An AGITG trial –A randomised phase II study of pre-operative cisplatin, fluorouracil and DOCetaxel +/-radioTherapy based on poOR early response to cisplatin and fluorouracil for resectable esophageal adenocarcinoma

Background Patients (pts) with esophageal/ gastro-esophageal junction (GOJ) adenocarcinoma (EAC) who do not show early metabolic response (as defined by 18-FDG PET) to chemotherapy have poorer survival and histological response rates <5%. We investigate if, by tailoring the neoadjuvant therapy, histological response can be improved in pts not showing early metabolic response after the first cycle of treatment. Methods Pts with resectable EAC were enrolled to a multicentre randomised phase II trial. PET scans were performed at baseline and after receiving one cycle of cisplatin and fluorouracil (CF) (day 15). If SUVmax decreased by ≥ 35%, pts were classified as early metabolic responders (EMR), otherwise as non-metabolic responders (NMR). EMR pts continued with a 2nd cycle of CF follow by surgery while the NMR pts were randomised 1:1 to receive either CF and docetaxel (DCF) for two cycles, or DCF concurrent with 45Gy radiotherapy (DCFRT). The primary endpoint was histological response (<10% residual tumor) in each randomised arm. Results 124 pts were enrolled (62% esophagus and 38% GOJ) and received one cycle of CF. At day 15, 45/122 (37%) pts were EMR and 77/122 (63%) were NMR and 2 did not have a day 15 PET. Of the NMR group, 31 were randomised to DCF; 35 to DCFRT; 11 were not randomised (2 pts progressed, 4 had toxicities and 5 refused). Grade 3 or 4 toxicities (AE) were seen in 19/58 (33%) pts on CF, 14/31 (45%) DCF pts and 25/35 (71%) DCFRT pts. No treatment related deaths were seen. Esophagectomy was performed in 45 (100%) EMR, 28/31 (90%) DCF (1 refused DCF, 1 refused surgery and 1 AE) and 33/35 (94%) DCFRT (2 progressed). R0 (>1mm margin) resection was achieved in 31/45(69%) of EMR, 18/28 (64%) of DCF, 31/33 (94%) of DCFRT pts. A major histological response was achieved in 3/45 (7%) EMR pts, 6/30 (20%) DCF pts, and 22/35 (63%) DCFRT pts. Conclusions Docetaxel added to CF, particularly DCFRT, can induce high rates of histological responses in NMR pts. Therefore tailoring multimodality therapy based on individual PET response is safe and feasible in EAC, although the impact on survival requires longer follow up.