Lung cancer prognostic index: a risk score to predict overall survival after the diagnosis of non-small-cell lung cancer

Real-World Treatment Patterns and Clinical Outcomes Among Patients with Metastatic or Unresectable EGFR-Mutated Non-Small Cell Lung Cancer Previously Treated with Osimertinib and Platinum-Based Chemotherapy

INTRODUCTION

For patients with epidermal growth factor receptor-mutated (EGFRm) locally advanced/metastatic non-small cell lung cancer (mNSCLC) whose disease has progressed on or after osimertinib and platinum-based chemotherapy (PBC), no uniformly accepted standard of care exists. Moreover, limited efficacy of standard treatments indicates an unmet medical need, which is being addressed by ongoing clinical investigations, including the HERTHENA-Lung01 (NCT04619004) study of patritumab deruxtecan (HER3‑DXd). However, because limited information is available on real-world clinical outcomes in such patients, early-phase trials of investigational therapies lack sufficient context for comparison. This study describes the real-world clinical characteristics, treatments, and outcomes for patients with EGFRm mNSCLC who initiated a new line of therapy following previous osimertinib and PBC, including a subset matched to the HERTHENA-Lung01 population.

METHODS

This retrospective analysis used a US database derived from deidentified electronic health records. The reference cohort included patients with EGFRm mNSCLC who had initiated a new line of therapy between November 13, 2015 and June 30, 2021, following prior osimertinib and PBC. A subset of patients resembling the HERTHENA-Lung01 population was then extracted from the reference cohort; this matched subset was optimized using propensity score (PS) weighting. Endpoints were real-world overall survival (rwOS) and real-world progression-free survival (rwPFS). Confirmed real-world objective response rate (rwORR; partial/complete response confirmed ≥ 28 days later) was calculated for the response-evaluable subgroups of patients (with ≥ 2 response assessments spaced ≥ 28 days apart).

RESULTS

In the reference cohort (N = 273), multiple treatment regimens were used, and none was predominant. Median rwPFS and rwOS were 3.3 and 8.6 months, respectively; confirmed rwORR (response evaluable, n = 123) was 13.0%. In the matched subset (n = 126), after PS weighting, median rwPFS and rwOS were 4.2 and 9.1 months, respectively; confirmed rwORR (response evaluable, n = 57) was 14.1%.

CONCLUSION

The treatment landscape for this heavily pretreated population of patients with EGFRm mNSCLC is fragmented, with no uniformly accepted standard of care. A high unmet need exists for therapeutic options that provide meaningful improvements in clinical benefit.

Real-world overview of therapeutic strategies and prognosis of older patients with advanced or metastatic non-small cell lung cancer from the ESME database

INTRODUCTION

In France, 40% of patients diagnosed with lung cancer are ≥70 years old, but these are under-represented in clinical trials. Using data from the French Epidemiological Strategy and Medical Economics (ESME) platform on Lung Cancer (LC), the objective is to provide an overview of the management and the prognosis of older patients with advanced or metastatic non-small cell lung cancer (AM-NSCLC) in a real-world context.

MATERIALS AND METHODS

From the ESME-LC database, we selected patients with AM-NSCLC (stage IIIB, IIIC, and IV), diagnosed between 2015 and 2019, and who received first-line systemic treatment. Demographics, tumour characteristics, and treatment received were described in patients ≥70, and compared to younger ones. Real-world progression-free survival (rwPFS) and overall survival (OS) were evaluated using the multivariable Cox model.

RESULTS

Among 10,002 patients with AM-NSCLC, the median age was 64 years, with 2,754 (27.5%) aged ≥70. In comparison with patients <70, older patients were more often male, with worse performance status and more comorbidities, but they were less underweight and more often non-smokers. The proportion of EGFR mutated non-squamous NSCLC was higher in older patients (25.0% vs 12.8%, p < 0.001), particularly among smokers and former smokers (12.7% vs 7.3%, p < 0.001). Among patients ≥70, 76.6% received first-line chemotherapy (including 67.0% treated with a platinum-based doublet), 15.0% received only targeted therapy, and 11.0% received immunotherapy (alone or in combination). Median first-line rwPFS was 5.1 months (95% confidence interval [CI] = [4.8;5.4]) for patients ≥70 and 4.6 months (95%CI = [4.4;4.8]) for patients <70, but age was not associated with rwPFS in multivariable analysis. Median OS was 14.8 months (95%CI = [13.9;16.1]) for patients ≥70 and 16.7 months (95%CI = [15.9;17.5]) for patients <70, with a significant effect of age in multivariable analysis for patients treated with chemotherapy and/or with targeted therapy, but not for patients treated with immunotherapy (alone or in combination with chemotherapy).

DISCUSSION

In this real-world cohort of patients with AM-NSCLC, age was not associated with first-line rwPFS regardless of treatment received, nor with OS for patients receiving immunotherapy. However, OS was significantly shorter for patients aged ≥70 treated with chemotherapy or with targeted therapy alone.

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.

The impact of pre-treatment smoking status on survival after chemoradiotherapy for locally advanced non-small-cell lung cancer

INTRODUCTION

Smoking is a risk factor for the development of lung cancer and reduces life expectancy within the general population. Retrospective studies suggest that non-smokers have better outcomes after treatment for lung cancer. We used a prospective database to investigate relationships between pre-treatment smoking status and survival for a cohort of patients with stage III non-small-cell lung cancer (NSCLC) treated with curative-intent concurrent chemoradiotherapy (CRT).

METHODS

All patients treated with CRT for stage III NSCLC at a major metropolitan cancer centre were prospectively registered to a database. A detailed smoking history was routinely obtained at baseline. Kaplan-Meier statistics were used to assess overall survival and progression-free survival in never versus former versus current smokers.

RESULTS

Median overall survival for 265 eligible patients was 2.21 years (95 % Confidence Interval 1.78, 2.84). It was 5.5 years (95 % CI 2.1, not reached) for 25 never-smokers versus 1.9 years (95 % CI 1.5, 2.7) for 182 former smokers and 2.2 years (95 % CI 1.3, 2.7) for 58 current smokers. Hazard ratio for death was 2.43 (95 % CI 1.32-4.50) for former smokers and 2.75 (95 % CI 1.40, 5.40) for current smokers, p = 0.006. Actionable tumour mutations (EGFR, ALK, ROS1) were present in more never smokers (14/25) than former (9/182) or current (3/58) smokers. TKI use was also higher in never smokers but this was not significantly associated with superior survival (Hazard ratio 0.71, 95 % CI 0.41, 1.26).

CONCLUSIONS

Never smokers have substantially better overall survival than former or current smokers after undergoing CRT for NSCLC.

The Clinical Frailty Scale for Risk Stratification in Patients With Fibrotic Interstitial Lung Disease

BACKGROUND

Previous studies have shown the importance of frailty in patients with fibrotic interstitial lung disease (ILD).

RESEARCH QUESTION

Is the Clinical Frailty Scale (CFS) a valid tool to improve risk stratification in patients with fibrotic ILD?

STUDY DESIGN AND METHODS

Patients with fibrotic ILD were included from the prospective multicenter Canadian Registry for Pulmonary Fibrosis. The CFS was assessed using available information from initial ILD clinic visits. Patients were stratified into fit (CFS score 1-3), vulnerable (CFS score 4), and frail (CFS score 5-9) subgroups. Cox proportional hazards and logistic regression models with mixed effects were used to estimate time to death or lung transplantation. A derivation and validation cohort was used to establish prognostic performance. Trajectories of functional tests were compared using joint models.

RESULTS

Of the 1,587 patients with fibrotic ILD, 858 (54%) were fit, 400 (25%) were vulnerable, and 329 (21%) were frail. Frailty was a risk factor for early mortality (hazard ratio, 5.58; 95% CI, 3.64-5.76, P < .001) in the entire cohort, in individual ILD diagnoses, and after adjustment for potential confounders. Adding frailty to established risk prediction parameters improved the prognostic performance in derivation and validation cohorts. Patients in the frail subgroup had larger annual declines in FVC % predicted than patients in the fit subgroup (-2.32; 95% CI, -3.39 to -1.17 vs -1.55; 95% CI, -2.04 to -1.15, respectively; P = .02).

INTERPRETATION

The simple and practical CFS is associated with pulmonary and physical function decline in patients with fibrotic ILD and provides additional prognostic accuracy in clinical practice.

Molecular alterations and clinical prognostic factors in resectable non-small cell lung cancer

BACKGROUND

EGFR inhibitor and immunotherapy have been approved for adjuvant treatment in resectable non-small cell lung cancer (NSCLC). Limited reports of molecular and clinical characteristics as prognostic factors in NSCLC have been published.

METHODS

Medical records of patients with resectable NSCLC stage I-III diagnosed during 2015-2020 were reviewed. Real time-PCR (RT-PCR) was performed for EGFR mutations (EGFRm). Immunohistochemistry staining was conducted for ALK and PD-L1 expression. Categorical variables were compared using chi-square test and Fisher's exact test. Survival analysis was done by cox-regression method.

