Faculty Bibliography

PURPOSE/OBJECTIVE:Women treated with radiation therapy (RT) for breast cancer have an increased risk of developing radiation-associated contralateral breast cancer (CBC). Predicting CBC events is challenging due to the complex interplay of genomic, treatment, personal, and clinical factors. This study investigated computational methods that integrate genome-wide single nucleotide polymorphisms (SNPs) and non-genomic data to develop a risk stratification model for developing CBC in women treated with RT for their first primary breast cancer. METHODS:This study used a subset of the population-based WECARE Study that included 633 CBC cases and 1,253 individually matched unilateral breast cancer (UBC) controls who were treated with RT and had SNP data available from a genome-wide association study (GWAS). The study population was split into training, validation, and test sets for rigorous modeling and validation. Three data integration methods were compared in terms of their ability to stratify CBC risk: 1) Naive integration, 2) Sequential integration, and 3) Sequential iterative integration. A biological analysis of the final model was performed using gene set enrichment analysis (GSEA) and protein-protein interaction (PPI) analysis with gene annotation information informed by the model. RESULTS:The best-performing integration method was the sequential iterative integration equipped with the mixed effect random forest (MERF) algorithm. This approach achieved an area under the curve of 0.64 to stratify CBC risk in the test set, representing moderate predictive power. Calibration analysis showed good agreement between the lowest and highest risk bins stratified using sorted predicted values in the test set, resulting in an odds ratio of 3.27 for both predicted and observed CBC occurrence. GSEA and PPI analysis revealed that genes with high importance scores were associated with pathways relevant to lipid and fatty acid metabolism as well as breast cancer sensitivity to tamoxifen. CONCLUSION/CONCLUSIONS:The MERF approach demonstrated the potential for integrating high-dimensional genomic and low-dimensional non-genomic data to stratify CBC risk.

BACKGROUND:Magnetic resonance (MR) imaging-linear accelerator-based real-time adaptive planning for delivering ultra-hypofractionated stereotactic body radiotherapy (SBRT) has advanced clinical accuracy yet added complexity to the radiotherapy workflow. We retrospectively evaluated whether the addition of a Parallel Automated Contouring module with Enhancement of AI (PACE-AI) reduced contouring time and overall SBRT duration, and its impact on intra-fractional target motion. METHODS:This study included 250 fractions from 90 prostate cancer patients from which fraction time were tracked. All patients received definitive SBRT to the prostate on the 1.5-Tesla Unity (Elekta©) system, through the Adapt-To-Shape (ATS) workflow, which required real-time re-contouring of the normal tissues by the dosimetrists and target contouring by the physician for each of the fractions. We compared the fraction durations for 125 consecutive fractions treated with the incorporation of PACE-AI with 125 fractions previously treated without PACE-AI to determine whether PACE-AI improved efficiency and reduced treatment session duration. RESULTS:For the cohort treated without PACE-AI, the overall median duration was 67.6 min, including 23.9 min contouring time. With the incorporation of PACE-AI, the overall median duration was 51.2 min, representing a 24.3% reduction. The average contouring time was reduced by 55% to 10.8 min. In addition, the extent of positional shifts prior to beam delivery was significantly reduced from an average of 1.8 mm to 1.3 mm (p < 0.001) in both superior/inferior (range reduced from 0 to 6.3 mm to 0-4.2 mm) and anterior/posterior directions (range reduced from 0 to 6.6 mm to 0-5.0 mm). CONCLUSIONS:The incorporation of PACE-AI improved workflow efficiency during SBRT. The reduction in treatment duration also helped reduce organ motion during real-time adaptive planning.

