Development and validation of a quantitative real-time polymerase chain reaction classifier for lung cancer prognosis. J Thorac Oncol. 2011;6:1481-1487. [PMID: 21792073 DOI: 10.1097/jto.0b013e31822918bd]

Prognostic Factors across Poorly Differentiated Neuroendocrine Neoplasms: A Pooled Analysis

INTRODUCTION

Poorly differentiated neuroendocrine carcinomas (NECs) are characterized by aggressive clinical course and poor prognosis. No reliable prognostic markers have been validated to date; thus, the definition of a specific NEC prognostic algorithm represents a clinical need. This study aimed to analyze a large NEC case series to validate the specific prognostic factors identified in previous studies on gastro-entero-pancreatic and lung NECs and to assess if further prognostic parameters can be isolated.

METHODS

A pooled analysis of four NEC retrospective studies was performed to evaluate the prognostic role of Ki-67 cut-off, the overall survival (OS) according to primary cancer site, and further prognostic parameters using multivariable Cox proportional hazards model and machine learning random survival forest (RSF).

RESULTS

422 NECs were analyzed. The most represented tumor site was the colorectum (n = 156, 37%), followed by the lungs (n = 111, 26%), gastroesophageal site (n = 83, 20%; 66 gastric, 79%) and pancreas (n = 42, 10%). The Ki-67 index was the most relevant predictor, followed by morphology (pure or mixed/combined NECs), stage, and site. The predicted RSF response for survival at 1, 2, or 3 years showed decreasing survival with increasing Ki-67, pure NEC morphology, stage III-IV, and colorectal NEC disease. Patients with Ki-67 <55% and mixed/combined morphology had better survival than those with pure morphology. Morphology pure or mixed/combined became irrelevant in NEC survival when Ki-67 was ≥55%. The prognosis of metastatic patients who did not receive any treatment tended to be worse compared to that of the treated group. The prognostic impact of Rb1 immunolabeling appears to be limited when multiple risk factors are simultaneously assessed.

CONCLUSION

The most effective parameters to predict OS for NEC patients could be Ki-67, pure or mixed/combined morphology, stage, and site.

Predictors of colorectal cancer survival using cox regression and random survival forests models based on gene expression data

Understanding and identifying the markers and clinical information that are associated with colorectal cancer (CRC) patient survival is needed for early detection and diagnosis. In this work, we aimed to build a simple model using Cox proportional hazards (PH) and random survival forest (RSF) and find a robust signature for predicting CRC overall survival. We used stepwise regression to develop Cox PH model to analyse 54 common differentially expressed genes from three mutations. RSF is applied using log-rank and log-rank-score based on 5000 survival trees, and therefore, variables important obtained to find the genes that are most influential for CRC survival. We compared the predictive performance of the Cox PH model and RSF for early CRC detection and diagnosis. The results indicate that SLC9A8, IER5, ARSJ, ANKRD27, and PIPOX genes were significantly associated with the CRC overall survival. In addition, age, sex, and stages are also affecting the CRC overall survival. The RSF model using log-rank is better than log-rank-score, while log-rank-score needed more trees to stabilize. Overall, the imputation of missing values enhanced the model’s predictive performance. In addition, Cox PH predictive performance was better than RSF.

Gene Expression Profiling as a Potential Tool for Precision Oncology in Non-Small Cell Lung Cancer

Recent technological advances and the application of high-throughput mutation and transcriptome analyses have improved our understanding of cancer diseases, including non-small cell lung cancer. For instance, genomic profiling has allowed the identification of mutational events which can be treated with specific agents. However, detection of DNA alterations does not fully recapitulate the complexity of the disease and it does not allow selection of patients that benefit from chemo- or immunotherapy. In this context, transcriptional profiling has emerged as a promising tool for patient stratification and treatment guidance. For instance, transcriptional profiling has proven to be especially useful in the context of acquired resistance to targeted therapies and patients lacking targetable genomic alterations. Moreover, the comprehensive characterization of the expression level of the different pathways and genes involved in tumor progression is likely to better predict clinical benefit from different treatments than single biomarkers such as PD-L1 or tumor mutational burden in the case of immunotherapy. However, intrinsic technical and analytical limitations have hindered the use of these expression signatures in the clinical setting. In this review, we will focus on the data reported on molecular classification of non-small cell lung cancer and discuss the potential of transcriptional profiling as a predictor of survival and as a patient stratification tool to further personalize treatments.

