Accelerated hand bone mineral density loss is associated with progressive joint damage in hands and feet in recent-onset rheumatoid arthritis.

INTRODUCTION: To investigate whether accelerated hand bone mineral density (BMD) loss is associated with progressive joint damage in hands and feet in the first year of rheumatoid arthritis (RA) and whether it is an independent predictor of subsequent progressive total joint damage after 4 years.
METHODS: In 256 recent-onset RA patients, baseline and 1-year hand BMD was measured in metacarpals 2-4 by digital X-ray radiogrammetry. Joint damage in hands and feet were scored in random order according to the Sharp-van der Heijde method at baseline and yearly up to 4 years.
RESULTS: 68% of the patients had accelerated hand BMD loss (>-0.003 g/cm2) in the first year of RA. Hand BMD loss was associated with progressive joint damage after 1 year both in hands and feet with odds ratios (OR) (95% confidence intervals [CI]) of 5.3 (1.3-20.9) and 3.1 (1.0-9.7). In univariate analysis, hand BMD loss in the first year was a predictor of subsequent progressive total joint damage after 4 years with an OR (95% CI) of 3.1 (1.3-7.6). Multivariate analysis showed that only progressive joint damage in the first year and anti-citrullinated protein antibody positivity were independent predictors of long-term progressive joint damage.
CONCLUSIONS: In the first year of RA, accelerated hand BMD loss is associated with progressive joint damage in both hands and feet. Hand BMD loss in the first year of recent-onset RA predicts subsequent progressive total joint damage, however not independent of progressive joint damage in the first year.

RCT Entities:   Population Interventions Outcomes

Introduction

Bone damage in rheumatoid arthritis (RA) includes joint damage and accelerated bone mineral density (BMD) loss [1]. Joint damage is provoked by an increased osteoclast and decreased osteoblast activation, leading to erosive damage, and by proteolytic pathways, leading to cartilage degradation. This is all mostly regulated by TNF-α, IL-1, IL-6, IL-17 and receptor activator of nuclear factor kappa B ligand (RANKL) [2-4]. It is believed that BMD loss, both localized and generalized, is also primarily the effect of increased osteoclast activity in RA [5]. In particular, bones in the proximity of inflamed joints are susceptible to BMD loss due to inflammation [6]. Furthermore, localized hand BMD loss occurs in an early phase of RA [7] and even in pre-RA undifferentiated arthritis [8], and might precede erosive damage on X-ray [9,10].

Dual energy X-ray absorptiometry (DEXA) is the gold standard for measuring BMD. Digital X-ray radiogrammetry (DXR) was developed as a method of radiogrammetry to estimate BMD in the metacarpals using standard hand radiographs [11]. BMD measured by DXR is highly correlated with DEXA measurements and DXR has a high precision for detecting changes in BMD [11,12]. Various clinical studies showed the association between hand BMD loss measured by DXR and RA severity, including disease activity, functional impairment and joint destruction [6,13-22]. Two clinical studies, one of them a pilot study, showed the potential value of BMD loss in hands measured by DXR to predict radiographic joint damage in hands [23,24]. However, to date, no data are available on the association between hand BMD loss and progressive joint damage in hands and feet and on the value of hand BMD loss as predictor of joint destruction in recent-onset RA patients who are treated intensively with disease modifying anti-rheumatic drugs (DMARDs) and TNF-α inhibitors in a tight control setting. We examined the association between accelerated hand BMD loss and progressive joint damage in hands and feet during the first year of recent-onset active RA to see whether both types of bone damage have common pathways in their pathogenesis, and we investigated whether accelerated hand BMD loss in the first year of RA was an independent predictor of subsequent progressive joint damage after four years in patients who are treated in a tight control setting.

Materials and methods

Patients

All measures were performed in the setting of the Behandel Strategieën (BeSt) study [25]. Patients aged 18 years and older, who met the definition of RA as defined by the American College of Rheumatology (ACR) 1987 revised criteria, with symptom duration of less than two years and active disease with 6 or more of 66 swollen joints and 6 or more of 68 tender joints and either an erythrocyte sedimentation rate (ESR) of 28 mm/hour or more or a visual analogue scale (VAS) global health of 20 mm or more, and who were DMARD naïve, were included in the trial from April 2000 to August 2002. Exclusion criteria have been reported previously [25]. Of the 508 patients, 236 were excluded from this study predominantly due to switch from analogue to digital radiographs. The other 272 patients had analogue radiographs at both baseline and after one year and were eligible for this study. The baseline and/or one year follow-up analogue radiographs of 16 patients could not be analysed by DXR due to underexposed images (13 patients) or improper positioning of the hands (3 patients). Hence, 256 patients were included in the current study.

Study design

The BeSt study was conducted by rheumatologists participating in the Foundation for Applied Rheumatology Research, in 18 peripheral and 2 university hospitals in the western part of the Netherlands. The medical ethics committee at each participating center approved the study protocol and all patients gave written informed consent prior to participation in the study.

After inclusion, patients were randomized to be treated according to one of four treatment strategies: sequential monotherapy starting with methotrexate (MTX); step-up combination therapy also starting with MTX; initial combination therapy with quickly tapered high-dose prednisone, MTX and sulphasalazine, or initial combination therapy with infliximab and MTX. For all groups, the protocol described a number of subsequent treatment steps for patients whose response to therapy was insufficient, based on the disease activity score (DAS) in 44 joints of more than 2.4. The treatment protocol and the effect of the different treatment strategies on hand BMD loss after one and two years are described earlier in detail [6,25].

