Radiosensitization by adjuvant chemotherapy can decrease the risk of distant relapse and locoregional failure (LRF) following resection of rectal cancer. Some studies have suggested the superiority of protracted venous infusion (PVI) of fluorouracil (FU) over bolus administration; others have found improved outcome associated with biochemically modulated FU treatment compared with FU alone.OBJECTIVE
To determine whether extended PVI of FU in conjunction with pelvic radiotherapy (RT) could improve outcome following surgery for rectal cancer compared with the use of bolus FU.DESIGN
Adult patients with curatively resected rectal adenocarcinoma with nodal spread, extension through the muscularis propria, or both (T3-4, N0 or T1-4, N1-3), but no signs of metastasis, were included in this phase III, multicenter trial. Patients who had previously received chemotherapy or radiotherapy for rectal cancer were excluded.INTERVENTION
Patients were registered between March 1994 and August 2000. Following surgery, patients were randomly assigned to one of the three study arms. Arm 1 comprised bolus FU in two cycles on days 1-5 every 28 days before and after pelvic RT, plus FU via PVI during RT, arm 2 comprised FU via PVI for 42 consecutive days before and 56 consecutive days after RT plus FU via PVI during RT, and arm 3 comprised bolus FU plus leucovorin (LV) in two 5-day cycles before and after RT, plus bolus FU with LV combined with RT. Levamisole was administered each cycle before and after RT.OUTCOME MEASURES
The primary end points of the study were overall survival (OS) and disease-free survival (DFS).RESULTS
Median follow-up was 5.7 years in 1,917 patients. No differences were found in OS or DFS between the three arms (P = 0.5 and P = 0.25 respectively). Three-year OS and DFS rates ranged from 81% to 83% and from 67% to 69%, respectively. Grade 3-4 hematologic toxicity was substantially higher in the bolus arms (55% for arm 1 and 49% for arm 3) than in the PVI-only arm (4%), and was higher in women than in men. Catheter-related toxicity was low. Grade 3-4 gastrointestinal toxicity was similar in all of the arms (41-44%). Overall, 15 patients died as a result of treatment, but lethal toxicity did not differ between arms. LRF ranged from 4.6% to 8%. LRF was 5% in arm 1, 3% in arm 2, and 5% in arm 3 for patients with T1-2, N1-2, or T3, N0-2 primaries who received low anterior resection (those most suitable for primary resection).CONCLUSION
The administration of FU via PVI throughout therapy did not improve OS and DFS compared with FU administration via PVI only during RT. A bolus-based regimen produced similar results to the two PVI-based arms. LRF with postoperative therapy is low, justifying initial resection for T1-2, N1-2 or T3, N0-2 candidates for anterior resection.