A risk model for lung cancer incidence

The HUNT lung-SNP model: genetic variants plus clinical variables improve lung cancer risk assessment over clinical models

PURPOSE

The HUNT Lung Cancer Model (HUNT LCM) predicts individualized 6-year lung cancer (LC) risk among individuals who ever smoked cigarettes with high precision based on eight clinical variables. Can the performance be improved by adding genetic information?

METHODS

A polygenic model was developed in the prospective Norwegian HUNT2 study with clinical and genotype data of individuals who ever smoked cigarettes (n = 30749, median follow up 15.26 years) where 160 LC were diagnosed within six years. It included the variables of the original HUNT LCM plus 22 single nucleotide polymorphisms (SNPs) highly associated with LC. External validation was performed in the prospective Norwegian Tromsø Study (n = 2663).

RESULTS

The novel HUNT Lung-SNP model significantly improved risk ranking of individuals over the HUNT LCM in both HUNT2 (p < 0.001) and Tromsø (p < 0.05) cohorts. Furthermore, detection rate (number of participants selected to detect one LC case) was significantly better for the HUNT Lung-SNP vs. HUNT LCM in both cohorts (42 vs. 48, p = 0.003 and 11 vs. 14, p = 0.025, respectively) as well as versus the NLST, NELSON and 2021 USPSTF criteria. The area under the receiver operating characteristic curve (AUC) was higher for the HUNT Lung-SNP in both cohorts, but significant only in HUNT2 (AUC 0.875 vs. 0.844, p < 0.001). However, the integrated discrimination improvement index (IDI) indicates a significant improvement of LC risk stratification by the HUNT Lung-SNP in both cohorts (IDI 0.019, p < 0.001 (HUNT2) and 0.013, p < 0.001 (Tromsø)).

CONCLUSION

The HUNT Lung-SNP model could have a clinical impact on LC screening and has the potential to replace the HUNT LCM as well as the NLST, NELSON and 2021 USPSTF criteria in a screening setting. However, the model should be further validated in other populations and evaluated in a prospective trial setting.

Lung Cancer Risk Prediction Models for Asian Ever-Smokers

INTRODUCTION

Although lung cancer prediction models are widely used to support risk-based screening, their performance outside Western populations remains uncertain. This study aims to evaluate the performance of 11 existing risk prediction models in multiple Asian populations and to refit prediction models for Asians.

METHODS

In a pooled analysis of 186,458 Asian ever-smokers from 19 prospective cohorts, we assessed calibration (expected-to-observed ratio) and discrimination (area under the receiver operating characteristic curve [AUC]) for each model. In addition, we developed the "Shanghai models" to better refine risk models for Asians on the basis of two well-characterized population-based prospective cohorts and externally validated them in other Asian cohorts.

RESULTS

Among the 11 models, the Lung Cancer Death Risk Assessment Tool yielded the highest AUC (AUC [95% confidence interval (CI)] = 0.71 [0.67-0.74] for lung cancer death and 0.69 [0.67-0.72] for lung cancer incidence) and the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial Model had good calibration overall (expected-to-observed ratio [95% CI] = 1.06 [0.90-1.25]). Nevertheless, these models substantially underestimated lung cancer risk among Asians who reported less than 10 smoking pack-years or stopped smoking more than or equal to 20 years ago. The Shanghai models were found to have marginal improvement overall in discrimination (AUC [95% CI] = 0.72 [0.69-0.74] for lung cancer death and 0.70 [0.67-0.72] for lung cancer incidence) but consistently outperformed the selected Western models among low-intensity smokers and long-term quitters.

CONCLUSIONS

The Shanghai models had comparable performance overall to the best existing models, but they improved much in predicting the lung cancer risk of low-intensity smokers and long-term quitters in Asia.

Risk factors for early-onset lung cancer in Korea: analysis of a nationally representative population-based cohort

OBJECTIVES

We examined the associations of socioeconomic factors, health behaviors, and comorbidities with early-onset lung cancer.

METHODS

The study included 6,794,287 individuals aged 20-39 years who participated in a Korean national health check-up program from 2009 to 2012. During the follow-up period, 4,684 participants developed lung cancer. Multivariable Cox regression analysis was used to estimate the independent associations of potential risk factors with incident lung cancer.

RESULTS

Older age (multivariable hazard ratio [mHR], 1.13; 95% confidence interval [CI], 1.12 to 1.14) and female sex (mHR, 1.62; 95% CI, 1.49 to 1.75) were associated with increased lung cancer risk. Current smoking was also associated with elevated risk (<10 pack-years: mHR, 1.12; 95% CI, 1.01 to 1.24; ≥10 pack-years: mHR, 1.30; 95% CI, 1.18 to 1.45), but past smoking was not. Although mild alcohol consumption (<10 g/day) was associated with lower lung cancer risk (mHR, 0.92; 95% CI, 0.86 to 0.99), heavier alcohol consumption (≥10 g/day) was not. Higher income (highest vs. lowest quartile: mHR, 0.86; 95% CI, 0.78 to 0.94), physical activity for at least 1,500 metabolic equivalent of task-min/wk (vs. non-exercisers: mHR, 0.83; 95% CI, 0.69 to 0.99) and obesity (vs. normal weight: mHR, 0.89; 95% CI, 0.83 to 0.96) were associated with lower lung cancer risk, whereas metabolic syndrome was associated with increased risk (mHR, 1.13; 95% CI, 1.03 to 1.24).

CONCLUSIONS

In young adults, age, female sex, smoking, and metabolic syndrome were risk factors for early-onset lung cancer, while high income, physical activity, and obesity displayed protective effects.

Machine Learning Study of SNPs in Noncoding Regions to Predict Non-small Cell Lung Cancer Susceptibility

Non-small cell lung cancer (NSCLC) is the most common pathological subtype of lung cancer. Both environmental and genetic factors have been reported to impact the lung cancer susceptibility. We conducted a genome-wide association study (GWAS) of 287 NSCLC patients and 467 healthy controls in a Chinese population using the Illumina Genome-Wide Asian Screening Array Chip on 712,095 SNPs (single nucleotide polymorphisms). Using logistic regression modeling, GWAS identified 17 new noncoding region SNP loci associated with the NSCLC risk, and the top three (rs80040741, rs9568547, rs6010259) were under a stringent p-value (<3.02e-6). Notably, rs80040741 and rs6010259 were annotated from the intron regions of MUC3A and MLC1, respectively. Together with another five SNPs previously reported in Chinese NSCLC patients and another four covariates (e.g., smoking status, age, low dose CT screening, sex), a predictive model by machine learning methods can separate the NSCLC from healthy controls with an accuracy of 86%. This is the first time to apply machine learning method in predicting the NSCLC susceptibility using both genetic and clinical characteristics. Our findings will provide a promising method in NSCLC early diagnosis and improve our understanding of applying machine learning methods in precision medicine.

Mortality Benefit of a Blood-Based Biomarker Panel for Lung Cancer on the Basis of the Prostate, Lung, Colorectal, and Ovarian Cohort

PURPOSE

To investigate the utility of integrating a panel of circulating protein biomarkers in combination with a risk model on the basis of subject characteristics to identify individuals at high risk of harboring a lethal lung cancer.

METHODS

Data from an established logistic regression model that combines four-marker protein panel (4MP) together with the Prostate, Lung, Colorectal, and Ovarian (PLCO) risk model (PLCOm2012) assayed in prediagnostic sera from 552 lung cancer cases and 2,193 noncases from the PLCO cohort were used in this study. Of the 552 lung cancer cases, 387 (70%) died of lung cancer. Cumulative incidence of lung cancer death and subdistributional and cause-specific hazard ratios (HRs) were calculated on the basis of 4MP + PLCOm2012 risk scores at a predefined 1.0% and 1.7% 6-year risk thresholds, which correspond to the current and former US Preventive Services Task Force screening criteria, respectively.

RESULTS

When considering cases diagnosed within 1 year of blood draw and all noncases, the area under receiver operation characteristics curve estimate of the 4MP + PLCOm2012 model for risk prediction of lung cancer death was 0.88 (95% CI, 0.86 to 0.90). The cumulative incidence of lung cancer death was statistically significantly higher in individuals with 4MP + PLCOm2012 scores above the 1.0% 6-year risk threshold (modified χ2, 166.27; P < .0001). Corresponding subdistributional and lung cancer death-specific HRs for test-positive cases were 9.88 (95% CI, 6.44 to 15.18) and 10.65 (95% CI, 6.93 to 16.37), respectively.

CONCLUSION

The blood-based biomarker panel in combination with PLCOm2012 identifies individuals at high risk of a lethal lung cancer.

