1640P - Response monitoring with ctDNA in metastatic pancreatic cancer

Background

Various studies about clinical usefulness of serial monitoring with liquid biopsy are being conducted actively in many types of cancer. However, there is not much in pancreatic cancer. We explored whether serial ctDNA monitoring in metastatic pancreatic cancer (MPC) can be beneficial for evaluating treatment response and predicting prognosis.

Methods

A total of 76 MPC patients were prospectively enrolled. Blood samples were collected at every response evaluation until progression of disease (PD) or 5^th evaluation. Quantitative ctDNA were measured by droplet digital polymerase chain reaction using KRAS G12/G13 screening multiplex Kit (Bio-Rad) and reported as variant allele fraction (VAF). Percent change of ctDNA VAF (DctDNA(%) = (current VAF - previous VAF)/previous VAF*100) were correlated to radiographic responses by RECIST 1.1 and tumor volume. Furthermore, Progression free survival (PFS) and overall survival (OS) were analyzed by DctDNA.

Results

Among the 76 enrolled patients, baseline ctDNA were detected in 65 patients. Patients with hepatic metastasis had higher baseline ctDNA level than other metastasis (p=0.038). During a median follow-up of 11.3 months, PD by radiographic response was observed in 43 patients. As expected, PD group showed higher DctDNA compared with non-PD group (p=0.013). The group with increased tumor volume exhibited higher DctDNA levels than the other groups (p=0.028). DctDNA for PD yielded Area Under Curve of 0.839 (p<0.001) with 73.3% sensitivity at 90% specificity. Patients who showed early ctDNA clearance had longer PFS (8.2 m vs 4.8 m; HR 0.45; 95% CI 0.20 to 1.03; p=0.059) and OS (16.6 m vs 10.8 m; HR 0.22; 95% CI 0.07 to 0.75; p=0.013) than those who did not. In multivariable analysis, early ctDNA clearance was still associated with significantly longer PFS (HR 0.29; p=0.036) and OS (HR 0.11; p=0.014).

Table: 1640P

Conclusions

Serial monitoring with ctDNA in MPC has considerable clinical potential in monitoring treatment response and predicting prognosis.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

The authors.

Funding

Has not received any funding

Disclosure

All authors have declared no conflicts of interest.