Integrated Biomarkers for the Management of Indeterminate Pulmonary Nodules.

Rationale: Patients with indeterminate pulmonary nodules (IPNs) at risk of cancer undergo high rates of invasive, costly, and morbid procedures. Objectives: To train and externally validate a risk prediction model that combined clinical, blood, and imaging biomarkers to improve the noninvasive management of IPNs. Methods: In this prospectively collected, retrospective blinded evaluation study, probability of cancer was calculated for 456 patient nodules using the Mayo Clinic model, and patients were categorized into low-, intermediate-, and high-risk groups. A combined biomarker model (CBM) including clinical variables, serum high sensitivity CYFRA 21-1 level, and a radiomic signature was trained in cohort 1 (n = 170) and validated in cohorts 2-4 (total n = 286). All patients were pooled to recalibrate the model for clinical implementation. The clinical utility of the CBM compared with current clinical care was evaluated in 2 cohorts. Measurements and Main Results: The CBM provided improved diagnostic accuracy over the Mayo Clinic model with an improvement in area under the curve of 0.124 (95% bootstrap confidence interval, 0.091-0.156; P < 2 × 10-16). Applying 10% and 70% risk thresholds resulted in a bias-corrected clinical reclassification index for cases and control subjects of 0.15 and 0.12, respectively. A clinical utility analysis of patient medical records estimated that a CBM-guided strategy would have reduced invasive procedures from 62.9% to 50.6% in the intermediate-risk benign population and shortened the median time to diagnosis of cancer from 60 to 21 days in intermediate-risk cancers. Conclusions: Integration of clinical, blood, and image biomarkers improves noninvasive diagnosis of patients with IPNs, potentially reducing the rate of unnecessary invasive procedures while shortening the time to diagnosis.

Long-term weight gain among Hispanic and non-Hispanic White women with and without breast cancer.

Weight gain following breast cancer diagnosis is common, but limited data exists on whether this gain is in excess of that gained during normal aging. This study investigated weight patterns among women with and without breast cancer to determine the effects of the breast cancer experience on weight change. Using the SHINE 4-Corners Breast Cancer Study, 305 women with breast cancer and 345 women without were followed prospectively. Weight change of ≥5% was defined as the difference between the self-reported weight measurements obtained at breast cancer diagnosis (or referent date for women without breast cancer) and about 6 yr later. Multiple logistic regression analyses were used. Within this cohort, 60% of women were overweight or obese and 37% of women gained weight. No significant greater weight gain was observed between women with vs. without breast cancer [adjusted odds ratio (ORadj) = 1.15, 95% CI 0.79-1.68] or between Hispanic vs. non-Hispanic White women (ORadj = 1.09, 95% CI 0.72-1.66) after adjustment. Weight gain was associated with being younger and having a lower body mass index. Among breast cancer survivors, cancer treatment factors were not associated with weight gain. These results suggest that weight management approaches are needed, especially those targeted to at-risk populations such as breast cancer survivors.

Differences in VA and Non-VA Pulmonary Nodules: All Evaluations Are not Created Equal.

BACKGROUND: Indeterminate pulmonary nodules present a common challenge for clinicians who must recommend surveillance or intervention based on an assessed risk of malignancy. PATIENTS AND METHODS: In this cohort study, patients presenting for indeterminate pulmonary nodule evaluation were enrolled at sites participating in the Colorado SPORE in Lung Cancer. They were followed prospectively and included for analysis if they had a definitive malignant diagnosis, benign diagnosis, or radiographic resolution or stability of their nodule for > 2 years. RESULTS: Patients evaluated at the Veterans Affairs (VA) and non-VA sites were equally as likely to have a malignant diagnosis (48%). The VA cohort represented a higher-risk group than the non-VA cohort regarding smoking history and chronic obstructive pulmonary disease (COPD). There were more squamous cell carcinoma diagnoses among VA malignant nodules (25% vs. 10%) and a later stage at diagnosis among VA patients. Discrimination and calibration of risk calculators produced estimates that were wide-ranging and different when comparing between risk score calculators as well as between VA/non-VA cohorts. Application of current American College of Chest Physicians guidelines to our groups could have resulted in inappropriate resection of 12% of benign nodules. CONCLUSION: Comparison of VA with non-VA patients shows important differences in underlying risk, histology of malignant nodules, and stage at diagnosis. This study highlights the challenge in applying risk calculators to a clinical setting, as the model discrimination and calibration were variable between calculators and between our higher-risk VA and lower-risk non-VA groups. MICROABSTRACT: Risk stratification and management of indeterminate pulmonary nodules (IPNs) is a common clinical problem. In this prospective cohort study of 282 patients with IPNs from Veterans Affairs (VA) and non-VA sites, we found differences in patient and nodule characteristics, histology and diagnostic stage, and risk calculator performance. Our findings highlight challenges and shortcomings of current IPN management guidelines and tools.