[Construction of a Risk Prediction Model for Lung Cancer Based on Lifestyle Behaviors in the UK Biobank Large-Scale Population Cohort]

Objective

To identify the risk factors related to lifestyle behaviors that affect the incidence of lung cancer, to build a lung cancer risk prediction model to identify, in the population, individuals who are at high risk, and to facilitate the early detection of lung cancer.

Methods

The data used in the study were obtained from the UK Biobank, a database that contains information collected from 502 389 participants between March 2006 and October 2010. Based on domestic and international guidelines for lung cancer screening and high-quality research literature on lung cancer risk factors, high-risk population identification criteria were determined. Univariate Cox regression was performed to screen for risk factors of lung cancer and a multifactor lung cancer risk prediction model was constructed using Cox proportional hazards regression. Based on the comparison of Akaike information criterion and Schoenfeld residual test results, the optimal fitted model assuming proportional hazards was selected. The multiple factor Cox proportional hazards regression was performed to consider the survival time and the population was randomly divided into a training set and a validation set by a ratio of 7:3. The model was built using the training set and the performance of the model was internally validated using the validation set. The area under the receiver operating characteristic (ROC) curve ( AUC) was used to evaluate the efficacy of the model. The population was categorized into low-risk, moderate-risk, and high-risk groups based on the probability of occurrence of 0% to <25%, 25% to <75%, and 75% to 100%. The respective proportions of affected individuals in each risk group were calculated.

Results

The study eventually covered 453 558 individuals, and out of the cumulative follow-up of 5 505 402 person-years, a total of 2 330 cases of lung cancer were diagnosed. Cox proportional hazards regression was performed to identify 10 independent variables as predictors of lung cancer, including age, body mass index (BMI), education, income, physical activity, smoking status, alcohol consumption frequency, fresh fruit intake, family history of cancer, and tobacco exposure, and a model was established accordingly. Internal validation results showed that 8 independent variables (all the 10 independent variables screened out except for BMI and fresh fruit intake) were significant influencing factors of lung cancer ( P<0.05). The AUC of the training set for predicting lung cancer occurrence at one year, five years, and ten years were 0.825, 0.785, and 0.777, respectively. The AUC of the validation set for predicting lung cancer occurrence at one year, five years, and ten years were 0.857, 0.782, and 0.765, respectively. 68.38% of the individuals who might develop lung cancer in the future could be identified by screening the high-risk population.

Conclusion

We established, in this study, a model for predicting lung cancer risks associated with lifestyle behaviors of a large population. Showing good performance in discriminatory ability, the model can be used as a tool for developing standardized screening strategies for lung cancer.

Lung cancer risk score for ever and never smokers in China

BACKGROUND

Most lung cancer risk prediction models were developed in European and North-American cohorts of smokers aged ≥ 55 years, while less is known about risk profiles in Asia, especially for never smokers or individuals aged < 50 years. Hence, we aimed to develop and validate a lung cancer risk estimate tool for ever and never smokers across a wide age range.

METHODS

Based on the China Kadoorie Biobank cohort, we first systematically selected the predictors and explored the nonlinear association of predictors with lung cancer risk using restricted cubic splines. Then, we separately developed risk prediction models to construct a lung cancer risk score (LCRS) in 159,715 ever smokers and 336,526 never smokers. The LCRS was further validated in an independent cohort over a median follow-up of 13.6 years, consisting of 14,153 never smokers and 5,890 ever smokers.

RESULTS

A total of 13 and 9 routinely available predictors were identified for ever and never smokers, respectively. Of these predictors, cigarettes per day and quit years showed nonlinear associations with lung cancer risk (Pnon-linear < 0.001). The curve of lung cancer incidence increased rapidly above 20 cigarettes per day and then was relatively flat until approximately 30 cigarettes per day. We also observed that lung cancer risk declined sharply within the first 5 years of quitting, and then continued to decrease but at a slower rate in the subsequent years. The 6-year area under the receiver operating curve for the ever and never smokers' models were respectively 0.778 and 0.733 in the derivation cohort, and 0.774 and 0.759 in the validation cohort. In the validation cohort, the 10-year cumulative incidence of lung cancer was 0.39% and 2.57% for ever smokers with low (< 166.2) and intermediate-high LCRS (≥ 166.2), respectively. Never smokers with a high LCRS (≥ 21.2) had a higher 10-year cumulative incidence rate than those with a low LCRS (< 21.2; 1.05% vs. 0.22%). An online risk evaluation tool (LCKEY; http://ccra.njmu.edu.cn/lckey/web) was developed to facilitate the use of LCRS.

CONCLUSIONS

The LCRS can be an effective risk assessment tool designed for ever and never smokers aged 30 to 80 years.

Identification of non-small cell lung cancer with chronic obstructive pulmonary disease using clinical symptoms and routine examination: a retrospective study

Background

Patients with non-small cell lung cancer (NSCLC) and patients with NSCLC combined with chronic obstructive pulmonary disease (COPD) have similar physiological conditions in early stages, and the latter have shorter survival times and higher mortality rates. The purpose of this study was to develop and compare machine learning models to identify future diagnoses of COPD combined with NSCLC patients based on the patient's disease and routine clinical data.

Methods

Data were obtained from 237 patients with COPD combined with NSCLC as well as NSCLC admitted to Ningxia Hui Autonomous Region People's Hospital from October 2013 to July 2022. Six machine learning algorithms (K-nearest neighbor, logistic regression, eXtreme gradient boosting, support vector machine, naïve Bayes, and artificial neural network) were used to develop prediction models for NSCLC combined with COPD. Sensitivity, specificity, positive predictive value, negative predictive value, accuracy, F1 score, Mathews correlation coefficient (MCC), Kappa, area under the receiver operating characteristic curve (AUROC)and area under the precision-recall curve (AUPRC) were used as performance indicators to evaluate the performance of the models.

Results

135 patients with NSCLC combined with COPD, 102 patients with NSCLC were included in the study. The results showed that pulmonary function and emphysema were important risk factors and that the support vector machine-based identification model showed optimal performance with accuracy:0.946, recall:0.940, specificity:0.955, precision:0.972, npv:0.920, F1 score:0.954, MCC:0.893, Kappa:0.888, AUROC:0.975, AUPRC:0.987.

Conclusion

The use of machine learning tools combining clinical symptoms and routine examination data features is suitable for identifying the risk of concurrent NSCLC in COPD patients.

Secondary prevention and treatment innovation of early stage non-small cell lung cancer: Impact on diagnostic-therapeutic pathway from a multidisciplinary perspective

Lung cancer (LC) is the leading cause of cancer-related death worldwide, mostly because the lack of a screening program so far. Although smoking cessation has a central role in LC primary prevention, several trials on LC screening through low-dose computed tomography (LDCT) in a high risk population showed a significant reduction of LC related mortality. Most trials showed heterogeneity in terms of selection criteria, comparator arm, detection nodule method, timing and intervals of screening and duration of the follow-up. LC screening programs currently active in Europe as well as around the world will lead to a higher number of early-stage Non Small Cell Lung Cancer (NSCLC) at the diagnosis. Innovative drugs have been recently transposed from the metastatic to the perioperative setting, leading to improvements in terms of resection rates and pathological responses after induction chemoimmunotherapy, and disease free survival with targeted agents and immune checkpoint inhibitors. The present review summarizes available evidence about LC screening, highlighting potential pitfalls and benefits and underlining the impact on the diagnostic therapeutic pathway of NSCLC from a multidisciplinary perspective. Future perspectives in terms of circulating biomarkers under evaluation for patients' risk stratification as well as a focus on recent clinical trials results and ongoing studies in the perioperative setting will be also presented.

Correlation between patient-specific quality assurance in volumetric modulated arc therapy and 2D dose image features

In radiotherapy, air-filled ion chamber detectors are ubiquitously used in routine dose measurements for treatment planning. However, its use has been restricted by intrinsic low spatial resolution barriers. We developed one procedure for patient-specific quality assurance (QA) in arc radiotherapy by coalescing two adjacent measurement images into a single image to improve spatial resolution and sampling frequency, and investigated how different spatial resolutions affect the QA results. PTW 729 and 1500 ion chamber detectors were used for dosimetric verification via coalescing two measurements with 5 mm-couch shift and the isocenter, and only isocenter measurement, which we call coalescence and standard acquisition (SA). Statistical process control (SPC), process capability analysis (PCA), and receiver operating characteristic (ROC) curve were used to compare the performance of the two procedures in determining tolerance levels and identifying clinically relevant errors. By analyzing 1256 γ values calculated on interpolated data points, our results indicated that detector 1500 showed higher averages in coalescence cohorts at different tolerance criteria and the dispersion degrees were spread out smaller. Detector 729 yielded a slightly lower process capability of 0.79, 0.76, 1.10, and 1.34, but detector 1500 exhibited somewhat different results of 0.94, 1.42, 1.19, and 1.60 in magnitude. The results of SPC individual control chart showed that cases in coalescence cohorts with γ values lowering its lower control limit (LCL) were greater than those in SA cohorts for detector 1500. A combination of the width of multi-leaf collimator (MLC) leaf, the cross-sectional area of the single detector, and the spacing between adjacent detectors might lead to discrepancies in percent γ values across diverse spatial resolution scenarios. The accuracy of reconstructed volume dose is mainly determined by the interpolation algorithm used in dosimetric systems. The magnitude of filling factor in the ion chamber detectors determined its ability to detect dose deviations. SPC and PCA results indicated that coalescence procedure could detect more potential failure QA results than SA while enhancing action thresholds.

