Search

Background. Cervical cancer ranks as the 4th most common female cancer worldwide. Early stage cervical cancer patients can be treated with operation, but clinical staging system is not a good predictor of patients’ survival. We aimed to develop a novel prognostic model to predict the prognosis for operable cervical cancer patients with better accuracy than clinical staging system. Methods. A total of 13,952 operable cervical cancer patients were retrospectively enrolled in this study. The whole dataset was randomly split into a training set ( n = 9,068 , 65%), validation set ( n = 2,442 , 17.5%), and testing set ( n = 2,442 , 17.5%). Cox proportional hazard (CPH) model and random survival forest (RSF) model were used as baseline models for the prediction of overall survival (OS). Then, a deep survival learning model (DSLM) was developed for OS prediction. Finally, a novel prognostic model was explored based on this DSLM. Results. The C-indexes for the CPH and RSF model were 0.731 and 0.753, respectively. DSLM, which had four layers that had 50 neurons in each layer, achieved a C-index of 0.782 in the validation set and a C-index of 0.758 in the testing set. The novel prognostic model based on DSLM showed better performances than the conventional clinical staging system (area under receiver operating curves were 0.826 and 0.689, respectively). Personalized survival curves for individual patient using this novel model also showed notably different survival slopes. Conclusions. Our study developed a novel, practical, personalized prognostic model for operable cervical cancer patients. This novel prognostic model may have the potential to provide a more prognostic information to oncologists.

Cancer treatment is gradually entering an era of precision, with multitude studies in gene testing and immunotherapy. Tumor cells can be recognized and eliminated by the immune system through the expression of tumor-associated antigens, but when the cancer escapes or otherwise suppresses immunity, the balance between cancer cell proliferation and immune-induced cancer cell killing may be interrupted, resulting in tumor proliferation and progression. There has been significant attention to combining conventional cancer therapies (i.e., radiotherapy) with immunotherapy as opposed to treatment alone. The combination of radio-immunotherapy has been demonstrated in both basic research and clinical trials to provide more effective anti-tumor responses. However, the absolute benefits of radio-immunotherapy are dependent on individual characteristics and not all patients can benefit from radio-immunotherapy. At present, there are numerous articles about exploring the optimal models for combination radio-immunotherapy, but the factors affecting the efficacy of the combination, especially with regard to radiosensitivity remain inconclusive. Radiosensitivity is a measure of the response of cells, tissues, or individuals to ionizing radiation, and various studies have shown that the radiosensitivity index (RSI) will be a potential biomarker for predicting the efficacy of combination radio-immunotherapy. The purpose of this review is to focus on the factors that influence and predict the radiosensitivity of tumor cells, and to evaluate the impact and predictive significance of radiosensitivity on the efficacy of radio-immunotherapy combination.

The development of novel treatments for breast invasive carcinoma (BC) has been stagnant. P2RX1, a member of the purinergic receptor family, has been found to have a prognostic impact in several tumors. Therefore, we analyzed the expression pattern of P2RX1 in pan-cancers including BC and its impact on survival and found that the expression level of P2RX1 was lower in BC compared with para-cancerous tissues, and higher P2RX1 expression indicated better prognoses. But real-time quantitative reverse transcription PCR (RT-qPCR) and Western blot detected that the P2RX1 expression in normal mammary epithelial cells was lower than that in tumor cells. Then we comprehensively analyzed the regulatory mechanism and protein-protein interaction network, and found that P2RX1 was significantly positively linked with immune cell infiltration and immune checkpoints.

Accurate prognostic prediction plays a crucial role in the clinical setting. However, the TNM staging system fails to provide satisfactory individual survival prediction for stage III non-small cell lung cancer (NSCLC). The performance of the deep learning network for survival prediction in stage III NSCLC has not been explored.This study aimed to develop a deep learning-based prognostic system that could achieve better predictive performance than the existing staging system for stage III NSCLC.In this study, a deep survival learning model (DSLM) for stage III NSCLC was developed based on the Surveillance, Epidemiology, and End Results (SEER) database and was independently tested with another external cohort from our institute. DSLM was compared with the Cox proportional hazard (CPH) and random survival forest (RSF) models. A new prognostic system for stage III NSCLC was also proposed based on the established deep learning model.The study included 16,613 patients with stage III NSCLC from the SEER database. DSLM showed the best performance in survival prediction, with a C-index of 0.725 in the validation set, followed by RSF (0.688) and CPH (0.683). DSLM also showed C-indices of 0.719 and 0.665 in the internal and real-world external testing datasets, respectively. In addition, the new prognostic system based on DSLM (AUROC = 0.744) showed better performance than the TNM staging system (AUROC = 0.561).In this study, a new, integrated deep learning-based prognostic model was developed and evaluated for stage III NSCLC. This novel approach may be valuable in improving patient stratification and potentially provide meaningful prognostic information that contributes to personalized therapy.

