Is PMR a separate entity?

Although characteristic for PMR, polymyalgic symptoms, including pain and stiffness in the shoulder and pelvic girdle, are non-specific clinical manifestation. They may serve as the initial presentation for a range of elderly-onset rheumatic (eg, rheumatoid arthritis, spondylarthritis, crystal-induced arthritis, vasculitis and connective tissue diseases) and non-rheumatic conditions (eg, infections and neoplasms).4 Consequently, it is essential to reevaluate the concept of PMR, viewing it not as a monolithic clinical entity but rather as a complex, multifaceted clinical syndrome (figure 1). However, although crucial, distinguishing between isolated PMR and PMR-like onset diseases is not always easy, given that similar symptoms may be the expression of different conditions (table 1) and given the absence of a specific diagnostic tests. Misclassification is, therefore, common even among experts, occurring in as many as one-third of cases.5

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

The spectrum of polymyalgia rheumatica (PMR) is a complex syndrome clinical entity with multiple facets. Within the realm of its characteristic symptoms, we encompass not only isolated PMR but also a range of elderly-onset rheumatological conditions that may overlap with (or mimic) PMR to varying degrees forming part of a disease spectrum (also see table 1). Rheumatic diseases other than GCA and non-rheumatological mimickers: see table 1. Yellow dotted line: GPSD. GCA, giant cell arteritis; GPSD, GCA-PMR spectrum disease.

PMR has long been perceived as a disease largely handled in primary care, owing to its swift and often complete response to glucocorticoid (GC) treatment. Referral to secondary care has typically been reserved to cases with an incomplete response or treatment-related adverse events. However, accumulating evidence suggests that PMR is part of a disease spectrum with giant cell arteritis (GCA), necessitating tailored diagnosis, treatment and monitoring. The term ‘GCA-PMR spectrum disease’ (GPSD) has recently been proposed,6 referring to a continuum from pure PMR to pure vasculitic patients, with varying degrees between the two ends of this spectrum. It has long been recognised that 40%–60% of GCA patients exhibit PMR characteristics and that approximately 15% of PMR patients may develop GCA signs and symptoms.7 However, only modern imaging techniques led to a deeper understanding of the overlap between PMR and (mostly subclinical) vasculitis of the large arteries. A metanalysis revealed that 29% of new-onset, GC-naïve PMR patients, who did not present any classical sign of systemic vasculitis (eg, headache, jaw or limb claudication), were identified with subclinical GCA by 18F-fluorodeoxyglucose positron emission tomography (FDG-PET) examination.8 Moreover, a recent cross-sectional study reported a similar percentage (23%) of vasculitis when ultrasound of superficial temporal, common carotid, subclavian and axillary arteries was performed.9 In the same article, it was shown that PMR patients with subclinical GCA exhibited a higher incidence of isolated extracranial (39% vs 19%) and lower frequencies of isolated cranial vessel involvement (28% vs 53%) in comparison with classic GCA.

Subclinical GCA in PMR patients: does it matter?

In a study by Blockmans et al, it was reported that PMR patients with FDG-PET-verified large-vessel vasculitis showed comparable response and relapse rates to those without vascular involvement.10 Sample size of that study was low, however, limiting the possibility to detect any true difference between the groups. In contrast, De Miguel et al showed, in an article published in the Annals of Rheumatic Disease, that individuals with ultrasound verified subclinical GCA experienced relapses four times more frequently than PMR patients without large-vessel involvement.11 Notably, faster GC tapering within the initial 3 months was associated with a higher relapse rate in patients with subclinical GCA, suggesting the potential necessity for a more aggressive treatment approach in this specific subgroup. A notable limitation of De Miguel et al’s study is the lack of blinding of the treating physicians with respect to the ultrasound result. This led not only to a higher initial GC dose in patients with subclinical GCA, but also raises concerns about the objectivity and consistency of monitoring and outcome assessment in the two groups.

Why did Blockmans et al and De Miguel et al come to different conclusions? Apart from potential variances in the patient populations and the methodological concerns raised above, the certainty of large-vessel inflammation may have differed between the studies. FDG-PET detects increased metabolism in the vessel wall, which does not necessarily reflect inflammation; rather it could be caused by arteriosclerosis or arterial wall remodelling. Besides, vascular tracer uptake in that study was lower than what we typically observe in classic GCA, raising concerns about the specificity of some positive FDG-PET scans to indicate true vasculitis.

Should we perform an imaging test to access vascular involvement in every PMR patient in routine clinical practice?

While it is tempting to respond with a simple ‘yes!’ especially given our enthusiasm for ultrasound, we must consider several factors. First, the results from DeMiguel et al’s study need to be confirmed in an independent cohort. Second, there are currently no data on the potential long-term impact of subclinical GCA for the prognosis of PMR patients. Unlike GCA, PMR does not result in disease-related tissue damage, and corresponding data for PMR patients with subclinical large-vessel vasculitis is lacking. Third, the optimal treatment approach for these patients to prevent relapses remains uncertain. Specialised trials stratifying patients within the GCA-PMR spectrum, particularly PMR patients based on the presence or absence of large-vessel vasculitis, are therefore necessary to investigate the impact of subclinical GCA on treatment outcomes.

