or individual radiographic features. Another potential limiting factor is that our study participants were all white women age 65 years and older, and our findings may not be generalizable to other populations. In summary, we found that the T/T genotype at a binding site for the Sp1 transcription factor in COL1A1 was associated with reduced risk of radiographic OA of the hip as defined by severe joint space narrowing. While our findings require confirmation in other populations, they suggest that the COL1A1 Sp1 genotype may be associated with a reduction in the development of hip OA, and that this effect may be mediated by changes in bone quantity and/or quality. In addition, our results highlight the importance of phenotypic definition in the genetic analyses of OA susceptibility. REFERENCES 1. Felson DT, Lawrence RC, Dieppe PA, Hirsch R, Helmick CG, Jordan JM, et al. Osteoarthritis: new insights. I. The disease and its risk factors [review]. Ann Intern Med 2000;133:635–46. 2. Spector TD, MacGregor AJ. Risk factors for osteoarthritis: genetics. Osteoarthritis Cartilage 2004;12 Suppl A:S39–44. 3. Loughlin J. Genetic epidemiology of primary osteoarthritis. Curr Opin Rheumatol 2001;13:111–6. 4. Kuivaniemi H, Tromp G, Prockop DJ. Mutations in collagen genes: causes of rare and some common diseases in humans. FASEB J 1991;5:2052–60. 5. Grant SF, Reid DM, Blake G, Herd R, Fogelman I, Ralston SH. Reduced bone density and osteoporosis associated with a polymorphic Sp1 binding site in the collagen type I ␣ 1 gene. Nat Genet 1996;14:203–5. 6. Uitterlinden AG, Burger H, Huang Q, Yue F, McGuigan FE, Grant SF, et al. Relation of alleles of the collagen type I␣1 gene to bone density and the risk of osteoporotic fractures in postmenopausal women. N Engl J Med 1998;338:1016–21. 7. Mann V, Hobson EE, Li B, Stewart TL, Grant SF, Robins SP, et al. A COL1A1 Sp1 binding site polymorphism predisposes to osteoporotic fracture by affecting bone density and quality. J Clin Invest 2001;107:889–907. 8. McGuigan FA, Reid DM, Ralston SH. Susceptibility to osteoporotic fracture is determined by allelic variation at the Sp1 site, 1436 9. 10. 11. 12. 13. rather than other polymorphic sites at the COL1A1 locus. Osteoporos Int 2000;11:338–43. Nevitt MC, Lane NE, Scott JC, Hochberg MC, Pressman AR, Genant AK, et al, and the Study of Osteoporotic Fractures Research Group. Radiographic osteoarthritis of the hip and bone mineral density. Arthritis Rheum 1995;38:907–16. Hart DJ, Cronin C, Daniels M, Worthy T, Doyle DV, Spector TD. The relationship of bone density and fracture to incident and progressive radiographic osteoarthritis of the knee: the Chingford Study. Arthritis Rheum 2002;46:92–9. Sowers M, Lachance L, Jamadar D, Hochberg MC, Hollis B, Crutchfield M, et al. The associations of bone mineral density and bone turnover markers with osteoarthritis of the hand and knee in pre- and perimenopausal women. Arthritis Rheum 1999;42:483–9. Loughlin J, Sinsheimer JS, Mustafa Z, Carr AJ, Clipsham K, Bloomfield VA, et al. Association analysis of the vitamin D receptor gene, the type I collagen gene COL1A1, and the estrogen receptor gene in idiopathic osteoarthritis. J Rheumatol 2000;27: 779–84. Radin EL, Paul IL, Rose RM. Role of mechanical factors in the pathogenesis of OA. Lancet 1972;1:519–22. LIAN ET AL 14. Aerssens J, Dequeker J, Peeters J, Breemans S, Boonen S. Lack of association between osteoarthritis of the hip and gene polymorphisms of VDR, COL1A1, and COL2A1 in postmenopausal women. Arthritis Rheum 1998;11:1946–50. 15. Cummings SR, Nevitt MC, Browner WS, Stone K, Fox KM, Ensrud KE, et al, and the Study of Osteoporotic Fractures Research Group. Risk factors for hip fracture in white women. N Engl J Med 1995;332:767–73. 16. Lane NE, Nevitt MC, Hochberg MC, Hung YY, Palermo L. Progression of radiographic hip osteoarthritis over eight years in a community sample of elderly white women. Arthritis Rheum 2004;50:1477–86. 17. Morin PA, Saiz R, Monjazeb A. High-throughput single nucleotide polymorphism genotyping by fluorescent 5⬘ exonuclease assay. Biotechniques 1999;27:538–52. 18. Germer S, Holland MJ, Higuchi R. High throughput SNP allelefrequency determination in pooled DNA samples by kinetic PCR. Genome Res 2000;10:258–66. 19. Burr JB. The importance of subchondral bone in pression of OA. J Rheumatol 2004;31 Suppl 70:77–82.