Colorectal cancer (CRC) is one of the most common gastrointestinal tract cancers in nowadays. It is the third prevalent malignancy and the second leading cause of cancer-related death worldwide. There are more than 1.9 million new patients diagnosed to have CRC and 935,000 deaths attributed to CRC in 2020 (1). Among these patients, more than half developed metastatic colorectal cancer (mCRC) eventually. The standard therapies for patients with mCRC includes chemotherapy regimens containing irinotecan, oxaliplatin, and fluoropyrimidines in combination with anti-vascular endothelial growth factor (VEGF) or anti-epidermal growth factor receptor (EGFR) antibodies (2). With these treatments, the overall survival of mCRC has improved gradually with an estimated median survival about 30 months and 5-year survival rate about 14% (3). However, the prognosis of mCRC drops sharply when in chemorefractory status. The median survival after chemorefractory is approximately 6 months. Hence, there is an urgent need to improve outcomes in patients with chemorefractory mCRC.

For chemorefractory mCRC, two oral agents, trifluridine/tipiracil (Lonsurf; TAS-102; Taiho Pharmaceutical, TTY Biopharm., Taiwan) and regorafenib (REG; Bayer AG, Berlin, Germany), have been proved as third to fourth-line treatment to prolong survival. TAS-102 is composed of an antineoplastic thymidine-based nucleoside analog (trifluridine) and an inhibitor of thymidine phosphorylase that degrades trifluridine (tipiracil) (4). The pivotal phase 3 RECOURSE trial compared TAS-102 with placebo in patients with refractory mCRC and demonstrated that TAS-102 significantly prolonged overall survival (OS) (7.1 months vs. 5.3 months, p < 0.001) and progression-free survival (PFS) (2.0 months vs. 1.7 months, p < 0.001) as compared with placebo (5). REG is an oral multi-kinase agent that inhibits activity of several stromal receptor tyrosine kinases associated with angiogenesis, oncogenesis, and the tumor microenvironment (6). The pivotal phase 3 CORRECT trial compared REG and placebo in patients with refractory mCRC and demonstrated that REG resulted in significantly longer OS (6.4 months vs. 5.0 months, p: 0.0052) and PFS (1.9 months vs. 1.7 months, p< 0.0001) as compared with placebo (7). Based on these results, both TAS-102 and REG gain the indication of refractory mCRC. Current guidelines also indicate that TAS-102 and REG are both effective regimens in later-line treatment of mCRC (2).

Nonetheless, there are no randomized control trials directly comparing the efficacy of TAS-102 and REG. Previous retrospective studies had published their prognosis of TAS-102 monotherapy and REG monotherapy. Kawakami et al. analyzed a nationwide database in which TAS-102 demonstrated significantly longer survival than REG (8), while other publications showed insignificant survival between TAS-102 and REG (9, 10). Moreover, recent evidences exhibited combination strategy of TAS-102 and REG with longer survival benefits numerically (11, 12). However, no comprehensive studies focused on the comparison between TAS-102 combination and REG combination. Given the inconclusive results, we conducted an observational study to compare the oncologic outcomes of TAS-102-containing regimen and REG-containing regimen in patients with mCRC.

To our best knowledge, this is the first study to demonstrate that combination is much better than monotherapy in mCRC patients treated with TAS-102 or REG. Previous literature all focused on the comparison between TAS-102 monotherapy and REG monotherapy. The phase 3 RECOURSE trial demonstrated that TAS-102 significantly prolonged survival (5) and the phase 3 CORRECT trial also demonstrated that REG resulted in significantly longer survival (7). The PFS was increased 0.3 months and the OS was increased 1.4 - 1.8 months. Although these two studies are significant, the survival differences were modest numerically. Moreover, Kuboki et al. conducted a phase 1/2 C-TASK FORCE trial to analyze the efficacy of TAS-102 in combination with bevacizumab (11). This study suggested that TAS-102 plus bevacizumab combination might become a potential treatment option in chemorefractory mCRC patients, with median PFS 5.6 months and median OS 11.4 months. Another phase Ib NIVOREGO trial also showed the combination of regorafenib plus nivolumab had a manageable safety profile and encouraging antitumor activity in patients with mCRC, with ORR 36% and median PFS 7.9 m (12). Our study was consistent with these conclusions. For total population, the median PFS were 3.7 months in TAS-102 and 2.0 months in REG (P = 0.006). The median OS were 9.2 months in TAS-102 and 6.8 months in REG (P = 0.048). After stratification, the survival is significantly longer in combination than monotherapy, no matter in TAS-102 or REG group. For patients treated with TAS-102, the median PFS and OS were 6.6 months versus 2.0 months (P < 0.001) and 16.7 months versus 6.5 months (P < 0.001) in combination and monotherapy groups, respectively. For patients treated with REG, the median PFS and OS were 4.8 months versus 1.8 months (P < 0.001) and 14.5 months versus 4.9 months (P < 0.001) in combination and monotherapy groups, respectively. Multivariate analysis demonstrated that combination therapy were independent predictor associated with better survival, no matter in TAS-102 or REG. Our study demonstrated that combination therapy can achieve the best prognosis, rather than monotherapy, in chemorefractory mCRC patients treated with TAS-102 or REG. Further prospective trials are warranted to confirm with our conclusions.

