MRI Tumor Regression Grade and Circulating Tumor DNA as Complementary Tools to Assess Response and Guide Therapy Adaptation in Rectal Cancer.

Purpose: Response to preoperative chemo-radiotherapy (CRT) varies. We assessed whether circulating tumor DNA (ctDNA) might be an early indicator of tumor response or progression to guide therapy adaptation in rectal cancer.

Experimental Design: A total of 243 serial plasma samples were analyzed from 47 patients with localized rectal cancer undergoing CRT. Up to three somatic variants were tracked in plasma using droplet digital PCR. RECIST and MRI tumor regression grade (mrTRG) evaluated response. Survival analyses applied Kaplan-Meier method and Cox regression.

Results: ctDNA detection rates were: 74% ( = 35/47) pretreatment, 21% ( = 10/47) mid CRT, 21% ( = 10/47) after completing CRT, and 13% ( = 3/23) after surgery. ctDNA status after CRT was associated with primary tumor response by mrTRG ( = 0.03). With a median follow-up of 26.4 months, metastases-free survival was shorter in patients with detectable ctDNA after completing CRT [HR 7.1; 95% confidence interval (CI), 2.4-21.5; < 0.001], persistently detectable ctDNA pre and mid CRT (HR 3.8; 95% CI, 1.2-11.7; = 0.02), and pre, mid, and after CRT (HR 11.5; 95% CI, 3.3-40.4; < 0.001) compared with patients with undetectable or nonpersistent ctDNA. In patients with detectable ctDNA, a fractional abundance threshold of ≥0.07% mid CRT or ≥0.13% after completing CRT predicted for metastases with 100% sensitivity and 83.3% specificity for mid CRT and 66.7% for CRT completion. All 3 patients with detectable ctDNA post-surgery relapsed compared with none of the 20 patients with undetectable ctDNA ( = 0.001).

Conclusions: ctDNA identified patients at risk of developing metastases during the neoadjuvant period and post-surgery, and could be used to tailor treatment.