Article Text
Abstract
Background Calcium crystals are below the spatial resolution limit of currently available imaging techniques, and only aggregates can be identified in vivo at more advanced stages of the disease. Although dual-energy computed tomography (DECT) has the potential to discriminate the various calcium crystal types owing to its biochemical signature assessment capabilities, it remains to be seen whether this technique would be able to identify early-stage calcium crystal deposition in vivo.
Objectives We aimed to assess whether DECT was able to identify calcium crystal deposition in the knee prior to the onset of chondrocalcinosis (CC), more specifically if DECT attenuation properties differed between patients with CC and controls without CC on DECT.
Methods Consecutive patients with clinical suspicion of crystal arthritis and knee DECT scans were retrospectively reviewed and assigned to either CPPD (n=50) or control (n=82) groups depending on the presence/absence of CC on DECT. Regions of interest (ROI) were drawn in the following knee zones on a specific coronal DECT image: hyaline cartilage of the patellofemoral and medial and lateral tibiofemoral joint spaces, as well as medial and lateral menisci. The presence or absence of CC in these predefined ROIs were noted. Five DECT parameters were obtained: CT numbers (HU) at 80 and 140 kV, dual-energy index (DEI), electron density (ρe), and effective atomic number (Zeff). Knee zones were compared between groups using mixed linear models adjusting for age and the presence of osteoarthritis. A subgroup analysis was performed excluding zones were calcifications were visible on DECT images.
Results Menisci from CPPD patients and controls had a mean Zeff of 7.9±0.4 and 7.6±0.2 (p<0.0001), mean ρe of 85±23 and 74±14 (p<0.0001) and mean DEI of 0.0036±0.0046 and -0.0001±0.0042 (p<0.001), respectively. DEI values differed significantly between patients and controls in tibiofemoral cartilage (0.0026±0.0041 in CPPD and 0.0023±0.0045)(p=0.013) but not in patellofemoral cartilage (p=0.57). When considering only the various regions from CPPD patients without CC in the selected ROIs, the ρe in menisci (n=79/185) did not differ between groups and differences in Zeff (p=0.15) and DEI (p=0.09) did not reach statistical significance after adjustment for age and osteoarthritis.
Conclusion DECT has the potential to discriminate between meniscal fibrocartilage and articular cartilage of CPPD patients and controls in predefined regions of interest. DECT’s ability to improve the sensitivity of conventional CT to identify invisible CPP deposits remains unclear as the trend did not reach statistical significance.
Disclosure of Interests None declared