Background
Multicentre, open-label, parallel-design, randomised phase 3 trial at 75 UK hospital sites. 589 patients randomly assigned; 564 in the modified intention-to-treat population (284 standard-of-care, 280 irinotecan). Eligible patients were aged ≥18 years with MRI-defined locally advanced rectal cancer threatening or involving the resection margin, without metastases. 370 (66%) male; median age 61 years (IQR 54-68); mrT3 in 223 (79%) and 212 (76%) and mrT4 in 44 (15%) and 45 (16%) of the standard-of-care and irinotecan groups respectively. Rationale: small studies had reported high pathological complete response rates with irinotecan-fluoropyrimidine chemoradiation, whereas standard fluoropyrimidine intensification had not improved outcomes. EudraCT 2008-005782-59.
Results
Interventions and follow up: Arm A: Preoperative radiotherapy 45 Gy in 25 daily fractions + oral capecitabine 650 mg/m2 twice daily on weekdays + intravenous irinotecan 60 mg/m2 once weekly in weeks 1-4 (n=280)
Arm B: Preoperative radiotherapy 45 Gy in 25 daily fractions + oral capecitabine 900 mg/m2 twice daily on weekdays (n=284)
Primary endpoint: Disease-free survival (modified intention-to-treat)
mFollow up: 78 months (95% CI 75-86)
Results: 36-month DFS: 68% (95% CI 63-73) irinotecan vs 67% (95% CI 61-72) standard of care; HR 0.91 (95% CI 0.68-1.23), P=.54
Deaths (any cause): 88 (31%) vs 92 (32%); rectal cancer was the leading cause in 59 (67%) and 67 (73%) respectively
pCR: NR
OS: NR
Treatment delivery: received the full 45 Gy in 208/276 (75%) irinotecan vs 251/283 (89%) standard of care; received ≥90% of planned capecitabine in 187/276 (68%) vs 253/283 (89%)
Arm B: Preoperative radiotherapy 45 Gy in 25 daily fractions + oral capecitabine 900 mg/m2 twice daily on weekdays (n=284)
Primary endpoint: Disease-free survival (modified intention-to-treat)
mFollow up: 78 months (95% CI 75-86)
Results: 36-month DFS: 68% (95% CI 63-73) irinotecan vs 67% (95% CI 61-72) standard of care; HR 0.91 (95% CI 0.68-1.23), P=.54
Deaths (any cause): 88 (31%) vs 92 (32%); rectal cancer was the leading cause in 59 (67%) and 67 (73%) respectively
pCR: NR
OS: NR
Treatment delivery: received the full 45 Gy in 208/276 (75%) irinotecan vs 251/283 (89%) standard of care; received ≥90% of planned capecitabine in 187/276 (68%) vs 253/283 (89%)
Adverse events
Grade ≥3 (any): 215 (78%) irinotecan vs 148 (52%) standard of care
Laboratory/haematological investigations: 235 (85%) vs 109 (39%); lymphocyte count decreased 181 (66%) vs 100 (35%); neutrophil count decreased 27 (10%) vs 3 (1%)
Gastrointestinal: any GI event 57 (21%) vs 35 (12%); diarrhoea 38 (14%) vs 10 (4%)
Deaths related to protocol treatment: 3 irinotecan (two thromboembolic events, one multiorgan failure with sepsis) vs 2 standard of care (one cardiac arrest, one febrile neutropenia)
Laboratory/haematological investigations: 235 (85%) vs 109 (39%); lymphocyte count decreased 181 (66%) vs 100 (35%); neutrophil count decreased 27 (10%) vs 3 (1%)
Gastrointestinal: any GI event 57 (21%) vs 35 (12%); diarrhoea 38 (14%) vs 10 (4%)
Deaths related to protocol treatment: 3 irinotecan (two thromboembolic events, one multiorgan failure with sepsis) vs 2 standard of care (one cardiac arrest, one febrile neutropenia)
Conclusions
Adding weekly irinotecan to capecitabine-based preoperative chemoradiotherapy did not improve disease-free survival in MRI-defined high-risk locally advanced rectal cancer, with 36-month DFS essentially identical between arms. Grade ≥3 toxicity rose from 52% to 78%, and patients in the irinotecan arm were substantially less likely to complete planned radiotherapy or capecitabine. Irinotecan should not be combined with radiotherapy plus capecitabine in this setting.
Key Limitations
Open-label with no masking of patients or treating clinicians, though the primary endpoint (DFS) is relatively objective. The comparison is confounded by design: capecitabine was dose-reduced to 650 mg/m2 in the irinotecan arm versus 900 mg/m2 in the control arm, so the trial tests a regimen rather than the isolated addition of irinotecan. Delivery was markedly worse in the experimental arm (75% vs 89% received full-dose radiotherapy; 68% vs 89% received ≥90% of planned capecitabine), meaning toxicity may have undercut any biological benefit. Accrual ran from 2011 to 2018 and therefore predates routine adoption of total neoadjuvant therapy and organ-preservation strategies, so the control arm no longer reflects current practice. Missing radiological and pathological response assessments were classified as non-response, a conservative but potentially distorting convention; pCR and OS are not reported in the primary publication. A separate post-hoc analysis using AI-derived tumour cell density reported differential benefit in high-density tumours, but that is hypothesis-generating and should not be used to select patients.
Clinical Context
ARISTOTLE joins a consistent run of negative chemoradiotherapy-intensification trials in locally advanced rectal cancer, reinforcing that adding cytotoxic agents to fluoropyrimidine-based chemoradiation increases toxicity without improving disease control. Contemporary standard of care has moved to total neoadjuvant therapy — either induction or consolidation chemotherapy around long-course chemoradiotherapy or short-course radiotherapy (RAPIDO, PRODIGE 23) — with non-operative watch-and-wait management for patients achieving a complete clinical response (OPRA). NCCN, ASCO, and ESMO recommend TNT approaches for high-risk disease and do not endorse irinotecan-containing chemoradiation. No regulatory action follows from this trial; its practical value is to close off irinotecan-based chemoradiation as an intensification strategy outside a clinical trial.