A Validated Prediction Model for Overall Survival From Stage III Non-Small Cell Lung Cancer: Toward Survival Prediction for Individual Patients

A nomogram-based survival prediction model for non-small cell lung cancer patients based on clinical risk factors and multiregion radiomics features

AIM

This study focuses on developing a nomogram-based overall survival (OS) prediction model for non-small cell lung cancer (NSCLC) patients by integrating clinical factors with multiregion radiomics features extracted from pretreatment CT images. The proposed nomogram aims to assist clinicians in stratifying patients into high- and low-risk groups for personalised treatment strategies.

MATERIALS AND METHODS

From 2008 to 2018, 77 NSCLC patients were included. The radiomics feature was extracted from the internal and peripheral tumour region of pretreatment computed tomography (CT) images. The least absolute shrinkage and selection operator (LASSO) and the univariable Cox regression model were used to select the radiomics features. The Rad-score was defined as a linear combination of the selected radiomics features and the Cox proportional hazards regression coefficients. The combined model was constructed based on the clinicopathological factors and the Rad-score. The discrimination capacity of the prediction model was evaluated by Harrell's concordance index (C-index), the calibration curve, and the Kaplan-Meier survival curve.

RESULTS

We found that nine radiomics features and histology were independent predictors. The combined model showed the best performance (C-index: 0.799 [95% CI: 0.726-0.872]) compared with the clinical model (C-index: 0.692 [95% CI: 0.625-0.759]) and Rad-score (C-index: 0.663 [95% CI: 0.580-0.746]), and could significantly stratify into high-risk and low-risk NSCLC patients. The calibration curve also showed good consistency between the observation and the prediction.

CONCLUSIONS

The multregion radiomics features have the potential for predicting OS in NSCLC patients. The nomogram-based survival prediction model demonstrates significant potential in guiding clinical decision-making, allowing for precise and personalised treatment for NSCLC patients.

Future of Team-based Basic and Translational Science in Radiation Oncology

To further optimize radiotherapy, a more personalized treatment towards individual patient's risk profiles, dissecting both patient-specific tumor and normal tissue response to multimodality treatments is needed. Novel developments in radiobiology, using in vitro patient-specific complex tissue resembling 3D models and multiomics approaches at a spatial single-cell level, may provide unprecedented insight into the radiation responses of tumors and normal tissue. Here, we describe the necessary team effort, including all disciplines in radiation oncology, to integrate such data into clinical prediction models and link the relatively "big data" from the clinical practice, allowing accurate patient stratification for personalized treatment approaches.

Post-radiotherapy stage III/IV non-small cell lung cancer radiomics research: a systematic review and comparison of CLEAR and RQS frameworks

BACKGROUND

Lung cancer, the second most common cancer, presents persistently dismal prognoses. Radiomics, a promising field, aims to provide novel imaging biomarkers to improve outcomes. However, clinical translation faces reproducibility challenges, despite efforts to address them with quality scoring tools.

OBJECTIVE

This study had two objectives: 1) identify radiomics biomarkers in post-radiotherapy stage III/IV nonsmall cell lung cancer (NSCLC) patients, 2) evaluate research quality using the CLEAR (CheckList_for_EvaluAtion_of_Radiomics_research), RQS (Radiomics_Quality_Score) frameworks, and formulate an amalgamated CLEAR-RQS tool to enhance scientific rigor.

MATERIALS AND METHODS

A systematic literature review (Jun-Aug 2023, MEDLINE/PubMed/SCOPUS) was conducted concerning stage III/IV NSCLC, radiotherapy, and radiomic features (RF). Extracted data included study design particulars, such as sample size, radiotherapy/CT technique, selected RFs, and endpoints. CLEAR and RQS were merged into a CLEAR-RQS checklist. Three readers appraised articles utilizing CLEAR, RQS, and CLEAR-RQS metrics.

RESULTS

Out of 871 articles, 11 met the inclusion/exclusion criteria. The Median cohort size was 91 (range: 10-337) with 9 studies being single-center. No common RF were identified. The merged CLEAR-RQS checklist comprised 61 items. Most unreported items were within CLEAR's "methods" and "open-source," and within RQS's "phantom-calibration," "registry-enrolled prospective-trial-design," and "cost-effective-analysis" sections. No study scored above 50% on RQS. Median CLEAR scores were 55.74% (32.33/58 points), and for RQS, 17.59% (6.3/36 points). CLEAR-RQS article ranking fell between CLEAR and RQS and aligned with CLEAR.

CONCLUSION

Radiomics research in post-radiotherapy stage III/IV NSCLC exhibits variability and frequently low-quality reporting. The formulated CLEAR-RQS checklist may facilitate education and holds promise for enhancing radiomics research quality.

CLINICAL RELEVANCE STATEMENT

Current radiomics research in the field of stage III/IV postradiotherapy NSCLC is heterogenous, lacking reproducibility, with no identified imaging biomarker. Radiomics research quality assessment tools may enhance scientific rigor and thereby facilitate radiomics translation into clinical practice.

KEY POINTS

There is heterogenous and low radiomics research quality in postradiotherapy stage III/IV nonsmall cell lung cancer. Barriers to reproducibility are small cohort size, nonvalidated studies, missing technical parameters, and lack of data, code, and model sharing. CLEAR (CheckList_for_EvaluAtion_of_Radiomics_research), RQS (Radiomics_Quality_Score), and the amalgamated CLEAR-RQS tool are useful frameworks for assessing radiomics research quality and may provide a valuable resource for educational purposes in the field of radiomics.

Clinical Significance and Molecular Annotation for PD-L1 Negative Advanced Non-Small Cell Lung Cancer with Sensitivity to Responsive to Dual PD-1/CTLA-4 Blockade

Background

Immunotherapy has become the standard treatment for driving gene-negative advanced non-small cell lung cancer (NSCLC). However, compared to PD-L1-positive patients, the efficacy of Anti-PD-(L)1 monotherapy is suboptimal in PD-L1-negative advanced NSCLC. In this study, we aim to analyze the optimal immunotherapy approach for PD-L1-negative NSCLC patients and develop a new nomogram to enhance the clinical predictability of immunotherapy for NSCLC patients.

Methods

In this study, we retrieved clinical information and genomic data from cBioPortal for NSCLC patients undergoing immunotherapy. Cox regression analyses were utilized to screen the clinical information and genomic data that related to survival. The prognostic-relate genes function was studied by comprehensive bioinformatics analyses. The Kaplan-Meier plot method was employed for survival analysis.

Results

A total of 199 PD-L1-negative NSCLC patients were included in this study. Among them, 165 patients received Anti-PD-(L)1 monotherapy, while 34 patients received Anti-PD-(L)1+Anti-CTLA-4 combination therapy. The Anti-PD-(L)1+Anti-CTLA-4 combination therapy demonstrated significantly higher PFS compared to the Anti-PD-(L)1 monotherapy. The mutation status of KRAS, ANO1, COL14A1, LTBP1. ERBB4 and PCSK5 were found to correlate with PFS. Utilizing the clinicopathological parameters and genomic data of the patients, a novel nomogram was developed to predict the prognosis of Anti-PD-(L)1+Anti-CTLA-4 combination therapy.

Conclusion

Our study revealed that KRAS, ANO1, COL14A1, LTBP1. ERBB4 and PCSK5 mutation could serve as predictive biomarkers for patients with Anti-PD-(L)1+Anti-CTLA-4 combination therapy. Our systematic nomogram demonstrates significant potential in predicting the prognosis for NSCLC patients with responsive to dual PD-1/CTLA-4 blockade.

A deep learning approach for overall survival prediction in lung cancer with missing values

BACKGROUND AND OBJECTIVE

In the field of lung cancer research, particularly in the analysis of overall survival (OS), artificial intelligence (AI) serves crucial roles with specific aims. Given the prevalent issue of missing data in the medical domain, our primary objective is to develop an AI model capable of dynamically handling this missing data. Additionally, we aim to leverage all accessible data, effectively analyzing both uncensored patients who have experienced the event of interest and censored patients who have not, by embedding a specialized technique within our AI model, not commonly utilized in other AI tasks. Through the realization of these objectives, our model aims to provide precise OS predictions for non-small cell lung cancer (NSCLC) patients, thus overcoming these significant challenges.

METHODS

We present a novel approach to survival analysis with missing values in the context of NSCLC, which exploits the strengths of the transformer architecture to account only for available features without requiring any imputation strategy. More specifically, this model tailors the transformer architecture to tabular data by adapting its feature embedding and masked self-attention to mask missing data and fully exploit the available ones. By making use of ad-hoc designed losses for OS, it is able to account for both censored and uncensored patients, as well as changes in risks over time.

RESULTS

We compared our method with state-of-the-art models for survival analysis coupled with different imputation strategies. We evaluated the results obtained over a period of 6 years using different time granularities obtaining a Ct-index, a time-dependent variant of the C-index, of 71.97, 77.58 and 80.72 for time units of 1 month, 1 year and 2 years, respectively, outperforming all state-of-the-art methods regardless of the imputation method used.

CONCLUSIONS

The results show that our model not only outperforms the state-of-the-art's performance but also simplifies the analysis in the presence of missing data, by effectively eliminating the need to identify the most appropriate imputation strategy for predicting OS in NSCLC patients.

A dual-radiomics model for overall survival prediction in early-stage NSCLC patient using pre-treatment CT images

Introduction

Radiation therapy (RT) is one of the primary treatment options for early-stage non-small cell lung cancer (ES-NSCLC). Therefore, accurately predicting the overall survival (OS) rate following radiotherapy is crucial for implementing personalized treatment strategies. This work aims to develop a dual-radiomics (DR) model to (1) predict 3-year OS in ES-NSCLC patients receiving RT using pre-treatment CT images, and (2) provide explanations between feature importanceand model prediction performance.

