Randomized Phase III Trial Comparing Infused Irinotecan/5-flurouracil (5-FU)/Folinic Acid (IF) versus 5-FU/FA (F) in Stage III Colon Cancer Patients (PETACC-3, v307)

Reviewer: S. Jack Wei, MD
University of Pennsylvania School of Medicine
Last Modified: May 17, 2005

Presenter: E. Van Cutsem
Presenter's Affiliation: PETACC-3 investigators
Type of Session: Scientific


  • Use of post-operative chemotherapy in patients with stage III colon cancer is well established.
  • Irinotecan and infusional 5-FU and FA are standard options for first-line therapy in metastatic colorectal cancer.
  • This study examines whether the addition of irinotecan to 5-FU and FA add significant benefit to patients with stage II and III colon cancer compared to 5-FU and FA alone.

Materials and Methods

  • Patients with stage II and III colon cancer who underwent surgical resection were randomized to:
    1. F: FA (200 mg/m2) + bolus 5-FU (400 mg/m2) OR continuous infusional 5-FU (600 mg/m2) repeated every 2 weeks (LV5FU2 regimen)
    2. IF: LV5FU2 + Irinotecan (180 mg/m2)
  • Patients were stratified by stage II vs. III and by treatment center.
  • Eligible patients had adenocarcinoma of the colon or intraperitoneal rectum, stage II or III by the UICC classification system, analysis of at least 8 lymph nodes if possible, sampling of at least 8 lymph nodes if stage II, no evidence of residual disease after resection, age 18-75, WHO performance status 0-1, and adequate hematologic, hepatic, and renal function.
  • Primary endpoint was disease-free survival (DFS) in stage III patients.  DFS is defined as relapse, second primary colon cancer, second primary cancer other than colon cancer, or death from any cause.
  • Secondary endpoints included relapse-free survival (RFS) in stage III patients, DFS in combined stage II & III patients, overall survival (OS), and safety.  RFS is defined as relapse, second primary colon cancer, or death from any cause and does not include other second non-colon primary cancers.


  • Between 1/00 and 4/02, 3005 patients from 340 sites in 32 countries were randomized.
  • Stage II: 894 patients (14 not treated); stage II: 2111 patients (17 not treated)
  • Patient groups were well-balanced for age, node positive vs. node negative, number of analyzed lymph nodes, evidence of obstruction or perforation, weight loss, performance status, histologic grade, and evidence of perineural or vascular invasion.
  • There was a significant imbalance in T4 stage: IF 17% vs. F 13% (p=0.006)
  • Median FU = 38 months.
  • For stage III patients, DFS hazard ratio (HR) = 0.89 (p=0.091)
  • For stage III patients, RFS HR = 0.86 (p=0.045)
  • Because of the imbalance between the two groups with regards to T stage, a post hoc analysis was performed adjusting for this difference between to two groups.  When doing so:
    • DFS HR = 0.85 (p=0.021)
    • RFS HR = 0.82 (p=0.009)
  • 3-year DFS in stage III patients:
Risk Adjusted
  • 3-year RFS in stage III patients:
Risk Adjusted
  • 3-year DFS in stage II&III patients combined:
Risk Adjusted
  • 3-year DFS in stage II patients:
  • Grade 3/4 Neutropenia: IF 28.2% vs. F 6.0%
  • Grade 3/4 Neutropenic Infection: IF 2.0% vs. F 0.1%
  • Higher rates of diarrhea, stomatitis, nausea, vomiting, and alopecia seen in the IF arm compared to F.
  • No difference in the number of treatment-related deaths.

Author's Conclusions

  • There were no unexpected toxicities with IF therapy.
  • Grade 3/4 diarrhea was 12% in the IF arm compared to 6% in the F arm.
  • There was a low incidence of febrile neutropenia and neutropenic infection.
  • 60 day all cause mortality and mortality within 30 days from last dose were low (<1%).
  • There was a trend towards improved 3-year DFS with IF; however, the primary endpoint was not met.
  • The secondary endpoint of improved 3-year RFS was met.
  • Adjusting for imbalances between T and N stages between the two treatment groups results in improved 3-year DFS and RFS for IF compared to F in a post hoc analysis.
  • In patients with stage III colon cancer, the addition of irinotecan to LV5FU2 reduces the risk of recurrence.

Clinical/Scientific Implications

This study follows two recently presented abstracts which also examined the use of irinotecan in addition to standard 5-FU-based chemotherapy as adjuvant therapy for stage II and III colon cancer.  In the previous trials, the benefit of irinotecan was marginal at best.  These results were somewhat disappointing given the clear benefit seen with the addition of irinotecan in the setting of metastatic colon cancer.  Unfortunately, the study presented here does not add significantly to this data.  Again, there was marginally significant benefit seen with irinotecan at the cost of increased hematologic and gastrointestinal toxicity as well as alopecia. 
Because there was a failure to stratify for T or N stage in this study, both of which are strongly linked to outcome, an imbalance in staging occurred between the two arms.  This significantly affects the ability to compare the two arms of this trial.  Although a post hoc analysis was performed to try to control for this balance, it should be kept in mind that this is merely a statistical reanalysis of the data.  The ultimate outcomes in the two arms showed no difference with the addition of irinotecan.  Given the clear benefit seen with the addition of oxaliplatin to 5-FU in this same population of patients, it would appear that oxaliplatin would remain the adjuvant treatment of choice, with irinotecan reserved only for those patients who cannot receive oxaliplatin, with the understanding that irinotecan adds marginal benefit at best.