Lung Cancer Risk Prediction Nomogram in Nonsmoking Chinese Women: Retrospective Cross-sectional Cohort Study

BACKGROUND

It is believed that smoking is not the cause of approximately 53% of lung cancers diagnosed in women globally.

OBJECTIVE

The study aimed to develop and validate a simple and noninvasive model that could assess and stratify lung cancer risk in nonsmoking Chinese women.

METHODS

Based on the population-based Cancer Screening Program in Urban China, this retrospective, cross-sectional cohort study was carried out with a vast population base and an immense number of participants. The training set and the validation set were both constructed using a random distribution of the data. Following the identification of associated risk factors by multivariable Cox regression analysis, a predictive nomogram was developed. Discrimination (area under the curve) and calibration were further performed to assess the validation of risk prediction nomogram in the training set, which was then validated in the validation set.

RESULTS

In sum, 151,834 individuals signed up to take part in the survey. Both the training set (n=75,917) and the validation set (n=75,917) were comprised of randomly selected participants. Potential predictors for lung cancer included age, history of chronic respiratory disease, first-degree family history of lung cancer, menopause, and history of benign breast disease. We displayed 1-year, 3-year, and 5-year lung cancer risk-predicting nomograms using these 5 factors. In the training set, the 1-year, 3-year, and 5-year lung cancer risk areas under the curve were 0.762, 0.718, and 0.703, respectively. In the validation set, the model showed a moderate predictive discrimination.

CONCLUSIONS

We designed and validated a simple and noninvasive lung cancer risk model for nonsmoking women. This model can be applied to identify and triage people at high risk for developing lung cancers among nonsmoking women.

Risk-stratified Approach for Never- and Ever-Smokers in Lung Cancer Screening: A Prospective Cohort Study in China

Rationale: Over 40% of lung cancer cases occurred in never-smokers in China. However, high-risk never-smokers were precluded from benefiting from lung cancer screening as most screening guidelines did not consider them. Objectives: We sought to develop and validate prediction models for 3-year lung cancer risks for never- and ever-smokers, named the China National Cancer Center Lung Cancer models (China NCC-LC_m2021 models). Methods: 425,626 never-smokers and 128,952 ever-smokers from the National Lung Cancer Screening program were used as the training cohort and analyzed using multivariable Cox models. Models were validated in two independent prospective cohorts: one included 369,650 never-smokers and 107,678 ever-smokers (841 and 421 lung cancers), and the other included 286,327 never-smokers and 78,469 ever-smokers (503 and 127 lung cancers). Measurements and Main Results: The areas under the receiver operating characteristic curves in the two validation cohorts were 0.698 and 0.673 for never-smokers and 0.728 and 0.752 for ever-smokers. Our models had higher areas under the receiver operating characteristic curves than other existing models and were well calibrated in the validation cohort. The China NCC-LC_m2021 ⩾0.47% threshold was suggested for never-smokers and ⩾0.51% for ever-smokers. Moreover, we provided a range of threshold options with corresponding expected screening outcomes, screening targets, and screening efficiency. Conclusion: The construction of the China NCC-LC_m2021 models can accurately reflect individual risk of lung cancer, regardless of smoking status. Our models can significantly increase the feasibility of conducting centralized lung cancer screening programs because we provide justified thresholds to define the high-risk population of lung cancer and threshold options to adapt different configurations of medical resources.

Comparison of discrimination performance of 11 lung cancer risk models for predicting lung cancer in a prospective cohort of screening-age adults from Germany followed over 17 years

Randomized trials have demonstrated considerable reduction in lung cancer (LC) mortality by screening pre-selected heavy smokers with low-dose computed tomography (LDCT). Newer screening guidelines recommend refined LC risk models for selecting the target population for screening. We aimed to evaluate and compare the discrimination performance of LC risk models and previously used trial criteria in predicting LC incidence and mortality in a large German cohort of screening-age adults. Within ESTHER, a population-based prospective cohort study conducted in Saarland, Germany, 4812 ever smokers aged 50-75 years were followed up with respect to LC incidence and mortality for up to 17 years. We quantified the performance of 11 different LC risk models by the area under the curve (AUC) and compared the proportion of correctly predicted LC cases between the best performing models and the LDCT trial criteria. Risk prediction of LC incidence in the ESTHER ever smokers was best for the Bach model, LCRAT and LCDRAT with AUCs ranging from 0.782 to 0.787, from 0.770 to 0.774, and from 0.765 to 0.771 for the follow-up time periods of cases identified at 6, 11, and 17 years, respectively. At cutoffs yielding comparable positivity rates as the LDCT trial criteria, these models would have identified between 11.8 (95% CI 3.0-20.5) and 17.6 (95% CI 10.1-25.2) percent units higher proportions of LC cases occurring during the initial 6 years of follow-up. Use of LC risk models is expected to result in substantially greater potential to identify people at highest risk of LC, suggesting enhanced potential for reducing LC mortality by LC screening.

Lung Cancer Screening

Lung cancer screening with low-dose computed tomography (LDCT) reduces lung cancer deaths by early detection. The United States Preventive Services Task Force recommends lung cancer screening with LDCT in adults of age 50 years to 80 years who have at least a 20 pack-year smoking history and are currently smoking or have quit within the past 15 years. The implementation of a lung-cancer-screening program is complex. High-quality screening requires the involvement of a multidisciplinary team. The aim of a screening program is to find balance between mortality reduction and avoiding potential harms related to false-positive findings, overdiagnosis, invasive procedures, and radiation exposure. Components and processes of a high-quality lung-cancer-screening program include the identification of eligible individuals, shared decision-making, performing and reporting LDCT results, management of screen-detected lung nodules and non-nodule findings, smoking cessation, ensuring adherence, data collection, and quality improvement.

The Journey for Lung Cancer Screening where we Stand Today

Background

Lung cancer remains a leading cause of cancer mortality worldwide with many patients presenting with advanced disease.

Objective

We reviewed the available literature for lung cancer screening using low dose computed tomography (LDCT). We reviewed the National Lung Screening Trial (NLST), Early Lung Cancer Action Program (ELCAP) and the (Nederlands-Leuvens Longkanker Screenings Onderzoek (NELSON) trials. We also look at different lung cancer risk prediction models that may aid in identifying target populations and also discuss the cost-effectiveness of LDCT screening in different groups of smokers and ex-smokers. Lastly, we discuss recent guideline changes that have occurred in line with new and emerging evidence on lung cancer screening.

Conclusion

LDCT has been shown reduce lung cancer mortality in certain groups of current and former smokers and should be considered to help in the early diagnosis of lung cancer.

Low-dose computed tomography lung cancer screening: Clinical evidence and implementation research

Lung cancer causes more deaths than breast, cervical, and colorectal cancer combined. Nevertheless, population-based lung cancer screening is still not considered standard practice in most countries worldwide. Early lung cancer detection leads to better survival outcomes: patients diagnosed with stage 1A lung cancer have a >75% 5-year survival rate, compared to <5% at stage 4. Low-dose computed tomography (LDCT) thorax imaging for the secondary prevention of lung cancer has been studied at length, and has been shown to significantly reduce lung cancer mortality in high-risk populations. The US National Lung Screening Trial reported a 20% overall reduction in lung cancer mortality when comparing LDCT to chest X-ray, and the Nederlands-Leuvens Longkanker Screenings Onderzoek (NELSON) trial more recently reported a 24% reduction when comparing LDCT to no screening. Hence, the focus has now shifted to implementation research. Consequently, the 4-IN-THE-LUNG-RUN consortium based in five European countries, has set up a large-scale multicenter implementation trial. Successful implementation of and accessibility to LDCT lung cancer screening are dependent on many factors, not limited to population selection, recruitment strategy, computed tomography screening frequency, lung-nodule management, participant compliance, and cost effectiveness. This review provides an overview of current evidence for LDCT lung cancer screening, and draws attention to major factors that need to be addressed to successfully implement standardized, effective, and accessible screening throughout Europe. Evidence shows that through the appropriate use of risk-prediction models and a more personalized approach to screening, efficacy could be improved. Furthermore, extending the screening interval for low-risk individuals to reduce costs and associated harms is a possibility, and through the use of volumetric-based measurement and follow-up, false positive results can be greatly reduced. Finally, smoking cessation programs could be a valuable addition to screening programs and artificial intelligence could offer a solution to the added workload pressures radiologists are facing.

