- 81 views
- 133 downloads
A Novel Comparative Analysis Approach to Personalize Chemotherapy Dose Calculation in Early Breast Cancer
-
- Author / Creator
- Perri, Melissa
-
Background
Worldwide, body surface area [BSA] is used to calculate chemotherapy dose. The BSA formula was originally developed in 1916, derived from height and weight, with no consideration of other patient characteristics. Most chemotherapy agents have a narrow therapeutic index and are distributed in lean body mass [LBM], leading to under- or over-dosing and deleterious effects to major organs, including the cardiovascular system, when body composition is not considered. While experts worldwide acknowledge the limitations and risks of BSA dosing, no practical approach to personalizing chemotherapy dose has been developed to this date. Ideally, body composition would be assessed by tests already routinely performed, avoiding unnecessary radiation exposure, clinic visits, discomfort to the patient, and cost. For example, most breast cancer patients undergo cardiac imaging prior to chemotherapy as per our clinical guidelines. We hypothesized that clinical parameters routinely performed prior to chemotherapy could predict LBM in early breast cancer patients.
Method
Early stage breast cancer patients (n = 45) enrolled in the Multidisciplinary Team Intervention in Cardio-Oncology (TITAN) study underwent pre-treatment cardiac MRI, body composition (iDEXA) and laboratory (complete blood cell count and chemistry). Cardiac MRI and iDEXA are considered 'gold standard' imaging modalities. The Pearson correlation was calculated to find the relationship between cardiac MRI values, total lean body mass, and routine chemistry.
Our modeling approach, which is novel in this area, aimed to select the best combination of parameters with the most predictive ability of total lean mass (iDEXA). The parameters included in this study were: cardiac MRI metrics (Left Ventricle (LV) mass, cardiac output), and laboratory parameters associated with major organ function (albumin, creatinine, bilirubin). All parameters were tested using univariate, multivariate and subset selection approach. Akaike's Information Criterion (AIC) was used to measure model quality, with lower AIC values indicating closer prediction.
Results
The univariate analysis of each parameter independently showed LV mass is most predictive with AIC 857.8, while combination of all parameter in multivariate fashion show improvement in prediction with AIC 851. The subset selection approach shows, Adjusted R2 with 4 parameters had AIC 849.14, Schwartz's information criterion (BIC) with 2 parameters had AIC 849.66 and Mallows' C Selection (Cp) model with 3 parameters had the least AIC 848.71 value (P < 0.001).
Conclusion
Our comparative analysis showed that the Cp model with 3 parameters (LV mass, cardiac output and bilirubin) has high prediction ability of LBM. This model can form the basis of a personalized formula for chemotherapy dose calculation. We expect this work to result in optimal cancer-specific outcomes while reducing short and long-term toxicities associated with necessary. -
- Subjects / Keywords
-
- Graduation date
- Fall 2018
-
- Type of Item
- Thesis
-
- Degree
- Master of Nursing
-
- License
- Permission is hereby granted to the University of Alberta Libraries to reproduce single copies of this thesis and to lend or sell such copies for private, scholarly or scientific research purposes only. Where the thesis is converted to, or otherwise made available in digital form, the University of Alberta will advise potential users of the thesis of these terms. The author reserves all other publication and other rights in association with the copyright in the thesis and, except as herein before provided, neither the thesis nor any substantial portion thereof may be printed or otherwise reproduced in any material form whatsoever without the author's prior written permission.