Original Article

*Correspondence to Margreet Kloppenburg, Leiden University Medical Centre, Department of Rheumatology, PO Box 9600, 2300 RC Leiden, The Netherlands

Funded by:
  Pfizer Inc.

Received: 9 March 2006; Accepted: 1 September 2006

10.1002/art.22680  About DOI

INTRODUCTION

Osteoarthritis (OA) is by far the most common joint disorder and is a leading cause of pain and disability, resulting in substantial medical and social costs. During the last decade the importance of genetic factors in the development of OA has been revealed ([1][2]). However, close to nothing is known about the contribution of genetic factors to OA disease progression, which from a clinical perspective is more relevant than just the presence of OA. Until now, no study has investigated whether there is evidence for transmission of the progression trait in families in order to prove hereditary influences on the progression of OA.

At present, there is no disease-modifying treatment available for OA. The identification of genetic loci that associate with OA progression could help to find genes that are responsible for driving OA progression. This would increase our knowledge on the pathophysiologic pathways involved in OA progression, which could provide potential targets for OA therapy. To gain more insight into whether familial factors play a role in disease progression of OA, we analyzed familial aggregation of radiologic progression of OA over 2 years in sibling pairs with OA at multiple joint sites.

PATIENTS AND METHODS

Study design and selection of patients.

The Genetics, Arthrosis, and Progression (GARP) study is a prospective longitudinal cohort study aimed at identifying determinants of OA susceptibility and progression. The study population consists of white sibling pairs of Dutch ancestry with OA at multiple sites. The GARP study was approved by the Medical Ethics Committee of the Leiden University Medical Center.

Details of recruitment and selection of patients have been described in detail elsewhere ([3]). In brief, probands ages 40-70 years were recruited from rheumatologists, orthopedic surgeons, and general practitioners. Subsequently, affected siblings were recruited via the probands. Probands and siblings were included at baseline if they had OA at multiple sites in the hands or in  2 of the following joint sites: hand, spine (cervical or lumbar), knee, or hip. If there was more than 1 affected sibling, the youngest sibling with OA at multiple joint sites was included. Both groups of patients were required to have symptomatic OA (as defined below) in at least 1 joint site. Patients with symptomatic OA in only 1 joint site were required to have symptoms and structural abnormalities, defined by the presence of radiographic OA, in at least 1 of the other 3 joint sites. Sibling pairs in which at least 1 patient had symptomatic hip or knee OA (but not in a radiologic end stage) were eligible for the 2-year followup study. Informed written consent was obtained from all patients.

Patients with secondary OA, familial syndromes with a clear Mendelian inheritance pattern, or a shortened life expectancy were excluded. OA developing under the following conditions was considered secondary: major congenital or developmental diseases and bone dysplasias, major local factors such as severe scoliosis and hypermobility, metabolic diseases associated with joint disease such as hemochromatosis and Wilson's disease, other bone diseases such as Paget's disease and osteochondritis, and intraarticular fractures. Crystal deposition arthropathies (unless in the case of severe polyarticular gout) and diabetes mellitus or thyroid conditions were not considered as exclusion criteria.

Diagnosis of OA.

Symptomatic OA in the knee and hip was defined according to the American College of Rheumatology (ACR) recommendations for knee and hip OA ([4][5]). Knee OA was defined as pain or stiffness on most days of the preceding month and osteophytes at joint margins of the tibiofemoral joint. Hip OA was defined as pain or stiffness in the groin and hip region on most days of the preceding month in addition to femoral or acetabular osteophytes or joint space narrowing (JSN) on radiograph. Prosthetic joints in hips or knees as a result of end-stage OA were defined as OA in that particular joint.