RESULTS

Total 441 patients were included. The prevalence of EGFRm, ALK fusion, and PD-L1 expression were 57.8%, 1.9%, and 20.5% (SP263), respectively. The most common EGFRm were Del19 (43%) and L858R (41%). There was no significant difference of recurrence free survival (RFS) by EGFRm status whereas patients with PD-L1 expression (PD-L1 positive patients) had lower RFS compared to without PD-L1 expression (PD-L1 negative patients) (HR = 1.75, P = 0.036). Patients with both EGFRm and PD-L1 expression had worse RFS compared with EGFRm and PD-L1 negative patients (HR = 3.38, P = 0.001). Multivariable analysis showed higher CEA at cut-off 3.8 ng/ml, pT4, pN2, pStage II, and margin were significant poor prognostic factors for RFS in the overall population, which was similar to EGFRm population (exception of pT and pStage). Only pStage was a significant poor prognostic factor for PD-L1 positive patients. The predictive score for predicting of recurrence were 6 for all population (63% sensitivity and 86% specificity) and 5 for EGFRm population (62% sensitivity and 93% specificity).

CONCLUSION

The prevalence and types of EGFRm were similar between early stage and advanced stage NSCLC. While lower prevalence of PD-L1 expression was found in early stage disease. Patients with both EGFRm and PD-L1 expression had poorer outcome. Thus PD-L1 expression would be one of the prognostic factor in EGFRm patients. Validation of the predictive score should be performed in a larger cohort.

The impact of ECOG performance status on efficacy of immunotherapy and immune-based combinations in cancer patients: the MOUSEION-06 study

ECOG performance status (PS) is a pivotal prognostic factor in a wide number of solid tumors. We performed a meta-analysis to assess the role of ECOG PS in terms of survival in patients with ECOG PS 0 or ECOG PS 1 treated with immunotherapy alone or combined with other anticancer treatments. Following the Preferred Reporting Items for Systematic Reviews and Meta-analyses, all phase II and III randomized clinical trials that compared immunotherapy or immune-based combinations in patients with solid tumors were retrieved. The outcomes of interest were overall survival (OS) and progression-free survival (PFS). We also performed subgroup analyses focused on type of therapy (ICI monotherapy or combinations), primary tumor type, setting (first line of treatment, subsequent lines). Overall, 60 studies were included in the analysis for a total of 35.020 patients. The pooled results showed that immunotherapy, either alone or in combination, reduces the risk of death or progression in both ECOG PS 0 and 1 populations. The survival benefit was consistent in all subgroups. Immune checkpoint inhibitors monotherapy or immune-based combinations are associated with improved survival irrespective of ECOG PS 0 or 1. Clinical trials should include more frail patients to assess the value of immunotherapy in these patients.

Detection method has independent prognostic significance in the PLCO lung screening trial

Prognostic models in cancer use patient demographic and tumor characteristics to predict survival and dynamic disease prognosis. Past work in breast cancer has shown that cancer detection method, screen-detected or symptom-detected, has prognostic significance. We investigate this phenomenon in the lung component of the Prostate, Lung, Colorectal, and Ovarian (PLCO) screening trial. Patients were randomized to intervention, receiving four annual chest x-rays (CXRs), or to control, receiving usual care. Patients were followed for a total of approximately 13 years. In PLCO, lung cancer detection method has independent prognostic value exceeding that of variables commonly used in lung cancer prognostic models, including sex, histology, and age. Results are robust to cohort selection and type of predictive model. These results imply that detection method should be considered when developing prognostic models in lung cancer studies, and cancer registries should routinely collect cancer detection method.

Survival with lung cancer in Finland has not improved during 2007-2019-a single center retrospective population-based real-world study

OBJECTIVES

According to the CONCORD-3 study, the 5-year survival rate of lung cancer patients in Finland has not improved during the twenty-first century. In the present study, we evaluated the survival trends of lung cancer patients diagnosed and treated in one of the five university hospitals in Finland to determine possible explanatory factors behind the lack of improved survival.

MATERIAL AND METHODS

This retrospective population-based study included all lung cancer patients diagnosed in Tampere University Hospital in 2007-2019 (N = 3041). The study population was divided into two subcohorts: the patients diagnosed in 2007-2012 and those diagnosed in 2013-2019. The two subcohorts were then compared to analyze the temporal changes in survival and the distribution of prognostic factors.

RESULTS

A comparison of the patients diagnosed in 2007-2012 and 2013-2019 showed that the patients' overall survival had remained unchanged. The median overall survival was 8.7 months in the earlier subcohort and 9.2 months in the later subcohort. The respective 5-year survival rates were 16.6% and 17.8%, and these differences were not statistically significant. The proportion of stage IV patients (approximately 59% in both subcohorts) and their risk of death were similar for the two subcohorts. According to the regression analysis, male gender, advanced stage, and poor Eastern Cooperative Oncology Group performance status were independent risk factors for death, while a never-smoking status and mutation-positive disease were associated with a decreased risk of death, but only in the later cohort.

CONCLUSION

Echoing the results of CONCORD-3, this study confirmed that the real-world survival of unselected lung cancer populations in Finland has not improved over the last 15 years, mainly because of the unchanged proportions of patients with late-stage lung cancer. This calls for earlier recognition of lung cancer, achieved by screening and increasing awareness of the disease.

The prognostic value of histopathology in invasive lung adenocarcinoma: a comparative review of the main proposed grading systems

INTRODUCTION

An accurate histological evaluation of invasive lung adenocarcinoma is essential for a correct clinical and pathological definition of the tumour. Different grading systems have been proposed to predict the prognosis of invasive lung adenocarcinoma.

AREAS COVERED

Invasive non mucinous lung adenocarcinoma is often morphologically heterogeneous, consisting of complex combinations of architectural patterns with different proportions. Several grading systems for non-mucinous lung adenocarcinoma have been proposed, being the main based on architectural differentiation and the predominant growth pattern. Herein we perform a thorough review of the literature using PubMed, Scopus and Web of Science and we highlight the peculiarities and the differences between the main grading systems and compare the data about their prognostic value. In addition, we carried out an evaluation of the proposed grading systems for less common histological variants of lung adenocarcinoma, such as fetal adenocarcinoma and invasive mucinous adenocarcinoma.

EXPERT OPINION

The current IASLC grading system, based on the combined score of predominant growth pattern plus high-grade histological pattern, shows the stronger prognostic significance than the previous grading systems in invasive non mucinous lung adenocarcinoma.

Prognostic significance of location index in resected T1-sized early-stage non-small cell lung cancer

BACKGROUND

While the central location is a known adverse prognostic factor in lung cancer, a precise definition of central lung cancer has not yet emerged.

PURPOSE

To determine the prognostic significance of central lung cancer (defined by location index) in resected T1-sized early-stage non-small cell lung cancer (NSCLC).

MATERIAL AND METHODS

Patients with resected T1-sized early-stage NSCLC between 2010 and 2015 at a single tertiary cancer center were retrospectively reviewed. Central lung cancer was defined by a location index of the second tertile or less. Kaplan-Meier analysis with log-rank test and multivariable Cox regression analysis were performed to analyze the relationship between central lung cancer and the prognosis of relapse-free survival (RFS) and overall survival (OS). Inter-observer agreement was assessed using Cohen's kappa value and intraclass correlation coefficient (ICC).

RESULTS

Overall, 289 patients (169 men; median age 65 years; interquartile range 58-70 years) were evaluated. Central lung cancer (defined by location index) was adversely associated with RFS (P = 0.005) and OS (P = 0.01). Multivariable Cox regression analysis showed that central lung cancer was independently associated with poor RFS (adjusted hazard ratio 1.91; 95% confidence interval [CI] 1.12-3.24; P = 0.017) and OS (adjusted hazard ratio 1.69; 95% CI 1.04-2.74; P = 0.033). Location index demonstrated excellent inter-observer agreement (Cohen's kappa value 0.88; 95% CI 0.82-0.93) with a high ICC (0.98; 95% CI 0.97-0.98).

CONCLUSION

Central lung cancer defined by a location index of the second tertile or lower is an independent adverse prognostic factor in resected T1-sized early-stage NSCLC.

Machine learning models for 180-day mortality prediction of patients with advanced cancer using patient-reported symptom data

PURPOSE

The objective of the current study was to develop and test the performances of different ML algorithms which were trained using patient-reported symptom severity data to predict mortality within 180 days for patients with advanced cancer.

METHODS

We randomly selected 630 of 689 patients with advanced cancer at our institution who completed symptom PRO measures as part of routine care between 2009 and 2020. Using clinical, demographic, and PRO data, we trained and tested four ML algorithms: generalized regression with elastic net regularization (GLM), extreme gradient boosting (XGBoost) trees, support vector machines (SVM), and a single hidden layer neural network (NNET). We assessed the performance of algorithms individually as well as part of an unweighted voting ensemble on the hold-out testing sample. Performance was assessed using area under the receiver-operating characteristic curve (AUROC), sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV).

RESULTS

The starting cohort of 630 patients was randomly partitioned into training (n = 504) and testing (n = 126) samples. Of the four ML models, the XGBoost algorithm demonstrated the best performance for 180-day mortality prediction in testing data (AUROC = 0.69, sensitivity = 0.68, specificity = 0.62, PPV = 0.66, NPV = 0.64). Ensemble of all algorithms performed worst (AUROC = 0.65, sensitivity = 0.65, specificity = 0.62, PPV = 0.65, NPV = 0.62). Of individual PRO symptoms, shortness of breath emerged as the variable of highest impact on the XGBoost 180-mortality prediction (1-AUROC = 0.30).

