Comparison of ultrasound attenuation by calcium pyrophosphate, hydroxyapatite and monosodium urate crystals: a proof-of-concept study
Annals of the Rheumatic Diseases, 2022
Over the past decade, ultrasound (US) has been extensively used for the diagnosis of crystal arth... more Over the past decade, ultrasound (US) has been extensively used for the diagnosis of crystal arthropathies, in particular gout and calcium pyrophosphate deposition disease. 2 US has proven to be accurate and reliable for the diagnosis of these diseases, and validated definitions for monosodium urate (MSU) and calcium pyrophosphate (CPP) crystal deposition in and around joints have been released by the Outcome Measures in Rheumatology US working group. On the other hand, although routinely used in the diagnosis and management of calcific tendinitis/periarthritis, the precise role of US in hydroxyapatite (HA) deposition disease, in particular its validation and reliability assessment, remains to be determined. The propensity to attenuate the US beam and generate acoustic shadowing are important US imaging features in distinguishing between the different types of crystal deposition. Generally, HA crystals are considered to generate acoustic shadowing, while MSU crystals tend to do the same depending mainly on the concentration and size of aggregates. In contrast, CPP crystals typically do not attenuate the US beam. Therefore, this proofofconcept study aimed to investigate the US attenuation characteristics of increasing concentrations of CPP, HA and MSU crystals. Sixteen synthetic crystal suspensions with known concentrations of CPP (26–109 mg/mL), HA (31–153 mg/mL) and MSU (90–500 mg/mL) were prepared. These specific concentrations were selected to replicate the Xray attenuation characteristics of those crystals when imaged by conventional radiography, CT and dualenergy CT in vivo (figure 1). The density of the agarbased lipogel background was intentionally increased to mimic the Xray attenuation of hyaline cartilage (ie, 100–120 HU at 120 kVp). Each crystal suspension was removed from its Eppendorf tube and placed in a plastic container filled with US gel, next to the control (ie, crystalfree background) calibration phantom. We acquired all US images using a Samsung RS80A system equipped with a highfrequency linear array transducer (4–18 MHz) set at the maximum frequency, by applying the same settings. US scans were performed by a single experienced sonographer (GF), blinded to the crystal type and concentration. For each of the 16 crystal suspensions, at least 2 images were recorded both in the longaxis and shortaxis views, the latter including the control phantom. Interpretation of US images for the extent of US beam attenuation and the presence of acoustic shadowing was performed in consensus with a second experienced sonographer (FB). No patients were involved in this study. Conventional radiography and CT confirmed the comparable and representative Xray attenuation of the three crystal suspension types (figure 1). None of the five CPP suspensions generated acoustic shadowing or substantial US attenuation, regardless of crystal concentration. In contrast, both HA and MSU suspensions substantially attenuated the US beam and generated acoustic shadowing from a different given crystal concentration. While HA did not attenuate the US beam at 31 mg/mL, US attenuation gradually increased with increasing crystal concentration from 62 to 123 mg/mL, eventually generating clear acoustic shadowing at 153 mg/dL. Similarly, MSU at 90 mg/mL showed no US attenuation, which gradually increased from 195 up to 345 mg/mL to finally generate clear acoustic shadowing from 420 mg/mL (figure 1). This proofofconcept study provides initial evidence that, at crystal concentrations encountered in and around patient joints (ie, with a CT attenuation range of 150–300 HU), CPP does not generate substantial acoustic shadowing, unlike HA and MSU which attenuate US in proportion to crystal concentration. Being an in vitro pilot study, US beam attenuation was assessed empirically and not quantitatively or semiquantitatively. However, our findings highlight the potential ability of US to distinguish between CPP, HA and MSU crystal deposition based on their appearance and variable attenuation on grayscale images. Future larger studies should then be performed with various US systems and settings, multiple ultrasonographers, and in correspondence with ex vivo samples in order to validate and assess the reliability of these initial findings because the surrounding human tissues and equipment characteristics could impact the attenuation level of the US beam. If validation studies were to confirm the different patterns of US attenuation of the various crystal deposits, this would not only enhance the diagnostic performance of US in crystal arthropathies overall, but also enable the development of a grading system for US attenuation of crystal deposition— which is currently lacking—and thereby to semiquantitatively estimate crystal concentration for monitoring purposes. Letter
Uploads
Papers by Silvia Sirotti