Methods

The publicly available TCIA Lung1 dataset with 132 ES-NSCLC patients received RT were studied: 89/43 patients in the under/over 3-year OS group. For each patient, two types of radiomic features were examined: 56 handcrafted radiomic features (HRFs) extracted within gross tumor volume, and 512 image deep features (IDFs) extracted using a pre-trained U-Net encoder. They were combined as inputs to an explainable boosting machine (EBM) model for OS prediction. The EBM's mean absolute scores for HRFs and IDFs were used as feature importance explanations. To evaluate identified feature importance, the DR model was compared with EBM using either (1) key or (2) non-key feature type only. Comparison studies with other models, including supporting vector machine (SVM) and random forest (RF), were also included. The performance was evaluated by the area under the receiver operating characteristic curve (AUCROC), accuracy, sensitivity, and specificity with a 100-fold Monte Carlo cross-validation.

Results

The DR model showed highestperformance in predicting 3-year OS (AUCROC=0.81 ± 0.04), and EBM scores suggested that IDFs showed significantly greater importance (normalized mean score=0.0019) than HRFs (score=0.0008). The comparison studies showed that EBM with key feature type (IDFs-only demonstrated comparable AUCROC results (0.81 ± 0.04), while EBM with non-key feature type (HRFs-only) showed limited AUCROC (0.64 ± 0.10). The results suggested that feature importance score identified by EBM is highly correlated with OS prediction performance. Both SVM and RF models were unable to explain key feature type while showing limited overall AUCROC=0.66 ± 0.07 and 0.77 ± 0.06, respectively. Accuracy, sensitivity, and specificity showed a similar trend.

Discussion

In conclusion, a DR model was successfully developed to predict ES-NSCLC OS based on pre-treatment CT images. The results suggested that the feature importance from DR model is highly correlated to the model prediction power.

Federated Learning Survival Model and Potential Radiotherapy Decision Support Impact Assessment for Non-small Cell Lung Cancer Using Real-World Data

AIMS

The objective of this study was to develop a two-year overall survival model for inoperable stage I-III non-small cell lung cancer (NSCLC) patients using routine radiation oncology data over a federated (distributed) learning network and evaluate the potential of decision support for curative versus palliative radiotherapy.

METHODS

A federated infrastructure of data extraction, de-identification, standardisation, image analysis, and modelling was installed for seven clinics to obtain clinical and imaging features and survival information for patients treated in 2011-2019. A logistic regression model was trained for the 2011-2016 curative patient cohort and validated for the 2017-2019 cohort. Features were selected with univariate and model-based analysis and optimised using bootstrapping. System performance was assessed by the receiver operating characteristic (ROC) and corresponding area under curve (AUC), C-index, calibration metrics and Kaplan-Meier survival curves, with risk groups defined by model probability quartiles. Decision support was evaluated using a case-control analysis using propensity matching between treatment groups.

RESULTS

1655 patient datasets were included. The overall model AUC was 0.68. Fifty-eight percent of patients treated with palliative radiotherapy had a low-to-moderate risk prediction according to the model, with survival times not significantly different (p = 0.87 and 0.061) from patients treated with curative radiotherapy classified as high-risk by the model. When survival was simulated by risk group and model-indicated treatment, there was an estimated 11% increase in survival rate at two years (p < 0.01).

CONCLUSION

Federated learning over multiple institution data can be used to develop and validate decision support systems for lung cancer while quantifying the potential impact of their use in practice. This paves the way for personalised medicine, where decisions can be based more closely on individual patient details from routine care.

Identifying optimal surgical approach among T1N2-3M0 non-small cell lung cancer patients: a population-based analysis

Background

Whether stage T1N2-3M0 non-small cell lung cancer (NSCLC) patients could benefit from surgery and the optimal surgical procedure have remained controversial and unclear. This study aimed to investigate whether stage T1N2-3M0 NSCLC can benefit from different surgery types and develop a tool for survival prediction.

Methods

The Surveillance, Epidemiology, and End Results (SEER) database was used to identify patients diagnosed with stage T1N2-3M0 NSCLC between 2000 and 2015. A 1:1 propensity score-matched (PSM) analysis was used to balance the distribution of clinical characteristics. Survival analyses were performed by using the Kaplan-Meier (KM) curves and Cox proportional hazards regression. All patients were randomly split at a ratio of 7:3 into training and validation cohorts. The nomogram was constructed by integrating all independent predictors for overall survival (OS) and cancer-specific survival (CSS). The model's performance was evaluated by discrimination, calibration ability, and risk stratification ability.

Results

A total of 4,671 patients were enrolled. After 1:1 PSM, the distribution proportions of clinical characteristics in 1,146 patients were balanced (all P>0.05). The non-surgical approach was associated with worse survival compared with sublobectomy and lobectomy in the unmatched and matched cohorts. The multivariate Cox analysis showed that sublobectomy and lobectomy were both related to better OS and CSS rates compared with no surgery (P<0.001). Moreover, the results of subgroup analyses based on age, N stage, and radiotherapy or chemotherapy strategy were consistent. A total of 801 patients were included in the training cohort and 345 cases constituted the validation cohort. The nomogram constructed for the 1-, 3-, and 5-year OS and CSS prediction showed good discrimination, performance, and calibration both in the training and validation sets. Significant distinctions in survival curves between different risk groups stratified by prognostic scores were also observed (all P<0.001).

Conclusions

Stage T1N2-3M0 NSCLC patients could benefit from sublobectomy or lobectomy, and lobectomy provides better survival benefits. We developed and validated nomograms, which could offer clinicians instructions for strategy making.

Identification of nomogram associated with durable clinical benefit gene for advanced non-small cell lung cancer with sensitivity to responsive to immunotherapy

Background

Immunotherapy has become the standard treatment for advanced non-small cell lung cancer (NSCLC). However, a subset of the most advanced NSCLC patients fails to respond adequately to Immune checkpoint inhibitors (ICIs). Developing new nomograms and integrating prognostic factors are crucial for improving the clinical predictability of NSCLC patients undergoing ICIs.

Methods

Clinical information and genomic data of NSCLC patients undergoing ICIs were retrieved from cBioPortal. Gene alterations associated with durable clinical benefit (DCB) were compared to those linked to no durable benefit (NDB). The Kaplan-Meier plot method was employed for survival analysis, and a novel nomogram was formulated by selecting pertinent clinical variables.

Results

For the NSCLC patients receiving immunotherapy, three subgroups were identified based on the treatment regimen, including anti-PD-1 monotherapy, anti-PD-1 combination with anti-CTLA-4, and first-line treatment. The mutation status of TP53, PGR, PTPRT, RELN, MUC19, LRP1B, and FAT3 was found to be associated with progression-free survival (PFS). Using the clinicopathological parameters and genomic data of the patients, we developed three novel nomograms to predict the prognosis of ICI treatment in different subgroups.

Conclusion

Our study revealed that PGR, PTPRD, RELN, MUC19, LRP1B, and FAT3 mutation could serve as predictive biomarkers. Our systematic nomograms demonstrate significant potential in predicting the prognosis for NSCLC patients with sensitivity to different ICI treatment strategies.

Calnexin as a dual-role biomarker: antibody-based diagnosis and therapeutic targeting in lung cancer

Lung cancer carries one of the highest mortality rates among all cancers. It is often diagnosed at more advanced stages with limited treatment options compared to other malignancies. This study focuses on calnexin as a potential biomarker for diagnosis and treatment of lung cancer. Calnexin, a molecular chaperone integral to N-linked glycoprotein synthesis, has shown some associations with cancer. However, targeted therapeutic or diagnostic methods using calnexin have been proposed. Through 1D-LCMSMS, we identified calnexin as a biomarker for lung cancer and substantiated its expression in human lung cancer cell membranes using Western blotting, flow cytometry, and immunocytochemistry. Anti-calnexin antibodies exhibited complement-dependent cytotoxicity to lung cancer cell lines, resulting in a notable reduction in tumor growth in a subcutaneous xenograft model. Additionally, we verified the feasibility of labeling tumors through in vivo imaging using antibodies against calnexin. Furthermore, exosomal detection of calnexin suggested the potential utility of liquid biopsy for diagnostic purposes. In conclusion, this study establishes calnexin as a promising target for antibody-based lung cancer diagnosis and therapy, unlocking novel avenues for early detection and treatment. [BMB Reports 2024; 57(3): 155-160].

Comparison of survival prediction models for bone metastases of the extremities following surgery

Aims

This study aimed to compare the performance of survival prediction models for bone metastases of the extremities (BM-E) with pathological fractures in an Asian cohort, and investigate patient characteristics associated with survival.

Methods

This retrospective cohort study included 469 patients, who underwent surgery for BM-E between January 2009 and March 2022 at a tertiary hospital in South Korea. Postoperative survival was calculated using the PATHFx3.0, SPRING13, OPTIModel, SORG, and IOR models. Model performance was assessed with area under the curve (AUC), calibration curve, Brier score, and decision curve analysis. Cox regression analyses were performed to evaluate the factors contributing to survival.

Results

The SORG model demonstrated the highest discriminatory accuracy with AUC (0.80 (95% confidence interval (CI) 0.76 to 0.85)) at 12 months. In calibration analysis, the PATHfx3.0 and OPTIModel models underestimated survival, while the SPRING13 and IOR models overestimated survival. The SORG model exhibited excellent calibration with intercepts of 0.10 (95% CI -0.13 to 0.33) at 12 months. The SORG model also had lower Brier scores than the null score at three and 12 months, indicating good overall performance. Decision curve analysis showed that all five survival prediction models provided greater net benefit than the default strategy of operating on either all or no patients. Rapid growth cancer and low serum albumin levels were associated with three-, six-, and 12-month survival.

