Introduction
Polymyalgia rheumatica (PMR) is one of the most common inflammatory rheumatic diseases among older people, especially in the 70 to 80 age group [1–3]. Its diagnosis is based on recognition of a clinical syndrome: typically, PMR patients complain of a sudden-onset bilateral pain in the shoulder and pelvic girdles limiting all self-care activities of daily living involving these parts of the body, associated with morning stiffness lasting more than 45 min. Neck ache may also be present [4, 5]. At present, no specific laboratory tests are available. High C-reactive protein (CRP) concentrations and/or erythrocyte sedimentation rate (ESR) are required criteria in the 2012 classification criteria proposed by a European Alliance of Associations for Rheumatology and American College of Rheumatology (EULAR/ACR) collaborative initiative [6].
As similar manifestations may be present in other conditions or diseases, differentiating between PMR and what looks like PMR (“PMR mimickers or PMR-like disease”) may not be a straightforward challenge in everyday clinical practice [7–9].
Calcium pyrophosphate deposition (CPPD) disease is a clinically heterogeneous condition that can present in several forms, including crowned dens syndrome (CDS) and proximal manifestations associated with raised inflammatory markers [10, 11]. Similarly to PMR, CPPD disease is common among older people, its prevalence being estimated at > 4% in the 70 to 80 age group, and very rare below 50 years of age [12]. Consequently, it is possible that PMR and CPPD may be associated by chance in the same patient.
Recently, a EULAR/ACR collaborative initiative proposed classification criteria for symptomatic CPPD disease.
The objective of this article was to discuss when CPPD disease could be considered a PMR-mimicking disease in the light of these criteria.
Material and methods
We performed a non-systematic literature search on PubMed with the following search terms: “polymyalgia rheumatica” AND “calcium pyrophosphate deposition disease” OR “chondrocalcinosis” OR “crown dens syndrome” OR “ultrasonography”, both MESH headings and free text (in each language in which they were written). Each paper’s reference list was scanned for additional publications meeting this study’s aim. When papers reported data partially presented in previous articles, we referred to the most recent published data. Abstracts submitted at conferences or from non-peer-reviewed sources were not included.
Results
In Table I, we list the main studies and reports that we found in published literature. In some studies, the percentages of CPPD among patients categorized as having PMR supported the inclusion of CPPD among the PMR-like diseases, excluding the possibility that it was mere coincidence. On the other hand, it is worth highlighting that CPPD disease was not diagnosed within 169 subjects in the non-PMR comparison group in the above-mentioned 2012 EULAR/ACR study [6].
Table I
Published studies on PMR and CPPD
| First author | Type | Number of patients | PMR and CPPD diagnosis | Length of follow-up | Imaging | Presence of definition of subset/subgroup/cluster | Significant characteristics of subset/subgroup/cluster | Suggested PMR-like condition | |
|---|---|---|---|---|---|---|---|---|---|
| Dieppe et al. [13] | Prospective monocentric | 8 PMR/CPPD over 105 CPPD | n.r. | 5 years of GC therapy | Rx | CPPD with concurrent PMR | CPPD/PMR: advanced age | Coexistence of PMR and CPPD for chance association or long-term GC therapy | |
| Aouba et al. [14] | Case series | 5 CDS PMR-like | Clinical | Max. 14 months | Rx and CT | CDS-CPPD with PMR-like presentation | Older, CC in Rx and CT, CDS in atlanto-axial CT. Responsive to NSAIDs and/or colchicine | Yes, different diagnosis and therapy between CPPD and PMR | |