Polygenic risk scores to stratify cancer screening should predict mortality not incidence

Population-based cancer screening programs such as mammography or colonscopy generally directed at all healthy individuals in a given age stratum. It has recently been proposed that cancer screening could be restricted to a high-risk subgroup based on polygenic risk scores (PRSs) using panels of single-nucleotide polymorphisms (SNPs). These PRSs were, however, generated to predict cancer incidence rather than cancer mortality and will not necessarily address overdiagnosis, a major problem associated with cancer screening programs. We develop a simple net-benefit framework for evaluating screening approaches that incorporates overdiagnosis. We use this methodology to demonstrate that if a PRS does not differentially discriminate between incident and lethal cancer, restricting screening to a subgroup with high scores will only improve screening outcomes in a small number of scenarios. In contrast, restricting screening to a subgroup defined as high-risk based on a marker that is more strongly predictive of mortality than incidence will often afford greater net benefit than screening all eligible individuals. If PRS-based cancer screening is to be effective, research needs to focus on identifying PRSs associated with cancer mortality, an unchartered and clinically-relevant area of research, with a much higher potential to improve screening outcomes.

Identification of sSIGLEC5 and sLAG3 as New Relapse Predictors in Lung Cancer

Lung cancer (LC) continues to be the leading cause of cancer-related deaths in both men and women worldwide. After complete tumour resection, around half of the patients suffer from disease relapse, emphasising the critical need for robust relapse predictors in this disease. In search of such biomarkers, 83 patients with non-microcytic lung cancer and 67 healthy volunteers were studied. Pre-operative levels of sSIGLEC5 along with other soluble immune-checkpoints were measured and correlated with their clinical outcome. Soluble SIGLEC5 (sSIGLEC5) levels were higher in plasma from patients with LC compared with healthy volunteers. Looking into those patients who suffered relapse, sSIGLEC5 and sLAG3 were found to be strong relapse predictors. Following a binary logistic regression model, a sSIGLEC5 + sLAG3 score was established for disease relapse prediction (area under the curve 0.8803, 95% confidence intervals 0.7955−0.9652, cut-off > 2.782) in these patients. Based on score cut-off, a Kaplan−Meier analysis showed that patients with high sSIGLEC5 + sLAG3 score had significantly shorter relapse-free survival (p ≤ 0.0001) than those with low sSIGLEC5 + sLAG3 score.Our study suggests that pre-operative sSIGLEC5 + sLAG3 score is a robust relapse predictor in LC patients.

Potential of Inflammatory Protein Signatures for Enhanced Selection of People for Lung Cancer Screening

Randomized trials have demonstrated a substantial reduction in lung cancer (LC) mortality by screening heavy smokers with low-dose computed tomography (LDCT). The aim of this study was to assess if and to what extent blood-based inflammatory protein biomarkers might enhance selection of those at highest risk for LC screening. Ever smoking participants were chosen from 9940 participants, aged 50-75 years, who were followed up with respect to LC incidence for 17 years in a prospective population-based cohort study conducted in Saarland, Germany. Using proximity extension assay, 92 inflammation protein biomarkers were measured in baseline plasma samples of ever smoking participants, including 172 incident LC cases and 285 randomly selected participants free of LC. Smoothly clipped absolute deviation (SCAD) penalized regression with 0.632+ bootstrap for correction of overoptimism was applied to derive an inflammation protein biomarker score (INS) and a combined INS-pack-years score in a training set, and algorithms were further evaluated in an independent validation set. Furthermore, the performances of nine LC risk prediction models individually and in combination with inflammatory plasma protein biomarkers for predicting LC incidence were comparatively evaluated. The combined INS-pack-years score predicted LC incidence with area under the curves (AUCs) of 0.811 and 0.782 in the training and the validation sets, respectively. The addition of inflammatory plasma protein biomarkers to established nine LC risk models increased the AUCs up to 0.121 and 0.070 among ever smoking participants from training and validation sets, respectively. Our results suggest that inflammatory protein biomarkers may have potential to improve the selection of people for LC screening and thereby enhance screening efficiency.

Construction and Validation of a Lung Cancer Risk Prediction Model for Non-Smokers in China

Background

About 15% of lung cancers in men and 53% in women are not attributable to smoking worldwide. The aim was to develop and validate a simple and non-invasive model which could assess and stratify lung cancer risk in non-smokers in China.

Methods

A large-sample size, population-based study was conducted under the framework of the Cancer Screening Program in Urban China (CanSPUC). Data on the lung cancer screening in Henan province, China, from October 2013 to October 2019 were used and randomly divided into the training and validation sets. Related risk factors were identified through multivariable Cox regression analysis, followed by establishment of risk prediction nomogram. Discrimination [area under the curve (AUC)] and calibration were further performed to assess the validation of risk prediction nomogram in the training set, and then validated by the validation set.

Results

A total of 214,764 eligible subjects were included, with a mean age of 55.19 years. Subjects were randomly divided into the training (107,382) and validation (107,382) sets. Elder age, being male, a low education level, family history of lung cancer, history of tuberculosis, and without a history of hyperlipidemia were the independent risk factors for lung cancer. Using these six variables, we plotted 1-year, 3-year, and 5-year lung cancer risk prediction nomogram. The AUC was 0.753, 0.752, and 0.755 for the 1-, 3- and 5-year lung cancer risk in the training set, respectively. In the validation set, the model showed a moderate predictive discrimination, with the AUC was 0.668, 0.678, and 0.685 for the 1-, 3- and 5-year lung cancer risk.

Conclusions

We developed and validated a simple and non-invasive lung cancer risk model in non-smokers. This model can be applied to identify and triage patients at high risk for developing lung cancers in non-smokers.

A risk prediction model for selecting high-risk population for computed tomography lung cancer screening in China

OBJECTIVE

Two large randomized controlled trials (RCTs) have demonstrated that low dose computed tomography (LDCT) screening reduces lung cancer mortality. Risk-prediction models have been proved to select individuals for lung cancer screening effectively. With the focus on established risk factors for lung cancer routinely available in general cancer screening settings, we aimed to develop and internally validated a risk prediction model for lung cancer.

MATERIALS AND METHODS

Using data from the Cancer Screening Program in Urban China (CanSPUC) in Henan province, China between 2013 and 2019, we conducted a prospective cohort study consisting of 282,254 participants including 126,445 males and 155,809 females. Detailed questionnaire, physical assessment and follow-up were completed for all participants. Using Cox proportional risk regression analysis, we developed the Henan Lung Cancer Risk Models based on simplified questionnaire. Model discrimination was evaluated by concordance statistics (C-statistics), and model calibration was evaluated by the bootstrap sampling, respectively.

RESULTS

By 2020, a total of 589 lung cancer cases occurred in the follow-up yielding an incident density of 64.91/100,000 person-years (pyrs). Age, gender, smoking, history of tuberculosis and history of emphysema were included into the model. The C-index of the model for 1-year lung cancer risk was 0.766 and 0.741 in the training set and validation set, respectively. In stratified analysis, the model showed better predictive power in males, younger participants, and former or current smoking participants. The model calibrated well across the deciles of predicted risk in both the overall population and all subgroups.

CONCLUSIONS

We developed and internally validated a simple risk prediction model for lung cancer, which may be useful to identify high-risk individuals for more intensive screening for cancer prevention.

Artificial Intelligence-Based Prediction of Lung Cancer Risk Using Nonimaging Electronic Medical Records: Deep Learning Approach

Background

Artificial intelligence approaches can integrate complex features and can be used to predict a patient’s risk of developing lung cancer, thereby decreasing the need for unnecessary and expensive diagnostic interventions.

Objective

The aim of this study was to use electronic medical records to prescreen patients who are at risk of developing lung cancer.

Methods

We randomly selected 2 million participants from the Taiwan National Health Insurance Research Database who received care between 1999 and 2013. We built a predictive lung cancer screening model with neural networks that were trained and validated using pre-2012 data, and we tested the model prospectively on post-2012 data. An age- and gender-matched subgroup that was 10 times larger than the original lung cancer group was used to assess the predictive power of the electronic medical record. Discrimination (area under the receiver operating characteristic curve [AUC]) and calibration analyses were performed.

