Background
Systemic chemotherapy for metastatic colorectal cancer (mCRC) has improved remarkably in recent years, currently allowing most mCRC patients to survive for >2 years. Standard treatments for patients with mCRC include chemotherapy regimens based on cytotoxic agents, such as fluoropyrimidine, irinotecan, oxaliplatin and the vascular endothelial growth factor (VEGF) antibody bevacizumab (Bev), with the addition of epithelial growth factor receptor (EGFR) antibodies cetuximab and panitumumab for RAS wild-type patients [
1,
2]. An additional file shows standard therapies for metastatic colorectal cancer in the various setting [see Additional file
1]. However, after failure of standard chemotherapy, the average survival rate is only 5–6 months. Therefore, there is a need for more effective treatment for patients with refractory mCRC who maintain a relatively good performance status (PS) and are willing to receive further treatment.
S-1 (TS-1; Taiho Pharmaceutical Co. Ltd., Tokyo, Japan) is an oral fluoropyrimidine anti-cancer agent that combines tegafur as the effector drug with the 2 modulators gimeracil and oteracil. Tegafur is a pro-drug of 5-fluorouracil (5-FU), gimeracil is an inhibitor of dihydropyrimidine dehydrogenase (DPD) which maintains the serum 5-FU level. Oteracil inhibits the gastrointestinal toxicity of 5-FU. Several phase-III trials on mCRC have demonstrated S-1 can be a substitute for infusional 5-FU [
3,
4]. Moreover, a phase-II study demonstrated promising S-1 activity in heavily pre-treated mCRC patients, including those treated with 5-FU [
5].
Leucovorin (LV) is a well-known enhancer of 5-FU activity by inhibiting thymidylate synthase (TS) [
6]. Concomitant 5-FU + LV therapy is used worldwide to treat patients with mCRC as either first-line or adjuvant therapy. In addition, oral tegafur/uracil (UFT) plus LV combination therapy has demonstrated similar clinical efficacy to that of intravenous 5-FU and LV, and it is associated with improved convenience of care because an infusion pump is not required [
7]. On the other hand, little data is available on S-1 plus oral LV (SL) combination therapy. Subsequently, Koizumi et al. conducted a phase-II study of SL treatment in patients with previously untreated mCRC, the results of which were promising [
8]. On the basis of these results, SL therapy is expected to yield high anti-tumour activity compared with S1 monotherapy in cases of refractory mCRC.
Bev is a humanized monoclonal antibody that inhibits VEGF and has demonstrated activity both as first-line and second-line therapy for mCRC in combination with fluoropyrimidine with or without irinotecan or oxaliplatin [
9‐
11]. However, the role of Bev in third-line or later-line therapy of mCRC remains unclear. The continuation of Bev beyond disease progression after first-line therapy has been demonstrated to improve progression-free survival (PFS) and overall survival (OS) in large phase-III trials [
12,
13]. These results suggest that Bev therapy after disease progression may have a clinical benefit even in refractory patients treated with Bev.
To test these hypotheses, we conducted a single-centre phase-II trial to assess the efficacy and safety of SL + Bev (SL/Bev) as a salvage therapy in patients with mCRC in whom prior chemotherapy with 5-FU, oxaliplatin, irinotecan, Bev and anti-EGFR antibodies has failed.
Discussion
In this study, first, we demonstrated SL/Bev has significant clinical activity, yielding 58 % disease stabilisation. This, together with 7 % partial response, resulted in 65 % DCR in heavily pre-treated mCRC. This trial met the pre-specified primary endpoint and also achieved favourable median PFS and OS of 5.3 and 9.9 months, respectively. Furthermore, SL/Bev was generally tolerated and appears to have little impact on QOL.
Several prospective studies have evaluated various cytotoxic and/or targeted agents in the setting of refractory mCRC, and potent new anti-cancer agents have been developed in recent years. The results of large phase-III study of regorafenib, oral multi-kinase inhibitor, showed prolonged survival compared with a placebo in patients with heavily pre-treated mCRC, with a DCR of 41 %, median PFS of 1.9 months and median OS of 6.4 months [
15]. In addition, TAS-102 reported encouraging results compared with placebo in randomized phase-III study, which is a new oral nucleoside anti-tumour agent consisting of trifluorothymidine and a thymidine phosphorylase inhibitor [
17]. The results showed a DCR of 44.0 %, median time PFS of 2.0 months and median OS of 7.1 months. Compared with the results of previous studies, the present study’s DCR of 65 % for refractory patients was higher and the median OS of 9.9 months was longer. Although, the favourable OS was partly due to post-study anti-cancer treatment; approximately half of the failures (16 of 28) underwent further treatment after stopping protocol therapy. Indeed, a PFS of 5.3 months is much better than that reported previously in salvage settings. Moreover, two patients achieved PR despite resistance to several chemotherapies. These results indicate SL/Bev has individual anti-cancer activity in this heavily pre-treated population.
In this study, SL/Bev showed an acceptable toxicity, considering that it was previously a heavily treated population. The most common adverse events in our study were non-hematological toxicities such as stomatitis and diarrhea. Grade-3/4 stomatitis reported in this study was 26 %, which is higher than that in a previous study for this schedule in patients with untreated mCRC, where stomatitis was reported in only 8.3 % patients [
18]. This discrepancy may be ascribed to the fact that patients included in the current study were more heavily treated. Although dose modification was required in 16 of the 31 patients because of mucositis, no discontinuation for this reason was reported, suggesting that mucositis was generally manageable with dose reduction/interruption. Adverse events potentially associated with Bev were well tolerated. Although epistaxis and hypertension were relatively frequent, both symptoms were ≤ grade 2. Moreover, serious adverse events, such as thrombosis and bowel perforation, were not observed. The safety of Bev addition appears to be high. However, it should be noted that one patient died of sudden death possibly related to therapy; this illustrates the need for more caution when treating such a fragile patient population.