Degeneration of the spine (cervical and lumbar) was defined as pain or stiffness in the spine on most days of the preceding month in addition to a Kellgren/Lawrence score of 2 in at least 1 disc or 1 apophyseal joint. OA in hand joints was defined, according to the ACR criteria ([6]), as pain or stiffness on most days of the preceding month in addition to 3 of the following 4 criteria: bony swelling of  2 of the 10 selected joints (bilateral distal interphalangeal [DIP] joints 2 and 3, bilateral proximal interphalangeal [PIP] joints 2 and 3, and carpometacarpal [CMC] joint 1), bony swelling of  2 DIP joints, <3 swollen metacarpophalangeal (MCP) joints, and deformity of at least 1 of the 10 selected joints.

Radiographs.

Standardized radiographs of the hands (anteroposterior), knees (posteroanterior weight bearing), and hips (posteroanterior weight bearing) were obtained in a single center at baseline and after 2 years. The standardized nonfluoroscopic fixed-flexion protocol was used to obtain radiographs of the knees ([7]). Uniform anatomic alignment of the knees was facilitated using a specifically designed calibration and positioning frame (Synaflexer; Synarc, San Francisco, CA) ([8]). An experienced radiographer obtained all radiographs using a standard protocol with a fixed film focus distance.

Assessment of radiologic progression of OA.

Radiographs were graded on a scale of 0-3 for JSN and osteophytes. The following joints were scored with the help of the Osteoarthritis Research Society International Atlas ([9]): the 8 DIP joints, the 8 PIP joints, the 10 MCP joints, the 2 interphalangeal (IP) joints of the thumbs, the right and left first CMC joints, the right and left trapezioscaphoid (TS) joints, the right and left medial and lateral compartments of the tibiofemoral joints, and the right and left hip joints.

All radiographs were scored by consensus opinion of 2 experienced readers (SAB-S and IW). In case of disagreement, the lower, more conservative score was recorded. Films were blinded for patient characteristics, and radiographs of the same patient from both time points were assessed together without knowledge of the chronological order. The reproducibility for JSN and osteophyte grading, depicted by the intraclass correlation coefficient (ICC), was very good: 0.93 and 0.96 for JSN and osteophyte total scores, respectively. ICCs for JSN and osteophytes were 0.92 and 0.98 in the hands, 0.88 and 0.87 in the knees, and 1.00 and 1.00 in the hips, respectively. The reproducibility was based on the repeat readings of a random sample of 20 radiographs, which were selected throughout the period of observation.

Definition of radiologic progression.

Radiologic progression of JSN and osteophytes was evaluated on an individual basis, including all hand, hip, and knee joints. In subanalyses, we also evaluated progression at the specific joint site level. The JSN and osteophyte scores of the following groups of joints were analyzed in combination: hands (DIPs, PIPs, MCPs, IPs, CMC1s, TSs), knees (medial and lateral compartments of the tibiofemoral joints), and hips. Total scores for JSN and osteophytes in the hands, knees, and hips were calculated by adding the scores of left and right sides. The maximum total JSN score for 1 patient was 114: 96 in the hands, 6 in the hips, and 12 in the knees. The maximum total osteophyte score for 1 patient was 126: 96 in the hands, 6 in the hips, and 24 in the knees. Radiologic outcome was the change in JSN and osteophyte scores after 2 years. Radiologic progression at the patient level was defined in 2 different ways: 1) by at least a 1-point score increase in JSN total score or osteophyte total score over 2 years, and 2) by at least a 1-point score increase in JSN score or osteophyte score in at least 1 joint site, including hands, hips, or knees. Radiologic progression at the joint site level was defined by at least a 1-point score increase in JSN score or osteophyte score over 2 years in a specific joint site.

Patients without radiologic end-stage disease in the knees and hips (grade 3) in terms of JSN or osteophytes at baseline who received hip or knee prostheses during the followup period were considered to have a 1-point score increase in JSN or osteophytes in that particular joint. In the subanalyses at the joint site level, patients with bilateral hip prostheses at baseline who were unable to progress further at the hip joint level were considered to have progressive disease in terms of JSN and osteophytes in their hip joints.