Background Many tools have been developed to predict the efficacy of immunotherapy, such as lung immune prognostic index (LIPI), EPSILoN [Eastern Cooperative Oncology Group performance status (ECOG PS), smoking, liver metastases, lactate dehydrogenase (LDH), neutrophil‐to‐lymphocyte ratio (NLR)], and modified lung immune predictive index (mLIPI) scores. The aim of this study was to determine the ability of three predictive scores to predict the outcomes in Chinese advanced non‐small cell lung cancer (aNSCLC) patients treated with immune checkpoint inhibitors (ICIs). Methods We retrospectively analyzed 429 patients with aNSCLC treated with ICIs at our institution. The predictive ability of these models was evaluated using area under the curve (AUC) in receiver operating characteristic curve (ROC) analysis. Calibration was assessed using the Hosmer–Lemeshow test (H–L test) and Spearman's correlation coefficient. Progression‐free survival (PFS) and overall survival (OS) curves were generated using the Kaplan–Meier method. Results The AUC values of LIPI, mLIPI, and EPSILoN scores predicting PFS at 6 months were 0.642 [95% confidence interval (CI):0.590–0.694], 0.720 (95% CI: 0.675–0.762), and 0.633 (95% CI: 0.585–0.679), respectively (p, Many tools have been developed to predict the efficacy of immunotherapy, such as lung immune prognostic index (LIPI), EPSILoN [Eastern Cooperative Oncology Group performance status (ECOG PS), smoking, liver metastases, lactate dehydrogenase (LDH), neutrophil‐to‐lymphocyte ratio (NLR)], and modified lung immune predictive index (mLIPI) scores. The aim of this study was to determine the ability of to predict outcomes in Chinese aNSCLC patients treated with immune checkpoint inhibitors (ICIs).

Cation/H+ -exchanger (CHX) perform diverse functions in plants, including being a part of the protective mechanisms to cope with salt stress. GmCHX1 has been previously identified as the causal gene in a major salt tolerance quantitative trait locus (QTL) in soybean, but little is known about another close paralog, GmCHX20a, found in the same QTL. In this study, GmCHX20a was characterized along with GmCHX1. The expression patterns of the two genes and the direction of Na+ flux directed by overexpression of these two transporters are different, suggesting that they are functionally distinct. The ectopic expression of GmCHX20a led to an increase in salt sensitivity and osmotic tolerance, which was consistent with its role in increasing Na+ uptake into the root. Although this seems counter-intuitive, it may in fact be part of the mechanism by which soybean could counter act the effects of osmotic stress which is commonly manifested in the initial stage of salinity stress. On the other hand, GmCHX1 from salt tolerant soybean was shown to protect plants via Na+ exclusion under salt stress. Taken together these results suggest that GmCHX20a and GmCHX1 might work complementally through a concerted effort to address both osmotic stress and ionic stress as a result of elevated salinity. This article is protected by copyright. All rights reserved.

Background The efficacy of bevacizumab in non-small cell lung cancer (NSCLC) patients is unsatisfactory, and the selection of suitable patients is still challenging. Given the epigenetic modifications can contribute to an aberrant regulation of angiogenesis and microenvironment, we investigated DNA methylation profiles to determine clinical benefit of bevacizumab in NSCLC patients. Methods Genome-wide DNA methylation profiling was performed in NSCLC patients treated with chemotherapy in combination with bevacizumab. Patients were divided into better prognosis group (A group) and inferior prognosis group (B group) based on their survival. The difference of methylation patterns and respective functional enrichment analysis were performed between two groups. Prognostic DNA methylation signature for bevacizumab was established with the least absolute shrinkage and selection operator regression analyses. TISIDB database was further used to infer immunological relationship for prognostic related DNA methylation. Results Twenty patients were included in this study, and significantly distinct methylation patterns were observed between patients with different prognosis. Related genes of different methylation regions were significantly enriched in the biological process of cell projection assembly, neutrophil mediated immunity, and pathway of VEGFA-VEGFR2 signaling pathway, neutrophil degranulation. A 10-gene DNA methylation signature for prognosis prediction was established with the C-index of 0.76. And host genes of signature were found to be related to the abundance of ActCD4, Th1, ActCD8, NKT and neutrophil cells. Conclusion The 10-gene DNA methylation signature could serve as a novel biomarker to predict the clinical benefit of bevacizumab therapy and improve this anti-tumor approach for NSCLC patients.

