Predicting benefit from immune checkpoint inhibitors in patients with non-small-cell lung cancer by CT-based ensemble deep learning: a retrospective study.

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.

Administration of Hypoxia-Activated Prodrug Evofosfamide after Conventional Adjuvant Therapy Enhances Therapeutic Outcome and Targets Cancer-Initiating Cells in Preclinical Models of Colorectal Cancer.

Purpose: Cancer-initiating cells (C-IC) have been described in multiple cancer types, including colorectal cancer. C-ICs are defined by their capacity to self-renew, thereby driving tumor growth. C-ICs were initially thought to be static entities; however, recent studies have determined these cells to be dynamic and influenced by microenvironmental cues such as hypoxia. If hypoxia drives the formation of C-ICs, then therapeutic targeting of hypoxia could represent a novel means to target C-ICs.Experimental Design: Patient-derived colorectal cancer xenografts were treated with evofosfamide, a hypoxia-activated prodrug (HAP), in combination with 5-fluorouracil (5-FU) or chemoradiotherapy (5-FU and radiation; CRT). Treatment groups included both concurrent and sequential dosing regimens. Effects on the colorectal cancer-initiating cell (CC-IC) fraction were assessed by serial passage in vivo limiting dilution assays. FAZA-PET imaging was utilized as a noninvasive method to assess intratumoral hypoxia.Results: Hypoxia was sufficient to drive the formation of CC-ICs and colorectal cancer cells surviving conventional therapy were more hypoxic and C-IC-like. Using a novel approach to combination therapy, we show that sequential treatment with 5-FU or CRT followed by evofosfamide not only inhibits tumor growth of xenografts compared with 5-FU or CRT alone, but also significantly decreases the CC-IC fraction. Furthermore, noninvasive FAZA-PET hypoxia imaging was predictive of a tumor's response to evofosfamide.Conclusions: Our data demonstrate a novel means to target the CC-IC fraction by adding a HAP sequentially after conventional adjuvant therapy, as well as the use of FAZA-PET as a biomarker for hypoxia to identify tumors that will benefit most from this approach. Clin Cancer Res; 24(9); 2116-27. ©2018 AACR.

Thermosensitive liposomal cisplatin in combination with local hyperthermia results in tumor growth delay and changes in tumor microenvironment in xenograft models of lung carcinoma.

Treatment efficacy of a heat-activated thermosensitive liposome formulation of cisplatin (CDDP), known as HTLC, was determined in xenograft models of non-small-cell lung carcinoma. The short-term impact of local hyperthermia (HT) on tumor morphology, microvessel density and local inflammatory response was also evaluated. The HTLC formulation in combination with local HT resulted in a significant advantage in therapeutic effect in comparison with free drug and a non-thermosensitive liposome formulation of CDDP (i.e. LipoplatinTM) when administered at their maximum tolerated doses. Local HT-induced widespread cell necrosis and a significant reduction in microvessel density in the necrotic regions of tumors. CD11b-expressing innate leukocytes were demonstrated to infiltrate and reside preferentially at the necrotic rim of tumors, likely as a means to phagocytose-damaged tissue. Colocalization of CD11b with a marker of DNA damage (i.e. γH2AX) revealed a small portion of CD11b-expressing leukocytes that were possibly undergoing apoptosis as a result of HT-induced damage and/or the short lifespan of leukocytes. Overall, HT-induced tissue damage (i.e. at 24-h post-treatment) alone did not result in significant improvements in treatment effect, rather, the enhancement in tumor drug availability was correlated with improved therapeutic outcomes.

Technology for Innovation in Radiation Oncology.

Radiation therapy is an effective, personalized cancer treatment that has benefited from technological advances associated with the growing ability to identify and target tumors with accuracy and precision. Given that these advances have played a central role in the success of radiation therapy as a major component of comprehensive cancer care, the American Society for Radiation Oncology (ASTRO), the American Association of Physicists in Medicine (AAPM), and the National Cancer Institute (NCI) sponsored a workshop entitled "Technology for Innovation in Radiation Oncology," which took place at the National Institutes of Health (NIH) in Bethesda, Maryland, on June 13 and 14, 2013. The purpose of this workshop was to discuss emerging technology for the field and to recognize areas for greater research investment. Expert clinicians and scientists discussed innovative technology in radiation oncology, in particular as to how these technologies are being developed and translated to clinical practice in the face of current and future challenges and opportunities. Technologies encompassed topics in functional imaging, treatment devices, nanotechnology, and information technology. The technical, quality, and safety performance of these technologies were also considered. A major theme of the workshop was the growing importance of innovation in the domain of process automation and oncology informatics. The technologically advanced nature of radiation therapy treatments predisposes radiation oncology research teams to take on informatics research initiatives. In addition, the discussion on technology development was balanced with a parallel conversation regarding the need for evidence of efficacy and effectiveness. The linkage between the need for evidence and the efforts in informatics research was clearly identified as synergistic.

