• osteoarthritis
• ultrasonography
• synovial fluid
• crystal arthropathies

Despite the high prevalence and symptom burden of osteoarthritis (OA), our understanding of its pathogenesis is incomplete. Intra-articular crystal deposits have been implicated in OA pathogenesis, with their ability to activate inflammatory pathways as well as cell proliferative and apoptotic reactions.1 2 For people with knee OA, basic calcium phosphate crystals in the synovial fluid are associated with worse patient-reported pain and function.3 So far, no study has investigated the association between intra-articular crystals and symptoms in hand OA. We obtained data as part of a cross-sectional single-centre study of people with hand OA undergoing phalangeal joint surgery (ClinicalTrials.gov NCT04585113). All participants signed informed consent before enrolment, met the American College of Rheumatology classification criteria for hand OA and had no other rheumatic diseases including gout.4 We also excluded people with skin psoriasis with or without psoriatic arthritis and individuals who had intra-articular injections within 3 months prior to enrolment. Within 4 weeks prior to surgery, participants underwent a standardised clinical examination and completed questionnaires including a Visual Analogue Scale (VAS) of finger pain, a VAS patient global assessment, the Australian/Canadian Hand Osteoarthritis Index (AUSCAN) and the European Quality of Life 5 Dimensions (EQ-5D). VAS was scored on a 100 mm with anchors 0=‘no pain’ and 100=‘worst possible pain’, the AUSCAN was scored on a Numerical Rating Scale by subscale pain (scored as 0–50) and function (0–90), and the EQ-5D score ranged from −0.624 (worst) to 1.000 (best). Clinically, we assessed the number of tender and swollen joints (present or absent) at second–fifth distal and proximal interphalangeal joints, first–fifth metacarpophalangeal joints, first interphalangeal joint and first carpometacarpal joint; grip strength, height, weight, pain medication use and symptom duration were recorded. Pain medication was recorded as intake in the week prior to assessment. For imaging, participants underwent ultrasound, cone-beam CT and radiographs of the hand with the joint scheduled for surgery. Synovial fluid and joint lavage fluid were obtained during surgery and examined within 60 min using light and compensated polarised light microscopy by a senior rheumatologist trained in crystal detection. Given the limited synovial fluid in the finger joints, we only obtained one to three drops from each participant. A drop of synovial fluid or joint lavage fluid was placed on a glass slide with a cover sheet and then systematically assessed for distinct crystal morphology. Negative and positive birefringent crystals were interpreted as monosodium urate and calcium pyrophosphate, respectively. We considered presence of ≥1 crystal as a positive reference test. Continuous variables are presented as mean with SD and skewed data as median with IQR. Categorical data are presented as absolute numbers with percentages. R studio V.3.6.1 was used for statistical analysis.

We screened 13 participants for eligibility and excluded 1 due to known gout, leaving 12 participants for inclusion in the study between 1 October 2020 and 22 August 2022. From synovial fluid and joint lavage fluid analyses, the prevalence of crystals was 42% (5 of 12); of these, 60% (3 of 5) were calcium pyrophosphate, and 40% (2 of 5) were monosodium urate. The characteristics of the participants are presented in table 1. No participants had elevated ionised calcium (>1.32 mmol/L), ferritin (>300 µg/L) or haemoglobin 1Ac (>48 mmol/mol). One participant with monosodium urate crystals had chondrocalcinosis in the triangular fibrocartilage, elevated plasma urate and was diagnosed with gout after participation in the trial. Statistically significant differences in VAS pain persisted after exclusion of participants with monosodium urate crystals identified in joint fluid.

Calcium-containing crystals have been associated with degenerative changes in knee OA, but the relationship with hand OA symptoms has not previously been described.3 5 In our study of participants with finger joint OA scheduled for joint surgery, we found for the first time that pain and function were worse among participants with calcium-containing crystals in the synovial fluid compared with participants without. The differences were not explained by radiographic structural differences of the hand or ultrasound inflammation. Whereas all participants without crystals in the synovial fluid had evidence of ultrasound inflammation, only two of five participants with crystals in the synovial fluid had evidence of ultrasound inflammation. The prevalence of crystals in the synovial fluid in this study is higher than a previous study on wrist and finger OA and may be explained by selecting people awaiting surgery.6 Regardless, the result suggests the presence of a crystal phenotype in hand OA associated with increased OA symptom severity. The phenotype can be defined as hand OA with comorbid crystal arthropathies. This phenotype may be suggested as a potential stratification factor in trials enrolling people with hand OA. The small sample size limits this study, and the observed associations should be confirmed in future studies, which may also assess the clinical impact of the crystal phenotype.