Machine-learning algorithms for predicting hospital re-admissions in sickle cell disease

Reducing preventable hospital re-admissions in Sickle Cell Disease (SCD) could potentially improve outcomes and decrease healthcare costs. In a retrospective study of electronic health records, we hypothesized Machine-Learning (ML) algorithms may outperform standard re-admission scoring systems (LACE and HOSPITAL indices). Participants (n = 446) included patients with SCD with at least one unplanned inpatient encounter between January 1, 2013, and November 1, 2018. Patients were randomly partitioned into training and testing groups. Unplanned hospital admissions (n = 3299) were stratified to training and testing samples. Potential predictors (n = 486), measured from the last unplanned inpatient discharge to the current unplanned inpatient visit, were obtained via both data-driven methods and clinical knowledge. Three standard ML algorithms, Logistic Regression (LR), Support-Vector Machine (SVM), and Random Forest (RF) were applied. Prediction performance was assessed using the C-statistic, sensitivity, and specificity. In addition, we reported the most important predictors in our best models. In this dataset, ML algorithms outperformed LACE [C-statistic 0·6, 95% Confidence Interval (CI) 0·57-0·64] and HOSPITAL (C-statistic 0·69, 95% CI 0·66-0·72), with the RF (C-statistic 0·77, 95% CI 0·73-0·79) and LR (C-statistic 0·77, 95% CI 0·73-0·8) performing the best. ML algorithms can be powerful tools in predicting re-admission in high-risk patient groups.

Machine learning approach to predict postoperative opioid requirements in ambulatory surgery patients

Opioids play a critical role in acute postoperative pain management. Our objective was to develop machine learning models to predict postoperative opioid requirements in patients undergoing ambulatory surgery. To develop the models, we used a perioperative dataset of 13,700 patients (≥ 18 years) undergoing ambulatory surgery between the years 2016–2018. The data, comprising of patient, procedure and provider factors that could influence postoperative pain and opioid requirements, was randomly split into training (80%) and validation (20%) datasets. Machine learning models of different classes were developed to predict categorized levels of postoperative opioid requirements using the training dataset and then evaluated on the validation dataset. Prediction accuracy was used to differentiate model performances. The five types of models that were developed returned the following accuracies at two different stages of surgery: 1) Prior to surgery—Multinomial Logistic Regression: 71%, Naïve Bayes: 67%, Neural Network: 30%, Random Forest: 72%, Extreme Gradient Boost: 71% and 2) End of surgery—Multinomial Logistic Regression: 71%, Naïve Bayes: 63%, Neural Network: 32%, Random Forest: 72%, Extreme Gradient Boost: 70%. Analyzing the sensitivities of the best performing Random Forest model showed that the lower opioid requirements are predicted with better accuracy (89%) as compared with higher opioid requirements (43%). Feature importance (% relative importance) of model predictions showed that the type of procedure (15.4%), medical history (12.9%) and procedure duration (12.0%) were the top three features contributing to model predictions. Overall, the contribution of patient and procedure features towards model predictions were 65% and 35% respectively. Machine learning models could be used to predict postoperative opioid requirements in ambulatory surgery patients and could potentially assist in better management of their postoperative acute pain.

Using machine learning to predict risk of incident opioid use disorder among fee-for-service Medicare beneficiaries: A prognostic study

OBJECTIVE

To develop and validate a machine-learning algorithm to improve prediction of incident OUD diagnosis among Medicare beneficiaries with ≥1 opioid prescriptions.

METHODS

This prognostic study included 361,527 fee-for-service Medicare beneficiaries, without cancer, filling ≥1 opioid prescriptions from 2011-2016. We randomly divided beneficiaries into training, testing, and validation samples. We measured 269 potential predictors including socio-demographics, health status, patterns of opioid use, and provider-level and regional-level factors in 3-month periods, starting from three months before initiating opioids until development of OUD, loss of follow-up or end of 2016. The primary outcome was a recorded OUD diagnosis or initiating methadone or buprenorphine for OUD as proxy of incident OUD. We applied elastic net, random forests, gradient boosting machine, and deep neural network to predict OUD in the subsequent three months. We assessed prediction performance using C-statistics and other metrics (e.g., number needed to evaluate to identify an individual with OUD [NNE]). Beneficiaries were stratified into subgroups by risk-score decile.

