Virtual Library
Start Your Search
A. Sengupta
Author of
-
+
P2.01 - Advanced NSCLC (ID 618)
- Event: WCLC 2017
- Type: Poster Session with Presenters Present
- Track: Advanced NSCLC
- Presentations: 1
- Moderators:
- Coordinates: 10/17/2017, 09:00 - 16:00, Exhibit Hall (Hall B + C)
-
+
P2.01-055 - Examining Metabolomics as a Prognostic Marker in Metastatic Non–Small Cell Lung Cancer Patients Undergoing First-Line Chemotherapy (ID 8685)
09:00 - 09:00 | Author(s): A. Sengupta
- Abstract
Background:
The metabolome represents the endpoint of many cellular events; hence patients' baseline metabolomic profile may reveal specific prognostic markers of overall survival. In this study, we sought to characterize the serum metabolite signatures in patients with metastatic non-small cell lung cancer (mNSCLC) who underwent first-line therapy, using nuclear magnetic resonance (1H-NMR) spectroscopy and liquid chromatography mass spectrometry (LC-MS), and to explore their potential prognostic impact.
Method:
Serum samples were collected prospectively as part of a clinical trial in which patients were treated with systemic therapy including platinum-doublet chemotherapy. For each method of analysis, samples were divided into training (3/5) and validation (2/5) sets stratified by treatment received, stage (III vs. IV), and ECOG PS (0, 1, vs. ≥ 2). Exploratory analyses were performed to characterize the relationships between baseline lipid and polar levels and overall survival. Kaplan-Meier curves were used to estimate the distributions of time to event outcomes, and a Cox regression model was used to correlate marker levels while adjusting for baseline characteristics.
Result:
Using 1H-NMR, 16 out of 43 metabolites were significantly correlated with overall survival (OS) by univariate analysis (p < 0.025) and 4 metabolites were included in the final multivariate model. The median OS was 11.4 months in the low risk group vs. 6.6 months in the high risk group (HR=1.99, 95% C.I. 1.45 – 2.68; p<0.0001). Using LC-MS, 53 lipid species were correlated with OS by univariate analysis. Variables were then subjected to hierarchical cluster analysis resulting in 12 branches which were moderately to significantly correlated with lipid features. Principle component analysis (PCA) was performed and the first PC from each such branch was used (n=9). Using Cox regression modeling, median OS was 5.7 months vs. 11. 9 months for the low and high risk groups respectively, even after adjusting for baseline characteristics (HR: 2.23, 95% C.I. 1.55 – 3.20; p< 0.0001).
Conclusion:
Metabolite profiles from baseline pre-treatment serum samples have the potential to act as prognostic markers in patients with mNSCLC undergoing first-line chemotherapy. Serial metabolite measurements pre- and post-treatment may yield additional information and provide enhanced data for predicting clinical outcomes.
