Abstract
Objective Acute and chronic calcium pyrophosphate (CPP) crystal arthritis is characterized by the presence of synovial CPP crystals within a clinically inflamed joint. CPP crystals may be situated intracellularly or extracellularly; however, the clinical significance of their location remains understudied. The objective of this retrospective cohort study was to assess the relevance of the CPP crystal location in diagnosing acute/chronic CPP crystal arthritis.
Methods Data were collected from Waikato District Health Board to identify a study population with synovial fluid samples positive for CPP crystals. The cohort was stratified into 2 groups based on crystal location: intracellular and extracellular. The proportions of acute/chronic CPP crystal arthritis cases were compared between these groups. Acute/chronic CPP crystal arthritis was diagnosed when synovial CPP crystals were present, with objective evidence of joint inflammation and no other alternative diagnosis. Further analysis was made with respect to demographics, other laboratory results, and cartilage calcification.
Results This study included 134 patients: 108 with intracellular CPP crystals and 26 with extracellular CPP crystals. Acute/chronic CPP crystal arthritis was diagnosed in 85% of cases in the intracellular and 50% in the extracellular group (P < 0.001). Following exclusion of septic arthritis cases, acute/chronic CPP crystal arthritis was diagnosed in 97% of patients in the intracellular group and in 62% of those in the extracellular group (P < 0.001).
Conclusion The presence of intracellular CPP crystals is more strongly associated with acute/chronic CPP crystal arthritis than with extracellular CPP crystals alone.
Calcium pyrophosphate crystal deposition (CPPD) disease is a broad term, encompassing a range of radiological and clinical manifestations including asymptomatic cartilage calcification (chondrocalcinosis), osteoarthritis with CPPD, acute calcium pyrophosphate (CPP) crystal arthritis, and chronic CPP crystal inflammatory arthritis.1 CPPD has an estimated prevalence of 5% in the general population,2 occurring on average in the eighth decade of life.3 Despite it being a common condition, many aspects of CPPD are incompletely understood, reflecting a paucity of published evidence.
Acute/chronic CPP crystal arthritis is characterized by the presence of CPP crystals in the synovial fluid of a clinically inflamed joint. In the clinical setting, it can be unclear if the CPP crystals are a significant finding as they are also found in asymptomatic and noninflamed joints. One study demonstrated that 20% of patients undergoing total knee joint replacement surgery for osteoarthritis (OA) had synovial fluid samples positive for CPP crystals.4 This is further complicated by the observation that CPP crystals can coexist with other joint pathology, including septic arthritis and gout.5-8
Synovial CPP crystals can be observed both intracellularly and extracellularly. There is no evidence to inform clinicians of the importance of the CPP crystal location in the workup of acute/chronic CPP crystal arthritis; however, it has been postulated that intracellular and extracellular CPP crystals have equal clinical significance.9 Our hypothesis was that intracellular CPP crystals are more strongly associated with episodes of acute/chronic CPP crystal arthritis than extracellular CPP crystals.
METHODS
Study design and population. This retrospective cohort study was performed at Waikato District Health Board (WDHB), New Zealand. The population analyzed in this article was the same as the one observed in a recently published study by Harris et al.10 All synovial fluid sample results at WDHB Laboratory that were positive for CPP crystals from November 2018 to July 2021 were obtained. Two groups were identified: one with intracellular CPP crystals, with or without extracellular crystals (termed the “intracellular group”); and another with extracellular CPP crystals only (termed the “extracellular group”).
Data collection and study group allocation. Data were retrospectively collected from WDHB’s clinical records. Collected data included age, sex, ethnicity, BMI (calculated as weight in kilograms divided by height in meters squared), documented clinical features consistent with acute/chronic CPP crystal arthritis, CPP crystal location (intracellular and/or extracellular), presence of non-CPP crystals, synovial white blood cell (WBC) count, serum WBC count, serum C-reactive protein (CRP) levels, synovial bacterial culture results, specific joint affected, and the presence of cartilage calcification at both the aspirated joint (on radiograph) and nonaspirated joints (on radiograph or computed tomography [CT] scan). The CPP crystals were identified by a WDHB medical laboratory scientist or technician using polarising microscopy with 1 filter plus a first-order red compensator. All results were double checked by another medical laboratory scientist or technician.