CONCLUSION

Our findings support ML models driven by patient-reported symptom severity as accurate predictors of short-term mortality in patients with advanced cancer, highlighting the opportunity to integrate these models prospectively into future studies of goal-concordant care.

Construction of IL-1 signalling pathway correlation model in lung adenocarcinoma and association with immune microenvironment prognosis and immunotherapy: Multi-data validation

Numerous studies have confirmed the inextricable link between inflammation and malignancy, which is also involved in developing lung adenocarcinoma, where IL-1 signalling is crucial. However, the predictive role of single gene biomarkers is insufficient, and more accurate prognostic models are needed. We downloaded data related to lung adenocarcinoma patients from the GDC, GEO, TISCH2 and TCGA databases for data analysis, model construction and differential gene expression analysis. The genes of IL-1 signalling-related factors were screened from published papers for subgroup typing and predictive correlation analysis. Five prognostic genes associated with IL-1 signalling were finally identified to construct prognostic prediction models. The K-M curves indicated that the prognostic models had significant predictive efficacy. Further immune infiltration scores showed that IL-1 signalling was mainly associated with enhanced immune cells, drug sensitivity of model genes was analysed using the GDSC database, and correlation of critical memories with cell subpopulation components was observed using single-cell analysis. In conclusion, we propose a predictive model based on IL-1 signalling-related factors, a non-invasive predictive approach for genomic characterisation, in predicting patients' survival outcomes. The therapeutic response has shown satisfactory and effective performance. More interdisciplinary areas combining medicine and electronics will be explored in the future.

Deep learning model integrating positron emission tomography and clinical data for prognosis prediction in non-small cell lung cancer patients

BACKGROUND

Lung cancer is the leading cause of cancer-related deaths worldwide. The majority of lung cancers are non-small cell lung cancer (NSCLC), accounting for approximately 85% of all lung cancer types. The Cox proportional hazards model (CPH), which is the standard method for survival analysis, has several limitations. The purpose of our study was to improve survival prediction in patients with NSCLC by incorporating prognostic information from F-18 fluorodeoxyglucose positron emission tomography (FDG PET) images into a traditional survival prediction model using clinical data.

RESULTS

The multimodal deep learning model showed the best performance, with a C-index and mean absolute error of 0.756 and 399 days under a five-fold cross-validation, respectively, followed by ResNet3D for PET (0.749 and 405 days) and CPH for clinical data (0.747 and 583 days).

CONCLUSION

The proposed deep learning-based integrative model combining the two modalities improved the survival prediction in patients with NSCLC.

The Effects of GCSF Primary Prophylaxis on Survival Outcomes and Toxicity in Patients with Advanced Non-Small Cell Lung Cancer on First-Line Chemoimmunotherapy: A Sub-Analysis of the Spinnaker Study

GCSF prophylaxis is recommended in patients on chemotherapy with a >20% risk of febrile neutropenia and is to be considered if there is an intermediate risk of 10−20%. GCSF has been suggested as a possible adjunct to immunotherapy due to increased peripheral neutrophil recruitment and PD-L1 expression on neutrophils with GCSF use and greater tumour volume decrease with higher tumour GCSF expression. However, its potential to increase neutrophil counts and, thus, NLR values, could subsequently confer poorer prognoses on patients with advanced NSCLC. This analysis follows on from the retrospective multicentre observational cohort Spinnaker study on advanced NSCLC patients. The primary endpoints were OS and PFS. The secondary endpoints were the frequency and severity of AEs and irAEs. Patient information, including GCSF use and NLR values, was collected. A secondary comparison with matched follow-up duration was also undertaken. Three hundred and eight patients were included. Median OS was 13.4 months in patients given GCSF and 12.6 months in those not (p = 0.948). Median PFS was 7.3 months in patients given GCSF and 8.4 months in those not (p = 0.369). A total of 56% of patients receiving GCSF had Grade 1−2 AEs compared to 35% who did not receive GCSF (p = 0.004). Following an assessment with matched follow-up, 41% of patients given GCSF experienced Grade 1−2 irAEs compared to 23% of those not given GCSF (p = 0.023). GCSF prophylaxis use did not significantly affect overall or progression-free survival. Patients given GCSF prophylaxis were more likely to experience Grade 1−2 adverse effects and Grade 1−2 immunotherapy-related adverse effects.

FDG PET and CT radiomics in diagnosis and prognosis of non-small-cell lung cancer

BACKGROUND

18F-FDG PET and CT radiomics has been the object of a wide research for over 20 years but its contribution to clinical practice remains not yet well established. We have investigated its impact versus that of only histo-clinical data, for the routine management of non-small-cell lung cancer (NSCLC).

METHODS

Our patients were retrospectively considered. They all had a FDG PET-CT and immuno-histo-chemistry (IHC) to assess PD-L1 expression at the beginning of the disease. A prognosis univariate and multivariate Cox survival analyses was performed for overall survival (OS) and progression free survival (PFS) prediction, including a training/testing procedure. Two sets of 47 PET and 47 CT radiomics features (RFs) were extracted. Difference between RFs according to PD-L1 expression, the histology status and the stage level were tested using suited non parametric statistical tests and the receiver operating characteristics (ROC) curve and the area under curve (AUC).

RESULTS

From 2017 to 2019, 212 NSCLC patients treated in our institution were included. The main conventional prognostic variables were stage and gender with a low added prognostic value in the models including PET and CT RFs. Neither PET nor CT RFs were significant to separate the different levels of PD-L1 expression. Several RFs differ between adenocarcinoma (ADC) and squamous cell carcinoma (SCC) tumours and a large number of PET and CT RFs are significantly linked to patient stage.

CONCLUSIONS

In our population, PET and CT RFs show their intrinsic power to predict survival but do not significantly improve OS and PFS prediction in the different multivariate models, in comparison to conventional data. It would seem necessary to carry out one's own survival analysis before determining a radiomics signature.

Infrastructure platform for privacy-preserving distributed machine learning development of computer-assisted theragnostics in cancer

INTRODUCTION

Emerging evidence suggests that data-driven support tools have found their way into clinical decision-making in a number of areas, including cancer care. Improving them and widening their scope of availability in various differing clinical scenarios, including for prognostic models derived from retrospective data, requires co-ordinated data sharing between clinical centres, secondary analyses of large multi-institutional clinical trial data, or distributed (federated) learning infrastructures. A systematic approach to utilizing routinely collected data across cancer care clinics remains a significant challenge due to privacy, administrative and political barriers.

METHODS

An information technology infrastructure and web service software was developed and implemented which uses machine learning to construct clinical decision support systems in a privacy-preserving manner across datasets geographically distributed in different hospitals. The infrastructure was deployed in a network of Australian hospitals. A harmonized, international ontology-linked, set of lung cancer databases were built with the routine clinical and imaging data at each centre. The infrastructure was demonstrated with the development of logistic regression models to predict major cardiovascular events following radiation therapy.

RESULTS

The infrastructure implemented forms the basis of the Australian computer-assisted theragnostics (AusCAT) network for radiation oncology data extraction, reporting and distributed learning. Four radiation oncology departments (across seven hospitals) in New South Wales (NSW) participated in this demonstration study. Infrastructure was deployed at each centre and used to develop a model predicting for cardiovascular admission within a year of receiving curative radiotherapy for non-small cell lung cancer. A total of 10417 lung cancer patients were identified with 802 being eligible for the model. Twenty features were chosen for analysis from the clinical record and linked registries. After selection, 8 features were included and a logistic regression model achieved an area under the receiver operating characteristic (AUROC) curve of 0.70 and C-index of 0.65 on out-of-sample data.

CONCLUSION

The infrastructure developed was demonstrated to be usable in practice between clinical centres to harmonize routinely collected oncology data and develop models with federated learning. It provides a promising approach to enable further research studies in radiation oncology using real world clinical data.

Application of quantitative bias analysis for unmeasured confounding in cost-effectiveness modelling

Due to uncertainty regarding the potential impact of unmeasured confounding, health technology assessment (HTA) agencies often disregard evidence from nonrandomized studies when considering new technologies. Quantitative bias analysis (QBA) methods provide a means to quantify this uncertainty but have not been widely used in the HTA setting, particularly in the context of cost-effectiveness modelling (CEM). This study demonstrated the application of an aggregate and patient-level QBA approach to quantify and adjust for unmeasured confounding in a simulated nonrandomized comparison of survival outcomes. Application of the QBA output within a CEM through deterministic and probabilistic sensitivity analyses and under different scenarios of knowledge of an unmeasured confounder demonstrates the potential value of QBA in HTA.

Treatment strategy, overall survival and associated risk factors among patients with unresectable stage IIIB/IV non-small cell lung cancer in China (2015-2017): A multicentre prospective study

BACKGROUND

There are limited studies on treatment and survival analysis among patients with unresectable Stage IIIB or IV non-small cell lung cancer (NSCLC) in routine practice in China. To address this gap, we conducted a prospective observational study in a cohort of patients treated at 11 hospitals in China.

