Post-Adjuvant Trifluridine/Tipiracil in ctDNA-Positive Colorectal Cancer

Key Takeaways

• Longer median disease-free survival was observed with trifluridine/tipiracil, but the primary endpoint was not met in the overall population.
• Hematologic toxicity was substantially more frequent with active treatment, and no treatment-related deaths were reported.
• Exploratory subgroup and ctDNA findings, secondary endpoints, patient-reported outcomes, and immature survival data added context without changing the main overall reading.

In the randomized, double-blind ALTAIR phase 3 trial within CIRCULATE-Japan, median disease-free survival was 9.30 months with trifluridine/tipiracil versus 5.55 months with placebo among patients with ctDNA-positive, radiographically disease-free resected colorectal cancer. The prespecified investigator-assessed disease-free survival endpoint was not met in the overall population, and the ctDNA-guided intervention strategy did not significantly improve the primary endpoint overall.

Patients had resected stage 0-IV colorectal cancer, became ctDNA positive after surgery, and had no radiological evidence of disease. Between July 2020 and June 2023, 243 patients were randomized 1:1 to trifluridine/tipiracil (n=122) or placebo (n=121) for 6 months. Treatment was given in 28-day courses, with oral trifluridine/tipiracil 35 mg/m2 per dose in the active arm. ctDNA testing used the tumor-informed Signatera assay, with positivity defined by plasma samples containing two or more SNVs and concentration reported in MTM/ml. Enrollment spanned MRD, on-treatment, and surveillance windows, with the MRD window defined as 2-10 weeks after surgery before adjuvant therapy when given. The trial therefore tested one intervention across heterogeneous ctDNA-detection time windows.

The primary endpoint was investigator-assessed disease-free survival. Numerical separation favored active treatment, with a hazard ratio of 0.79, a 95% confidence interval of 0.60-1.05, and P=0.107. Six-month disease-free survival rates were 70.5% with trifluridine/tipiracil and 45.5% with placebo. Statistical significance was not reached in the prespecified primary analysis, leaving the overall randomized comparison negative for the primary endpoint.

Exploratory subgroup analyses showed a stage IV signal, with HR 0.53 and P=0.012, whereas stage I-III disease showed HR 0.86 and P=0.375. These subgroup comparisons were exploratory and were not adjusted for multiplicity. DFS2 numerically favored trifluridine/tipiracil, with HR 0.78, 95% CI 0.58-1.02, and P=0.073, while ctDNA clearance after treatment was 17.2% versus 12.4%. ctDNA dynamics were associated with prognosis, and a post hoc blinded central radiology review was included as an exploratory sensitivity analysis. These analyses added context without overturning the primary result.

Safety findings showed the main tradeoff, as grade 3 or higher hematologic adverse events occurred in 73.0% of trifluridine/tipiracil-treated patients and 3.3% of placebo-treated patients. Any-grade adverse events occurred in 98.4% and 57.0%, respectively, while neutrophil count decrease and white blood cell count decrease reached 56.6% and 17.2% in the active arm. Dose reduction due to adverse events occurred in 37.7% of active-arm patients, while treatment completion rates were similar between groups and no treatment-related deaths were reported. EORTC QLQ-C30 and EQ-5D-5L assessments showed temporary worsening in global health status and quality of life during active treatment, with minimal differences after completion. Overall survival remained immature, with 24 events reported at a median follow-up of 23.29 months.