Predicting Treatment Outcome Of Rectal Cancer Patients With Neoadjuvant Chemoradiotherapy By Ctdna: The Potential Use Of Ctdna Monitoring As Organ-Sparing Approach

Neoadjuvant chemoradiotherapy (nCRT) is widely accepted for the treatment of locally advanced rectal cancer (LARC). Watch & wait strategy can avoid surgery to improve life quality, but needs strict patient selection. Here, we investigated the ability of circulating tumor DNA (ctDNA) to predict treatment outcome and improve risk stratification in LARC. Between 2015 and 2016, we enrolled LARC (T3/T4 and/or N+) patients planned for nCRT. Plasma samples were collected pretreatment, in the middle of nCRT, post-nCRT, before surgery and one week after surgery. Somatic mutations in individual patient's tumor and ctDNA were identified via massively parallel sequencing of 455 genes commonly mutated in colorectal cancer. Responders (ypTRG 0–1) or non-responders (ypTRG 2–3) were recorded. Chi-square test was used to compared ctDNA levels to clinical, radiological and pathological response. We analyzed 440 serial plasma samples from 88 patients. Ten cases were pathological complete response (pCR). CtDNA was detectable in 65.9% (58/88), 14.7% (13/88) and 3.4% (3/88) of pretreatment, post-nCRT and post-surgery samples, with most frequent mutations as TP53, APC, KRAS. The rates of responders and non-responders are 24.1% (14/58) and 75.9% (44/58), however, no difference was found in the baseline ctDNA between them. CtDNA monitoring showed that 65.5% (38/58) of the patients’ ctDNA level decreased sharply to zero before the end of nCRT, which was consistent to their radiological and pathological changes. CtDNA was detectable in 60% (6/10) pCR cases, most of their ctDNA vanishing at the middle of nCRT. 29.3% (17/58) of the patients’ ctDNA status changed in irregular patterns. However, 5.2% (3/58) of the patients’ ctDNA status goes up during the course of nCRT, all of the three developed metastatic disease during the follow up period. Increased level of ctDNA post-nCRT indicated high rates of disease progression, which was not appropriate for W&W strategy. The baseline levels were not helpful in distinguishing responders from non-responders. Post-nCRT analysis stratifies patients with LARC into subsets at high or low risk of progression. CtDNA analysis could potentially be used to guide patient selection for W&W strategy or adjuvant chemotherapy.