ABSTRACT
BACKGROUND: Accurate assessment of the probability of lung cancer (pCA) is critical in patients with pulmonary nodules (PNs) to help guide decision-making. We sought to validate a clinical-genomic classifier developed using whole-transcriptome sequencing of nasal epithelial cells from patients with a PN ≤ 30 mm who smoke or have previously smoked. RESEARCH QUESTION: Can the pCA in individuals with a PN and a history of smoking be predicted by a classifier that uses clinical factors and genomic data from nasal epithelial cells obtained by cytologic brushing? STUDY DESIGN AND METHODS: Machine learning was used to train a classifier using genomic and clinical features on 1,120 patients with PNs labeled as benign or malignant established by a final diagnosis or a minimum of 12 months of radiographic surveillance. The classifier was designed to yield low-, intermediate-, and high-risk categories. The classifier was validated in an independent set of 312 patients, including 63 patients with a prior history of cancer (other than lung cancer), comparing the classifier prediction with the known clinical outcome. RESULTS: In the primary validation set, sensitivity and specificity for low-risk classification were 96% and 42%, whereas sensitivity and specificity for high-risk classification was 58% and 90%, respectively. Sensitivity was similar across stages of non-small cell lung cancer, independent of subtype. Performance compared favorably with clinical-only risk models. Analysis of 63 patients with prior cancer showed similar performance as did subanalyses of patients with light vs heavy smoking burden and those eligible for lung cancer screening vs those who were not. INTERPRETATION: The nasal classifier provides an accurate assessment of pCA in individuals with a PN ≤ 30 mm who smoke or have previously smoked. Classifier-guided decision-making could lead to fewer diagnostic procedures in patients without cancer and more timely treatment in patients with lung cancer.
Subject(s)
Carcinoma, Non-Small-Cell Lung , Lung Neoplasms , Multiple Pulmonary Nodules , Humans , Lung Neoplasms/pathology , Carcinoma, Non-Small-Cell Lung/diagnosis , Early Detection of Cancer , Multiple Pulmonary Nodules/diagnosis , Multiple Pulmonary Nodules/pathology , ProbabilityABSTRACT
BACKGROUND: Bronchoscopy is commonly utilized for non-surgical sampling of indeterminant pulmonary lesions, but nondiagnostic procedures are common. Accurate assessment of the risk of malignancy is essential for decision making in these patients, yet we lack tools that perform well across this heterogeneous group of patients. We sought to evaluate the accuracy of three previously validated risk models and physician-assessed risk (PAR) in patients with a newly identified lung lesion undergoing bronchoscopy for suspected lung cancer where the result is nondiagnostic. METHODS: We performed an analysis of prospective data collected for the Percepta Bronchial Genomic Classifier Multicenter Registry. PAR and three previously validated risk models (Mayo Clinic, Veteran's Affairs, and Brock) were used to determine the probability of lung cancer (low, intermediate, or high) in 375 patients with pulmonary lesions who underwent bronchoscopy for possible lung cancer with nondiagnostic pathology. Results were compared to the actual adjudicated prevalence of malignancy in each pre-test risk group, determined with a minimum of 12 months follow up after bronchoscopy. RESULTS: PAR and the risk models performed poorly overall in the assessment of risk in this patient population. PAR most closely matched the observed prevalence of malignancy in patients at 12 months after bronchoscopy, but all modalities had a low area under the curve, and in all clinical models more than half of all the lesions labeled as high risk were truly or likely benign. The studied risk model calculators overestimate the risk of malignancy compared to PAR, particularly in the subset in older patients, irregularly bordered nodules, and masses > 3 cm. Overall, the risk models perform only slightly better when confined to lung nodules < 3 cm in this population. CONCLUSION: The currently available tools for the assessment of risk of malignancy perform suboptimally in patients with nondiagnostic findings following a bronchoscopic evaluation for lung cancer. More accurate and objective tools for risk assessment are needed. TRIAL REGISTRATION: not applicable.