Objectives We aimed to identify the risk factors and provide a nomogram for the prediction of radiotherapy-related esophageal fistula in patients with esophageal cancer (EC) using a case-control study. Patients and methods Patients with esophageal fistula who received radiotherapy or chemoradiotherapy between 2003 and 2017 were retrospectively collected in two institutions. In the training cohort (TC), clinical, pathologic, and serum data of 136 patients (cases) who developed esophageal fistula during or after radiotherapy were enrolled and compared with 272 controls (1:2 matched with the diagnosis time of EC, sex, marriage, and race). After the univariable and multivariable logistic regression analyses, the independent risk factors were identified and incorporated into a nomogram. Then the nomogram for the risk prediction was externally validated in the validation cohort (VC; 47 cases and 94 controls) using bootstrap resampling. Results Multivariable analyses demonstrated that ECOG PS, BMI, T4, N2/3 and re-radiotherapy were independent factors for esophageal fistula. Then a nomogram was constructed with the C-index of 0.805 (95% CI, 0.762–0.848) for predicting the risk of developing esophageal fistula in EC patients receiving radiotherapy. Importantly, the C-index maintained 0.764 (95% CI, 0.683–0.845) after the external validation. Conclusions We created and externally validated the first risk nomogram of esophageal fistula associated with radiotherapy. This will aid individual risk stratification of patients with EC developing esophageal fistula.

Background: Bevacizumab combined with chemotherapy is still one of the standard options for treatment of advanced non-small cell lung cancer (NSCLC) patients without driver mutations. Serum inflammatory factors, representing the systemic immune status, are shown to have complicated relationships with tumor angiogenesis, and proved to be associated with survival of advanced NSCLC patients. However, the information from the baseline factors is relatively limited, which cannot reflect the dynamic changes of systemic immune status during bevacizumab treatment. We, thus, attempted to evaluate longitudinal kinetics of systemic inflammatory factors during treatment of bevacizumab and to explore their predictive role in treatment response and patient outcomes in advanced NSCLC. Method: Systemic inflammatory factors (neutrophil/lymphocyte (NLR), platelet/lymphocyte (PLR), neutrophil×platelet/lymphocyte (SII) and lymphocyte/monocyte (LMR)) and clinical variables were collected and analyzed from 161 advanced NSCLC patients treated with bevacizumab. Mixed effect regression models were first performed for longitudinal analysis of the changes of serum inflammatory factors during bevacizumab treatment. Then, univariate and multivariate Cox models were used for overall survival (OS) and progression free survival (PFS) analyses to determine the independent prognostic factors. Results: In the first 6 cycles of bevacizumab treatment, patients with complete response/partial response (CR/PR) had a -0.11, -0.066, -0.15, and 0.073 change every 2 cycles in transformed NLR (95%CI: -0.19--0.03, p=0.008), PLR (95%CI: -0.12--0.013, p=0.015), SII (95%CI: -0.23--0.05, p

Purpose/Objective(s) This study aims to propose an adaptive integration model to predict lymph node metastasis (LNM) in esophageal squamous cell carcinoma (ESCC) patients by learning from preoperative CT imaging and non-imaging clinical parameters. Materials/Methods Contrast enhanced CT (CECT) scans taken two weeks before surgery and 20 clinical factors including general, pathological, hematological and diagnostic information were collected from 397 ESCC patients between October 2013 and November 2018. Of these, there are 924 lymph nodes (LN) (798 negative: 126 positive) with pathologically confirmed status after surgery. All LNs were randomly divided into a training set (n = 663) and validation set (n = 185). Data augmentations generated 1326 negative and 1140 positive LN samples used for training. FAGNN was composed of: a) deep image feature extraction by 3D CNN encoder and high-level non-imaging factors representation by a feed-forward neural network; b) a feature-wise attention module for feature embedding with learnable adaptive weights; and c) a graph convolutional layer to integrate the embedded features for final LN level metastasis prediction. Results FAGNN achieved area under ROC curve (AUC) of 0.816, sensitivity (sen) of 0.794, specificity (spe) of 0.861 in validation, which outperformed our model without adaptive embedding (AUC 0.810, sen 0.785, spe 0.857), our model using CT alone (AUC 0.802, sen 0.765, and spe 0.826), radiomics integration model (AUC 0.762, sen 0.739, spe 0.818), CT radiomics model (AUC 0.733, sen 0.689, spe 0.765). The improvement was statistically significant (P Conclusion Our adaptive integration model improved the LNM prediction performance based on CT and clinical parameters. Our model has the potential to foster effective fusion of multi-sourced parameters and support early prognosis and personalized surgery or radiotherapy planning in ESCC patients.

