Localization accuracy from automatic and semi-automatic rigid registration of locally-advanced lung cancer targets during image-guided radiation therapy.

PURPOSE: To evaluate localization accuracy resulting from rigid registration of locally-advanced lung cancer targets using fully automatic and semi-automatic protocols for image-guided radiation therapy. METHODS: Seventeen lung cancer patients, fourteen also presenting with involved lymph nodes, received computed tomography (CT) scans once per week throughout treatment under active breathing control. A physician contoured both lung and lymph node targets for all weekly scans. Various automatic and semi-automatic rigid registration techniques were then performed for both individual and simultaneous alignments of the primary gross tumor volume (GTV(P)) and involved lymph nodes (GTV(LN)) to simulate the localization process in image-guided radiation therapy. Techniques included "standard" (direct registration of weekly images to a planning CT), "seeded" (manual prealignment of targets to guide standard registration), "transitive-based" (alignment of pretreatment and planning CTs through one or more intermediate images), and "rereferenced" (designation of a new reference image for registration). Localization error (LE) was assessed as the residual centroid and border distances between targets from planning and weekly CTs after registration. RESULTS: Initial bony alignment resulted in centroid LE of 7.3 ± 5.4 mm and 5.4 ± 3.4 mm for the GTV(P) and GTV(LN), respectively. Compared to bony alignment, transitive-based and seeded registrations significantly reduced GTV(P) centroid LE to 4.7 ± 3.7 mm (p = 0.011) and 4.3 ± 2.5 mm (p < 1 × 10(-3)), respectively, but the smallest GTV(P) LE of 2.4 ± 2.1 mm was provided by rereferenced registration (p < 1 × 10(-6)). Standard registration significantly reduced GTV(LN) centroid LE to 3.2 ± 2.5 mm (p < 1 × 10(-3)) compared to bony alignment, with little additional gain offered by the other registration techniques. For simultaneous target alignment, centroid LE as low as 3.9 ± 2.7 mm and 3.8 ± 2.3 mm were achieved for the GTV(P) and GTV(LN), respectively, using rereferenced registration. CONCLUSIONS: Target shape, volume, and configuration changes during radiation therapy limited the accuracy of standard rigid registration for image-guided localization in locally-advanced lung cancer. Significant error reductions were possible using other rigid registration techniques, with LE approaching the lower limit imposed by interfraction target variability throughout treatment.

Radio-morphology: Parametric shape-based features in radiotherapy.

PURPOSE: In radiotherapy, it is necessary to characterize dose over the patient anatomy to target areas and organs at risk. Current tools provide methods to describe dose in terms of percentage of volume and magnitude of dose, but are limited by assumptions of anatomical homogeneity within a region of interest (ROI) and provide a non-spatially aware description of dose. A practice termed radio-morphology is proposed as a method to apply anatomical knowledge to parametrically derive new shapes and substructures from a normalized set of anatomy, ensuring consistently identifiable spatially aware features of the dose across a patient set. METHODS: Radio-morphologic (RM) features are derived from a three-step procedure: anatomy normalization, shape transformation, and dose calculation. Predefined ROI's are mapped to a common anatomy, a series of geometric transformations are applied to create new structures, and dose is overlaid to the new images to extract dosimetric features; this feature computation pipeline characterizes patient treatment with greater anatomic specificity than current methods. RESULTS: Examples of applications of this framework to derive structures include concentric shells based around expansions and contractions of the parotid glands, separation of the esophagus into slices along the z-axis, and creating radial sectors to approximate neurovascular bundles surrounding the prostate. Compared to organ-level dose-volume histograms (DVHs), using derived RM structures permits a greater level of control over the shapes and anatomical regions that are studied and ensures that all new structures are consistently identified. Using machine learning methods, these derived dose features can help uncover dose dependencies of inter- and intra-organ regions. Voxel-based and shape-based analysis of the parotid and submandibular glands identified regions that were predictive of the development of high-grade xerostomia (CTCAE grade 2 or greater) at 3-6 months post treatment. CONCLUSIONS: Radio-morphology is a valuable data mining tool that approaches radiotherapy data in a new way, improving the study of radiotherapy to potentially improve prognostic and predictive accuracy. Further applications of this methodology include the use of parametrically derived sub-volumes to drive radiotherapy treatment planning.

