Review: Outcome Measures in CPPD

Abstract

This review highlights outcomes for patients with calcium pyrophosphate deposition (CPPD) reported in prior studies and underscores challenges to assessing outcomes of this condition. Prior clinical studies of interventions for CPPD focused on joint damage and calcification on imaging tests, joint pain and swelling, and inflammatory biomarkers. Qualitative interviews with patients with CPPD and healthcare providers additionally identified flares, overall function, and use of analgesic medications as important outcomes. Imaging evidence of joint damage and calcification are likely to be outcomes in future clinical studies of CPPD, though reliability and sensitivity to change in CPPD require further testing for several imaging modalities. Challenges to outcome measurement in CPPD include questions of attribution of signs and symptoms to CPPD versus co-existing forms of arthritis, lack of therapies to prevent or dissolve calcium pyrophosphate crystal deposition, absence of validated patient- or physician-reported CPPD outcome measures, and scarcity of large cohorts in which to study outcomes of different clinical presentations of CPPD.

Introduction

Calcium pyrophosphate deposition (CPPD) disease is one of the most common forms of inflammatory arthritis with an estimated prevalence of between 4-7% in population studies (1-3). First described in 1962 by two independent groups (4, 5), the condition results from the deposition of calcium pyrophosphate (CPP) crystals, leading to multiple different clinical presentations. The most widely recognized manifestation is a flare of acute inflammatory arthritis that shares some similarities to gout, which led to the term “pseudogout”. Another common clinical manifestation is osteoarthritis in a joint with cartilage calcification (historically called “chondrocalcinosis”); while the knee is a common site of osteoarthritis with cartilage calcification, this sometimes occurs in joints for which primary osteoarthritis is uncommon, such as the shoulder, and can be associated with prominent subchondral cysts. A smaller portion of patients with CPPD have chronic inflammatory arthritis that may present similarly to rheumatoid arthritis (RA), which led to the term “pseudo-RA”. Because CPPD disease commonly co-exists with other types of arthritis – including OA, RA, and gout – attribution of symptoms is particularly challenging, as described by a patient: “I've never known which part of my problem is the pseudogout and which part is the osteoarthritis. I've always assumed that the pain was just the osteoarthritis, and the swelling was the pseudogout, but I have no idea how to tease the two apart”.

Clinicians often use the terms CPPD, pseudogout, and chondrocalcinosis interchangeably to refer to clinical or radiographic manifestations of CPPD. Compounding this lack of clarity, the International Classification of Diseases (ICD) diagnostic coding system contains codes for “chondrocalcinosis” (a radiographic finding) but not for “pseudogout” (a clinical manifestation) (6).

Noting challenges with nomenclature used to describe this condition, the European Alliance of Associations for Rheumatology (EULAR) proposed recommended terminology in 2011. Consistent with these recommendations, as we defined above, CPPD stands for calcium pyrophosphate (CPP) deposition. Manifestations of CPPD include:

• acute calcium pyrophosphate (CPP) crystal arthritis (also known as “pseudogout”)

• chronic CPP crystal inflammatory arthritis

• osteoarthritis (OA) with CPPD (CPPD + OA): “CPPD in a joint that also shows changes of OA, on imaging or histological examination” (7)

• asymptomatic cartilage calcification on imaging (traditionally termed “chondrocalcinosis” and detected on conventional radiograph or ultrasound) or histological examination (7).

We use the term “CPPD disease” to indicate clinical manifestations of CPPD, which frequently occur together with radiographic manifestations of CPPD, as distinguished from radiographic manifestations without clinical symptoms (e.g. asymptomatic cartilage calcification on imaging).

Although clinical manifestations of CPPD are common, very few clinical studies of this disease have been published relative to other forms of arthritis. As an example, a search of articles indexed in PubMed from 2011-2021 revealed 8377 articles with the search term “gout” and only 285 articles with the search term “CPPD”. Accordingly, basic questions about CPPD pathogenesis, natural history, and long-term extra-articular consequences remain unanswered.

Another challenge to studying CPPD disease is the lack of validated clinical outcome measures for this condition. The Outcome Measures in Rheumatology (OMERACT) CPPD Ultrasound Subtask Force, established in 2014, has developed and validated definitions for ultrasonographic findings of CPPD (8-10). The OMERACT CPPD Working Group, established in 2018, is in the process of establishing a set of core outcome domains to be measured in CPPD disease clinical trials, following established methodology (11). The OMERACT CPPD Working Group recently performed a scoping review of outcomes reported in previous CPPD studies (12) and a qualitative interview study to generate outcomes relevant to people with CPPD as well as better understand the experience of people living with this condition (13, 14).