Nurse-led triage of new sleep referrals is associated with lower risk of potentially contraindicated sleep testing: a retrospective cohort study.

BACKGROUND: The volume of specialty care referrals often outstrips specialists' capacity. The Department of Veterans Affairs launched a system of referral coordination to augment our workforce, empowering registered nurses to use decision support tools to triage specialty referrals. While task shifting may improve access, there is limited evidence regarding the relative quality of nurses' triage decisions to ensure such management is safe. OBJECTIVE: Within the specialty of sleep medicine, we compared receipt of contraindicated testing for obstructive sleep apnoea (OSA) between patients triaged to sleep testing by nurses in the referral coordination system (RCS) relative to our traditional specialist-led system (TSS). METHODS: Patients referred for OSA evaluation can be triaged to either home sleep apnoea testing (HSAT) or polysomnography, and existing guidelines specify patients for whom HSAT is contraindicated. In RCS, nurses used a decision support tool to make triage decisions for sleep testing but were instructed to seek specialist oversight in complex cases. In TSS, specialists made triage decisions themselves. We performed a single-centre retrospective cohort study of patients without OSA who were referred to sleep testing between September 2018 and August 2019. Patients were assigned to triage by RCS or TSS in quasirandom fashion based on triager availability at time of referral. We compared receipt of contraindicated sleep tests between groups using a generalised linear model adjusted for day of the week and time of day of referral. RESULTS: RCS triaged 793 referrals for OSA evaluation relative to 1787 by TSS. Patients with RCS triages were at lower risk of receiving potentially contraindicated sleep tests relative risk 0.52 (95% CI 0.29 to 0.93). CONCLUSION: Our results suggest that incorporating registered nurses into triage decision-making may improve the quality of diagnostic care for OSA.

Quantifying the incremental value of deep learning: Application to lung nodule detection.

We present a case study for implementing a machine learning algorithm with an incremental value framework in the domain of lung cancer research. Machine learning methods have often been shown to be competitive with prediction models in some domains; however, implementation of these methods is in early development. Often these methods are only directly compared to existing methods; here we present a framework for assessing the value of a machine learning model by assessing the incremental value. We developed a machine learning model to identify and classify lung nodules and assessed the incremental value added to existing risk prediction models. Multiple external datasets were used for validation. We found that our image model, trained on a dataset from The Cancer Imaging Archive (TCIA), improves upon existing models that are restricted to patient characteristics, but it was inconclusive about whether it improves on models that consider nodule features. Another interesting finding is the variable performance on different datasets, suggesting population generalization with machine learning models may be more challenging than is often considered.

Clinical Utility of Chromosomal Aneusomy in Individuals at High Risk of Lung Cancer.

INTRODUCTION: Low-dose computed tomography screening for lung cancer has a high false-positive rate with frequent discovery of indeterminate pulmonary nodules. Noninvasive biomarkers are needed to reduce false positives and improve risk stratification. A retrospective longitudinal evaluation was performed to assess chromosomal aneusomy in sputum by fluorescence in situ hybridization (CA-FISH) in four nested case-control studies. METHODS: Receiver operating characteristic analysis resulted in two grouped cohorts: a high-risk cohort (Colorado High-Risk Cohort and Colorado Nodule Cohort [68 case patients and 69 controls]) and a screening cohort (American College of Radiology Imaging Network/National Lung Screening Trial and Pittsburgh Lung Screening Study [97 case patients and 185 controls]). The CA-FISH assay was a four-target DNA panel encompassing the EGFR and v-myc avian myelocytomatosis viral oncogene homolog (MYC) genes, and the 5p15 and centromere 6 regions or the fibroblast growth factor 1 gene (FGFR1) and phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha gene (PIK3CA). A four-category scale (normal, probably normal, probably abnormal, and abnormal) was applied. Sensitivity, specificity, and positive and negative likelihood ratios (LRs) (with 95% confidence intervals [CIs]) were estimated for each cohort. RESULTS: Sensitivity and specificity were, respectively, 0.67 (95% CI: 0.55-0.78) and 0.94 (95% CI: 0.85-0.98) for high-risk participants and 0.20 (95% CI: 0.13-0.30) and 0.84 (95% CI: 0.78-0.89) for screening participants. The positive and negative LRs were, respectively, 11.66 (95% CI: 4.44-30.63) and 0.34 (95% CI: 0.24-0.48) for high-risk participants and 1.36 (95% CI: 0.81-2.28) and 0.93 (95% CI: 0.83-1.05) for screening participants. CONCLUSION: The high positive LR of sputum CA-FISH indicates that it could be a useful adjunct to low-dose computed tomography for lung cancer in high-risk settings. For screening, however, its low positive LR limits clinical utility. Prospective assessment of CA-FISH in the incidentally identified indeterminate nodule setting is ongoing in the Colorado Pulmonary Nodule Biomarker Trial.

