Why is PMR classified as 'provisional' in these 2012 criteria?▼
The 2012 EULAR/ACR criteria were explicitly labeled provisional at the time of publication because they had been developed and validated in a single derivation cohort rather than an independent external validation cohort. The criteria were derived from a 6-month prospective cohort of 125 new-onset PMR patients and 169 comparison subjects at multiple centers in Europe and the United States. The authors acknowledged that while the c-statistic of 81% exceeded the conventional threshold of 80% for clinical utility, further validation in diverse populations and settings was needed before they could be considered fully endorsed. The word "provisional" also acknowledged that PMR remains a clinical diagnosis without a true gold standard biomarker, and that the scoring algorithm - while useful for research standardization - reflects probability of PMR rather than definitive disease classification.
What conditions most commonly mimic PMR and drive misclagnosis?▼
The derivation cohort included 169 non-PMR subjects with conditions that most commonly mimic PMR in clinical practice. These were grouped into: inflammatory shoulder conditions (rotator cuff pathology, adhesive capsulitis, glenohumeral OA, bilateral rotator cuff syndrome), rheumatoid arthritis (particularly seronegative RA with proximal presentation - the hardest mimicker, as the criteria have only 65% specificity against RA), late-onset spondyloarthritis and other forms of inflammatory arthritis, remitting seronegative symmetric synovitis with pitting edema (RS3PE), fibromyalgia, calcium crystal arthropathy (CPPD), hypothyroidism and paraneoplastic syndromes. Seronegative RA masquerading as PMR is the most clinically important distinction because misdiagnosis leads to inappropriate glucocorticoid monotherapy that initially responds but later fails.
Why does negative serology (RF and ACPA) score 2 points - the same as prolonged morning stiffness?▼
In the multivariable logistic regression analysis underpinning the criteria, absence of both RF and ACPA had one of the strongest abilities to distinguish PMR from RA specifically. Because seronegative RA is the most clinically problematic PMR mimic, and because positive serology essentially points away from PMR toward RA, negative serology was weighted heavily. The 2-point weight reflects that bilateral shoulder pain with prolonged stiffness but with positive serology should NOT be classified as PMR even if the symptom pattern fits. Conversely, when a patient has the typical PMR clinical picture AND is seronegative, the combination is strongly supportive. The criteria were designed to have their maximum performance when both morning stiffness and negative serology are present (4 points from these two items alone, meeting the threshold even if hip and peripheral items are absent).
Why does the ultrasound add-on increase specificity but not sensitivity?▼
Adding ultrasound raises the classification threshold from ≥4 to ≥5, which means a higher total score is required to classify as PMR. This additional requirement naturally increases specificity (fewer false positives) at the cost of sensitivity (slightly fewer true positives are captured). Ultrasound findings like bilateral subdeltoid bursitis are highly characteristic of PMR but are not universal - some patients with genuine PMR may not demonstrate these findings on ultrasound at the time of assessment, particularly early in the disease or after treatment has begun. The ultrasound criterion adds +1 point when the characteristic PMR bursitis pattern is present, but the higher threshold means patients who have the clinical picture but a non-diagnostic US score less favorably. This tradeoff makes the ultrasound algorithm more useful for research contexts requiring higher confidence in PMR classification.
What exactly constitutes the PMR-specific ultrasound findings?▼
PMR is now understood as predominantly a periarticular inflammatory disease centered on bursal and tenosynovial structures rather than synovitis. The key ultrasound findings include: Subdeltoid (subacromial) bursitis - thickening of the subacromial/subdeltoid bursa (>2mm) with or without Doppler signal indicating hyperemia; Bicipital tenosynovitis - fluid in the tendon sheath of the long head of biceps; Hip trochanteric bursitis - fluid or thickened bursa over the greater trochanter; and Hip synovitis - increased fluid in the hip joint capsule. The 2012 criteria specifically credit bilateral subdeltoid bursitis (the most characteristic bilateral shoulder finding) or the combination of unilateral subdeltoid bursitis with hip involvement. Importantly, a normal or equivocal ultrasound does not exclude PMR - it simply does not add the bonus point.
