IMPACT-NSCLC: Innovative machine-learning protocols for accurate clinical biomarker testing in non-small cell lung cancer (NSCLC).

Authors

null

Muzammil Dastagir

University of Alabama at Birmingham, Birmingham, AL

Muzammil Dastagir , Shashi Shankar , Aakash Desai

Organizations

University of Alabama at Birmingham, Birmingham, AL, Novellia, New York, NY

Research Funding

Daiichi Sankyo

Background: Lung cancer is often detected in advanced stages and can benefit from biomarker testing to identify targeted therapeutics. The IMPACT-NSCLC study was designed to identify barriers to equitable biomarker testing for lung cancer patients and develop targeted interventions to help improve biomarker testing rates using a digital personal health record and AI-enabled analysis. Methods: IMPACT-NSCLC unites the O'Neal Comprehensive Cancer Center (UAB) and Novellia, a leading personal health record tool, in a two-part single-institution study of adult NSCLC patients. Novellia uses proprietary technology and AI-enabled services to help patients digitally collect retrospective health records from providers and track health records prospectively, offering researchers a novel method of collecting and evaluating longitudinal health data. Part 1 of the study involves a retrospective study of patients with advanced NSCLC, with advanced data modeling to identify barriers and predictive factors in biomarker testing. Part 2 focuses on developing, implementing, and evaluating targeted interventions based on Part 1’s findings with the goal of improving biomarker testing rates in both academic and community settings. Results: In Part 1, data were collected from 35 patients diagnosed with NSCLC from 2020-2024. Most were white (80.77%) and male (53.84%). Median age was 65 (range 51-79). Most received care at an academic (NCI-designated) cancer center (74.28%). Most were not on commercial insurance plans (85.71%), with the most common form being BCBS/Medicare B and Viva Medicare. Most patients had Stage IV NSCLC (82.85%). More patients received liquid NGS testing at diagnosis (91.42%) than tissue NGS testing at diagnosis (68.57%). Nearly all patients (82.85%) received combination therapy, such as pemetrexed, and pembrolizumab, following biomarker testing. Oncologists initiated treatment while biomarker testing results were pending or incomplete more often in non-academic settings (55.55% of cases) vs. academic settings (26.92%). Conclusions: The pilot analysis found that patients with NSCLC typically received biomarker testing at diagnosis, more often with liquid biopsy. However, oncologists, especially in non-academic settings, often started treatment before full biomarker testing was completed. Comprehensive biomarker testing was frequently not done at progression. Further analysis leveraging machine learning and advanced modeling of Part 1 data may reveal more predictive factors for these variations, guiding interventions in Part 2. Developing a scalable best-practice guide for comprehensive biomarker testing could improve health system efficiencies and outcomes for patients with NSCLC.

Disclaimer

This material on this page is ©2024 American Society of Clinical Oncology, all rights reserved. Licensing available upon request. For more information, please contact licensing@asco.org

Abstract Details

Meeting

2024 ASCO Quality Care Symposium

Session Type

Poster Session

Session Title

Poster Session B

Track

Health Care Access, Equity, and Disparities,Technology and Innovation in Quality of Care,Survivorship

Sub Track

Access to Timely Detection and Referral

Citation

JCO Oncol Pract 20, 2024 (suppl 10; abstr 84)

DOI

10.1200/OP.2024.20.10_suppl.84

Abstract #

84

Poster Bd #

B8

Abstract Disclosures