Endobronchial miRNAs as biomarkers in lung cancer chemoprevention.

Lung cancers express lower levels of prostacyclin than normal lung tissues. Prostacyclin prevents lung cancer in a variety of mouse models. A randomized phase II trial comparing oral iloprost (a prostacyclin analog) with placebo in high-risk subjects showed improvement in bronchial histology in former, but not current, smokers. This placebo-controlled study offered the opportunity for investigation of other potential intermediate endpoint and predictive biomarkers to incorporate into chemoprevention trials. Matched bronchial biopsies were obtained at baseline and at 6-month follow-up from 125 high-risk individuals who completed the trial: 31/29 and 37/28 current/former smokers in the iloprost and placebo arm, respectively. We analyzed the expression of 14 selected miRNAs by Real Time PCR in 496 biopsies. The expression of seven miRNAs was significantly correlated with histology at baseline. The expression of miR-34c was inversely correlated with histology at baseline (P < 0.0001) and with change in histology at follow-up (P = 0.0003), independent of treatment or smoking status. Several miRNAs were also found to be differentially expressed in current smokers as compared with former smokers. In current smokers, miR-375 was upregulated at baseline (P < 0.0001) and downregulated after treatment with iloprost (P = 0.0023). No miRNA at baseline reliably predicted a response to iloprost. No biomarker predictive of response to iloprost was found. MiR-34c was inversely correlated with baseline histology and with histology changes. Mir-34c changes at follow-up could be used as a quantitative biomarker that parallels histologic response in formalin-fixed bronchial biopsies in future lung cancer chemoprevention studies.

Loss of transforming growth factor beta type II receptor increases aggressive tumor behavior and reduces survival in lung adenocarcinoma and squamous cell carcinoma.

PURPOSE: Lung adenocarcinoma and lung squamous cell carcinoma (SCC) are the most common non-small cell lung cancer (NSCLC) subtypes. This study was designed to determine whether reduced expression of TGFß type II receptor (TGFßRII) promotes lung adenocarcinoma and SCC carcinogenesis. EXPERIMENTAL DESIGN: We examined TGFßRII expression at the protein and mRNA levels in human NSCLC samples and assessed the relationship between TGFßRII expression and clinicopathologic parameters. To determine whether sporadic TGFßRII deletion in airway epithelial cells induces NSCLC formation, we targeted TGFßRII deletion alone and in combination with oncogenic Kras(G12D) to murine airways using a keratin 5 (K5) promoter and inducible Cre recombinase. RESULTS: Reduced TGFßRII expression in human NSCLC is associated with male gender, smoking, SCC histology, reduced differentiation, increased tumor stage, increased nodal metastasis, and reduced survival. Homozygous or heterozygous TGFßRII deletion in mouse airway epithelia increases the size and number of Kras(G12D)-initiated adenocarcinoma and SCC. TGFßRII deletion increases proliferation, local inflammation, and TGFß ligand elaboration; TGFßRII knockdown in airway epithelial cells increases migration and invasion. CONCLUSIONS: Reduced TGFßRII expression in human NSCLC is associated with more aggressive tumor behavior and inflammation that is, at least partially, mediated by increased TGFß1 expression. TGFßRII deletion in mouse airway epithelial cells promotes adenocarcinoma and SCC formation, indicating that TGFßRII loss plays a causal role in lung carcinogenesis. That TGFßRII shows haploid insufficiency suggests that a 50% TGFßRII protein reduction would negatively impact lung cancer prognosis.

Defining a gene promoter methylation signature in sputum for lung cancer risk assessment.

PURPOSE: To evaluate the methylation state of 31 genes in sputum as biomarkers in an expanded nested, case-control study from the Colorado cohort, and to assess the replication of results from the most promising genes in an independent case-control study of asymptomatic patients with stage I lung cancer from New Mexico. EXPERIMENTAL DESIGN: Cases and controls from Colorado and New Mexico were interrogated for methylation of up to 31 genes using nested, methylation-specific PCR. Individual genes and methylation indices were used to assess the association between methylation and lung cancer with logistic regression modeling. RESULTS: Seventeen genes with ORs of 1.4 to 3.6 were identified and selected for replication in the New Mexico study. Overall, the direction of effects seen in New Mexico was similar to Colorado with the largest increase in case discrimination (ORs, 3.2-4.2) seen for the PAX5α, GATA5, and SULF2 genes. Receiver operating characteristic (ROC) curves generated from seven-gene panels from Colorado and New Mexico studies showed prediction accuracy of 71% and 77%, respectively. A 22-fold increase in lung cancer risk was seen for a subset of New Mexico cases with five or more genes methylated. Sequence variants associated with lung cancer did not improve the accuracy of this gene methylation panel. CONCLUSIONS: These studies have identified and replicated a panel of methylated genes whose integration with other promising biomarkers could initially identify the highest risk smokers for computed tomographic screening for early detection of lung cancer.

