Sequencing fruquintinib and trifluridine/tipiracil to improve outcomes in the real-world management of MMR-proficient, chemo-refractory colorectal cancer in Hong Kong.

173 Background: The optimal treatment sequence of chemo-refractory MMR-proficient colorectal cancers (CRC) is unknown. Trifluridine/tipiracil, a fixed-dose nucleoside analogue/thymidine phosphorylase inhibitor and fruquintinib, a VEGFR1-3 kinase inhibitor, are FDA-approved, but the effect of sequence with respect to clinical outcomes remains unreported. Methods: We retrospectively identified patients who received fruquintinib at Queen Mary Hospital Hong Kong between 2021 and 2025 through structured electronic medical records and collected clinicopathological data including sequence of therapy (if applicable), tumor sidedness, RAS-mutation status, and overall survival (OS). Among patients who received sequential trifluridine/tipiracil (L) and fruquintinib (F) (L-F or F-L), we report OS and real-world progression free survival 1 and 2 (rwPFS1, rwPFS2) defined as time from treatment sequence initiation to first progression and second progression/death respectively. Kaplan–Meier methods and log-rank tests were used for survival analyses. Results: Between 2021 and 30/6/2025, 78 patients (43 male, 35 female) received one or more doses of fruquintinib at a median age of 65.5 (range 30.3-89.4). All patients had chemotherapy and targeted therapy for recurrent/metastatic disease: including 87.2% prior anti-VEGF use, 12.8% prior regorafenib, and a KRAS/NRAS mutation frequency of 51.2%. At a median of 3.5 treatment lines, median OS was 8.25 months (95%CI 5.91 -9.40) among fruquintinib users. Of 30 patients treated sequentially with L-F or F-L (22 vs. 8; median treatment line 3.0 vs. 3.5), median rwPFS1 was 4.1 vs. 3.7 months (P = 0.16), and rwPFS2 was 9.5 vs. 4.7 months (P = 0.003). Median OS was 10.8 months (95%CI 9.56-21.6); 14.6 vs 9.3 months (P=0.003) for L-F and F-L respectively. Notably, 12 of 22 patients in the L-F arm (vs. 0 of 8 in F-L) received concurrent trifluridine/tipiracil with bevacizumab, which was associated with superior rwPFS2 (10.1 vs. 7.1months; P = 0.006) and OS (21.7 vs. 10.2 months; P = 0.05). Among patients receiving L-F or F-L as monotherapy (10 vs. 8), rwPFS2 (7.1 vs. 4.7 months; P = 0.08) was not significantly different, while OS (10.8 vs. 9.3 months; P = 0.018) was significantly longer with L-F. Conclusions: In our cohort of pMMR chemorefractory CRC, sequencing trifluridine/tipiracil followed by fruquintinib as monotherapy was associated with superior overall survival. The longest PFS2 and OS were observed in patients who received trifluridine/tipiracil plus bevacizumab followed by fruquintinib. Further analyses exploring sequence outcomes in relation to prior VEGF exposure will be presented.