The per-protocol population analyzed was defined as all randomized patients who had no major violations of eligibility criteria or study conduct, and who received at least 1 dose of treatment. TOSCA indicates the Three or Six Colon Adjuvant randomized clinical trial.
FOLFOX/CAPOX indicates fluorouracil, leucovorin, and oxaliplatin/capecitabine plus oxaliplatin.
CAPOX indicates capecitabine plus oxaliplatin; FOLFOX, fluorouracil, leucovorin, and oxaliplatin; ECOG PS, Eastern Cooperative Oncology Group performance status; and HR, hazard ratio. Dashed line indicates predefined noninferiority boundary (HR, 1.2).
eTable 1. Patients and Tumor Characteristics
eTable 2. Relapse-Free Survival, Cox Regression Models—Subgroup Analysis
eAppendix. All Italian Cooperative Groups (With Their Associated Centers) That Participated in the Global Enrollment of the TOSCA trial
Data Sharing Statement
Customize your JAMA Network experience by selecting one or more topics from the list below.
Petrelli F, Labianca R, Zaniboni A, et al. Assessment of Duration and Effects of 3 vs 6 Months of Adjuvant Chemotherapy in High-Risk Stage II Colorectal Cancer: A Subgroup Analysis of the TOSCA Randomized Clinical Trial. JAMA Oncol. 2020;6(4):547–551. doi:10.1001/jamaoncol.2019.6486
What is the optimal duration of oxaliplatin-based adjuvant chemotherapy with capecitabine plus oxaliplatin (CAPOX) or fluorouracil, leucovorin, and oxaliplatin (FOLFOX) for patients with high-risk stage II resected colorectal cancer?
In this subgroup analysis of a randomized clinical trial of 1254 patients with high-risk stage II resected colorectal cancer, 5-year relapse-free survival was 82.2% for the 3-month arm and 88.2% for the 6-month arm, with noninferiority not shown. Differences in favor of 6-month administration for CAPOX were 0.76% and were 8.56% for FOLFOX although in the 3-month arm, the treatment was significantly less toxic.
Although noninferiority of the 3-month arm was not shown, a 3-month CAPOX regimen or a 6-month FOLFOX regimen can be selected when choosing adjuvant chemotherapy in stage II colorectal cancer.
The addition of oxaliplatin to the standard 6-month fluorouracil-based adjuvant chemotherapy in stage II colorectal cancer has been reported to reduce the risk of relapse although it does not increase survival. The Three or Six Colon Adjuvant (TOSCA) trial compared 3 months with 6 months of adjuvant fluoropyrimidine and oxaliplatin-based chemotherapy in patients with stage III colon cancer. The utility remains unknown.
To assess the noninferiority and toxic effects of 3 vs 6 months of FOLFOX (fluorouracil, leucovorin, and oxaliplatin) or CAPOX (capecitabine plus oxaliplatin) adjunct chemotherapy among patients with high-risk stage II resected colorectal cancer enrolled in the TOSCA trial.
Design, Setting, and Participants
The TOSCA study was a noninferiority phase 3 randomized clinical trial conducted from June 2007 to March 2013 in 130 Italian centers. Included patients had resected colorectal cancer located 12 cm from the anal verge by endoscopy or above the peritoneal reflection at surgery. In this preplanned study assessing the per-protocol population, 5-year relapse-free survival was evaluated in 1254 patients with high-risk stage II resected colorectal cancer who had received adjuvant FOLFOX (fluorouracil, leucovorin, and oxaliplatin) or CAPOX (capecitabine plus oxaliplatin).
Patients were originally randomized (1:1) in the TOSCA trial to receive 3 months (experimental group) or 6 months (control) of standard doses of FOLFOX or CAPOX at the discretion of the treating physician.
Main Outcome and Measures
A hazard ratio of at least 1.2 between the 3-month and 6-month chemotherapy groups was set to reject the null hypothesis of noninferiority.
Overall, 1254 patients (mean [SD] age, 62.4 [9.8] years; 565 women [45.1%]) with clinical high-risk stage II resected colorectal cancer were analyzed at a median follow-up of 62 months (interquartile range, 53-71) months. Of them, 301 patients (24.0%) had pT4N0M0 tumors, and the remaining 953 patients (76.0%) had high-risk pT3N0M0 tumors; 776 patients (61.9%) received FOLFOX and 478 (38.1%) received CAPOX. The 5-year relapse-free survival was 82.2% for the 3-month arm and 88.2% for the 6-month arm, with an estimated hazard ratio of 1.41 (95% CI, 1.05-1.89; P = .86 for noninferiority). For CAPOX, the 5-year relapse-free survival was similar in the 2 arms (difference, 0.76% favoring the 6-month arm; 95% CI, −6.28% to 7.80%), whereas for FOLFOX, the difference was pronounced: 8.56% in favor of the longer-duration arm (95% CI, 3.45%-13.67%). Nevertheless, the test for an interaction between duration and regimen was not statistically significant. Neurotoxicity was approximately 5 times lower in the shorter duration arm than in the longer duration arm.
