Huntsman Cancer Institute at the University of Utah Salt Lake City, Utah
F. Ticona1, A. Tai2, C. Dial3, J. D. Tward1, and M. M. Poppe3; 1Department of Radiation Oncology, Huntsman Cancer Institute, University of Utah, Salt Lake City, UT, 2Department of Radiation Oncology, Medical College of Wisconsin, Milwaukee, WI, 3Huntsman Cancer Institute, University of Utah, Salt Lake City, UT
Purpose/Objective(s): Our study investigates renal dysfunction in pediatric patients undergoing abdominal radiation (RT). Other studies have report average estimated kidney exposure, none have utilized full dose volume histograms (DVH). Our primary objective assesses the long-term impact of RT on renal function by glomerular filtration rate (GFR). Inclusion criteria incorporated detailed medical records inclusive of laboratory results, physical examinations, disease history, chemotherapy, and full dosimetric data of RT. Materials/
Methods: A systematic review of medical records was conducted for 221 pediatric patients who underwent abdominal RT at a single institution between 2007 and 2020. The inclusion criteria required the last lab analysis to be > 3 years post-treatment. A total of 138 eligible patients were identified. Endpoints were categorized by Chronic Kidney Dysfunction (CKD) grades (= 1, = 2, = 3) following the National Kidney Foundation guidelines. Full volume-dose plan data was used to model potential causal relationships between dose to the kidneys and post-exposure renal dysfunction. The Lyman Kutcher Burman (LKB) model was used to perform NTCP analysis for this data set. The DVHs were first converted to EQD2 with a/ß=3.4 Gy. The model parameters were extracted using a maximum-likelihood method. Results: Patients were ages 1 to 21 years at the time of RT and with numerous diagnoses involving the abdomen. Prescribed treatment doses ranged from 800 cGy to 6400 cGy. We performed a univariate regression analysis and found the dose delivered to the lowest exposed kidney (only kidney in the setting of nephrectomy) to be a significant risk factor for = G2 CKD, OR 1.0095 per cGy (p=0.011). Although, age and use or chemotherapy in general were not independent risk factors for CKD in our analysis, carboplatin use was significant with an OR 2.27 (p=0.027). We found a baseline risk of 30% for grade = 2 CKD with no RT exposure to one kidney. In estimating the additional toxicity risk from RT, the extracted LKB model parameters are m=0.42+- 0.28, TD50=14.9 +-4.2 Gy and n=1.5+-1.0. The NTCP for moderate renal toxicity was found significantly correlated with the equivalent uniform dose (EUD) (p=0.03). EUD at 5% over the baseline risk is 3.9 Gy. We were unable to design a model for severe toxicity given a limited number of data point. Conclusion: Given the lack of pediatric partial kidney toxicity modeling in the literature, this adds significantly to pediatric RT late effects literature. It would appear from this analysis, that mean kidney dose (EUD) is the best partial kidney constraint guideline. Although intuitive, protection to at one kidney seems most prognostic for CKD, demonstrated by our modeling. The baseline risk for CKD is multi-factorial in pediatric cancer, as has been published elsewhere. RT adds to the risk of CKD to a small degree, with a recommendation to keep the mean dose to the lowest exposed kidney to < 4Gy to keep the risk of CKD < 5% over the baseline risk of significant CKD.