Targeted therapy-chemotherapy combinations have been recognized as the optimal regimens in patients with mCRC. You et al. conducted a comprehensive meta-analysis enrolling 16 first-line clinical trials to evaluate the efficacy between chemotherapy plus targeted therapies and chemotherapy alone in mCRC patients (13). The meta-analysis suggested that the right-sided mCRC patients benefited more from chemotherapy plus bevacizumab comparing with chemotherapy alone. Arnold et al. also performed an retrospective study to compare chemotherapy plus EGFR antibody therapy with chemotherapy alone in patients with left-side mCRC (14). This study demonstrated that a greater effect of chemotherapy plus EGFR antibody therapy was observed in comparison with chemotherapy alone for patients with left-side mCRC. Our results were consistent with these conclusions that the survival is significantly longer in combination than those in monotherapy, no matter in TAS-102 or REG group. TAS-102 plus targeted therapy has a greater clinical benefit than TAS-102 alone, and REG plus chemotherapy also has a longer survival than REG alone. Immunotherapy is emerging treatment in nowadays. The indication of immunotherapy in mCRC was mainly in microsatellite instability patients. Previous studies also tested immunotherapy combination therapy in mCRC. Patel et al. conducted a phase 2 trial adding nivolumab to TAS-102 in patients with heavily pretreated microsatellite-stable (MSS) mCRC patients (15). The results showed nivolumab plus TAS-102 failed to extend clinical benefits in patients with refractory MSS mCRC. Median PFS was only 2.2 months. Another immunotherapy combination is nivolumab plus REG. Fukuoka et al. conducted a phase Ib trial of regorafenib plus nivolumab for patients with mCRC. The efficacy is promising with ORR 36% and median PFS 7.9 months. However, further prospective trials are warranted. Taking together, our study suggests that the optimal treatment strategy for mCRC patients is combination therapy like targeted therapy plus chemotherapy, rather than targeted therapy or chemotherapy alone.

Current guidelines all suggested that TAS-102 and REG are standard treatment in patients with chemorefractory mCRC (2). However, little is known about the priority and treatment sequences. Several literatures had compared the oncologic outcomes between TAS-102 monotherapy and REG monotherapy retrospectively. Chida et al. recruited 550 mCRC patients and revealed prolonged survival in patients treated with both REG and TAS-102, as compared with either REG or TAS-102 alone (16). Unseld et al. investigated the optimal treatment sequence for late-line mCRC and showed a tendency for longer PFS and OS for TAS-102 treated patients. The PFS and OS for patients treated with TAS-102 was 2.8 months and 15.9 months, respectively (17). Nakashima et al. identified 7279 patients (REG: 1501, regorafenib/TAS-102: 973, TAS-102: 3777, and TAS-102/REG: 1028) with corresponding median OS was 6.4, 16.4, 10.2, and 16.9 months, respectively. They concluded that it might be better to prescribe TAS-102 first and sequential administration of TAS-102 and REG ought to be considered because of longer OS (8). Moriwaki et al. conducted a propensity score analysis and demonstrated that no significant difference in OS between TAS-102 and REG was observed in patients with mCRC. The median OS was 7.9 months in the REG and 7.4 months in the TAS-102 (18). Ogata et al. reported a multi-institute retrospective study and found TAS-102 and REG had similar efficacy. The median OS was 9.9 and 11.4 months in the REG and TAS-102, respectively and the median PFS was 2.0 and 3.3 months in the REG and TAS-102, respectively (9). Huemer et al. performed an effectiveness and safety analysis and indicated that REG was associated with an increased hospitalization probability during palliative therapy in chemorefractory mCRC. The corresponding hospitalization probability was 30% with regorafenib versus 18% with TAS-102 at five weeks and 41% versus 28% at ten weeks, respectively (19). Patel et al. showed that mCRC patients taking TAS-102 were significantly more likely to adhere to and comply with therapy compared with those taking REG. The TAS-102 patients were 37% less likely to discontinue their treatment compared with the REG patients (20, 21). Our study was consistent with these publications that oncologic outcomes were insignificant between TAS-102 monotherapy and REG monotherapy, accounting for 2.0 vs 1.8 months, respectively. Further prospective trials with head to head comparison are ongoing.

There are several potential limitations in our work, which are inherent to any retrospective studies. Chemotherapy regimen, combination or monotherapy were decided at the discretion of physicians and patients. These might be major biases in this study. Meanwhile, a single institutional experience, a small sample size, heterogeneity of our patients and inconsistent follow-up interval may also limit the power of our study. Given that, our study first identified that combination therapy is much better than monotherapy in chemorefractory mCRC patients treated with TAS-102 or REG. Moreover, our study also confirmed that the oncologic outcomes of TAS-102 monotherapy and REG monotherapy were consistently insignificant. To date, there are no prospective randomized controlled trials with larger cohorts focusing on the comparison between TAS-102 and REG. Thus, in spite of a retrospective study with inevitable selection bias, our study remains clinically valuable.

Our study investigated the oncologic outcomes of TAS-102 and REG in patients with chemorefractory mCRC. Based on our results, we suggested that combination therapy is much better than monotherapy. Furthermore, the efficacy of TAS-102 monotherapy and REG monotherapy were consistently similar. In our multivariate analysis, combination therapy were strong prognostic factors related with survival. These conclusions are clinical valuable and pave the way for the treatment of chemorefractory mCRC. Further prospective randomized controlled trials are warranted to validate our conclusions.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by E-Da Hospital Institutional Review Board (EMPR-109-012). Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

M-CH, K-MR, and H-PC wrote manuscript and performed clinical/genetic investigation. S-EL, K-WL, C-CC, C-IC, and L-CS performed clinical/genetic investigation. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.