| Pego- Reigosa et al. [15] | Prospective monocentric | 118 PMR – 36/118 (31%) CPPD/PMR | PMR Chuang and Healey CPPD Mc Carty | 1 year | Radiologic imaging and SFA | CPPD with concurrent PMR | CPPD/PMR: 31%, older, more frequent peripheral arthritis, more advanced knee OA, more frequent tendinous calcifications and ankle and wrist arthritis Shorter GC course and disease duration (not signif.) | Yes, shorter GC course and disease duration (not signif.) | |
| Salaffi et al. [16] | Case series | 2 PMR-like over 25 CPPD with CDS | Mc Carty for CPPD | T | CDS-CPPD with PMR-like presentation | Older, CC in Rx and CT, CDS in atlanto-axial CT | Yes, different diagnosis between CPPD and PMR | ||
| Yanai et al. [17] | Case series | 1 CPPD over 10 PMR | SFA | SFA Rx | CPPD with PMR-like presentation | Older, higher CRP, prompt response to NSAIDs | Yes, different diagnosis and therapy between CPPD and PMR | ||
| Siau et al. [18] | Case report | 1 PMR-like CPPD | Clinical | n.r. | CT | CDS-CPPD with PMR-like presentation | Higher CRP, prompt response to NSAIDs and GC | Yes, different diagnosis between CPPD and PMR | |
| Ceccato et al. [9] | Retrospective multicentric | 200 PMR – 16 (8%) had other diagnosis in follow-ups – 2 had CDS | Chuang criteria for PMR | 1 year | Radiologic imaging | CPPD with concurrent PMR | PMR/CPPD: 1%, calcifications at C1–C2, typical chondrocalcinosis, good response to NSAID | Yes, other diagnosis with different therapy and better outcome | |
| Falsetti et al. [19] | Prospective monocentric | 61 PMR at onset – 9/61 (15%) PMR/CPPD | Bird criteria | 1 year | MSK US | PMR with US diagnosis of other conditions | PMR/CPPD: 15%, more frequent females, higher frequency of knee menisci calcifications and tendinous calcaneal calcifications. Lower ESR, CRP, PLT vs. others (not signif.) | Yes, suggestion for lower dosage of GC and different course | |
| Oka et al. [20] | Case series and review | 72 CDS published cases 7 (19.4%) PMR-like | CT | CDS-CPPD with PMR-like presentation | Higher CRP (mean 12.6 mg/dl), frequent fever (80.4%), prompt response to only NSAIDs 67.5% | Yes, different diagnosis and therapy between CPPD and PMR | |||
| Manzo et al. [21] | Prospective monocentric | 134 PMR (by GP) 41 PMR 93 not PMR | Healey | 18 months | Not specified | PMR with diagnosis of other conditions | PMR/CPPD: 11% | Yes, but authors consider the possible overlap between the two diseases | |
| Ottaviani et al. [22] | Prospective monocentric | 94 Sy PMR – 52 PMR diagnosis – 25/52 (48%) PMR/CPPD | ACR/EULAR 2012 crit. and McCarty/Zhang ACR/EULAR recommend. | n.r. | US and SFA | Yes: PMR patients with concurrent diagnosis of CPPD (on US and synovial fluid analysis) | PMR/CPPD: 48%, more frequent older females, with humeral head erosions, synovitis and calcifications of AC joint. Lesser frequency of SAD bursitis | Yes, PMR/CPPD is considered other condition, suggesting shorter courses of GC | |
| Conticini et al. [23] | Retrospective multicentric | 204 PMR 31 CPPD over 104 evaluated by MSUS (22%) | Bird 2012 ACR EULAR | Range 12–60 months | MSUS and CDUS vs. only clinical | Yes: PMR with early or late diagnostic shift in other diagnosis | PMR/CPPD: 22%, patients with more frequent peripheral synovitis, lesser frequent flares, lesser dependence on GC, more frequent use of DMARDs | Yes, PMR/CPPD is considered other condition, suggesting different management. Only PMR with MSUS at onset can have a change on diagnosis | |
[i] AC – acromion-clavicular, ACR – American College of Rheumatology, CDUS – color duplex ultrasound, CPPD – calcium pyrophosphate deposition, CRP – C-reactive protein, CT – computed tomography, DMARDs – disease-modyfing antirheumatic drugs, ESR – erythrocyte sedimentation rate, EULAR – European Alliance of Associations for Rheumatology, GC – glucocorticosteroid, MSUS – musculoskeletal ultrasound, n.r. – not reported, NSAID – nonsteroidal anti-inflammatory drugs, PLT – platelet count, PMR – polymyalgia rheumatica, SAD – subacromial deltoid, US – ultrasound.
