Short-term outcomes of a multicentre randomized clinical trial comparing D2 versus D3 lymph node dissection for colonic cancer (COLD trial)

Abstract

Abstract Background It remains unclear whether extended lymphadenectomy provides oncological advantages in colorectal cancer. This multicentre RCT aimed to address this issue. Methods Patients with resectable primary colonic cancer were enrolled in four hospitals registered in the COLD trial, and randomized to D2 or D3 dissection in a 1 : 1 ratio. Data were analysed to assess the safety of D3 dissection. Results The study included the first 100 patients randomized in this ongoing trial. Ninety-nine patients were included in the intention-to-treat (ITT) analysis (43 D2, 56 D3). Ninety-two patients received the allocated treatment and were included in the per-protocol (PP) analysis: 39 of 43 in the D2 group and 53 of 56 in the D3 group. There were no deaths. The 30-day postoperative morbidity rate was 47 per cent in the D2 group and 48 per cent in the D3 group, with a risk ratio of 1·04 (95 per cent c.i. 0·68 to 1·58) (P = 0·867). There were two anastomotic leaks (5 per cent) in the D2 group and none in the D3 group. Postoperative recovery, complication and readmission rates did not differ between the groups in ITT and PP analyses. Mean lymph node yield was 26·6 and 27·8 in D2 and D3 procedures respectively. Good quality of complete mesocolic excision was more frequently noted in the D3 group (P = 0·048). Three patients in the D3 group (5 per cent) had metastases in D3 lymph nodes. D3 was never the only affected level of lymph nodes. N-positive status was more common in the D3 group (46 per cent versus 26 per cent in D2), with a risk ratio of 1·81 (95 per cent c.i. 1·01 to 3·24) (P = 0·044). Conclusion D3 lymph node dissection is feasible and may be associated with better N staging. Registration number: NCT03009227 (http://www.clinicaltrials.gov).