|Home | About | Journals | Submit | Contact Us | Français|
The primary goal of this multicenter phase III trial was to determine whether overall survival (OS) of fluorouracil (FU) -refractory patients was noninferior when treated with second-line infusional fluorouracil, leucovorin, and oxaliplatin (FOLFOX4; arm B) versus irinotecan (arm A). Cross-over to the other treatment on disease progression was mandated.
Patients who experienced treatment failure with one prior FU-based therapy and had not received prior irinotecan or oxaliplatin, either for metastatic disease or within 6 months of adjuvant FU therapy, were randomly assigned to arm A (irinotecan 350 or 300 mg/m2 every 3 weeks) or arm B (FOLFOX4).
A total of 491 patients were randomly assigned (arm A, n = 245; arm B, n = 246); 288 (59%) had experienced treatment failure with FU for metastatic colorectal cancer. Two hundred twenty-seven patients (46%) received protocol-mandated third-line therapy (arm A, 43%; arm B, 57%). Median survival was 13.8 months (95% CI, 12.2 to 15.0 months) for initial treatment with FOLFOX4 and 14.3 months (95% CI, 12.0 to 15.9 months) for irinotecan (P = .38; hazard ratio = 0.92; 95% CI, 0.8 to 1.1). Response rates (RR; 28% v 15.5%; P = .0009) and time to progression (TTP; 6.2 v 4.4 months; P = .0009) were significantly superior with FOLFOX4. In the nonrandom subset of patients who crossed over, RR and TTP improvements with FOLFOX4 continued into third-line treatment. Irinotecan therapy was associated with more grade 3 nausea, vomiting, diarrhea, and febrile neutropenia; FOLFOX4 was associated with more neutropenia and paresthesias.
In patients who experienced treatment failure with front-line FU therapy, OS does not significantly differ whether second-line therapy begins with irinotecan or FOLFOX4. FOLFOX4 produces higher RR and longer TTP. Both arms had notable OS in patients who experienced treatment failure with first-line FU therapy.
In 2008, colorectal cancer was anticipated to be diagnosed in 148,810 people in the United States, of whom approximately one quarter would have advanced disease.1 Over the past decade, advances in treatment have occurred with the introduction of active conventional chemotherapies (irinotecan, oxaliplatin) as well as biologic agents (bevacizumab, cetuximab). Before formal approval and widespread acceptance of these drugs, patients were treated first line with the then-standard fluorouracil (FU); the role of irinotecan- or oxaliplatin-based regimens as salvage therapy required definition. In the present treatment of patients with advanced colorectal cancer, the optimal use of these agents and how best to combine them with biologics remains unsettled.
Irinotecan gained US Food and Drug Administration approval in 2000 for first-line use in combination with bolus FU.2 Several irinotecan schedules and combinations were subsequently tested,3,4 and a widely accepted schedule is irinotecan administration at 350 mg/m2 given every 3 weeks. In a multicenter phase III study,5 patients with FU-refractory disease randomly assigned to second-line irinotecan experienced a significantly superior median survival (10.8 v 8.5 months) and 1-year survival rate (44.8% v 32.4%) compared with infusional FU. Oxaliplatin, when combined with infusional FU such as in the regimen of infusional fluorouracil, leucovorin, and oxaliplatin (FOLFOX4), has also demonstrated significantly higher response rates (RR) and improved progression-free survival (PFS)7–9 compared with FU alone. FOLFOX4 was subsequently shown to be superior to irinotecan with bolus FU as first-line therapy in Intergroup N9741.10–12
With the addition of active drugs to the treatment of colorectal cancer, challenges remain regarding how best to combine these agents and in which sequence they should be delivered. This is particularly relevant in the context of noncurative, palliative treatment. It has been reported that the greatest benefit is observed when all three active cytotoxic drugs (FU, irinotecan, and oxaliplatin) are given to patients with advanced colorectal cancer.13 In an effort to address this issue, this Intergroup phase III trial evaluated whether the overall survival (OS) of patients with FU-refractory, advanced colorectal carcinoma treated with FOLFOX4 followed by irinotecan monotherapy was noninferior to that of patients treated with irinotecan followed by FOLFOX4.