RESULTS

The training (n = 120,474), testing (n = 120,556), and validation (n = 120,497) samples had similar characteristics (age ≥65 years = 81.1%; female = 61.3%; white = 83.5%; with disability eligibility = 25.5%; 1.5% had incident OUD). In the validation sample, the four approaches had similar prediction performances (C-statistic ranged from 0.874 to 0.882); elastic net required the fewest predictors (n = 48). Using the elastic net algorithm, individuals in the top decile of risk (15.8% [n = 19,047] of validation cohort) had a positive predictive value of 0.96%, negative predictive value of 99.7%, and NNE of 104. Nearly 70% of individuals with incident OUD were in the top two deciles (n = 37,078), having highest incident OUD (36 to 301 per 10,000 beneficiaries). Individuals in the bottom eight deciles (n = 83,419) had minimal incident OUD (3 to 28 per 10,000).

CONCLUSIONS

Machine-learning algorithms improve risk prediction and risk stratification of incident OUD in Medicare beneficiaries.

Evaluation of Machine-Learning Algorithms for Predicting Opioid Overdose Risk Among Medicare Beneficiaries With Opioid Prescriptions

IMPORTANCE

Current approaches to identifying individuals at high risk for opioid overdose target many patients who are not truly at high risk.

OBJECTIVE

To develop and validate a machine-learning algorithm to predict opioid overdose risk among Medicare beneficiaries with at least 1 opioid prescription.

DESIGN, SETTING, AND PARTICIPANTS

A prognostic study was conducted between September 1, 2017, and December 31, 2018. Participants (n = 560 057) included fee-for-service Medicare beneficiaries without cancer who filled 1 or more opioid prescriptions from January 1, 2011, to December 31, 2015. Beneficiaries were randomly and equally divided into training, testing, and validation samples.

EXPOSURES

Potential predictors (n = 268), including sociodemographics, health status, patterns of opioid use, and practitioner-level and regional-level factors, were measured in 3-month windows, starting 3 months before initiating opioids until loss of follow-up or the end of observation.

MAIN OUTCOMES AND MEASURES

Opioid overdose episodes from inpatient and emergency department claims were identified. Multivariate logistic regression (MLR), least absolute shrinkage and selection operator-type regression (LASSO), random forest (RF), gradient boosting machine (GBM), and deep neural network (DNN) were applied to predict overdose risk in the subsequent 3 months after initiation of treatment with prescription opioids. Prediction performance was assessed using the C statistic and other metrics (eg, sensitivity, specificity, and number needed to evaluate [NNE] to identify one overdose). The Youden index was used to identify the optimized threshold of predicted score that balanced sensitivity and specificity.

RESULTS

Beneficiaries in the training (n = 186 686), testing (n = 186 685), and validation (n = 186 686) samples had similar characteristics (mean [SD] age of 68.0 [14.5] years, and approximately 63% were female, 82% were white, 35% had disabilities, 41% were dual eligible, and 0.60% had at least 1 overdose episode). In the validation sample, the DNN (C statistic = 0.91; 95% CI, 0.88-0.93) and GBM (C statistic = 0.90; 95% CI, 0.87-0.94) algorithms outperformed the LASSO (C statistic = 0.84; 95% CI, 0.80-0.89), RF (C statistic = 0.80; 95% CI, 0.75-0.84), and MLR (C statistic = 0.75; 95% CI, 0.69-0.80) methods for predicting opioid overdose. At the optimized sensitivity and specificity, DNN had a sensitivity of 92.3%, specificity of 75.7%, NNE of 542, positive predictive value of 0.18%, and negative predictive value of 99.9%. The DNN classified patients into low-risk (76.2% [142 180] of the cohort), medium-risk (18.6% [34 579] of the cohort), and high-risk (5.2% [9747] of the cohort) subgroups, with only 1 in 10 000 in the low-risk subgroup having an overdose episode. More than 90% of overdose episodes occurred in the high-risk and medium-risk subgroups, although positive predictive values were low, given the rare overdose outcome.