Cases of acute/chronic CPP crystal arthritis were predefined as patients with a synovial fluid sample positive for intracellular and/or extracellular CPP crystals in a symptomatic joint, plus objective evidence of joint inflammation and no evidence of an alternative pathology. Objective joint inflammation was defined as a synovial WBC count of ≥ 2000 cells/μL. If the synovial WBC count was below this or unavailable, patients required (1) clinical features of acute/chronic CPP crystal arthritis (ie, documented swelling, synovitis, or erythema), and/or (2) imaging appearances consistent with active joint inflammation (ie, ultrasound and/or magnetic resonance imaging [MRI]). These predefined criteria met the definition of definite acute/chronic CPP crystal arthritis according to the 2023 ACR/EULAR CPPD classification criteria.11
A diagnosis of septic arthritis was made based on a positive synovial fluid bacterial culture with evidence of objective joint inflammation or a documented diagnosis of septic arthritis by the treating doctor. The diagnoses in the miscellaneous group were established on a retrospective review of the clinical records. Cases were excluded if urate crystals were present or there were insufficient clinical details.
Outcomes. The primary outcome of this study was the proportion of acute/chronic CPP crystal arthritis cases in the intracellular vs the extracellular groups. The secondary outcome was the proportion of acute/chronic CPP crystal arthritis cases in these 2 groups following exclusion of the septic arthritis cases. Further analyses were made on the demographics, other laboratory results, and radiological investigations.
Ethics approval. Ethics approval was obtained from New Zealand’s Northern B Health and Disability Ethics Committee (21/NTB/234) and a waiver for patient consent was provided. This research was consistent with principles from the Declaration of Helsinki and the Council for International Organizations of Medical Sciences.
Statistical methods. SPSS version 29.0.1.0 (IBM) was used for calculating the results and performing the statistical analysis. Continuous variables are reported as means and SDs. In all cases, P < 0.05 was taken as the level of statistical significance. Multivariate logistic regression analysis using the single-step enter method was performed with acute/chronic CPP arthritis as the dependent variable; and age, sex, presence of cartilage calcification at any site, and CPP crystal location as the independent variables.
RESULTS
Patient characteristics. One hundred forty patients were identified as having synovial fluid samples positive for CPP crystals from November 2018 to July 2021 at WDHB. Six patients were excluded: 5 for concurrent monosodium urate (MSU) crystals and 1 for insufficient details. The study population consisted of 134 patients, including 108 in the intracellular group and 26 in the extracellular group.
There were 105 cases of acute/chronic CPP crystal arthritis (103 acute CPP crystal arthritis and 2 chronic CPP crystal arthritis cases), 18 cases of septic arthritis, and 11 miscellaneous cases. The miscellaneous cases consisted of osteoarthritis with CPPD (4), traumatic hemarthrosis (1), avascular necrosis of the hip (1), psoriatic arthritis (1), patellar bursitis (1), infected ankle graft (1) and undifferentiated, noninflammatory joint pathology (2).
Objective joint inflammation was based on a synovial WBC count ≥ 2000 cells/μL in 94 cases. In the remaining 11 cases, clinical evidence of inflammation (all having swelling plus pain, erythema, or a raised CRP) was noted in 10 cases and MRI evidence of synovitis in 1 case. Of the 11 cases without a synovial WBC count ≥ 2000 cells/μL, 3 of these had a synovial WBC count < 2000 cells/μL and the other 8 had an insufficient specimen to determine the WBC count. However, all had increased synovial fluid polymorphonuclear cells per semi-quantitative microscopic examination, and a raised CRP (range 34-310 mg/L).
The mean age of the intracellular group was significantly higher than that of the extracellular group (76 yrs vs 65 yrs, respectively; P < 0.001). The baseline characteristics did not significantly differ otherwise between the groups based on sex, ethnicity, BMI, and obesity rates (Table).