METHODS

This was a multicentre, prospective cohort study including patients with newly diagnosed unresectable Stage IIIB or IV NSCLC from June 26th, 2015 to April 28th, 2017. Patient baseline characteristics, disease characteristics, and anti-cancer treatments were obtained by medical chart review. The overall survival (OS) from the initiation of first-line treatment was analysed by the Kaplan-Meier method. Factors associated with survival were analysed by univariate and multivariate Cox regression models.

FINDINGS

Among 1324 patients enrolled with median follow-up duration of 15·0 (range: 0·0-42·1) months, 83·5% (1105/1324) of them received first-line chemotherapy of which platinum-based compounds were the dominated agents. Overall, 30·9% (409/1324) of patients received targeted therapy as 1st-line treatment including 65·0% (266/409) EGFR-TKIs and 5·1% (21/409) ALK-TKIs. Of all eligible patients, gene testing rates were 44·0% (583/1324) for EGFR mutations, 17·0% (225/1324) for EML4-ALK gene fusions, and 8·3% (110/1324) for ROS1 gene fusions. The EGFR-TKIs were administered to 63·9% (179/280) of EGFR mutated patients as first-line treatment. The overall median OS was 23·2 (95%CI 19·5-25·5) months, and patients treated at tier 1 cities had better OS than that of tier 2 cities. Also, the OS in patients with EGFR mutation was longer than those with EGFR wild type. Multivariate Cox regression models suggested that male, education below high school, tier 2 cities, smoking history, and multiple metastases were associated with poor survival.

INTERPRETATION

The gene test coverage was relatively low among the studied population, and over half of EGFR mutated patients received EGFR-TKIs, suggesting that the result of genetic tests in real-world settings may not always indicate the selection of treatment. The OS benefit observed from patients treated in tier 1 cities and those with EGFR mutation may indicate a need for broader gene test coverage, providing NSCLC patients with personalized treatment according to the results of genetic tests.

FUNDING

Roche Holding AG.TRANSLATED ABSTRACT: This translation in Chinese was submitted by the authors and we reproduce it as supplied. It has not been peer reviewed. Our editorial processes have only been applied to the original abstract in English, which should serve as reference for this manuscript.:IIIBIV(NSCLC)., ,, 11.:,, 20156262017428IIIBIVNSCLC.,.Kaplan-Meier(OS), Cox.:1324, 15.0(:0.0-42.1), 83.5%(1105/1324), ., 30.9%(409/1324), 65.0%(266/409)EGFR-TKI5.1%(21/409)ALK-TKI., EGFR,EML4-ALKROS144.0%(583/1324),17.0%(225/1324)8.3%(110/1324).63.9%(179/280)EGFREGFR-TKI.23.2 (95% 19·5-25·5) , ., EGFREGFR.Cox, ,,,.:, EGFREGFR-TKI, , .EGFR, , NSCLC.

Pan-cancer prognostic genetic mutations and clinicopathological factors associated with survival outcomes: a systematic review

Cancer is a leading cause of death, accounting for almost 10 million deaths annually worldwide. Personalised therapies harnessing genetic and clinical information may improve survival outcomes and reduce the side effects of treatments. The aim of this study is to appraise published evidence on clinicopathological factors and genetic mutations (single nucleotide polymorphisms [SNPs]) associated with prognosis across 11 cancer types: lung, colorectal, breast, prostate, melanoma, renal, glioma, bladder, leukaemia, endometrial, ovarian. A systematic literature search of PubMed/MEDLINE and Europe PMC was conducted from database inception to July 1, 2021. 2497 publications from PubMed/MEDLINE and 288 preprints from Europe PMC were included. Subsequent reference and citation search was conducted and a further 39 articles added. 2824 articles were reviewed by title/abstract and 247 articles were selected for systematic review. Majority of the articles were retrospective cohort studies focusing on one cancer type, 8 articles were on pan-cancer level and 6 articles were reviews. Studies analysing clinicopathological factors included 908,567 patients and identified 238 factors, including age, gender, stage, grade, size, site, subtype, invasion, lymph nodes. Genetic studies included 210,802 patients and identified 440 gene mutations associated with cancer survival, including genes TP53, BRCA1, BRCA2, BRAF, KRAS, BIRC5. We generated a comprehensive knowledge base of biomarkers that can be used to tailor treatment according to patients' unique genetic and clinical characteristics. Our pan-cancer investigation uncovers the biomarker landscape and their combined influence that may help guide health practitioners and researchers across the continuum of cancer care from drug development to long-term survivorship.

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.

Clinical Implications of Circulating Circular RNAs in Lung Cancer

Circular RNAs (circRNAs) are single-stranded RNAs with a covalently closed-loop structure that increases their stability; thus, they are more advantageous to use as liquid biopsy markers than linear RNAs. circRNAs are thought to be generated by back-splicing of pre-mRNA transcripts, which can be facilitated by reverse complementary sequences in the flanking introns and trans-acting factors, such as splicing regulatory factors and RNA-binding factors. circRNAs function as miRNA sponges, interact with target proteins, regulate the stability and translatability of other mRNAs, regulate gene expression, and produce microproteins. circRNAs are also found in the body fluids of cancer patients, including plasma, saliva, urine, and cerebrospinal fluid, and these “circulating circRNAs” can be used as cancer biomarkers. In lung cancer, some circulating circRNAs have been reported to regulate cancer progression and drug resistance. Circulating circRNAs have significant diagnostic value and are associated with the prognosis of lung cancer patients. Owing to their functional versatility, heightened stability, and practical applicability, circulating circRNAs represent promising biomarkers for lung cancer diagnosis, prognosis, and treatment monitoring.

Effectiveness of crizotinib versus entrectinib in ROS1-positive non-small-cell lung cancer using clinical and real-world data

Aims: To compare clinical trial results for crizotinib and entrectinib in ROS1-positive non-small-cell lung cancer and compare clinical trial data and real-world outcomes for crizotinib. Patients & methods: We analyzed four phase I-II studies using a simulated treatment comparison (STC). A STC of clinical trial versus real-world evidence compared crizotinib clinical data to real-world outcomes. Results: Adjusted STC found nonsignificant trends favoring crizotinib over entrectinib: objective response rate, risk ratio = 1.04 (95% CI: 0.85-1.28); median duration of response, mean difference = 16.11 months (95% CI: -1.57- 33.69); median progression-free survival, mean difference = 3.99 months (95% CI: -6.27-14.25); 12-month overall survival, risk ratio = 1.01 (95% CI: 0.90-1.12). Nonsignificant differences were observed between the trial end point values and the real-world evidence for crizotinib. Conclusions: Crizotinib and entrectinib have comparable efficacy in ROS1-positive non-small-cell lung cancer.

Dynamic Prediction of Near-Term Overall Survival in Patients with Advanced NSCLC Based on Real-World Data

Simple Summary

Patients near the end of life often receive aggressive care, which may be of low value. For patients with advanced cancers, it is standard clinical practice to estimate the prognosis to inform treatment decisions and improve end-of-life care. However, clinical estimates of prognosis may be imprecise and rapidly become out-of-date if clinical factors that evolve over time are not incorporated. Patient prognosis is commonly estimated based on a clinician’s subjective assessment of patient reserve, such as performance status. We propose a spline-smoothed landmarking approach to dynamically estimate survival probabilities based on objective, evolving patient features. The proposed method allows predictions at any time during the patient disease course and demonstrates dramatically improved prediction accuracy compared to methods using clinical features at a fixed time. The proposed approaches can assist clinicians and patients in appropriately regulating treatments to improve outcomes and quality of life.

Abstract

Patients with terminal cancers commonly receive aggressive and sub-optimal treatment near the end of life, which may not be beneficial in terms of duration or quality of life. To improve end-of-life care, it is essential to develop methods that can accurately predict the short-term risk of death. However, most prediction models for patients with cancer are static in the sense that they only use patient features at a fixed time. We proposed a dynamic prediction model (DPM) that can incorporate time-dependent predictors. We apply this method to patients with advanced non-small-cell lung cancer from a real-world database. Inverse probability of censoring weighted AUC with bootstrap inference was used to compare predictions among models. We found that increasing ECOG performance status and decreasing albumin had negative prognostic associations with overall survival (OS). Moreover, the negative prognostic implications strengthened over the patient disease course. DPMs using both time-independent and time-dependent predictors substantially improved short-term prediction accuracy compared to Cox models using only predictors at a fixed time. The proposed model can be broadly applied for prediction based on longitudinal data, including an estimation of the dynamic effects of time-dependent features on OS and updating predictions at any follow-up time.

