Role of 18F-FDG-PET/CT in Combination With Neutrophil–Lymphocyte Ratio in the Diagnosis of Upper Urinary Tract Lesion: Can We Accurately Predict Malignant Tumor?

OBJECTIVE: To explore whether preoperative 18Fluorine-Fluorodeoxyglucose-positron emission tomography/computed tomography (18F-FDG-PET/CT) in combination with neutrophil–lymphocyte ratio (NLR) could accurately predict malignant lesions of upper urinary tract (UUT). METHODS AND MATERIALS: The clinicopathologic data of a total of 252 patients with UUT lesions receiving surgical treatment at our center from January 2012 to November 2019 were retrospectively analyzed. All patients performed routine preoperative hematological examination, urine cytology, computed tomography urography (CTU), and 18F-FDG-PET/CT. Clinicopathologic data between 179 cases with malignancy (Group 1) and 73 cases with benign lesions (Group 2) were compared. Multivariate logistic regression analysis was used to explore the independent predictors of malignant UUT lesions. Receiver operating characteristic (ROC) curve was used to evaluate the predictive ability. RESULTS: Among all patients, univariate analysis showed that NLR, hydronephrosis, CTU indicating malignancy, and PET/CT indicating malignancy were significantly associated with malignant UUT lesions; multivariate analysis revealed that NLR, CTU indicating malignancy, and PET/CT indicating malignancy were independent predictors of malignant UUT lesions; the area under ROC curve (AUC) of NLR, CTU, PET/CT, combining CTU and NLR, combining PET/CT and NLR, and combining PET/CT and CTU were 0.735, 0.788, 0.857, 0.863, 0.913, and 0.919, respectively, for postoperative pathological malignancy. Among 68 patients undergoing ureteroscopy biopsy, univariate analysis suggested that NLR, positive urine exfoliation cytology, CTU indicating malignancy, and PET/CT indicating malignancy were significantly associated with malignant UUT lesions; multivariate analysis demonstrated that positive urine cytology, PET/CT indicating malignancy, and NLR were independent predictors of malignant UUT lesions; the AUC of NLR, ureteroscopy biopsy, and combining PET/CT and NLR were 0.768, 0.853, and 0.839, respectively, for postoperative pathological malignancy. CONCLUSIONS: Combining preoperative NLR and PET/CT performed well in differentiating benign from malignant UUT lesions, which could not be identified by traditional imaging or urine cytology. Combining preoperative NLR and PET/CT could be used to reduce unnecessary ureteroscopy biopsy, which might result in tumor cell dissemination and risk of associated complications.