CONCLUSIONS AND RELEVANCE

Machine-learning algorithms appear to perform well for risk prediction and stratification of opioid overdose, especially in identifying low-risk subgroups that have minimal risk of overdose.

A Predictive 7-Gene Assay and Prognostic Protein Biomarkers for Non-small Cell Lung Cancer

PURPOSE

This study aims to develop a multi-gene assay predictive of the clinical benefits of chemotherapy in non-small cell lung cancer (NSCLC) patients, and substantiate their protein expression as potential therapeutic targets.

PATIENTS AND METHODS

The mRNA expression of 160 genes identified from microarray was analyzed in qRT-PCR assays of independent 337 snap-frozen NSCLC tumors to develop a predictive signature. A clinical trial JBR.10 was included in the validation. Hazard ratio was used to select genes, and decision-trees were used to construct the predictive model. Protein expression was quantified with AQUA in 500 FFPE NSCLC samples.

RESULTS

A 7-gene signature was identified from training cohort (n = 83) with accurate patient stratification (P = 0.0043) and was validated in independent patient cohorts (n = 248, P < 0.0001) in Kaplan-Meier analyses. In the predicted benefit group, there was a significantly better disease-specific survival in patients receiving adjuvant chemotherapy in both training (P = 0.035) and validation (P = 0.0049) sets. In the predicted non-benefit group, there was no survival benefit in patients receiving chemotherapy in either set. The protein expression of ZNF71 quantified with AQUA scores produced robust patient stratification in separate training (P = 0.021) and validation (P = 0.047) NSCLC cohorts. The protein expression of CD27 quantified with ELISA had a strong correlation with its mRNA expression in NSCLC tumors (Spearman coefficient = 0.494, P < 0.0088). Multiple signature genes had concordant DNA copy number variation, mRNA and protein expression in NSCLC progression.

CONCLUSIONS

This study presents a predictive multi-gene assay and prognostic protein biomarkers clinically applicable for improving NSCLC treatment, with important implications in lung cancer chemotherapy and immunotherapy.

Application of random survival forests in understanding the determinants of under-five child mortality in Uganda in the presence of covariates that satisfy the proportional and non-proportional hazards assumption

BACKGROUND

Uganda just like any other Sub-Saharan African country, has a high under-five child mortality rate. To inform policy on intervention strategies, sound statistical methods are required to critically identify factors strongly associated with under-five child mortality rates. The Cox proportional hazards model has been a common choice in analysing data to understand factors strongly associated with high child mortality rates taking age as the time-to-event variable. However, due to its restrictive proportional hazards (PH) assumption, some covariates of interest which do not satisfy the assumption are often excluded in the analysis to avoid mis-specifying the model. Otherwise using covariates that clearly violate the assumption would mean invalid results.

METHODS

Survival trees and random survival forests are increasingly becoming popular in analysing survival data particularly in the case of large survey data and could be attractive alternatives to models with the restrictive PH assumption. In this article, we adopt random survival forests which have never been used in understanding factors affecting under-five child mortality rates in Uganda using Demographic and Health Survey data. Thus the first part of the analysis is based on the use of the classical Cox PH model and the second part of the analysis is based on the use of random survival forests in the presence of covariates that do not necessarily satisfy the PH assumption.

RESULTS

Random survival forests and the Cox proportional hazards model agree that the sex of the household head, sex of the child, number of births in the past 1 year are strongly associated to under-five child mortality in Uganda given all the three covariates satisfy the PH assumption. Random survival forests further demonstrated that covariates that were originally excluded from the earlier analysis due to violation of the PH assumption were important in explaining under-five child mortality rates. These covariates include the number of children under the age of five in a household, number of births in the past 5 years, wealth index, total number of children ever born and the child's birth order. The results further indicated that the predictive performance for random survival forests built using covariates including those that violate the PH assumption was higher than that for random survival forests built using only covariates that satisfy the PH assumption.

CONCLUSIONS

Random survival forests are appealing methods in analysing public health data to understand factors strongly associated with under-five child mortality rates especially in the presence of covariates that violate the proportional hazards assumption.