Conclusion

State-of-art survival prediction models for BM-E (PATHFx3.0, SPRING13, OPTIModel, SORG, and IOR models) are useful clinical tools for orthopaedic surgeons in the decision-making process for the treatment in Asian patients, with SORG models offering the best predictive performance. Rapid growth cancer and serum albumin level are independent, statistically significant factors contributing to survival following surgery of BM-E. Further refinement of survival prediction models will bring about informed and patient-specific treatment of BM-E.

The Prognostic Value of the Novel Global Immune-Nutrition-Inflammation Index (GINI) in Stage IIIC Non-Small Cell Lung Cancer Patients Treated with Concurrent Chemoradiotherapy

BACKGROUND

We sought to determine the prognostic value of the newly developed Global Immune-Nutrition-Inflammation Index (GINI) in patients with stage IIIC non-small cell lung cancer (NSCLC) who underwent definitive concurrent chemoradiotherapy (CCRT).

METHODS

This study was conducted on a cohort of 802 newly diagnosed stage IIIC NSCLC patients who underwent CCRT. The novel GINI created first here was defined as follows: GINI = [C-reactive protein × Platelets × Monocytes × Neutrophils] ÷ [Albumin × Lymphocytes]. The receiver operating characteristic (ROC) curve analysis was used to determine the optimal pre-CCRT GINI cut-off value that substantially interacts with the locoregional progression-free (LRPFS), progression-free (PFS), and overall survival (OS).

RESULTS

The optimal pre-CCRT GINI cutoff was 1562 (AUC: 76.1%; sensitivity: 72.4%; specificity: 68.2%; Youden index: 0.406). Patients presenting with a GINI ≥ 1562 had substantially shorter median LRPFS (13.3 vs. 18.4 months; p < 0.001), PFS (10.2 vs. 14.3 months; p < 0.001), and OS (19.1 vs. 37.8 months; p < 0.001) durations than those with a GINI < 1562. Results of the multivariate analysis revealed that the pre-CCRT GINI ≥ 1562 (vs. <1562), T4 tumor (vs. T3), and receiving only 1 cycle of concurrent chemotherapy (vs. 2-3 cycles) were the factors independently associated with poorer LRPS (p < 0.05 for each), PFS (p < 0.05 for each), and OS (p < 0.05 for each).

CONCLUSION

The newly developed GINI index efficiently divided the stage IIIC NSCLSC patients into two subgroups with substantially different median and long-term survival outcomes.

A deep learning-based cancer survival time classifier for small datasets

Cancer survival time prediction using Deep Learning (DL) has been an emerging area of research. However, non-availability of large-sized annotated medical imaging databases affects the training performance of DL models leading to their arguable usage in many clinical applications. In this research work, a neural network model is customized for small sample space to avoid data over-fitting for DL training. A set of prognostic radiomic features is selected through an iterative process using average of multiple dropouts which results in back-propagated gradients with low variance, thus increasing the network learning capability, reliable feature selection and better training over a small database. The proposed classifier is further compared with erasing feature selection method proposed in the literature for improved network training and with other well-known classifiers on small sample size. Achieved results which were statistically validated show efficient and improved classification of cancer survival time into three intervals of 6 months, between 6 months up to 2 years, and above 2 years; and has the potential to aid health care professionals in lung tumor evaluation for timely treatment and patient care.

Calibration and validation of modeled 5-year survival predictions among people with cystic fibrosis treated with the cystic fibrosis transmembrane conductance regulator modulator ivacaftor using United States registry data

OBJECTIVES

Cystic fibrosis (CF) is a rare genetic disease characterized by life-shortening lung function decline. Ivacaftor, a CF transmembrane conductance regulator modulator (CFTRm), was approved in 2012 for people with CF with specific gene mutations. We used real-world evidence of 5-year mortality impacts of ivacaftor in a US registry population to validate a CF disease-progression model that estimates the impact of ivacaftor on survival.

METHODS

The model projects the impact of ivacaftor vs. standard care in people with CF aged ≥6 years with CFTR gating mutations by combining parametric equations fitted to historical registry survival data, with mortality hazards adjusted for fixed and time-varying person-level characteristics. Disease progression with standard care was derived from published registry studies and the expected impact of ivacaftor on clinical characteristics was derived from clinical trials. Individual-level baseline characteristics of the registry ivacaftor-treated population were entered into the model; 5-year model-projected mortality with credible intervals (CrIs) was compared with registry mortality to evaluate the model's validity.

RESULTS

Post-calibration 5-year mortality projections closely approximated registry mortality in populations treated with standard care (6.4% modeled [95% CrI: 5.3% to 7.6%] vs. 6.0% observed) and ivacaftor (3.4% modeled [95% CrI: 2.7% to 4.4%] vs. 3.1% observed). The model accurately predicted 5-year relative risk of mortality (0.53 modeled [0.47 to 0.60] vs. 0.51 observed) in people treated with ivacaftor vs. standard care.

CONCLUSIONS

Modeled 5-year survival projections for people with CF initiating ivacaftor vs. standard care align closely with real-world registry data. Findings support the validity of modeling CF to predict long-term survival and estimate clinical and economic outcomes of CFTRm.

Interpretable deep learning survival predictive tool for small cell lung cancer

Background

Small cell lung cancer (SCLC) is an aggressive and almost universally lethal neoplasm. There is no accurate predictive method for its prognosis. Artificial intelligence deep learning may bring new hope.

Methods

By searching the Surveillance, Epidemiology, and End Results database (SEER), 21,093 patients' clinical data were eventually included. Data were then divided into two groups (train dataset/test dataset). The train dataset (diagnosed in 2010-2014, N = 17,296) was utilized to conduct a deep learning survival model, validated by itself and the test dataset (diagnosed in 2015, N = 3,797) in parallel. According to clinical experience, age, sex, tumor site, T, N, M stage (7th American Joint Committee on Cancer TNM stage), tumor size, surgery, chemotherapy, radiotherapy, and history of malignancy were chosen as predictive clinical features. The C-index was the main indicator to evaluate model performance.

Results

The predictive model had a 0.7181 C-index (95% confidence intervals, CIs, 0.7174-0.7187) in the train dataset and a 0.7208 C-index (95% CIs, 0.7202-0.7215) in the test dataset. These indicated that it had a reliable predictive value on OS for SCLC, so it was then packaged as a Windows software which is free for doctors, researchers, and patients to use.

Conclusion

The interpretable deep learning survival predictive tool for small cell lung cancer developed by this study had a reliable predictive value on their overall survival. More biomarkers may help improve the prognostic predictive performance of small cell lung cancer.

Risk-aware survival time prediction from whole slide pathological images

Deep-learning-based survival prediction can assist doctors by providing additional information for diagnosis by estimating the risk or time of death. The former focuses on ranking deaths among patients based on the Cox model, whereas the latter directly predicts the survival time of each patient. However, it is observed that survival time prediction for the patients, particularly with close observation times, possibly has incorrect orders, leading to low prediction accuracy. Therefore, in this paper, we present a whole slide image (WSI)-based survival time prediction method that takes advantage of both the risk as well as time prediction. Specifically, we propose to combine these two approaches by extracting the risk prediction features and using them as guides for the survival time prediction. Considering the high resolution of WSIs, we extract tumor patches from WSIs using a pre-trained tumor classifier and apply the graph convolutional network to aggregate information across these patches effectively. Extensive experiments demonstrate that the proposed method significantly improves the time prediction accuracy when compared with direct prediction of the survival times without guidance and outperforms existing methods.

Elevated Levels of Circulating Hsp70 and an Increased Prevalence of CD94+/CD69+ NK Cells Is Predictive for Advanced Stage Non-Small Cell Lung Cancer

Non-small cell lung cancer (NSCLC) is the second most frequently diagnosed tumor worldwide. Despite the clinical progress which has been achieved by multimodal therapies, including radiochemotherapy, and immune checkpoint inhibitor blockade, the overall survival of patients with advanced-stage NSCLC remains poor, with less than 16 months. It is well established that many aggressive tumor entities, including NSCLC, overexpress the major stress-inducible heat shock protein 70 (Hsp70) in the cytosol, present it on the plasma membrane in a tumor-specific manner, and release Hsp70 into circulation. Although high Hsp70 levels are associated with tumor aggressiveness and therapy resistance, membrane-bound Hsp70 can serve as a tumor-specific antigen for Hsp70-primed natural killer (NK) cells, expressing the C-type lectin receptor CD94, which is part of the activator receptor complex CD94/NKG2C. Therefore, we investigated circulating Hsp70 levels and changes in the composition of peripheral blood lymphocyte subsets as potential biomarkers for the advanced Union for International Cancer Control (UICC) stages in NSCLC. As expected, circulating Hsp70 levels were significantly higher in NSCLC patients compared to the healthy controls, as well as in patients with advanced UICC stages compared to those in UICC stage I. Smoking status did not influence the circulating Hsp70 levels significantly. Concomitantly, the proportions of CD4+ T helper cells were lower compared to the healthy controls and stage I tumor patients, whereas that of CD8+ cytotoxic T cells was progressively higher. The prevalence of CD3-/CD56+, CD3-/NKp30, CD3-/NKp46+, and CD3-/NKG2D+ NK cells was higher in stage IV/IIIB of the disease than in stage IIIA but were not statistically different from that in healthy individuals. However, the proportion of NK cells expressing CD94 and the activation/exhaustion marker CD69 significantly increased in higher tumor stages compared with stage I and the healthy controls. We speculate that although elevated circulating Hsp70 levels might promote the prevalence of CD94+ NK cells in patients with advanced-stage NSCLC, the cytolytic activity of these NK cells also failed to control tumor growth due to insufficient support by pro-inflammatory cytokines from CD4+ T helper cells. This hypothesis is supported by a comparative multiplex cytokine analysis of the blood in lung cancer patients with a low proportion of CD4+ T cells, a high proportion of NK cells, and high Hsp70 levels versus patients with a high proportion of CD4+ T cells exhibiting lower IL-2, IL-4, IL-6, IFN-γ, granzyme B levels.