Discussion
Polymyalgia rheumatica and CPPD are two diseases of the elderly population, especially in the 70 to 80 age group. Consequently, it is possible that PMR and CPPD may be associated by chance in the same patient. Their pathogenesis is significantly different. Specifically, CPPD belongs to the group of crystal-related arthropathies, and its clinical manifestations are a consequence of activation of the NLRP3 inflammasome in response to the pathological presence of CPP crystals inside joints [10]. On the other hand, the pathogenesis of PMR is not at all linear, and some working hypotheses are awaiting confirmation [24, 25]. The evidence that PMR and CPPD occur almost exclusively in individuals aged over 50 may indicate that age-related immune modifications in genetically predisposed subjects contribute to development of these diseases [26, 27]. Proximal involvement is reported in patients with chronic pyrophosphate arthropathy.
The possibility that CPPD disease may be a PMR-mimicking disease was first discussed by Dieppe et al. [13] in a 1982 clinical and radiological study of 105 patients diagnosed with CPPD. Among 92 consecutive CPPD patients with chronic joint disease, they found 8 patients who had been treated with glucocorticosteroids (GCs) for presumed PMR. As chronic GC therapy may predispose to crystal deposition, the authors were unable to distinguish whether pyrophosphate arthropathy had occurred with polymyalgic symptoms (so-called “CPPD/PMR”) or if it was favoured by GCs.
In a 2005 prospective study, Pego-Reigosa et al. [15] involved 36 patients with CPPD and polymyalgic manifestations (5 met McCarthy’s revised criteria for definite CPPD, and 31 for probable CPPD), and pointed out that presence of tibiofemoral osteoarthritis, tendinous calcifications, and ankle arthritis should have raised suspicion for CPPD in patients with polymyalgic manifestations. For these patients, they suggested a “pseudo-PMR/CPPD” pattern, distinct from the other CPPD disease patterns previously described in the literature. According to their suggestion, CPPD must be included among the rheumatic diseases, with which PMR can be confused.
Other studies agree to propose a PMR/CPPD subset with some different characteristics compared with pure PMR: older females, lower levels of acute inflammatory markers at onset, higher frequency of peripheral arthritis, and lower exudation on bursal sites of shoulders. A possible overlap between PMR and CPPD disease was not excluded in a 2018 monocentric cohort study performed in a primary care setting [21].
More recently, Ottaviani et al. [22] found a surprisingly high percentage of CPPD disease (48.07%) in patients considered to have PMR according to the 2012 EULAR/ACR criteria for PMR. In their observational study, CPPD of the acromion-clavicular (AC) joint had the best ratio of sensitivity to specificity (sensitivity: 85.2%; specificity: 97.1%). Consequently, they proposed ultrasonographic (US) assessment of the AC joints as a more effective way to distinguish CPPD from true PMR.
Crowned dens syndrome can be a clinical-radiological manifestation for CPPD disease. It occurs with acute or subacute upper-neck pain limiting rotation of the cervical spine. Usually, fever and elevated inflammatory markers are present. The presence of calcification around the dental process of the second cervical vertebra (the “crowned dens”) on cervical computed tomography (CT) is almost pathognomonic [28, 29]. Crowned dens syndrome is estimated to represent about 5% of acute presentations of CPPD disease, and it is therefore considered rare. However, in some countries (such as Japan) it is not uncommon, being diagnosed within 1 day of its presentation during hospitalization [30].
Crowned dens syndrome clinical manifestations can mimic meningitis, infectious spondylodiscitis, or septic arthritis of the facet joint. Is CDS a potential PMR mimicker? Painful stiffness of the pelvic girdle is always absent in CDS patients, suggesting that this is a discriminating manifestation in clinical practice. According to the 2012 EULAR/ACR classification criteria, patients aged 50 years or older presenting with bilateral shoulder pain and elevated CRP and/or ESR can be classified as having PMR in the presence of morning stiffness > 45 min and new hip pain, in the absence of peripheral synovitis or positive rheumatoid arthritis (RA) serology [6]. Therefore, the possibility that CDS may be a PMR-like disease seems more like speculative reasoning than actual diagnostic uncertainty.