Subject(s)
Bronchoscopy , Lung Neoplasms , Humans , Aged , Bronchoscopy/methods , Prospective Studies , Lung/pathology , Lung Neoplasms/pathology , Risk AssessmentABSTRACT
INTRODUCTION: Bronchoscopic sampling of pulmonary lesions suspicious for lung cancer is frequently nondiagnostic. A genomic sequencing classifier utilizing bronchial brushings obtained at the time of the bronchoscopy has been shown to provide an accurate reclassification of the risk of malignancy (ROM) based on pre-procedure risk. Our objectives for this study were to determine the frequency with which the classifier up- or down-classifies risk in regular clinical practice and to model the potential clinical utility of that reclassification. METHODS: This observational study retrospectively assessed data from four clinical sites that regularly use the genomic classifier in the bronchoscopic evaluation of indeterminate lesions. Demographics and pre-bronchoscopy ROM were recorded. The frequency of up- and down-classification was calculated. Modeling based on reclassification rates and the performance characteristics of the classifier was performed to demonstrate the potential clinical utility of the result. RESULTS: 86 patients who underwent classifier testing following a nondiagnostic bronchoscopy were included. 45% of patients with high ROM prior to bronchoscopy were reclassified very high-risk. 38% of patients with intermediate ROM were up-or down-classified. 56% of patients with low ROM were reclassified to very low-risk. Overall, 42% of patients had a change in classification. 35% of the study cohort could potentially have avoided additional unnecessary procedures with subsequent guideline-adherent management. CONCLUSIONS: The classifier can guide decision-making following a nondiagnostic bronchoscopy, reclassifying risk in a significant percentage of cases. Use of the classifier should allow more patients with early-stage cancer to proceed directly to curative therapy while helping more patients with benign disease avoid further unnecessary procedures.
Subject(s)
Bronchoscopy , Lung Neoplasms , Humans , Bronchoscopy/methods , Retrospective Studies , Lung Neoplasms/diagnosis , Lung Neoplasms/genetics , Lung Neoplasms/pathology , Genomics/methods , Lung/pathologyABSTRACT
The Percepta Genomic Sequencing Classifier (GSC) was developed to up-classify as well as down-classify the risk of malignancy for lung lesions when bronchoscopy is non-diagnostic. We evaluated the performance of Percepta GSC in risk re-classification of indeterminate lung lesions. This multicenter study included individuals who currently or formerly smoked undergoing bronchoscopy for suspected lung cancer from the AEGIS I/ II cohorts and the Percepta Registry. The classifier was measured in normal-appearing bronchial epithelium from bronchial brushings. The sensitivity, specificity, and predictive values were calculated using predefined thresholds. The ability of the classifier to decrease unnecessary invasive procedures was estimated. A set of 412 patients were included in the validation (prevalence of malignancy was 39.6%). Overall, 29% of intermediate-risk lung lesions were down-classified to low-risk with a 91.0% negative predictive value (NPV) and 12.2% of intermediate-risk lesions were up-classified to high-risk with a 65.4% positive predictive value (PPV). In addition, 54.5% of low-risk lesions were down-classified to very low risk with >99% NPV and 27.3% of high-risk lesions were up-classified to very high risk with a 91.5% PPV. If the classifier results were used in nodule management, 50% of patients with benign lesions and 29% of patients with malignant lesions undergoing additional invasive procedures could have avoided these procedures. The Percepta GSC is highly accurate as both a rule-out and rule-in test. This high accuracy of risk re-classification may lead to improved management of lung lesions.