Cost-effectiveness of proton radiotherapy versus photon radiotherapy for non-small cell lung cancer patients: Exploring the model-based approach

INTRODUCTION

Proton radiotherapy (PT) is a promising but more expensive strategy than photon radiotherapy (XRT) for the treatment of non-small cell lung cancer (NSCLC). PT is probably not cost-effective for all patients. Therefore, patients can be selected using normal tissue complication probability (NTCP) models with predefined criteria. This study aimed to explore the cost-effectiveness of three treatment strategies for patients with stage III NSCLC: 1. photon radiotherapy for all patients (XRT_All); 2. PT for all patients (PT_All); 3. PT for selected patients (PT_{Individualized}).

METHODS

A decision-analytical model was constructed to estimate and compare costs and QALYs of all strategies. Three radiation-related toxicities were included: dyspnea, dysphagia and cardiotoxicity. Costs and QALY's were incorporated for grade 2 and ≥ 3 toxicities separately. Incremental Cost-Effectiven Ratios (ICERs) were calculated and compared to a threshold value of €80,000. Additionally, scenario, sensitivity and value of information analyses were performed.

RESULTS

PT_All yielded most QALYs, but was also most expensive. XRT_All was the least effective and least expensive strategy, and the most cost-effective strategy. For thresholds higher than €163,467 per QALY gained, PT_{Individualized} was cost-effective. When assuming equal minutes per fraction (15 minutes) for PT and XRT, PT_{Individualized} was considered the most cost-effective strategy (ICER: €76,299).

CONCLUSION

Currently, PT is not cost-effective for all patients, nor for patient selected on the current NTCP models used in the Dutch indication protocol. However, with improved clinical experience, personnel and treatment costs of PT can decrease over time, which potentially leads to PT_{Individualized}, with optimal patient selection, will becoming a cost-effective strategy.

Radiomic and Volumetric Measurements as Clinical Trial Endpoints—A Comprehensive Review

Simple Summary

The extraction of quantitative data from standard-of-care imaging modalities offers opportunities to improve the relevance and salience of imaging biomarkers used in drug development. This review aims to identify the challenges and opportunities for discovering new imaging-based biomarkers based on radiomic and volumetric assessment in the single-site solid tumor sites: breast cancer, rectal cancer, lung cancer and glioblastoma. Developing approaches to harmonize three essential areas: segmentation, validation and data sharing may expedite regulatory approval and adoption of novel cancer imaging biomarkers.

Abstract

Clinical trials for oncology drug development have long relied on surrogate outcome biomarkers that assess changes in tumor burden to accelerate drug registration (i.e., Response Evaluation Criteria in Solid Tumors version 1.1 (RECIST v1.1) criteria). Drug-induced reduction in tumor size represents an imperfect surrogate marker for drug activity and yet a radiologically determined objective response rate is a widely used endpoint for Phase 2 trials. With the addition of therapies targeting complex biological systems such as immune system and DNA damage repair pathways, incorporation of integrative response and outcome biomarkers may add more predictive value. We performed a review of the relevant literature in four representative tumor types (breast cancer, rectal cancer, lung cancer and glioblastoma) to assess the preparedness of volumetric and radiomics metrics as clinical trial endpoints. We identified three key areas—segmentation, validation and data sharing strategies—where concerted efforts are required to enable progress of volumetric- and radiomics-based clinical trial endpoints for wider clinical implementation.

Radioresistance of Non-Small Cell Lung Cancers and Therapeutic Perspectives

Survival in unresectable locally advanced stage non-small cell lung cancer (NSCLC) patients remains poor despite chemoradiotherapy. Recently, adjuvant immunotherapy improved survival for these patients but we are still far from curing most of the patients with only a 57% survival remaining at 3 years. This poor survival is due to the resistance to chemoradiotherapy, local relapses, and distant relapses. Several biological mechanisms have been found to be involved in the chemoradioresistance such as cancer stem cells, cancer mutation status, or the immune system. New drugs to overcome this radioresistance in NSCLCs have been investigated such as radiosensitizer treatments or immunotherapies. Different modalities of radiotherapy have also been investigated to improve efficacity such as dose escalation or proton irradiations. In this review, we focused on biological mechanisms such as the cancer stem cells, the cancer mutations, the antitumor immune response in the first part, then we explored some strategies to overcome this radioresistance in stage III NSCLCs with new drugs or radiotherapy modalities.

Identification of a TGF-β signaling-related gene signature for prediction of immunotherapy and targeted therapy for lung adenocarcinoma

BACKGROUND

Transforming growth factor (TGF)-β signaling functions importantly in regulating tumor microenvironment (TME). This study developed a prognostic gene signature based on TGF-β signaling-related genes for predicting clinical outcome of patients with lung adenocarcinoma (LUAD).

METHODS

TGF-β signaling-related genes came from The Molecular Signature Database (MSigDB). LUAD prognosis-related genes were screened from all the genes involved in TGF-β signaling using least absolute shrinkage and selection operator (LASSO) Cox regression analysis and then used to establish a risk score model for LUAD. ESTIMATE and CIBERSORT analyzed infiltration of immune cells in TME. Immunotherapy response was analyzed by the TIDE algorithm.

RESULTS

A LUAD prognostic 5-gene signature was developed based on 54 TGF-β signaling-related genes. Prognosis of high-risk patients was significantly worse than low-risk patients. Both internal validation and external dataset validation confirmed a high precision of the risk model in predicting the clinical outcomes of LUAD patients. Multivariate Cox analysis demonstrated the model independence in OS prediction of LUAD. The risk model was significantly related to the infiltration of 9 kinds of immune cells, matrix, and immune components in TME. Low-risk patients tended to respond more actively to anti-PD-1 treatment, while high-risk patients were more sensitive to chemotherapy and targeted therapy.

CONCLUSIONS

The 5-gene signature based on TGF-β signaling-related genes showed potential for LUAD management.

Prognosis of lung cancer with simple brain metastasis patients and establishment of survival prediction models: a study based on real events

Objectives

The aim of this study was to explore risk factors for the prognosis of lung cancer with simple brain metastasis (LCSBM) patients and to establish a prognostic predictive nomogram for LCSBM patients.

Materials and methods

Three thousand eight hundred and six cases of LCSBM were extracted from the Surveillance, Epidemiology, and End Results (SEER) database from 2010 to 2015 using SEER Stat 8.3.5. Lung cancer patients only had brain metastasis with no other organ metastasis were defined as LCSBM patients. Prognostic factors of LCSBM were analyzed with log-rank method and Cox proportional hazards model. Independent risk and protective prognostic factors were used to construct nomogram with accelerated failure time model. C-index was used to evaluate the prediction effect of nomogram.

Results and conclusion

The younger patients (18–65 years old) accounted for 54.41%, while patients aged over 65 accounted for 45.59%.The ratio of male: female was 1:1. Lung cancer in the main bronchus, upper lobe, middle lobe and lower lobe were accounted for 4.91%, 62.80%, 4.47% and 27.82% respectively; and adenocarcinoma accounted for 57.83% of all lung cancer types. The overall median survival time was 12.2 months. Survival rates for 1-, 3- and 5-years were 28.2%, 8.7% and 4.7% respectively. We found female (HR = 0.81, 95% CI 0.75–0.87), the married (HR = 0.80; 95% CI 0.75–0.86), the White (HR = 0.90, 95% CI 0.84–0.95) and primary site (HR = 0.45, 95% CI 0.39–0.52) were independent protective factors while higher age (HR = 1.51, 95% CI 1.40–1.62), advanced grade (HR = 1.19, 95% CI 1.12–1.25) and advanced T stage (HR = 1.09, 95% CI 1.05–1.13) were independent risk prognostic factors affecting the survival of LCSBM patients. We constructed the nomogram with above independent factors, and the C-index value was 0.634 (95% CI 0.622–0.646). We developed a nomogram with seven significant LCSBM independent prognostic factors to provide prognosis prediction.

Predicting 2-year survival in stage I-III non-small cell lung cancer: the development and validation of a scoring system from an Australian cohort

Background

There are limited data on survival prediction models in contemporary inoperable non-small cell lung cancer (NSCLC) patients. The objective of this study was to develop and validate a survival prediction model in a cohort of inoperable stage I-III NSCLC patients treated with radiotherapy.

Methods

Data from inoperable stage I-III NSCLC patients diagnosed from 1/1/2016 to 31/12/2017 were collected from three radiation oncology clinics. Patient, tumour and treatment-related variables were selected for model inclusion using univariate and multivariate analysis. Cox proportional hazards regression was used to develop a 2-year overall survival prediction model, the South West Sydney Model (SWSM) in one clinic (n = 117) and validated in the other clinics (n = 144). Model performance, assessed internally and on one independent dataset, was expressed as Harrell’s concordance index (c-index).

Results

The SWSM contained five variables: Eastern Cooperative Oncology Group performance status, diffusing capacity of the lung for carbon monoxide, histological diagnosis, tumour lobe and equivalent dose in 2 Gy fractions. The SWSM yielded a c-index of 0.70 on internal validation and 0.72 on external validation. Survival probability could be stratified into three groups using a risk score derived from the model.

