Background
A high rate of glomerulosclerosis serves as an important signal of poor response to treatment and a high risk of disease progression or adverse prognosis in transplanted kidneys. We hypothesized that contrast-enhanced ultrasound (CEUS) could serve as a novel imaging biomarker in the early prediction of glomerulosclerosis rate by evaluating renal allograft microcirculation.
Methods
A retrospective analysis was performed on 143 transplanted kidney recipients with confirmed pathology, including 100 in the training group and 43 in the validation group. All patients underwent conventional ultrasound (CUS) and CEUS examinations. The patients were divided into two groups: those with >50% glomerulosclerosis and those with ≤50% glomerulosclerosis. The nomograms derived from independent predictors identified by multivariate logistic analysis were assessed using receiver operating characteristic (ROC) curve analysis, 1,000 bootstrap resamples, calibration curves, and decision curve analysis (DCA).
Results
The patients with >50% glomerulosclerosis and those with ≤50% glomerulosclerosis showed statistically significant differences in CEUS parameters, including in peak intensity (PI) (25
vs.
30; P<0.001), absolute time to peak (ATTP) (10
vs.
9; P=0.004), and time to peak (TTP) (22
vs.
19.5; P=0.026). Multivariate analysis revealed that PI [odds ratio (OR) =0.852; 95% confidence interval (CI): 0.737–0.986], peak systolic velocity (PSV) of the interlobar artery (OR =0.850; 95% CI: 0.758–0.954), cortical echogenicity (OR =38.429; 95% CI: 3.695–399.641), and time since transplantation (OR =1.017; 95% CI: 1.006–1.028) were independent predictors of whether the glomerulosclerosis rate was >50% and were incorporated into the construction of a nomogram. The area under the curve (AUC) of the nomogram in the training and validation groups was 0.914 (95% CI: 0.840–0.960) and 0.909 (95% CI: 0.781–0.975), respectively, with a bootstrap resampling AUC of 0.877. The calibration curve and DCA confirmed the diagnostic performance of the nomogram model.
Conclusions
The nomogram, which combined CUS, CEUS, and clinical indicators, exhibited notable predictive efficacy for the glomerulosclerosis rate in transplanted kidneys, thereby demonstrating the potential to improve clinical decision-making.