What is the role of CRP and ESR in PMR diagnosis beyond the entry criterion?▼
In the 2012 criteria, abnormal CRP and/or ESR is an entry criterion rather than a scored domain. This was by design: the working group found that essentially all PMR patients have elevated acute-phase reactants at diagnosis (only 5 of 125 PMR patients in the cohort had normal CRP and ESR), making it a near-universal feature. However, because it doesn't discriminate between PMR and mimicking conditions (which also raise inflammatory markers), it was included as an obligate entry criterion rather than given discriminative weight in the scoring algorithm. In clinical practice, significantly elevated CRP (>40-50 mg/L) or ESR (>50 mm/hr) at presentation is highly supportive, and a dramatic response of these markers to glucocorticoid therapy within 1-2 weeks is both diagnostically confirmatory and therapeutically reassuring.
How does hip pain in PMR differ from hip OA or trochanteric bursitis?▼
Hip involvement in PMR classically presents as proximal thigh aching and difficulty with activities requiring hip rotation (getting in/out of a car, rolling over in bed). On physical examination, passive internal and external rotation of the hip may be restricted or painful. This contrasts with hip OA, where pain is typically groin-based, mechanical, associated with crepitus and radiographic changes, and worse with weight-bearing rather than worst in the morning. Trochanteric bursitis causes lateral hip pain with tenderness over the greater trochanter and is often worse lying on the affected side. In PMR, hip imaging with ultrasound or MRI demonstrates hip synovitis and/or trochanteric bursitis as periarticular structures rather than cartilage/bone pathology, distinguishing it from degenerative disease. The periarticular pattern is key: PMR affects structures around joints, not the joints themselves as in primary synovitis.
What is the typical glucocorticoid response pattern used to support PMR diagnosis?▼
The rapid and dramatic response to low-dose glucocorticoids is considered a near-essential feature of PMR in clinical diagnosis, though it is not part of the 2012 classification criteria because it requires retrospective evaluation and was not included in the prospective cohort study design. Clinically, PMR patients classically experience near-complete resolution of stiffness and pain within 24-72 hours of starting prednisolone 15-20 mg/day, along with normalization of CRP and ESR within 2-4 weeks. This response is so characteristic that failure to respond to adequate glucocorticoids within 1 week should raise concern for an alternative diagnosis (particularly RS3PE, late-onset RA, or malignancy). The EULAR/ACR 2015 management recommendations establish this rapid response as an expected treatment outcome, and persistent disease activity at 3-4 weeks on appropriate doses warrants diagnostic re-evaluation.
When should GCA (giant cell arteritis) be suspected in a patient classified as PMR?▼
Up to 10-20% of patients classified as having PMR will have concurrent giant cell arteritis (GCA), and all PMR patients should be screened at diagnosis for GCA features. Red flags requiring urgent GCA evaluation include: new-onset headache (particularly temporal or occipital), jaw or tongue claudication, scalp tenderness, visual symptoms (amaurosis fugax, diplopia, blurred vision - ophthalmic emergency), polymyalgia without response to standard doses, or constitutional symptoms disproportionate to the PMR picture. GCA requires higher glucocorticoid doses (40-60 mg/day) and may require temporal artery biopsy or large-vessel imaging (PET-CT, MR angiography) for confirmation. The 2012 PMR criteria explicitly excluded patients with GCA features from the PMR classification, reflecting that PMR with GCA is a distinct entity requiring different management.
What are the main limitations of these criteria for routine clinical use?▼
The 2012 criteria have several important limitations for clinical practice. First, the specificity against RA is only 65% - meaning seronegative RA presenting with proximal stiffness will frequently meet PMR criteria, leading to diagnostic errors. Second, the criteria were derived predominantly from European white populations at specialist centers, limiting generalizability to primary care and diverse ethnic backgrounds. Third, the criteria require new-onset symptoms evaluated before glucocorticoid treatment begins - retrospective application or evaluation after steroid initiation is unreliable. Fourth, the ultrasound component requires trained operators and appropriate equipment, limiting its universal applicability. Fifth, as a research tool, the criteria intentionally sacrifice some sensitivity to achieve specificity - meaning some genuine PMR cases will score below threshold. The criteria are best used to ensure research cohort homogeneity, not to confirm or exclude individual diagnoses.