A comparison of the conditional inference survival forest model to random survival forests based on a simulation study as well as on two applications with time-to-event data

Background

Random survival forest (RSF) models have been identified as alternative methods to the Cox proportional hazards model in analysing time-to-event data. These methods, however, have been criticised for the bias that results from favouring covariates with many split-points and hence conditional inference forests for time-to-event data have been suggested. Conditional inference forests (CIF) are known to correct the bias in RSF models by separating the procedure for the best covariate to split on from that of the best split point search for the selected covariate.

Methods

In this study, we compare the random survival forest model to the conditional inference model (CIF) using twenty-two simulated time-to-event datasets. We also analysed two real time-to-event datasets. The first dataset is based on the survival of children under-five years of age in Uganda and it consists of categorical covariates with most of them having more than two levels (many split-points). The second dataset is based on the survival of patients with extremely drug resistant tuberculosis (XDR TB) which consists of mainly categorical covariates with two levels (few split-points).

Results

The study findings indicate that the conditional inference forest model is superior to random survival forest models in analysing time-to-event data that consists of covariates with many split-points based on the values of the bootstrap cross-validated estimates for integrated Brier scores. However, conditional inference forests perform comparably similar to random survival forests models in analysing time-to-event data consisting of covariates with fewer split-points.

Conclusion

Although survival forests are promising methods in analysing time-to-event data, it is important to identify the best forest model for analysis based on the nature of covariates of the dataset in question.

Using machine learning to examine medication adherence thresholds and risk of hospitalization

BACKGROUND

Quality improvement efforts are frequently tied to patients achieving ≥80% medication adherence. However, there is little empirical evidence that this threshold optimally predicts important health outcomes.

OBJECTIVE

To apply machine learning to examine how adherence to oral hypoglycemic medications is associated with avoidance of hospitalizations, and to identify adherence thresholds for optimal discrimination of hospitalization risk.

METHODS

A retrospective cohort study of 33,130 non-dual-eligible Medicaid enrollees with type 2 diabetes. We randomly selected 90% of the cohort (training sample) to develop the prediction algorithm and used the remaining (testing sample) for validation. We applied random survival forests to identify predictors for hospitalization and fit survival trees to empirically derive adherence thresholds that best discriminate hospitalization risk, using the proportion of days covered (PDC).

OUTCOMES

Time to first all-cause and diabetes-related hospitalization.

RESULTS

The training and testing samples had similar characteristics (mean age, 48 y; 67% female; mean PDC=0.65). We identified 8 important predictors of all-cause hospitalizations (rank in order): prior hospitalizations/emergency department visit, number of prescriptions, diabetes complications, insulin use, PDC, number of prescribers, Elixhauser index, and eligibility category. The adherence thresholds most discriminating for risk of all-cause hospitalization varied from 46% to 94% according to patient health and medication complexity. PDC was not predictive of hospitalizations in the healthiest or most complex patient subgroups.

CONCLUSIONS

Adherence thresholds most discriminating of hospitalization risk were not uniformly 80%. Machine-learning approaches may be valuable to identify appropriate patient-specific adherence thresholds for measuring quality of care and targeting nonadherent patients for intervention.

Integrating Tumor and Stromal Gene Expression Signatures With Clinical Indices for Survival Stratification of Early-Stage Non-Small Cell Lung Cancer

BACKGROUND

Accurate survival stratification in early-stage non-small cell lung cancer (NSCLC) could inform the use of adjuvant therapy. We developed a clinically implementable mortality risk score incorporating distinct tumor microenvironmental gene expression signatures and clinical variables.

METHODS

Gene expression profiles from 1106 nonsquamous NSCLCs were used for generation and internal validation of a nine-gene molecular prognostic index (MPI). A quantitative polymerase chain reaction (qPCR) assay was developed and validated on an independent cohort of formalin-fixed paraffin-embedded (FFPE) tissues (n = 98). A prognostic score using clinical variables was generated using Surveillance, Epidemiology, and End Results data and combined with the MPI. All statistical tests for survival were two-sided.