Three National Cancer Institute cooperative oncology groups collaborated in this study: North Central Cancer Treatment Group (coordinating center), Eastern Cooperative Oncology Group (ECOG), and the Southwest Oncology Group. Patients were eligible if they had experienced treatment failure with one prior FU-based chemotherapy regimen for metastatic disease or experienced disease progression during or within 6 months of FU-based adjuvant therapy. Detailed inclusion/exclusion criteria are enumerated in Table 1.
Patients were randomly assigned to treatment through dynamic allocation14 designed to balance assignment for the following stratification factors: ECOG performance status (PS) score (0/1 v 2), primary indicator lesion (hepatic v pulmonary v other), age (< 65 v ≥ 65 years), alkaline phosphatase (grade < v ≥ 2), FU treatment failure (adjuvant v metastatic), membership (membership within North Central Cancer Treatment Group v other Intergroup). Institutional review board approval from each participating institution was required.
The regimens (doses in milligrams per square meters) were as follows: Arm A, irinotecan 350 mg/m2 on day 1, every 3 weeks (300 mg/m2 for patients with ECOG PS 2, ≥ 70 years of age, or prior pelvic radiotherapy); arm B, oxaliplatin 85 mg/m2 on day 1 and bolus FU 400 mg/m2 plus leucovorin 200 mg/m2 followed by FU 600 mg/m2 in 22-hour infusions on days 1 and 2, every 2 weeks. Treatment continued until progression, unmanageable toxic effects, or withdrawal of consent. Toxic effects (except paresthesias, dysesthesias) were graded using the National Cancer Institute Common Toxicity Criteria version 2.0. Functional impairments that interfered with daily activities or caused disability were classified as grade 3 or 4 paresthesias, respectively. Any grade 3 or 4 toxic effect resulted in a dose reduction for subsequent cycles. Persistent grade 2 or worse toxic effects delayed therapy until toxicity resolved. If the toxic effect exceeded grade 1 after 2 weeks or if more than two dose reductions were required, protocol therapy was discontinued.
Study enrollment required either at least one measurable lesion (> 2 cm in diameter) or disease that could be serially evaluated to establish whether the disease was getting better or worse (assessable disease). Tumor assessments were performed every 6 weeks. Complete response required that all disease disappear. Partial response required at least a 50% reduction in the sum of the products of the longest perpendicular diameters of all measurable lesions. Disease progression required ≥ 25% increase in measurable tumor or an increase in tumor size in patients with nonmeasurable but assessable disease. After partial response, tumor measurements exceeding 50% of the maximal extent of a previously observed reduction constituted progression. New lesions constituted progression. Patients who did not meet the definitions of response or progression were classified as having stable disease.
Time to progression (TTP) was calculated from study entry to disease progression. Patients were censored for TTP on initial treatment discontinuation. TTP for third-line therapy was calculated from the date of cross-over (as opposed to initial randomization). Patients who died within 30 days of treatment discontinuation were considered as having progressive disease. OS was calculated from enrollment to death or last contact. Without confirmatory data, patients who died or were lost to follow-up were assumed to have experienced disease progression at the time they were last documented to be progression free.