Conclusions

A 2-year survival model with good discrimination was developed. The model included tumour lobe as a novel variable and has the potential to guide treatment decisions. Further validation is needed in a larger patient cohort.

A comparison of machine learning methods for predicting recurrence and death after curative-intent radiotherapy for non-small cell lung cancer: Development and validation of multivariable clinical prediction models

BACKGROUND

Surveillance is universally recommended for non-small cell lung cancer (NSCLC) patients treated with curative-intent radiotherapy. High-quality evidence to inform optimal surveillance strategies is lacking. Machine learning demonstrates promise in accurate outcome prediction for a variety of health conditions. The purpose of this study was to utilise readily available patient, tumour, and treatment data to develop, validate and externally test machine learning models for predicting recurrence, recurrence-free survival (RFS) and overall survival (OS) at 2 years from treatment.

METHODS

A retrospective, multicentre study of patients receiving curative-intent radiotherapy for NSCLC was undertaken. A total of 657 patients from 5 hospitals were eligible for inclusion. Data pre-processing derived 34 features for predictive modelling. Combinations of 8 feature reduction methods and 10 machine learning classification algorithms were compared, producing risk-stratification models for predicting recurrence, RFS and OS. Models were compared with 10-fold cross validation and an external test set and benchmarked against TNM-stage and performance status. Youden Index was derived from validation set ROC curves to distinguish high and low risk groups and Kaplan-Meier analyses performed.

FINDINGS

Median follow-up time was 852 days. Parameters were well matched across training-validation and external test sets: Mean age was 73 and 71 respectively, and recurrence, RFS and OS rates at 2 years were 43% vs 34%, 54% vs 47% and 54% vs 47% respectively. The respective validation and test set AUCs were as follows: 1) RFS: 0·682 (0·575-0·788) and 0·681 (0·597-0·766), 2) Recurrence: 0·687 (0·582-0·793) and 0·722 (0·635-0·81), and 3) OS: 0·759 (0·663-0·855) and 0·717 (0·634-0·8). Our models were superior to TNM stage and performance status in predicting recurrence and OS.

INTERPRETATION

This robust and ready to use machine learning method, validated and externally tested, sets the stage for future clinical trials entailing quantitative personalised risk-stratification and surveillance following curative-intent radiotherapy for NSCLC.

FUNDING

A full list of funding bodies that contributed to this study can be found in the Acknowledgements section.

Multitask Learning Radiomics on Longitudinal Imaging to Predict Survival Outcomes following Risk-Adaptive Chemoradiation for Non-Small Cell Lung Cancer

Medical imaging provides quantitative and spatial information to evaluate treatment response in the management of patients with non-small cell lung cancer (NSCLC). High throughput extraction of radiomic features on these images can potentially phenotype tumors non-invasively and support risk stratification based on survival outcome prediction. The prognostic value of radiomics from different imaging modalities and time points prior to and during chemoradiation therapy of NSCLC, relative to conventional imaging biomarker or delta radiomics models, remains uncharacterized. We investigated the utility of multitask learning of multi-time point radiomic features, as opposed to single-task learning, for improving survival outcome prediction relative to conventional clinical imaging feature model benchmarks. Survival outcomes were prospectively collected for 45 patients with unresectable NSCLC enrolled on the FLARE-RT phase II trial of risk-adaptive chemoradiation and optional consolidation PD-L1 checkpoint blockade (NCT02773238). FDG-PET, CT, and perfusion SPECT imaging pretreatment and week 3 mid-treatment was performed and 110 IBSI-compliant pyradiomics shape-/intensity-/texture-based features from the metabolic tumor volume were extracted. Outcome modeling consisted of a fused Laplacian sparse group LASSO with component-wise gradient boosting survival regression in a multitask learning framework. Testing performance under stratified 10-fold cross-validation was evaluated for multitask learning radiomics of different imaging modalities and time points. Multitask learning models were benchmarked against conventional clinical imaging and delta radiomics models and evaluated with the concordance index (c-index) and index of prediction accuracy (IPA). FDG-PET radiomics had higher prognostic value for overall survival in test folds (c-index 0.71 [0.67, 0.75]) than CT radiomics (c-index 0.64 [0.60, 0.71]) or perfusion SPECT radiomics (c-index 0.60 [0.57, 0.63]). Multitask learning of pre-/mid-treatment FDG-PET radiomics (c-index 0.71 [0.67, 0.75]) outperformed benchmark clinical imaging (c-index 0.65 [0.59, 0.71]) and FDG-PET delta radiomics (c-index 0.52 [0.48, 0.58]) models. Similarly, the IPA for multitask learning FDG-PET radiomics (30%) was higher than clinical imaging (26%) and delta radiomics (15%) models. Radiomics models performed consistently under different voxel resampling conditions. Multitask learning radiomics for outcome modeling provides a clinical decision support platform that leverages longitudinal imaging information. This framework can reveal the relative importance of different imaging modalities and time points when designing risk-adaptive cancer treatment strategies.

Evaluation of Prognostic and Predictive Models in the Oncology Clinic

Predictive and prognostic models hold great potential to support clinical decision making in oncology and could ultimately facilitate a paradigm shift to a more personalised form of treatment. While a large number of models relevant to the field of oncology have been developed, few have been translated into clinical use and assessment of clinical utility is not currently considered a routine part of model development. In this narrative review of the clinical evaluation of prediction models in oncology, we propose a high-level process diagram for the life cycle of a clinical model, encompassing model commissioning, clinical implementation and ongoing quality assurance, which aims to bridge the gap between model development and clinical implementation.

Prognosis prediction of extremity and trunk wall soft-tissue sarcomas treated with surgical resection with radiomic analysis based on random survival forest

Many researches have applied machine learning methods to find associations between radiomic features and clinical outcomes. Random survival forests (RSF), as an accurate classifier, sort all candidate variables as the rank of importance values. There was no study concerning on finding radiomic predictors in patients with extremity and trunk wall soft-tissue sarcomas using RSF. This study aimed to determine associations between radiomic features and overall survival (OS) by RSF analysis. To identify radiomic features with important values by RSF analysis, construct predictive models for OS incorporating clinical characteristics, and evaluate models' performance with different method. We collected clinical characteristics and radiomic features extracted from plain and contrast-enhanced computed tomography (CT) from 353 patients with extremity and trunk wall soft-tissue sarcomas treated with surgical resection. All radiomic features were analyzed by Cox proportional hazard (CPH) and followed RSF analysis. The association between radiomics-predicted risks and OS was assessed by Kaplan-Meier analysis. All clinical features were screened by CPH analysis. Prognostic clinical and radiomic parameters were fitted into RSF and CPH integrative models for OS in the training cohort, respectively. The concordance indexes (C-index) and Brier scores of both two models were evaluated in both training and testing cohorts. The model with better predictive performance was interpreted with nomogram and calibration plots. Among all 86 radiomic features, there were three variables selected with high importance values. The RSF on these three features distinguished patients with high predicted risks from patients with low predicted risks for OS in the training set (P < 0.001) using Kaplan-Meier analysis. Age, lymph node involvement and grade were incorporated into the combined models for OS (P < 0.05). The C-indexes in both two integrative models fluctuated above 0.80 whose Brier scores maintained less than 15.0 in the training and testing datasets. The RSF model performed little advantages over the CPH model that the calibration curve of the RSF model showed favorable agreement between predicted and actual survival probabilities for the 3-year and 5-year survival prediction. The multimodality RSF model including clinical and radiomic characteristics conducted high capacity in prediction of OS which might assist individualized therapeutic regimens. Level III, prognostic study.

Development and validation of a prediction model using molecular marker for long-term survival in unresectable stage III non-small cell lung cancer treated with chemoradiotherapy

Background

This study aimed to establish a predictive nomogram integrating epidermal growth factor receptor (EGFR) mutation status for 3‐ and 5‐year overall survival (OS) in unresectable/inoperable stage III non‐small cell lung cancer (NSCLC) treated with definitive chemoradiotherapy.

Methods

A total of 533 stage III NSCLC patients receiving chemoradiotherapy from 2013 to 2017 in our institution were included and divided into training and testing sets (2:1). Significant factors impacting OS were identified in the training set and integrated into the nomogram based on Cox proportional hazards regression. The model was subject to bootstrap internal validation and external validation within the testing set and an independent cohort from a phase III trial. The accuracy and discriminative capacity of the model were examined by calibration plots, C‐indexes and risk stratifications.

Results

The final multivariate model incorporated sex, smoking history, histology (including EGFR mutation status), TNM stage, planning target volume, chemotherapy sequence and radiation pneumonitis grade. The bootstrapped C‐indexes in the training set were 0.688, 0.710 for the 3‐ and 5‐year OS. For external validation, C‐indexes for 3‐ and 5‐year OS were 0.717, 0.720 in the testing set and 0.744, 0.699 in the external testing cohort, respectively. The calibration plots presented satisfying accuracy. The derivative risk stratification strategy classified patients into distinct survival subgroups successfully and performed better than the traditional TNM staging.

Conclusions

The nomogram incorporating EGFR mutation status could facilitate survival prediction and risk stratification for individual stage III NSCLC, providing information for enhanced immunotherapy decision and future trial design.

A narrative review of prognosis prediction models for non-small cell lung cancer: what kind of predictors should be selected and how to improve models?

Objective

To discover potential predictors and explore how to build better models by summarizing the existing prognostic prediction models of non-small cell lung cancer (NSCLC).