RESULTS

The MPI stratified stage I patients into prognostic categories in three microarray and one FFPE qPCR validation cohorts (HR = 2.99, 95% CI = 1.55 to 5.76, P < .001 in stage IA patients of the largest microarray validation cohort; HR = 3.95, 95% CI = 1.24 to 12.64, P = .01 in stage IA of the qPCR cohort). Prognostic genes were expressed in distinct tumor cell subpopulations, and genes implicated in proliferation and stem cells portended poor outcomes, while genes involved in normal lung differentiation and immune infiltration were associated with superior survival. Integrating the MPI with clinical variables conferred greatest prognostic power (HR = 3.43, 95% CI = 2.18 to 5.39, P < .001 in stage I patients of the largest microarray cohort; HR = 3.99, 95% CI = 1.67 to 9.56, P < .001 in stage I patients of the qPCR cohort). Finally, the MPI was prognostic irrespective of somatic alterations in EGFR, KRAS, TP53, and ALK.

CONCLUSION

The MPI incorporates genes expressed in the tumor and its microenvironment and can be implemented clinically using qPCR assays on FFPE tissues. A composite model integrating the MPI with clinical variables provides the most accurate risk stratification.

Incorporating auxiliary information for improved prediction using combination of kernel machines

With evolving genomic technologies, it is possible to get different measures of the same underlying biological phenomenon using different technologies. The goal of this paper is to build a prediction model for an outcome variable Y from covariates X. Besides X, we have surrogate covariates W which are related to X. We want to utilize the information in W to boost the prediction for Y using X. In this paper, we propose a kernel machine-based method to improve prediction of Y by X by incorporating auxiliary information W. By combining single kernel machines, we also propose a hybrid kernel machine predictor, which can yield a smaller prediction error than its constituents. The prediction error of our kernel machine predictors is evaluated using simulations. We also apply our method to a lung cancer dataset and an Alzheimer's disease dataset.

C4.4A as a biomarker in pulmonary adenocarcinoma and squamous cell carcinoma

The high prevalence and mortality of lung cancer, together with a poor 5-year survival of only approximately 15%, emphasize the need for prognostic and predictive factors to improve patient treatment. C4.4A, a member of the Ly6/uPAR family of membrane proteins, qualifies as such a potential informative biomarker in non-small cell lung cancer. Under normal physiological conditions, it is primarily expressed in suprabasal layers of stratified squamous epithelia. Consequently, it is absent from healthy bronchial and alveolar tissue, but nevertheless appears at early stages in the progression to invasive carcinomas of the lung, i.e., in bronchial hyperplasia/metaplasia and atypical adenomatous hyperplasia. In the stages leading to pulmonary squamous cell carcinoma, expression is sustained in dysplasia, carcinoma in situ and invasive carcinomas, and this pertains to the normal presence of C4.4A in squamous epithelium. In pulmonary adenocarcinomas, a fraction of cases is positive for C4.4A, which is surprising, given the origin of these carcinomas from mucin-producing and not squamous epithelium. Interestingly, this correlates with a highly compromised patient survival and a predominant solid tumor growth pattern. Circumstantial evidence suggests an inverse relationship between C4.4A and the tumor suppressor LKB1. This might provide a link to the prognostic impact of C4.4A in patients with adenocarcinomas of the lung and could potentially be exploited for predicting the efficacy of treatment targeting components of the LKB1 pathway.

Metformin and Not Diabetes Influences the Survival of Resected Early Stage NSCLC Patients

BACKGROUND

Published data suggest that diabetes influences survival of patients with lung cancer. The anti-cancer effect of metformin confounds this association. We sought to study the association of diabetes and metformin with survival in patients undergoing resection of stage I non-small cell lung cancer (NSCLC).

METHODS

Pathologic stage I NSCLC patients undergoing anatomic resection from 2002 to 2011 were studied. A diagnosis of diabetes and diabetic medication use were identified through records. Univariate and multivariate analyses examined the association of diabetes and metformin usage with overall survival (OS).

RESULTS

409 eligible patients were included in the analysis - excluding patients with neoadjuvant therapy, more than one lung cancer, or resection less than lobectomy. 71 (17.4%) patients were diabetics and 41 (10.0%) used metformin. With a median follow up of 44 months, univariate analysis demonstrates that diabetes had no effect on OS (P=0.75); however, metformin use was associated with improved OS (median survival not reached vs. 60 months; P=0.02). Metformin use remained an important predictor of good survival in multivariate analysis (HR=3.08; P<0.01) after adjusting for age, gender, pathologic stage, histology and smoking status.

