Start Your Search
Lisa I. Wang
P1.01 - Advanced NSCLC (Not CME Accredited Session) (ID 933)
- Event: WCLC 2018
- Type: Poster Viewing in the Exhibit Hall
- Presentations: 1
- Coordinates: 9/24/2018, 16:45 - 18:00, Exhibit Hall
P1.01-04 - Treatment Patterns and Overall Survival Following Biomarker Testing in Real-World Advanced NSCLC Patients (ID 12743)
16:45 - 18:00 | Author(s): Lisa I. Wang
Foundation Medicine (FMI) comprehensive genomic profiling and other next-generation sequencing (NGS) tests are gaining importance in routine clinical management of non-small cell lung cancer (NSCLC). They assess multiple genetic alterations that drive sensitivity or resistance to treatment, enabling optimal therapeutic decisions. We evaluated the effect of biomarker testing on treatment patterns and overall survival (OS) in real-world advanced NSCLC (aNSCLC) patients receiving different test types, and in non-tested patients.a9ded1e5ce5d75814730bb4caaf49419 Method
The Flatiron Health (FH) Database comprises patient-level electronic health records from a large network of US cancer clinics. Patients had aNSCLC diagnoses between 01/2013 and 05/2017, ≥2 clinic visits in the FH network, first treatment starting ≤90 days after aNSCLC diagnosis, and biomarker tests before first treatment. Testing data were abstracted for five biomarkers (EGFR, ALK, KRAS, ROS1, and PD-L1). Patients were hierarchically categorized into three testing groups: FMI, other NGS, and single-biomarker non-NGS. Biomarker status and patterns in first treatment were described. Cox proportional hazards models were used to compare OS among testing groups and non-tested patients.4c3880bb027f159e801041b1021e88e8 Result
As of 11/30/2017, 355 patients had ≥1 FMI test, 780 had ≥1 other NGS test, and 6,363 had ≥1 non-NGS test prior to first treatment; 5,148 patients were never tested. Table 1 summarizes biomarker status, treatment patterns, and results of multivariate survival models adjusted for baseline demographic and clinical differences among testing groups. Patients with FMI tests were more likely to receive NCCN-recommended targeted treatments. Better OS was observed for FMI, other NGS, and non-NGS compared with non-tested patients.
(n = 355)
(n = 780)
(n = 6,363)
(n = 5,148)
Patterns in first treatment
Non NCCN-recommended targeted therapy2,4
Non NCCN-recommended ICI2,6
Multivariate Cox proportional hazards model to compare OS, hazard ratio (95% CI)
aNSCLC, non-EGFR-mutated, non-ALK-rearranged, non-squamous cell histology11
ECOG, Eastern Cooperative Oncology Group; ICI, immune checkpoint inhibitor; NCCN, National Comprehensive Cancer Network.
1 Denotes biomarker status overall prior to starting first treatment and represents overall status from all test-types. In case of multiple tests, the following hierarchy is used: positive>negative>pending/unsuccessful/indeterminate/unknown.
2 Based on the NSCLC NCCN Guidelines, Version 3. 2018; 02/21/2018.
3 NCCN-recommended targeted therapy implies treatment regimens containing at least one of the following: erlotinib, afatinib, gefitinib, osimertinib, crizotinib, ceritinib, alectinib, brigatinib, dabrafenib+trametinib, cabozantinib, vandetanib, ado-trastuzumab emtansine.
4 Non NCCN-recommended targeted therapy implies treatment regimens containing at least one of the following: necitumumab, cetuximab, panitumumab, vemurafenib, dabrafenib, trametinib, trastuzumab, pertuzumab+trastuzumab, venetoclax.
5 NCCN-recommended ICI implies treatment regimens containing at least one of the following: pembrolizumab, nivolumab, atezolizumab.
6 Non NCCN-recommended ICI implies treatment regimens containing at least one of the following: ipilimumab, avelumab.
7 Adjusted for age, sex, race, clinic type, payer type, smoking history, stage at initial diagnosis, ECOG performance status, histology, and year of advanced diagnosis.
8 Adjusted for age, sex, race, clinic type, payer type, smoking history, stage at initial diagnosis, ECOG performance status, and year of advanced diagnosis.
9 Adjusted for age, sex, race, clinic type, payer type, smoking history, stage at initial diagnosis, ECOG performance status, histology, year of advanced diagnosis, sample type used for the test, and biomarker status.
10 Adjusted for age, sex, race, clinic type, payer type, smoking history, stage at initial diagnosis, ECOG performance status, year of advanced diagnosis, sample type used for test, and biomarker status.
11 Adjusted for age, sex, race, clinic type, smoking history, stage at initial diagnosis, ECOG performance status, and year of advanced diagnosis.
* Indicates a statistically significant estimate (p<0.05).
Complexity of real-world aNSCLC biomarker testing and associated treatments creates challenges when comparing OS among testing groups. In the future, as more treatments targeting a wider array of genomic alterations become available and accessible, the utility of NGS-based assays to guide NCCN-recommended treatments with actionable targets and differences in OS may become more apparent.6f8b794f3246b0c1e1780bb4d4d5dc53