Later-line treatment for metastatic colorectal cancer remains challenging, particularly for patients who have already received multiple systemic therapies. While the SUNLIGHT trial established trifluridine/tipiracil (FTD/TPI) plus bevacizumab as a new later-line standard in refractory mCRC, real-world evidence is needed to understand how this regimen performs in broader, less selected patient populations.
The study was published in ESMO Gastrointestinal Oncology in June 2026.
The article was titled “Effectiveness of FTD/TPI plus bevacizumab and impact of prior bevacizumab exposure in patients with mCRC: the Italian FLOWER study.”
Authors: G. Trovato, M. Carullo, A. Nicastro, C. Spoto, E. Perissinotto, G. Giordano, E. Di Giacomo, M.A. Calegari, V. Pasquinucci, T. Troiani, S. Tamberi, M. Schirripa, M. Scartozzi, F. Morano, G. Arrivi, C. Morelli, M. Basso, M. Landi, E. Martinelli, F. Zoratto, F. Bergamo, M. Landriscina, J. Lucchetti, C. Cremolini, G. Tortora, and L. Salvatore.
The Italian FLOWER study evaluated the effectiveness and safety of FTD/TPI plus bevacizumab in patients with refractory metastatic colorectal cancer treated across 15 Italian centers. It also explored whether prior bevacizumab exposure was associated with outcomes on this later-line combination.
Why this study matters
Treatment options become increasingly limited for patients with refractory metastatic colorectal cancer. The phase III SUNLIGHT trial established FTD/TPI plus bevacizumab as a later-line standard by showing improved progression-free and overall survival compared with FTD/TPI alone. However, patients enrolled in clinical trials are often more selected than those treated in routine oncology practice.
The FLOWER study was designed to address this gap by evaluating the effectiveness and safety of FTD/TPI plus bevacizumab in a broad, unselected real-world population. The study also explored whether previous exposure to bevacizumab was associated with outcomes with this combination.
Read more about Lonsurf (trifluridine/tipiracil) on OncoDaily.
Methods
FLOWER was a retrospective, multicenter study including patients with refractory metastatic colorectal adenocarcinoma treated with FTD/TPI plus bevacizumab across 15 Italian centers. Patients received FTD/TPI at a starting dose of 35 mg/m² orally twice daily on days 1-5 and 8-12 of each 28-day cycle, combined with bevacizumab 5 mg/kg intravenously every 14 days.
The primary endpoint was overall survival (OS). Secondary endpoints included progression-free survival (PFS), investigator-assessed overall response rate according to RECIST v1.1, and safety.
Patient Population
A total of 297 patients were included. The median age was 61.8 years, and 41.1% of patients were aged 65 years or older. Most patients had good performance status, with 53.5% having ECOG PS 0 and 40.4% having ECOG PS 1, while 6.1% had ECOG PS 2 — a subgroup that had been excluded from the SUNLIGHT trial.
The cohort reflected a clinically heterogeneous real-world population. Most patients had multiple metastatic sites, reported in 79.1% of the cohort. The liver and lung were the most frequent metastatic locations, reported in 62.9% and 52.5% of patients, respectively.
Prior treatment exposure was substantial. Patients had received a median of two previous systemic treatment lines, and 35.4% had received more than two previous lines. Prior bevacizumab exposure was very common, documented in 89.9% of patients.
Key Findings
FTD/TPI plus bevacizumab showed clinically meaningful activity in this real-world cohort. The overall response rate was 7.4%, with 22 patients achieving partial response. Stable disease was reported in 46.8% of patients, resulting in a disease control rate of 54.2%.
After a median follow-up of 11.0 months, median PFS was 5.3 months. After a median follow-up of 13.1 months, median OS was 11.9 months. On multivariate analysis, four factors remained independently associated with overall survival: ECOG performance status, tumor sidedness, number of metastatic sites, and time from metastatic diagnosis to treatment initiation. Better ECOG performance status was consistently linked with longer PFS and OS. Patients with ECOG PS 0 had a median OS of 14.3 months, compared with 9.0 months for patients with ECOG PS ≥1.
Tumor sidedness also remained clinically relevant. Patients with right-sided tumors had inferior outcomes compared with those with left-sided tumors, with median OS of 8.8 months versus 13.3 months, respectively. Having two or more metastatic sites was associated with worse survival. Patients with at least two metastatic sites had a median OS of 11.5 months, compared with 16.0 months among those with a single metastatic site.
The interval from metastatic diagnosis to treatment initiation was another important prognostic factor. Patients who started FTD/TPI plus bevacizumab more than 18 months after metastatic diagnosis had longer PFS and OS than those who started within 18 months.
Prior Bevacizumab Exposure
One of the key questions addressed by FLOWER was whether previous bevacizumab exposure reduced the effectiveness of FTD/TPI plus bevacizumab.
Most patients in the cohort had already received bevacizumab before starting this regimen. However, prior bevacizumab use was not statistically associated with worse PFS or OS. Specifically, bevacizumab exposure in the first-line setting, as part of second-line treatment, and the length of the bevacizumab-free interval before starting FTD/TPI plus bevacizumab were each not significantly associated with outcomes.
These findings support the rationale for continued VEGF pathway inhibition in later-line mCRC, including in patients previously treated with bevacizumab-containing regimens.
Activity Across KRAS Subgroups
The combination showed similar activity across RAS molecular subgroups, with no statistically significant difference in outcomes based on KRAS/NRAS mutational status. This finding aligns with a recent post-hoc analysis of SUNLIGHT, which showed that KRAS G12 mutations were neither prognostic nor predictive of reduced benefit from FTD/TPI plus bevacizumab — a particularly relevant observation given emerging KRAS-targeted therapies.
Safety
The safety profile was manageable and consistent with the known toxicity pattern of FTD/TPI plus bevacizumab. Any-grade adverse events occurred in 81.8% of patients, while grade 3-4 adverse events occurred in 52.2%. The most frequent toxicity was neutropenia, reported in 57.9% of patients, including grade 3-4 neutropenia in 35.4%.
Other common adverse events included fatigue, anemia, nausea, thrombocytopenia, and diarrhea. Non-hematologic toxicities were mostly low grade. Bleeding and thromboembolic events were uncommon.
ESMO Guideline Context: Later-Line Treatment Sequencing in mCRC
According to the updated ESMO Clinical Practice Guideline for metastatic colorectal cancer, treatment beyond the second line is increasingly guided by molecular profiling, prior therapy, and patient fitness. For molecularly unselected patients, trifluridine–tipiracil, with or without bevacizumab, represents a standard later-line approach, while regorafenib and fruquintinib are recommended options after progression.
The guideline also emphasizes targeted strategies for patients with actionable alterations, including anti-EGFR rechallenge in selected RAS wild-type patients guided by ctDNA, and therapies directed against HER2 amplification, KRAS G12C mutations, NTRK and RET fusions, or POLE/POLD1-mutated disease.
Read more about ESMO 2026 Guideline about Management of Metastatic Colorectal Cancer on OncoDaily.
Conclusion
The Italian FLOWER study provides important real-world evidence supporting FTD/TPI plus bevacizumab as an active and tolerable later-line treatment option for patients with metastatic colorectal cancer.
In a broad, unselected population, the combination achieved a median PFS of 5.3 months and median OS of 11.9 months, outcomes broadly consistent with those reported in the SUNLIGHT trial. Importantly, prior bevacizumab exposure did not appear to reduce treatment effectiveness.
These findings strengthen the real-world evidence for FTD/TPI plus bevacizumab and support its use as a later-line treatment option in metastatic colorectal cancer, including in patients previously treated with bevacizumab.