CONCLUSION

Metformin use rather than diabetes is associated with improved long-term survival in Stage I NSCLC patients.

Epithelial-mesenchymal transition-associated secretory phenotype predicts survival in lung cancer patients

In cancer cells, the process of epithelial-mesenchymal transition (EMT) confers migratory and invasive capacity, resistance to apoptosis, drug resistance, evasion of host immune surveillance and tumor stem cell traits. Cells undergoing EMT may represent tumor cells with metastatic potential. Characterizing the EMT secretome may identify biomarkers to monitor EMT in tumor progression and provide a prognostic signature to predict patient survival. Utilizing a transforming growth factor-β-induced cell culture model of EMT, we quantitatively profiled differentially secreted proteins, by GeLC-tandem mass spectrometry. Integrating with the corresponding transcriptome, we derived an EMT-associated secretory phenotype (EASP) comprising of proteins that were differentially upregulated both at protein and mRNA levels. Four independent primary tumor-derived gene expression data sets of lung cancers were used for survival analysis by the random survival forests (RSF) method. Analysis of 97-gene EASP expression in human lung adenocarcinoma tumors revealed strong positive correlations with lymph node metastasis, advanced tumor stage and histological grade. RSF analysis built on a training set (n = 442), including age, sex and stage as variables, stratified three independent lung cancer data sets into low-, medium- and high-risk groups with significant differences in overall survival. We further refined EASP to a 20 gene signature (rEASP) based on variable importance scores from RSF analysis. Similar to EASP, rEASP predicted survival of both adenocarcinoma and squamous carcinoma patients. More importantly, it predicted survival in the early-stage cancers. These results demonstrate that integrative analysis of the critical biological process of EMT provides mechanism-based and clinically relevant biomarkers with significant prognostic value.

BAYESIAN SHRINKAGE METHODS FOR PARTIALLY OBSERVED DATA WITH MANY PREDICTORS

Motivated by the increasing use of and rapid changes in array technologies, we consider the prediction problem of fitting a linear regression relating a continuous outcome Y to a large number of covariates X , eg measurements from current, state-of-the-art technology. For most of the samples, only the outcome Y and surrogate covariates, W , are available. These surrogates may be data from prior studies using older technologies. Owing to the dimension of the problem and the large fraction of missing information, a critical issue is appropriate shrinkage of model parameters for an optimal bias-variance tradeoff. We discuss a variety of fully Bayesian and Empirical Bayes algorithms which account for uncertainty in the missing data and adaptively shrink parameter estimates for superior prediction. These methods are evaluated via a comprehensive simulation study. In addition, we apply our methods to a lung cancer dataset, predicting survival time (Y) using qRT-PCR ( X ) and microarray ( W ) measurements.

Ly6/uPAR-related protein C4.4A as a marker of solid growth pattern and poor prognosis in lung adenocarcinoma

INTRODUCTION

We have recently shown that the protein C4.4A is induced in early precursor lesions of pulmonary adenocarcinomas and squamous cell carcinomas. In the present study, we aimed at analyzing the impact of C4.4A on the survival of non-small cell lung cancer patients and determining whether its unexpected expression in adenocarcinomas could be attributed to a specific growth type (lepidic, acinar, papillary, micropapillary, solid).

METHODS

Sections from the center and periphery of the primary tumor, as well as N2-positive lymph node metastases, were stained by immunohistochemistry for C4.4A and scored semi-quantitatively for intensity and frequency of positive tumor cells.

RESULTS

C4.4A score (intensity × frequency) in the tumor center was a highly significant prognostic factor in adenocarcinomas (n = 88), both in univariate (p = 0.004; hazard ratio [95% confidence interval] = 1.44 [1.12-1.85]) and multivariate statistical analysis (p = 0.0005; hazard ratio = 1.65 [1.24-2.19]), demonstrating decreasing survival with increasing score. In contrast, C4.4A did not provide prognostic information in squamous cell carcinomas (n = 104). Pathological stage was significant in both groups. In the adenocarcinomas, C4.4A expression was clearly associated with, but a stronger prognostic factor than, solid growth.