LncRNA LINC00460: Function and mechanism in human cancer

Long non-coding RNAs (LncRNAs), which are more than 200 nucleotides in length and with limited protein-coding potential, play vital roles in the pathogenesis, tumorigenesis, and angiogenesis of cancers. Aberrant expression of lncRNAs has been detected in various carcinomas and may be correlated with oncogenesis by affecting related genes expression. Recently, an increasing number of studies have reported on long intergenic non-protein coding RNA 460 (LINC00460) in human tumor fields. LINC00460 is upregulated in diverse cancer tissues and cells. The upregulated expression level of LINC00460 is correlated with larger tumor size, tumor node metastasis (TNM) stage, lymph node metastasis, and shorter overall survival. The regulatory mechanism of LINC00460 was complex and diverse. LINC00460 could act as a competitive endogenous RNA (ceRNA), directly bind with proteins or regulate multiple pathways, which affected tumor progression. Moreover, LINC00460 was also identified to increase drug resistance, and therefore, weaken the effectiveness of tumor treatment. It has become increasingly important to investigate the roles of LINC00460 in various cancers by different mechanisms. Therefore, a more comprehensive understanding of LINC00460 is crucial to expound on the cellular function and molecular mechanism of human cancers. In this review, we refer to studies concerning LINC00460 and provide the basis for the evaluation of LINC00460 as a predicted biomarker or potential therapeutic target in malignancies, and also provide ideas for the future research of lncRNAs similar to LINC00460.

Implementation of the Australian Computer-Assisted Theragnostics (AusCAT) network for radiation oncology data extraction, reporting and distributed learning

INTRODUCTION

There is significant potential to analyse and model routinely collected data for radiotherapy patients to provide evidence to support clinical decisions, particularly where clinical trials evidence is limited or non-existent. However, in practice there are administrative, ethical, technical, logistical and legislative barriers to having coordinated data analysis platforms across radiation oncology centres.

METHODS

A distributed learning network of computer systems is presented, with software tools to extract and report on oncology data and to enable statistical model development. A distributed or federated learning approach keeps data in the local centre, but models are developed from the entire cohort.

RESULTS

The feasibility of this approach is demonstrated across six Australian oncology centres, using routinely collected lung cancer data from oncology information systems. The infrastructure was used to validate and develop machine learning for model-based clinical decision support and for one centre to assess patient eligibility criteria for two major lung cancer radiotherapy clinical trials (RTOG-9410, RTOG-0617). External validation of a 2-year overall survival model for non-small cell lung cancer (NSCLC) gave an AUC of 0.65 and C-index of 0.62 across the network. For one centre, 65% of Stage III NSCLC patients did not meet eligibility criteria for either of the two practice-changing clinical trials, and these patients had poorer survival than eligible patients (10.6 m vs. 15.8 m, P = 0.024).

CONCLUSION

Population-based studies on routine data are possible using a distributed learning approach. This has the potential for decision support models for patients for whom supporting clinical trial evidence is not applicable.

Deep learning model to predict the need for mechanical ventilation using chest X-ray images in hospitalised patients with COVID-19

OBJECTIVES

There exists a wide gap in the availability of mechanical ventilator devices and their acute need in the context of the COVID-19 pandemic. An initial triaging method that accurately identifies the need for mechanical ventilation in hospitalised patients with COVID-19 is needed. We aimed to investigate if a potentially deteriorating clinical course in hospitalised patients with COVID-19 can be detected using all X-ray images taken during hospitalisation.

METHODS

We exploited the well-established DenseNet121 deep learning architecture for this purpose on 663 X-ray images acquired from 528 hospitalised patients with COVID-19. Two Pulmonary and Critical Care experts blindly and independently evaluated the same X-ray images for the purpose of validation.

RESULTS

We found that our deep learning model predicted the need for mechanical ventilation with a high accuracy, sensitivity and specificity (90.06%, 86.34% and 84.38%, respectively). This prediction was done approximately 3 days ahead of the actual intubation event. Our model also outperformed two Pulmonary and Critical Care experts who evaluated the same X-ray images and provided an incremental accuracy of 7.24%-13.25%.

CONCLUSIONS

Our deep learning model accurately predicted the need for mechanical ventilation early during hospitalisation of patients with COVID-19. Until effective preventive or treatment measures become widely available for patients with COVID-19, prognostic stratification as provided by our model is likely to be highly valuable.

Durvalumab after Chemoradiotherapy for PD-L1 Expressing Inoperable Stage III NSCLC Leads to Significant Improvement of Local-Regional Control and Overall Survival in the Real-World Setting

Concurrent chemoradiotherapy (CRT) followed by maintenance treatment with the PD-L1 inhibitor durvalumab is a new standard of care for inoperable stage III NSCLC. The present study compares the oncological outcome of patients treated with CRT to those treated with CRT and durvalumab (CRT-IO) in the real-world setting. The analysis was performed based on the retro- and prospectively collected data of 144 consecutive inoperable stage III NSCLC patients treated between 2011-2020. Local-regional-progression-free-survival (LRPFS-defined as progression in the mediastinum, hilum and/or supraclavicular region at both sites and the involved lung), progression-free survival (PFS), and overall survival (OS) were evaluated from the last day of thoracic radiotherapy (TRT). Median follow-up for the entire cohort was 33.1 months (range: 6.3-111.8) and median overall survival was 27.2 (95% CI: 19.5-34.9) months. In the CRT-IO cohort after a median follow-up of 20.9 (range: 6.3-27.4) months, median PFS was not reached, LRPFS (p = 0.002), PFS (p = 0.018), and OS (p = 0.005) were significantly improved vs. the historical cohort of conventional CRT patients. After propensity-score matching (PSM) analysis with age, gender, histology, tumor volume, and treatment mode, and exact matching for T-and N-stage, 22 CRT-IO patients were matched 1:2 to 44 CRT patients. Twelve-month LRPFS, PFS, and OS rates in the CRT-IO vs. CRT cohort were 78.9 vs. 45.5% (p = 0.002), 60.0 vs. 31.8% (p = 0.007), and 100 vs. 70.5% (p = 0.003), respectively. This real-world analysis demonstrated that durvalumab after CRT led to significant improvement of local-regional control, PFS, and OS in PD-L1 expressing inoperable stage III NSCLC patients compared to a historical cohort.

Real-world prospective analysis of treatment patterns in durvalumab maintenance after chemoradiotherapy in unresectable, locally advanced NSCLC patients

The aim of this prospective study is to evaluate the clinical use and real-world efficacy of durvalumab maintenance treatment after chemoradiotherapy (CRT) in unresectable stage, locally advanced non-small cell lung cancer (NSCLC). All consecutive patients with unresectable, locally advanced NSCLC and PD-L1 expression (≥1%) treated after October 2018 were included. Regular follow up, including physical examination, PET/CT and/or contrast-enhanced CT-Thorax/Abdomen were performed every three months after CRT. Descriptive treatment pattern analyses, including reasons of discontinuation and salvage treatment, were undertaken. Statistics were calculated from the last day of thoracic irradiation (TRT). Twenty-six patients were included. Median follow up achieved 20.6 months (range: 1.9–30.6). Durvalumab was initiated after a median of 25 (range: 13–103) days after completion of CRT. In median 14 (range: 2–24) cycles of durvalumab were applied within 6.4 (range 1–12.7) months. Six patients (23%) are still in treatment and seven (27%) have completed treatment with 24 cycles. Maintenance treatment was discontinued in 13 (50%) patients: 4 (15%) patients developed grade 3 pneumonitis according to CTCAE v5 after a median of 3.9 (range: 0.5–11.6) months and 7 (range: 2–17) cycles of durvalumab. Four (15%) patients developed grade 2 skin toxicity. One (4%) patient has discontinued treatment due to incompliance. Six and 12- month progression-free survival (PFS) rates were 82% and 62%, median PFS was not reached. No case of hyperprogression was documented. Eight (31%) patients have relapsed during maintenance treatment after a median of 4.8 (range: 2.2–11.3) months and 11 (range: 6–17) durvalumab cycles. Two patients (9%) developed a local-regional recurrence after 14 and 17 cycles of durvalumab. Extracranial distant metastases and brain metastases as first site of failure were detected in 4 (15%) and 2 (8%) patients, respectively. Three (13%) patients presented with symptomatic relapse. Our prospective study confirmed a favourable safety profile of durvalumab maintenance treatment after completion of CRT in unresectable stage, locally advanced NSCLC in a real-world setting. In a median follow-up time of 20.6 months, durvalumab was discontinued in 27% of all patients due to progressive disease. All patients with progressive disease were eligible for second-line treatment.

Preclinical Metrics Correlate With Drug Activity in Phase II Trials of Targeted Therapies for Non-Small Cell Lung Cancer

Novel oncology drugs often fail to progress from preclinical experiments to FDA approval. Therefore, determining which preclinical or clinical factors associate with drug activity could accelerate development of effective therapies. We investigated whether preclinical metrics and patient characteristics are associated with objective response rate (ORR) in phase II clinical trials of targeted therapies for non-small cell lung cancer (NSCLC). We developed a reproducible process to select a single phase II trial and supporting preclinical publication for a given drug-indication pair, which we defined as the pairing of a small molecule inhibitor (e.g., crizotinib) with the specific patient population for which it was designed to work (e.g., patients with an ALK aberration). We demonstrated that robust drug activity in mice, as measured by change in tumor size, is independently associated with improved ORR in phase II clinical trials. The number of mice utilized in experiments, the number of publications referencing the drug for NSCLC before the phase II clinical trial, and whether the drug was approved for a cancer other than NSCLC also significantly correlated with ORR. Among clinical characteristics, sex, race, histology, and smoking history were significantly associated with ORR. Further research into metrics that correlate with drug activity has the potential to optimize selection of novel therapies for clinical trials and enrich the drug development pipeline, particularly for patients with targetable genetic aberrations and rare cancers.