Results

The analysis included 11,617 patients with lung cancer and 1,423,154 control patients. The model achieved AUCs of 0.90 for the overall population and 0.87 in patients ≥55 years of age. The AUC in the matched subgroup was 0.82. The positive predictive value was highest (14.3%) among people aged ≥55 years with a pre-existing history of lung disease.

Conclusions

Our model achieved excellent performance in predicting lung cancer within 1 year and has potential to be deployed for digital patient screening. Convolution neural networks facilitate the effective use of EMRs to identify individuals at high risk for developing lung cancer.

Individual 5-Year Lung Cancer Risk Prediction Model in Korea Using a Nationwide Representative Database

Early detection of lung cancer by screening has contributed to reduce lung cancer mortality. Identifying high risk subjects for lung cancer is necessary to maximize the benefits and minimize the harms followed by lung cancer screening. In the present study, individual lung cancer risk in Korea was presented using a risk prediction model. Participants who completed health examinations in 2009 based on the Korean National Health Insurance (KNHI) database (DB) were eligible for the present study. Risk scores were assigned based on the adjusted hazard ratio (HR), and the standardized points for each risk factor were calculated to be proportional to the b coefficients. Model discrimination was assessed using the concordance statistic (c-statistic), and calibration ability assessed by plotting the mean predicted probability against the mean observed probability of lung cancer. Among candidate predictors, age, sex, smoking intensity, body mass index (BMI), presence of chronic obstructive pulmonary disease (COPD), pulmonary tuberculosis (TB), and type 2 diabetes mellitus (DM) were finally included. Our risk prediction model showed good discrimination (c-statistic, 0.810; 95% CI: 0.801-0.819). The relationship between model-predicted and actual lung cancer development correlated well in the calibration plot. When using easily accessible and modifiable risk factors, this model can help individuals make decisions regarding lung cancer screening or lifestyle modification, including smoking cessation.

Comparative performance of lung cancer risk models to define lung screening eligibility in the United Kingdom

BACKGROUND

The National Health Service England (NHS) classifies individuals as eligible for lung cancer screening using two risk prediction models, PLCOm2012 and Liverpool Lung Project-v2 (LLPv2). However, no study has compared the performance of lung cancer risk models in the UK.

METHODS

We analysed current and former smokers aged 40-80 years in the UK Biobank (N = 217,199), EPIC-UK (N = 30,813), and Generations Study (N = 25,777). We quantified model calibration (ratio of expected to observed cases, E/O) and discrimination (AUC).

RESULTS

Risk discrimination in UK Biobank was best for the Lung Cancer Death Risk Assessment Tool (LCDRAT, AUC = 0.82, 95% CI = 0.81-0.84), followed by the LCRAT (AUC = 0.81, 95% CI = 0.79-0.82) and the Bach model (AUC = 0.80, 95% CI = 0.79-0.81). Results were similar in EPIC-UK and the Generations Study. All models overestimated risk in all cohorts, with E/O in UK Biobank ranging from 1.20 for LLPv3 (95% CI = 1.14-1.27) to 2.16 for LLPv2 (95% CI = 2.05-2.28). Overestimation increased with area-level socioeconomic status. In the combined cohorts, USPSTF 2013 criteria classified 50.7% of future cases as screening eligible. The LCDRAT and LCRAT identified 60.9%, followed by PLCOm2012 (58.3%), Bach (58.0%), LLPv3 (56.6%), and LLPv2 (53.7%).

CONCLUSION

In UK cohorts, the ability of risk prediction models to classify future lung cancer cases as eligible for screening was best for LCDRAT/LCRAT, very good for PLCOm2012, and lowest for LLPv2. Our results highlight the importance of validating prediction tools in specific countries.

Personalising lung cancer screening: An overview of risk-stratification opportunities and challenges

Randomised clinical trials have shown the efficacy of computed tomography lung cancer screening, initiating discussions on whether and how to implement population‐based screening programs. Due to smoking behaviour being the primary risk‐factor for lung cancer and part of the criteria for determining screening eligibility, lung cancer screening is inherently risk‐based. In fact, the selection of high‐risk individuals has been shown to be essential in implementing lung cancer screening in a cost‐effective manner. Furthermore, studies have shown that further risk‐stratification may improve screening efficiency, allow personalisation of the screening interval and reduce health disparities. However, implementing risk‐based lung cancer screening programs also requires overcoming a number of challenges. There are indications that risk‐based approaches can negatively influence the trade‐off between individual benefits and harms if not applied thoughtfully. Large‐scale implementation of targeted, risk‐based screening programs has been limited thus far. Consequently, questions remain on how to efficiently identify and invite high‐risk individuals from the general population. Finally, while risk‐based approaches may increase screening program efficiency, efficiency should be balanced with the overall impact of the screening program. In this review, we will address the opportunities and challenges in applying risk‐stratification in different aspects of lung cancer screening programs, as well as the balance between screening program efficiency and impact.

Lung cancer screening: who pays? Who receives? The European perspectives

Lung cancer is the leading cause of cancer-related death worldwide, and its early detection is critical to achieving a curative treatment and to reducing mortality. Low-dose computed tomography (LDCT) is a highly sensitive technique for detecting noninvasive small lung tumors in high-risk populations. We here analyze the current status of lung cancer screening (LCS) from a European point of view. With economic burden of health care in most European countries resting on the state, it is important to reduce costs of screening and improve its effectiveness. Current cost-effectiveness analyses on LCS have indicated a favorable economic profile. The most recently published analysis reported an incremental cost-effectiveness ratio (ICER) of €3,297 per 1 life-year gained adjusted for the quality of life (QALY) and €2,944 per life-year gained, demonstrating a 90% probability of ICER being below €15,000 and a 98.1% probability of being below €25,000. Different risk models have been used to identify the target population; among these, the PLCOM2012 in particular allows for the selection of the population to be screened with high sensitivity. Risk models should also be employed to define screening intervals, which can reduce the general number of LDCT scans after the baseline round. Future perspectives of screening in a European scenario are related to the will of the policy makers to implement policy on a large scale and to improve the effectiveness of a broad screening of smoking-related disease, including cardiovascular prevention, by measuring coronary calcium score on LDCT. The employment of artificial intelligence (AI) in imaging interpretation, the use of liquid biopsies for the characterization of CT-detected undetermined nodules, and less invasive, personalized surgical treatments, will improve the effectiveness of LCS.

A 10-Year Probability Deep Neural Network Prediction Model for Lung Cancer

Cancer is the leading cause of death in Taiwan. According to the Cancer Registration Report of Taiwan's Ministry of Health and Welfare, a total of 13,488 people suffered from lung cancer in 2016, making it the second-most common cancer and the leading cancer in men. Compared with other types of cancer, the incidence of lung cancer is high. In this study, the National Health Insurance Research Database (NHIRDB) was used to determine the diseases and symptoms associated with lung cancer, and a 10-year probability deep neural network prediction model for lung cancer was developed. The proposed model could allow patients with a high risk of lung cancer to receive an earlier diagnosis and support the physicians' clinical decision-making. The study was designed as a cohort study. The subjects were patients who were diagnosed with lung cancer between 2000 and 2009, and the patients' disease histories were back-tracked for a period, extending to ten years before the diagnosis of lung cancer. As a result, a total of 13 diseases were selected as the predicting factors. A nine layers deep neural network model was created to predict the probability of lung cancer, depending on the different pre-diagnosed diseases, and to benefit the earlier detection of lung cancer in potential patients. The model is trained 1000 times, the batch size is set to 100, the SGD (Stochastic gradient descent) optimizer is used, the learning rate is set to 0.1, and the momentum is set to 0.1. The proposed model showed an accuracy of 85.4%, a sensitivity of 72.4% and a specificity of 85%, as well as an 87.4% area under ROC (AUROC) (95%, 0.8604-0.8885) model precision. Based on data analysis and deep learning, our prediction model discovered some features that had not been previously identified by clinical knowledge. This study tracks a decade of clinical diagnostic records to identify possible symptoms and comorbidities of lung cancer, allows early prediction of the disease, and assists more patients with early diagnosis.

Lung Screening Benefits and Challenges: A Review of The Data and Outline for Implementation

Lung cancer is the leading cause of cancer-related deaths worldwide, accounting for almost a fifth of all cancer-related deaths. Annual computed tomographic lung cancer screening (CTLS) detects lung cancer at earlier stages and reduces lung cancer-related mortality among high-risk individuals. Many medical organizations, including the U.S. Preventive Services Task Force, recommend annual CTLS in high-risk populations. However, fewer than 5% of individuals worldwide at high risk for lung cancer have undergone screening. In large part, this is owing to delayed implementation of CTLS in many countries throughout the world. Factors contributing to low uptake in countries with longstanding CTLS endorsement, such as the United States, include lack of patient and clinician awareness of current recommendations in favor of CTLS and clinician concerns about CTLS-related radiation exposure, false-positive results, overdiagnosis, and cost. This review of the literature serves to address these concerns by evaluating the potential risks and benefits of CTLS. Review of key components of a lung screening program, along with an updated shared decision aid, provides guidance for program development and optimization. Review of studies evaluating the population considered "high-risk" is included as this may affect future guidelines within the United States and other countries considering lung screening implementation.