Defining relevant clinical outcomes in CPPD can be difficult due to this condition having variable clinical presentations including single or recurrent flares of acute CPP crystal arthritis, often with long asymptomatic periods, chronic inflammatory arthritis, and/or CPPD + OA (Figure 1) (15). Typical joints affected include the knee and wrist, but involvement of other peripheral joints and the axial skeleton has also been described (16). CPPD disease is strongly associated with OA, which in some patients is characterised by an atypical distribution of joint involvement, large subchondral cysts, and prominent osteophytes compared to OA without CPPD.

Figure 1.

CPPD clinical presentations and manifestations (reprinted from “Towards development of core domain sets for short term and long term studies of calcium pyrophosphate crystal deposition (CPPD) disease: A framework paper by the OMERACT CPPD working group” by Cai et al., 2021, Semin Arthritis Rheum., with permission from Elsevier). The size of each circle approximates the relative prevalence of each manifestation.

The variability in clinical presentations along with lack of uniformity of CPPD definitions and subsequent inaccuracies in clinical coding have led to difficulties establishing large CPPD cohorts for studies (6). Additionally, there have been very few clinical trials including patients with CPPD disease as an explicit condition. Future studies may evaluate novel anti-inflammatory agents for treating acute CPP arthritis flares as well as long-term therapies, such as disease-modifying and/or anti-inflammatory treatments to prevent flares. Disease-modifying therapies that inhibit or reverse CPP crystal formation and deposition could have profound clinical impact, as targeted treatments for CPPD do not currently exist. Clinical trials testing these agents would be greatly facilitated by agreement on relevant imaging outcomes and clinical outcome measures and instruments for measurement. Due to the heterogeneity of this condition, the OMERACT CPPD Working Group has proposed to develop outcome measures for “long term” studies of CPPD disease that may include participants with any clinical manifestation of CPPD (acute CPP crystal arthritis, chronic CPP crystal inflammatory arthritis, and/or CPPD + OA), and for “short term” studies for an individual flare of acute CPP crystal arthritis (15).

This review aims to describe the outcome measures used in previous CPPD studies, discuss the challenges in developing outcome measures for clinical or radiographic manifestations of CPPD, and highlight important areas for future research.

Outcome Measures in CPPD: Previous Studies

Observational clinical studies and clinical trials

Currently, there are no validated clinical outcome measures for CPPD. Recently, the OMERACT CPPD Working Group performed a scoping review of 112 clinical studies of CPPD, which included studies of CPPD disease and asymptomatic cartilage calcification on imaging (12). Outcome measures identified from these studies were mapped to domains and core areas of the OMERACT Filter 2.1: (i) Manifestations/Abnormalities, (ii) Life Impact, (iii) Longevity, and (iv) Societal/Resource Use (17). There were 22 outcome domains, which were predominantly in the area of disease Manifestations/Abnormalities and Life Impact (Table 1). The majority of previous studies have focused primarily on imaging outcomes. The most frequently reported outcomes within the area of Manifestations/Abnormalities included joint damage on imaging tests (59 studies), joint calcification on imaging tests (28 studies), joint pain (26 studies), joint swelling (12 studies), soluble biomarkers of inflammation (12 studies) and related medical conditions such as osteoarthritis (9 studies). Within the category of Life Impact, the most frequently reported outcomes were response to treatment (in 23 studies, outcome measures detailed below) and overall function (in 14 studies, outcome measures also detailed below) (12). There were 15 studies that included outcomes related to side effects of treatments. Very few studies reported outcomes related to the core areas of Longevity (only two studies) or Societal/Resource Use. When the included studies were analysed by predominant CPPD clinical presentation, there was substantial heterogeneity in the predominant CPPD clinical presentation (the predominant CPPD clinical presentation was unclear in 21 studies) and study sample size (range from 2 to 25,157 subjects) as well as preponderance of previous research primarily in acute CPP crystal arthritis and CPPD + OA (47 studies of CPPD + OA, and 24 studies of acute CPP crystal arthritis). The majority of CPPD + OA studies comprised subjects with symptomatic knee OA and cartilage calcification on conventional radiography.

Table 1.

The most frequently reported clinical outcomes and instruments from a scoping review of 112 CPPD studies categorised by Manifestations/Abnormalities, Life Impact, Longevity and Societal/Resource Use (adapted from “Outcome domains reported in calcium pyrophosphate deposition studies: A scoping review by the OMERACT CPPD working group” by Cai et al. (2020), Semin Arthritis Rheum., 50(4):719-27, with permission from Elsevier).

