Functional data analysis applied to modeling of severe acute mucositis and dysphagia resulting from head and neck radiation therapy

Publication date: Available online 22 August 2016
Source:International Journal of Radiation Oncology*Biology*Physics
Author(s): Jamie A. Dean, Kee H. Wong, Hiram Gay, Liam C. Welsh, Ann-Britt Jones, Ulrike Schick, Jung Hun Oh, Aditya Apte, Kate L. Newbold, Shreerang A. Bhide, Kevin J. Harrington, Joseph O. Deasy, Christopher M. Nutting, Sarah L. Gulliford
PurposeCurrent normal tissue complication probability (NTCP) modeling using logistic regression suffers from bias and high uncertainty in the presence of highly correlated radiation therapy (RT) dose data. This hinders robust estimates of dose-response associations and, hence, optimal normal tissue-sparing strategies from being elucidated. Using functional data analysis (FDA) to reduce the dimensionality of the dose data could overcome this limitation.Methods and MaterialsFDA was applied to modeling of severe acute mucositis and dysphagia resulting from head and neck RT. Functional partial least squares regression (FPLS) and functional principal component analysis (FPCA) were used for dimensionality reduction of the dose-volume histogram data. The reduced dose data were input into functional logistic regression models (FPLS-LR and FPC-LR) along with clinical data. This approach was compared with penalized logistic regression (PLR) in terms of predictive performance and the significance of treatment covariate-response associations, assessed using bootstrapping.ResultsThe area under the receiver operating characteristic curves (AUC) for the PLR, FPC-LR and FPLS-LR models were 0.65, 0.69 and 0.67 for mucositis (internal validation) and 0.81, 0.83 and 0.83 for dysphagia (external validation), respectively. The calibration slopes/intercepts for the PLR, FPC-LR and FPLS-LR models were 1.6/-0.67, 0.45/0.47 and 0.40/0.49 for mucositis (internal validation) and 2.5/-0.96, 0.79/-0.04 and 0.79/0.00 for dysphagia (external validation). The bootstrapped odds ratios indicated significant associations between RT dose and severe toxicity in the mucositis and dysphagia FDA models. Cisplatin was significantly associated with severe dysphagia in the FDA models. None of the covariates was significantly associated with severe toxicity in the PLR models. Dose levels greater than approximately 1.0 Gy/fraction were most strongly associated with severe acute mucositis and dysphagia in the FDA models.ConclusionsFPLS and FPCA marginally improved predictive performance compared with PLR and provided robust dose-response associations. FDA is recommended for use in NTCP modeling.

Teaser

Normal tissue complication probability modeling using logistic regression (LR) suffers from bias and uncertainty in the presence of highly correlated radiation therapy dose data. Consequently robust estimates of the dose levels most strongly associated with toxicity and, potentially, predictive performance are limited. To overcome this limitation, functional data analysis (FDA), which describes the dose-volume histogram as a continuous curve, was applied to modeling of severe acute mucositis and dysphagia and compared with LR. FDA models demonstrated slightly better predictive performance and more robust dose-response estimates than LR.


from Cancer via ola Kala on Inoreader http://ift.tt/2bcJWol
via IFTTT