Radiotherapy plays a vital role in treating patients with esophageal cancer (EC), whereas potential complications such as esophageal fistula (EF) can be devastating and even life-threatening. Therefore, predicting EF risks prior to radiotherapies for EC patients is crucial for their clinical treatment and quality of life. We propose a novel method of combining thoracic Computerized Tomography (CT) scans and clinical tabular data to improve the prediction of EF risks in EC patients. The multimodal network includes encoders to extract salient features from images and clinical data, respectively. In addition, we devise a self-attention module, named VisText, to uncover the complex relationships and correlations among different features. The associated multimodal features are integrated with clinical features by aggregation to further enhance prediction accuracy. Experimental results indicate that our method classifies EF status for EC patients with an accuracy of 0.8366, F1 score of 0.7337, specificity of 0.9312 and AUC of 0.9119, outperforming other methods in comparison.

Lymph node metastasis (LNM) is the most critical prognosis factor in esophageal squamous cell carcinoma (ESCC). Effective and adaptive integration of preoperative CT images and multi-sourced non-imaging clinical factors is a challenging issue. In this work, we propose a graph-based reasoning model to learn new representations from multi-categorical clinical parameters for LNM prediction. Given CT, general, diagnostic, pathological, and hematological clinical information, we firstly propose a graph construction strategy with category-wise contextual attention to embed multi-categorical features as graph node attributes. Secondly, we introduce a co-graph attention layer composed of a conventional graph attention network (con-GAT) and a correlation-based GAT (corr-GAT) to learn new representations. Corr-GAT complements con-GAT by difference-based correlations across image regions in global spectral space. Experimental results of ablation studies and comparison with others over 924 lymph nodes demonstrated improved performance and contributions of our major innovations. Our model has the potential to foster early prognosis and personalized surgery or radiotherapy planning in ESCC patients.

e20591 Background: Alectinib and thoracic radiotherapy (TRT) are important modalities in the management of ALK-rearranged non-small cell lung cancer. We therefore aimed to assess the incidence and risk factors of pneumonitis when these two treatments were combined. Methods: Patients with ALK-rearranged non-small cell lung cancer, receiving alectinib and TRT from January 2018 to December 2020 were reviewed, and the clinical and dosimetry data were collected. Logistic regression analyses were performed to evaluate risk factors associated with treatment-related pneumonitis (TRP). The prediction ability of dosimetry parameters for pneumonitis were examined by receiver-operating characteristic (ROC) curve analyses. Results: Of the 25 enrolled patients, 16 (64%) developed pneumonitis, and 12 (48%) developed grade 2 or higher pneumonitis. Logistics regression analyses revealed chemotherapy (OR = 8.750, 95% CI 1.207 - 63.428, p = 0.032), higher ipsilateral V5 (OR = 1.060, 95%CI 1.002 - 1.122, p = 0.043) and total lung V5 (OR = 1.083, 95%CI 1.000 - 1.174, p = 0.05) to be risk factors for TRP. Total lung V10, total lung V5, and ipsilateral V5 were strong predictors of TRP among the dosimetry parameters (AUC = 0.800, 0.786 and 0.781, respectively). After developing pneumonitis, 52% of patients recovered or improved, but one patient died due to respiratory failure. Conclusions: The combined use of alectinib and TRT significantly increased the risk of pneumonitis. Clinicians should consider the elevated risks and related dosimetric factors when deciding on combination treatment for ALK-rearranged NSCLC patients.

Schistosoma japonicum (S. japonicum) is one of the etiological agents of schistosomiasis, a widespread zoonotic parasitic disease. However, the mechanism of the balanced co-existence between the host immune system and S. japonicum as well as their complex interaction remains unclear. In this study, 16S rRNA gene sequencing, combined with metagenomic sequencing approach as well as ultraperformance liquid chromatography–mass spectrometry metabolic profiling, was applied to demonstrate changes in the gut microbiome community structure during schistosomiasis progression, the functional interactions between the gut bacteria and S. japonicum infection in BALB/c mice, and the dynamic metabolite changes of the host. The results showed that both gut microbiome and the metabolites were significantly altered at different time points after the infection. Decrease in richness and diversity as well as differed composition of the gut microbiota was observed in the infected status when compared with the uninfected status. At the phylum level, the gut microbial communities in all samples were dominated by Firmicutes, Bacteroidetes, Proteobacteria, and Deferribacteres, while at the genus level, Lactobacillus, Lachnospiraceae NK4A136 group, Bacteroides, Staphylococcus, and Alloprevotella were the most abundant. After exposure, Roseburia, and Ruminococcaceae UCG-014 decreased, while Staphylococcus, Alistipes, and Parabacteroides increased, which could raise the risk of infections. Furthermore, LEfSe demonstrated several bacterial taxa that could discriminate between each time point of S. japonicum infection. Besides that, metagenomic analysis illuminated that the AMP-activated protein kinase (AMPK) signaling pathway and the chemokine signaling pathway were significantly perturbed after the infection. Phosphatidylcholine and colfosceril palmitate in serum as well as xanthurenic acid, naphthalenesulfonic acid, and pimelylcarnitine in urine might be metabolic biomarkers due to their promising diagnostic potential at the early stage of the infection. Alterations of glycerophospholipid and purine metabolism were also discovered in the infection. The present study might provide further understanding of the mechanisms during schistosome infection in aspects of gut microbiome and metabolites, and facilitate the discovery of new targets for early diagnosis and prognostic purposes. Further validations of potential biomarkers in human populations are necessary, and the exploration of interactions among S. japonicum, gut microbiome, and metabolites is to be deepened in the future.