Conclusions and Relevance
In the 3-month arm, the treatment was significantly less toxic than in the 6-month arm. Noninferiority was not shown for 5-year relapse-free survival. However, a possible regimen effect was observed, suggesting that either 3 months of CAPOX or 6 months of FOLFOX therapy can be used whenever an oxaliplatin doublet is indicated for treatment of patients with stage II colorectal cancer.
ClinicalTrials.gov Identifier: NCT0064660
The high cure rate with surgery alone for stage II colon cancer limits the absolute benefit of fluorouracil-based adjuvant chemotherapy to approximately 5% in high-risk colorectal cancer (CC).1,2 The addition of oxaliplatin to the standard 6-month fluorouracil-based adjuvant chemotherapy in stage II CC reduces the risk of relapse although it does not increase survival in the first 6 to 10 years after surgery.3,4
Given this background, guidelines suggest discussing and offering oxaliplatin-based adjuvant chemotherapy to patients with high-risk stage II CC.5,6 A recent study (the International Duration Evaluation of Adjuvant [IDEA] collaboration) have examined the duration of adjuvant therapy in high-risk stage III CC.7 In the primary analysis of the Italian Three or Six Colon Adjuvant (TOSCA) trial that contributed to the IDEA combined analysis, the hazard ratio (HR) for relapse-free survival (RFS) in the stage III subgroup favored the 6-month arm.8,9 Here we report a preplanned subgroup analysis of patients with stage II CC enrolled in the Italian TOSCA trial.
The TOSCA study was an open-label, multicenter, noninferiority phase 3 trial involving patients with resected CC located 12 cm from the anal verge by endoscopy or above the peritoneal reflection at surgery. After stratification by center and stage, patients were randomized in a 1:1 ratio to receive 3 months (experimental group) or 6 months (control) of standard doses of FOLFOX (fluorouracil, leucovorin, and oxaliplatin) or CAPOX (capecitabine plus oxaliplatin). Patients were included with histologically confirmed high-risk stage II CC, defined as the presence of at least one of the following criteria: pT4 tumor; G3; bowel obstruction or perforation; vascular, lymphatic, or perineural invasion on histologic specimens; and fewer than 12 nodes examined.8-10 The primary end point was RFS, assessed from the time of treatment allocation to relapse or to death from any cause, whichever came first. An HR of 1.2 or higher between the 3-month and 6-month treatment groups was set to reject the null hypothesis of noninferiority. The per-protocol population analyzed was defined as all randomized patients who had no major violations of eligibility criteria or study conduct, and who received at least 1 dose of treatment. Time-to-event curves were described by the Kaplan-Meier method. A Cox proportional hazards model was used to estimate and test the treatment effect. Analysis of the interaction between treatment effects and selected subgroups was conducted using a χ2 test for heterogeneity and described with forest plots. The trial protocol is available in Supplement 1. This study followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline. Approval for the trial was obtained from local ethics committee for each participating site. All patients provided written informed consent in a manner consistent with the regulations of Italy before participating in the study. No one received compensation or was offered any incentive for participating in this study.
From June 2007 to March 2013, 3614 patients, recruited by 130 Italian centers, were included in the primary analysis of TOSCA trial. Of these, 1254 patients (overall mean [SD] age, 62.4 [9.8] years; 565 (45.1%) women) had clinical high-risk stage II CC: 301 patients (24.0%) had pT4N0M0 tumors and the remaining 953 patients (76.0%) had high-risk pT3N0M0 tumors. In total, 621 patients (49.6%) received 3 months of chemotherapy and 633 patients (50.4%) received 6 months of chemotherapy (Figure 1; eTable 1 in Supplement 2). Overall, 776 patients (61.9%) received the FOLFOX regimen, and 478 patients (38.1%) received the CAPOX regimen. When the patients and tumor characteristics were analyzed by the type of treatment received (FOLFOX or CAPOX), no differences were observed between the populations.