OMERACT Core Area	Outcomes	Instruments
Manifestations/Abnormalities	Joint damage on imaging tests [59]	Cartilage loss on plain radiography [50], ultrasound [4], computed tomography [1], MRI [2] and arthroscopy [2].
	Joint calcification on imaging tests [28]	Joint calcification on plain radiography [22], ultrasound [4], computed tomography [1] and MRI [2]
	Joint pain [26]	Dichotomous scale (present/absent) [9] or visual analogue scale (0-10cm) [7]
	Joint swelling [12]	Dichotomous scale (present/absent) [9]
	Soluble biomarkers of inflammation [12]	ESR or CRP [10]
Life Impact	Response to treatment [23]	Dichotomous scale (response/no response) [11] or trichotomous scale (full response/partial response/no response) [6]
	Overall function [14]	Patient reported disability scale (0-5 for persona care and indoor/outdoor mobility) [3]
Longevity	Survival [2]	Survival at 10 years [1] or number of deaths [1]
Resource/Societal Use	Need for joint surgery [7]	Dichotomous scale (surgery/no surgery) [7]

Note: [n] = number of studies with the reported outcome or instrument. Total of 112 studies included in scoping review.

While not the focus of the paper, the authors also identified substantial heterogeneity in the instruments used to measure these reported clinical outcomes (12). Joint pain, which was reported in 26 studies, was measured most frequently on a dichotomous (present/absent; 9 studies) or visual analogue scale (0-10cm; 7 studies) with a small number of other pain instruments. Joint swelling, which was reported in 12 studies, was measured most frequently on a dichotomous scale (present/absent; 9 studies). Response to treatment, which was reported in 23 studies, was measured most frequently using a dichotomous scale (response/no response; 11 studies), trichotomous scale (full response/partial response/no response; 6 studies) or time to major improvement in symptoms (5 studies). Overall function, which was reported in 14 studies, was measured using a variety of methods, the most frequent was which a patient reported disability scale (0-5 scale for personal care, indoor/outdoor mobility; 3 studies).

Qualitative Studies

Prior to the publication of an interview study of patients, caregivers, healthcare professionals and stakeholders (13, 14), we are not aware of any other previous qualitative studies of CPPD. This study was undertaken to generate outcome domains as part of the OMERACT methodology for the development of Core Domain Sets (11). A total of 36 interviews were conducted with 28 patients with CPPD disease (six accompanied by a care giver), seven health care professionals and one stakeholder to identify outcome domains that were most relevant to patients (13). These domains were also mapped to core areas of: (i) Manifestations/Abnormalities, (ii) Life Impact, (iii) Longevity, and (iv) Societal/Resource Use (Table 2). The most frequently reported outcomes within the area of Manifestations/Abnormalities included joint pain (35 interviews), joint swelling (27 interviews), joint stiffness (25 interviews) and CPPD flares (25 interviews). Within the category of Life Impact, the most frequently reported outcome was overall function (35 interviews), highlighted by the domain of ability to complete daily tasks (25 interviews). Within the category of Societal/Resource Use, use of analgesic medicines was also reported in 26 interviews. These interviews highlighted the differences between relevant outcomes for patients with CPPD disease compared to what has been reported in previous CPPD studies, particularly with respect to the importance of joint pain, flares, overall function and use of analgesic medicines.

Table 2.

The most frequently generated outcomes from two qualitative studies of patients with CPPD (adapted from “Outcome domains reported by patients, caregivers, healthcare professionals and stakeholders for calcium pyrophosphate deposition (CPPD): A content analysis based on semi-structured qualitative interviews from the OMERACT CPPD working group” by Fuller et al. (2021), Semin Arthritis Rheum.,51(3):650-4 and “Experience and impact of crystal pyrophosphate deposition (CPPD) from a patient and caregiver perspective: A qualitative exploration from the OMERACT CPPD working group”, by Fuller et al. (2021), Semin Arthritis Rheum., with permission from Elsevier)

OMERACT Core Area	Outcomes	Personal Quotes
Manifestations/Abnormalities	Joint pain [35]	“I just woke up, and I couldn’t really walk properly… It was stabbing pain. It was like nothing I’ve ever really experienced… it was out of nowhere”
	Joint swelling [27]	“When it affects my ankle, my foot, I have to wear open shoes. I can ’t stand normal shoes. My foot swells, so I can’t use them”
	Joint stiffness [25]
	CPPD flares [25]	“When flares happen, then I get very moody because I can’t do what I had planned with my family, or my job, or hobbies, whatever. Flares ruin every plan. I can’t walk, so I am not able to enjoy anything. That is very annoying”
Life Impact	Overall function [35]	“When flares happen, then I get very moody because I can’t do what I had planned with my family, or my job, or hobbies, whatever. Flares ruin every plan. I can’t walk, so I am not able to enjoy anything. That is very annoying”
	Ability to complete daily tasks [25]	“She can’t wash herself, she can’t cook, etc. It has an impact on her surroundings”
Longevity	No outcome domains reported
Resource/Societal Use	Use of analgesic medicines [26]	“The use of analgesics, well, it is a pain to have to keep on taking them.”