Background

Research on clinical prediction models of NSCLC has experienced explosive growth in recent years. As more predictors of prognosis are discovered, the choice of predictors to build models is particularly important, and in the background of more applications of next-generation sequencing technology, gene-related predictors are widely used. As it is more convenient to obtain samples and follow-up data, the prognostic model is preferred by researchers.

Methods

PubMed and the Cochrane Library were searched using the items “NSCLC”, “prognostic model”, “prognosis prediction”, and “survival prediction” from 1 January 1980 to 5 May 2021. Reference lists from articles were reviewed and relevant articles were identified.

Conclusions

The performance of gene-related models has not obviously improved. Relative to the innovation and diversity of predictors, it is more important to establish a highly stable model that is convenient for clinical application. Most of the prevalent models are highly biased and referring to PROBAST at the beginning of the study may be able to significantly control the bias. Existing models should be validated in a large external dataset to make a meaningful comparison.

Lymph Node Parameters Predict Adjuvant Chemoradiotherapy Efficacy and Disease-Free Survival in Pathologic N2 Non-Small Cell Lung Cancer

Pathologic N2 non-small cell lung cancer (NSCLC) is prominently intrinsically heterogeneous. We aimed to identify homogeneous prognostic subgroups and evaluate the role of different adjuvant treatments. We retrospectively collected patients with resected pathologic T1-3N2M0 NSCLC from the Shanghai Chest Hospital as the primary cohort and randomly allocated them (3:1) to the training set and the validation set 1. We had patients from the Fudan University Shanghai Cancer Center as an external validation cohort (validation set 2) with the same inclusion and exclusion criteria. Variables significantly related to disease-free survival (DFS) were used to build an adaptive Elastic-Net Cox regression model. Nomogram was used to visualize the model. The discriminative and calibration abilities of the model were assessed by time-dependent area under the receiver operating characteristic curves (AUCs) and calibration curves. The primary cohort consisted of 1,312 patients. Tumor size, histology, grade, skip N2, involved N2 stations, lymph node ratio (LNR), and adjuvant treatment pattern were identified as significant variables associated with DFS and integrated into the adaptive Elastic-Net Cox regression model. A nomogram was developed to predict DFS. The model showed good discrimination (the median AUC in the validation set 1: 0.66, range 0.62 to 0.71; validation set 2: 0.66, range 0.61 to 0.73). We developed and validated a nomogram that contains multiple variables describing lymph node status (skip N2, involved N2 stations, and LNR) to predict the DFS of patients with resected pathologic N2 NSCLC. Through this model, we could identify a subtype of NSCLC with a more malignant clinical biological behavior and found that this subtype remained at high risk of disease recurrence after adjuvant chemoradiotherapy.

Privacy preserving distributed learning classifiers - Sequential learning with small sets of data

BACKGROUND

Artificial intelligence (AI) typically requires a significant amount of high-quality data to build reliable models, where gathering enough data within a single institution can be particularly challenging. In this study we investigated the impact of using sequential learning to exploit very small, siloed sets of clinical and imaging data to train AI models. Furthermore, we evaluated the capacity of such models to achieve equivalent performance when compared to models trained with the same data over a single centralized database.

METHODS

We propose a privacy preserving distributed learning framework, learning sequentially from each dataset. The framework is applied to three machine learning algorithms: Logistic Regression, Support Vector Machines (SVM), and Perceptron. The models were evaluated using four open-source datasets (Breast cancer, Indian liver, NSCLC-Radiomics dataset, and Stage III NSCLC).

FINDINGS

The proposed framework ensured a comparable predictive performance against a centralized learning approach. Pairwise DeLong tests showed no significant difference between the compared pairs for each dataset.

INTERPRETATION

Distributed learning contributes to preserve medical data privacy. We foresee this technology will increase the number of collaborative opportunities to develop robust AI, becoming the default solution in scenarios where collecting enough data from a single reliable source is logistically impossible. Distributed sequential learning provides privacy persevering means for institutions with small but clinically valuable datasets to collaboratively train predictive AI while preserving the privacy of their patients. Such models perform similarly to models that are built on a larger central dataset.

Hsp70 in Liquid Biopsies-A Tumor-Specific Biomarker for Detection and Response Monitoring in Cancer

In contrast to normal cells, tumor cells of multiple entities overexpress the Heat shock protein 70 (Hsp70) not only in the cytosol, but also present it on their plasma membrane in a tumor-specific manner. Furthermore, membrane Hsp70-positive tumor cells actively release Hsp70 in small extracellular vesicles with biophysical characteristics of exosomes. Due to conformational changes of Hsp70 in a lipid environment, most commercially available antibodies fail to detect membrane-bound and vesicular Hsp70. To fill this gap and to assess the role of vesicular Hsp70 in circulation as a potential tumor biomarker, we established the novel complete (comp)Hsp70 sandwich ELISA, using two monoclonal antibodies (mAbs), that is able to recognize both free and lipid-associated Hsp70 on the cell surface of viable tumor cells and on small extracellular vesicles. The epitopes of the mAbs cmHsp70.1 (aa 451-461) and cmHsp70.2 (aa 614-623) that are conserved among different species reside in the substrate-binding domain of Hsp70 with measured affinities of 0.42 nM and 0.44 nM, respectively. Validation of the compHsp70 ELISA revealed a high intra- and inter-assay precision, linearity in a concentration range of 1.56 to 25 ng/mL, high recovery rates of spiked liposomal Hsp70 (>84%), comparable values between human serum and plasma samples and no interference by food intake or age of the donors. Hsp70 concentrations in the circulation of patients with glioblastoma, squamous cell or adeno non-small cell lung carcinoma (NSCLC) at diagnosis were significantly higher than those of healthy donors. Hsp70 concentrations dropped concomitantly with a decrease in viable tumor mass upon irradiation of patients with approximately 20 Gy (range 18-22.5 Gy) and after completion of radiotherapy (60-70 Gy). In summary, the compHsp70 ELISA presented herein provides a sensitive and reliable tool for measuring free and vesicular Hsp70 in liquid biopsies of tumor patients, levels of which can be used as a tumor-specific biomarker, for risk assessment (i.e., differentiation of grade III vs. IV adeno NSCLC) and monitoring of therapeutic outcomes.

Towards the Interpretability of Machine Learning Predictions for Medical Applications Targeting Personalised Therapies: A Cancer Case Survey

Artificial Intelligence is providing astonishing results, with medicine being one of its favourite playgrounds. Machine Learning and, in particular, Deep Neural Networks are behind this revolution. Among the most challenging targets of interest in medicine are cancer diagnosis and therapies but, to start this revolution, software tools need to be adapted to cover the new requirements. In this sense, learning tools are becoming a commodity but, to be able to assist doctors on a daily basis, it is essential to fully understand how models can be interpreted. In this survey, we analyse current machine learning models and other in-silico tools as applied to medicine-specifically, to cancer research-and we discuss their interpretability, performance and the input data they are fed with. Artificial neural networks (ANN), logistic regression (LR) and support vector machines (SVM) have been observed to be the preferred models. In addition, convolutional neural networks (CNNs), supported by the rapid development of graphic processing units (GPUs) and high-performance computing (HPC) infrastructures, are gaining importance when image processing is feasible. However, the interpretability of machine learning predictions so that doctors can understand them, trust them and gain useful insights for the clinical practice is still rarely considered, which is a factor that needs to be improved to enhance doctors' predictive capacity and achieve individualised therapies in the near future.

A systematic review and meta-analysis of treatment-related toxicities of curative and palliative radiation therapy in non-small cell lung cancer

Treatment-related toxicity is an important component in non-small cell lung cancer (NSCLC) management decision-making. Our aim was to evaluate and compare the toxicity rates of curative and palliative radiotherapy with and without chemotherapy. This meta-analysis provides better quantitative estimates of the toxicities compared to individual trials. A systematic review of randomised trials with > 50 unresectable NSCLC patients, treated with curative or palliative conventional radiotherapy (RT) with or without chemotherapy. Data was extracted for oesophagitis, pneumonitis, cardiac events, pulmonary fibrosis, myelopathy and neutropenia by any grade, grade ≥ 3 and treatment-related deaths. Mantel-Haenszel fixed-effect method was used to obtain pooled risk ratio. Forty-nine trials with 8609 evaluable patients were included. There was significantly less grade ≥ 3 acute oesophagitis (6.4 vs 22.2%, p < 0.0001) and any grade oesophagitis (70.4 vs 79.0%, p = 0.04) for sequential CRT compared to concurrent CRT, with no difference in pneumonitis (grade ≥ 3 or any grade), neutropenia (grade ≥ 3), cardiac events (grade ≥ 3) or treatment-related deaths. Although the rate of toxicity increased with intensification of treatment with RT, the only significant difference between treatment regimens was the rate of oesophagitis between the use of concurrent and sequential CRT. This can aid clinicians in radiotherapy decision making for NSCLC.

Predicting the Risk of Disease Recurrence and Death Following Curative-intent Radiotherapy for Non-small Cell Lung Cancer: The Development and Validation of Two Scoring Systems From a Large Multicentre UK Cohort

AIMS

There is a paucity of evidence on which to produce recommendations on neither the clinical nor the imaging follow-up of lung cancer patients after curative-intent radiotherapy. In the 2019 National Institute for Health and Care Excellence lung cancer guidelines, further research into risk-stratification models to inform follow-up protocols was recommended.

MATERIALS AND METHODS

A retrospective study of consecutive patients undergoing curative-intent radiotherapy for non-small cell lung cancer from 1 October 2014 to 1 October 2016 across nine UK trusts was carried out. Twenty-two demographic, clinical and treatment-related variables were collected and multivariable logistic regression was used to develop and validate two risk-stratification models to determine the risk of disease recurrence and death.