The primary objective of the trial was to test noninferiority in OS between the two treatment arms using a one-sided generalized Wilcoxon test. Analyses were based on initial treatment assignment, irrespective of crossing over to the other treatment. We assumed that the irinotecan arm (control arm) would demonstrate a median survival of 10 months. If at the time of the analysis, the one-sided P value was more than .05 in favor of the OS for the irinotecan arm,we would conclude that FOLFOX4 therapy is at least equivalent to irinotecan in regard to median survival. Under these assumptions, if the true hazard ratio (HR) for survival comparing FOLFOX4 with irinotecan was 1.33 (ie, median survival for FOLFOX4 is 7.8 months), there was a 5% chance of declaring noninferiority. A total of 560 patients were to be enrolled, with the analysis occurring after 405 patient deaths. All randomly assigned patients were included for efficacy analyses according to intention-to-treat principle. If the FOLFOX4 arm was determined to be noninferior, a subsequent Wilcoxon test was planned to test for superiority of FOLFOX4, using a two-sided test with P < .05 in favor of the FOLFOX arm to declare superiority. Secondary end points include TTP, objective tumor response rate (RR), and toxicity. Summary statistics, graphical methods, and frequency tables were used to assess toxicity and response.
In patients who crossed over to third-line therapy, additional exploratory analyses for the secondary end points were conducted. These analyses should be interpreted cautiously and descriptively, as comparisons between treatment arms based solely on the patients who crossed over were not protected by randomization. Quality-of-life data were collected but will be reported separately. All P values reported are two-sided, unless otherwise noted.
Between October1999 and December 2003, 491 patients were randomly assigned to receive second-line therapy with either irinotecan (n = 245) or FOLFOX4 (n = 246). The arms were well balanced with respect to stratification factors and other baseline characteristics (Table 2). Two hundred eighty-eight patients (59%) experienced treatment failure with FU therapy as first-line treatment of metastatic disease. With the cross-over design, 227 patients (46%) proceeded to third-line therapy. Six patients (three on irinotecan and three on FOLFOX4) canceled before receiving treatment. Eighteen patients randomly assigned to receive irinotecan and, having either ECOG PS 2, age ≥ 70 years, or prior pelvic radiotherapy, received an initial dose of irinotecan at 350 mg/m2 rather than a protocol-specified 300 mg/m2, constituting a major protocol violation.
Four hundred eighty patients completed a median of six cycles of second-line therapy (range, one to 69 cycles). Reasons for discontinuation included disease progression (62%), refusal (15%), adverse reactions (11%), and miscellaneous reasons, including alternative therapy and death during treatment (12%). The reasons for treatment discontinuation significantly differed between the two treatments: more patients discontinued FOLFOX4 as a result of adverse reactions (17% v 5%; P ≤ .0001), and more discontinued irinotecan because of disease progression (67% v 57%; P = .027). For the patients receiving FOLFOX4 who stopped treatment because of adverse reaction, 40% had grade 3 or worse neuropathy at the time of FOLFOX4 discontinuation.
Table 3 lists adverse events during initial protocol therapy considered at least possibly related to study treatment. Grade 3 or worse nausea, diarrhea (both P < .01), and febrile neutropenia (P = .03) were two to three times more common in the irinotecan arm, whereas neutropenia and paresthesias were more strongly associated with FOLFOX4 (P = .0012 and < .0001).
A total of 2.1% of patients died during treatment. Although a higher rate of grade 5 events were reported for patients treated with irinotecan (2.9% v 1.2%), the rates were not significantly different between the two treatments (P = .19). Sixteen patients died within 60 days of initiating treatment (4.5% irinotecan, 2% FOLFOX4; P = .13).
A total of 6.1% patients are still alive, with a median follow-up of 3.0 years (range, 0 to 5.6 years). Estimated median survival (Fig 1) was 13.8 months (95% CI, 12.2 to 15.0 months) for FOLFOX4-treated patients and 14.3 months (95% CI, 12.0 to 15.9 months) for patients on the irinotecan arm (HR = 0.92; 95% CI, 0.8 to 1.1; Wilcoxon P = .57; log-rank P = .38), clearly meeting the protocol-specified criteria for noninferiority in OS of FOLFOX4 compared with irinotecan. In multivariate Cox proportional hazards modeling, factors significantly impacting improved OS included ECOG PS of 0 or 1 (P = .028), baseline alkaline phosphate level less than grade 2 (P = .0029), and pulmonary indicator lesion (P = .015). In multivariate analyses, only the first two covariates identified in univariate analyses (Table 4) were significantly associated with OS. A nearly two-fold increased confirmed RR was seen with FOLFOX4 treatment (28.0% v 15.5%; P = .0009). TTP differed significantly (P = .0009;Fig 2) between FOLFOX4 (median, 6.2 months; 95% CI, 5.3 to 7.4 months) and irinotecan (median, 4.4 months; 95% CI, 3.3 to 5.5 months) treated patients (HR = 0.73; 95% CI, 0.6 to 0.9). Cox models identified treatment assignment as the only factor significantly associated with TTP (Table 4).