Treatment compliance and toxic effects of the TOSCA study have already been published.9 The treatment completion rate was 87% (3-month arm) vs 56% (6-month arm). Median treatment durations were 6.0 and 3.0 months in the FOLFOX group and 5.6 and 2.8 months in the CAPOX group. The rates of severe toxic effects were higher in the 6-month arm than in the 3-month arm, especially for neutropenia (27.6% vs 20.7%, P < .001), diarrhea (6.4% vs 5.0%, P < .001), and allergic reactions (2.0% vs 0.5%, P < .001). Grade 3 to 4 neuropathy was higher in the 6-month arm vs the 3-month arm (8.4% vs 1.3%, P < .001).
At a median follow-up of 62 months (interquartile range, 53-71 months), 138 relapses, 109 deaths, and 183 relapses or deaths were observed. The HR for RFS of the 3-month vs 6-month arms was 1.41 (95% CI, 1.05-1.89; P = .02 for superiority), and the 95% CI crossed the noninferiority limit of 1.20; thus, the null hypothesis of noninferiority of the 3-month arm could not be rejected. The absolute 5-year RFS difference between the arms was 6.0% (95% CI, 1.9%-10.2%; 88.2% for the 6-month arm vs 82.2% for the 3-month arm; noninferiority margin for 5-year difference, 2.2%) (Figure 2). The HR of 3 vs 6 months was 1.13 (95% CI, 0.70-1.84) when CAPOX was used as the regimen, with a similar 5-year RFS for the 3-month (84.6%) and 6-month (85.4%) arms (5-year RFS difference, 0.8%; 95% CI, −6.3% to 7.8%; noninferiority margin for 5-year difference, 2.8%). Conversely, the effect of treatment duration in patients treated with FOLFOX was pronounced (HR, 1.58; 95% CI, 1.09-2.28), with 5-year RFS of 81.1% for the 3-month arm and 89.6% for the 6-month arm (5-year RFS difference, 8.6%; 95% CI, 3.4%-13.7%; noninferiority margin for 5-year difference, 3.3%).
A statistically significant interaction was found between treatment and sex (female: HR, 1.03; 95% CI, 0.71-1.49; vs male: HR, 2.39; 95% CI, 1.45-3.95; P = .008) and between treatment and tumor site (ascending HR, 0.87; 95% CI, 0.55-1.37; descending HR, 2.11; 95% CI, 1.339-3.19; and multiple HR, 1.60; 95% CI, 0.38-6.70; P = .01; eTable 2 in Supplement 2; Figure 3). Median overall survival was not reached in either arm.
The most relevant results of the TOSCA trial assessing patients with resected high-risk stage II colon cancer are summarized as follows: (1) the study was not able to demonstrate the noninferiority of 3-month treatment duration arm; (2) toxic effects were substantially reduced by the shorter (3 vs 6 months) treatment duration; (3) a possible regimen effect was observed here as it was in all other IDEA studies; (4) the difference in favor of 6 months’ treatment duration was relatively large (6%) in stage II CC, much larger than that reported in the same TOSCA study for patients with stage III CC; and (5) there was a significant interaction of sex and of tumor site with treatment duration.
There are several potential explanations for the main findings of the TOSCA trial for the population with stage II vs stage III CC. We may reasonably postulate that stage II needs a longer adjuvant treatment than stage III. The study by Naxerova et al10 provides a rationale for our counterintuitive results: in node-negative cancers, malignant cells may move directly from the primary site to colonize distant sites, and these cells may be more resistant to chemotherapy; therefore, more prolonged chemotherapy administration may be more effective. Other authors have reported that patients with pT4N0M0 CCs have worse survival than patients with stage IIIA CCs.11-13 The former may be relatively more aggressive in terms of grade, size, higher proportions of perineural invasion, microsatellite instability, and obstruction or perforation.11-13 Another potential explanation for these findings may be associated with the main differential drug regimen effect. Because the outcome difference between the FOLFOX duration arms is relatively large, it is highly plausible that the overall results of the study are mainly driven by the FOLFOX subgroup. In addition, our hypothesis is that microsatellite instability may represent a major factor in the neutral effect of duration seen in stage II, right-sided CCs. Conversely, sex is a significant unfavorable factor of survival in patients with CC; therefore, males may need a more prolonged chemotherapy duration.14 Nevertheless, results of subgroup analyses must be interpreted with caution because the study was not powered for these analyses.
These data are in line with those of the overall IDEA combined analysis of stage II CC disease.12 Despite this, the utility of oxaliplatin in stage II CC remains unclear, and the choice between 6 months of fluoropyrimidine-based chemotherapy and 3 or 6 months of oxaliplatin-based chemotherapy must be made on an individual basis.
