Diagnostic accuracy of musculoskeletal ultrasound in suspected polymyalgia rheumatica: a multicentre, cohort study.

Background

Diagnosing polymyalgia rheumatica is challenging, and musculoskeletal ultrasound is primarily used in diagnostically difficult situations. Previous musculoskeletal ultrasound studies in patients with polymyalgia rheumatica used case-control designs or non-representative cohorts, reducing assessment of diagnostic accuracy. This study aimed to assess the diagnostic accuracy of musculoskeletal ultrasound in two independent cohorts of patients with suspected polymyalgia rheumatica.

Methods

This multicentre, diagnostic accuracy study included consecutive, glucocorticoid-naive patients with suspected polymyalgia rheumatica, referred for rheumatology evaluation at the University Medical Center Groningen, the Netherlands (retrospective cohort), between April 30, 2019, and Feb 13, 2024, and at Aarhus University Hospital, Silkeborg Regional Hospital, and Horsens Regional Hospital, Denmark (prospective cohort), between Sept 14, 2020, and June 30, 2022. The Aarhus cohort included patients older than 50 years with proximal muscle or joint pain and excluded patients with cranial symptoms suggestive of giant cell arteritis, cancer within the past 5 years, or pre-existing inflammatory rheumatic disease. No explicit referral criteria were applied in the Groningen cohort. Participants underwent protocolised clinical examination and musculoskeletal ultrasound. The reference standard was clinical diagnosis after 6 months (Groningen) or 12 months (Aarhus) of follow-up. Musculoskeletal ultrasound assessment included subacromial-subdeltoid bursitis, biceps tenosynovitis, and hip synovitis in both cohorts, plus glenohumeral synovitis and trochanteric bursitis in the Groningen cohort. A polymyalgia rheumatica ultrasound lesion count was calculated based on six (Aarhus) and ten (Groningen) sites. Diagnostic performance of individual and combined lesions was assessed using sensitivity, specificity, likelihood ratios, and receiver operating characteristic (ROC) analysis. There was no lived experience involvement in the study design or conduct. The Aarhus cohort study is registered with ClinicalTrials.gov, NCT04519580.

Findings

The Groningen cohort included 92 patients patients with suspected polymyalgia rheumatica (41 [45%] male, 51 [55%] female, mean age 69 years [SD 9]), of whom 58 (63%) were diagnosed with polymyalgia rheumatica after 6-month follow-up. The Aarhus cohort included 93 patients (53 [57%] male, 40 [43%] female, mean age 71 years [SD 8]), 66 (71%) of whom had a diagnosis of polymyalgia rheumatica at 12 months. No single musculoskeletal ultrasound finding was pathognomonic for polymyalgia rheumatica. The combination of musculoskeletal ultrasound findings with the highest sensitivity was the presence of at least one inflammatory lesion (Groningen: 10-site polymyalgia rheumatica ultrasound lesion count 93% [95% CI 83-98]; Aarhus: 6-site polymyalgia rheumatica ultrasound lesion count 92% [83-98]). The combination with highest specificity was the presence of at least one inflammatory lesion in both the shoulder and hip girdles (Groningen: 85% [95% CI 69-95]; Aarhus: 93% [76-99]). ROC analysis showed an area under the curve of 0·779 (95% CI 0·675-0·883) for the 10-site count in Groningen and 0·769 (0·664-0·875) for the 6-site count in Aarhus. Musculoskeletal ultrasound performed better in distinguishing polymyalgia rheumatica from non-inflammatory conditions than from other inflammatory rheumatic diseases.

Interpretation

Absence of shoulder and hip lesions on musculoskeletal ultrasound makes a polymyalgia rheumatica diagnosis unlikely. However, their presence requires careful clinical evaluation, as similar findings occur in other inflammatory rheumatic diseases. These findings support the use of musculoskeletal ultrasound as a potential adjunct to clinical assessment when evaluating patients with suspected polymyalgia rheumatica.

Funding

FOREUM, Danish Rheumatism Association, Independent Research Fund Denmark, Ketty and Ejvind Lyngsbæks Foundation, Frimodt-Heineke's Foundation, Aase and Ejnar Danielsen's Foundation, AP Møller Foundation, Health Research Foundation of Central Denmark Region, and Regional Hospital Central Jutland Research Foundation.