Forty-six percent (227 of 491) of patients experiencing treatment failure with second-line therapy crossed over to receive third-line therapy (ie, irinotecan to FOLFOX4, n = 130; FOLFOX4 to irinotecan, n = 97). No significant differences were detected in either patient characteristics or stratification factors recorded at initial randomization among patients who crossed over (Table 2). Ninety-five percent of patients started third-line treatment within 6 weeks of discontinuing second-line therapy. Patients completed a median of 14 cycles of third-line therapy (range, three to 60 cycles). Reasons for discontinuation included disease progression (68%), adverse reactions (12%), patient refusal (9%), and miscellaneous reasons, including alternative treatment, other medical problems, and death on study (11%).
All 227 patients receiving third-line therapy were assessable for analysis of adverse events. Adverse event rates for third-line therapy paralleled those for the same regimen as second-line therapy. There were no statistically significant differences between the third-line treatments in terms of adverse events, except for a trend toward more grade 3 or worse paresthesias in FOLFOX4-treated patients (P = .046).
In this nonrandomized comparison, a three-fold higher confirmed RR was observed in patients treated with third-line FOLFOX4 versus irinotecan (13.6% v 5.2%, respectively; P = .039). Patients achieving response to second-line therapy were twice as likely to experience response to third-line therapy (18% v 9%); this was of borderline significance (P = .07). Confirmed response was also independent of whether patients at initial randomization had experienced treatment failure with prior adjuvant therapy or had FU-refractory metastatic disease (9% v 10%, respectively; Breslow-Day test, P = .75).
With a median of 3.5 years of follow-up after cross-over (range, 0.4 to 4.9 years), 94% of patients have died since starting third-line therapy. In this cohort of patients, FOLFOX4-treated patients had significantly longer TTP (median, 5.8 months; 95% CI, 4.9 to 6.7 months; P = .0022) than those receiving irinotecan (median, 3.3 months; 95% CI, 2.6 to 4.1 months; HR = 0.65, 95% CI, 0.5 to 0.9). The estimated median OS for the entire study period among those who initiated third-line therapy did not differ for those who were initially treated with FOLFOX4 (15.9 months; 95% CI, 14.3 to 17.5 months; P = .7691) versus irinotecan (14.9 months; 95% CI, 13.2 to 17.5 months).
In this randomized cooperative group trial, second-line treatment with either irinotecan or FOLFOX4 resulted in similar survival when administered to patients with advanced colorectal cancer refractory to FU. Survival was noteworthy at approximately 14 months in both arms. For those patients able to receive third-line therapy, the median OS was an additional 10 months. In its entirety, OS was not significantly different between the two sequencing strategies. Treatment with the FOLFOX4 regimen led to statistically significant improvements in several clinical end points, including RR and TTP. These comparisons were made before the introduction of bevacizumab and cetuximab and the identification of the predictive role of KRAS .
Importantly, the toxicity profiles for the two regimens were as expected. Irinotecan was associated with significantly higher rates of diarrhea and vomiting, whereas FOLFOX4 was associated with a higher rate of neutropenia and neuropathy. Whether the choice of a better tolerated every-2-weeks irinotecan (180 mg/m2) regimen would have influenced these toxicity observations needs to be considered. Toxicities observed during second-line treatment were similar to those experienced during third-line treatment. In terms of overall tolerance, nearly half of the patients who experienced disease progression while receiving second-line therapy received third-line treatment. These individuals completed twice the median number of cycles in third-line treatment as the average patient on second-line therapy (12 v six cycles), clearly suggesting a selection effect among those willing and able to receive third-line therapy.