RESULTS

In total, 898 patients were included in the study. The mean age was 72 years, 63% (562/898) had a good performance status (0-1) and 43% (388/898), 15% (134/898) and 42% (376/898) were clinical stage I, II and III, respectively. Thirty-six per cent (322/898) suffered disease recurrence and 41% (369/898) died in the first 2 years after radiotherapy. The ASSENT score (age, performance status, smoking status, staging endobronchial ultrasound, N-stage, T-stage) was developed, which stratifies the risk for disease recurrence within 2 years, with an area under the receiver operating characteristic curve (AUROC) for the total score of 0.712 (0.671-0.753) and 0.72 (0.65-0.789) in the derivation and validation sets, respectively. The STEPS score (sex, performance status, staging endobronchial ultrasound, T-stage, N-stage) was developed, which stratifies the risk of death within 2 years, with an AUROC for the total score of 0.625 (0.581-0.669) and 0.607 (0.53-0.684) in the derivation and validation sets, respectively.

CONCLUSIONS

These validated risk-stratification models could be used to inform follow-up protocols after curative-intent radiotherapy for lung cancer. The modest performance highlights the need for more advanced risk prediction tools.

Prognostic factors for overall survival of stage III non-small cell lung cancer patients on computed tomography: A systematic review and meta-analysis

INTRODUCTION

Prognosis prediction is central in treatment decision making and quality of life for non-small cell lung cancer (NSCLC) patients. However, conventional computed tomography (CT) related prognostic factors may not apply to the challenging stage III NSCLC group. The aim of this systematic review was therefore to identify and evaluate CT-related prognostic factors for overall survival (OS) of stage III NSCLC.

METHODS

The Medline, Embase, and Cochrane electronic databases were searched. After study selection, risk of bias was estimated for the included studies. Meta-analysis of univariate results was performed when sufficient data were available.

RESULTS

1595 of the 11,996 retrieved records were selected for full text review, leading to inclusion of 65 studies that reported data of 144,513 stage III NSCLC patients andcompromising 26 unique CT-related prognostic factors. Relevance and validity varied substantially, few studies had low relevance and validity. Only four studies evaluated the added value of new prognostic factors compared with recognized clinical factors. Included studies suggested gross tumor volume (meta-analysis: HR = 1.22, 95%CI: 1.05-1.42), tumor diameter, nodal volume, and pleural effusion, are prognostic in patients treated with chemoradiation. Clinical T-stage and location (right/left) were likely not prognostic within stage III NSCLC. Inconclusive are several radiomic features, tumor volume, atelectasis, location (pulmonary lobes, central/peripheral), interstitial lung abnormalities, great vessel invasion, pit-fall sign, and cavitation.

CONCLUSIONS

Tumor-size and nodal size-related factors are prognostic for OS in stage III NSCLC. Future studies should carefully report study characteristics and contrast factors with guideline recognized factors to improve evidence evaluation and validation.

Targeted Natural Killer Cell-Based Adoptive Immunotherapy for the Treatment of Patients with NSCLC after Radiochemotherapy: A Randomized Phase II Clinical Trial

PURPOSE

Non-small cell lung cancer (NSCLC) is a fatal disease with poor prognosis. A membrane-bound form of Hsp70 (mHsp70) which is selectively expressed on high-risk tumors serves as a target for mHsp70-targeting natural killer (NK) cells. Patients with advanced mHsp70-positive NSCLC may therefore benefit from a therapeutic intervention involving mHsp70-targeting NK cells. The randomized phase II clinical trial (EudraCT2008-002130-30) explores tolerability and efficacy of ex vivo-activated NK cells in patients with NSCLC after radiochemotherapy (RCT).

PATIENTS AND METHODS

Patients with unresectable, mHsp70-positive NSCLC (stage IIIa/b) received 4 cycles of autologous NK cells activated ex vivo with TKD/IL2 [interventional arm (INT)] after RCT (60-70 Gy, platinum-based chemotherapy) or RCT alone [control arm (CTRL)]. The primary objective was progression-free survival (PFS), and secondary objectives were the assessment of quality of life (QoL, QLQ-LC13), toxicity, and immunobiological responses.

RESULTS

The NK-cell therapy after RCT was well tolerated, and no differences in QoL parameters between the two study arms were detected. Estimated 1-year probabilities for PFS were 67% [95% confidence interval (CI), 19%-90%] for the INT arm and 33% (95% CI, 5%-68%) for the CTRL arm (P = 0.36, 1-sided log-rank test). Clinical responses in the INT group were associated with an increase in the prevalence of activated NK cells in their peripheral blood.

CONCLUSIONS

Ex vivo TKD/IL2-activated, autologous NK cells are well tolerated and deliver positive clinical responses in patients with advanced NSCLC after RCT.

Artificial intelligence in radiation oncology

Artificial intelligence (AI) has the potential to fundamentally alter the way medicine is practised. AI platforms excel in recognizing complex patterns in medical data and provide a quantitative, rather than purely qualitative, assessment of clinical conditions. Accordingly, AI could have particularly transformative applications in radiation oncology given the multifaceted and highly technical nature of this field of medicine with a heavy reliance on digital data processing and computer software. Indeed, AI has the potential to improve the accuracy, precision, efficiency and overall quality of radiation therapy for patients with cancer. In this Perspective, we first provide a general description of AI methods, followed by a high-level overview of the radiation therapy workflow with discussion of the implications that AI is likely to have on each step of this process. Finally, we describe the challenges associated with the clinical development and implementation of AI platforms in radiation oncology and provide our perspective on how these platforms might change the roles of radiotherapy medical professionals.

Deep Learning of Imaging Phenotype and Genotype for Predicting Overall Survival Time of Glioblastoma Patients

Glioblastoma (GBM) is the most common and deadly malignant brain tumor. For personalized treatment, an accurate pre-operative prognosis for GBM patients is highly desired. Recently, many machine learning-based methods have been adopted to predict overall survival (OS) time based on the pre-operative mono- or multi-modal imaging phenotype. The genotypic information of GBM has been proven to be strongly indicative of the prognosis; however, this has not been considered in the existing imaging-based OS prediction methods. The main reason is that the tumor genotype is unavailable pre-operatively unless deriving from craniotomy. In this paper, we propose a new deep learning-based OS prediction method for GBM patients, which can derive tumor genotype-related features from pre-operative multimodal magnetic resonance imaging (MRI) brain data and feed them to OS prediction. Specifically, we propose a multi-task convolutional neural network (CNN) to accomplish both tumor genotype and OS prediction tasks jointly. As the network can benefit from learning tumor genotype-related features for genotype prediction, the accuracy of predicting OS time can be prominently improved. In the experiments, multimodal MRI brain dataset of 120 GBM patients, with as many as four different genotypic/molecular biomarkers, are used to evaluate our method. Our method achieves the highest OS prediction accuracy compared to other state-of-the-art methods.

Development and External Validation of a Nomogram for Predicting Survival in Patients With Stage IA Non-small Cell Lung Cancer ≤2 cm Undergoing Sublobectomy

Background: Postoperative prognosis of early stage non-small cell lung cancer (NSCLC) undergoing sublobectomy is heterogeneous. Therefore, we sought to construct a novel survival prediction model for stage IA NSCLC ≤2 cm undergoing sublobectomy.

Methods: Based on the data from the Surveillance, Epidemiology, and End Results (SEER) program, we successfully determined and incorporated independent prognostic markers to construct the nomogram. Internal validation of the constructed nomogram was conducted through 1,000 bootstrap resamples. The constructed nomogram was further subjected to external validation with an independent cohort of patients from two Chinese institutions. The performance of the survival prediction model was assessed by concordance index, calibration plots, and risk subgroup classification.

Results: A total of 3,238 patients from SEER registries (development cohort), as well as 769 patients from two Chinese institutions (validation cohort) was included. Gender, age, size, histologic type, grade, and examined lymph nodes count were identified as significant prognostic parameters. A novel nomogram was developed and externally validated. Concordance index of constructed nomogram was significantly better than that of the current TNM staging system. Calibration plots demonstrated an optimal consistency between the nomogram predicted and actual observed probability of survival. Survival curves of different risk subgroups within respective TNM stage demonstrated significant distinctions.

Conclusion: We developed and externally validated a survival prediction model for patients with stage IA NSCLC ≤2 cm undergoing sublobectomy. This novel nomogram outperforms the conventional TNM staging system and could help clinicians in postoperative surveillance and future clinical trial design.

A nomogram to predict overall survival of patients with early stage non-small cell lung cancer

Background

Nomograms have been widely used for estimating cancer prognosis. The aim of this study was to construct a clinical nomogram that would well predict overall survival of early stage non-small cell lung cancer (NSCLC) patients after surgery resection.

Methods

A total of 443 patients diagnosed with pathologic stage I and II NSCLC who had undergone curative resection without neoadjuvant chemotherapy or radiotherapy were recruited and analyzed. The log-rank test and multivariate Cox regression analysis were used to select the most significant predictors in the final nomogram for predicting overall survival. Furthermore, the model was validated by bootstrap methods and measured by concordance index (C-index) and calibration plots.

Results

Four independent predictors for overall survival were identified and included into the delineation of the nomogram (tumor differentiation, station of sampled lymph nodes, pathologic T and pathologic N). The model showed comparatively stable discrimination (bootstrap-corrected C-index =0.622, 95% CI: 0.572-0.672) and good calibration.

Conclusions

We successfully developed a nomogram incorporating available clinicopathological variables to predict overall survival of early stage NSCLC patients after surgery resection, which might help clinician select better appropriate treatment decisions.

