Background/Purpose:

Cartilage loss, represented by joint space width (JSW) on radiographs, is an important outcome measure in hand OA disease course. JSW can be assessed quantitatively in a valid, reliable and feasible way. However, it is not known whether this method is sensitive over time and was therefore subject of this 2-year follow-up study.

Methods:

We used data of 276 participants (mean age 61 yrs, 88% women, mean BMI 27 kg/m², 93% fulfilled ACR criteria, median number of joints with pain 2 (range 0-24)) in the ongoing HOSTAS (Hand OSTeoArthritis in Secondary care) study, including consecutive primary hand OA patients diagnosed by their treating rheumatologist.

During physical examination (PE) in the 2^nd to 5^th DIP and PIP joints the number of joints painful upon palpation, with bony swellings, soft-tissue swelling, limited Range Of Motion (ROM) and deformity was assessed (all 0-16). Digital hand radiographs (dorsal-volar views) were taken at baseline and after two years.

JSW was measured in 2^nd to 5^th DIP, PIP and MCP joints (n=24) with a semi-automatic quantification method (van ‘t Klooster Osteoarthritis Cartilage 2008; 16: 18-25. Openly available through www.lkeb.nl (software downloads)). ICC were calculated based on repeat measurement in 40 radiographs to determine reliability.  Smallest Detectable Differences (SDDs) of all joints and per joint group were calculated by measuring two radiographs from 21 patients acquired within a 6-months (max 196 days) time-interval, assuming no progression over such short time.  

Progression was defined as a change more than the SDD. Progression on patient level was defined  when at least 1  joint had progressed.

Generalized Estimating Equations (GEE) were used to test on joint level in 2^nd to 5^th DIP and PIP joints which factors associate with progression. Adjustments were made for patient (age, sex, BMI) and joint characteristics (all five PE assessments and baseline JSW).

Results:

After two years, of 192 patients both baseline and follow-up radiographs were available to assess progression. Of the 4570 joints measured, 8.8% progressed, which was seen most often in the DIP joints. On patient-level, 76.6% progressed. The results are depicted in table 1. The reliability was high with ICC 0.97.

Variables at joint level associated with progression are JSW at baseline OR 9.9 (95%CI 5.1;19.0), pain upon palpation dichotomized to present/absent 1.4 (1.02;1.9) and limited ROM 1.9 (1.4;2.5). All the other variables had no significant association.

Conclusion:

Semi-automated JSW measurement is sensitive to change over 2 years. Risk factors for progression on joint level are: a relatively large JSW on baseline, pain upon palpation and limited ROM.

Table 1: JSW, SDD and progression in 195 patients with hand osteoarthritis