Statistical analyses.

All analyses were based on the sibling pairs in which both the proband and the sibling had complete radiographic followup after 2 years. For all patients, radiologic progression in JSN or osteophytes at the individual patient level and joint site level (hands, hips, and knees) was dichotomized according to its presence or absence. The possible effect of familial factors on radiologic disease progression was estimated by calculating odds ratios (ORs) comparing siblings of probands who progressed with siblings of probands who did not progress. ORs were calculated for radiologic progression in the siblings, given progression in the probands, using both abovementioned definitions of progression. If familial factors had no influence on progression of OA, it would occur randomly in probands and siblings and no association (OR   1) would be found. An OR >1 indicated that siblings and probands both had radiologic progression more often than expected from the overall distribution of radiologic progression of OA. ORs are presented with 95% confidence intervals (95% CIs). Additionally, we estimated the dose-response relationship between progression in the siblings and the amount of progression among the probands per increase in proband total scores (an increase of  0, 1, or  2 in total scores) or per number of joint sites with progression (0, 1, or  2 joint sites). Multivariate logistic regression analyses were used to adjust for age, sex, and body mass index (BMI). All analyses were conducted using SPSS statistical software, version 11.0 (SPSS, Chicago, IL).

RESULTS

Patients.

Between August 2000 and March 2003, 191 probands and 202 siblings were included in the GARP study. Of these patients, 105 sibling pairs were eligible for the 2-year followup study. Due to a concurrent magnetic resonance imaging (MRI) study of the knee, one sibling pair was excluded due to a contraindication for MRI investigation of the knee. This resulted in the inclusion of 104 sibling pairs, of which 100 probands (96.2%) and 93 siblings (89.4%) returned for followup radiographic examinations after 2 years. In 92 sibling pairs, both the proband and the sibling had complete radiographic followup.

Baseline characteristics of these patients are presented in Table 1. The majority of the probands (84%) and the siblings (75%) were women. Age and BMI were similar in probands and siblings. Seventy-three percent of the probands and 67% of the siblings had symptomatic hand OA. Symptomatic hip and knee OA was present in 33% and 46% of the probands and in 31% and 43% of the siblings, respectively. The patients with incomplete followup did not differ significantly from the patients with complete followup (data not shown).

Radiologic score changes.

The distribution of the changes in total scores in the 92 sibling pairs in which both the proband and the sibling had complete radiographic followup after 2 years is shown in Table 2. An increase of at least 1 point in JSN total score was present in 43 (47%) probands and 31 (34%) siblings, and 39 (42%) probands and 34 (37%) siblings had an increase of at least 1 point in osteophyte total score.

Increases in JSN scores in probands and siblings occurred most often in the hands and less often in the knees and hips. Similar results were found for increases in osteophyte scores, occurring most often in the hands of probands and siblings and less often in the knees and hips. During the 2-year followup period, 4 probands received knee prostheses and 3 probands and 5 siblings received hip prostheses. Of the 92 followed sibling pairs, 7 probands and 3 siblings had bilateral hip prostheses at baseline and were considered to have OA progression in terms of JSN and osteophytes in their hip joints in the subanalyses at the hip joint level.

Familial aggregation of progression at the patient level according to 2 different definitions.

Progression defined by at least a 1-point score increase in total scores.

The OR (95% CI), adjusted for age, sex, and BMI, for a sibling having at least a 1-point score increase in JSN total score if the proband had at least a 1-point score increase in JSN total score was 3.0 (1.2-7.8) (Table 3). The adjusted OR (95% CI) for a sibling having at least a 1-point score increase in osteophyte total score if the proband had at least a 1-point score increase in osteophyte total score was 1.5 (0.6-3.6) (Table 3).

Progression in at least 1 joint site.

