Search

Clustering is the task of identifying groups of similar subjects according to certain criteria. The AJCC staging system can be thought as a clustering mechanism that groups patients based on their disease stage. This grouping drives prognosis and influences treatment. The goal of this work is to evaluate the efficacy of machine learning algorithms to cluster the patients into discriminative groups to improve prognosis for overall survival (OS) and relapse free survival (RFS) outcomes. We apply clustering over a retrospectively collected data from 644 head and neck cancer patients including both clinical and radiomic features. In order to incorporate outcome information into the clustering process and deal with the large proportion of censored samples, the feature space was scaled using the regression coefficients fitted using a proxy dependent variable, martingale residuals, instead of follow-up time. Two clusters were identified and evaluated using cross validation. The Kaplan Meier (KM) curves between the two clusters differ significantly for OS and RFS (p-value < 0.0001). Moreover, there was a relative predictive improvement when using the cluster label in addition to the clinical features compared to using only clinical features where AUC increased by 5.7% and 13.0% for OS and RFS, respectively. [ABSTRACT FROM AUTHOR]

Background: The incidence of oropharyngeal squamous cell carcinoma (OPSCC) in the US is rapidly increasing, driven largely by the epidemic of human papillomavirus (HPV)-mediated OPSCC. Although survival for patients with HPV mediated OPSCC (HPV+ OPSCC) is generally better than that of patients with non-virally mediated OPSCC, this effect is not uniform. We hypothesized that tobacco exposure remains a critical modifier of survival for HPV+ OPSCC patients.Methods: We conducted a retrospective analysis of 611 OPSCC patients with concordant p16 and HPV testing treated at a single institute (2002-2013). Survival analysis was performed using Kaplan-Meier analysis and Cox regression. Recursive partitioning analysis (RPA) was used to define tobacco exposure associated with survival (p < 0.05).Results: Tobacco exposure impacted overall survival (OS) for HPV+ patients on univariate and multivariate analysis (p = 0.002, p = 0.003 respectively). RPA identified 30 pack-years (PY) as a threshold at which survival became significantly worse in HPV+ patients. OS and disease-free survival (DFS) for HPV+ > 30 PY patients didn't differ significantly from HPV- patients (p = 0.72, p = 0.27, respectively). HPV+ > 30 PY patients had substantially lower 5-year OS when compared to their ≤30 PYs counterparts: 78.4% vs 91.6%; p = 0.03, 76% vs 88.3%; p = 0.07, and 52.3% vs 74%; p = 0.05, for stages I, II, and III (AJCC 8th Edition Manual), respectively.Conclusions: Tobacco exposure can eliminate the survival benefit associated with HPV+ status in OPSCC patients. Until this effect can be clearly quantified using prospective datasets, de-escalation of treatment for HPV + OPSCC smokers should be avoided. [ABSTRACT FROM AUTHOR]

• The majority of patients with oropharyngeal cancer undergoing radiotherapy develop Grade 3 or higher lymphopenia. • 1 in 4 develop Grade 4 lymphopenia during radiotherapy. • In this large cohort of patients with oropharyngeal cancer, the development of lymphopenia during radiotherapy did not impact outcomes. Radiation-induced lymphopenia has been associated with poor survival outcomes in certain solid tumors such as esophageal, lung, cervical and pancreatic cancers. We aim to determine the effect of treatment-related lymphopenia during radiotherapy on outcomes of patients with oropharyngeal cancer. A retrospective analysis of all patients who completed definitive radiotherapy for oropharyngeal cancer at The University of Texas MD Anderson Cancer Center and had blood counts taken during radiotherapy from 2002 to 2013 were included. Patient, tumor and treatment characteristics, clinical outcomes and lymphocyte counts during radiotherapy were recorded. Lymphopenia was graded according to the CTCAE v4.0. Survival rates were estimated using the Kaplan-Meier method and compared with log-rank tests. 850 patients were evaluated. The median age was 57 years. The majority of the cohort had p16/HPV-positive disease (71%), 8% had HPV-negative disease and 21% were unknown. The median radiation total dose was 70 Gy. 45% of patients had induction chemotherapy, and 87% had concurrent chemotherapy. 703 (83%) patients developed ≥grade 3 (G3) lymphopenia and 209 (25%) had grade 4 (G4) lymphopenia during radiotherapy. The median follow-up was 59 months; the 5-year overall survival rate was 81%. There were no significant differences in overall survival rates nor in disease control rates, in those who developed G3/G4 lymphopenia compared with those who did not. No significant effect of lymphopenia on survival was observed when analyzed according to p16/HPV status. In this large cohort of patients with oropharyngeal cancer, the development of lymphopenia during radiotherapy did not impact outcomes. [ABSTRACT FROM AUTHOR]

