LBA28 The PEGASUS trial: Post-surgical liquid biopsy-guided treatment of stage III and high-risk stage II colon cancer patients

Retrospective studies demonstrated an impressive prognostic impact of circulating tumor DNA (ctDNA, LB) positivity on colon cancer (CC) relapse after radical resection as proxy of minimal residual disease (MRD). Initial evidence is accumulating that in some patients (pts) adjuvant treatment does not clear ctDNA and this associates with relapse. With these premises we designed the PEGASUS trial to prove the feasibility of using LB to guide the post-surgical and post-adjuvant clinical management of Stage-IIT4N0/III CC pts. LB was performed using Guardant Reveal assay (v L1.2). In the post-surgical phase, LB+ pts received 3 months (mos) of CAPOX and LB- pts 6 mos of CAPE. To mitigate the risk of false negatives a LB was repeated after one cycle of CAPE and treatment escalated to CAPOX if LB+. At the end of adjuvant treatments, LB was repeated: CAPOX-treated pts were switched to FOLFIRI if LB+ or de-escalated to CAPE if LB-; CAPE-treated pts received CAPOX if LB+ or started the standard follow-up phase if LB-. Primary endpoint was the number of post-surgery false negative cases, defined as pts with two consecutive LB- experiencing disease relapse (accepted <14 relapses). From July 2020 to July 2022, 135 pts were included in 11 cancer centers in Italy and Spain. At data cut-off, median follow-up was 20,8 months, with data completeness of 91% in LB- and 92% in LB+. Post-surgery ctDNA was detected in 35/135 pts (26%), of which 12 (34%) experienced relapses, while 9 out of 100 LB- pts (9%) relapsed, being defined as false negatives. ctDNA+ was associated with a clear increase in the risk of relapse (HR 4.37, log rank P=0.0003). After 3 months of CAPOX, 11/35 LB+ pts were converted to LB- (31%), but 8/11 relapsed or LB re-positivized. Of the 23 LB+ pts receiving FOLFIRI after CAPOX, 12 remained LB+ (52%) among which 6 relapsed, while 11 were converted to LB- remaining relapse-free at the time of analysis (48%), thus suggesting an effect of FOLFIRI in the MRD setting. LB may be used to guide the post-surgical clinical management of CC pts by reducing unnecessary toxicity and by improving the response to standard chemotherapy. Ongoing and newly designed randomized trials trials will define the clinical utility of LB for MRD in colon cancer.