PURPOSE/OBJECTIVE:This study evaluated the extent of prostate displacement during SBRT on an MRI Linac using comprehensive motion management (CMM) and identified variables associated with intrafraction motion (IFM). METHODS:212 patients with clinically localized prostate cancer were treated with 5-fraction SBRT on a 1.5 T MR-Linac where IFM was continuously tracked and gated by CMM. Pre-beam positional shifts were identified from MRI registration prior to beam delivery. Intrafraction positional variability during beam delivery was evaluated, and multivariable analysis identified variables associated with IFM. RESULTS:In 614 fractions (62.7%), a > 1.5 mm pre-beam positional shift led to an adapt-to-position (ATP) plan correction. Mean anterior-posterior and superior-inferior pre-beam shifts were 2.2 mm and 2.1 mm, respectively. For 962 evaluable fractions, the median beam-on-time was 13.7 min with a mean duty cycle of 95.8%. Sustained > 3 mm displacement was observed in 520 fractions (54.1%) with a median cumulative duration of 24 s; >5 mm displacement was observed in 209 fractions (21.7%) with a median duration of 12.4 s. The ATS + ATP workflow was associated with reduced odds of sustained > 3 mm motion (p = 0.035), while older age was associated with increased odds (p = 0.011). CONCLUSIONS:Significant prostate shifts can occur immediately prior to and during radiation beam delivery, frequently exceeding applied margins and potentially leading to tumor underdosage. Continuous motion tracking and gating during prostate SBRT is an important tool in reducing IFM and enhance treatment delivery accuracy.

BACKGROUND AND PURPOSE:Late radiation-induced hematuria can develop in prostate cancer patients undergoing radiotherapy and can negatively impact the quality-of-life of survivors. If a genetic component of risk could be modeled, this could potentially be the basis for modifying treatment for high-risk patients. We therefore investigated whether a previously developed machine learning-based modeling method using genome-wide common single nucleotide polymorphisms (SNPs) can stratify patients in terms of the risk of radiation-induced hematuria. MATERIALS AND METHODS:We applied a two-step machine learning algorithm that we previously developed for genome-wide association studies called pre-conditioned random forest regression (PRFR). PRFR includes a pre-conditioning step, producing adjusted outcomes, followed by random forest regression modeling. Data was from germline genome-wide SNPs for 668 prostate cancer patients treated with radiotherapy. The cohort was stratified only once, at the outset of the modeling process, into two groups: a training set (2/3 of samples) for modeling and a validation set (1/3 of samples). Post-modeling bioinformatics analysis was conducted to identify biological correlates plausibly associated with the risk of hematuria. RESULTS:The PRFR method achieved significantly better predictive performance compared to other alternative methods (all p < 0.05). The odds ratio between the high and low risk groups, each of which consisted of 1/3 of samples in the validation set, was 2.87 (p = 0.029), implying a clinically useful level of discrimination. Bioinformatics analysis identified six key proteins encoded by CTNND2, GSK3B, KCNQ2, NEDD4L, PRKAA1, and TXNL1 genes as well as four statistically significant biological process networks previously shown to be associated with the bladder and urinary tract. CONCLUSION:The risk of hematuria is significantly dependent on common genetic variants. The PRFR algorithm resulted in a stratification of prostate cancer patients at differential risk levels of post-radiotherapy hematuria. Bioinformatics analysis identified important biological processes involved in radiation-induced hematuria.