CONCLUSIONS

The present results substantiate the potential value of C4.4A as a prognostic marker in pulmonary adenocarcinomas seen earlier in a smaller, independent patient cohort. Importantly, we also show that C4.4A is a surrogate marker for adenocarcinoma solid growth. Recent data suggest that C4.4A is negatively regulated by the tumor suppressor liver kinase B1, which is inactivated in some adenocarcinomas, providing a possible link to the impact of C4.4A on the survival of these patients.

Update on recent prognostic markers in adenocarcinoma

SUMMARY Biomarkers have evolved from individual markers to gene signatures validated in multi-institutional studies and independent datasets. Multiple gene signatures have been associated with survival, as well as defining outcome following specific adjuvant chemotherapy. The heterogeneity of lung cancer has necessitated the incorporation of multiple genes in mRNA signatures. The use of plasma miRNA profiles in the diagnosis or prognosis of early-stage lung carcinoma also appears promising. However, challenges remain, with questions regarding reproducibility, standardized techniques, cost and differences between gene signatures. Before they can be widely adopted, molecular profiles must be validated in large, prospective trials evaluating the specific patient population for which the profiles are to be utilized.

Incorporating auxiliary information for improved prediction in high-dimensional datasets: an ensemble of shrinkage approaches

With advancement in genomic technologies, it is common that two high-dimensional datasets are available, both measuring the same underlying biological phenomenon with different techniques. We consider predicting a continuous outcome Y using X, a set of p markers which is the best available measure of the underlying biological process. This same biological process may also be measured by W, coming from a prior technology but correlated with X. On a moderately sized sample, we have (Y,X,W), and on a larger sample we have (Y,W). We utilize the data on W to boost the prediction of Y by X. When p is large and the subsample containing X is small, this is a p>n situation. When p is small, this is akin to the classical measurement error problem; however, ours is not the typical goal of calibrating W for use in future studies. We propose to shrink the regression coefficients β of Y on X toward different targets that use information derived from W in the larger dataset. We compare these proposals with the classical ridge regression of Y on X, which does not use W. We also unify all of these methods as targeted ridge estimators. Finally, we propose a hybrid estimator which is a linear combination of multiple estimators of β. With an optimal choice of weights, the hybrid estimator balances efficiency and robustness in a data-adaptive way to theoretically yield a smaller prediction error than any of its constituents. The methods, including a fully Bayesian alternative, are evaluated via simulation studies. We also apply them to a gene-expression dataset. mRNA expression measured via quantitative real-time polymerase chain reaction is used to predict survival time in lung cancer patients, with auxiliary information from microarray technology available on a larger sample.

Characterization of vitamin D receptor (VDR) in lung adenocarcinoma

PURPOSE

The anti-proliferative effects of 1α,25-dihydroxyvitamin D(3) (1,25-D(3), calcitriol, the active form of vitamin D) are mediated by the nuclear vitamin D receptor (VDR). In the present study, we characterized VDR expression in lung adenocarcinoma (AC).

EXPERIMENTAL DESIGN

We examined VDR mRNA expression using a quantitative real-time PCR (qRT-PCR) in 100 patients who underwent surgery for lung AC. In a subset of these patients (n=89), we examined VDR protein expression using immunohistochemistry. We also examined the association of VDR protein expression with circulating serum levels of 25-hydroxyvitamin D(3) (25-D(3)) and 1,25-D(3). The antiproliferative effects and cell cycle arrest of 1,25-D(3) were examined using lung cancer cell lines with high (SKLU-1) as well as low (A549) expression of VDR mRNA.

RESULTS

Higher VDR expression correlates with longer survival after adjusting for age, sex, disease stage and tumor grade (HR 0.73, 95% CI 0.58-0.91). In addition, there was a positive correlation (r=0.38) between serum 1,25-D(3) and tumor VDR protein expression. A greater anti-proliferative effect of 1,25-D(3) was observed in high compared to low VDR-expressing cell lines; these effects corresponded to G1 cell cycle arrest; this was associated with a decline in cyclin D1, S-phase kinase protein 2 (Skp2), retinoblastoma (Rb) and minichromosome maintenance 2 (MCM2) proteins involved in S-phase entry.

CONCLUSIONS

Increased VDR expression in lung AC is associated with improved survival. This may relate to a lower proliferative status and G1 arrest in high VDR-expressing tumors.