LUNG NODULE INTEGRATED CLASSIFIER BIOMARKER: FIRST DATA WITH REAL-WORLD CLINICAL USE

Abstract

SESSION TITLE: Lung Cancer Posters SESSION TYPE: Original Investigation Posters PRESENTED ON: October 18-21, 2020 PURPOSE: Predicting cancer risk (pCA) of lung nodules presents a challenge. Many nodules are either over or under investigated leading to unnecessary procedures or delays in diagnosis. A clinical validation study (PANOPTIC, CHEST 2018) predicted that a blood-based biomarker, when used for 8-30 mm nodules with a pCA ≤50%, could potentially reduce invasive procedures on benign nodules by up to 40% with a false negative rate of 3%. Here we present clinical use (CU) experience with the integrated classifier (IC) with a focus on the changes in classification of pCA from ACCP category moderate risk (pCA 5-65%) to very low risk (pCA <5%) to determine if the clinical test use appears to be generalizable. METHODS: The CU group consists of consecutive tests (n=1000) recently ordered by 308 US physicians. Clinical characteristics of nodules were collected to confirm pCA ≤50% using the Mayo calculator. The IC (Nodify XL2™ test, Biodesix, Inc.) reports a calculated post-IC pCA using a negative likelihood ratio. Pre-test information and post-IC results are summarized and compared to a similar group from the clinical validation study. This subset (PANOPTIC study (PS), n=226) was derived from the original patient cohort by restricting criteria to intended use (Mayo pCA ≤50%). RESULTS: The CU group was slightly older (median: 67 vs. 64 years, P=0.0003), with smaller nodules (median: 11 vs. 13 mm, P<0.0001) and more patients who never smoked (28% vs. 20%, P=0.02). Nodule location, spiculation and history of previous cancers were similar. The pCAs were similar in both groups pre-IC (median: CU 20% vs. PS 22%, P=0.07) and post-IC (median: CU 6% vs. PS 10%, P=0.10). The percentage of nodules at baseline with a very low risk (pCA <5%) was 3% and 5% for PS and CU groups, respectively. Post IC, more nodules were reclassified to pCA <5% (P<0.0001) with a similar change in both groups (nodules with pCA <5%: PS 44% vs. CU 49%). CONCLUSIONS: These data show that using the IC, lung nodules are reclassified similarly from the moderate to the very low risk pCA category of ACCP guidelines in both predicted clinical validation and real-world clinical use. A limitation is that outcome data is not known for the CU group, however, the similarity of patient characteristics and distribution of pCA between the two groups suggests good generalizability of testing from research to clinical practice. The significant reclassification of nodules into the very low risk category may result in more nodules being sent to CT surveillance. Clinical trials are ongoing to determine if reclassification is a surrogate marker for clinical utility, including the reduction of invasive procedures and low false negative rate. CLINICAL IMPLICATIONS: This classifier integrates the patientʼs biological signature with clinical risk factors to reclassify a large proportion of nodules to the very low risk category and potentially improve management of likely benign nodules. DISCLOSURES: Consultant relationship with Biodesix Please note: $20001 - $100000 Added 05/31/2020 by James Jett, source=Web Response, value=Consulting fee Employee relationship with Biodesix Inc. Please note: >$100000 Added 06/01/2020 by Kerstin Pohl, source=Web Response, value=Salary No relevant relationships by Gerard Silvestri, source=Web Response Employee relationship with Biodesix Please note: >$100000 Added 06/02/2020 by Steven Springmeyer, source=Web Response, value=Salary