Cation/H

Early prognosis of the radiotherapy-related esophageal fistula is of great significance in making personalized stratification and optimal treatment plans for esophageal cancer (EC) patients. The effective fusion of diagnostic consideration guided multi-level radiographic visual descriptors is a challenging task. We propose an end-to-end clinical knowledge enhanced multi-level cross-channel feature extraction and aggregation model. Firstly, clinical attention is represented by contextual CT, segmented tumor and anatomical surroundings from nine views of planes. Then for each view, a Cross-Channel-Atten Network is proposed with CNN blocks for multi-level feature extraction, cross-channel convolution module for multi-domain clinical knowledge embedding at the same feature level, and attentional mechanism for the final adaptive fusion of multi-level cross-domain radiographic features. The experimental results and ablation study on 558 EC patients showed that our model outperformed the other methods in comparison with or without multi-view, multi-domain knowledge, and multi-level attentional features. Visual analysis of attention maps shows that the network learns to focus on tumor and organs of interests, including esophagus, trachea, and mediastinal connective tissues.

It has long been a puzzle in cancer treatment that despite the initial appearance of apoptosis, the process could be reversed in some cancer cells and often results in more aggressive tumors and metastasis. The mechanism for this recurrence is yet unknown. Here we report that human mammary carcinoma cells induced to undergo apoptosis could recover with increased tumorigenicity in vitro and in vivo, and induced lymph node metastasis. Specifically, the reversed cells underwent epithelial-to-mesenchymal transitions in the primary tumors in situ, and mesenchymal-to-epithelial transitions in the metastatic cells. Flow cytometry confirmed an elevated percentage of cells carrying cancer stem cells (CSCs) markers (CD44+/CD24-) in the reversed breast cancer cell population, with hypomethylated CD44 promoters and hypermethylated CD24 promoters. More importantly, CSCs were generated anew from non-stem cancer cells after apoptosis reversal possibly through epigenetic modifications. The results from this study can open doors to discovering more effective cancer treatments by suppressing apoptosis reversal.

Cancer recurrence has long been studied by oncologists while the underlying mechanisms remain unclear. Recently, we and others found that a phenomenon named apoptosis reversal leads to increased tumorigenicity in various cell models under different stimuli. Previous studies have been focused on tracking this process in vitro and in vivo; however, the isolation of real reversed cells has yet to be achieved, which limits our understanding on the consequences of apoptosis reversal. Here, we take advantage of a Caspase-3/7 Green Detection dye to label cells with activated caspases after apoptotic induction. Cells with positive signals are further sorted out by fluorescence-activated cell sorting (FACS) for recovery. Morphological examination under confocal microscopy helps confirm the apoptotic status before FACS. An increase in tumorigenicity can often be attributed to the elevation in the percentage of cancer stem cell (CSC)-like cells. Also, given the heterogeneity of breast cancer, identifying the origin of these CSC-like cells would be critical to cancer treatment. Thus, we prepare breast non-stem cancer cells before triggering apoptosis, isolating caspase-activated cells and performing the apoptosis reversal procedure. Flow cytometry analysis reveals that breast CSC-like cells re-appear in the reversed group, indicating breast CSC-like cells are transited from breast non-stem cancer cells during apoptosis reversal. In summary, this protocol includes the isolation of apoptotic breast cancer cells and detection of changes in CSC percentage in reversed cells by flow cytometry.