Performance of Risk Factor-Based Guidelines and Model-Based Chest CT Lung Cancer Screening in World Trade Center-Exposed Fire Department Rescue/Recovery Workers

BACKGROUND

Lung cancer is a leading cause of cancer incidence and death in the United States. Risk factor-based guidelines and risk model-based strategies are used to identify patients who could benefit from low-dose chest CT (LDCT) screening. Few studies compare guidelines or models within the same cohort. We evaluate lung cancer screening performance of two risk factor-based guidelines (US Preventive Services Task Force 2014 recommendations [USPSTF-2014] and National Comprehensive Cancer Network Group 2 [NCCN-2]) and two risk model-based strategies, Prostate Lung Colorectal and Ovarian Cancer Screening (PLCOm2012) and the Bach model) in the same occupational cohort.

RESEARCH QUESTION

Which risk factor-based guideline or model-based strategy is most accurate in detecting lung cancers in a highly exposed occupational cohort?

STUDY DESIGN AND METHODS

Fire Department of City of New York (FDNY) rescue/recovery workers exposed to the September 11, 2001 attacks underwent LDCT lung cancer screening based on smoking history and age. The USPSTF-2014, NCCN-2, PLCOm2012 model, and Bach model were retrospectively applied to determine how many lung cancers were diagnosed using each approach.

RESULTS

Among the study population (N = 3,953), 930 underwent a baseline scan that met at least one risk factor or model-based LDCT screening strategy; 73% received annual follow-up scans. Among the 3,953, 63 lung cancers were diagnosed, of which 50 were detected by at least one LDCT screening strategy. The NCCN-2 guideline was the most sensitive (79.4%; 50/63). When compared with NCCN-2, stricter age and smoking criteria reduced sensitivity of the other guidelines/models (USPSTF-2014 [44%], PLCOm2012 [51%], and Bach[46%]). The 13 missed lung cancers were mainly attributable to smoking less and quitting longer than guideline/model eligibility criteria. False-positive rates were similar across all four guidelines/models.

INTERPRETATION

In this cohort, our findings support expanding eligibility for LDCT lung cancer screening by lowering smoking history from ≥30 to ≥20 pack-years and age from 55 years to 50 years old. Additional studies are needed to determine its generalizability to other occupational/environmental exposed cohorts.

Chronic non-communicable disease risk calculators - An overview, part I

This review identifies the different risk assessment tools that stratify the individual's risk of four of the eight leading causes of death in women: breast cancer, lung cancer, colorectal cancer and osteoporosis. It will be followed by the publication of a second paper that summarizes the risk assessment tools for the other four leading causes of death (myocardial infarction, stroke, diabetes mellitus type 2 and dementia). The different tools were compared by their use of different variables and validation criteria. To corroborate the validation process, validation study papers were considered for each risk assessment tool. Four tables, one for each illness, were designed. The tables provide an outline for each risk assessment tool, which includes its inventor/company, required variables, advantages, disadvantages and validity. These tables simplify the comparison of the different tools and enable the identification of the most suitable one for each patient.

Recommendations for Implementing Lung Cancer Screening with Low-Dose Computed Tomography in Europe

Lung cancer screening (LCS) with low-dose computed tomography (LDCT) was demonstrated in the National Lung Screening Trial (NLST) to reduce mortality from the disease. European mortality data has recently become available from the Nelson randomised controlled trial, which confirmed lung cancer mortality reductions by 26% in men and 39-61% in women. Recent studies in Europe and the USA also showed positive results in screening workers exposed to asbestos. All European experts attending the "Initiative for European Lung Screening (IELS)"-a large international group of physicians and other experts concerned with lung cancer-agreed that LDCT-LCS should be implemented in Europe. However, the economic impact of LDCT-LCS and guidelines for its effective and safe implementation still need to be formulated. To this purpose, the IELS was asked to prepare recommendations to implement LCS and examine outstanding issues. A subgroup carried out a comprehensive literature review on LDCT-LCS and presented findings at a meeting held in Milan in November 2018. The present recommendations reflect that consensus was reached.

Recommendations for Implementing Lung Cancer Screening with Low-Dose Computed Tomography in Europe

Lung cancer screening (LCS) with low-dose computed tomography (LDCT) was demonstrated in the National Lung Screening Trial (NLST) to reduce mortality from the disease. European mortality data has recently become available from the Nelson randomised controlled trial, which confirmed lung cancer mortality reductions by 26% in men and 39-61% in women. Recent studies in Europe and the USA also showed positive results in screening workers exposed to asbestos. All European experts attending the "Initiative for European Lung Screening (IELS)"-a large international group of physicians and other experts concerned with lung cancer-agreed that LDCT-LCS should be implemented in Europe. However, the economic impact of LDCT-LCS and guidelines for its effective and safe implementation still need to be formulated. To this purpose, the IELS was asked to prepare recommendations to implement LCS and examine outstanding issues. A subgroup carried out a comprehensive literature review on LDCT-LCS and presented findings at a meeting held in Milan in November 2018. The present recommendations reflect that consensus was reached.

Risk prediction model for lung cancer incorporating metabolic markers: Development and internal validation in a Chinese population

BACKGROUND

Low-dose computed tomography screening has been proved to reduce lung cancer mortality, however, the issues of high false-positive rate and overdiagnosis remain unsolved. Risk prediction models for lung cancer that could accurately identify high-risk populations may help to increase efficiency. We thus sought to develop a risk prediction model for lung cancer incorporating epidemiological and metabolic markers in a Chinese population.

METHODS

During 2006 and 2015, a total of 122 497 people were observed prospectively for lung cancer incidence with the total person-years of 976 663. Stepwise multivariable-adjusted logistic regressions with P_entry  = .15 and P_stay  = .20 were conducted to select the candidate variables including demographics and metabolic markers such as high-sensitivity C-reactive protein (hsCRP) and low-density lipoprotein cholesterol (LDL-C) into the prediction model. We used the C-statistic to evaluate discrimination, and Hosmer-Lemeshow tests for calibration. Tenfold cross-validation was conducted for internal validation to assess the model's stability.

RESULTS

A total of 984 lung cancer cases were identified during the follow-up. The epidemiological model including age, gender, smoking status, alcohol intake status, coal dust exposure status, and body mass index generated a C-statistic of 0.731. The full model additionally included hsCRP and LDL-C showed significantly better discrimination (C-statistic = 0.735, P = .033). In stratified analysis, the full model showed better predictive power in terms of C-statistic in younger participants (<50 years, 0.709), females (0.726), and former or current smokers (0.742). The model calibrated well across the deciles of predicted risk in both the overall population (P_HL  = .689) and all subgroups.

CONCLUSIONS

We developed and internally validated an easy-to-use risk prediction model for lung cancer among the Chinese population that could provide guidance for screening and surveillance.

Accounting for established predictors with the multistep elastic net

Multivariable models for prediction or estimating associations with an outcome are rarely built in isolation. Instead, they are based upon a mixture of covariates that have been evaluated in earlier studies (eg, age, sex, or common biomarkers) and covariates that were collected specifically for the current study (eg, a panel of novel biomarkers or other hypothesized risk factors). For that context, we present the multistep elastic net (MSN), which considers penalized regression with variables that can be qualitatively grouped based upon their degree of prior research support: established predictors vs unestablished predictors. The MSN chooses between uniform penalization of all predictors (the standard elastic net) and weaker penalization of the established predictors in a cross-validated framework and includes the option to impose zero penalty on the established predictors. In simulation studies that reflect the motivating context, we show the comparability or superiority of the MSN over the standard elastic net, the Integrative LASSO with Penalty Factors, the sparse group lasso, and the group lasso, and we investigate the importance of not penalizing the established predictors at all. We demonstrate the MSN to update a prediction model for pediatric ECMO patient mortality.

