Daily aspirin did not improve disease-free survival after curative-intent treatment of colorectal cancer liver metastases and was linked to more serious adverse events in a randomized phase 3 trial published in The Lancet Gastroenterology & Hepatology.
Aspirin has shown preventive and antitumor effects in colorectal cancer, but its role after treatment of metastatic disease has been uncertain, the authors wrote. Among 428 patients who started treatment, serious adverse events occurred in 8% of those receiving aspirin vs. 2% of those receiving placebo.
“Aspirin is inexpensive and widely available, and observational and laboratory studies provided a strong rationale for benefit,” first author Sheraz Yaqub, MD, PhD, of Oslo University Hospital and the University of Oslo in Norway, told GI & Hepatology News. “However, when tested in a randomized, double-blind, placebo-controlled phase 3 trial in patients treated for colorectal cancer liver metastases, we found no clinical benefit. This also highlights that results from nonmetastatic colorectal cancer should not automatically be extrapolated to patients who have had metastatic disease.”
Dr. Yaqub and colleagues conducted the investigator-initiated ASAC trial at 14 hepatobiliary centers in Norway, Sweden, and Denmark. Eligible adults had undergone surgery, ablation, or both with curative intent for colorectal cancer liver metastases and had no evidence of remaining disease on imaging. Patients were excluded if they had contraindications to aspirin or a clinically relevant bleeding risk, or if they were receiving antithrombotic treatment incompatible with the protocol.
The investigators randomly assigned 466 patients to aspirin or placebo. The primary analysis included 428 patients who started treatment: 217 received aspirin and 211 received placebo. Their mean age was 62 years, and 64% were men. Patients received aspirin 160 mg or matching placebo once daily for up to three years or until recurrence or death. Surveillance included chest and abdominal CT scans every four to six months.
Enrollment stopped at 466 of a planned 800 patients because the prespecified number of primary endpoint events accrued faster than expected, the authors reported. Recruitment had also been slowed by the COVID-19 pandemic and by more patients than expected being ineligible because of ongoing antiplatelet or anticoagulant therapy.
The primary outcome was disease-free survival, measured from the start of treatment until colorectal cancer recurrence or death from any cause. Secondary outcomes included overall survival, time to recurrence, and safety; health-related quality of life, also a secondary outcome, will be reported separately. Median follow-up for overall survival was 1,067 days, or about 2.9 years.
A total of 269 disease-free survival events occurred: 138 with aspirin and 131 with placebo. Aspirin did not reduce the risk of recurrence or death. Median disease-free survival was 1.13 years with aspirin vs. 1.40 years with placebo. A secondary analysis of mean disease-free time over three years was consistent with the primary result.
Overall survival, a secondary outcome, was lower with aspirin. At 36 months, 76% of patients receiving aspirin were alive vs. 85% receiving placebo. There were 42 deaths with aspirin and 27 with placebo, most from metastatic colorectal cancer. Time to recurrence and the location of the first recurrence did not differ between the groups.
The authors noted that overall survival was a secondary endpoint the trial was not powered to assess, and that the finding should be interpreted cautiously.
More patients in the aspirin group had tumors with BRAF V600E mutations (8% vs. 3% of those tested); the authors described that mutation as carrying an established poor prognosis in colorectal cancer liver metastases. In a post hoc analysis adjusting for BRAF status, the overall survival estimate was modestly attenuated, though its direction was unchanged. The cause-of-death analysis also did not suggest that the survival difference was primarily driven by fatal treatment-related toxicity, according to the paper.
Among clinically relevant and serious adverse events, myocardial infarction, duodenal ulcer, and cerebral hemorrhage were each reported twice with aspirin and not with placebo. Nosebleeds were reported four times with aspirin and not with placebo, and cerebral infarction twice with placebo and not with aspirin. No deaths were considered related to treatment.
The trial also found no clear evidence that any molecular or clinical subgroup benefited from aspirin. Among patients with available tumor data, aspirin did not improve disease-free survival in those with mutated or wild-type PIK3CA or KRAS. However, molecular data were incomplete, and the study was not powered to detect differences within these subgroups.
Previous randomized trials, including ALASCCA, have suggested a benefit from adjuvant aspirin in patients with localized colorectal cancer and PI3K pathway alterations, and National Comprehensive Cancer Network guidelines now recommend it for stage II and III colon cancer with those alterations, the authors noted. They cautioned that benefits seen in localized disease should not be extrapolated to patients treated for metastatic disease.
The authors acknowledged several limitations. The trial was not powered for secondary outcomes or subgroup analyses, and PIK3CA status was available for only 61% of patients. In addition, 38 randomized patients never started treatment, data on treatment after recurrence were not collected, and follow-up was limited to three years. The Scandinavian setting also may limit the generalizability of the overall survival findings.
“These findings indicate that the anticancer effects of aspirin are context dependent and do not extend to the adjuvant treatment of resected metastatic colorectal cancer,” the authors concluded.
The Research Council of Norway, Norwegian Cancer Society, Therapy Research in the Specialist Health Services Norway, and Oslo University Hospital Fondsstiftelsen funded the trial. Dr. Yaqub reported lecture honoraria from Bayer and AstraZeneca and serving as a scientific board member of Node Pharma. Other authors reported relationships with companies including AstraZeneca, Novo Nordisk, Medtronic, Isofol Medical, and Serca Pharmaceuticals.
Expert insight
Dr. Yaqub spoke with GI & Hepatology News about the study findings.
How should these findings affect the use of aspirin after curative-intent treatment of colorectal cancer liver metastases?
Dr. Yaqub: Our findings do not support routinely starting aspirin to prevent recurrence after curative-intent treatment of colorectal cancer liver metastases. Importantly, our findings apply to starting aspirin for this indication and should not be interpreted as a reason to stop aspirin in patients who are already taking it for cardiovascular or other established medical indications.
How should physicians interpret the lower overall survival seen with aspirin, given that overall survival was a secondary endpoint and the trial was not powered for it?
Dr. Yaqub: The overall survival finding should be interpreted cautiously because it was a secondary endpoint and the trial was not powered to detect differences in overall survival. We also did not collect detailed information on treatment after recurrence. We therefore cannot establish that aspirin caused the observed difference in survival. Nevertheless, the finding should not be ignored and provides an additional reason not to initiate aspirin in this setting in the absence of demonstrated benefit.
Do the PIK3CA and KRAS subgroup findings suggest that biomarkers could still identify patients with metastatic colorectal cancer who might benefit from aspirin?
Dr. Yaqub: We found no evidence of benefit from aspirin according to PIK3CA or KRAS mutation status. However, these subgroup analyses were limited by sample size and do not exclude the possibility that other molecular features could identify biologically distinct subgroups. We are planning more comprehensive molecular analyses of tumor samples from ASAC to explore this further.