Additional file 1. Characteristics of patients.

e20519 Background: Immune checkpoint inhibitors have revolutionized the patterns of tumor treatment with significantly improved survival. However, it’s worth noting that the objective response rate of PD-1/PD-L1 inhibitors monotherapy is just 20% in non-small cell lung cancer (NSCLC) patients, and biomarkers for the prognosis of patients receiving PD-1/PD-L1 inhibitors are urgently needed. LILRB2 (ILT4) is known as an immunosuppressive molecule and promotes the development and progression of NSCLC. However, the relationship between LILRB2 expression and immune microenvironment remains unclear in patients with lung adenocarcinoma. Methods: Expression profile and corresponding clinicopathological data of patients with lung adenocarcinoma were obtained from The Cancer Genome Atlas (TCGA) database. TIMER 2.0 was used to investigate the relationship between LILRB2 expression and immune infiltrates, including T cell subset, macrophage and so on. Cox proportional hazard model was also performed to evaluate the role LILRB2 expression and immune infiltrations in prognostic prediction of patients with lung adenocarcinoma. Results: 515 patients with lung adenocarcinoma were included in analysis from TCGA dataset. LILRB2 expression was found to have significant negative correlation with CD8+ T cell infiltration (P = 0.0012), and positive correlation with CD4+ T cell infiltration (P < 0.001). There were also significant positive correlations between LILRB2 expression and Treg (P = 0.0083), cancer associated fibroblast (P < 0.001), monocyte (P < 0.001), M2 macrophage (P < 0.001) infiltration, which was regarded as the marker of immunosuppressive microenvironment. Notably, there were also significant association between the expression of LILRB2 and PD-1 (P < 0.001) and PD-L1(P < 0.001). In terms of survival analysis, high level of Treg (HR = 0.86, P = 0.03) and M2 macrophage (HR = 0.84, P = 0.017) infiltration was associated with poor prognosis of patients with lung adenocarcinoma. However, the level of CD8+T cell (HR = 0.93, P = 0.298), CD4+ T cell (HR = 0.933, P = 0.321), and LILRB2 expression (P = 0.56) was not associated with the survival of patient with lung adenocarcinoma. Conclusions: The overexpression of LILRB2 is significantly correlated with immunosuppressive immune infiltration and microenvironment in lung adenocarcinoma and might be the potential predictive biomarkers for the PD-1/PD-L1 inhibitors.

e12594 Background: Triple-negative breast cancer (TNBC) is the subtype of breast cancer with the worst prognosis. There is no reliable model for survival prediction of TNBC patients. The traditional Cox regression analysis with poor prediction power cannot satisfy the clinical needs. The purpose was to establish a deep learning model and develop a new prognostic system for TNBC patients. Methods: This study collected data of TNBC patients from the Surveillance, Epidemiology, and End Results (SEER) program between 2010 and 2016. 70% were used to develop the deep learning model, 15% were used as the validation set, and 15% as the independent testing set. Then the concordance-index (c-index) and Brier score (IBS) were calculated and compared with the Cox regression analysis and random forest. Finally, according to the classification of the deep survival model, an individualized prognosis system was established. Results: A total of 37,818 patients were enrolled in this study. In the validation set, the c-index of the deep learning was 0.799, which was better than the traditional Cox regression model (0.774) and random forest (0.763). The independent testing set further proved the robustness of the deep survival model (c-index 0.788). The new prognosis system based on the deep survival model reached an area under the curve (AUC) of 0.805, which was better than the Tumor, Node, Metastases (TNM) staging system (0.771). Conclusions: Deep learning model had better prediction power than the Cox regression analysis and the random forest. The established prognosis system can better predict prognosis and aid individual risk stratification for TNBC patients patients.

e16093 Background: Lymph node (LN) metastasis is the most important factor for decision making in esophageal squamous cell carcinoma (ESCC). A more accurate prediction model for LN metastatic status in ESCC patients is needed. Methods: In this retrospective study, 397 ESCC patients who took Contrast-Enhanced CT (CECT) within 15 days before surgery between October 2013 and November 2018 were collected. There are 924 (798 negative and 126 positive) LNs with pathologically confirmed status after surgery. All LNs were randomly divided into a training set (n = 663) and validation set (n = 185). Data augmentation including shifting and rotation was performed in the training set, resulting in 1326 negative and 1140 positive LN samples. The GACNN model was trained over CT volumetric patches centred at manually segmented LN samples. GACNN was composed of a 3D UNet encoder to extract deep features, and a graph attention layer to integrate morphological features extracted from segmented LN. The model was validated using the validation set (135 negative and 50 positive) and measured by area under ROC curve (auc), sensitivity (sen), and specificity (spe). Results: GACNN achieved better auc, sen, and spe of 0.802, 0.765, and 0.826, when compared to 3 other models including CT radiomics model (auc 0.733, sen 0.689, spe 0.765), 3D UNet encoder (auc 0.778, sen 0.722, spe 0.767), and our model without morphological features (auc 0.796, sen 0.754, spe 0.803). The improvement was statistically significant (p < 0.001). Conclusions: Our prediction model improved the prediction of LN metastasis, which has the potential to assist LN metastasis risk evaluation and personalized treatment planning in ESCC patients for surgery or radiotherapy.

