High attenuation value in non-contrast computer tomography can predict pyonephrosis in patients with upper urinary tract stones

Abstract

To evaluate whether the higher attenuation value [Hounsfield unit (HU)] in non-contrast CT can predict pyonephrosis in patients with upper urinary tract stones (UTS). Between October 2019 and October 2021, patients with hydronephrosis or pyonephrosis secondary to upper UTS were retrospectively searched in our study. All patients with UTS were treated with percutaneous nephrostomy, percutaneous nephrolithotomy, retrograde ureteral stent or transurethral ureteroscope lithotripsy. We excluded patients treated with extracorporeal shock-wave lithotripsy. Patients whose CT was not performed in our hospital or treated in another hospital were also excluded. Clinical data regarding basic information, clinical feature, Calculi-related indicators, HU values of the renal pelvis, the thick wall of the renal pelvis on CT were collected. Univariate and multivariate logistic analyses were performed. Receiver operative characteristic curves were drawn to predict pyonephrosis. A total of 240 patients with UTS were retrospected in this research, 191 patients had hydronephrosis (Group 1), and 49 patients had hydronephrosis with pyonephrosis (Group 2). The HU value of the renal collecting system in Group 2 (mean, 15.46; range, +1/+30) was significantly higher than that in Group 1 (mean, 5.5; 5 range −6/+24) ( P = .02); the receiver operative characteristic curve analysis revealed that the best cut-off value of 9.5 could predict the presence of pyonephrosis, with 71.4% sensitivity and 70.2% specificity (area under the curve = 0.613; 95% CI: 0.514–0.713). In this study, we found the HU attenuation value of the renal collecting system can be used to distinguish pyonephrosis from hydronephrosis in patients with UTS.

Construction of an alignment diagram model for predicting calculous obstructive pyonephrosis before PNL

“…Yuruk et al [ 14 ] calculated that the HU of the renal collecting system in patients with pyonephrosis was significantly higher than that of patients with hydronephrosis; the sensitivity for diagnosing pyonephrosis with a HU of ≥9.21 was 65.96%; and the specificity was 87.93%. A study by Erdogan et al reached similar conclusions [ 15 , 16 ]. Many studies have focused on finding risk factors for complications of SIRS and sepsis after PNL [ [17] , [18] , [19] , [20] , [21] , [22] , [23] , [24] ], including female sex, diabetes mellitus, volume of renal stones, degree of hydronephrosis, density of hydronephrosis, urinary leukocytes, positive urine cultures and elevated blood neutrophils.…”