Expanding global access to radiotherapy.

Radiotherapy is a critical and inseparable component of comprehensive cancer treatment and care. For many of the most common cancers in low-income and middle-income countries, radiotherapy is essential for effective treatment. In high-income countries, radiotherapy is used in more than half of all cases of cancer to cure localised disease, palliate symptoms, and control disease in incurable cancers. Yet, in planning and building treatment capacity for cancer, radiotherapy is frequently the last resource to be considered. Consequently, worldwide access to radiotherapy is unacceptably low. We present a new body of evidence that quantifies the worldwide coverage of radiotherapy services by country. We show the shortfall in access to radiotherapy by country and globally for 2015-35 based on current and projected need, and show substantial health and economic benefits to investing in radiotherapy. The cost of scaling up radiotherapy in the nominal model in 2015-35 is US$26·6 billion in low-income countries, $62·6 billion in lower-middle-income countries, and $94·8 billion in upper-middle-income countries, which amounts to $184·0 billion across all low-income and middle-income countries. In the efficiency model the costs were lower: $14·1 billion in low-income, $33·3 billion in lower-middle-income, and $49·4 billion in upper-middle-income countries-a total of $96·8 billion. Scale-up of radiotherapy capacity in 2015-35 from current levels could lead to saving of 26·9 million life-years in low-income and middle-income countries over the lifetime of the patients who received treatment. The economic benefits of investment in radiotherapy are very substantial. Using the nominal cost model could produce a net benefit of $278·1 billion in 2015-35 ($265·2 million in low-income countries, $38·5 billion in lower-middle-income countries, and $239·3 billion in upper-middle-income countries). Investment in the efficiency model would produce in the same period an even greater total benefit of $365·4 billion ($12·8 billion in low-income countries, $67·7 billion in lower-middle-income countries, and $284·7 billion in upper-middle-income countries). The returns, by the human-capital approach, are projected to be less with the nominal cost model, amounting to $16·9 billion in 2015-35 (-$14·9 billion in low-income countries; -$18·7 billion in lower-middle-income countries, and $50·5 billion in upper-middle-income countries). The returns with the efficiency model were projected to be greater, however, amounting to $104·2 billion (-$2·4 billion in low-income countries, $10·7 billion in lower-middle-income countries, and $95·9 billion in upper-middle-income countries). Our results provide compelling evidence that investment in radiotherapy not only enables treatment of large numbers of cancer cases to save lives, but also brings positive economic benefits.

Heat-activated thermosensitive liposomal cisplatin (HTLC) results in effective growth delay of cervical carcinoma in mice.

Cisplatin (CDDP) has been identified as the primary chemotherapeutic agent for the treatment of cervical cancer, but dose limiting toxicity is a key issue associated with its clinical application. A suite of liposome formulations of CDDP has been developed in efforts to reduce systemic toxicity, but their therapeutic advantage over the free drug has been modest due to insufficient drug release at the tumor site. This report describes the development of a novel heat-activated thermosensitive liposome formulation containing CDDP (HTLC) designed to release approximately 90% of the loaded drug in less than 5min under mild heating conditions (42°C). Physico-chemical characteristics of HTLC were assessed in terms of gel to liquid crystalline phase transition temperature (Tm), drug loading efficiency, particle size, and stability. The pharmacokinetic profile and biodistribution of HTLC in non-tumor-bearing mice were evaluated over a 24h period. A sophisticated spatio-temporal elucidation of HTLC release in tumor-bearing mice was achieved by way of real-time monitoring using a magnetic resonance (MR) imaging protocol, wherein a custom-built laser-based conformal heat source was applied at the tumor volume to trigger the release of HTLC co-encapsulated with the MR contrast agent gadoteridol (Gd-HP-DO3A). MR thermometry (MRT) demonstrated that a relatively uniform temperature distribution was achieved in the tumor volume using the external laser-based heating setup. In mice bearing subcutaneously-implanted ME-180 cervical tumors, the combination of HTLC and heat resulted in a 2-fold increase in tumor drug levels at 1h post-administration compared to HTLC without heating. Furthermore, the overall tumor accumulation levels for the HTLC groups (with and without heat) at 1h post-injection were significantly higher than the corresponding free CDDP group. This translated into a significant improvement in therapeutic efficacy evaluated as tumor growth delay (p<0.05) for the heated HTLC treatment group compared to the unheated HTLC, heated or unheated free CDDP, and saline groups. Overall, findings from this study demonstrate that a heat-activated, triggered release formulation of CDDP results in a significant enhancement in the therapeutic index of this drug.

