Dataset Information

Added Value of Systemic Inflammation Markers in Predicting Clinical Stage T1 Renal Cell Carcinoma Pathologically Upstaged to T3a.

ABSTRACT:

Objectives

We aimed to determine preoperative risk factors associated with pathologic T3a (pT3a) upstaging of clinical T1 (cT1) renal cell carcinomas (RCCs) and develop a novel model capable of accurately identifying those patients at high risk of harboring occult pT3a characteristics.

Methods

A retrospective analysis of 1324 cT1 RCC patients who underwent partial nephrectomy (PN) or radical nephrectomy (RN) was performed. The study cohort was divided into training and testing datasets in a 70:30 ratio for further analysis. Univariable and multivariable logistic regression analyses were performed to identify predictors associated with cT1 to pT3a upstaging and subsequently, those significant risk factors were used to construct models. We used the area under the curve (AUC) to determine the model with the highest discrimination power. Decision curve analyses (DCAs) were applied to evaluate clinical net benefit associated with using the predictive models.

Results

The rates of upstaging were 6.1% (n = 81), 5.8% (n = 54) and 6.8% (n = 27) in the total population, training cohort and validation cohort, respectively. Tumor size, clinical T stage, R.E.N.A.L. (radius, exophytic/endophytic properties, nearness of tumor to collecting system or sinus, anterior/posterior) nephrometry score, lymphocyte to monocyte ratio (LMR), prognostic nutrition index (PNI) and albumin to globulin ratio (AGR) were significantly associated with pT3a upstaging. The model that consisted of R.E.N.A.L. score, LMR, AGR and PNI achieved the highest AUC of 0.70 in the validation cohort and yielded the highest net benefit. In the subpopulation with complete serum lipid profile, the inclusion of low-density lipoprotein cholesterol (LDL-C) and Castelli risk index-I (CRI-I) significantly improved the discrimination of model (AUC = 0.86).

Conclusions

Our finding highlights the importance of systemic inflammation response markers and serum lipid parameters in predicting pT3a upstaging. Our model had relatively good discrimination in predicting occult pT3a disease among patients with cT1 renal lesions, and the use of the model may be greatly beneficial to urologists in risk stratification and management decisions.

SUBMITTER: Liu H

PROVIDER: S-EPMC8202414 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BackgroundModels that predict postoperative complications often ignore important intraoperative events and physiological changes. This study tested the hypothesis that accuracy, discrimination, and precision in predicting postoperative complications would improve when using both preoperative and intraoperative data input data compared with preoperative data alone.MethodsThis retrospective cohort analysis included 43,943 adults undergoing 52,529 inpatient surgeries at a single institution during a 5-y period. Random forest machine learning models in the validated MySurgeryRisk platform made patient-level predictions for seven postoperative complications and mortality occurring during hospital admission using electronic health record data and patient neighborhood characteristics. For each outcome, one model trained with preoperative data alone; one model trained with both preoperative and intraoperative data. Models were compared by accuracy, discrimination (expressed as area under the receiver operating characteristic curve), precision (expressed as area under the precision-recall curve), and reclassification indices.ResultsMachine learning models incorporating both preoperative and intraoperative data had greater accuracy, discrimination, and precision than models using preoperative data alone for predicting all seven postoperative complications (intensive care unit length of stay >48 h, mechanical ventilation >48 h, neurologic complications including delirium, cardiovascular complications, acute kidney injury, venous thromboembolism, and wound complications), and in-hospital mortality (accuracy: 88% versus 77%; area under the receiver operating characteristic curve: 0.93 versus 0.87; area under the precision-recall curve: 0.21 versus 0.15). Overall reclassification improvement was 2.4%-10.0% for complications and 11.2% for in-hospital mortality.ConclusionsIncorporating both preoperative and intraoperative data significantly increased the accuracy, discrimination, and precision of machine learning models predicting postoperative complications and mortality.