Note: [n] = number of interviewees with the reported outcome. Total of 36 people interviewed, of which 28 were patients with CPPD.

The lived experience of patients with CPPD disease and their caregivers has been analyzed in more depth (14). Of the 28 patients with CPPD disease interviewed, 19 patients (68%) had more than one clinical presentation of CPPD: 21 patients had acute CPP crystal arthritis, 21 had CPPD + OA and 9 had chronic CPP crystal inflammatory arthritis. This highlights the challenges for patients and clinicians in defining this condition and differentiating whether their joint pain may be a manifestation of CPPD disease or another cause (“At the beginning the diagnosis was different. They used to tell it was one thing, then a different one… When the x-rays showed calcium lines, they started to identify the problem. But it seems to be poorly known. Not all doctors knew about it, it depended who, they said one thing or a different”). CPPD disease was found to impact patients in a multitude of ways. There were physical limitations ([i]“I just woke up, and I couldn’t really walk properly… It was stabbing pain. It was like nothing I’ve ever really experienced… it was out of nowhere”, [ii] “She can’t wash herself, she can’t cook, etc. It has an impact on her surroundings”, [iii] “This impacts also my relatives’ life. They have to plan everything (holidays, work transfers etc.) so that someone is always near to me, in case of need”) but also psychological limitations, which were just as impactful ([i]“When flares happen, then I get very moody because I can’t do what I had planned with my family, or my job, or hobbies, whatever. Flares ruin every plan. I can’t walk, so I am not able to enjoy anything. That is very annoying.”, [ii] “When you are in pain all the time you are worn down by it and it is complicated to manage. I feel diminished by it because sometimes I am not able to be independent and so psychologically it is upsetting.”). The interviews also showed the challenges patients experienced in establishing a diagnosis (“They used to tell it was one thing, then a different one… When the x-rays showed calcium lines, they started to identify the problem. But it seems to be poorly known. Not all doctors knew about it, it depended who, they said one thing or a different. Three or four years ago they defined clearly as chondrocalcinosis”) and uncertainty about their prognosis (“To me there isn’t a plan here and saying, you know this is, this is what we will follow up, or we will do. That sort of, that’s what I can’t see there”).

Considerations on Using Imaging Outcomes for CPPD Research

The aforementioned scoping review identified imaging outcomes as the most commonly measured outcome in previous CPPD studies (12). Imaging outcomes included joint damage on imaging tests or joint calcification on imaging tests, the majority of which were related to findings on plain radiography (50 of 59 studies for joint damage on imaging tests; 22 of 28 studies for joint calcification on imaging tests). Key imaging features of CPP deposition include characteristic calcific deposits in and around joint structures. Evidence of CPP deposition on conventional radiograph (x-ray) has historically been considered fundamental to CPPD diagnosis and is required in Ryan and McCarty’s proposed diagnostic criteria for CPPD (18).

In the past several decades, ultrasound has emerged as an accurate imaging modality for identifying CPP deposits, with superior sensitivity and slightly less specificity compared to conventional radiograph (19). Work performed by the OMERACT CPPD Ultrasound Subtask Force has been critical for developing validated ultrasound definitions for CPPD, and define characteristic features of deposits in fibrocartilage, hyaline cartilage, tendon, and synovial fluid (8-10). Ultrasound and conventional radiograph are currently widely used in clinical care of CPPD disease, but cannot definitively distinguish between types of calcium crystals - though the location and morphology of calcification may hint at crystal composition. For example, on conventional radiograph, CPP deposits are generally fine linear or punctate, as distinct from basic calcium phosphate (BCP) deposits that are generally larger and homogenous. On ultrasound, dense deposits within hyaline cartilage (“pseudo-double contour sign”) are characteristic of CPP deposits, in contrast to deposits atop hyaline cartilage (“double contour sign”) characteristic of monosodium urate crystals.

Sensitivity and specificity of conventional computed tomography (CT) have not been formally studied with regard to performance for identifying CPP deposits, but this modality can identify characteristic calcifications (20-22). Dual-energy CT (DECT) and multi-energy spectral photon-counting CT are emerging techniques to distinguish between CPP and BCP deposits, but are not widely used at present to evaluate calcific deposits (20, 23). We direct readers to several excellent reviews summarizing the relevant literature on the performance of conventional radiograph, ultrasound, and DECT in CPPD (19, 24, 25).