INTRODUCTION/BACKGROUND:Paraspinal stereotactic body radiotherapy (SBRT) involves risks of severe complications. We evaluated the safety of the paraspinal SBRT program in a large academic hospital by applying failure modes and effects analysis. METHODS:The analysis was conducted by a multidisciplinary committee (two therapists, one dosimetrist, four physicists, and two radiation oncologists). The paraspinal SBRT workflow was segmented into four phases (simulation, treatment planning, delivery, and machine quality assurance (QA)). Each phase was further divided into a sequence of sub-processes. Potential failure modes (PFM) were identified from each subprocess and scored in terms of the frequency of occurrence, severity and detectability, and a risk priority number (RPN). High-risk PFMs were identified based on RPN and were studied for root causes using fault tree analysis. RESULTS:Our paraspinal SBRT process was characterized by eight simulations, 11 treatment planning, nine delivery, and two machine QA sub-processes. There were 18, 29, 19, and eight PFMs identified from simulation, planning, treatment, and machine QA, respectively. The median RPN of the PFMs was 62.9 for simulation, 68.3 for planning, 52.9 for delivery, and 22.0 for machine QA. The three PFMs with the highest RPN were: previous radiotherapy outside the institution is not accurately evaluated (RPN: 293.3), incorrect registration between diagnostic magnetic resonance imaging and simulation computed tomography causing incorrect contours (273.0), and undetected patient movement before ExacTrac baseline (217.8). Remedies to the high RPN failures were implemented, including staff education, standardized magnetic resonance imaging acquisition parameters, and an image fusion process, and additional QA on beam steering. CONCLUSIONS:A paraspinal SBRT workflow in a large clinic was evaluated using a multidisciplinary and systematic risk analysis, which led to feasible solutions to key root causes. Treatment planning was a major source of PFMs that systematically affect the safety and quality of treatments. Accurate evaluation of external treatment records remains a challenge.

PURPOSE/OBJECTIVE:To evaluate the accuracy of surface-guided radiotherapy (SGRT) in cranial patient setup by direct comparison between optical surface imaging (OSI) and cone-beam computed tomography (CBCT), before applying SGRT-only setup for conventional radiotherapy of brain and nasopharynx cancer. METHODS AND MATERIALS/METHODS:Using CBCT as reference, SGRT setup accuracy was examined based on 269 patients (415 treatments) treated with frameless cranial stereotactic radiosurgery (SRS) during 2018-2019. Patients were immobilized in customized head molds and open-face masks and monitored using OSI during treatment. The facial skin area in planning CT was used as OSI region of interest (ROI) for automatic surface alignment and the skull was used as the landmark for automatic CBCT/CT registration. A 6 degrees of freedom (6DOF) couch was used. Immediately after CBCT setup, an OSI verification image was captured, recording the SGRT setup differences. These differences were analyzed in 6DOFs and as a function of isocenter positions away from the anterior surface to assess OSI-ROI bias. The SGRT in-room setup time was estimated and compared with CBCT and orthogonal 2D kilovoltage (2DkV) setups. RESULTS:The SGRT setup difference (magnitude) is found to be 1.0 ± 2.5 mm and 0.1˚±1.4˚ on average among 415 treatments and within 5 mm/3˚ with greater than 95% confidence level (P < 0.001). Outliers were observed for very-posterior isocenters: 15 differences (3.6%) are >5.0mm and 9 (2.2%) are >3.0˚. The setup differences show minor correlations (|r| < 0.45) between translational and rotational DOFs and a minor increasing trend (<1.0 mm) in the anterior-to-posterior direction. The SGRT setup time is 0.8 ± 0.3 min, much shorter than CBCT (5 ± 2 min) and 2DkV (2 ± 1 min) setups. CONCLUSION/CONCLUSIONS:This study demonstrates that SGRT has sufficient accuracy for fast in-room patient setup and allows real-time motion monitoring for beam holding during treatment, potentially useful to guide radiotherapy of brain and nasopharynx cancer with standard fractionation.

