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.

Esophageal adenocarcinoma with any component of signet ring cells portends poor prognosis and response to neoadjuvant therapy.

BACKGROUND: Multiple investigations have shown inferior outcomes for esophageal cancer patients with signet ring cell (SRC) histology. Traditionally, SRC adenocarcinoma has been defined by ≥50% of the tumor composed of SRC. We hypothesized that patients with SRC even <50% would show resistance to standard multimodality therapy with poorer long-term outcomes. METHODS: Patients treated with trimodality therapy for adenocarcinoma from 2006 to 2018 were evaluated for SRC on pretreatment biopsy specimens. Available hematoxylin and eosin slides containing SRC tumors were re-reviewed by an esophageal pathologist to quantify the percent composition of SRC. RESULTS: SRC histology was identified on at least 1 pathologic specimen in 106 of 819 (13%) patients. Rates of pathologic complete response (pCR) among usual-type and SRC tumors were 25% (177/713) and 10% (11/106), respectively (P = .006). The pretreatment SRC components did not independently affect the rate of pCR (1%-10% SRC: 4% [2/46] pCR; 11%-49% SRC: 25% [7/28] pCR; 50%-100% SRC: 7% [2/30] pCR). Kaplan-Meier analysis demonstrated worse survival among patients with any degree of SRC present on pretreatment biopsy, as compared with usual-type esophageal adenocarcinoma (P < .0001). Cox multivariable analysis failed to identify a relationship between increasing SRC component and poorer survival. CONCLUSIONS: We present the only known evaluation of the percentage of SRC component in esophageal carcinoma. Our data support the hypothesis that esophageal adenocarcinoma with any component of SRC are more resistant to chemoradiation with poorer survival. Pathologic reporting of esophageal adenocarcinoma should include any component of SRC. Alternative therapies in patients with any SRC component may be indicated.

Preoperative Nomogram to Risk Stratify Patients for the Benefit of Trimodality Therapy in Esophageal Adenocarcinoma.

PURPOSE: To develop a nomogram that estimates 1-year recurrence-free survival (RFS) after trimodality therapy for esophageal adenocarcinoma and to assess the overall survival (OS) benefit of esophagectomy after chemoradiotherapy (CRT) on the basis of 1-year recurrence risk. METHODS: In total, 568 consecutive patients with potentially resectable esophageal adenocarcinoma who underwent CRT were included for analysis, including 373 patients who underwent esophagectomy after CRT (trimodality therapy), and 195 who did not undergo surgery (bimodality therapy). A nomogram for 1-year RFS was created using a Cox regression model. The upper tertile of the nomogram score was used to stratify patients in low-risk and high-risk groups for 1-year recurrence. The 5-year OS was compared between trimodality and bimodality therapy in low-risk and high-risk patients after propensity score matching, respectively. RESULTS: Median follow-up for the entire cohort was 62 months. The 5-year OS in the trimodality and bimodality treatment groups was 56.3% (95% confidence interval [CI] 47.9-64.7) and 36.9% (95% CI 31.4-42.4), respectively. The final nomogram for the prediction of 1-year RFS included male gender, poor histologic grade, signet ring cell adenocarcinoma, cN1, cN2-3, and baseline SUVmax, with accurate calibration and reasonable discrimination (C-statistic: 0.66). Trimodality therapy was associated with improved 5-year OS in low-risk patients (p = 0.003), whereas it showed no significant survival benefit in high-risk patients (p = 0.302). CONCLUSIONS: The proposed nomogram estimates early recurrence risk. The addition of surgery to CRT provides a clear OS benefit in low-risk patients. The OS benefit of surgery in high-risk patients is less pronounced.

Biological responses of human solid tumor cells to X-ray irradiation within a 1.5-Tesla magnetic field generated by a magnetic resonance imaging-linear accelerator.

Devices that combine magnetic resonance imaging with linear accelerators (MRL) represent a novel tool for MR-guided radiotherapy. However, whether magnetic fields (MFs) generated by these devices affect the radiosensitivity of tumors is unknown. We investigated the influence of a 1.5-T MF on cell viability and radioresponse of human solid tumors. Human head/neck cancer and lung cancer cells were exposed to single or fractionated 6-MV X-ray radiation; effects of the MF on cell viability were determined by cell plating efficiency and on radioresponsiveness by clonogenic cell survival. Doses needed to reduce the fraction of surviving cells to 37% of the initial value (D0s) were calculated for multiple exposures to MF and radiation. Results were analyzed using Student's t-tests. Cell viability was no different after single or multiple exposures to MRL than after exposure to a conventional linear accelerator (Linac, without MR-generated MF) in 12 of 15 experiments (all P > 0.05). Single or multiple exposures to MF had no influence on cell radioresponse (all P > 0.05). Cells treated up to four times with an MRL or a Linac further showed no changes in D0s with MF versus without MF (all P > 0.05). In conclusion, MF within the MRL does not seem to affect in vitro tumor radioresponsiveness as compared with a conventional Linac. Bioelectromagnetics. 37:471-480, 2016. © 2016 Wiley Periodicals, Inc.

Outcomes of adjuvant chemoradiation after pancreaticoduodenectomy with mesenterico-portal vein resection for adenocarcinoma of the pancreas.

PURPOSE: Surgery followed by chemotherapy and radiation (CRT) offers patients with pancreatic adenocarcinoma a chance for extended survival. In some patients, however, resection is difficult because of vascular involvement by the carcinoma, necessitating resection and grafting of the mesenterico-portal vessels. The purpose of this study was to compare outcomes between pancreaticoduodenectomy (PD) with and without mesenterico-portal vein resection (VR) in patients receiving adjuvant CRT for pancreatic adenocarcinoma. METHODS AND MATERIALS: Between 1993 and 2005, 160 patients underwent PD with 5-FU-based adjuvant CRT followed by maintenance chemotherapy at the Johns Hopkins Hospital; 20 (12.5%) of the 160 underwent VR. Clinical outcomes, including median survival, overall survival, and complication rates were assessed for both groups. RESULTS: Patients who underwent VR had significantly longer operative times (p = 0.009), greater intraoperative blood loss (p = 0.01), and longer postoperative lengths of stay (p = 0.03). However, postoperative morbidity, median survival, and overall survival rates were similar between the two groups. Most patients (70%) from both groups were able to complete CRT, and a subgroup analysis demonstrated no appreciable differences in terms of complications. None of the VR patients who received adjuvant CRT developed veno-oclusive disease or graft failure/leakage. CONCLUSION: In a cohort of patients treated with adjuvant 5-FU-based CRT at the Johns Hopkins Hospital, having a VR at the time of PD resulted in similar complication rates and survival. These data support the feasibility and safety of adjuvant CRT in patients undergoing VR at the time of PD.