Evaluation of classification and regression tree (CART) model in weight loss prediction following head and neck cancer radiation therapy.

OBJECTIVE: We explore whether a knowledge-discovery approach building a Classification and Regression Tree (CART) prediction model for weight loss (WL) in head and neck cancer (HNC) patients treated with radiation therapy (RT) is feasible. METHODS AND MATERIALS: HNC patients from 2007 to 2015 were identified from a prospectively collected database Oncospace. Two prediction models at different time points were developed to predict weight loss ≥5 kg at 3 months post-RT by CART algorithm: (1) during RT planning using patient demographic, delineated dose data, planning target volume-organs at risk shape relationships data and (2) at the end of treatment (EOT) using additional on-treatment toxicities and quality of life data. RESULTS: Among 391 patients identified, WL predictors during RT planning were International Classification of Diseases diagnosis; dose to masticatory and superior constrictor muscles, larynx, and parotid; and age. At EOT, patient-reported oral intake, diagnosis, N stage, nausea, pain, dose to larynx, parotid, and low-dose planning target volume-larynx distance were significant predictive factors. The area under the curve during RT and EOT was 0.773 and 0.821, respectively. CONCLUSIONS: We demonstrate the feasibility and potential value of an informatics infrastructure that has facilitated insight into the prediction of WL using the CART algorithm. The prediction accuracy significantly improved with the inclusion of additional treatment-related data and has the potential to be leveraged as a strategy to develop a learning health system.

Improving prediction of surgical resectability over current staging guidelines in patients with pancreatic cancer who receive stereotactic body radiation therapy.

PURPOSE: For patients with localized pancreatic cancer (PC) with vascular involvement, prediction of resectability is critical to define optimal treatment. However, the current definitions of borderline resectable (BR) and locally advanced (LA) disease leave considerable heterogeneity in outcomes within these classifications. Moreover, factors beyond vascular involvement likely affect the ability to undergo resection. Herein, we share our experience developing a model that incorporates detailed radiologic, patient, and treatment factors to predict surgical resectability in patients with BR and LA PC who undergo stereotactic body radiation therapy (SBRT). METHODS AND MATERIALS: Patients with BR or LA PC who were treated with SBRT between 2010 and 2016 were included. The primary endpoint was margin negative resection, and predictors included age, sex, race, treatment year, performance status, initial staging, tumor volume and location, baseline and pre-SBRT carbohydrate antigen 19-9 levels, chemotherapy regimen and duration, and radiation dose. In addition, we characterized the relationship between tumors and key arteries (superior mesenteric, celiac, and common hepatic arteries), using overlap volume histograms derived from computed tomography data. A classification and regression tree was built, and leave-one-out cross-validation was performed. Prediction of surgical resection was compared between our model and staging in accordance with the National Comprehensive Care Network guidelines using McNemar's test. RESULTS: A total of 191 patients were identified (128 patients with LA and 63 with BR), of which 87 patients (46%) underwent margin negative resection. The median total dose was 33 Gy. Predictors included the chemotherapy regimen, amount of arterial involvement, and age. Importantly, radiation dose that covers 95% of gross tumor volume (GTV D95), was a key predictor of resectability in certain subpopulations, and the model showed improved accuracy in the prediction of margin negative resection compared with National Comprehensive Care Network guideline staging (75% vs 63%; P < .05). CONCLUSIONS: We demonstrate the ability to improve prediction of surgical resectabiliy beyond the current staging guidelines, which highlights the value of assessing vascular involvement in a continuous manner. In addition, we show an association between radiation dose and resectability, which suggests the potential importance of radiation to allow for resection in certain populations. External data are needed for validation and to increase the robustness of the model.