Impact of musculoskeletal degradation on cancer outcomes and strategies for management in clinical practice

The prevalence of malnutrition in patients with cancer is one of the highest of all patient groups. Weight loss (WL) is a frequent manifestation of malnutrition in cancer and several large-scale studies have reported that involuntary WL affects 50-80% of patients with cancer, with the degree of WL dependent on tumour site, type and stage of disease. The study of body composition in oncology using computed tomography has unearthed the importance of both low muscle mass (sarcopenia) and low muscle attenuation as important prognostic indications of unfavourable outcomes including poorer tolerance to chemotherapy; significant deterioration in performance status and quality of life (QoL), poorer post-operative outcomes and shortened survival. While often hidden by excess fat and high BMI, muscle abnormalities are highly prevalent in patients with cancer (ranging from 10 to 90%). Early screening to identify individuals with sarcopenia and decreased muscle quality would allow for earlier multimodal interventions to attenuate adverse body compositional changes. Multimodal therapies (combining nutritional counselling, exercise and anti-inflammatory drugs) are currently the focus of randomised trials to examine if this approach can provide a sufficient stimulus to prevent or slow the cascade of tissue wasting and if this then impacts on outcomes in a positive manner. This review will focus on the aetiology of musculoskeletal degradation in cancer; the impact of sarcopenia on chemotherapy tolerance, post-operative complications, QoL and survival; and outline current strategies for attenuation of muscle loss in clinical practice.

Value of Precision Medicine in Advanced Non-Small Cell Lung Cancer: Real-World Outcomes Associated with the Use of Companion Diagnostics

BACKGROUND

Companion diagnostic (CDx) testing for patients with advanced non-small cell lung cancer (aNSCLC) identifies patients more likely to benefit from biomarker-driven treatments.

METHODS

Patients with nonsquamous cell (non-Sq) aNSCLC from the Flatiron Health database (diagnosed January 1, 2011-May 31, 2018) who had CDx testing were compared with those who had no reported evidence of testing. The association between CDx testing and overall survival was evaluated by unadjusted and adjusted Cox proportional hazards regression models. Logistic regression analysis identified characteristics associated with CDx testing. The revised modified Lung Cancer Prognostic Index and other factors identified a priori were included in the adjusted models.

RESULTS

A total of 17,555 patients with non-Sq aNSCLC (CDx, n = 14,732; no CDx, n = 2,823) with mean ± SD age of 67.2 ± 10.0 years were included. Most were insured (91.7%) and white (67.1%). Asian patients and those who were never-smokers were more likely to undergo CDx testing. Those with CDx testing lived longer than those without (median [95% confidence interval (CI)] survival, 13.04 [12.62-13.40] vs. 6.01 [5.72-6.24] months) and had a decreased mortality risk (adjusted hazard ratio [95% CI], 0.72 [0.69-0.76]). A survival advantage was also seen for patients with CDx testing who received biomarker-driven first-line therapy.

CONCLUSION

Patients with non-Sq aNSCLC who had CDx testing had a greater survival benefit than those without, supporting broader use of CDx testing in routine clinical practice to identify patients more likely to benefit from precision medicine.

IMPLICATIONS FOR PRACTICE

Companion diagnostic (CDx) testing coupled with biomarker-driven treatment offers a greater survival benefit for patients with advanced non-small cell lung cancer (aNSCLC). In this study, patients with nonsquamous aNSCLC from Flatiron Health, a large, real-world oncology database, with CDx testing had a reduced mortality risk and lived longer than patients without reported evidence of CDx testing; those who received biomarker-driven therapy as their first line of treatment were likely to survive three times longer than those who did not. These results demonstrate the clinical utility of CDx testing as the first step in treating nonsquamous aNSCLC in real-world clinical practice.

Prognostic value of PD-L1 expression in patients with unresectable stage III non-small cell lung cancer treated with chemoradiotherapy

Background

Expression of PD-L1 is the most investigated predictor of benefit from immune checkpoint blockade in advanced NSCLC but little is known about the association of PD-L1 expression and clinicopathological parameters of patients with unresectable stage III NSCLC.

Methods

National registry data was searched for medical records of consecutive inoperable stage III NSCLC patients treated with ChT and RT from January 2012 to December 2017. Totally 249 patients were identified that met inclusion criteria and of those 117 patients had sufficient tissue for PD-L1 immunohistochemical staining.

Results

Eighty patients (68.4%) expressed PD-L1 of ≥ 1% and 29.9% of more than 50%. Median PFS was 15.9 months in PD-L1 negative patients and 16.1 months in patients with PD-L1 expression ≥ 1% (p = 0.696). Median OS in PD-L1 negative patients was 29.9 months compared to 28.5 months in patients with PD-L1 expression ≥ % (p = 0.888). There was no difference in median OS in patients with high PD-L1 expression (≥ 50%) with 29.8 months compared to 29.9 months in those with low (1–49%) or no PD-L1 expression (p = 0.694). We found that patients who received a total dose of 60 Gy or more had significantly better median OS (32 months vs. 17.5 months, p < 0.001) as well as patients with PS 0 (33.2 vs. 20.3 months, p = 0.005).

Conclusions

In our patients PD-L1 expression had no prognostic value regarding PFS and OS. Patients with good performance status and those who received a total radiation dose of more than 60 Gy had significantly better mOS.

A systematic review and meta-analysis of the prognostic value of radiomics based models in non-small cell lung cancer treated with curative radiotherapy

BACKGROUND AND PURPOSE

Radiomics allows extraction of quantifiable features from imaging. This study performs a systematic review and meta-analysis of the performance of radiomics based prognostic models in non-small cell lung cancer (NSCLC).

MATERIALS AND METHODS

A literature review was performed following PRISMA guidelines. Medline, EMBASE and Cochrane databases were searched for articles investigating radiomics features predictive of overall survival (OS) in NSCLC treated with curative intent radiotherapy. A random-effects meta-analysis of Harrell's Concordance Index (C-index) was performed on the performance of radiomics models.

RESULTS

Of the 2746 articles retrieved, 40 studies of 55 datasets and 6223 patients were eligible for inclusion in the systematic review. There was significant heterogeneity in the methodology for feature selection and model development. Twelve datasets reported the C-index of radiomics based models in predicting OS and were included in the meta-analysis. The C-index random effects estimate was 0.57 (95% CI 0.53-0.62). There was significant heterogeneity (I² = 70.3%).

CONCLUSIONS

Based on this review, radiomics based models for lung cancer have to date demonstrated modest prognostic capabilities. Future research should consider using standardised radiomics features, robust feature selection and model development, and deep learning techniques, absolving the need for pre-defined features, to improve imaging-based models.

Systemic Inflammation and Activation of Haemostasis Predict Poor Prognosis and Response to Chemotherapy in Patients with Advanced Lung Cancer

Systemic inflammation and activation of haemostasis are common in patients with lung cancer. Both conditions support tumour growth and metastasis. Therefore, inflammatory and haemostatic biomarkers might be useful for prediction of survival and therapy response. Patients with unresectable/metastatic lung cancer initiating 1st-line chemotherapy (n = 277, 83% non-small cell lung cancer) were followed in a prospective observational cohort study. A comprehensive panel of haemostatic biomarkers (D-dimer, prothrombin fragment 1+2, soluble P-selectin, fibrinogen, coagulation factor VIII, peak thrombin generation), blood count parameters (haemoglobin, leucocytes, thrombocytes) and inflammatory markers (neutrophil-lymphocyte ratio, lymphocyte-monocyte ratio, platelet-lymphocyte ratio, C-reactive protein) were measured at baseline. We assessed the association of biomarkers with mortality, progression-free-survival (PFS) and disease-control-rate (DCR). A biomarker-based prognostic model was derived. Selected inflammatory and haemostatic biomarkers were strong and independent predictors of mortality and therapy response. The strongest predictors (D-dimer, LMR, CRP) were incorporated in a unified biomarker-based prognostic model (1-year overall-survival (OS) by risk-quartiles: 79%, 69%, 51%, 24%; 2-year-OS: 53%, 36%, 23%, 8%; log-rank p < 0.001). The biomarker-based model further predicted shorter PFS and lower DCR. In conclusion, inflammatory and haemostatic biomarkers predict poor prognosis and treatment-response in patients with advanced lung cancer. A biomarker-based prognostic score efficiently predicts mortality and disease progression beyond clinical characteristics.

Machine-Learning and Stochastic Tumor Growth Models for Predicting Outcomes in Patients With Advanced Non-Small-Cell Lung Cancer

PURPOSE

The prediction of clinical outcomes for patients with cancer is central to precision medicine and the design of clinical trials. We developed and validated machine-learning models for three important clinical end points in patients with advanced non-small-cell lung cancer (NSCLC)-objective response (OR), progression-free survival (PFS), and overall survival (OS)-using routinely collected patient and disease variables.

METHODS

We aggregated patient-level data from 17 randomized clinical trials recently submitted to the US Food and Drug Administration evaluating molecularly targeted therapy and immunotherapy in patients with advanced NSCLC. To our knowledge, this is one of the largest studies of NSCLC to consider biomarker and inhibitor therapy as candidate predictive variables. We developed a stochastic tumor growth model to predict tumor response and explored the performance of a range of machine-learning algorithms and survival models. Models were evaluated on out-of-sample data using the standard area under the receiver operating characteristic curve and concordance index (C-index) performance metrics.