Will That Pulmonary Nodule Become Cancerous? A Risk Prediction Model for Incident Lung Cancer

This prospective investigation derived a prediction model for identifying risk of incident lung cancer among patients with visible lung nodules identified on computed tomography (CT). Among 2,924 eligible patients referred for evaluation of a pulmonary nodule to the Stony Brook Lung Cancer Evaluation Center between January 1, 2002 and December 31, 2015, 171 developed incident lung cancer during the observation period. Cox proportional hazard models were used to model time until disease onset. The sample was randomly divided into discovery (n = 1,469) and replication (n = 1,455) samples. In the replication sample, concordance was computed to indicate predictive accuracy and risk scores were calculated using the linear predictions. Youden index was used to identify high-risk versus low-risk patients and cumulative lung cancer incidence was examined for high-risk and low-risk groups. Multivariable analyses identified a combination of clinical and radiologic predictors for incident lung cancer including ln-age, ln-pack-years smoking, a history of cancer, chronic obstructive pulmonary disease, and several radiologic markers including spiculation, ground glass opacity, and nodule size. The final model reliably detected patients who developed lung cancer in the replication sample (C = 0.86, sensitivity/specificity = 0.73/0.81). Cumulative incidence of lung cancer was elevated in high-risk versus low-risk groups [HR = 14.34; 95% confidence interval (CI), 8.17-25.18]. Quantification of reliable risk scores has high clinical utility, enabling physicians to better stratify treatment protocols to manage patient care. The final model is among the first tools developed to predict incident lung cancer in patients presenting with a concerning pulmonary nodule.

CT screening for lung cancer: Are we ready to implement in Europe?

Lung cancer screening with low-dose CT (LDCT) is already available in certain parts of the world, such as the United States, but not yet in Europe. The recently published European position statement on lung cancer screening has recommended planning for implementation of screening to start within 18-months [1]. Pilot European programmes are already underway, primarily in the United Kingdom (UK), delivering lung cancer screening to their local populations. This review article acknowledges the evidence base for LDCT screening and will discuss the challenges that still need to be overcome in an attempt to answer the question: are we ready to implement in Europe?

Prediction of the 1-Year Risk of Incident Lung Cancer: Prospective Study Using Electronic Health Records from the State of Maine

BACKGROUND

Lung cancer is the leading cause of cancer death worldwide. Early detection of individuals at risk of lung cancer is critical to reduce the mortality rate.

OBJECTIVE

The aim of this study was to develop and validate a prospective risk prediction model to identify patients at risk of new incident lung cancer within the next 1 year in the general population.

METHODS

Data from individual patient electronic health records (EHRs) were extracted from the Maine Health Information Exchange network. The study population consisted of patients with at least one EHR between April 1, 2016, and March 31, 2018, who had no history of lung cancer. A retrospective cohort (N=873,598) and a prospective cohort (N=836,659) were formed for model construction and validation. An Extreme Gradient Boosting (XGBoost) algorithm was adopted to build the model. It assigned a score to each individual to quantify the probability of a new incident lung cancer diagnosis from October 1, 2016, to September 31, 2017. The model was trained with the clinical profile in the retrospective cohort from the preceding 6 months and validated with the prospective cohort to predict the risk of incident lung cancer from April 1, 2017, to March 31, 2018.

RESULTS

The model had an area under the curve (AUC) of 0.881 (95% CI 0.873-0.889) in the prospective cohort. Two thresholds of 0.0045 and 0.01 were applied to the predictive scores to stratify the population into low-, medium-, and high-risk categories. The incidence of lung cancer in the high-risk category (579/53,922, 1.07%) was 7.7 times higher than that in the overall cohort (1167/836,659, 0.14%). Age, a history of pulmonary diseases and other chronic diseases, medications for mental disorders, and social disparities were found to be associated with new incident lung cancer.

CONCLUSIONS

We retrospectively developed and prospectively validated an accurate risk prediction model of new incident lung cancer occurring in the next 1 year. Through statistical learning from the statewide EHR data in the preceding 6 months, our model was able to identify statewide high-risk patients, which will benefit the population health through establishment of preventive interventions or more intensive surveillance.

Referral Decision Making of General Practitioners: A Signal Detection Study

BACKGROUND

Signal detection theory (SDT) describes how respondents categorize ambiguous stimuli over repeated trials. It measures separately "discrimination" (ability to recognize a signal amid noise) and "criterion" (inclination to respond "signal" v. "noise"). This is important because respondents may produce the same accuracy rate for different reasons. We employed SDT to measure the referral decision making of general practitioners (GPs) in cases of possible lung cancer.

METHODS

We constructed 44 vignettes of patients for whom lung cancer could be considered and estimated their 1-year risk. Under UK risk-based guidelines, half of the vignettes required urgent referral. We recruited 216 GPs from practices across England. Practices differed in the positive predictive value (PPV) of their urgent referrals (chance of referrals identifying cancer) and the sensitivity (chance of cancer patients being picked up via urgent referral from their practice). Participants saw the vignettes online and indicated whether they would refer each patient urgently or not. We calculated each GP's discrimination ( d ') and criterion ( c) and regressed these on practice PPV and sensitivity, as well as on GP experience and gender.

RESULTS

Criterion was associated with practice PPV: as PPV increased, GPs' c also increased, indicating lower inclination to refer ( b = 0.06 [0.02-0.09]; P = 0.001). Female GPs were more inclined to refer than male GPs ( b = -0.20 [-0.40 to -0.001]; P = 0.049). Average discrimination was modest ( d' = 0.77), highly variable (range, -0.28 to 1.91), and not associated with practice referral performance.

CONCLUSIONS

High referral PPV at the organizational level indicates GPs' inclination to avoid false positives, not better discrimination. Rather than bluntly mandating increases in practice PPV via more referrals, it is necessary to increase discrimination by improving the evidence base for cancer referral decisions.

Risk prediction models for lung cancer: Perspectives and dissemination

Objective

The objective was to systematically assess lung cancer risk prediction models by critical evaluation of methodology, transparency and validation in order to provide a direction for future model development.

Methods

Electronic searches (including PubMed, EMbase, the Cochrane Library, Web of Science, the China National Knowledge Infrastructure, Wanfang, the Chinese BioMedical Literature Database, and other official cancer websites) were completed with English and Chinese databases until April 30th, 2018. Main reported sources were input data, assumptions and sensitivity analysis. Model validation was based on statements in the publications regarding internal validation, external validation and/or cross-validation.

Results

Twenty-two studies (containing 11 multiple-use and 11 single-use models) were included. Original models were developed between 2003 and 2016. Most of these were from the United States. Multivariate logistic regression was widely used to identify a model. The minimum area under the curve for each model was 0.57 and the largest was 0.87. The smallest C statistic was 0.59 and the largest 0.85. Six studies were validated by external validation and three were cross-validated. In total, 2 models had a high risk of bias, 6 models reported the most used variables were age and smoking duration, and 5 models included family history of lung cancer.

Conclusions

The prediction accuracy of the models was high overall, indicating that it is feasible to use models for high-risk population prediction. However, the process of model development and reporting is not optimal with a high risk of bias. This risk affects prediction accuracy, influencing the promotion and further development of the model. In view of this, model developers need to be more attentive to bias risk control and validity verification in the development of models.

Implications of Nine Risk Prediction Models for Selecting Ever-Smokers for Computed Tomography Lung Cancer Screening

BACKGROUND

Lung cancer screening guidelines recommend using individualized risk models to refer ever-smokers for screening. However, different models select different screening populations. The performance of each model in selecting ever-smokers for screening is unknown.

OBJECTIVE

To compare the U.S. screening populations selected by 9 lung cancer risk models (the Bach model; the Spitz model; the Liverpool Lung Project [LLP] model; the LLP Incidence Risk Model [LLPi]; the Hoggart model; the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial Model 2012 [PLCOM2012]; the Pittsburgh Predictor; the Lung Cancer Risk Assessment Tool [LCRAT]; and the Lung Cancer Death Risk Assessment Tool [LCDRAT]) and to examine their predictive performance in 2 cohorts.

DESIGN

Population-based prospective studies.

SETTING

United States.

PARTICIPANTS

Models selected U.S. screening populations by using data from the National Health Interview Survey from 2010 to 2012. Model performance was evaluated using data from 337 388 ever-smokers in the National Institutes of Health-AARP Diet and Health Study and 72 338 ever-smokers in the CPS-II (Cancer Prevention Study II) Nutrition Survey cohort.

MEASUREMENTS

Model calibration (ratio of model-predicted to observed cases [expected-observed ratio]) and discrimination (area under the curve [AUC]).

RESULTS

At a 5-year risk threshold of 2.0%, the models chose U.S. screening populations ranging from 7.6 million to 26 million ever-smokers. These disagreements occurred because, in both validation cohorts, 4 models (the Bach model, PLCOM2012, LCRAT, and LCDRAT) were well-calibrated (expected-observed ratio range, 0.92 to 1.12) and had higher AUCs (range, 0.75 to 0.79) than 5 models that generally overestimated risk (expected-observed ratio range, 0.83 to 3.69) and had lower AUCs (range, 0.62 to 0.75). The 4 best-performing models also had the highest sensitivity at a fixed specificity (and vice versa) and similar discrimination at a fixed risk threshold. These models showed better agreement on size of the screening population (7.6 million to 10.9 million) and achieved consensus on 73% of persons chosen.