Effects of perineural invasion on biochemical recurrence and prostate cancer-specific survival in patients treated with definitive external beam radiotherapy.

OBJECTIVES: Perineural invasion (PNI) has not yet gained universal acceptance as an independent predictor of adverse outcomes for prostate cancer treated with external beam radiotherapy (EBRT). We analyzed the prognostic influence of PNI for a large institutional cohort of prostate cancer patients who underwent EBRT with and without androgen deprivation therapy (ADT). MATERIAL AND METHODS: We, retrospectively, reviewed prostate cancer patients treated with EBRT from 1993 to 2007 at our institution. The primary endpoint was biochemical failure-free survival (BFFS), with secondary endpoints of metastasis-free survival (MFS), prostate cancer-specific survival (PCSS), and overall survival (OS). Univariate and multivariable Cox proportional hazards models were constructed for all survival endpoints. Hazard ratios for PNI were analyzed for the entire cohort and for subsets defined by NCCN risk level. Additionally, Kaplan-Meier survival curves were generated for all survival endpoints after stratification by PNI status, with significant differences computed using the log-rank test. RESULTS: Of 888 men included for analysis, PNI was present on biopsy specimens in 187 (21.1%). PNI was associated with clinical stage, pretreatment PSA level, biopsy Gleason score, and use of ADT (all P<0.01). Men with PNI experienced significantly inferior 10-year BFFS (40.0% vs. 57.8%, P = 0.002), 10-year MFS (79.7% vs. 89.0%, P = 0.001), and 10-year PCSS (90.9% vs. 95.9%, P = 0.009), but not 10-year OS (67.5% vs. 77.5%, P = 0.07). On multivariate analysis, PNI was independently associated with inferior BFFS (P<0.001), but not MFS, PCSS, or OS. In subset analysis, PNI was associated with inferior BFFS (P = 0.04) for high-risk patients and with both inferior BFFS (P = 0.01) and PCSS (P = 0.05) for low-risk patients. Biochemical failure occurred in 33% of low-risk men with PNI who did not receive ADT compared to 8% for low-risk men with PNI treated with ADT (P = 0.01). CONCLUSION: PNI was an independently significant predictor of adverse survival outcomes in this large institutional cohort, particularly for patients with NCCN low-risk disease. PNI should be carefully considered along with other standard prognostic factors when treating these patients with EBRT. Supplementing EBRT with ADT may be beneficial for select low-risk patients with PNI though independent validation with prospective studies is recommended.

Using Big Data Analytics to Advance Precision Radiation Oncology.

Big clinical data analytics as a primary component of precision medicine is discussed, identifying where these emerging tools fit in the spectrum of genomics and radiomics research. A learning health system (LHS) is conceptualized that uses clinically acquired data with machine learning to advance the initiatives of precision medicine. The LHS is comprehensive and can be used for clinical decision support, discovery, and hypothesis derivation. These developing uses can positively impact the ultimate management and therapeutic course for patients. The conceptual model for each use of clinical data, however, is different, and an overview of the implications is discussed. With advancements in technologies and culture to improve the efficiency, accuracy, and breadth of measurements of the patient condition, the concept of an LHS may be realized in precision radiation therapy.

Very High-Risk Localized Prostate Cancer: Outcomes Following Definitive Radiation.