Our trial and two recently reported European studies15,16 provide insight into the actual percentages of patients able to initiate multiple lines of therapy. It has previously been observed in clinical trials that 50% to 70% of patients receive second-line therapy, and this is consistent within these three more recent trials (our trial, FOCUS [Fluorouracil, Oxaliplatin, and CPT-11 Use and Sequencing trial], and CAIRO [CApecitabine, IRinotecan, Oxaliplatin trial]; range, 52% to 62%). In addition, approximately 49% of these patients with second-line refractory disease (range, 41% to 57%), or approximately one quarter to one third of the patients initially treated, are able to proceed to third-line therapy. It is worth highlighting that roughly 30% to 35% of patients elect to receive no additional chemotherapy after progression with first-line treatment. Implicit in these observations is that patient selection influences whether individuals are sufficiently fit and can withstand the adverse events of all three lines of therapy. These data forewarn that not all patients who initiate first-line chemotherapy can continue on to second- and third-line treatments; the assumption that they will is tenuous.
The tolerability of specific sequences of therapy, for example, proceeding from an oxaliplatin- to an irinotecan-based regimen, or vice versa, was assessed in this trial. In a small study by Tournigand et al,17 treatment with FOLFOX followed by fluorouracil, leucovorin, and irinotecan (FOLFIRI) or the opposite sequence resulted in no significant OS difference. In that trial, roughly 68% of patients committed to the obligatory cross-over, which was completed in more patients randomly assigned to the sequence of FOLFIRI-FOLFOX (74% v 62%) than to FOLFOX-FOLFIRI. This finding was also observed in our trial in which more patients received FOLFOX after irinotecan versus the opposite sequence (57% v 43%). In the FOCUS trial, the number of patients who could receive the alternate regimen did not differ.15
Presently, debate continues regarding whether combination versus sequential chemotherapy in the first-line setting determines overall outcome. Although our trial and the two recent European trials offer some support for the sequential approach, final resolution of this issue requires the ability to identify patients whose disease will be adequately controlled by single-agent first-line therapy and who will be able to sufficiently tolerate multiple lines of therapy. Good performance status, less aggressive tumor behavior, and tolerance of side effects are likely attributes of these individuals. The FOCUS trial15 is similar in design to our trial in which patients who experienced treatment failure with initial FU monotherapy were randomly assigned to either single-agent irinotecan or FU combination regimens with irinotecan or oxaliplatin. A third group of patients were given combination chemotherapy as first-line and then second-line therapy. In the arms similar to our trial, the median survival did not differ. In addition, in comparing the second-line irinotecan and FU-oxaliplatin arms with those of our trial, the oxaliplatin-containing regimen again had a higher RR (23% v 11%; P < .001), although progression-free survival did not differ (4.8 v 4.3 months; P = .74). It is important to emphasize that the sequential approach pertains to patients for whom curative resection has been excluded. In addition, at present, it is not possible to a priori predict which patients will have a positive experience with first-line treatment that then allows them to proceed through multiple lines of treatment. It is these patients in whom a sequential approach may be appropriate—the challenge remains to identify them.
The incorporation of biologic agents, such as bevacizumab18 or cetuximab,19 into sequential strategies may further extend OS. For example, in the Tournigand et al study,17 the median delay between progression on first-line therapy and initiation of second-line therapy was only 2 to 3 weeks. In both the FOCUS and CAIRO trials, patients without disease progression or unacceptable toxicity were permitted to pause treatment for up to 3 months. The contribution of this flexibility to the longer median OS observed in the Tournigand et al17 study versus FOCUS and CAIRO is unknown. In our trial, in patients in whom data were available (54%), second-line therapy began within a median of 11 weeks and third-line therapy within 3 weeks. The impact of minimally toxic chemotherapy (ie, a fluoropyrimidine alone or with a biologic agent) as maintenance to extend these intervals requires further investigation. This approach when coupled with a sequential therapy maybe a more cost-effective, less toxic strategy for palliation of patients with poor PS and unresectable colorectal cancer.