In this study, in the stage II CC subgroup of the TOSCA trial, noninferiority was not shown for RFS. Similar to what has been reported in the IDEA combined analysis for stage III CC, a possible regimen effect was also observed in the present study, suggesting that either 3 months of CAPOX or 6 months of FOLFOX treatment can be used whenever an oxaliplatin doublet is indicated for use in patients with stage II CC.
Accepted for Publication: November 6, 2019.
Corresponding Author: Fausto Petrelli, MD, Medical Oncology Unit, Medical Science Department, ASST Bergamo Ovest, Piazzale Ospedale 1, 24047 Treviglio (BG), Italy (email@example.com).
Published Online: February 13, 2020. doi:10.1001/jamaoncol.2019.6486
Author Contributions: Dr Roberto Labianca had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Petrelli, Labianca, Zaniboni, Rulli, Zampino, Marchetti, Sobrero.
Acquisition, analysis, or interpretation of data: Labianca, Zaniboni, Lonardi, Galli, Rulli, Rosati, Corallo, Ronzoni, Cardellino, Mattioli, Mambrini, Ciuffreda, Banzi, Pusceddu, Maiello, Zampino, Zagonel, Marchetti, Corsi, Rimassa, Cinieri, Sobrero.
Drafting of the manuscript: Petrelli, Labianca, Galli, Rulli, Corallo, Mambrini, Pusceddu, Marchetti, Cinieri, Sobrero.
Critical revision of the manuscript for important intellectual content: Petrelli, Labianca, Zaniboni, Lonardi, Galli, Rulli, Rosati, Ronzoni, Cardellino, Mattioli, Ciuffreda, Banzi, Maiello, Zampino, Zagonel, Marchetti, Corsi, Rimassa, Sobrero.
Statistical analysis: Labianca, Rulli, Zagonel, Sobrero.
Obtained funding: Labianca, Sobrero.
Administrative, technical, or material support: Lonardi, Corallo, Zagonel, Rimassa.
Supervision: Petrelli, Labianca, Zaniboni, Lonardi, Galli, Rosati, Ciuffreda, Zampino, Marchetti, Cinieri, Sobrero.
Conflict of Interest Disclosures: Dr Labianca reported receiving grants from Agenzia Italiana del Farmaco (AIFA) during the conduct of the study and receiving personal fees from Roche, Merck & Co, and Servier outside the submitted work. Dr Zaniboni reported receiving grants from Bayer, Eli Lilly and Company, Saboti, and Servier outside the submitted work. Dr Lonardi reported being a paid consultant and receiving research funding from Amge; being a paid consultant, receiving research funding, and serving on the speakers bureau for Merck Serono; being a paid consultant, and serving on the speakers bureau for Eli Lilly and Company; and serving on the speakers bureau for Roche, Bristol-Myers Squibb, and Servier. Drs Galli and Rulli reported receiving grants from AIFA during the conduct of the study. Dr Zagonel reported receiving personal fees and nonfinancial support from Bristol-Myers Squibb; personal fees and nonfinancial support from Merck & Co, Celgene; Bayer, and Roche; and receiving personal fees from Janssen and from Pfizer outside the submitted work. Dr Rimassa reported receiving personal fees from AbbVie, Amgen, AstraZeneca, Bayer, Baxter, Celgene, Eisai Co Ltd, Eli Lilly and Company, Exelixis, Gilead Sciences Inc, Hengrui Therapeutics, Incyte, Italfarmaco SpA, Merck Sharp & Dohme Corp, Roche, and Sanofi outside the submitted work; and receiving personal fees and nonfinancial support from Arqule and from Ipsen outside the submitted work. No other disclosures were reported.
Funding/Support: This study was sponsored by the Italian Group for the Study of Digestive Tract Cancers (GISCAD) Foundation and supported by the grant FARM5RWTWZ from Agenzia Italiana del Farmaco.
Role of the Funder/Sponsor: The GISCAD Foundation supported the trial in design and conduct of the study; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: The Gruppo Italiano per lo Studio dei Carcinomi dell’Apparato Digerente–Ufficio Operativo members Luciano Frontini, MD, Silvia Rota, and Lorena Cozzi provided administrative support. No one was financially compensated for the stated contribution.
Data Sharing Statement: See Supplement 3.
Additional Information: Istituto di Ricerche Farmacologiche Mario Negri IRCCS (Drs Galli and Rulli) conducted the collection, management, analysis, and interpretation of the data. The institute has access to all data and analyses. All Italian cooperative groups (with their associated centers) that participated in the global enrollment of the TOSCA trial are given in the eAppendix in Supplement 2.