Concomitant treatment with non-steroidal anti-inflammatory drugs and intra-articular corticosteroids were permitted but not parenteral corticosteroids. In case of calcium intake of less than 1,000 mg/day and serum vitamin D level below the local reference value at baseline, suppletion of 500 to 1,000 mg/day calcium and 400 IE/day vitamin D (colecalciferol), respectively, was advised. Antiresorptive therapy with oral alendronate, 10 mg/day or 70 mg/week, or risedronate, 5 mg/day or 35 mg/week, was advised if the BMD measurement at baseline showed at a T-score -2.5 standard deviations (SD) or less in the spine and/or hip in non-corticosteroid users or a T-score -1 SD or less in corticosteroids users (ACR recommendations) [6].

Hand BMD measurements

Analogue radiographs of both hands in the posterior-anterior view were digitized by a high-resolution 300 DPI scanner (Canon Vidar VXR-12 plus, Amstelveen, North-Holland, The Netherlands) and analysed under blind conditions using the DXR-online (Pronosco X-posure system, Sectra, Sweden). According to the manufacturer, there is a very good agreement between BMD measured by DXR on original analogue radiographs and on digitalized versions. Patients who switched from analogue to digital radiographs were excluded due to lack of comparability between the different imaging devices.

DXR is a computerized version of the traditional technique of radiogrammetry originally proposed by Barnett and Nordin [26]. The digitized hand radiograph is subjected to a number of image processing algorithms to measure the cortical thickness of three regions of interest around the narrowest part of the second, third and fourth metacarpal bones [11]. A mean surrogate BMD, based on the mean volume per area, was calculated in g/cm2 with correction for the estimated porosity. Both hands were measured and the mean was used for the analyses.

Hand BMD loss after one year was categorized in two groups using the cut-off of -0.003 g/cm2/year, equal to the upper limit of normal BMD loss in the metacarpals according to specifications by the manufacturer.

Radiographic assessment of progressive joint damage

Radiographic progression of joint damage in hands and feet at baseline and after one to four years were independently scored by two readers, blinded for patient characteristics, treatment group and time order, using the Sharp-van der Heijde score (SHS) [27]. The inter-class correlation coefficient between the readers was 0.98. The mean score of the two observers were used for the analyses.

Progressive joint damage in the first year was defined as an increase in total SHS of 5 units or more at year one compared with baseline. In subanalyses, progressive joint damage in hands was defined as delta SHS 0 to 1 year 5 units or more, whereas progressive joint damage in feet was defined as delta SHS 0 to 1 year 3 units or more, due to a 0.6 times lower maximum score in feet than in hands. Subsequent progressive total joint damage in hands and feet was defined as delta SHS 1 to 4 years 5 units or more.

Demographic and clinical variables

The following variables were collected at baseline: age; sex; and symptom duration. At baseline and after one year the following variables were collected: postmenopausal status; body mass index (BMI); DAS; based on the number of swollen joints and the Ritchie articular index (RAI) for pain in tender joints; the VAS for patient's global assessment of disease activity (0 to 100 mm); ESR; C-reactive protein (CRP); serum IgM rheumatoid factor (RF), defined as positive or negative according to locally applied assays and cut-off units; and functional disability by the Dutch validated health assessment questionnaire (HAQ). The presence of anti-citrullinated protein antibodies (ACPA) was determined from serum samples obtained at baseline or during follow up. The presence or absence of ACPA is a stable characteristic [28].

Statistical analyses

All analyses were performed in an intention-to-treat method using all available data.

To determine the associations between hand BMD loss and progressive joint damage in hands and feet after one year, multivariate logistic regression analyses were performed adjusted for age, gender, postmenopausal status, BMI, HAQ, baseline SHS, treatment group, and the use of intra-articular steroids and antiresorptive drugs (bisphosphonates, vitamin D and calcium supplements and hormone replacement therapy (HRT)).

The sensitivity, specificity, and positive and negative predictive value of hand BMD loss with regard to total progressive joint damage in the first year were calculated. Various baseline demographic and disease-related factors and one-year follow-up disease-related factors were analysed regarding prediction of subsequent progressive total joint damage after four years by univariate logistic regression analyses adjusted for age, gender, postmenopausal status, BMI and HAQ, and additionally adjusted for the treatment group and use of antiresorptive drugs and intra-articular steroids during the first year follow-up in case of one-year follow up variables. The following factors were analysed: baseline demographic factors (gender, age ≥ 50 years, postmenopausal status and BMI ≥ 25 kg/m2), baseline disease-related factors (symptom duration ≥ 6 months, presence ACPA and RF, number of swollen joints ≥ 10, RAI ≥ 10, ESR ≥ 30 mm/hr, CRP ≥ 10 mg/L, HAQ ≥ 1.057 units [29] and SHS ≥ 1 unit) and one-year follow-up disease-related factors (high area under the curve (AUC) of number of swollen joints, RAI, ESR and CRP and delta HAQ ≤ -0.22 units [30], total SHS ≥ 5 units and hand BMD loss >-0.003 g/cm2). Both significant (p-value < 0.05) and borderline significant (0.05