Genome-wide significant association between a sequence variant at 15q15.2 and lung cancer risk.

Genome-wide association studies (GWAS) have identified 3 genomic regions, at 15q24-25.1, 5p15.33, and 6p21.33, which associate with the risk of lung cancer. Large meta-analyses of GWA data have failed to find additional associations of genome-wide significance. In this study, we sought to confirm 7 variants with suggestive association to lung cancer (P < 10(-5)) in a recently published meta-analysis. In a GWA dataset of 1,447 lung cancer cases and 36,256 controls in Iceland, 3 correlated variants on 15q15.2 (rs504417, rs11853991, and rs748404) showed a significant association with lung cancer, whereas rs4254535 on 2p14, rs1530057 on 3p24.1, rs6438347 on 3q13.31, and rs1926203 on 10q23.31 did not. The most significant variant, rs748404, was genotyped in an additional 1,299 lung cancer cases and 4,102 controls from the Netherlands, Spain, and the United States and the results combined with published GWAS data. In this analysis, the T allele of rs748404 reached genome-wide significance (OR = 1.15, P = 1.1 × 10(-9)). Another variant at the same locus, rs12050604, showed association with lung cancer (OR = 1.09, 3.6 × 10(-6)) and remained significant after adjustment for rs748404 and vice versa. rs748404 is located 140 kb centromeric of the TP53BP1 gene that has been implicated in lung cancer risk. Two fully correlated, nonsynonymous coding variants in TP53BP1, rs2602141 (Q1136K) and rs560191 (E353D) showed association with lung cancer in our sample set; however, this association did not remain significant after adjustment for rs748404. Our data show that 1 or more lung cancer risk variants of genome-wide significance and distinct from the coding variants in TP53BP1 are located at 15q15.2.

The detection of chromosomal aneusomy by fluorescence in situ hybridization in sputum predicts lung cancer incidence.

Lung cancer usually is disseminated (advanced) and has a poor prognosis at diagnosis. Current and former smokers are at a high risk for lung cancer and are candidates for prevention and early detection strategies. Sputum is a potential source of biomarkers that might determine either lung cancer risk or the presence of early lung cancer, but no current sputum test is sufficiently sensitive and specific for effective screening. We used fluorescence in situ hybridization (FISH) to measure chromosomal aneusomy (CA) in sputum samples collected prospectively from 100 incident lung cancer cases and 96 controls (matched on age, gender, and date of collection) nested within an ongoing high-risk cohort. The CA-FISH assay was aimed at four DNA targets: epidermal growth factor receptor, MYC, 5p15, and CEP 6. The sensitivity of a positive CA-FISH assay (abnormal for two or more of the four markers) for lung cancer was substantially higher for samples collected within 18 months (76% sensitivity) than for samples collected more than 18 months (31%) before lung cancer diagnosis. Sensitivity was higher for squamous cell cancers (94%) than for other histologic types (69%). CA-FISH specificity based on samples collected within 18 months before diagnosis was 88%. The adjusted odds ratio (OR) of lung cancer for specimens collected within 18 months before a cancer diagnosis was higher for the CA-FISH assay [OR, 29.9; 95% confidence interval (95% CI), 9.5-94.1] than for previously studied ORs of cytologic atypia (OR, 1.8; 95% CI, 1.3-2.6) and gene promoter methylation (OR, 6.5; 95% CI, 1.2-35.5). Whether CA-FISH is an indicator of extreme risk for incident lung cancer or detects exfoliated cancer cells is unknown. The apparent promise of CA-FISH in sputum for assessing lung cancer risk and/or for lung cancer early detection now needs to be validated in a clinical screening trial.

CALCULATING AVERAGE POWER FOR THE BENJAMINI-HOCHBERG PROCEDURE.

Using exact, analytic results for the average power of the Benjamini-Hochberg (1995) procedure, we provide example power analyses useful for scientists planning studies involving multiple comparisons. The power results are based on the distribution of the p-value under the alternative for the Pearson's χ2, and for the Hotelling-Lawley trace, the Wilks' lambda, and the Pillai-Bartlett trace, all tests for the general linear multivariate model. Detailed example power analyses are given for a planned mammography experiment with categorical data and a study that tests the association of a single nucleotide polymorphism with insulin resistance and visceral adiposity.