Abstract LB031: Plasma based metabolic profiling in metastatic gastrointestinal stromal tumors (GIST)

Abstract

Background: Discovery of oncogenic mutations as targets for cancer therapy revolutionized treatment of GIST and other cancers. However, nearly all patients ultimately progress, which emphasizes the need for development of new tools to assess cancer prognosis and factors associated with benefit of cancer therapies. Altered metabolism is a hallmark of cancer, enabling tumors to proliferate, survive and metastasize. By measuring the complete set of metabolites in an individual (metabolome) it is possible to identify biomarkers that correlate with disease status, prognosis, and therapeutic response. Methods: We performed untargeted NMR and MS-based machine learning metabolomic analysis (Olaris, Waltham, MA) of serial plasma samples collected at baseline and during experimental systemic therapies in 39 patients with advanced/metastatic GIST. Results were compared to clinical outcomes. Results: In serial plasma samples from 39 patients with advanced/metastatic GIST using untargeted NMR and MS-based machine learning (ML) metabolomic analysis, we identified metabolic signatures to build Biomarker-of-Response (BoR) ML models that could accurately differentiate patients with response, intrinsic and adaptive resistance to experimental systemic therapies. The BoR ML model also correlated with tumor growth or tumor reduction in patients with response, intrinsic and adaptive resistance. Finally, we identified metabolic pathways associated with response and resistance. Conclusions: Comprehensive metabolomic profiling of serially collected plasma is feasible and detects metabolic signatures associated with therapeutic response in advanced GIST. Citation Format: Chandrashekhar Honrao, Srihari Raghavendra Rao, Nathalie Teissier, S. Greg Call, Elizabeth M. ODay, Filip Janku. Plasma based metabolic profiling in metastatic gastrointestinal stromal tumors (GIST) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr LB031.

Gadolinium-Based Paramagnetic Relaxation Enhancement Agent Enhances Sensitivity for NUS Multidimensional NMR-Based Metabolomics

“…Indeed, altered metabolite levels have been observed in many diseases, including diabetes [39], neurodegeneration [40], cancer [4], cardiovascular disease [6], and even aging [3]. Furthermore, in a series of separate studies, we have identified metabolite biomarkers of response (BoRs) that correlate with drug responsiveness for metastatic breast cancer patients treated with CDK4/6 inhibitors as well as the anti-HER2 therapy trastuzumab; and for gastrointestinal stromal tumor (GIST) patients treated with tyrosine kinase inhibitors [41][42][43]. While additional validation studies are required, these preliminary results suggest the exciting possibility that metabolite-based biomarkers have for designing optimal treatment strategies for individual patients, which is a major goal of precision medicine.…”

Plasma Metabolite Signature Classifies Male LRRK2 Parkinson’s Disease Patients

“…We have previously demonstrated that using a global, unbiased, multi-dimensional 1 H- 13 C heteronuclear single quantum coherence spectroscopy (HSQC) NMR spectroscopy metabolomics platform coupled with machine learning (ML) techniques, we are able to uncover metabolite-based biomarkers that correlate with disease status and response to therapy [32][33][34]. Further, as metabolomics studies are widely acknowledged to suffer from reproducibility challenges [35][36][37][38], we documented the reproducibility of our platform and how to translate our results for vitro diagnostic development [25,32,33]. In this study, we used our NMR and ML platform to measure plasma metabolites from approximately age/sex-matched PD patients with G2019S LRRK2 mutations and non-PD controls wildtype for LRRK2.…”