Project Summary
Title: Early change in circulating tumor DNA as a patient- and trial-level diagnostic in advanced lung cancer
Primary Investigator: Geoffrey R. Oxnard, MD
Genomic analysis of plasma circulating tumor DNA (ctDNA) is transforming the care of advanced non-
small cell lung cancer (NSCLC). Plasma ctDNA genotyping approaches (also known as “liquid biopsies”) are
now used routinely to noninvasively test for key cancer genotypes (EGFR, KRAS, ALK, ROS1, etc), both to
guide initial systemic treatment or to effectively target drug resistance. The quantitative nature of this
diagnostic creates obvious potential for use as a noninvasive tool for monitoring treatment response, a
phenomenon we and others have shown retrospectively but an approach which is inadequately being
leveraged clinically. Our preliminary data suggests that plasma ctDNA response is a highly dynamic marker,
offering rapid insight into treatment effect, earlier and potentially more sensitively than standard response
imaging. In an era with an increasing number and variety of cancer therapies, nimble tools for evaluating
treatment benefit may help us better assess therapies and get patients onto the best treatments.
Our group at DFCI is uniquely positioned to translate plasma genomics into a clinical response
biomarker for NSCLC care. We have led the clinical validation of numerous ctDNA genotyping assays, while
simultaneously characterizing the complexities and limitations of ctDNA genomics. In parallel, we have a
strong track record in studying radiographic response as a translational tool and trial endpoint. Our group is
unique in its focus on the clinical application of novel diagnostics, with a deep understanding of the clinical
decision points and the necessary diagnostic performance. Building off this experience, we have an
opportunity now to address a critical scientific gap and inform the scientific community on (1) what constitutes a
plasma response, (2) what does a plasma response signify, and (3) how should plasma response be used as a
patient-level and trial-level diagnostic.
In this application, we first plan to quantify the clinical variation of ctDNA levels in patients with NSCLC
starting a new therapy and, using existing trial and clinical cohorts, identify and validate a plasma response
cutpoint which reliably and robustly predicts for benefit from therapy. In parallel, we will perform a trial-level
analysis of plasma response in a cohort of 11 expansion cohorts from 8 clinical trials in EGFR- and KRAS-
mutant NSCLC, all with plasma being banked at DFCI, to understand the optimal timing of plasma response as
well as its ability to predict trial outcome (radiographic response rate). Finally, we will launch a therapeutic trial
of first-line immunotherapy in advanced NSCLC which uses plasma response assessment to adaptively guide
intensification of treatment, a paradigm-shifting approach which, if successful, has the potential to motivate a
new generation of trials using plasma ctDNA monitoring to effectively evolve the treatment of cancer patients.
Together, these studies will lay the groundwork for ctDNA-based response assessment as a widely available
tool for guiding patient care and for clinical trial analysis.