Comparison of the intracellular and extracellular CPP crystal location groups.
Multivariate logistic regression analysis was performed to assess the effect of age, sex, cartilage calcification at any site, and CPP crystal location on the likelihood of presenting with acute/chronic CPP crystal arthritis. The overall model was statistically significant when compared to the null model (χ2 = 13.7; P = 0.008), correctly predicting 79% of cases. The CPP crystal location was significant (P = 0.004), but age (P = 0.26), sex (P = 0.82), and cartilage calcification at any site (P = 0.88) were not.
Intracellular CPP crystals appear more likely to be associated with a diagnosis of acute/chronic CPP crystal arthritis than extracellular CPP crystals. Acute/chronic CPP crystal arthritis cases made up 85% (92/108) of the intracellular and 50% (13/26) of the extracellular groups (P < 0.001; Figure 1). The remaining cases consisted of septic arthritis (intracellular 12% [13/108], extracellular 19% [5/26]; P = 0.93) and miscellaneous cases (intracellular 3% [3/108], extracellular 31% [8/26]; P < 0.001; Table).
Comparison of the acute/chronic CPP crystal arthritis and nonacute/chronic CPP crystal arthritis cases in the intracellular vs extracellular groups. CPP: calcium pyrophosphate.
When excluding septic arthritis cases, there were 116 patients. In this nonseptic population, acute/chronic CPP crystal arthritis cases made up 97% (92/95) of the intracellular and 62% (13/21) of the extracellular groups (P < 0.001). The remaining miscellaneous cases made up 3% (3/95) of the intracellular and 38% (8/21) of the extracellular groups (P < 0.001; Figure 2).
Comparison of the acute/chronic CPP crystal arthritis and nonacute/chronic CPP crystal arthritis cases in the intracellular vs extracellular groups following exclusion of septic arthritis cases. CPP: calcium pyrophosphate.
Other laboratory and radiological investigations. The mean levels of the serum CRP (115 mg/L vs 102 mg/L; P = 0.51), serum WBC count (10.8 cells × 109/L vs 12.3 cells × 109/L; P = 0.13) and synovial WBC count (57,010 cells/μL vs 35,190 cells/μL; P = 0.47) did not significantly differ between the intracellular and extracellular groups. The proportion of patients with normal CRP levels did not significantly differ either. Significantly more patients in the intracellular group had an inflammatory synovial fluid aspirate, defined as a synovial fluid count ≥ 2000 cells/μL (93% intracellular vs 71% extracellular; P < 0.001).
Cartilage calcification on radiograph at the aspirated joint with CPP crystals was found in 57% of the intracellular and 41% of the extracellular cases (P = 0.23). There was a nonsignificantly higher proportion of cartilage calcification at any joint on radiograph or CT imaging in the intracellular vs the extracellular group (63% vs 44%; P = 0.08).
DISCUSSION
An intracellular CPP crystal location is a statistically significant finding associated with acute/chronic CPP crystal arthritis. These results may prompt clinicians to use this variable in the clinical setting. The CPP crystal location must be interpreted in the context of the clinical assessment, other laboratory tests, and radiological investigations to ensure diagnostic accuracy. This is a novel finding, and further research is important to demonstrate reproducibility of these results.
Synovial CPP crystals are commonly present in the absence of clinical joint inflammation.4 The phagocytosis of synovial CPP crystals plays a key role in initiating the inflammatory process in acute/chronic CPP crystal arthritis by triggering the NALP-3 inflammasome and upregulating proinflammatory cytokines.12,13 One study observed a positive relationship between a greater percentage of synovial polymorphonuclear cells containing intracellular CPP crystals and elevated markers of local inflammation (WBC and interleukin 1β).14 This is consistent with our observed association between an intracellular CPP crystal location and episodes of acute/chronic CPP crystal arthritis.