Subject(s)
Bronchoscopy , Lung Neoplasms , Biopsy , Bronchoscopy/methods , Chromosome Mapping , Humans , Lung Neoplasms/diagnosis , Lung Neoplasms/genetics , Respiratory MucosaABSTRACT
Rationale: The diagnosis of idiopathic pulmonary fibrosis (IPF) remains challenging and can result in delayed or misdiagnosis. IPF diagnosis is based on the presence of either a radiographic or histologic usual interstitial pneumonia (UIP) pattern in the absence of an identifiable etiology. The Envisia Genomic Classifier is a clinically validated molecular diagnostic test that identifies UIP in transbronchial biopsies. Objectives: To determine the impact of the Envisia Genomic Classifier on physicians' clinical decision-making in the diagnosis and management of IPF. Methods: This prospective randomized decision impact survey was designed to test the hypothesis that including an Envisia UIP-positive result will increase IPF diagnoses, diagnostic confidence, and the recommendation for antifibrotic therapy. The survey included patients from the BRAVE (Bronchial Sample Collection for a Novel Genomic Test) study who had a high-resolution computed tomographic scan without a typical UIP pattern, an Envisia UIP-positive result, and a final diagnosis of IPF by multidisciplinary team discussion. Each case was presented in three different formats: a pre-post cohort, where each case is presented initially without and then with Envisia, and two independent cohorts, where each case is presented without and with Envisia, respectively. Results: U.S.-based pulmonologists from community and academic centers in geographically diverse practices were approached for inclusion in this study. 103 (65%) U.S.-based pulmonologists met the inclusion criteria and provided 605 case reviews of 11 patient cases. The number of IPF diagnoses increased with Envisia by an absolute difference of 39% from 47 (30%) before Envisia to 107 (69%) after Envisia in the pre-post cohort and by 13% in the independent cohorts. High confidence (⩾90%) of interstitial lung disease diagnoses was more commonly seen with Envisia in both the pre-post cohort and in the independent cohorts. Recommendation for antifibrotic treatment increased with Envisia by an absolute difference of 36% from 15 (10%) before Envisia to 72 (46.4%) after Envisia in the pre-post cohort and by 11% in the independent cohorts. Conclusions: This decision impact survey suggests the clinical utility of the Envisia Classifier by demonstrating a significant increase in IPF diagnoses, diagnostic confidence, and recommendation for antifibrotic therapies to assist physicians in effectively managing patients to improve outcomes of patients with IPF.
Subject(s)
Idiopathic Pulmonary Fibrosis , Lung Diseases, Interstitial , Biopsy/methods , Genomics/methods , Humans , Idiopathic Pulmonary Fibrosis/diagnosis , Idiopathic Pulmonary Fibrosis/genetics , Idiopathic Pulmonary Fibrosis/therapy , Lung/diagnostic imaging , Lung/pathology , Lung Diseases, Interstitial/diagnosis , Prospective StudiesABSTRACT
Rationale: Usual interstitial pneumonia (UIP) is the defining morphology of idiopathic pulmonary fibrosis (IPF). Guidelines for IPF diagnosis conditionally recommend surgical lung biopsy for histopathology diagnosis of UIP when radiology and clinical context are not definitive. A "molecular diagnosis of UIP" in transbronchial lung biopsy, the Envisia Genomic Classifier, accurately predicted histopathologic UIP.Objectives: We evaluated the combined accuracy of the Envisia Genomic Classifier and local radiology in the detection of UIP pattern.Methods: Ninety-six patients who had diagnostic lung pathology as well as a transbronchial lung biopsy for molecular testing with Envisia Genomic Classifier were included in this analysis. The classifier results were scored against reference pathology. UIP identified on high-resolution computed tomography (HRCT) as documented by features in local radiologists' reports was compared with histopathology.Measurements and Main Results: In 96 patients, the Envisia Classifier achieved a specificity of 92.1% (confidence interval [CI],78.6-98.3%) and a sensitivity of 60.3% (CI, 46.6-73.0%) for histology-proven UIP pattern. Local radiologists identified UIP in 18 of 53 patients with UIP histopathology, with a sensitivity of 34.0% (CI, 21.5-48.3%) and a specificity of 96.9% (CI, 83.8-100%). In conjunction with HRCT patterns of UIP, the Envisia Classifier results identified 24 additional patients with UIP (sensitivity 79.2%; specificity 90.6%).Conclusions: In 96 patients with suspected interstitial lung disease, the Envisia Genomic Classifier identified UIP regardless of HRCT pattern. These results suggest that recognition of a UIP pattern by the Envisia Genomic Classifier combined with HRCT and clinical factors in a multidisciplinary discussion may assist clinicians in making an interstitial lung disease (especially IPF) diagnosis without the need for a surgical lung biopsy.