Table 3 summarizes the sensitivity and specificity of imaging modalities for identifying CPP deposits (19, 22, 24, 26-29). It is worth noting that many of these studies employed different definitions of a positive imaging study, used different reference standards (e.g. synovial fluid crystal analysis, tissue biopsy with crystal analysis, or Ryan and McCarty’s diagnostic criteria), and imaged different joints (e.g. knee, wrist, hip) and joint structures (e.g. fibrocartilage, hyaline cartilage). Generally, the knee was the most commonly imaged joint (wrist second most common), and synovial fluid crystal analysis was the most common reference standard for studies of the knee while Ryan and McCarty’s diagnostic criteria was the most common reference standard for the wrist. While synovial fluid crystal analysis is generally considered the gold standard for CPPD diagnosis, its sub-optimal sensitivity is a recognized limitation (30, 31). EULAR recommends synovial fluid crystal analysis and/or (less commonly) identification of CPP crystals on tissue biopsy as the gold standard (7).

Table 3.

Sensitivity and specificity of imaging modalities for CPP deposition

Imaging Modality	Sensitivity	Specificity
Conventional radiograph (19, 24)*	Pooled sensitivity 47 to 60%	Pooled specificity 95 to 96%
Ultrasound (19, 26-28)*	Pooled sensitivity 81 to 88%	Pooled specificity 87 to 92%
Dual-energy CT (22, 29)	78% to 100% (reported in two studies)	94% (reported in one study)

^*Most studies included in meta-analyses evaluated the knee; fewer studies evaluated the wrist, shoulder, or other anatomic sites

The utility of these imaging modalities as outcome measures for CPPD is a separate issue that has not been well-studied. Several features of radiographic outcome measures for musculoskeletal diseases are summarized in Table 4 (32-36). Radiographic outcome measures reported in clinical research should be valid, feasible, and have discriminative ability (37). Reliability and sensitivity to change represent two key features of discrimination worth discussing for CPPD.

Table 4.

Examples of radiographic outcome measures in musculoskeletal diseases and hypothetical application to CPPD

	Rheumatoid arthritis	Gout	CPPD (hypothetical)
Imaging Modalities	Conventional radiograph (33)	Conventional radiograph (35), ultrasound (34, 36), DECT (32)	Conventional radiograph, ultrasound, conventional CT, DECT
Radiographic Abnormalities	Erosions, osteoporosis, soft tissue swelling, joint space narrowing, subluxation	Tophi, erosion, joint space narrowing, double-contour sign, aggregates, volume of MSU deposits	Calcification with particular characteristics (e.g. linear, punctate) in particular locations (e.g. fibro/hyaline cartilage), subchondral cysts, patellofemoral joint space narrowing
Scoring Method	Continuous count of abnormalities or semi-quantitative scoring	Sum of erosion and joint space narrowing scores	Continuous count of joints with characteristic calcifications (e.g. right knee, left knee, right wrist); sum of “calcification score” in various joints; joint space narrowing
Examples of Joints Included	PIP, MCP, wrist, MTP	1st MTP, knee	Knee, wrist, 2nd and 3rd MCP

Reliability

Reliability of radiographic scoring for CPPD between readers (inter-reader) and within readers (intra-reader) will be critical to establish. For CPPD, ultrasound is likely to play an important role as an outcome measure given its excellent sensitivity and high specificity for CPP deposits (19). Reliability of the OMERACT CPPD ultrasound definitions, which specify the shape, echogenicity, location, and behavior in dynamic scanning of calcific deposits in the fibrocartilage, hyaline cartilage, tendon, and synovial fluid, was recently assessed. Reliability of the OMERACT CPPD ultrasound definitions was high, with good inter-reader reliability at the triangular fibrocartilage of the wrists (kappa 0.82), hyaline cartilage of the knees (kappa 0.58), and fibrocartilage of the knees, i.e. meniscus (kappa 0.65), as well as high intra-reader reliability at these sites (9). Standardized radiographic outcome definitions for conventional radiograph and DECT have not been established for CPPD and will require development and testing, akin to previous validation studies for ultrasound by the OMERACT CPPD Ultrasound Subtask Force (9).

Sensitivity to Change

Sensitivity to change refers to “whether a radiographic scoring method can detect a real change over time”; this may be represented by the standardized response mean, minimal detectable change, or G coefficient (33). In CPPD, a key question is whether and to what degree radiographic evidence of cartilage calcification might change following an intervention. Change in cartilage calcification might be measured in various ways, including binary scoring (present/absent), semi-quantitative scoring, or percentage change from baseline.

Approaches used to report radiographic features in gout, another crystal deposition disease, may be relevant for CPPD imaging outcomes. One gout study assessed presence vs. absence of ultrasound features (double-contour sign, tophus, aggregates, and erosion) in multiple joints and summed these on a patient level before and during urate-lowering therapy (34). Double-contour sign and tophus were sensitive to change after 3 months of urate-lowering therapy (34). In the case of CPPD, prior studies have not demonstrated whether cartilage calcification can be eradicated. It is therefore unclear whether a binary measure of present vs. absent would be reasonable for chondrocalcinosis. In gout, a semi-quantitative DECT urate scoring method based on relative amounts of urate deposits in four anatomic regions demonstrated better discrimination than urate volume on DECT (32). This type of semi-quantitative measure may be well-suited for CPPD, in which a gradient of response might be expected after an intervention, such as a percentage decrease in the quantity of CPP deposits in cartilage, tendon, or synovial capsule.