The adjusted OR (95% CI) for a sibling having progression of JSN in at least 1 joint site (hands, hips, or knees) if the proband had progression of JSN in at least 1 joint site was 2.8 (1.1-7.1) (Table 4). The adjusted OR (95% CI) for a sibling having radiologic progression of osteophytes in at least 1 joint site if the proband had progression of osteophytes in at least 1 joint site was 1.3 (0.6-3.3) (Table 4).

Dose-response relationship.

A dose-response relationship was found between the amount of increase in JSN total score among the probands and progression in the siblings (Table 5). A crude dose-response relationship was also found between the number of joint sites with progression among the probands and progression of JSN in at least 1 joint in the siblings (Table 5).

Familial aggregation of progression per joint site.

The adjusted ORs (95% CI) for a sibling having progression of JSN in the same joint group as the proband were 1.3 (0.4-4.0) for the hands, 2.2 (0.3-13.6) for the hips, and 4.3 (1.2-15.0) for the knees (Table 6). The adjusted ORs (95% CI) for a sibling having progression of osteophytes in the same joint group as the proband were 1.2 (0.4-3.8) for the hands, 3.6 (0.8-16.5) for the hips, and 2.2 (0.5-10.4) for the knees (Table 6).

DISCUSSION

This prospective cohort study of middle-aged sibling pairs with familial OA at multiple sites demonstrates that familial factors play a role in radiologic disease progression in OA in terms of JSN reflecting articular cartilage loss. A 3-fold increase in risk for radiologic progression of JSN was found in siblings of patients (probands) with progression of JSN. This association showed a dose-response effect. Familial aggregation tended to be more prominent for progression of OA in the knees and hips than in the hands. Interestingly, this effect was mainly expressed in progression of JSN reflecting cartilage degeneration rather than in progression of osteophytes.

A hereditary influence in the development of OA was already documented in the 1940s by Stecher in family studies of hand OA ([10]) and was subsequently confirmed for generalized OA ([11]) and OA in other specific joint groups ([1]). More recently, a number of genetic loci have been associated with the presence of such different OA phenotypes ([2]). Until now, no study has investigated the hereditary influences on the progression of OA. The first step consists of genetic-epidemiologic research, starting with the question of whether there is evidence for transmission of the progression trait in families. For this purpose, the strength of familial aggregation of progression can be assessed or the heritability of progression can be estimated. The present study fills this gap by providing the first evidence that familial factors are important not only in progression of OA at multiple sites, but also in progression of knee and hip OA. The next step would be to search for genetic loci associated with OA progression. A previous association study demonstrated that genetic variants of 4 genes (ADAM12, CILP, OPG, and TNA) correlated with progression of knee OA over 10 years ([12]). However, in this association study, performing a large number of tests (multiple testing) could have lead to false-positive findings, and replication of those findings in different populations is required ([13]). Identifying genetic factors that predispose to disease progression would give more insight into the pathophysiologic pathways involved in OA progression, which could help in developing treatment strategies to prevent or delay the progression of OA.

The familial aggregation of disease progression that was observed in the present study is likely to be explained by both genetic factors and shared environmental influences. This effect was minimized in the present study by including only 1 sibling per proband and by adjusting for BMI.

Assessing JSN in serial radiographs, as a surrogate for the thinning of articular cartilage, is currently the recommended outcome measure for longitudinal studies on progression of OA. We examined progression of JSN as well as osteophytes and found familial aggregation of progression of JSN, but this was not so clear for progression of osteophytes. This finding may suggest different genetic effects on progression of cartilage degeneration and bone proliferation. These findings are in line with those of previous animal studies on the mechanism of OA development suggesting that cartilage loss and bone proliferation, either in the subchondral area or at the margins of the bone, are independent processes ([14]). Further evidence comes from studies that demonstrated different genes to be associated with the radiographic prevalence of JSN and osteophytes ([15-17]). However, in the present study, probands had higher baseline JSN and osteophyte scores than siblings, suggesting that they were at a slightly later stage of OA. Therefore, because osteophyte formation in patients with OA is considered to be a secondary process following cartilage loss, this could also explain the lack of familial aggregation found between progression of osteophytes in the probands and siblings.