Additionally, in their 2004 case series, Aouba et al. [14] reported on 4 patients first diagnosed with PMR whose final diagnosis was CDS. In these patients, GCs were ineffective, and a fast improvement followed treatment with nonsteroidal anti-inflammatory drugs (NSAIDs) and colchicine.
Finally, an interesting aspect should be underlined regarding the prognosis and clinical course of patients with PMR and CPPD. In fact, in all the studies on this topic, no patient with PMR/CPPD has had a late diagnosis of GCA (at least within the term of follow-up of each study) [15, 23]. On the other hand, patients with pure PMR not infrequently experience late GCA (16–47% within 2–24 months of PMR diagnosis). This remarkable divergence in clinical course raises the question of whether PMR/CPPD may constitute a peculiar non-vasculitic subset.
Ultrasonographic assessment is included as an option in the 2012 classification criteria for PMR, where US criteria (Table II) increased the specificity (from 81.5 to 91.3%). Additionally, US serves as a valuable tool for excluding alternative diagnoses such as non-RA shoulder conditions and subjects without shoulder conditions. Ultrasonography has proved helpful in the diagnosis of CPPD [31]. In 2023, an international multidisciplinary working group developed a framework for diagnosing CPPD based on imaging modalities, including US [32]. Typical US findings were considered as follows: 1) crystal deposition must be found in the fibrocartilage or hyaline cartilage to qualify as CPP deposition; 2) crystal deposition must be described as hyperechoic deposits with variable shape and size and not creating posterior shadowing; 3) hyperechoic deposits may also be visualized on US in the synovial membrane, joint capsule, or tendons, and 4) in dynamic scanning, the deposits situated within hyaline cartilages move solidly in a synchronous way and in the same direction of joint movement, whereas CPP deposits of the synovial membrane and joint capsule move in opposition to joint movement, as cartilage slides under the capsule and synovial membrane. Basic calcium phosphate (BCP) crystals – mostly partially carbonate substituted hydroxyapatite crystals – are associated with BCP deposition (BCPD) disease, in tendons (calcific tendinopathy of the shoulder and hip), bursae, or joints, where they can lead to a severe destructive arthropathy associated with osteoarthritis (mostly in large joints such as the shoulders, where the pathology is called Milwaukee shoulder syndrome) [33]. In BCP arthropathy, a recent review concluded that there are limited data on the utility of US in differentiating the crystals of BCP and CPP [34], whereas in the 2023 EULAR recommendations on imaging for diagnosis of crystal-induced arthropathies, Recommendation 5 states that in the diagnostic assessment of BCPD, imaging is necessary, and conventional radiology or US is the recommended modality [35]. In any case, a risk of misidentification should always be taken into account in clinical practice for some consequences: as an example, the response to local intra-articular GCS may be impressive in patients with CPPD disease, whereas it should be used with caution in patients with BCD disease because of the risk of disrupting the hydroxyapatite deposit or seeding further calcification.
Finally, new classification criteria for CPPD disease have been recently proposed. In the 1960s, Ryan and McCarty first proposed diagnostic criteria for CPPD disease. According to these criteria, the definite diagnosis was when CPP crystals were found (presence of both typical linear or punctate calcification on radiography and findings consistent with CPP crystals on synovial fluid polarised light microscopy) [36].
In 2023, a EULAR/ACR collaborative initiative proposed classification criteria for symptomatic CPPD disease [37]. These criteria would apply to CPPD disease as a whole, which meant that separate classification criteria for each clinical presentation were not within the aim of this initiative. The identification of CPP crystals in synovial fluid from a symptomatic joint was sufficient for classification as CPPD as long as the exclusion criteria were not met (e.g. another condition did not explain the entire presentation). Crowned dens syndrome was another sufficient criterion. Attribution of symptoms to CPPD disease can be challenging, particularly in patients with coexisting conditions. Those patients for whom all symptoms are better explained by another coexisting condition cannot be categorized as having CPPD disease.