OBJECTIVES/OBJECTIVE:Compare a quantitative, algorithm-driven, and qualitative, pathologist-driven, scoring of radiation-induced pulmonary fibrosis (RIPF). And using these scoring models to derive preliminary comparisons on the effects of different mesenchymal stem cell (MSC) administration modalities in reducing RIPF. METHODS:25 rats were randomized into 5 groups: non-irradiated control (CG), irradiated control (CR), intraperitoneally administered granulocyte-macrophage colony stimulating factor or GM-CSF (Drug), intravascularly administered MSC (IV), and intratracheally administered MSC (IT). All groups, except CG, received an 18 Gy conformal dose to the right lung. Drug, IV and IT groups were treated immediately after irradiation. After 24 weeks of observation, rats were euthanized, their lungs excised, fixed and stained with Masson's Trichrome. Samples were anonymized and RIPF was scored qualitatively by a certified pathologist and quantitatively using ImageScope. An analysis of association was conducted, and two binary classifiers trained to validate the integrity of both qualitative and quantitative scoring. Differences between the treatment groups, as assessed by the pathologist score, were then tested by variance component analysis and mixed models for differences in RIPF outcomes. RESULTS:= 0.037, 95% CI = (0.062, 1.800]. CONCLUSION/CONCLUSIONS:Quantitative image analysis may help in the assessment of therapeutic interventions for RIPF and can serve as a scoring surrogate in differentiating between severe and mild cases of RIPF. Preliminary data demonstrate that the use of GM-CSF was best correlated with lower RIPF severity. ADVANCES IN KNOWLEDGE/CONCLUSIONS:Quantitative image analysis can be a viable supplemental system of quality control and triaging in situations where pathologist work hours or resources are limited. The use of different MSC administration methods can result in different degrees of MSC efficacy and study outcomes.

BACKGROUND:We aimed at predicting fatigue after breast cancer treatment using machine learning on clinical covariates and germline genome-wide data. METHODS: < .001) variables, a multivariable preconditioned random-forest regression model was built and validated on a hold-out subset to predict fatigue. Gene set enrichment analysis identified key biological correlates (MetaCore). All statistical tests were 2-sided. RESULTS:). CONCLUSIONS:Genomic analyses in this large cohort of breast cancer survivors suggest a possible genetic role for severe Cognitive Fatigue that warrants further exploration.

The purpose of this study was to identify germline single nucleotide polymorphisms (SNPs) that optimally predict radiation-associated contralateral breast cancer (RCBC) and to provide new biological insights into the carcinogenic process. Fifty-two women with contralateral breast cancer and 153 women with unilateral breast cancer were identified within the Women's Environmental Cancer and Radiation Epidemiology (WECARE) Study who were at increased risk of RCBC because they were ≤ 40 years of age at first diagnosis of breast cancer and received a scatter radiation dose > 1 Gy to the contralateral breast. A previously reported algorithm, preconditioned random forest regression, was applied to predict the risk of developing RCBC. The resulting model produced an area under the curve (AUC) of 0.62 (p = 0.04) on hold-out validation data. The biological analysis identified the cyclic AMP-mediated signaling and Ephrin-A as significant biological correlates, which were previously shown to influence cell survival after radiation in an ATM-dependent manner. The key connected genes and proteins that are identified in this analysis were previously identified as relevant to breast cancer, radiation response, or both. In summary, machine learning/bioinformatics methods applied to genome-wide genotyping data have great potential to reveal plausible biological correlates associated with the risk of RCBC.

PURPOSE/OBJECTIVE:Late genitourinary (GU) toxicity after radiation therapy limits the quality of life of prostate cancer survivors; however, efforts to explain GU toxicity using patient and dose information have remained unsuccessful. We identified patients with a greater congenital GU toxicity risk by identifying and integrating patterns in genome-wide single nucleotide polymorphisms (SNPs). METHODS AND MATERIALS/METHODS:We applied a preconditioned random forest regression method for predicting risk from the genome-wide data to combine the effects of multiple SNPs and overcome the statistical power limitations of single-SNP analysis. We studied a cohort of 324 prostate cancer patients who were self-assessed for 4 urinary symptoms at 2 years after radiation therapy using the International Prostate Symptom Score. RESULTS:The predictive accuracy of the method varied across the symptoms. Only for the weak stream endpoint did it achieve a significant area under the curve of 0.70 (95% confidence interval 0.54-0.86; P = .01) on hold-out validation data that outperformed competing methods. Gene ontology analysis highlighted key biological processes, such as neurogenesis and ion transport, from the genes known to be important for urinary tract functions. CONCLUSIONS:We applied machine learning methods and bioinformatics tools to genome-wide data to predict and explain GU toxicity. Our approach enabled the design of a more powerful predictive model and the determination of plausible biomarkers and biological processes associated with GU toxicity.