PURPOSE: Existing definitions of high-risk prostate cancer consist of men who experience significant heterogeneity in outcomes. As such, criteria that identify a subpopulation of National Comprehensive Cancer Network (NCCN) high-risk prostate cancer patients who are at very high risk (VHR) for poor survival outcomes following prostatectomy were recently developed at our institution and include the presence of any of the following disease characteristics: multiple NCCN high-risk factors, primary Gleason pattern 5 disease and/or ≥5 biopsy cores with Gleason sums of 8 to 10. Whether these criteria also apply to men undergoing definitive radiation is unclear, as is the optimal treatment regimen in these patients. METHODS AND MATERIALS: All men consecutively treated with definitive radiation by a single provider from 1993 to 2006 and who fulfilled criteria for NCCN high-risk disease were identified (n=288), including 99 patients (34%) with VHR disease. Multivariate-adjusted competing risk regression models were constructed to assess associations between the VHR definition and biochemical failure (BF), distant metastasis (DM), and prostate cancer-specific mortality (PCSM). Multivariate-adjusted Cox regression analysis assessed the association of the VHR definition with overall mortality (OM). Cumulative incidences of failure endpoints were compared between VHR men and other NCCN high-risk men. RESULTS: Men with VHR disease compared to other NCCN high-risk men experienced a higher 10-year incidence of BF (54.0% vs 35.4%, respectively, P<.001), DM (34.9% vs 13.4%, respectively, P<.001), PCSM (18.5% vs 5.9%, respectively, P<.001), and OM (36.4% vs 27.0%, respectively, P=.04). VHR men with a detectable prostate-specific antigen (PSA) concentration at the end of radiation (EOR) remained at high risk of 10-year PCSM compared to VHR men with an undetectable EOR PSA (31.0% vs 13.7%, respectively, P=.05). CONCLUSIONS: NCCN high-risk prostate cancer patients who meet VHR criteria experience distinctly worse outcomes following definitive radiation and long-term androgen deprivation therapy, particularly if an EOR PSA is detectable. Optimal use of local therapies for VHR patients should be explored further, as should novel agents.

A data-mining framework for large scale analysis of dose-outcome relationships in a database of irradiated head and neck cancer patients.

PURPOSE: To develop a hypothesis-generating framework for automatic extraction of dose-outcome relationships from an in-house, analytic oncology database. METHODS: Dose-volume histograms (DVH) and clinical outcomes have been routinely stored to the authors' database for 684 head and neck cancer patients treated from 2007 to 2014. Database queries were developed to extract outcomes that had been assessed for at least 100 patients, as well as DVH curves for organs-at-risk (OAR) that were contoured for at least 100 patients. DVH curves for paired OAR (e.g., left and right parotids) were automatically combined and included as additional structures for analysis. For each OAR-outcome combination, only patients with both OAR and outcome records were analyzed. DVH dose points, DVt, at a given normalized volume threshold Vt were stratified into two groups based on severity of toxicity outcomes after treatment completion. The probability of an outcome was modeled at each Vt = [0%, 1%, …, 100%] by logistic regression. Notable OAR-outcome combinations were defined as having statistically significant regression parameters (p < 0.05) and an odds ratio of at least 1.05 (5% increase in odds per Gy). RESULTS: A total of 57 individual and combined structures and 97 outcomes were queried from the database. Of all possible OAR-outcome combinations, 17% resulted in significant logistic regression fits (p < 0.05) having an odds ratio of at least 1.05. Further manual inspection revealed a number of reasonable models based on either reported literature or proximity between neighboring OARs. The data-mining algorithm confirmed the following well-known OAR-dose/outcome relationships: dysphagia/larynx, voice changes/larynx, esophagitis/esophagus, xerostomia/parotid glands, and mucositis/oral mucosa. Several surrogate relationships, defined as OAR not directly attributed to an outcome, were also observed, including esophagitis/larynx, mucositis/mandible, and xerostomia/mandible. CONCLUSIONS: Prospective collection of clinical data has enabled large-scale analysis of dose-outcome relationships. The current data-mining framework revealed both known and novel dosimetric and clinical relationships, underscoring the potential utility of this analytic approach in hypothesis generation. Multivariate models and advanced, 3D dosimetric features may be necessary to further evaluate the complex relationship between neighboring OAR and observed outcomes.

A block matching-based registration algorithm for localization of locally advanced lung tumors.