LIMITATION

No consensus on risk thresholds for screening.

CONCLUSION

The 9 lung cancer risk models chose widely differing U.S. screening populations. However, 4 models (the Bach model, PLCOM2012, LCRAT, and LCDRAT) most accurately predicted risk and performed best in selecting ever-smokers for screening.

PRIMARY FUNDING SOURCE

Intramural Research Program of the National Institutes of Health/National Cancer Institute.

A Validated Clinical Risk Prediction Model for Lung Cancer in Smokers of All Ages and Exposure Types: A HUNT Study

Lung cancer causes >1·6 million deaths annually, with early diagnosis being paramount to effective treatment. Here we present a validated risk assessment model for lung cancer screening.

The prospective HUNT2 population study in Norway examined 65,237 people aged >20 years in 1995–97. After a median of 15·2 years, 583 lung cancer cases had been diagnosed; 552 (94·7%) ever-smokers and 31 (5·3%) never-smokers. We performed multivariable analyses of 36 candidate risk predictors, using multiple imputation of missing data and backwards feature selection with Cox regression. The resulting model was validated in an independent Norwegian prospective dataset of 45,341 ever-smokers, in which 675 lung cancers had been diagnosed after a median follow-up of 11·6 years.

Our final HUNT Lung Cancer Model included age, pack-years, smoking intensity, years since smoking cessation, body mass index, daily cough, and hours of daily indoors exposure to smoke. External validation showed a 0·879 concordance index (95% CI [0·866–0·891]) with an area under the curve of 0·87 (95% CI [0·85–0·89]) within 6 years. Only 22% of ever-smokers would need screening to identify 81·85% of all lung cancers within 6 years.

Our model of seven variables is simple, accurate, and useful for screening selection.

•

Applying this risk model in adults, screening 22% of ever-smokers would identify 81·85% of all lung cancers within 6 years.

•

Two novel highly significant factors were identified, periodical daily cough, and hours of daily indoors exposure to smoke.

•

The HUNT Lung Cancer Model is an accurate risk predictor useful in prospective screening studies for lung cancer.

The National Lung Screening Trial used selection criteria that failed to include three quarters of people who went on to develop lung cancer because they only screened heavy smokers of a certain age group (55-74). In a European Union position statement recently published in Lancet Oncology, risk stratification was identified as one of the keys to ensuring the successful implementation of future low-dose CT screening programmes in Europe. The current study has developed a new, simple and accurate model, named HUNT Lung Cancer Model, including seven clinical variables that can pick the high-risk population even among the young and the light smokers.

Development of a risk prediction model for lung cancer: The Japan Public Health Center-based Prospective Study

Although the impact of tobacco consumption on the occurrence of lung cancer is well‐established, risk estimation could be improved by risk prediction models that consider various smoking habits, such as quantity, duration, and time since quitting. We constructed a risk prediction model using a population of 59 161 individuals from the Japan Public Health Center (JPHC) Study Cohort II. A parametric survival model was used to assess the impact of age, gender, and smoking‐related factors (cumulative smoking intensity measured in pack‐years, age at initiation, and time since cessation). Ten‐year cumulative probability of lung cancer occurrence estimates were calculated with consideration of the competing risk of death from other causes. Finally, the model was externally validated using 47 501 individuals from JPHC Study Cohort I. A total of 1210 cases of lung cancer occurred during 986 408 person‐years of follow‐up. We found a dose‐dependent effect of tobacco consumption with hazard ratios for current smokers ranging from 3.78 (2.00‐7.16) for cumulative consumption ≤15 pack‐years to 15.80 (9.67‐25.79) for >75 pack‐years. Risk decreased with time since cessation. Ten‐year cumulative probability of lung cancer occurrence estimates ranged from 0.04% to 11.14% in men and 0.07% to 6.55% in women. The model showed good predictive performance regarding discrimination (cross‐validated c‐index = 0.793) and calibration (cross‐validated χ² = 6.60; P‐value = .58). The model still showed good discrimination in the external validation population (c‐index = 0.772). In conclusion, we developed a prediction model to estimate the probability of developing lung cancer based on age, gender, and tobacco consumption. This model appears useful in encouraging high‐risk individuals to quit smoking and undergo increased surveillance.

Applying Risk Prediction Models to Optimize Lung Cancer Screening: Current Knowledge, Challenges, and Future Directions

Purpose of review

Risk prediction models may be useful for facilitating effective and high-quality decision-making at critical steps in the lung cancer screening process. This review provides a current overview of published lung cancer risk prediction models and their applications to lung cancer screening and highlights both challenges and strategies for improving their predictive performance and use in clinical practice.

Recent findings

Since the 2011 publication of the National Lung Screening Trial results, numerous prediction models have been proposed to estimate the probability of developing or dying from lung cancer or the probability that a pulmonary nodule is malignant. Respective models appear to exhibit high discriminatory accuracy in identifying individuals at highest risk of lung cancer or differentiating malignant from benign pulmonary nodules. However, validation and critical comparison of the performance of these models in independent populations are limited. Little is also known about the extent to which risk prediction models are being applied in clinical practice and influencing decision-making processes and outcomes related to lung cancer screening.

Summary

Current evidence is insufficient to determine which lung cancer risk prediction models are most clinically useful and how to best implement their use to optimize screening effectiveness and quality. To address these knowledge gaps, future research should be directed toward validating and enhancing existing risk prediction models for lung cancer and evaluating the application of model-based risk calculators and its corresponding impact on screening processes and outcomes.

Lung cancer screening-low dose CT for lung cancer screening: recent trial results and next steps

Screening for lung cancer using low-dose CT has already been implemented in North America following the results of the National Lung Screening Trial. Outside North America, clinicians and researchers are addressing issues that may have a major impact on the success of screening programmes by reviewing results of existing trials and by designing new research and pilot programmes. This review summarizes the work that has been done to try to answer the remaining questions and highlights potential barriers which may affect screening uptake and cost-effectiveness.

Evaluating Molecular Biomarkers for the Early Detection of Lung Cancer: When Is a Biomarker Ready for Clinical Use? An Official American Thoracic Society Policy Statement

BACKGROUND

Molecular biomarkers have the potential to improve the current state of early lung cancer detection. The goal of this project was to develop a policy statement that provides guidance about the level of evidence required to determine that a molecular biomarker, used to support early lung cancer detection, is appropriate for clinical use.

METHODS

An ad hoc project steering committee was formed, to include individuals with expertise in the early detection of lung cancer and molecular biomarker development, from inside and outside of the Assembly on Thoracic Oncology. Key questions, generated from the results of a survey of the project steering committee, were discussed at an in-person meeting. Results of the discussion were summarized in a policy statement that was circulated to the steering committee and revised multiple times to achieve consensus.

RESULTS

With a focus on the clinical applications of lung cancer screening and lung nodule evaluation, the policy statement outlines categories of results that should be reported in the early phases of molecular biomarker development, discusses the level of evidence that would support study of the clinical utility, describes the outcomes that should be proven to consider a molecular biomarker clinically useful, and suggests study designs capable of assessing these outcomes.

CONCLUSIONS

The application of molecular biomarkers to assist with the early detection of lung cancer has the potential to substantially improve our ability to select patients for lung cancer screening, and to assist with the characterization of indeterminate lung nodules. We have described relevant considerations and have suggested standards to apply when determining whether a molecular biomarker for the early detection of lung cancer is ready for clinical use.

Lung cancer screening

PURPOSE OF REVIEW

Lung cancer screening with low-dose chest computed tomography is now recommended for high-risk individuals by the US Preventive Services Task Force. This recommendation was informed by several randomized controlled trials, the largest of which, the National Lung Screening Trial, demonstrated a 20% relative reduction in lung cancer mortality with annual low-dose chest computed tomography compared with chest radiography.

RECENT FINDINGS

The benefit of lung cancer screening must be balanced against potential harms, including a high false-positive rate with consequent further evaluative studies and invasive testing. It is critical that harms be minimized as screening generalizes to the broad community. Informed decision making between providers and patients should include individualized risk assessment, a discussion of both potential benefit and harm, and tobacco treatment. Given the multiple components required for high quality, screening should ideally occur in the context of a multidisciplinary program.

SUMMARY

We are in the early days of lung cancer screening, still with much to learn. Ongoing studies are necessary to refine the definition of a positive screen and develop better methods of distinguishing between true positive and false-positive results. Novel approaches, including the development of multicomponent lung cancer biomarkers, will likely inform and improve our future screening practice.