Another explanation for the apparent increased specificity of intracellular CPP crystals might be related to the accuracy of their correct identification. Several studies have highlighted the poor interlaboratory reliability of CPP crystal identification, with substantial false-negative and false-positive rates.15-18 Urate crystals have a very characteristic and prominent appearance of highly birefringent needle-shaped crystals. In contrast, the identification of CPP crystals is more challenging as they are polymorphic, varying in size, shape (rod to rhomboid) and birefringence (weak or absent).19 CPP crystal interpretation is vulnerable to artefact interference since they can be difficult to distinguish from contaminants; therefore, the extracellular group might contain more misidentified crystals. An intracellular CPP crystal result may be more reliable as the presence of intracellular matter suggests phagocytosis occurred in vivo, increasing the probability the observed crystals are a true-positive result and less likely to be artefact.
The mean age was higher in the intracellular vs the extracellular group (76 years vs 65 years, respectively; P < 0.001). This is likely explained by the intracellular group having a higher proportion of acute/chronic CPP crystal arthritis cases, which is more likely to occur at an advanced age. Multivariate logistic regression analysis demonstrated this difference in age did not significantly contribute to the identified association between the CPP crystal location and diagnosis of acute/chronic CPP crystal arthritis. The other baseline characteristics did not significantly differ between the 2 groups.
The levels of the acute-phase reactants were predominantly influenced by the underlying diagnosis. The majority of the intracellular and extracellular groups comprised inflammatory diagnoses including acute/chronic CPP crystal arthritis or septic arthritis. This likely explains why there was no significant difference identified in the mean levels of the serum WBC count, serum CRP, and synovial WBC count. The intracellular group had a significantly greater proportion of inflammatory synovial fluid aspirates than the extracellular group, as defined by a synovial WBC count ≥ 2000 cells/μL (93% vs 71%, respectively; P < 0.001). This difference was explained by the extracellular group having a greater proportion of noninflammatory diagnoses than the intracellular group.
The observed rates of septic arthritis were low in both the intracellular and extracellular groups (12% vs 19%, respectively). This is because the finding of synovial fluid crystals appears to significantly reduce the probability of septic arthritis. One study demonstrated that when MSU or CPP crystals were present, septic arthritis occurred in 0% of cases with a synovial WBC count < 50,000 cells/μL and in 6% of cases with a synovial WBC count ≥ 50,000 cells/μL; however, when crystals were absent, septic arthritis occurred in 12% and 64% of cases, respectively.20
Overall, radiographic cartilage calcification was present at the aspirated joint in 55% of the total study population, and this did not significantly differ between the intracellular and extracellular groups. This is consistent with a previous study showing that plain radiographs have moderate sensitivity for CPPD.21
The strengths of this study include access to a set of comprehensive clinical, radiological, and laboratory data. This includes the CPP crystal location, which was reported and verified by a minimum of 2 qualified scientists and technicians at WDHB Laboratory. WDHB Laboratory leads a quality assurance program with 24 other laboratories in New Zealand. They have a strong focus on ensuring the reproducible and accurate interpretation of synovial fluid CPP crystals.
Although this study population was identified prior to publication of the 2023 ACR/EULAR CPPD classification criteria, our predefined criteria in this study fulfilled the classification criteria’s definition for definite acute/chronic CPP crystal arthritis.11
A limitation of this study includes its retrospective nature. As the authors did not clinically assess the patients, this does introduce the possibility of error. However, this risk was felt to be low because the available clinical records were comprehensive and enabled us to collect the required data.
In conclusion, an intracellular CPP crystal location appears more likely than an extracellular CPP crystal location to be associated with clinical episodes of acute/chronic CPP crystal arthritis. The CPP crystal location should be interpreted in the context of the clinical assessment, other laboratory tests, and radiological investigations to ensure diagnostic accuracy.
Footnotes
UMA is listed as an inventor in a patent application for “Nucleoside 5΄-phosporothioate analogues and uses thereof ” (publication no. 20140378408). All other authors have no conflicts of interest to declare.
- Accepted for publication September 13, 2024.
- Copyright © 2024 by the Journal of Rheumatology
REFERENCES
DATA AVAILABILITY
An anonymized set of these data may be shared upon reasonable request to the corresponding author, provided this is appropriately acknowledged.