The rationale for this combination therapy was based on prior success with S-1 monotherapy against refractory mCRC and the synergistic effects of LV + Bev combination therapy. In a phase-II study, S-1 monotherapy showed promising activity in heavily pre-treated patients with mCRC who had previously failed a regimen containing irinotecan and oxaliplatin. A DCR of 42.9 %, median PFS of 3–4 months and a median OS of 10–13 months were achieved, which suggest that S-1 has potent tumor activity even in 5-FU-refractory mCRC [
5]. Several preclinical reports have presented evidence to support the effectiveness of S-1. For example, S-1 showed higher tumor growth inhibition than UFT did in an orthotopic implantation model of colon cancer, and it promoted anti-tumor activity in 5-FU-resistant cancer cell lines [
19]. In addition, it has been suggested that DPD inhibition plays a significant role of S-1 in chemo-resistant cancer cells. According to the results of the randomized phase-II/III study, which verify the non-inferiority of IRIS regimen (irinotecan + S-1) to FOLFIRI regimen (fluorouracil + LV + irinotecan) as second-line chemotherapy, IRIS regimen was superior to the FOLFIRI regimen for mCRC patients who previously received an oxaliplatin-based regimen [
20]. This difference was explained by the fact that patients previously treated with an oxaliplatin-containing regimen had a higher level of DPD gene expression in the tumor tissue than those not treated with oxaliplatin, and this mechanism was related to the stronger effects of S-1 [
21].
In the present study, the enhancement made in the treatment regimen for refractory mCRC was the S-1/LV combination therapy. The addition of LV enhanced the anti-tumor activity of S-1 by TS inhibition. Compared with S-1 monotherapy for previously untreated mCRC, S-1/LV combination therapy demonstrated an improvement in objective response rate from 35 to 57 % and improvement in time to progression from 5.3 to 6.7 months [
8]. Moreover, it is reported that acquisition of resistance to 5FU is related to increased TS expression. 5FU-resistant cell lines show increased TS mRNA expression, protein expression, and activity compared with their respective parent cells in
in vitro and
in vivo assays [
22,
23]. On the basis of this finding, concurrent treatment with LV may have contributed in overcoming resistance to 5FU and yielded some efficacy in the present study.
Little comparative data are available regarding the activity of Bev after second-line therapy. Bev therapy in the later-line setting has been reported in several phase-II studies and in retrospective series [
24‐
29]. According to these reports, Bev does not show a tendency for reduction in tumour size but results in tumour stabilisation and improved survival. However, studies evaluating its effect in later-line treatment of Bev re-introduction are limited. Recently, international multicentre study (ML18147 study) revealed that continuation of Bev after initial tumour progression significantly improved PFS and OS [
12]. Similar result was also observed in an Italian multicentre study (BEBYP study), which demonstrated a significant improvement in PFS and OS continuing Bev plus second-line chemotherapy [
13]. These results imply disease may still partially depend on VEGF after disease progression and raises the possibility that the angiogenic signal may continue throughout the tumor lifespan. With this speculation, Bev re-introduction may still contribute to enhance anti-tumour activity that has already proved to be resistant. Although we must note that prospective and randomized clinical trials are lacking regarding the role of Bev in chemo-refractory mCRC patients, we could speculate that a combination with Bev provides some efficacy in a heavily treated population.
With regard to the difference in patients in
KRAS mutation status in this study, no significant difference was observed in either disease stabilization or survival benefit between
KRAS mutation status. Another finding in this study which is noteworthy, is 1 patient who had
BRAF mutation achieved PR.
BRAF mutation is associated with poor prognosis because of more aggressive and rapidly progressing disease and is also predictive of a lack of response to chemotherapy in mCRC. However, in this patient, the size of hepatic metastases remarkably decreased after resistance to all standard chemotherapies. This fact suggests that in patients with
BRAF mutation, SL/Bev may have potent anti-tumour activity based on some specific effects. A possible explanation for this is the role played by γ-glutamyl hydrolase (GGH) in regulating intracellular folate levels. Low GGH expression is associated with higher folate levels, leading to the enhancement of anti-tumor activity in 5-FU with LV [
30]. In the
BRAF-mutated oncogene, the CpG island methylator phenotype (CIMP+) frequently occurs within gene promoter regions, and CIMP+ is associated with low GGH expression. For this reason, SL/Bev may be more effective in
BRAF-mutant patients than in
BRAF-wild-type ones. Although this phenomenon may have occurred by chance, it remains noteworthy.
This study has some limitations. One is that it is a single-arm study design with no control group for comparison. Therefore, we cannot rule out some potential bias, such as the selection of patients with good prognosis. In fact, a majority of patients except two in our study (94 %) had an ECOG PS status of <2. The second limitation is it is unclear whether the efficacy and safety of S-1 in Japanese patients would be similar to those in Western patients. Despite these limitations above, SL/Bev may provide an additional advantage. This regimen considers that no cross-resistance to new anti-cancer agents for refractory mCRC, such as regorafenib and TAS102, exists. Therefore, SL/Bev may provide a therapeutic option even after failure with these agents. To confirm its efficacy, further prospective randomized control trial is necessary to compare SL/Bev with BSC in patient with refractory mCRC.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
KY and HT designed the study, analyzed the data, and drafted the manuscript. KY, HT, AK, YN, SN, MO, SK, DT, TU and KMuro acquired clinical data. KMori reviewed and revised statistical points of the manuscript. MA, KMuro and YI reviewed and revised manuscript. All authors read and approved the final manuscript.