PURPOSE: To implement and evaluate a block matching-based registration (BMR) algorithm for locally advanced lung tumor localization during image-guided radiotherapy. METHODS: Small (1 cm(3)), nonoverlapping image subvolumes ("blocks") were automatically identified on the planning image to cover the tumor surface using a measure of the local intensity gradient. Blocks were independently and automatically registered to the on-treatment image using a rigid transform. To improve speed and robustness, registrations were performed iteratively from coarse to fine image resolution. At each resolution, all block displacements having a near-maximum similarity score were stored. From this list, a single displacement vector for each block was iteratively selected which maximized the consistency of displacement vectors across immediately neighboring blocks. These selected displacements were regularized using a median filter before proceeding to registrations at finer image resolutions. After evaluating all image resolutions, the global rigid transform of the on-treatment image was computed using a Procrustes analysis, providing the couch shift for patient setup correction. This algorithm was evaluated for 18 locally advanced lung cancer patients, each with 4-7 weekly on-treatment computed tomography scans having physician-delineated gross tumor volumes. Volume overlap (VO) and border displacement errors (BDE) were calculated relative to the nominal physician-identified targets to establish residual error after registration. RESULTS: Implementation of multiresolution registration improved block matching accuracy by 39% compared to registration using only the full resolution images. By also considering multiple potential displacements per block, initial errors were reduced by 65%. Using the final implementation of the BMR algorithm, VO was significantly improved from 77% ± 21% (range: 0%-100%) in the initial bony alignment to 91% ± 8% (range: 56%-100%;p < 0.001). Left-right, anterior-posterior, and superior-inferior systematic BDE were 3.2, 2.4, and 4.4 mm, respectively, with random BDE of 2.4, 2.1, and 2.7 mm. Margins required to include both localization and delineation uncertainties ranged from 5.0 to 11.7 mm, an average of 40% less than required for bony alignment. CONCLUSIONS: BMR is a promising approach for automatic lung tumor localization. Further evaluation is warranted to assess the accuracy and robustness of BMR against other potential localization strategies.

Tumor, lymph node, and lymph node-to-tumor displacements over a radiotherapy series: analysis of interfraction and intrafraction variations using active breathing control (ABC) in lung cancer.

PURPOSE: To estimate errors in soft tissue-based image guidance due to relative changes between primary tumor (PT) and affected lymph node (LN) position and volume, and to compare the results with bony anatomy-based displacements of PTs and LNs during radiotherapy of lung cancer. METHODS AND MATERIALS: Weekly repeated breath-hold computed tomography scans were acquired in 17 lung cancer patients undergoing radiotherapy. PTs and affected LNs were manually contoured on all scans after rigid registration. Interfraction and intrafraction displacements in the centers of mass of PTs and LNs relative to bone, as well as LNs relative to PTs (LN-PT), were calculated. RESULTS: The mean volume after 5 weeks was 65% for PTs and 63% for LNs. Systematic and random interfraction displacements were 2.6 to 4.6 mm and 2.7 to 2.9 mm, respectively, for PTs; 2.4 to 3.8 mm and 1.4 to 2.7 mm, respectively, for LNs; and 2.3 to 3.9 mm and 1.9 to 2.8 mm, respectively, for LN-PT. Systematic and random intrafraction displacements were less than 1 mm except in the superoinferior direction. Interfraction LN-PT displacements greater than 3 mm were observed in 67% of fractions and require a safety margin of 12 mm in the lateral direction, 11 mm in the anteroposterior direction, and 9 mm in the superoinferior direction. LN-PT displacements displayed significant time trends (p < 0.0001) and depended on the presence of pathoanatomic conditions of the ipsilateral lung, such as atelectasis. CONCLUSION: Interfraction LN-PT displacements were mostly systematic and comparable to bony anatomy-based displacements of PTs or LNs alone. Time trends, large volume changes, and the influence of pathoanatomic conditions underline the importance of soft tissue-based image guidance and the potential of plan adaptation.