Search

Summary: Background: Only around 20–30% of patients with non-small-cell lung cancer (NCSLC) have durable benefit from immune-checkpoint inhibitors. Although tissue-based biomarkers (eg, PD-L1) are limited by suboptimal performance, tissue availability, and tumour heterogeneity, radiographic images might holistically capture the underlying cancer biology. We aimed to investigate the application of deep learning on chest CT scans to derive an imaging signature of response to immune checkpoint inhibitors and evaluate its added value in the clinical context. Methods: In this retrospective modelling study, 976 patients with metastatic, EGFR/ALK negative NSCLC treated with immune checkpoint inhibitors at MD Anderson and Stanford were enrolled from Jan 1, 2014, to Feb 29, 2020. We built and tested an ensemble deep learning model on pretreatment CTs (Deep-CT) to predict overall survival and progression-free survival after treatment with immune checkpoint inhibitors. We also evaluated the added predictive value of the Deep-CT model in the context of existing clinicopathological and radiological metrics. Findings: Our Deep-CT model demonstrated robust stratification of patient survival of the MD Anderson testing set, which was validated in the external Stanford set. The performance of the Deep-CT model remained significant on subgroup analyses stratified by PD-L1, histology, age, sex, and race. In univariate analysis, Deep-CT outperformed the conventional risk factors, including histology, smoking status, and PD-L1 expression, and remained an independent predictor after multivariate adjustment. Integrating the Deep-CT model with conventional risk factors demonstrated significantly improved prediction performance, with overall survival C-index increases from 0·70 (clinical model) to 0·75 (composite model) during testing. On the other hand, the deep learning risk scores correlated with some radiomics features, but radiomics alone could not reach the performance level of deep learning, indicating that the deep learning model effectively captured additional imaging patterns beyond known radiomics features. Interpretation: This proof-of-concept study shows that automated profiling of radiographic scans through deep learning can provide orthogonal information independent of existing clinicopathological biomarkers, bringing the goal of precision immunotherapy for patients with NSCLC closer. Funding: National Institutes of Health, Mark Foundation Damon Runyon Foundation Physician Scientist Award, MD Anderson Strategic Initiative Development Program, MD Anderson Lung Moon Shot Program, Andrea Mugnaini, and Edward L C Smith.

Purpose: The contributory effects of radiation dose to different brain regions on neurocognitive performance after radiation therapy (RT) for primary brain tumors is not well known. Methods and Materials: In this retrospective cohort study, 30 patients with brain tumors treated with photon RT were identified, and radiation dosimetric parameters across brain regions were calculated. All patients had longitudinal neurocognitive evaluations at baseline and after treatment. Generalized estimating equations were used to model each neurocognitive endpoint over time in a multivariable analysis, while adjusted for multiple comparisons of brain regions. Results: Median follow-up from RT to last assessment was 4.1 years. Fewer years of formal education and older age at the time of RT were associated with lower scores in language, verbal memory, and working memory, after adjustment for baseline scores in multivariable analyses. Higher radiation dose to specific brain regions was not associated with declines in any of the evaluated cognitive domains. On average, there was no clinically significant decline (magnitude of z score change >1) between first and last neurocognitive evaluation. Across each individual cognitive domain, fewer than 15% of patients were impaired at most recent follow-up. Conclusions: In this small study of 30 patients treated with RT for a primary brain tumor, brain region dosimetry was not associated with decline in cognitive performance. Older age at time of RT and fewer years of formal education were associated with declines in cognitive performance, suggesting that effects of nondosimetric factors on cognitive performance should be considered alongside treatment factors and dosimetry in neuro-oncology research.

Purpose: We present the case of a 48-year-old patient with recurrent World Health Organization grade II meningioma in the left occipital region who underwent a subtotal resection, followed by postoperative proton therapy to residual disease and the resection cavity. Fourteen months after radiation treatment completion, surveillance imaging revealed numerous ring-enhancing infratentorial lesions, both within and outside of the high-dose field, of concern for viable tumor. We describe the use of advanced imaging and proton biologic effectiveness analyses to enable the diagnosis of radiation necrosis (RN) and ascertain intrinsic physical factors contributing to the development of RN in this patient. Methods and Materials: Multiparametric magnetic resonance imaging (MRI) and Monte Carlo predictions of linear energy transfer (LET) and variable relative biologic effectiveness dose were performed. Results: The dosimetric analysis revealed that of the 10 lesions, 9 were located outside of the clinical treatment volume and 6 received a dose of

Purpose: This study aimed to report on the safety, feasibility, and workflow of using magnetic resonance imaging (MRI) simulation, while immobilized in the treatment position, for radiation therapy treatment planning in the pediatric population. Methods and Materials: Between May and December 2017, 10 pediatric patients completed both MRI and computed tomography imaging simulation in treatment immobilization for radiation therapy planning for central nervous system disease. We report our initial institutional experience and workflow of the use of MRI simulation in immobilization for treatment planning in this population. Results: Ten pediatric patients successfully underwent MRI and computed tomography imaging simulation for CNS disease. Two patients required anesthesia for sedation during the simulations. From our initial experience, MRI simulation was tolerated by all 10 pediatric patients without any safety or clinical issues, including those who required anesthesia. Conclusions: Our initial experience supports the use of MRI simulation for radiation treatment planning in the pediatric population, with and without anesthetic sedation, as a safe and feasible image-guidance tool. This is particularly useful in the treatment of pediatric patients because MRI simulation enables superior, soft-tissue, anatomic imaging for a more robust delineation of organs at risk and target volumes without increasing radiation exposure.

Objectives: Stereotactic radiosurgery is a common treatment for brain metastases and is typically planned on magnetic resonance imaging (MRI). However, the MR acquisition parameters used for patient selection and treatment planning for stereotactic radiosurgery can vary within and across institutions. In this work, we investigate the effect of MRI slice thickness on the detection and contoured volume of metastatic lesions in the brain. Methods and Materials: A retrospective cohort of 28 images acquired with a slice thickness of 1 mm were resampled to simulate acquisitions at 2- and 3-mm slice thickness. A total of 102 metastases ranging from 0.0030 cc to 5.08 cc (75-percentile 0.36 cc) were contoured on the original images. All 3 sets of images were recontoured by experienced physicians. Results: Of all the images detected and contoured on the 1 mm images, 3% of lesions were missed on the 2 mm images, and 13% were missed on the 3 mm images. One lesion that was identified on both the 2 mm and 3 mm images was determined to be a blood vessel on the 1 mm images. Additionally, the lesions were contoured 11% larger on the 2 mm and 43% larger on the 3 mm images. Conclusions: Using images with a slice thickness >1 mm effects detection and segmentation of brain lesions, which can have an important effect on patient management and treatment outcomes.