EARLY DIAGNOSIS OF PULMONARY NODULES: BIOMARKER PERFORMANCE IN PATIENTS WITH SOLITARY OR MULTIPLE NODULES IN THE PANOPTIC STUDY

Abstract

SESSION TITLE: Lung Cancer SESSION TYPE: Original Investigation Posters PRESENTED ON: 10/10/2018 01:00 PM - 02:00 PM PURPOSE: Patients can present with a solitary pulmonary nodule, but, often they present with multiple nodules with a specific nodule of concern to the pulmonologist. Recently, a proteomic-based integrated classifier which uses clinical risk factors along with protein abundance from a patient’s blood sample has been shown to perform well as a predictor of the likelihood that a patient has cancer. The large prospective trial (PANOPTIC) done for clinical validation, provided an opportunity to assess the tests performance in solitary and multiple nodule patients. METHODS: The integrated classifier (Xpresys Lung, Integrated Diagnostics, Seattle, WA) is a blood test designed to evaluate patients 40 years old and older with an 8-30mm nodule, integrating the protein levels with 5 clinical factors in those with a pre-test probability of cancer (pCA) being ≤ 50%. One-year results were categorized using an adjudication algorithm that required tissue diagnosis for malignant nodules and included nodule resolution or stability over a defined period to be considered benign. The entire population and the subsets of patients with solitary and multiple nodules were evaluated for test performance. RESULTS: Total patients eligible for analyses were 392 and 178 were in the lower risk group (pCA≤50%) with a cancer prevalence of 16% and mean nodule size of 16.8 ± 0.6 mm. The performance of XL: sensitivity 97%, specificity 44%, NPV 98%. Clinical utility calculations were: reduction in invasive procedures ≥36%, 3% of malignant nodules to surveillance (compared to 45% actual). Nodules were solitary in 77 (43%) patients and 101 (57%) were recorded as having multiple nodules with a mean nodule number of 3.2 (range 1-10). The average age of patients with multiple nodules (67.69 +/- 1.91) was statistically higher than solitary nodule patients (62.68 +/- 2.45) with a p-value of 0.002. Otherwise, there were no statistically significant differences in terms of gender, smoking status, nodule size or cancer prevalence. Similarly, the integrated classifier performance showed no statistically significant difference between the multiple and solitary nodule patients.Table: Table of patients with multiple and solitary nodules All Patients Multiple Nodules Solitary Nodules p valuePatients 178 101 77 Age 65.52(+/- 1.55) 67.69(+/- 1.91) 62.68(+/- 2.45) 0.002Gender Male 95 (53%) 52 (51%) 43 (56%) 0.564 Female 83 (47%) 49 (49%) 34 (44%) Lung Nodule Size 13.95(+/- 0.76) 13.72(+/- 0.96) 14.26(+/- 1.24) 0.493Pathology Cancer 29 (16%) 16 (16%) 13 (17%) Benign 149 (84%) 85 (84%) 64 (83%) 0.852Test Performance Sensitivity 97% 100% 92% Specificity 44% 47% 39% PPV 25% 26% 24% NPV 98% 100% 96% 0.164 CONCLUSIONS: Analysis of PANOPTIC study data does not show any difference in classifier performance for the patients with multiple nodules compared to solitary nodules. CLINICAL IMPLICATIONS: An integrated classifier comprised of a patient's clinical factors and protein measurement may help in determining whether an incidental nodule is benign regardless of nodule count. DISCLOSURES: Employee relationship with Integrated Diagnostics Please note: >$100000 Added 03/05/2018 by Paul Kearney, source=Web Response, value=Salary Employee relationship with Integrated Diagnostics Please note: >$100000 Added 03/02/2018 by alexander porter, source=Web Response, value=Salary Employee relationship with Integrated Diagnostics Please note: >$100000 Added 03/02/2018 by Steven Springmeyer, source=Web Response, value=Salary