RESULTS

Our models achieved promising out-of-sample predictive performances of 0.79 area under the receiver operating characteristic curve (95% CI, 0.77 to 0.81), 0.67 C-index (95% CI, 0.66 to 0.69), and 0.73 C-index (95% CI, 0.72 to 0.74) for OR, PFS, and OS, respectively. The calibration plots for PFS and OS suggested good agreement between actual and predicted survival probabilities. In addition, the Kaplan-Meier survival curves showed that the difference in survival between the low- and high-risk groups was significant (log-rank test P < .001) for both PFS and OS.

CONCLUSION

Biomarker status was the strongest predictor of OR, PFS, and OS in patients with advanced NSCLC treated with immune checkpoint inhibitors and targeted therapies. However, single biomarkers have limited predictive value, especially for programmed death-ligand 1 immunotherapy. To advance beyond the results achieved in this study, more comprehensive data on composite multiomic signatures is required.

Could venous thromboembolism and major bleeding be indicators of lung cancer mortality? A nationwide database study

Background

Venous thromboembolism (VTE) is highly prevalent in cancer patients and can cause severe morbidity. VTE treatment is essential, but anticoagulation increases the risk of major bleeding. The purpose was to evaluate the impact of VTE and major bleeding on survival and to identify significant risk factors for these events in lung cancer patients.

Methods

Data were extracted from a permanent sample of the French national health information system (including hospital and out-of-hospital care) from 2009 to 2016. All episodes of VTE and major bleeding events within one year after cancer diagnosis were identified. A Cox model was used to analyse the effect of VTE and major bleeding on the patients’ one-year survival. VTE and major bleeding risk factors were analysed with a Fine and Gray survival model.

Results

Among the 2553 included patients with lung cancer, 208 (8%) had a VTE episode in the year following diagnosis and 341 (13%) had major bleeding. Almost half of the patients died during follow-up. Fifty-six (60%) of the patients presenting with pulmonary embolism (PE) died, 48 (42%) of the patients presenting with deep vein thrombosis (DVT) alone died and 186 (55%) of those presenting with a major bleeding event died. The risk of death was significantly increased following PE and major bleeding events. VTE concomitant with cancer diagnosis was associated with an increased risk of VTE recurrence beyond 6 months after the first VTE event (sHR = 4.07 95% CI: 1.57–10.52). Most major bleeding events did not appear to be related to treatment.

Conclusion

VTE is frequent after a diagnosis of lung cancer, but so are major bleeding events. Both PE and major bleeding are associated with an increased risk of death and could be indicators of lung cancer mortality.

Integrating machine-generated mortality estimates and behavioral nudges to promote serious illness conversations for cancer patients: Design and methods for a stepped-wedge cluster randomized controlled trial

INTRODUCTION

Patients with cancer often receive care that is not aligned with their personal values and goals. Serious illness conversations (SICs) between clinicians and patients can help increase a patient's understanding of their prognosis, goals and values.

METHODS AND ANALYSIS

In this study, we describe the design of a stepped-wedge cluster randomized trial to evaluate the impact of an intervention that employs machine learning-based prognostic algorithms and behavioral nudges to prompt oncologists to have SICs with patients at high risk of short-term mortality. Data are collected on documented SICs, documented advance care planning discussions, and end-of-life care utilization (emergency room and inpatient admissions, chemotherapy and hospice utilization) for patients of all enrolled clinicians.

CONCLUSION

This trial represents a novel application of machine-generated mortality predictions combined with behavioral nudges in the routine care of outpatients with cancer. Findings from the trial may inform strategies to encourage early serious illness conversations and the application of mortality risk predictions in clinical settings.

TRIAL REGISTRATION

Clinicaltrials.gov Identifier: NCT03984773.

Comparison of the outcomes of stereotactic body radiotherapy versus surgical treatment for elderly (≥70) patients with early-stage non-small cell lung cancer after propensity score matching

Background

The optimal treatment for elderly patients with early-stage non-small cell lung cancer (NSCLC) remains inconclusive. Previous studies have shown that stereotactic body radiotherapy (SBRT) provides encouraging local control though higher incidence of toxicity in elderly than younger populations. The objective of this study was to compare the outcomes of SBRT and surgical treatment in elderly patients with clinical stage I-II NSCLC.

Methods

This retrospective analysis included 205 patients aged ≥70 years with clinical stage I NSCLC who underwent SBRT or surgery at Zhejiang Cancer Hospital (Hangzhou, China) from January 2012 to December 2017. A propensity score matching analysis was performed between the two groups. In addition, we compared outcomes and related toxicity in both study arms.

Results

Each group included 35 patients who met the inclusion criteria. Median follow-up was 50.1 (0.8–74.4) months for surgery and 35.5 (11.5–71.4) months for SBRT. The rate of cancer-specific survival was similar between the two treatment arms (p = 0.958). In patients who underwent surgery, the corresponding 3- and 5-year cancer-specific survival rates were 85.3 and 81.7%, respectively. In those who received radiotherapy, these rates were 91.3 and 74.9%, respectively. Moreover, the 3- and 5-year locoregional control in patients who underwent surgery were 90.0 and 80.0%, respectively. In those who received radiotherapy, these rates were 91.1 and 84.1%, respectively. Notably, the observed differences in progression-free survival were not statistically significant (p = 0.934). In the surgery group, grade 1–2 complications were observed in eleven patients (31%). One patient died due to perioperative infection within 30 days following surgery. There was no grade 3–5 toxicity observed in the SBRT group.

Conclusions

The outcomes of surgery and SBRT in elderly patients with early-stage NSCLC were similar.

Machine Learning Approaches to Predict 6-Month Mortality Among Patients With Cancer

Importance

Machine learning algorithms could identify patients with cancer who are at risk of short-term mortality. However, it is unclear how different machine learning algorithms compare and whether they could prompt clinicians to have timely conversations about treatment and end-of-life preferences.

Objectives

To develop, validate, and compare machine learning algorithms that use structured electronic health record data before a clinic visit to predict mortality among patients with cancer.

Design, Setting, and Participants

Cohort study of 26 525 adult patients who had outpatient oncology or hematology/oncology encounters at a large academic cancer center and 10 affiliated community practices between February 1, 2016, and July 1, 2016. Patients were not required to receive cancer-directed treatment. Patients were observed for up to 500 days after the encounter. Data analysis took place between October 1, 2018, and September 1, 2019.

Exposures

Logistic regression, gradient boosting, and random forest algorithms.

Main Outcomes and Measures

Primary outcome was 180-day mortality from the index encounter; secondary outcome was 500-day mortality from the index encounter.

Results

Among 26 525 patients in the analysis, 1065 (4.0%) died within 180 days of the index encounter. Among those who died, the mean age was 67.3 (95% CI, 66.5-68.0) years, and 500 (47.0%) were women. Among those who were alive at 180 days, the mean age was 61.3 (95% CI, 61.1-61.5) years, and 15 922 (62.5%) were women. The population was randomly partitioned into training (18 567 [70.0%]) and validation (7958 [30.0%]) cohorts at the patient level, and a randomly selected encounter was included in either the training or validation set. At a prespecified alert rate of 0.02, positive predictive values were higher for the random forest (51.3%) and gradient boosting (49.4%) algorithms compared with the logistic regression algorithm (44.7%). There was no significant difference in discrimination among the random forest (area under the receiver operating characteristic curve [AUC], 0.88; 95% CI, 0.86-0.89), gradient boosting (AUC, 0.87; 95% CI, 0.85-0.89), and logistic regression (AUC, 0.86; 95% CI, 0.84-0.88) models (P for comparison = .02). In the random forest model, observed 180-day mortality was 51.3% (95% CI, 43.6%-58.8%) in the high-risk group vs 3.4% (95% CI, 3.0%-3.8%) in the low-risk group; at 500 days, observed mortality was 64.4% (95% CI, 56.7%-71.4%) in the high-risk group and 7.6% (7.0%-8.2%) in the low-risk group. In a survey of 15 oncology clinicians with a 52.1% response rate, 100 of 171 patients (58.8%) who had been flagged as having high risk by the gradient boosting algorithm were deemed appropriate for a conversation about treatment and end-of-life preferences in the upcoming week.

Conclusions and Relevance

In this cohort study, machine learning algorithms based on structured electronic health record data accurately identified patients with cancer at risk of short-term mortality. When the gradient boosting algorithm was applied in real time, clinicians believed that most patients who had been identified as having high risk were appropriate for a timely conversation about treatment and end-of-life preferences.

Survival score to characterize prognosis in inoperable stage III NSCLC after chemoradiotherapy

Background

Stage III non-small cell lung cancer (NSCLC) represents a heterogeneous disease regarding principal patient- and tumor characteristics. A simple score may aid in personalizing multimodal therapy.

Methods

The data of 99 consecutive patients with performance status ECOG 0-1 treated until the end of 2016 with multimodal approach for inoperable NSCLC (UICC 7^th edition stage IIIA/B) were evaluated. Patient- and tumor-related factors were examined for their impact on overall survival. Factors showing a negative association with prognosis were then included in the score. Three subgroups with low, intermediate and high-risk score were defined. The results were then validated in the prospective cohort, which includes 45 patients.