Familial aggregation in different joint groups tended to be more prominent for progression of OA in the hips and knees than in the hands. The inability to detect a familial effect on progression in the hands in the present study cannot exclude a potentially important role of genetic factors in the progression of hand OA. It is well recognized that radiologic assessment of progression of hand OA has several problems. It remains unclear which joints should be evaluated for progression and which is the best scoring system to use.

The results of the present study should be interpreted within the context of a population in which symptomatic familial OA with a relatively early onset (between 40 and 70 years of age) occurs at multiple joint sites. In this population,  40% of patients showed progression of JSN or osteophytes over 2 years. Little is known about radiologic disease progression in this OA phenotype. The only other study on OA progression at multiple joint sites was a retrospective cohort study of women over the age of 40 conducted by Cerhan et al ([18]). The mean difference in full-body OA score, including 10 joint groups (range 0-10), between baseline and followup after 1-9 years was only 0.9. It is difficult, however, to compare this finding with the results of our study due to differences in patient populations, studied joint groups, and definitions of radiologic outcome.

A potential limitation of the present study was the possibility to misclassify patients for the presence or absence of OA progression. Currently, there are no recommendations for the assessment of OA progression at multiple joint sites on an individual basis. We defined radiologic progression of OA in 2 ways: as a 1-point score increase in JSN and osteophyte total scores or as a 1-point score increase in JSN and osteophyte scores in at least 1 joint site on standardized radiographs of the hands, hips, and knees. Both definitions yielded similar results. However, the maximal potential total score for JSN or osteophytes in the hands was higher than in the hips or knees. It is difficult to weight the responses in the different joint groups. It is not clear if progression characterized as one additional hand joint is as significant as one additional knee joint. Any misclassification of progression, however, will be nondifferential because it was assessed in the siblings blinded to and therefore unrelated to progression in the probands. Such nondifferential misclassification will tend to dilute the effect, producing estimates of the effect that are closer to a no-effect value than the actual effect.

In 11 patients who already had bilateral hip prostheses at inclusion into the study, we altered our definition of OA progression at the hip joint level. Clearly these patients were unable to progress further at the hip joint level. Therefore, we considered these patients to have progressive disease in terms of JSN and osteophytes in their hip joints. This approach was used only in the subanalyses at the hip joint level.

In conclusion, our study provides evidence for familial aggregation of radiologic progression of JSN, even over a relatively short followup period of 2 years. These results suggest that familial factors play a role in the progression of cartilage loss in the hand, hip, and knee joints of patients with OA at multiple joint sites. Our understanding of the complex mechanisms driving OA progression may be increased by identifying genetic loci contributing to the familial effects on cartilage degeneration, which may reveal potential pathophysiologic targets for treatment of OA.

AUTHOR CONTRIBUTIONS

Dr. Kloppenburg had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Slagboom, Meulenbelt, Breedveld, Kloppenburg.

Acquisition of data. Botha-Scheepers, Watt, Kloppenburg.

Analysis and interpretation of data. Botha-Scheepers, Watt, Slagboom, Meulenbelt, Rosendaal, Kloppenburg.

Manuscript preparation. Botha-Scheepers, Watt, Rosendaal, Breedveld, Kloppenburg.

Statistical analysis. Botha-Scheepers, Rosendaal, Kloppenburg.

ROLE OF THE STUDY SPONSOR

Pfizer did not have a role in the design and conduct of the study, collection, management, analysis, and interpretation of the data, or preparation of the manuscript.

Acknowledgements

The authors would like to acknowledge the support of the cooperating hospitals and referring rheumatologists, orthopedic surgeons, and general practitioners in our region. We are very grateful to N. Riyazi for inclusion of the patients, J. Krol-van Berkel for data management, and C. van Driel-Kroesbergen for performing the radiologic examinations in the GARP study.