Utility of tumor-informed circulating tumor DNA for detection of minimal residual disease after curative-intent therapy in localized pancreatic cancer

Abstract

Introduction

Patients with localized PDAC have high recurrence rates even after curative-intent therapies. Detection of minimal residual disease (MRD) can inform prognosis and may be therapeutically actionable. Tumor-informed circulating tumor (ct)DNA has been shown to be useful for MRD detection in other cancers, but its utility in localized PDAC patients undergoing curative-intent therapy is not well established.

Methods

ctDNA samples (total 106) from 32 patients, following completion of all curative-intent therapy, were subjected to Signatera™ analysis (Natera, Inc.). Recurrence-free survival (RFS) data was calculated using Kaplan-Meier estimates. Data from three previously presented studies using the same platform were pooled for validation.

Results

In our cohort (n = 32), ctDNA positivity rate was 28.1 % (9/32) with a median follow-up time of 17.7 months (range 4–62). Median RFS was significantly lower in patients with positive ctDNA (3.6 vs. 29.0 months, p < 0.001; HR: 72.1 [8.6—604.9]). Correlation of positive ctDNA with radiographic recurrence showed a sensitivity of 47.4 % (9/19), specificity of 100 % (13/13), PPV of 100 % (9/9), and NPV of 56.5 % (13/23). In the pooled cohort (n = 172), sensitivity was 66.7 % (50/75), specificity 77.3 % (75/97), PPV 69.4 % (50/72), and NPV 75.0 % (75/100).

Conclusions

Positive tumor-informed ctDNA test shows a high specificity and PPV for radiographic recurrence and is associated with significantly worse RFS. However, sensitivity of the test remains low.

Synopsis

In patients with localized PDAC completing curative-intent therapies, tumor-informed ctDNA assessment shows high specificity for radiographic recurrence and is associated with worse RFS. However, sensitivity remains low and presents an opportunity for improved calibration of this platform.