Consideration of Joints, Joint Structures, Imaging Techniques, and Patient Characteristics

Radiographic outcome measures in CPPD might demonstrate variable reliability and sensitivity to change between different joints and in different joint structures. Studies of ultrasound detection of CPP deposits report varying reliability in the knee and wrist, and in fibrocartilage compared to hyaline cartilage (9). Some of these differences may have to do with joint size and the volume of visible cartilage; consider the large amount of cartilage visible at the knee, versus the relatively small amount visible at the metacarpophalangeal joints. Other differences may result from constraints of the imaging modality itself. Conventional radiograph can measure two-dimensional area of chondrocalcinosis; ultrasound can measure the two-dimensional area of hyperechoic deposits in the femoral trochlear cartilage of the knee in maximum flexion; and DECT can measure three-dimensional volume of CPP deposits. Cartilage calcification of the knee may be easily visible on conventional radiograph in patients with preserved joint space, but difficult to visualize in knees with advanced osteoarthritis and severe cartilage loss. The type of radiographic projection, such as AP vs PA techniques for conventional radiograph, may also affect the performance of radiographic outcome measures.

Outcome Measures in CPPD: Challenges

Attribution

Attribution of signs and symptoms presents a problem for CPPD, as CPPD frequently co-exists with other forms of arthritis. OA and RA are most notable, though CPPD can also co-exist with gout; approximately 5% of synovial fluid samples with monosodium urate crystals also have CPP crystals on polarized light microscopy (38). When CPPD co-exists with other forms of arthritis, the question of attributing a sign or symptom to CPPD arises. OA poses a particular challenge regarding attribution of symptoms. The knee is a common site of OA and is the most common site for CPPD, making it difficult to differentiate the cause of knee symptoms in a patient with both (7). While OA can occur in atypical joints (i.e. radiocarpal joints and elbows) and has been thought to associate with prominent osteophytes and subchondral cysts, these findings are not always present in joints with cartilage calcification and OA (CPPD + OA). Joint pain, joint swelling, fatigue, and functional limitation provide examples of signs and symptoms that could be due to CPPD primarily, or OA primarily, or might be the product of both. The concept of effect modification comes to mind, namely, whether joint pain is worse in a patient with concurrent osteoarthritis and CPPD compared to a patient with just one of those conditions. This is a challenging concept to study in CPPD, as pinpointing the time of onset of cartilage calcification is difficult to prove and disease-modifying therapies are lacking.

Lack of Targeted Therapies

Lack of targeted therapies for CPPD poses another challenge to studying outcomes in CPPD. Presently, targeted therapies are lacking to either prevent formation of CPP crystals or dissolve these crystals once they have formed. Common treatments for CPPD include colchicine, joint aspiration, intra-articular glucocorticoids, and non-steroidal anti-inflammatory drugs (NSAIDs) (39). Acute CPP crystal arthritis is commonly treated with those modalities as well as oral glucocorticoids if needed. These therapies improve joint pain, joint swelling, and joint function in many forms of arthritis including OA, RA, and gout. Improvement in symptoms following use of these therapies is thus non-specific for CPPD and may point to treatment of pain from another type of inflammatory or non-inflammatory arthritis.

Absence of Validated Outcome Measures for CPPD

Whilst disease-specific measures have been developed and validated in gout [Gout Assessment Questionnaire 2.0, Tophus Impact Questionnaire, and Gout Attack Intensity Score (40, 41), Gout Attack Intensity Score (42)], RA [DAS-28 (43), CDAI (44), RADAI (45)], and OA [WOMAC (46), KOOS (47)], validated physician-reported and patient-reported instruments to measure outcomes in CPPD have not been developed to date. This presents another challenge to clinical research studies measuring outcomes in CPPD, of which considerations for radiographic outcome measures have been discussed previously.

Development and validation of outcome measures requires rigorous testing of psychometric properties including scalability, reliability, and validity. Until these instruments are developed for CPPD disease, it will be tempting to adapt or apply instruments to CPPD. Pain visual analogue scale (VAS), ranging from 0 (no pain) to 100 (unbearable pain), might be reasonable to employ in a CPPD clinical study despite lack of testing specifically in CPPD. As noted in Table 1, 23% of studies on interventions for CPPD disease reported on joint pain using scales that have not been tested in CPPD. OA instruments might be included as endpoints in some CPPD clinical studies if they require OA as an inclusion criterion and/or are interested in studying potential effect modification of CPP deposits on OA outcomes. However, distinct radiographic metrics that can capture features attributable to CPPD are needed. These might include measurement or number of subchondral cysts, or volume or location of CPP deposits on imaging modalities.