Results

Most Patients were treated with concurrent (78%) or sequential (11%) chemoradiotherapy. 53% received induction chemotherapy. Median survival for the entire cohort was 20.8 (range: 15.3-26.3) months. Age (P=0.020), gender (P=0.007), pack years (P=0.015), tumor-associated atelectasis (P=0.004) and histology (P=0.004) had a significant impact on overall survival and were scored with one point each. Twelve, 59 and 28 patients were defined to have a low (0-1 points), intermediate (2-3 points) and high-risk (4-5 points) score. Median survival, 1-, 2- and 3-year survival rates were not reached, 100%, 83% and 67% in the low, 22.9 months, 80%, 47% and 24% intermediate and 13.7 months, 57%, 25% and 18% high-risk patients, respectively (P<0.001). Median survival was not reached in prospective cohort; analysis has revealed a trend for the 1-year survival rates with 100% for the low, 93% intermediate and 69% high-risk patients (P=0.100).

Conclusions

The score demonstrated remarkable survival differences in inoperable stage III NSCLC patients with good performance status receiving multimodal therapy.

Current landscape of palliative radiotherapy for non-small-cell lung cancer

Radiotherapy (RT) is a cornerstone in the management of advanced stage III and stage IV non-small-cell lung cancer (NSCLC) patients. Despite international guidelines, clinical practice remains heterogeneous. Additionally, the advent of stereotactic ablative RT (SABR) and new systemic treatments such as immunotherapy have shaken up dogmas in the approach of these patients. This review will focus on palliative thoracic RT for NSCLC but will also discuss the role of stereotactic radiotherapy, endobronchial brachytherapy (EBB), the interest of concomitant treatments (chemotherapy and immunotherapy), and the role of RT in lung cancer emergencies with palliative intent.

Trans-omics biomarker model improves prognostic prediction accuracy for early-stage lung adenocarcinoma

Limited studies have focused on developing prognostic models with trans-omics biomarkers for early-stage lung adenocarcinoma (LUAD). We performed integrative analysis of clinical information, DNA methylation, and gene expression data using 825 early-stage LUAD patients from 5 cohorts. Ranger algorithm was used to screen prognosis-associated biomarkers, which were confirmed with a validation phase. Clinical and biomarker information was fused using an iCluster plus algorithm, which significantly distinguished patients into high- and low-mortality risk groups (Pdiscovery = 0.01 and Pvalidation = 2.71×10-3). Further, potential functional DNA methylation-gene expression-overall survival pathways were evaluated by causal mediation analysis. The effect of DNA methylation level on LUAD survival was significantly mediated through gene expression level. By adding DNA methylation and gene expression biomarkers to a model of only clinical data, the AUCs of the trans-omics model improved by 18.3% (to 87.2%) and 16.4% (to 85.3%) in discovery and validation phases, respectively. Further, concordance index of the nomogram was 0.81 and 0.77 in discovery and validation phases, respectively. Based on systematic review of published literatures, our model was superior to all existing models for early-stage LUAD. In summary, our trans-omics model may help physicians accurately identify patients with high mortality risk.

Mining Prognosis Index of Brain Metastases Using Artificial Intelligence

This study is to identify the optimum prognosis index for brain metastases by machine learning. Seven hundred cancer patients with brain metastases were enrolled and divided into 446 training and 254 testing cohorts. Seven features and seven prediction methods were selected to evaluate the performance of cancer prognosis for each patient. We used mutual information and rough set with particle swarm optimization (MIRSPSO) methods to predict patient’s prognosis with the highest accuracy at area under the curve (AUC) = 0.978 ± 0.06. The improvement by MIRSPSO in terms of AUC was at 1.72%, 1.29%, and 1.83% higher than that of the traditional statistical method, sequential feature selection (SFS), mutual information with particle swarm optimization(MIPSO), and mutual information with sequential feature selection (MISFS), respectively. Furthermore, the clinical performance of the best prognosis was superior to conventional statistic method in accuracy, sensitivity, and specificity. In conclusion, identifying optimal machine-learning methods for the prediction of overall survival in brain metastases is essential for clinical applications. The accuracy rate by machine-learning is far higher than that of conventional statistic methods.

Association between single nucleotide polymorphisms of NOTCH signaling pathway-related genes and the prognosis of NSCLC

Objective

In this study, we analyzed the association between genetic variants of genes in the NOTCH signaling pathway and the prognosis of non-small-cell lung cancer (NSCLC) in the Chinese population. We also explored the interaction between genetic and epidemiological factors for the test group.

Methods

We performed genotyping of 987 NSCLC patients. Then, we used Cox proportional hazard models to analyze the associations between single-nucleotide polymorphisms (SNPs) and the prognosis of NSCLC. We employed Stata software to test the heterogeneity of associations between subgroups, and we analyzed the additive and multiplicative interactions between SNPs and epidemiologic factors.

Results

This work revealed the important prognostic and predictive value of rs915894 in the NOTCH4 gene, which may be regarded as a promising prognosis biomarker of NSCLC. Cox regression analysis indicated that the C allele of rs915894 is associated with longer survival and decreased risk of death in NSCLC (codominant model: adjusted HR =0.83, 95% CI =0.70-0.99; dominant model: adjusted HR =0.83, 95% CI =0.71-0.98). Additional stepwise regression analysis suggested that this SNP is an independently favorable factor for the prognosis of NSCLC (dominant model: adjusted HR =0.85, 95% CI =0.72-0.99). This protective effect is more pronounced for patients who are not smokers, have a history of other lung diseases, or have a family history of cancer. We also detected statistically significant additive and multiplicative interactions between rs915894 and smoking, rs915894 and history of lung diseases, and rs915894 and family history of cancer, which all affect NSCLC survival.

Conclusion

This study demonstrated that rs915894 in NOTCH 4 may be a genetic marker for NSCLC prognosis in the Chinese population and that rs915894 may have an interactive relationship with epidemiologic factors.

Presence of Chronic Obstructive Pulmonary Disease (COPD) Impair Survival in Lung Cancer Patients Receiving Epidermal Growth Factor Receptor-Tyrosine Kinase Inhibitor (EGFR-TKI): A Nationwide, Population-Based Cohort Study

The emergence of epidermal growth factor receptor-tyrosine kinase inhibitor (EGFR-TKI) caused a paradigm shift in the treatment of non-small cell lung cancer (NSCLC). Although several clinicopathologic factors to predict the response to and survival on EGFR-TKI were recognized, its efficacy has not been confirmed for patients with underlying pulmonary disease, such as chronic obstructive pulmonary disease (COPD). We conducted the study to evaluate the impact of COPD on survival for NSCLC patients that underwent EGFR-TKI treatment. The nationwide study obtained clinicopathologic data from the National Health Insurance Research Database in Taiwan between 1995 and 2013. Patients receiving EGRR-TKI were divided into COPD and non-COPD groups, and adjusted for age, sex, comorbidities, premium level and cancer treatments. Overall survival (OS) and progression-free survival (PFS) were calculated by Kaplan–Meier analysis. In total, 21,026 NSCLC patients were enrolled, of which 47.6% had COPD. After propensity score (PS) matching, all covariates were adjusted and balanced except for age (p < 0.001). In the survival analysis, the median OS (2.04 vs. 2.28 years, p < 0.001) and PFS (0.62 vs. 0.69 years, p < 0.001) of lung cancer with COPD were significantly worse than those without COPD. Lung cancer patients on EGFR-TKI treatment had a worse survival outcome if patients had pre-existing COPD.

Video-assisted thoracic lobectomy versus stereotactic body radiotherapy for stage I nonsmall cell lung cancer in elderly patients: a propensity matched comparative analysis

Comparative studies of the overall survival (OS) in elderly patients with nonsmall cell lung cancer (NSCLC) after surgery or stereotactic body radiotherapy (SBRT) have been limited by mixed extents of resection and different surgical approaches.

792 patients aged ≥65 years with clinical stage I NSCLC underwent video-assisted thoracic surgery (VATS) lobectomy or SBRT between 2010 and 2015. The propensity score-matched primary analysis included data from the full cohort; the secondary analysis included data from a subgroup of patients with data on pulmonary function.

Median OS for unmatched patients was 77 months for patients undergoing VATS lobectomy and 38 months for SBRT. The 1-, 3- and 5-year OS rates after VATS lobectomy were 92%, 76% and 65%, and after SBRT were 90%, 52% and 29% (p<0.001). Median OS for matched patients in the primary analysis was 77 months for patients undergoing VATS lobectomy and 33 months for SBRT. The 1-, 3- and 5-year OS rates after VATS lobectomy were 91%, 68% and 58%, and after SBRT were 87%, 46% and 29% (p<0.001). The survival advantage with VATS lobectomy persisted in the secondary analysis after adjusting for non-matched variables (p=0.034).

We suggest that elderly patients with stage I NSCLC undergoing VATS lobectomy have a better rate of OS than patients undergoing SBRT, irrespective of matching. This could be clinically important in decision-making for elderly patients who can tolerate surgery.