Abstract
Objective Conventional radiography (CR) and ultrasound (US) are used interchangeably for identification of calcium pyrophosphate deposition (CPPD). The aim of this study was to assess whether combining US and CR offers greater accuracy over either modality alone for the identification of CPPD.
Methods Consecutive patients scheduled for knee replacement surgery for osteoarthritis were enrolled. Before surgery, patients underwent CR and US of the knee. Menisci and hyaline cartilage were collected and analyzed using polarized light microscopy to confirm the presence of CPPD (gold standard). CR and US were assessed for absence/presence of CPPD by expert radiologists and sonographers. Diagnostic performance statistics were calculated. Poisson models with robust variance estimators were used to determine the likelihood of identifying CPPD.
Results Fifty-one patients (63% female, mean age 71.4 [SD 8] years) were enrolled. US demonstrated higher overall accuracy than CR for CPPD identification (0.78 vs 0.73). Sequential use of both modalities provided an advantage when only 1 knee site was positive in 1 of the 2 techniques; however, when 2 or 3 sites were positive, no additional advantage was observed. When US was negative, subsequent CR did not improve CPPD detection, but in cases of a negative CR, a positive US increased the likelihood of CPPD by 4.21 times, whereas a negative US substantially reduced the probability of CPPD, increasing the likelihood of its absence by 76%.
Conclusion US was more accurate than CR for identification of CPPD. Performing both exams can be an added value for CPPD identification only in a few specific cases.