The definition of a flare of acute CPP crystal arthritis has not been formally established. Presence of a flares may be an inclusion criterion for some clinical trials, as well as an important outcome of various clinical studies. Once a definition of a flare has been established, tools to measure flare recurrence, duration, and severity will be needed.

In gout, tophus regression is an important outcome measure (48). While a number of physical examination and imaging methods have been proposed for measuring tophus in gout, physical measurement using Vernier calipers meet most of the OMERACT filter components of feasibility, truth, and discrimination (19, 20)(49). In CPPD, tophaceous CPPD crystal deposition is rare and is unlikely to be a core outcome domain. Other physical examination findings in CPPD may overlap with osteoarthritis physical examination features, again raising the challenge of attribution.

Osteoarthritis-Related Outcomes

Given the prevalence of CPPD, the potential impact of CPPD on OA presents an important question with public health implications and will likely be an important outcome measure in future clinical trials. To date, our understanding of the possible effect of CPPD on progression of OA is limited to studies focused on the knee. Several prospective observational studies compared subjects with vs. without cartilage calcification on a plain radiography of the knee at baseline, and reported mixed results regarding OA-related outcomes in the presence vs. absence of cartilage calcification. All subjects in these studies had knee symptoms; asymptomatic cartilage calcification (chondrocalcinosis) was not assessed. The Knee & Hip OA Long-term Assessment cohort (KHOALA) reported no association between baseline chondrocalcinosis and risk of total knee replacement (HR 1.3, 95% CI: 0.7-2.2) (50). There was also no association between WOMAC pain (OR 1.1, 95% CI: 0.7-1.4) or WOMAC function (OR 0.9, 95% CI: 0.4-2.0) scores. The Boston OA Knee Study (BOKS) reported a lower risk of progressive cartilage loss on MRI in subjects with vs. without cartilage calcification (RR 0.4, 95% CI: 0.2-0.7) (51). Two cohort studies identified no association between baseline cartilage calcification and progressive cartilage loss on MRI (Health ABC: RR 0.9, 95% CI: 0.6-1.5; KHOALA: OR 0.9, 95% CI: 0.4-1.7) (50, 51).

Several challenging aspects of study design are worth considering. First, given the aforementioned concerns about suboptimal sensitivity of plain radiography for CPP deposits, it is possible that these studies misclassified CPPD as non-CPPD (false-negative for cartilage calcification) which would be expected to move the effect estimate toward the null. Second, cartilage calcification was a prevalent exposure with an unknown duration, rather than a new-onset/incident exposure. It is possible that the duration of cartilage calcification might impact OA outcomes, but this could not be tested in those prior studies. Identifying incident cartilage calcification would require serial imaging of subjects without baseline cartilage calcification to identify incident cartilage calcification, such as in the Multicenter Osteoarthritis (MOST) Study, and then follow prospectively to assess long-term outcomes (52). The potential impact of asymptomatic cartilage calcification on OA progression would also be of interest; all subjects in these studies had knee pain, attributed at least in part to knee OA. It remains unknown whether incidentally detected cartilage calcification of the knee might lead to development of OA. Finally, these studies included participants with CPPD+OA and were not designed to assess the occurrence of other CPPD disease manifestations such flares of acute CPP crystal arthritis. Isolating and assessing the potential contributions of flares of acute CPP crystal arthritis or impact of chronic CPP inflammatory arthritis would not be possible with these cohorts.

Beyond Osteoarthritis: Identifying Cohorts with Different Manifestations of CPPD

Large clinic-based CPPD cohorts do not exist, and our knowledge of risk factors for CPPD and long-term outcomes in CPPD stem from studies that identified patients using billing codes and other electronic health record data. Epidemiologic research on CPPD has largely centered around a broad definition of CPPD, encompassing all of its presentations – both symptomatic and asymptomatic (53, 54). Acute CPP crystal arthritis represents a discrete inflammatory episode that recurs in approximately 25% of patients after the initial episode, and may be associated with different outcomes compared to a broader definition of CPPD (55). Several studies identified a cohort of patients with acute CPP crystal arthritis (pseudogout) using a single diagnosis code from a general practitioner (56, 57). More advanced bioinformatics methods, including machine learning, offer tools for more accurate identification of acute CPP crystal arthritis using electronic health record data (58). A machine learning-based algorithm was able to identify patients with definite or probable acute CPP crystal arthritis (synovitis with either synovial fluid CPP crystals or radiographic evidence of chondrocalcinosis) with a positive predictive value of 81% in a large electronic health record database (58). The algorithm can be applied to identify a large cohort of patients with this specific CPPD manifestation, facilitating observational studies of long-term outcomes. Well-phenotyped cohorts of patients with the various clinical presentations of CPPD are needed for evaluating long-term outcomes of this crystalline arthritis.

