Determination of the Urinary Stone Chemical Composition Using the Hounsfield Unit

Abstract

Objectives To determine the chemical composition of pure and mixed urinary calculi by multislice helical computed tomography (CT) in an in-vitro setting. CT is becoming the preferred radiologic examination in diagnosis and management of calculus disease. The management of calculus disease can be facilitated by ascertaining stone composition. Recent in-vitro studies have suggested that CT can be used to predict the composition of urinary stones on the basis of differences in radiodensity (measured in Hounsfield units, HU) supporting the clinician in selecting the more opportune therapeutic strategy. Aim of our study is to compare the radiodensity and chemical composition of urinary stones with the purpose of identifying the cut-off value of HU with high specificity. Materials e Methods A total of 76 patients with urinary stones, candidate to treatment, were assessed in a multislice helical scanner within an air-filled environment (Ge Healthcare light Speed, 120 KV; 150–250 mA 1.5 pitch thick speed 2.5; interval 2.5) differentiating dimensions and radiodensity. The chemical compositions of urinary stones were assessed on the basis of the differences in densities measured in Hounsfield units. Chemical analysis of stone has been done, using colorimetrical kit Ecoline (DiaSys Diagnostic System GmbH, Germany) validated versus IR Spectrometry. HU density values are expressed as median and interquantile range (IQR). Statistic analysis has been done using SPSS 15.0 (SPSS Inc., Chicago, IL, USA). Results The stones were assigned to seven different groups according to the chemical composition: calcium oxalate in 57%, calcium oxalate and phosphate in 16%, calcium urate-oxalate 12%, uric acid in 5%, struvite and calcium phosphate in 4%, struvite and calcium oxalate in 5%, cystine in 1%. Uric acid stone density (Ur+Ur/Ox-Ca, n=13; HU 487, IQR 352 - 594) was significantly lower than both Ox-Ca pure (n=44; HU 639, IQR 460 -942) (Bonferroni correction p=0012) and phosphate (Ox / P-Ca+struvite, n=19; HU 801, IQR 502 -1205) (Bonferroni correction p=0006), while no difference was evident between oxalate and phosphate stones. Areas under ROC curve for uric acid stone identification (toward phosphate and oxalate) or for phosphate stone identification (toward uric acid and oxalate) were respectively 0.73 (95%IC 0.59–0.87) and 0.62 (95%IC 0.46–0.79). Cut-off values of 352 HU (first quartile uric distribution) and 1205 HU (third quartile phosphates distribution) allow to identify a specificity of 91% and 96%, respectively, for uric acid or phosphates stones. Conclusions Although our data underline an association between radiodensity and chemical composition, confirming literature, the big overlap of distributions of HU values limits clinical application. It is possible to identify specific HU cut-off values for uric acid and phosphates stone, however it is necessary to verify these results in a larger sample. The spiral TC is useful to study the chemical composition of the urinary stone, but a larger sample is necessary to calculate more accurate radiodensity cut-off values.