Subject(s)
Genomics/methods , Idiopathic Pulmonary Fibrosis/diagnosis , Idiopathic Pulmonary Fibrosis/genetics , Tomography, X-Ray Computed , Adult , Aged , Aged, 80 and over , Female , Genetic Markers , Humans , Idiopathic Pulmonary Fibrosis/classification , Male , Middle Aged , Prospective Studies , Reproducibility of Results , Sensitivity and Specificity , Tomography, X-Ray Computed/methodsABSTRACT
BACKGROUND: The Percepta genomic classifier has been clinically validated as a complement to bronchoscopy for lung nodule evaluation. RESEARCH QUESTION: The goal of this study was to examine the impact on clinical management decisions of the Percepta result in patients with low- and intermediate-risk lung nodules. STUDY DESIGN AND METHODS: A prospective "real world" registry was instituted across 35 US centers to observe physician management of pulmonary nodules following a nondiagnostic bronchoscopy. To assess the impact on management decisions of the Percepta genomic classifier, a subset of patients was analyzed who had an inconclusive bronchoscopy for a pulmonary nodule, a Percepta result, and an adjudicated lung diagnosis with at least 1 year of follow-up. In this cohort, change in the decision to pursue additional invasive procedures following Percepta results was assessed. RESULTS: A total of 283 patients met the study eligibility criteria. In patients with a low/intermediate risk of malignancy for whom the clinician had designated a plan for a subsequent invasive procedure, a negative Percepta result down-classified the risk of malignancy in 34.3% of cases. Of these down-classified patients, 73.9% had a change in their management plan from an invasive procedure to surveillance, and the majority avoided a procedure up to 12 months following the initial evaluation. In patients with confirmed lung cancers, the time to diagnosis was not significantly delayed when comparing Percepta down-classified patients vs patients who were not down-classified (P = .58). INTERPRETATION: The down-classification of nodule malignancy risk with the Percepta test decreased additional invasive procedures without a delay in time to diagnosis among those with lung cancer.
Subject(s)
Clinical Decision-Making , Genomics , Lung Neoplasms/diagnosis , Solitary Pulmonary Nodule/diagnosis , Aged , Bronchoscopy , Female , Genetic Markers , Humans , Lung Neoplasms/genetics , Lung Neoplasms/therapy , Male , Middle Aged , Patient Selection , Prospective Studies , Registries , Solitary Pulmonary Nodule/genetics , Solitary Pulmonary Nodule/therapy , United StatesABSTRACT
BACKGROUND: In the appropriate clinical setting, the diagnosis of idiopathic pulmonary fibrosis (IPF) requires a pattern of usual interstitial pneumonia to be present on high-resolution chest CT (HRCT) or surgical lung biopsy. A molecular usual interstitial pneumonia signature can be identified by a machine learning algorithm in less-invasive transbronchial lung biopsy samples. We report prospective findings for the clinical validity and utility of this molecular test. METHODS: We prospectively recruited 237 patients for this study from those enrolled in the Bronchial Sample Collection for a Novel Genomic Test (BRAVE) study in 29 US and European sites. Patients were undergoing evaluation for interstitial lung disease and had had samples obtained by clinically indicated surgical or transbronchial biopsy or cryobiopsy for pathology. Histopathological diagnoses were made by experienced pathologists. Available HRCT scans were reviewed centrally. Three to five transbronchial lung biopsy samples were collected from all patients specifically for this study, pooled by patient, and extracted for transcriptomic sequencing. After exclusions, diagnostic histopathology and RNA sequence data from 90 patients were used to train a machine learning algorithm (Envisia Genomic Classifier, Veracyte, San Francisco, CA, USA) to identify a usual interstitial pneumonia pattern. The primary study endpoint was validation of the classifier in 49 patients by comparison with diagnostic histopathology. To assess clinical utility, we compared the agreement and confidence level of diagnosis made by central multidisciplinary teams based on anonymised clinical information and radiology results plus either molecular classifier or histopathology results. FINDINGS: The classifier identified usual interstitial pneumonia in transbronchial lung biopsy samples from 49 patients with 88% specificity (95% CI 70-98) and 70% sensitivity (47-87). Among 42 of these patients who had possible or inconsistent usual interstitial pneumonia on HRCT, the classifier showed 81% positive predictive value (95% CI 54-96) for underlying biopsy-proven usual interstitial pneumonia. In the clinical utility analysis, we found 86% agreement (95% CI 78-92) between clinical diagnoses using classifier results and those using histopathology data. Diagnostic confidence was improved by the molecular classifier results compared with histopathology results in 18 with IPF diagnoses (proportion of diagnoses that were confident or provisional with high confidence 89% vs 56%, p=0·0339) and in all 48 patients with non-diagnostic pathology or non-classifiable fibrosis histopathology (63% vs 42%, p=0·0412). INTERPRETATION: The molecular test provided an objective method to aid clinicians and multidisciplinary teams in ascertaining a diagnosis of IPF, particularly for patients without a clear radiological diagnosis, in samples that can be obtained by a less invasive method. Further prospective clinical validation and utility studies are planned. FUNDING: Veracyte.
