A prospective study of dual-energy CT scanning, US and X-ray in acute calcium pyrophosphate crystal arthritis

Rheumatology key message

• Dual-energy CT demonstrated excellent sensitivity for acute calcium pyrophosphate crystal arthritis.

Dear Editor, CPPD crystal deposition disease, a common inflammatory arthritis, is notoriously underdiagnosed [1, 2]. Acute calcium pyrophosphate (CPP) crystal arthritis typically presents as acute monoarthritis, representing the most dramatic subtype of CPPD [2]. The CPPD diagnostic criteria proposed >30 years ago require both demonstration of CPP crystals in synovial fluid and X-ray chondrocalcinosis [3]. US and dual-energy CT (DECT) have been adopted for use in crystalline arthritis diagnosis and research since these CPPD diagnostic criteria were proposed. US is highly sensitive and specific for CPPD and demonstrates high inter-observer reliability [4, 5]. DECT can distinguish CPP from monosodium urate crystals and demonstrates high sensitivity and specificity for CPPD in ex vivo meniscectomy samples [6, 7]. We performed a pilot study to assess the sensitivity of DECT for acute CPP crystal arthritis as compared with US and X-ray.

We identified eligible patients at an academic medical centre, from March to November 2018, and reviewed the electronic medical record for the following entry criteria: age ⩾18 years; acute monoarthritis of the wrist, hand, elbow, knee, ankle or foot; and recent arthrocentesis with synovial fluid crystal analysis positive for CPP crystals. Patients were ineligible if synovial fluid crystal analysis was positive for both monosodium urate and CPP crystals; surgical hardware was present in the aspirated joint; pregnant; unable to lie flat; or weight was ⩾204 kg. The Partners HealthCare Institutional Review Board approved this study and patients provided informed consent prior to study procedures.

The study visit occurred as soon as possible after joint aspiration. Subjects completed a questionnaire regarding the current episode and underwent DECT, US and X-ray of the aspirated joint and a standardized joint (right wrist) following a standardized protocol (see supplementary material, section on Imaging modalities, protocols, and DECT software settings, available at Rheumatology online). The primary DECT outcome was presence of colour-coded changes indicating CPP deposits. We considered two thresholds for the volume of colour-coded changes defining a positive scan after examining the data. The primary US outcome was calcium deposits evidenced by linear hyperechoic deposits or bands in hyaline cartilage and/or hyperechoic sparkling spots in fibrocartilage or tendon [4]. The primary X-ray outcome was chondrocalcinosis, defined as calcific deposits in hyaline cartilage or fibrocartilage. For each imaging modality, location(s) of the primary outcome were recorded. Presence of the primary radiologic outcome at any location defined a positive scan.

We estimated sensitivity of each imaging modality using the uniformly minimum-variance unbiased estimate, equal to the percentage of subjects with a positive scan in the aspirated joint. We used the exact test to calculate 95% CIs. We calculated prevalence of positive scans in the standardized joint, rather than sensitivity, because a reference standard did not exist. DECT sensitivity and prevalence were calculated using two volume thresholds.

Ten of 27 eligible patients were enrolled. Mean (s.d.) age was 73.0 (9.7) years and 40% were female (supplementary Table S1, available at Rheumatology online). Acute CPP crystal arthritis occurred in the knee in eight subjects and in the wrist in two subjects. Mean interval between arthrocentesis and enrolment was 16.9 (8.7) days. Sixty percent of subjects had received an IA CS injection.

Location of radiologic outcomes in the aspirated joint and volume of colour-coded changes on DECT are presented in Table 1; data for the standardized joint are in supplementary Table S2, available at Rheumatology online. In the aspirated joint, the minimum volume of colour-coded changes was 0.02 cm³; 9 of 10 had a colour-coded volume >0.40 cm³. Based on these data, we chose a lower threshold (>0.01 cm³) that defined all aspirated joints as positive, and a higher threshold (>0.40 cm³) that defined most but not all aspirated joints as positive.

Table 1.

Presence and location of radiographic outcomes in the aspirated joint

Subject	DECT colour-coded changes (cm3)	Conventional CT chondrocalcinosis	US calcific deposits	X-ray chondrocalcinosis
1	Patellofemoral Post. superior femoral condyle Meniscus Tibiofibular (0.48)	Patellofemoral Post. superior femoral condyle Meniscus	Meniscus	Meniscus
2	Intercondylar notch Meniscus (0.73)	Meniscus	Femoral cartilage Meniscus	Meniscus
3	Intercondylar notch Meniscus Patellofemoral Tibiofibular (2.85)	Intercondylar notch Meniscus Patellofemoral Post. superior femoral condyle Tibiofibular	Femoral cartilage	Intercondylar notch Meniscus Patellofemoral
4	Intercondylar notch Meniscus Patellofemoral Post. superior femoral condyle Popliteus tendon insertion Quadriceps tendon (1.05)	Meniscus Patellofemoral Post. superior femoral condyle Popliteus tendon insertion Quadriceps insertion	Femoral cartilage	Meniscus
5	Meniscus Post. superior femoral condyle Quadriceps tendon (0.50)	Meniscus Post. superior femoral condyle	Femoral cartilage	Meniscus Patellofemoral
6	Meniscus Patellofemoral (0.67)	Meniscus Patellofemoral	Femoral cartilage	None
7	Meniscus Patellofemoral Post. superior femoral condyle (1.33)	Meniscus Patellofemoral Post. superior femoral condyle	Femoral cartilage Meniscus	Meniscus Patellofemoral
8a	Intercarpal (0.02)		Intercarpal TFCC	None
9	Meniscus (0.45)	Meniscus	Meniscus	None
10	1st CMC Intercarpal 1st to 5th MCP TFCC (0.82)	1st CMC Intercarpal 1st to 5th MCP Radiocarpal TFCC	TFCC	Intercarpal 1st to 2nd MCP Radiocarpal TFCC