In conclusion, this study demonstrates the noninferiority, in terms of OS, of FOLFOX4 to irinotecan as second-line therapy in patients with FU-refractory disease. In addition, OS is not significantly different based on sequencing strategies of second- and third-line regimens (FOLFOX4 followed by CPT-11 or vice versa). FOLFOX4 has superior activity on earlier end points, such as RR and TTP, regardless of line of therapy. The observed toxicities are similar to those of previous reports and are agent-specific. This North American cooperative group phase III trial confirms earlier observations that survival is optimized when patients with metastatic colorectal cancer are treated with all three active drugs.
Supported in part by Public Health Service Grants No. CA-25224, CA-37404, CA-35269, CA-60276, CA-52352, CA-63849, CA-37417, CA-35195, CA-35101, CA-35103, CA-35448, CA-35113, and CA-35415 from the National Cancer Institute Department of Health and Human Services.
This study was conducted as a trial of the North Central Cancer Treatment Group and Mayo Clinic.
Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
Clinical trial information can be found for the following: NCT00005036.
Although all authors completed the disclosure declaration, the following author(s) indicated a financial or other interest that is relevant to the subject matter under consideration in this article. Certain relationships marked with a “U” are those for which no compensation was received; those relationships marked with a “C” were compensated. For a detailed description of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.
Employment or Leadership Position: None Consultant or Advisory Role: Daniel J. Sargent, sanofi-aventis (C), Pfizer Inc (C); Edith Mitchell, Pfizer Inc (C), sanofi-aventis (C); Richard M. Goldberg, Pfizer Inc (C), sanofi-aventis (C)Stock Ownership: None Honoraria: Daniel J. Sargent, sanofi-aventis, Pfizer Inc; Edith Mitchell, Pfizer Inc, sanofi-aventis; Richard M. Goldberg, Pfizer Inc, sanofi-aventisResearch Funding: Edith Mitchell, Pfizer Inc, sanofi-aventis; Steven R. Alberts, Pfizer Inc, sanofi-aventisExpert Testimony: None Other Remuneration: None
Conception and design: George P. Kim, Daniel J. Sargent, Michelle R. Mahoney, Kendrith M. Rowland Jr, Richard M. Goldberg, Steven R. Alberts, Henry C. Pitot
Administrative support: George P. Kim, Daniel J. Sargent, Richard M. Goldberg
Provision of study materials or patients: George P. Kim, Kendrith M. Rowland Jr, Philip A. Philip, Richard M. Goldberg, Henry C. Pitot
Collection and assembly of data: George P. Kim, Daniel J. Sargent, Michelle R. Mahoney, Edith Mitchell, Abraham P. Mathews, Tom R. Fitch, Henry C. Pitot
Data analysis and interpretation: George P. Kim, Daniel J. Sargent, Michelle R. Mahoney, Kendrith M. Rowland Jr, Philip A. Philip, Edith Mitchell, Abraham P. Mathews, Tom R. Fitch, Richard M. Goldberg, Steven R. Alberts, Henry C. Pitot
Manuscript writing: George P. Kim, Daniel J. Sargent, Michelle R. Mahoney, Kendrith M. Rowland Jr, Tom R. Fitch, Steven R. Alberts, Henry C. Pitot
Final approval of manuscript: George P. Kim, Daniel J. Sargent, Michelle R. Mahoney, Kendrith M. Rowland Jr, Edith Mitchell, Abraham P. Mathews, Tom R. Fitch, Richard M. Goldberg, Steven R. Alberts, Henry C. Pitot