Clinical, Dosimetric, and Location-Related Factors to Predict Local Control in Non-Small Cell Lung Cancer

Cancer Treatment and Research

2008

Radiation therapy dose distributions to eradicate tumor cells are typically constrained in extent or intensity to minimize the risk of injury to nearby critical normal tissues. With the widespread use of 3D image-based treatment planning systems, the question naturally arises how patient-specific anatomy and treatment differences affect outcome. It has long been known, that for many organs, variations in the fractional volume irradiated to high doses greatly alters the dose to achieve a given complication level (the "isoeffective dose") [1]. Smaller irradiated fractional volumes often lead to a much lower risk of complication; this is often referred to as the "volume-effect" in the literature, but would be more correctly referred to as the "dose-volume" effect. Normal tissue complication probability (NTCP) modeling is simply the ongoing effort to understand the risk of normal tissue injury as a function of the 3D dose distribution. Recently, there has been a steady accumulation of NTCP studies [2], and this is expected to continue or even accelerate in the future. NTCP models are particularly needed when the "volume-effect" becomes important (i.e., injury depends on the detailed dose distribution), such as for skin, lung, or liver. In this chapter we will review the basic principles of NTCP modeling, as well as publications related to selected endpoints (xerostomia, radiation pneumonitis, late rectal toxicity), and several issues related to the use of NTCP models, especially relating to their safe use. Other recent reviews which further discuss data on endpoints of interest in treatment planning include Deasy and Fowler [3], Moiseenko et al. [4], and the slightly older but still invaluable Seminars in Radiation Oncology issue, edited by Randy Ten Haken, devoted to dose-volume effects in normal tissues [2]. A useful review of models and model principles are the chapters by Jackson and Yorke [5], and Yorke [6]. Many technical issues in modeling dose-volume outcomes were also discussed by Deasy et al. [7].This chapter describes standard NTCP models as well as our own approach, which is more data-driven and image-based [8]. This contrasts with the more common approach of assuming the validity of a specific model and then attempting to fit the model parameters to a given data set. The term "image-based" indicates

Section: Why Image-based Ntcp Analysis?supporting

confidence: 56%

Section: Why Image-based Ntcp Analysis?mentioning

confidence: 99%

Image-Based Modeling of Normal Tissue Complication Probability for Radiation Therapy

Cancer Treatment and Research

2008

“…For instance, De Crevoisier et al reported that rectal distension on the planning computed tomography (CT) scan is associated with an increased risk of biochemical and local failure in patients of prostate cancer when treated without daily image-guided localization of the prostate [25]. Similarly, we found that as tumor distance to the spinal cord decreased, the rate of local tumor failure increased in patients receiving definitive radiotherapy for lung cancer [26]. Moreover, biological markers were found to be predictive of biochemical failure in prostate cancer or radiation-induced lung injury post-radiotherapy treatment [27,28].…”

mentioning

confidence: 52%

Datamining approaches for modeling tumor control probability

et al. 2010

Acta Oncologica

Background Tumor control probability (TCP) to radiotherapy is determined by complex interactions between tumor biology, tumor microenvironment, radiation dosimetry, and patient-related variables. The complexity of these heterogeneous variable interactions constitutes a challenge for building predictive models for routine clinical practice. We describe a datamining framework that can unravel the higher order relationships among dosimetric dose-volume prognostic variables, interrogate various radiobiological processes, and generalize to unseen data before when applied prospectively. Material and methods Several datamining approaches are discussed that include dose-volume metrics, equivalent uniform dose, mechanistic Poisson model, and model building methods using statistical regression and machine learning techniques. Institutional datasets of non-small cell lung cancer (NSCLC) patients are used to demonstrate these methods. The performance of the different methods was evaluated using bivariate Spearman rank correlations (rs). Over-fitting was controlled via resampling methods. Results Using a dataset of 56 patients with primary NCSLC tumors and 23 candidate variables, we estimated GTV volume and V75 to be the best model parameters for predicting TCP using statistical resampling and a logistic model. Using these variables, the support vector machine (SVM) kernel method provided superior performance for TCP prediction with an rs = 0.68 on leave-one-out testing compared to logistic regression (rs = 0.4), Poisson-based TCP (rs = 0.33), and cell kill equivalent uniform dose model (rs = 0.17). Conclusions The prediction of treatment response can be improved by utilizing datamining approaches, which are able to unravel important non-linear complex interactions among model variables and have the capacity to predict on unseen data for prospective clinical applications.

“…Therefore, recent approaches have focused more on datadriven models, in which dosimetric metrics are mixed with other patient or disease-based prognostic factors [4]. This approach is motivated by recent reports of imagespecific outcomes findings [5,6]. In de Crevoisier et al, it was reported that rectal distension on the planning computed tomography (CT) scan is associated with an increased risk of biochemical and local failure in patients of prostate cancer when treated without daily imageguided localization of the prostate [5].…”

Section: Introductionmentioning

confidence: 99%

“…In de Crevoisier et al, it was reported that rectal distension on the planning computed tomography (CT) scan is associated with an increased risk of biochemical and local failure in patients of prostate cancer when treated without daily imageguided localization of the prostate [5]. Similarly, in Hope el al., it was found that tumor distance to the spinal cord was a significant predictor of failure in irradiated lung cancer patients [6].…”

Section: Introductionmentioning

confidence: 99%

Nonlinear Kernel-Based Approaches for Predicting Normal Tissue Toxicities

Bradley

2008 Seventh International Conference on Machine Learning and Applications

2008

Since the early demonstration of the curative potential of radiation therapy for tumor sterilization, normal tissue toxicity continues to be dose limiting. Accurate prediction of patient's complication risk would allow personalization of treatment planning decisions. Nonlinear kernel methods can provide a robust framework for learning complex interactions between observed toxicities and treatment, anatomical, and patient-related variables. However, proper application of these powerful methods would require better understanding of a high-dimensional feature space that is spanned by all these variables. In this work, we investigate methods for visualization of this high-dimensional space and compare different approaches for extracting discriminant features. Our preliminary results demonstrate that principle component analysis is a valuable tool for visualizing high dimensional data and for determining proper kernel type. In addition, variable selection based on resampling methods within the logistic regression framework seemed to yield improved prediction performance compared to the recursive-feature elimination method.