Online planning and delivery technique for radiotherapy of spinal metastases using cone-beam CT: image quality and system performance.

PURPOSE: To assess the feasibility of an online strategy for palliative radiotherapy (RT) of spinal bone metastasis, which integrates imaging, planning, and treatment delivery in a single step at the treatment unit. The technical challenges of this approach include cone-beam CT (CBCT) image quality for target definition, online planning, and efficient process integration. METHODS AND MATERIALS: An integrated imaging, planning, and delivery system was constructed and tested with phantoms. The magnitude of CBCT image artifacts following the use of an antiscatter grid and a nonlinear scatter correction was quantified using phantom data and images of patients receiving conventional palliative RT of the spine. The efficacy of online planning was then assessed using corrected CBCT images. Testing of the complete process was performed on phantoms with assessment of timing and dosimetric accuracy. RESULTS: The use of image corrections reduced the cupping artifact from 30% to 4.5% on CBCT images of a body phantom and improved the accuracy of CBCT numbers (water: +/- 20 Hounsfield unit [HU], and lung and bone: to within +/- 130 HU). Bony anatomy was clearly visible and was deemed sufficient for target definition. The mean total time (n = 5) for application of the online approach was 23.1 min. Image-guided dose placement was assessed using radiochromic film measurements with good agreement (within 5% of dose difference and 2 mm of distance to agreement). CONCLUSIONS: The technical feasibility of CBCT-guided online planning and delivery for palliative single treatment has been demonstrated. The process was performed in one session equivalent to an initial treatment slot (<30 min) with dosimetric accuracy satisfying accepted RT standards.

A magnetic resonance imaging study of prostate deformation relative to implanted gold fiducial markers.

PURPOSE: To describe prostate deformation during radiotherapy and determine the margins required to account for prostate deformation after setup to intraprostatic fiducial markers (FM). METHODS AND MATERIALS: Twenty-five patients with T1c-T2c prostate cancer had three gold FMs implanted. The patients presented with a full bladder and empty rectum for two axial magnetic resonance imaging (MRI) scans using a gradient recalled echo (GRE) sequence capable of imaging the FMs. The MRIs were done at the time of radiotherapy (RT) planning and a randomly assigned fraction. A single observer contoured the prostate surfaces. They were entered into a finite element model and aligned using the centroid of the three FMs. RESULTS: During RT, the prostate volume decreased by 0.5%/fraction (p = 0.03) and the FMs in-migrated by 0.05 mm/fraction (p < 0.05). Prostate deformation was unrelated to differential bladder and bowel filling, but was related to a transurethral resection of the prostate (TURP) (p = 0.003). The standard deviation for systematic uncertainty of prostate surface contouring was 0.8 mm and for FM centroid localization was 0.4 mm. The standard deviation of random interfraction prostate deformation was 1.5 mm and for FM centroid variability was 1.1 mm. These uncertainties from prostate deformation can be incorporated into a margin recipe to determine the total margins required for RT. CONCLUSIONS: During RT, the prostate exhibited: volume decrease, deformation, and in-migration of FMs. Patients with TURPs were prone to prostate deformation.

Characterization of scattered radiation in kV CBCT images using Monte Carlo simulations.

Kilovoltage (kV) cone beam computed tomography (CBCT) images suffer from a substantial scatter contribution. In this study, Monte Carlo (MC) simulations are used to evaluate the scattered radiation present in projection images. These predicted scatter distributions are also used as a scatter correction technique. Images were acquired using a kV CBCT bench top system. The EGSnrc MC code was used to model the flat panel imager, the phantoms, and the x-ray source. The x-ray source model was validated using first and second half-value layers (HVL) and profile measurements. The HVLs and the profile were found to agree within 3% and 6%, respectively. MC simulated and measured projection images for a cylindrical water phantom and for an anthropomorphic head phantom agreed within 8% and 10%. A modified version of the DOSXYZnrc MC code was used to score phase space files with identified scattered and primary particles behind the phantoms. The cone angle, the source-to-detector distance, the phantom geometry, and the energy were varied to determine their effect on the scattered radiation distribution. A scatter correction technique was developed in which the MC predicted scatter distribution is subtracted from the projections prior to reconstruction. Preliminary testing of the procedure was done with an anthropomorphic head phantom and a contrast phantom. Contrast and profile measurements were obtained for the scatter corrected and noncorrected images. An improvement of 3% for contrast between solid water and a liver insert and 11% between solid water and a Teflon insert were obtained and a significant reduction in cupping and streaking artifacts was observed.