4553 Background: We aimed to propose a risk model based on MM-Atten-CNN for predicting esophageal fistula in patients with esophageal cancer (EC) from computerized tomography (CT) -based radiomics. Methods: EC patients who didn’t received esophageal surgery between July 2014 and August 2019 were collected. Of these, 186 patients (cases) who developed esophageal fistula were enrolled and compared with 372 controls (1:2 matched with the diagnosis time of EC, sex, marriage, and race). All 558 patients were divided into training set (n = 390) and validation set (n = 168) randomly. The MM-Atten-CNN risk model was trained over 2D slices from nine views of planes, where there were three patches of contextual CT, segmented tumor and neighbouring information in each view. In the training set (130 cases and 260 controls), data augmentation was performed including pixel shifting [-10, -5, +5, +10] and rotation [-10, +10]. In total, there were (130+260) *16*2 = 12480 subjects used for training. Finally, the risk model was validated in the validation set (56 cases and 112 controls) and measured by accuracy (acc), sensitivity (sen), and specificity (spe). Results: The developed risk model achieved (acc, sen, spe) of (0.839, 0.807, 0.926), which were more predictive for the occurrence of esophageal fistula when compared to CNN models using single coronal view (acc 0.763, sen 0.581, spe 0.837), multi-view 2D contextual CT slices (acc 0.779, sen 0.656, spe 0.896), and 3D CNN using contextual CT volumes (acc 0.781, sen 0.689, spe 0.852). Conclusions: MM-Atten-CNN CT-based model improved the performance of esophageal fistula risk prediction, which has the potential to assist individualized stratification and treatment planning in EC patients.

Shuheng Shang,1,2 Yi Su,3 Zhaofeng Zhu,4 Butuo Li,2,5 Meiying Guo,1,2 Yiyue Xu,1,2 Xindong Sun,2 Linlin Wang,2 Jinming Yu2 1Department of Radiation Oncology, School of Medicine, Shandong University, Jinan, China; 2Department of Radiation Oncology, Shandong Cancer Hospital Affiliated to Shandong University, Shandong Academy of Medical Science, Jinan, China; 3Department of Radiotherapy, The Affiliated Yantai Yuhuangding Hospital of Qingdao University Institution, Yantai, China; 4Department of Radiotherapy, Tai’an Central Hospital, Tai’an, China; 5Department of Radiation Oncology, Tianjin Medical University, Tianjin, China Background: The optimal treatment strategy for patients with non-small-cell lung cancer (NSCLC) with postoperative oligometastases is poorly defined. This two-institution analysis sought to retrospectively compare the efficacy and toxicity of local ablative treatment plus chemotherapy vs local treatment alone in patients with NSCLC who developed oligometastases after surgery.Patients and methods: Among patients who underwent surgery for stage I–III NSCLC, 163 patients with oligometastases were enrolled between 2005 and 2016 in this study. All patients had ≤5 metachronous metastases with a disease-free interval (DFI) of ≥6months after surgery. Patients with a second primary cancer, local recurrence, or driver mutations were excluded. Overall survival (OS), progression-free survival (PFS), objective response rate (ORR), failure patterns, and treatment-related toxicities were compared between groups receiving local ablative treatment plus chemotherapy and local treatment alone.Results: A total of 105 patients who underwent local ablative therapy combined with chemotherapy and 58 patients who received local ablative therapy alone were included in this study. The median follow-up was 19 (range, 1.5–107) months. The combination therapy group had a higher ORR than the local therapy alone group (66.7% vs 46.5%, P=0.012), while the median PFS was 10 vs 7months (P=0.006) and the median OS was 19 vs 18.5months (P=0.498), respectively. By multivariate analysis, combination therapy and DFI ≥24months were associated with superior PFS. Age was the only independent prognostic factor for OS (P

Introduction Locally advanced nonsmall‐cell lung cancer (LA‐NSCLC) represented a highly heterogeneous group, with more than half of the patients aged ≥65 years at the time of diagnosis. However, the optimal treatment for elderly LA‐NSCLC patients was still not defined. Methods A total of 33530 elderly patients (≥65 years) diagnosed with LA‐NSCLC from 2004 to 2014 were identified from Surveillance, Epidemiology, and End Results (SEER) database. Results Locally advanced nonsmall‐cell lung cancer patients aged 65‐74 years were more frequently treated with chemoradiotherapy (CRT) (40%), while patients aged ≥75 years received more best supportive care (BSC) (36%). For age group of 65‐74 years, patients who had surgery with or without (neo)adjuvant therapy had a median survival of 28 months, CRT 15 months, radiotherapy (RT) alone 6 months, chemotherapy alone 11 months, and BSC 3 months; while for patients aged ≥ 75 years, the median OS was 20, 13, 7, 9, and 2, respectively. Besides, independent clinicopathological factors were integrated into nomograms for OS and CSS prediction, C‐indexes achieved 0.692 and 0.698, respectively. Importantly, the discrimination of nomogram was superior to that of the American Joint Committee on Cancer TNM classification (0.742 vs 0.572 for training set and 0.731 vs 0.565 for validation set). Conclusion For elderly patients with LA‐NSCLC, the curative‐intent treatment (surgery or CRT) conferred better survival compared to chemotherapy alone, RT alone and BSC. The proposed nomograms based on independent clinicopathological variables may be practical and helpful for precise evaluation of patient prognosis, and guiding the individualized treatment for elderly LA‐NSCLC.