Can PMR still be seen as a disease of primary care?

Considering that ultrasound-based screening of large arteries would foster stratification and prognostic evaluation of patients according to the GPSD concept, it is questionable whether this can be afforded in primary care. The idea of referring all PMR patients to specialist evaluation has already been part of the British society of Rheumatology recommendations for the management of PMR,12 and has recently been incorporated in the early referral recommendations of an international PMR/GCA study group.13 Whether imaging modalities (ultrasound or other imaging) are available at the specialist level in order to conduct vascular screening of all PMR patients, and whether rheumatology workforce is sufficient to guarantee a timely visit to all these patients are unclear yet. PMR is one of the most common inflammatory rheumatic diseases in adults aged ≥50 years, with an incidence of up to 113:100 000.4 We expect that with the demographic changes in the population, the incidence of this disease will increase further. If all suspected PMR patients were sent to specialist care for ultrasound evaluation and disease stratification, thousands of additional new cases would have to be covered by specialists, while waiting lists for a first rheumatological appointment are already long in most European countries. The solution to this problem could either be the development of PMR-fast track clinics, assigning additional resources to PMR, the development of screening tools for general practitioners to identify patients at risk for subclinical large-vessel inflammation or a combination of the two. Effective clinical stratification instruments for primary care are yet to be developed, and it is possible that certain manifestations, such as constitutional symptoms or markedly elevated inflammatory markers, may play a pivotal role in identifying PMR patients with subclinical vasculitis. This assumption is supported by the association of these parameters with large-vessel inflammation in GCA.14

Should PMR patients with subclinical vasculitis be managed as GCA or as PMR?

Patients with GCA currently receive at least twice the initial GC dose than PMR patients, regardless of whether GCA manifestations are predominantly cranial or extracranial.15 Individuals with extracranial large-vessel vasculitis generally face a lower risk of ocular complications (and hence, are less of a clinical emergency); however, they are at a higher risk of relapse.16 Speculating that GCA patients with extracranial large-vessel involvement and PMR patients with subclinical vasculitis share more commonalities than differences in terms of clinical presentation and prognosis, the question arises whether both entities should be managed in a similar way with regard to GC dose, application of GC sparing agents and monitoring strategies. A related aspect is the need to understand the extend of concomitant aortic involvement in PMR patients with subclinical GCA and the potential for the development of aortic aneurysms. A recent study on (classical) GCA patients revealed a strong association between ultrasound verified extracranial involvement and the presence of aortitis, in 92% of cases when a PET-CT was performed.17 It will be vital to determine if this association holds true also for PMR patients with subclinical GCA and what will be the corresponding risk for developing aortic aneurysms or arterial stenoses. Although several aspects require further clarification, the emergence of new, effective therapeutic alternatives for both diseases is reassuring. While the Food and Drug Administration (FDA) and European Medicines Agency endorsed the use of the IL-6 inhibitor tocilizumab for GCA some years ago,18 the IL-6R inhibitor sarilumab has recently received FDA approval for PMR.19

What are the next steps?

First, we need to understand what is the minimum GC dose to effectively treat PMR with and without subclinical vasculitis, and whether patients in the latter group benefit from early use of GC-sparing agents. These questions can only be answered by strategic trials in which patients are stratified according to the presence or absence of subclinical vasculitis. A challenge of such trials will be the assessment of response: should criteria of PMR or GCA be used, or both? Should imaging be part of outcome assessment in such trials? Universally accepted criteria for response exist neither for PMR nor GCA, even though international study groups of European Alliance of Associations for Rheumatology (EULAR)/American College of Rheumatologoy (ACR) and Outcome Measures for Arthritis Clinical Trials (OMERACT). Whether response criteria for PMR with subclinical GCA are developed, however, is questionable. Second, we need clinical and laboratory predictors of subclinical GCA in order to facilitate rapid referral of these patients to specialist care. Lastly, it is imperative to enhance access to and training in vascular imaging. Ultrasound is an instrument that is readily available, yields low costs, good sensitivity and specificity and is not burdened by potential risks from contrast agents and radiation. In addition, this technique revealed excellent diagnostic performance in patients with suspected GCA.20 Despite the increasing utilisation of vascular ultrasound in European countries, and the efforts of EULAR and other groups to foster training with this technique, its implementation into the curricula of rheumatologists still remains an unmet need.21

To conclude, behind an apparent simplicity, PMR syndrome hides a significant complexity, leaving several questions unanswered.22 As the notion of disease stratification within GPSD gains recognition, De Miguel et al’s research suggests that PMR patients with subclinical GCA may experience a poorer prognosis compared with those without. This underscores the potential need for vascular screening in specialist care in every PMR patient, necessitating a substantial expansion of the rheumatological infrastructure. However, additional evidence is required to validate these findings and to determine whether tailoring treatment based on the presence or absence of subclinical large-vessel vasculitis leads to better short-term and long-term outcomes of these patients.