• Retrospective analysis of 695 HPV+ oropharyngeal cancer patients. • Risk prediction using machine learning to predict toxicity and survival. • Pre-treatment anatomical information improves standard (AJCC-8) risk prediction. • Machine-learning derived patient stratifications improve outcome prediction. Evaluate the effectiveness of machine learning tools that incorporate spatial information such as disease location and lymph node metastatic patterns-of-spread, for prediction of survival and toxicity in HPV+ oropharyngeal cancer (OPC). 675 HPV+ OPC patients that were treated at MD Anderson Cancer Center between 2005 and 2013 with curative intent IMRT were retrospectively collected under IRB approval. Risk stratifications incorporating patient radiometric data and lymph node metastasis patterns via an anatomically-adjacent representation with hierarchical clustering were identified. These clusterings were combined into a 3-level patient stratification and included along with other known clinical features in a Cox model for predicting survival outcomes, and logistic regression for toxicity, using independent subsets for training and validation. Four groups were identified and combined into a 3-level stratification. The inclusion of patient stratifications in predictive models for 5-yr Overall survival (OS), 5-year recurrence free survival, (RFS) and Radiation-associated dysphagia (RAD) consistently improved model performance measured using the area under the curve (AUC). Test set AUC improvements over models with clinical covariates, was 9 % for predicting OS, and 18 % for predicting RFS, and 7 % for predicting RAD. For models with both clinical and AJCC covariates, AUC improvement was 7 %, 9 %, and 2 % for OS, RFS, and RAD, respectively. Including data-driven patient stratifications considerably improve prognosis for survival and toxicity outcomes over the performance achieved by clinical staging and clinical covariates alone. These stratifications generalize well to across cohorts, and sufficient information for reproducing these clusters is included. [ABSTRACT FROM AUTHOR]

Purpose: Automating and standardizing the contouring of clinical target volumes (CTVs) can reduce interphysician variability, which is one of the largest sources of uncertainty in head and neck radiation therapy. In addition to using uniform margin expansions to auto-delineate high-risk CTVs, very little work has been performed to provide patient- and disease-specific high-risk CTVs. The aim of the present study was to develop a deep neural network for the auto-delineation of high-risk CTVs.Methods and Materials: Fifty-two oropharyngeal cancer patients were selected for the present study. All patients were treated at The University of Texas MD Anderson Cancer Center from January 2006 to August 2010 and had previously contoured gross tumor volumes and CTVs. We developed a deep learning algorithm using deep auto-encoders to identify physician contouring patterns at our institution. These models use distance map information from surrounding anatomic structures and the gross tumor volume as input parameters and conduct voxel-based classification to identify voxels that are part of the high-risk CTV. In addition, we developed a novel probability threshold selection function, based on the Dice similarity coefficient (DSC), to improve the generalization of the predicted volumes. The DSC-based function is implemented during an inner cross-validation loop, and probability thresholds are selected a priori during model parameter optimization. We performed a volumetric comparison between the predicted and manually contoured volumes to assess our model.Results: The predicted volumes had a median DSC value of 0.81 (range 0.62-0.90), median mean surface distance of 2.8 mm (range 1.6-5.5), and median 95th Hausdorff distance of 7.5 mm (range 4.7-17.9) when comparing our predicted high-risk CTVs with the physician manual contours.Conclusions: These predicted high-risk CTVs provided close agreement to the ground-truth compared with current interobserver variability. The predicted contours could be implemented clinically, with only minor or no changes. [ABSTRACT FROM AUTHOR]