^aRight wrist was both the aspirated and standardized joint. TFCC: triangular fibrocartilage complex.

In the knee, DECT colour-coded changes in the both the medial and lateral menisci were present in 8/8 knees (Table 1 and supplementary Fig. S1, available at Rheumatology online). US calcific deposits were observed in 8/8 knees; the meniscus had deposits in 4/8. X-ray chondrocalcinosis of the meniscus was present in 6/8. In the two aspirated wrists, DECT colour-coded changes were present in the intercarpal joints and triangular fibrocartilage complex but not the radiocarpal joint (i.e. aspiration site). In the standardized joint (wrist), DECT colour-coded changes were most frequent in the intercarpal joints (7/10) (supplementary Table S2, available at Rheumatology online). Calcification on conventional CT and colour-coded changes on DECT were strongly correlated in the knee and wrist, with the exception of the radiocarpal joint and first CMC joint (Cohen’s kappa ∼0.4) (supplementary Table S3, available at Rheumatology online).

In the aspirated joint, DECT sensitivity was 90% (62–100%) for volume >0.40 cm³ and 100% (74–100%) for volume >0.01 cm³. US sensitivity was 100% (74–100%) and X-ray sensitivity was 70% (42–91%). In the standardized joint, DECT was positive in 20% (4–49%) of joints using a volume threshold >0.40 cm³; 90% (62–100%) had a positive DECT with a threshold >0.01 cm³. US was positive in 80% (51–96%) and X-ray chondrocalcinosis was present in 30% (9–59%) of standardized joints.

This study adds to a small but growing body of literature on DECT sensitivity for CPPD in vivo. Tanikawa et al. reported the performance of DECT ex vivo using surgically excised menisci from 25 patients with severe OA [6]. DECT sensitivity was 77.8% (7/9) and specificity was 93.8% (15/16) using synovial fluid CPP crystals as the reference standard. Pascart et al. recently demonstrated that DECT distinguishes CPP deposits from hydroxyapatite in vivo [8].

We identified higher sensitivity of X-ray chondrocalcinosis for acute CPP crystal arthritis than two small studies in CPPD [6, 9]; US sensitivity was similar to prior reports [5]. DECT, US and X-ray comprise different fields of view, hence their sensitivities may not be directly comparable. Differences in meniscus visualization might explain why CPP deposits were noted in the meniscus in 8/8 knees on DECT, but only 6/8 on X-ray and 4/8 on US.

Limitations to this pilot study include small sample size and the lack of a control group. We required acute CPP crystal arthritis in case crystal location (synovial fluid vs cartilage) impacted results. It is possible that DECT sensitivity differs across CPPD subtypes; in gout, DECT was less sensitive in new-onset acute gout and non-tophaceous gout than tophaceous gout (supplementary material, section Additional references related to DECT in gout, available at Rheumatology online). Removing CPP crystals during aspiration might affect results, although arthrocentesis before DECT did not impact results in gout. Six subjects received an IA CS injection before enrolment; gout studies have not commented on whether this impacts DECT. DECT sensitivity changed slightly when different volume thresholds were used to define a positive scan; in gout, a DECT scoring system provided greater discrimination than volume of colour-coded changes. Given the small number of aspirated wrists, we were unable to consider joint-specific volume thresholds. Use of a 12 MHz US probe may have limited our ability to detect CPP deposits in superficial structures such as tendons.

DECT scanning is becoming increasingly widespread and has proven useful in gout. Future work with a larger sample and controls with mimicking conditions (e.g. gout, OA) will be critical to determine the sensitivity, specificity, positive and negative predictive value of DECT in CPPD. Establishing a scoring system to define a positive DECT and evaluating DECT performance in other CPPD subtypes will be of paramount interest.

Funding: This work was supported by the Brigham and Women’s Hospital Brigham Research Institute Cross-Collaborative Pilot Award in Musculoskeletal Research and the National Institutes of Health (K23 AR075070, K24 AR055989, L30 AR070514 and P30 AR072577).

Disclosure statement: The authors have declared no conflicts of interest.