The presence of erosive joints is a strong predictor of radiological progression in hand osteoarthritis: results of a 2-year prospective follow-up of the Liège Hand Osteoarthritis Cohort (LIHOC)

BACKGROUND: This study measured the magnitude and determinants of clinical and radiological progression in patients with hand osteoarthritis (HOA) over a 2-year prospective follow-up to gain a greater understanding of the disease time course. METHODS: Two hundred three consecutive outpatients diagnosed with HOA were followed for 2 years (183 women, median age 69 years). Pain and function were evaluated using the Australian/Canadian Osteoarthritis Hand Index (AUSCAN), and clinical examination recorded the number of painful/swollen joints and nodes. X-rays were scored using Kellgren-Lawrence (KL) and Verbruggen-Veys scales. Clinical progression was defined as deterioration in AUSCAN ≥ the minimal clinically important difference. Radiographic progression was defined as (a) one new erosive/remodeled joint, (b) progression of ≥ one anatomical stage in one joint, or (c) change in KL total score above the smallest detectable difference. Logistic regression was performed to determine whether patient characteristics influenced clinical and radiological progression. RESULTS: After 2 years, all radiographic scores deteriorated significantly in the study population (p < 0.05), and the number of proximal and distal interphalangeal nodes was significantly higher (p < 0.01). The AUSCAN, number of painful joints at rest or at pressure, number of swollen joints, and pain measure on a visual analog scale remained unchanged. At the individual level, the number of patients with clinically meaningful progression ranged from 25 to 42% (clinical progression) and from 22 to 76% (radiological progression). The only significant predictor of worsening of total AUSCAN was AUSCAN pain subscale < 74.5 (odds ratio [OR] 1.02 [1.01, 1.03]; p < 0.01). The presence of ≥ four swollen joints (OR 2.78 [1.21, 6.39]; p = 0.02) and erosive osteoarthritis (OR 13.23 [5.07, 34.56]; p < 0.01) at baseline predicted a new erosive joint. A meaningful change in KL was more frequent with painful joints at baseline (OR 3.43 [1.68, 7.01]; p < 0.01). CONCLUSIONS: Evidence of radiological progression over 2 years was observed in patients with HOA in the LIHOC population even without clinical worsening of disease. For individual patients, baseline pain level is predictive for clinical progression and the presence of erosive or swollen joints are significant predictors of radiological progression. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s13075-020-02390-x.