In the absence of the aforementioned two sufficient criteria, patients could be categorized by scoring some remaining imaging and clinical criteria through 8 domains and several levels: the face validity of a threshold score of > 56 was assessed. In other words, according to the 2023 EULAR/ACR criteria, a patient without sufficient criteria may be equally classified as having CPPD disease if the sum of points according to the criteria presented in Table III is > 56 points.
Table III
Domains, levels and points in the 2023 EULAR/ACR classification criteria for CPPD disease [37]
[i] CMCJ – carpometacarpal joint, CPPD – calcium pyrophosphate disease, CR – conventional radiography, CT – computed tomography, DECT – dual-energy computed tomography, MCPJ – metacarpophalangeal joint, MTPJ – metatarsophalangeal joint, OA – osteoarthritis, STTJ – scaphotrapeziotrapezoid joint, US – ultrasound.
Among the 148 definite mimickers forming the derivation cohort, 6 were affected by PMR. The classification criteria demonstrated high sensitivity and specificity in an independent validation cohort. Among the 162 definite mimickers forming the validation cohort, only 1 was affected by PMR [37].
Given the above data, when should we consider CPPD disease as a differential diagnosis in a patient with polymyalgic manifestations categorized as having PMR according to the 2012 EULAR/ACR criteria? Schematically, there are 3 possible scenarios that may occur in clinical practice.
Polymyalgic manifestations categorized as PMR in a patient with already diagnosed CPPD disease: PMR/CPPD or pseudo-PMR CPPD pattern.
Proximal involvement is reported as common in patients with chronic pyrophosphate arthropathy. In these patients, US assessment may be useful for differentiating true PMR from chronic CPPD disease with polymyalgic manifestations. Additionally, peripheral joints are commonly involved in chronic CPPD disease. In contrast, absence of peripheral synovitis is one of the 2012 classification criteria for PMR. In these patients with pseudo-PMR CPPD pattern, oral GCs can be used. However, GCs rarely give the rapid improvement that the clinician usually observes in a patient with true, isolated PMR. NSAIDs associated with colchicine are conversely more effective [38].
A previous diagnosis of CPPD disease does not exclude a true PMR: PMR with concurrent symptomatic CPPD disease. In this case, the therapeutic recommendations for PMR should be used. However, CPPD disease must also be treated if widespread and massive, to avoid its complications [39].
Polymyalgic manifestations categorized as PMR in patients with undiagnosed chronic CPPD disease.
In this case, the critical point is whether the polymyalgic manifestations can be explained by PMR or by CPPD disease. Ultrasonography assessment may be crucial for a correct diagnosis, and consequently for a proper therapeutic approach.
Conclusions
The possibility that CPPD disease and PMR may be present in the same patient is not negligible, especially if the patient is aged over 70 years. On the other hand, CPPD disease is often included in the list of common PMR-mimicking diseases, because CPPD patients with marked proximal involvement and elevated acute inflammatory markers may be categorized as having PMR.
The relationship between PMR and CPPD disease should be reviewed in the light of the 2023 EULAR/ACR classification criteria for symptomatic CPPD disease. Specifically, presence of CPP crystals in synovial fluid and/or a score > 56 points (using additional weighted criteria taking into account clinical features, associated metabolic disorders, and results of laboratory and imaging investigations) can classify patients as having CPPD disease with polymyalgic manifestations. In these patients, a diagnosis of PMR-like CPPD is therefore possible. On the other hand, when these criteria for symptomatic CPPD are absent (no CPP crystals and/or score < 56 points), CPPD should be considered a coincidental finding in patients with PMR (that is, PMR without symptomatic CPPD).
We look forward to future prospective studies applying the recent EULAR/ACR criteria, in order to better classify patients with polymyalgic manifestations and CPPD.
Likewise, further studies are required to confirm the possibility that PMR/CPPD may be a non-vasculitic pattern of disease.