Research Agenda

This is an exciting time for CPPD clinical research, with several international collaborations underway to develop research infrastructure for this common form of arthritis (Figure 2). OMERACT is planning a Delphi exercise to identify the core outcome domains for short-term studies of acute CPP crystal arthritis and long-term studies of CPPD disease, similar to the dichotomy used for studies of acute inflammatory gout and hyperuricemia. Once core outcome domains are established, the next step will be to validate generic or CPPD-specific instruments to measure these domains. Meanwhile, ACR/EULAR are jointly sponsoring development of classification criteria for CPPD, which is underway (59). The final validated classification criteria will facilitate recruitment of clinic-based cohorts of CPPD patients, some of which may be used to test and validate CPPD-specific outcome metrics. CPPD cohorts will provide ripe opportunity to study the natural history of different CPPD manifestations, including recurrence and distribution of affected joints with acute CPP crystal arthritis, and risk factors for development of particular manifestations of CPPD disease. Cohorts will also enable study of long-term articular outcomes of asymptomatic cartilage calcification, either in patients with no joint symptoms at all, and/or in patients with symptomatic CPPD in one joint and asymptomatic cartilage calcification in another joint. Long-term extra-articular outcomes of CPPD disease, such as risk of cardiovascular disease, a common comorbidity in other forms of inflammatory arthritis, can be followed in such cohorts. Genome-wide association studies in CPPD cohorts may uncover novel polymorphisms that confer increased risk for this common form of arthritis, and/or particular manifestations of CPPD disease. CPPD cohorts will also help answer questions about the societal impact of CPPD, which has not been studied to date. Future work related to attribution of symptoms to CPPD vs. other type of arthritis will also be paramount; studies might include tissue histopathology in patients with rheumatoid arthritis that have ongoing synovitis while on traditional rheumatoid arthritis therapies and imaging findings of CPPD in the symptomatic joints. Validated metrics for CPPD clinical research will facilitate clinical trials of therapies to prevent recurrent flares, decrease the burden of CPP deposits, and reduce joint pain – ultimately leading to improved clinical care for patients with CPPD.

Figure 2.

Areas for development in CPPD clinical research.

Summary

Outcomes of CPPD disease, a common crystalline arthritis in adults for which no targeted treatments exist, are not well characterized. Two ongoing international collaborations are establishing a framework for CPPD clinical studies: the OMERACT CPPD Working Group is developing a core outcome domain set for CPPD, which complements efforts by the ACR/EULAR CPPD Classification Criteria working group to establish criteria for classifying CPPD patients for clinical research. Prior clinical studies of interventions for CPPD (n=112) generally focused on joint damage and calcification on imaging tests, joint pain and swelling, and inflammatory biomarkers. A recent qualitative interview study including patients with CPPD and healthcare providers identified additional outcomes of importance including flares, overall function, and use of analgesic medications. Imaging outcomes, including joint damage and calcification, are likely to feature prominently for this crystalline arthritis though several major steps remain until these are ready for use in clinical research, including the need to demonstrate reliability and sensitivity to change for plain radiography and dual-energy CT. Challenges to outcome measurement in CPPD include questions of attribution of signs and symptoms to CPPD versus co-existing forms of arthritis, a current lack of therapies to prevent or dissolve calcium pyrophosphate crystal deposition, absence of validated patient- or physician-reported clinical outcome measures for CPPD, and sparse existence of large cohorts of CPPD patients for studying outcomes of the various clinical presentations of CPPD. Prior research on osteoarthritis-related outcomes in CPPD provides concrete examples of some of the challenges to studying outcomes in CPPD. CPPD disease is clinically heterogeneous and many questions about outcomes of this disease remain unanswered. Ongoing international efforts to establish a framework for CPPD clinical research will lay the groundwork for measuring outcomes of this crystalline arthritis.

Practice points.

• Challenges to outcome measurement in CPPD include questions of attribution of signs and symptoms to CPPD versus co-existing forms of arthritis, lack of therapies to prevent or dissolve calcium pyrophosphate crystal deposition, absence of validated patient- or physician-reported CPPD outcome measures, and scarcity of large cohorts in which to study outcomes of different clinical presentations of CPPD.

• Imaging evidence of joint damage and calcification are likely to be outcome measures in future clinical studies of CPPD, though reliability and sensitivity to change in CPPD require further testing for several imaging modalities.