The effects of co-exposure to ultraviolet (UV) irradiation (with either low- or medium-pressure UV lamps) and free chlorine (chloramine) at practical relevant conditions on changes in natural organic matter (NOM) properties were investigated using four waters. The changes were characterized using the specific disinfection by-product formation potential (SDBPFP), specific total organic halogen formation potential (STOXFP), differential UV absorbance (∆UVA), and size-exclusion chromatography (SEC). The results for exposure to UV irradiation alone and for samples with no exposure were also obtained. The SDBPFPs in all UV-irradiated NOM waters observed were higher than those of non-irradiated samples. UV irradiation led to increases in STOXFPs as a result of chlorination, but no changes, or only small decreases, from chloramination. UV irradiation alone led to positive ∆UVA spectra of the four NOM waters; co-exposure to UV and chlorine gave larger negative ∆UVA spectra than those obtained by chlorine exposure alone. No obvious changes in SEC results were observed for samples only irradiated with UV light; co-exposure gave no detectable changes in the abundances of small fractions for exposure to chlorine only. Both UV photooxidation and photocatalytic oxidation appear to affect the reactivity of the NOM toward subsequent chlorination, and the magnitude of the changes is generally greater for medium-pressure lamps than for low-pressure lamps. These results suggest that applying UV disinfection technology to a particular source may not always be disinfection by-product-problem-free, and the interactions between UV light, chlorine, and NOM may need to be considered.

Background Schistosomiasis is considered second only to malaria as the most devastating parasitic disease in tropical countries. Schistosome cercariae invade the host by penetrating the skin and migrate though the lungs and portal circulation to their final destination in the hepatic portal system and eventually the mesenteric veins. Previous studies have shown that the cytotoxic pathways that target schistosomulum in the lung-stage involve nitric oxide (NO) produced by macrophages. By contrast, skin-stage schistosomulas can evade clearance, indicating that they might be freed from macrophage NO-mediated cytotoxicity to achieve immune evasion; however, the critical molecules and mechanisms involved remain unknown. Methods Recombinant SjCa8 (rSjCa8), an 8-kDa calcium-binding protein that is stage-specifically expressed in cercaria and early skin-stage schistosomulas of Schistosoma japonicum, was incubated with mouse RAW264.7 macrophages. Effects on macrophage proliferation were determined using Cell Counting Kit-8. Next, transwell assay was carried out to further investigate the role of rSjCa8 in macrophage migration. The effects of rSjCa8 on macrophage apoptosis were evaluated using confocal microscopy and flow cytometry. Additional impacts of rSjCa8 on NO release by lipopolysaccharide (LPS)-stimulated macrophages as well as the underlying mechanisms were explored using fluorescent probe, nitric oxide signaling pathway microarray, quantitative real-time PCR, mutagenesis, and neutralizing antibody approaches. Results rSjCa8 exhibited a striking inhibitory effect on macrophage migration, but did not markedly increase cell proliferation or apoptosis. Additionally, rSjCa8 potently inhibited NO release by LPS-stimulated macrophages in a dose- and time-dependent manner, and the inhibitory mechanism was closely associated with intracellular Ca2+ levels, the up-regulation of catalase expression, and the down-regulation of the expression of 47 genes, including Myc, Gadd45a, Txnip, Fas, Sod2, Nos2, and Hmgb1. Vaccination with rSjCa8 increased NO concentration in the challenging skin area of infected mice and reduced the number of migrated schistosomula after skin penetration by cercariae. Conclusions Our findings indicate that SjCa8 might be a novel molecule that plays a critical role in immune evasion by S. japonicum cercaria during the process of skin penetration. The inhibitory impacts of rSjCa8 on macrophage migration and [Ca2+]i-dependent NO release suggest it might represent a novel vaccine candidate and chemotherapeutic target for the prevention and treatment of schistosomiasis. Electronic supplementary material The online version of this article (doi:10.1186/s13071-015-1119-4) contains supplementary material, which is available to authorized users.