Subject(s)
Algorithms , Biopsy/statistics & numerical data , Idiopathic Pulmonary Fibrosis/diagnosis , Machine Learning/statistics & numerical data , Tomography, X-Ray Computed/statistics & numerical data , Aged , Biopsy/methods , Diagnosis, Differential , Female , Humans , Lung/diagnostic imaging , Lung/pathology , Male , Middle Aged , Predictive Value of Tests , Prospective Studies , Reproducibility of Results , Sensitivity and Specificity , Tomography, X-Ray Computed/methodsABSTRACT
BACKGROUND: AMH is widely used for assessing ovarian reserve, and it is particularly convenient, because it is thought to have minimal variability throughout the menstrual cycle. However, studies assessing the stability of AMH over the menstrual cycle have been conflicting. PURPOSE: The purpose of this study is to determine whether AMH levels vary across the normal menstrual cycle. DESIGN: A multi-center, prospective cohort study conducted at three US centers. METHODS: Fifty females with regular menstrual cycles aged 18-45 underwent serial venipuncture every 3-5 days starting in the early follicular phase and lasting up to 10 collections. AMH was tested using the Access 2 immunoassay system. RESULTS: Age-adjusted mixed-effect models utilizing data from 384 samples from 50 subjects demonstrated a within subject standard deviation of 0.81 (95% CI 0.75-0.88) with a coefficient of variation of 23.8% across the menstrual cycle and between subject standard deviation of 2.56 (95% CI 2.13-3.21) with a coefficient of variation of 75.1%. Intra-class correlation (ICC) of AMH across the menstrual cycle was 0.91. CONCLUSION: Overall, AMH levels, using the automated Access AMH assay, appear to be relatively stable across the menstrual cycle. Fluctuations, if any, appear to be small, and therefore, clinicians may advise patients to have AMH levels drawn at any time in the cycle.
Subject(s)
Anti-Mullerian Hormone/blood , Menstrual Cycle/physiology , Adolescent , Adult , Cohort Studies , Female , Humans , Middle Aged , Young AdultABSTRACT
BACKGROUND: Deciding when to biopsy a man with non-suspicious DRE findings and tPSA in the 4-10 ng/ml range can be challenging, because two-thirds of such biopsies are typically found to be benign. The Prostate Health Index (phi) exhibits significantly improved diagnostic accuracy for prostate cancer detection when compared to tPSA and %fPSA, however only one published study to date has investigated its impact on biopsy decisions in clinical practice. METHODS: An IRB approved observational study was conducted at four large urology group practices using a physician reported two-part questionnaire. Physician recommendations were recorded before and after receiving the phi test result. A historical control group was queried from each site's electronic medical records for eligible men who were seen by the same participating urologists prior to the implementation of the phi test in their practice. 506 men receiving a phi test were prospectively enrolled and 683 men were identified for the historical control group (without phi). Biopsy and pathological findings were also recorded for both groups. RESULTS: Men receiving a phi test showed a significant reduction in biopsy procedures performed when compared to the historical control group (36.4% vs. 60.3%, respectively, P < 0.0001). Based on questionnaire responses, the phi score impacted the physician's patient management plan in 73% of cases, including biopsy deferrals when the phi score was low, and decisions to perform biopsies when the phi score indicated an intermediate or high probability of prostate cancer (phi ≥36). CONCLUSIONS: phi testing significantly impacted the physician's biopsy decision for men with tPSA in the 4-10 ng/ml range and non-suspicious DRE findings. Appropriate utilization of phi resulted in a significant reduction in biopsy procedures performed compared to historical patients seen by the same participating urologists who would have met enrollment eligibility but did not receive a phi test.