Phase II study of maintenance trifluridine/tipiracil (TAS-102) plus bevacizumab after induction chemotherapy in metastatic colorectal cancer.

Trial information
UMIN trial ID: UMIN000031317

Principal Investigators: Katsuya Ota https://orcid.org/0000-0003-4897-7032

Sponsor: None

IRB Approved: Yes

General toxicity profile
All patients were evaluable for safety using CTCAE v5.0 (Table 1). The observed major grade 3 or 4 hematologic toxicities were neutropenia (n = 38, 73%), white blood cell decreased (n = 18, 35%), platelet count decreased (n = 5, 10%) and anemia (n = 4, 8%). The grade 3 non-hematologic toxicities included hypertension (n = 4, 8%), proteinuria (n = 4, 8%), fatigue (n = 3, 6%) and fever (n = 2, 4%). There was no grade 3 or higher peripheral sensory neuropathy (any-grade incidence, n = 16, 31%) or thromboembolic event (any-grade incidence, n = 1, 2%). There was no treatment-related death.

Discussion
This prospective, multicenter study evaluated the efficacy and safety of a complete switch to first-line maintenance therapy (TAS+Bev) after induction chemotherapy (fluoropyrimidine plus oxaliplatin) in previously untreated patients with mCRC. The study was completed as planned. The strategy specifically omits both oxaliplatin and fluoropyrimidines during the maintenance phase, aiming to reduce cumulative toxicity while maintaining disease control. Maintenance therapy with TAS+Bev achieved a median PFS1 of 8.7 months, with a high DCR of 94.2% and RR of 59.6%. Oxaliplatin reintroduced in 53.8% of patients after maintenance therapy. Total first-line treatment duration was 16.2 months. The median OS among 52 patients was 37.1 months.
These findings suggest that the ability to sustain long-term maintenance treatment may not be easily predicted by conventional baseline factors, thereby needing further investigation into novel clinical or molecular predictors. Nevertheless, first-line maintenance therapy with TAS+Bev can effectively bridge induction and reintroduction phases while preserving therapeutic efficacy. This approach circumvents cumulative peripheral neuropathy, a common and often dose-limiting toxicity in oxaliplatin-containing chemotherapy that negatively affects quality of life and frequently necessitates early treatment discontinuation.1 By transitioning to TAS+Bev, patients were able to avoid neurotoxicity while remaining on active therapy.
Previous studies have explored similar de-escalation strategies. The GERCOR OPTIMOX1 trial introduced a “Stop-and-Go” approach with intermittent oxaliplatin administration, which prolonged the treatment duration without compromising efficacy.2 The OPTIMOX2 trial later emphasized the importance of maintenance therapy by demonstrating that complete treatment cessation had inferior outcomes.3 Likewise, the CAIRO34 and CCOG09025 studies supported maintenance strategies with capecitabine plus bevacizumab after induction therapy. Collectively, these findings highlight the clinical value of structured maintenance approaches. In refractory mCRC, the efficacy of TAS-102, a combination of trifluridine and tipiracil hydrochloride, has been established9. Moreover, its combination with bevacizumab has demonstrated synergistic efficacy in the C-TASK FORCE6 and SUNLIGHT7 trials, leading to improved PFS and OS compared to TAS-102 alone. In the TASCO1 study, this regimen was tolerable and effective even among patients unfit for intensive therapy, suggesting its utility beyond late-line settings. Remarkably, the present study revealed several patients are now continuing maintenance therapy beyond 2 years without disease progression. However, due to protocol-defined constraints, observation beyond this period exceeded the data cut-off.
From a biological standpoint, this strategy may also leverage mechanisms of drug sensitivity restoration. Sharma et al. demonstrated that chromatin remodeling plays a role in transient drug tolerance, suggesting that temporal withdrawal of certain agents might resensitize tumor cells to previously used therapies.8 Furthermore, drug rechallenges and treatment beyond progression have also been explored as valid therapeutic strategy in drug-resistant cancers.10 Our study builds on this foundation by introducing TAS+Bev earlier in the treatment course as a proactive maintenance strategy rather than a salvage option.
The strengths of this study include its prospective, multicenter design and the two-step registration protocol, which ensured that only patients with stable disease or better on imaging after induction were enrolled into the maintenance phase. This rigorous patient selection likely contributed to the robustness of the results. Conversely, in CAIRO3 study, reintroduction of CapeOX+Bev was permitted in both arms. In that study, 61% of patients in the observation group (ie, did not undergo chemotherapy) underwent reintroduction therapy, compared to only 47% in the Capecitabin plus bevacizumab maintenance group.4 This difference suggests that maintenance treatment can delay disease progression, consequently reducing the need for the early reintroduction of combination chemotherapy. In contrast, patients in the observation group, who discontinued therapy tended to progress earlier, thereby requiring reintroduction more frequently. These findings support the clinical utility of maintenance therapy in prolonging PFS and minimizing the need for more intensive treatment over time. Using this regimen allowed for the reintroduction of oxaliplatin in 53.8% of patients in this study, which is a higher rate than the Cap+Bev maintenance group.
The limitations of this study must be acknowledged. Despite being a prospective study, selection bias may still be present because patients transitioning to maintenance may have had more favorable prognostic features. The study also did not include a comparator arm, such as capecitabine plus bevacizumab or observation, limiting our ability to assess the relative efficacy of TAS+Bev. Furthermore, since the OS analysis is ongoing, no definitive conclusions regarding survival benefit can be drawn yet. As with other maintenance therapy trials, OS can be influenced by multiple confounding factors, including post-progression treatments and crossover.4
In previous clinical trials, approximately 68% to 79% of patients with mCRC transition from second-line to third-line,11–13 but when accounting for patients who deviate from first-line treatment protocols or discontinue early, around 40% to 60% of patients treated with multidrug regimens may not reach third- or fourth-line therapy. Thus, agents that are typically reserved for later-line therapy, such as TAS-102, are rarely ever utilized in such populations. Since this strategy switches away from fluoropyrimidine and oxaliplatin during the maintenance phase, key cytotoxic agents can be preserved for future use. Our OS analysis indicated that maintenance therapy with TAS+Bev was not detrimental, and when oxaliplatin was reintroduced, patients experienced notably improved outcomes. This highlights the clinical relevance of a structured treatment strategy that includes planned reintroduction after maintenance therapy. The favorable prognosis of this subgroup suggests that preserved sensitivity to previously administered agents contributes to the success of maintenance therapy. Nevertheless, the efficacy of this “induction–maintenance–reintroduction” strategy needs to be validated in a phase III randomized controlled trial. If proven non-inferior to conventional maintenance therapies and associated with an OS benefit, this strategy could offer a patient-centered, tolerable, and effective alternative to conventional maintenance strategies, especially for those unfit for long-term oxaliplatin-based therapy. In future seeking to replicate these results, a randomized design including active comparators could enhance the robustness of the findings, particularly in terms of OS and patient-reported outcomes. Additionally, future large-scale prospective trials are needed to confirm and establish TAS+Bev as a standard maintenance option, as well as to better define the optimal patient population using molecular or clinical biomarkers.