Are unsatisfactory outcomes after concurrent chemoradiotherapy for locally advanced non-small cell lung cancer due to treatment-related immunosuppression?

BACKGROUND AND PURPOSE

We test the hypothesis that unsatisfactory outcomes after concurrent chemoradiotherapy (RT) for locally advanced non-small cell lung cancer (LA-NSCLC) are due to treatment-related immunosuppression.

MATERIALS AND METHODS

White blood cells (WBCs) data were retrospectively collected for all stage IIIA/B LA-NSCLC patients before and after (after RT: two weeks, two months, four months) concurrent chemotherapy and intensity-modulated RT in which patients were treated to a median of 63 Gy (1.8-2.0 Gy/fractions) in 2004-2014 (N = 155). Nine WBC variables were generated from pre-RT normalized absolute number of lymphocytes and neutrophils (L, N) and the N/L thereof. A WBC variable was considered a predictor for overall survival and recurrence (distant/local/nodal/regional) if p ≤ 0.006 (corrected for 9 variables) from Cox regression and competing risk analyses, respectively; both conducted using bootstrap resampling. Finally, a WBC variable predicting any of the outcomes was linearly associated with each of eleven disease/patient/treatment characteristics (p ≤ 0.005; corrected for 11 characteristics).

RESULTS

At the three post-RT time points both L and N significantly decreased (p < 0.0003). Overall survival was associated with N and N/L four months post-RT (p = 0.00001, 0.0003); regional recurrence was associated with L two months post-RT (p < 0.0001). None of the disease/patient/treatment characteristics was significantly associated with any of the three WBC variables that predicted OS or recurrence (lowest p-value: p = 0.006 for tumour stage,).

CONCLUSION

Significantly lower WBC levels after concurrent chemo-RT for LA-NSCLC are associated with worse long-term outcomes. The mechanism behind this treatment-related immunosuppression requires further analysis likely including other characteristics as no statistically significant association was established between any WBC variable and the disease/patient/treatment characteristics.

Tumor regression during radiotherapy for non-small cell lung cancer patients using cone-beam computed tomography images

PURPOSE

Previous literature has reported contradicting results regarding the relationship between tumor volume changes during radiotherapy treatment for non-small cell lung cancer (NSCLC) patients and locoregional recurrence-free rate or overall survival. The aim of this study is to validate the results from a previous study by using a different volume extraction procedure and evaluating an external validation dataset.

METHODS

For two datasets of 94 and 141 NSCLC patients, gross tumor volumes were determined manually to investigate the relationship between tumor volume regression and locoregional control using Kaplan-Meier curves. For both datasets, different subgroups of patients based on histology and chemotherapy regimens were also investigated. For the first dataset (n = 94), automatically determined tumor volumes were available from a previously published study to further compare their correlation with updated clinical data.

RESULTS

A total of 70 out of 94 patients were classified into the same group as in the previous publication, splitting the dataset based on median tumor regression calculated by the two volume extraction methods. Non-adenocarcinoma patients receiving concurrent chemotherapy with large tumor regression show reduced locoregional recurrence-free rates in both datasets (p < 0.05 in dataset 2). For dataset 2, the opposite behavior is observed for patients not receiving chemotherapy, which was significant for overall survival (p = 0.01) but non-significant for locoregional recurrence-free rate (p = 0.13).

CONCLUSION

The tumor regression pattern observed during radiotherapy is not only influenced by irradiation but depends largely on the delivered chemotherapy schedule, so it follows that the relationship between patient outcome and the degree of tumor regression is also largely determined by the chemotherapy schedule. This analysis shows that the relationship between tumor regression and outcome is complex, and indicates factors that could explain previously reported contradicting findings. This, in turn, will help guide future studies to fully understand the relationship between tumor regression and outcome.

Dose to the cardio-pulmonary system and treatment-induced electrocardiogram abnormalities in locally advanced non-small cell lung cancer

•

ECG abnormalities after chemo-RT in LA-NSCLC are common (35%–67%).

•

Non-specific ECG abnormalities are associated with a high superior vena cava dose.

•

Reducing cardiopulmonary dose is likely to lead to less radiation-induced cardiac toxicity.

Introduction

High dose radiotherapy (RT) has been associated with unexpectedly short survival times for locally advanced Non-Small Cell Lung Cancer (LA-NSCLC) patients. Here we tested the hypothesis that cardiac substructure dose is associated with electrocardiography (ECG) assessed abnormalities after RT for LA-NSCLC.

Materials and methods

Pre- and post-RT ECGs were analyzed for 155 LA-NSCLC patients treated to a median of 64 Gy in 1.8–2.0 Gy fractions using intensity-modulated RT plus chemotherapy (concurrent/sequential: 64%/36%) between 2004 and 2014. ECG abnormalities were classified as new Arrhythmic, Ischemic/Pericardial, or Non-specific (A_ΔECG, I/P_ΔECG, or NS_ΔECG) events. Abnormalities were modeled as time to ECG events considering death a competing risk, and the variables considered for analysis were fractionation-corrected dose-volume metrics (α/β = 3 Gy) of ten cardio-pulmonary structures (aorta, heart, heart chambers, inferior and superior vena cava, lung, pulmonary artery) and 15 disease, patient and treatment characteristics. Each abnormality was modelled using bootstrapping and a candidate predictor was suggested by a median multiple testing-adjusted p-value ≤0.05 across the 1000 generated samples. Forward-stepwise multivariate analysis was conducted in case of more than one candidate.

Results

At a median of eight months post-RT, the rate of A_ΔECG, I/P_ΔECG, and NS_ΔECG was 66%, 35%, and 67%. Both A_ΔECG and I/P_ΔECG were associated with worse performance status (p = 0.007, 0.03), while a higher superior vena cava minimum dose was associated with NS_ΔECG (p = 0.002).

Conclusion

This study suggests that higher radiation doses to the cardio-pulmonary system lead to non-specific ECG abnormalities. Reducing dose to this system, along with effective tumor control, is likely to decrease radiation-induced cardiac toxicity.

Is tumour sphericity an important prognostic factor in patients with lung cancer?

BACKGROUND AND PURPOSE

Quantitative tumour shape features extracted from radiotherapy planning scans have shown potential as prognostic markers. In this study, we investigated if sphericity of the gross tumour volume (GTV) on planning computed tomography (CT) is an independent predictor of overall survival (OS) in lung cancer patients treated with standard radiotherapy. In the analysis, we considered whether tumour sphericity is correlated with clinical prognostic factors or influenced by the inclusion of lymph nodes in the GTV.

MATERIALS AND METHODS

Sphericity of single GTV delineation was extracted for 457 lung cancer patients. Relationships between sphericity, and common patient and tumour characteristics were investigated via correlation analysis and multivariate Cox regression to assess prognostic value of GTV sphericity. A subset analysis was performed for 290 nodal stage N0 patients to determine prognostic value of primary tumour sphericity.

RESULTS

Sphericity is correlated with clinical variables: tumour volume, mean lung dose, N stage, and T stage. Sphericity is strongly associated with OS (p < 0.001, hazard ratio (HR) (95% confidence interval (CI)) = 0.13 (0.04-0.41)) in univariate analysis. However, this association did not remain significant in multivariate analysis (p = 0.826, HR (95% CI) = 0.83 (0.16-4.31), and inclusion of sphericity to a clinical model did not improve model performance. In addition, no significant relationship between sphericity and OS was detected in univariate (p = 0.072) or multivariate (p = 0.920) analysis of N0 subset.

CONCLUSION

Sphericity correlates clearly with clinical prognostic factors, which are often unaccounted for in radiomic studies. Sphericity is also influenced by the presence of nodal involvement within the GTV contour.

Development and validation of nomograms for predicting survival probability of patients with advanced adenocarcinoma in different EGFR mutation status

INTRODUCTION

Molecular markers are important variables in the selection of treatment for cancer patients and highly associated with their survival. Therefore, a nomogram that can predict survival probability by incorporating epidermal growth factor receptor mutation status and treatments for patients with advanced adenocarcinoma would be highly valuable. The aim of the study is to develop and validate a novel nomogram, incorporating epidermal growth factor receptor mutation status and treatments, for predicting 1-year and 2-year survival probability of patients with advanced adenocarcinoma.

MATERIAL AND METHODS

Data on 13,043 patients between June 1, 2011, and December 31, 2014 were collected. Seventy percent of them were randomly assigned to the training cohort for nomogram development, and the remaining 30% assigned to the validation cohort. The most important factors for constructing the nomogram were identified using multivariable Cox regression analysis. The discriminative ability and calibration of the nomograms were tested using C-statistics, calibration plots, and Kaplan-Meier curves.

RESULTS

In the training cohort, 1-year and 2-year OS were 52.8% and 28.5% in EGFR(-) patients, and 73.9% and 44.1% in EGFR(+) patients, respectively. In EGFR(+) group, factors selected were age, gender, congestive heart failure, renal disease, number of lymph node examined, tumor stage, surgical intervention, radiotherapy, first-line chemotherapy, ECOG performance status, malignant pleural effusion, and smoking. In EGFR(-) group, factors selected were age, gender, myocardial infarction, cerebrovascular disease, chronic pulmonary disease, number of lymph node examined, tumor stage, surgical intervention, radiotherapy, ECOG performance status, malignant pleural effusion, and a history of smoking. Two nomograms show good accuracy in predicting OS, with a concordance index of 0.83 in EGFR(+) and of 0.88 in EGFR(-).

CONCLUSIONS

The survival prediction models can be used to make individualized predictions with different EGFR mutation status and a useful tool for selecting regimens for treating advanced adenocarcinoma.