Preoperative versus Postoperative Chemoradiotherapy for Rectal Cancer

Authors: Rolf Sauer MD, Heinz Becker MD, Werner Hohenberger MD, et al. for the German Rectal Cancer Study Group
Source:New England Journal of Medicine 351(17): 1731-1340 (2004)

Background

• Adjuvant radiotherapy (RT) with or without chemotherapy has significantly improved outcomes in patients with rectal cancer undergoing resection.
• The National Institutes of Health (NIH) consensus conference has recommended postoperative chemoradiotherapy (CTRT) for patients with stage II or III rectal cancer.
• Several randomized studies and a meta-analysis have suggested that pre-operative RT may also be of significant benefit in terms of local control and possibly overall survival.
• The current trial was designed to test whether preoperative CTRT would improve outcomes compared to postoperative CTRT for stage II and III rectal cancer

Methods

• Patients were enrolled from 2/95 to 9/02.
• Patients were eligible if they had histologically-proven respectable adenocarcinoma of the rectum with the inferior margin of the tumor within 16 cm of the anal verge.
• Endorectal ultrasound (EUS) and computed tomography (CT) scanning was used to exclude patients with stage I tumors or with distant metastasis.
• Patients were also excluded if they were older than 75 years, previously had a non-melanoma skin cancer, had previously received RT or chemotherapy or were unable to receive chemotherapy for medical reason
• Patients were randomly assigned to either:
• 1. Arm A: RT (5040 cGy in 180 cGy fractions) with concurrent 5-fluorouracil (5-FU) chemotherapy (1000 mg/m2/d as 120-hour infusion) on weeks 1 and 5 of RT followed by surgical resection 6 weeks after CTRT followed by 4 cycles of bolus 5-FU (500 mg/m2/d, 5 times weekly) q4 weeks.
  2. Arm B: Surgical resection followed by the same CTRT and the same adjuvant chemotherapy 4 weeks after CTRT
• Informed consent was obtained only after the patient was randomly assigned to one of the two treatment groups and informed of which group they were assigned to.
• The primary endpoint was overall survival (OS) with secondary endpoints of disease-free survival (DFS), postoperative complications, acute and long-term toxic effects, and sphincter preservation.

Results

• The median follow-up was 45.8 months.
• A total of 823 patients were randomized (421 Arm A, 402 Arm B) of which 24 were not eligible (16 in Arm A, 8 in Arm B), mostly because they withdrew consent or did not meet inclusion criteria.
• 9 patients in Arm A and 19 patients in Arm B requested a change in treatment group, resulting in 415 patients treated according to Arm A, and 384 patients treated according to Arm B.
• Patients were more likely to receive the prescribed RT in Arm A than in Arm B (92% vs. 54%, p<0.001) and more likely to receive chemotherapy as planned (89% vs. 50%, p<0.001)
• In Arm B, 28% were excluded post-operatively because they were found to have either pathologic stage I disease (18%) or distant metastasis or post-operative complications or death (10%)
• A significant amount of downstaging was seen in the preoperative CTRT arm with 25% of patients having stage I disease (compared to 18%) in this group and 25% having lymph node involvement (stage III disease) compared to 40% in Arm B (p<0.001)
• Rates of complete resection and sphincter-sparing surgery were no different between the two arms.
• Among patients who were deemed by the surgeon to require an abdominoperineal resection (APR) prior to therapy, sphincter-sparing surgery was more common in Arm A compared to Arm B (39% vs. 19%, p=0.004)
• There was no difference between the two arms with regards to rate of in-hospital mortality, overall rate of post-operative complications, anastomotic leakage, postoperative bleeding, or ileus.
• The rates of any acute grade 3 or 4 toxicity was lower in Arm A than in Arm B (27% vs. 40%, p=0.001), including a lower rate of diarrhea (12% vs. 18%, p=0.04)
• The rates of any long-term grade 3 or 4 toxicity was also lower in Arm A (14% vs. 24%, p=0.01) including a decreased rate of strictures at the anastomotic site in Arm A compared to Arm B (4% vs. 12%, p=0.003)
• There was no difference between the two groups with regards to five-year OS (76% vs. 74%, p=0.80), five-year DFS (68% vs. 65%, p=0.32), or incidence of distant recurrences at five years (36% vs. 38%, p=0.84)
• Five-year local recurrence were reduced in Arm A (6% vs. 13%, p=0.006)

Author's Conclusions

• Preoperative CTRT, when given as planned, results in reduced rates of local failure and acute and long-term toxicities compared to postoperative CTRT for locally advanced, respectable rectal adenocarcinoma.
• The rate of sphincter-sparing surgery is improved with preoperative CTRT for patients who are initially judged by the surgeon to require APR.
• Postponing surgery for a six-week course of CTRT and six-week interval before surgery did not result in increased surgical complications or incidence of tumor progression.
• Preoperative CTRT is the preferred treatment for locally advanced, resectable rectal cancer

Discussion

The current study addresses an important question in the optimal treatment of rectal cancer. Previously, two studies in the United States attempted to address this exact issue; however, both were closed due to lack of patient accrual. This study shows that in patients who are judged to require postoperative CTRT by preoperative staging, the use of preoperative therapy results in improved local control and decreased rates of complications compared to postoperative therapy. This study utilized chemotherapy and radiation that is widely used in the United States and should be widely applicable to treatment of these patients.

Two important caveats should be considered when interpreting the results of this study. First, the study was not analyzed in an intent-to-treat manner. Instead, patients were allowed to switch treatment groups after being informed of the treatment arm to which they were randomized. The analysis was performed by the actual treatment received, rather than by the group to which they were randomized. This introduces a potential bias in the results as an uneven distribution of patients may switch from one group to the other. Second, almost one-fifth of patients in the postoperative CTRT group were found to have either stage I disease or distant metastatic disease at the time of surgery. This indicates that a similar percentage of patients underwent CTRT in the preoperative therapy arm that otherwise may not have if their pathologic stage had been available. The potential for overtreating a number of patients clearly exists when treating patients with preoperative CTRT. Despite this fact, the relatively low rates of complications with preoperative treatment, the improvement in sphincter-preservation in patients deemed to need an APR, and the improvement in local control would indicate that preoperative CTRT would be preferred in these patients, particularly for those with low-lying tumors who are at risk for requiring an APR.