O2156: Predictors of Adverse Events Following Nephrectomy for Renal Tumors in Children: Analysis from the NSQIP-P Database

Background: Renal Tumors (RTs), including Wilms tumors, are among the most common solid tumors in children. Surgical resection is the cornerstone of treatment, typically in the form of a radical nephrectomy. Recently, with the expanded use of neo adjuvant treatments, alternative surgical approaches, such as nephron-sparing surgery (NSS), have become more common. In this study we aimed to characterize the 30-day outcomes of nephrectomy for RTs and identify predictors of adverse outcomes using the National Surgical Quality Improvement Program-Pediatric (NSQIP-P) database.

Methods: We queried NSQIP-P for children < 18 years with Malignant Neoplasms of the Kidney (ICD-9: 189; ICD-10: C64) who underwent nephrectomy from 2012-2021. Relevant clinical variables relating to patient demographics, outcomes, and type of nephrectomy were extracted. Due to the low overall rate of complications, the most prevalent adverse outcomes were compiled into a binary Composite Adverse Outcome (CAO) variable for further analysis. Univariate and multivariate logistic regression was conducted to identify predictors of CAO for all types of nephrectomies. A secondary analysis was conducted to compare the rates of individual adverse events between partial and total nephrectomy.

Results: We identified 1883 patients who underwent total (85.6%) or partial nephrectomy (13.5%) for RTs. Mean patient age was 4.53 ± 3.61 years and 52.2% were female. The most common complications overall were Mortality (0.2%), Length of Stay >30 days (1.2%), Organ Space Infection (0.5%), Prolonged Ventilation >48 hours (1.7%), Renal Insufficiency (0.4%) and Postoperative Transfusion (30.1%), which were used to construct the binary CAO variable. On multivariate regression, Previous Hematologic Disorder (aOR=2.061, 95% C.I.=1.114-3.811) and Preoperative hematocrit (HCT) (aOR=0.863, 95% C.I.=0.818-0.920) were independent predictors of morbidity after adjusting for age and covariates. An ROC curve was constructed and used to determine that HCT < 31.00 had the best sensitivity and specificity for predicting morbidity (AUC=0.674, p< 0.001). On secondary analysis, partial nephrectomy had a higher rate of postoperative acute renal failure (p=0.050), and bleeding requiring transfusion (p=0.005) than total nephrectomy.

Conclusion: In this analysis of a large, nationally validated dataset, we found overall low rates of mortality and complications, demonstrating that surgical intervention is safe in the short-term irrespective of the extent of resection. Patients with preoperative Hematocrit < 31.00 are more likely to have morbidity and should considered for preoperative transfusion. Partial nephrectomy appears to have similar 30-day outcomes to total nephrectomy, with the exception that it is associated with higher risk of renal injury and bleeding likely due to patient selection for higher risk. Further analysis, including the use of data from SIOP may be helpful in understanding the role of nephron sparing surgery.