Colorectal cancer (CRC) is one of the most common gastrointestinal malignancies in China, and the incidence and mortality rate are on the rise. According to the data of the 2015 China Tumor Registry Annual Report released in March 2015, colorectal cancer ranks third in the incidence of malignant tumors nationwide, among which the top five are lung cancer, stomach cancer, colorectal cancer, liver cancer and esophageal cancer. It can be seen that colorectal cancer poses a great threat to the health of Chinese residents.
For early-stage colorectal cancer without distant metastasis (UICC TNM stage I-III), radical surgical resection is the main treatment, and adjuvant chemotherapy based on fluorouracil, as another effective treatment option, has become an important treatment for early-stage CRC since it was proved to be able to improve the efficacy of radical surgery to some extent in the 1990s.
I. Status of adjuvant chemotherapy for colon cancer.
(i) Indications: i.e., which patients need to receive postoperative adjuvant chemotherapy The selection of adjuvant chemotherapy requires weighing factors such as survival benefit, treatment toxicity, and cost. Among them, the survival benefit of chemotherapy is of course the most important consideration.
From the perspective of TNM staging, stage I (T1-2N0M0) colon cancer has a fairly good prognosis with an OS (overall survival) of nearly 90% at 5 years after radical surgery alone, and the additional benefit of adjuvant chemotherapy is rather marginal; therefore, the current consensus is that adjuvant chemotherapy is not recommended for stage I colon cancer.
Stage III colon cancer (any TN+M0), due to the occurrence of lymph node metastasis, has a relatively poor prognosis, and the additional benefit from adjuvant chemotherapy is generally 10%-20%; therefore, stage III colon cancer is currently the standard indication for adjuvant chemotherapy.
The situation of stage II colon cancer (T3-4N0M0) is much more complicated, because the group of patients in this stage is very heterogeneous and the prognosis is very different, the overall benefit of adjuvant chemotherapy is about 3%-6%, and it is believed that adjuvant chemotherapy should not be recommended as a whole, but different subgroups should be screened out, and some groups need adjuvant chemotherapy, while some groups may not.
(ii) Drugs and regimens: i.e., which drugs or regimens can be used for adjuvant chemotherapy of colon cancer According to earlier studies, the only two drugs that can be used for adjuvant chemotherapy of colon cancer are fluorinated pyrimidine and oxaliplatin. The former can be used as a single agent, including intravenous fluorouracil (5-FU) and the oral fluorouracil precursors capecitabine and eflornithine (UFT, mainly used in Japan, etc.), while oxaliplatin cannot be used alone, according to MOSAIC (FOLFOX4 vs. infusion 5-FU/LV), NSABPC-07 (FLOX vs. push 5-FU/LV ) and XELOXA (XELOX versus push 5-FU/LV), which established oxaliplatin’s place in adjuvant chemotherapy for colon cancer.
With the exception of these, all other drugs that can be used in advanced bowel cancer have not been shown to benefit patients in adjuvant chemotherapy and, therefore, are not recommended.
All clinical studies of targeted agents in adjuvant chemotherapy for colon cancer have yielded negative results; therefore, there is currently no evidence that targeted agents provide additional survival benefit and should not be used.
So, what are the standard regimens recommended for adjuvant chemotherapy for colon cancer at present?
1.Single-agent regimen All of them are fluorinated pyrimidine drugs.
(1) 5-FU/LV biweekly infusion regimen (sLV5FU2): also known as deGramont biweekly regimen, i.e., the 5-FU/LV portion of the mFOLFOX6 or FOLFIRI regimen that remains after the removal of oxaliplatin or irinotecan. Due to the specific mechanism of action, intravenous infusion of 5-FU has been shown to be no less effective than push 5-FU, but with significantly lower bone marrow toxicity and GI toxicity. Therefore, intravenous fluorouracil has fully replaced the use of push infusion in the field of intestinal cancer chemotherapy. It should be noted that the push 5-FU/LV regimen (Mayo regimen) was removed from the NCCN guidelines as early as 2010, and therefore should no longer be used in adjuvant chemotherapy.
(2) Capecitabine: The X-ACT study has demonstrated that in stage III colon cancer, the efficacy of capecitabine is not inferior to that of pushed 5-FU/LV, and there is a trend of prolonging DFS (tumor-free survival) and OS, therefore, capecitabine has become the standard regimen for postoperative adjuvant chemotherapy in colon cancer.
2. Combination regimen
(1) mFOLFOX6 (infusion of 5-FU/LV/oxaliplatin): The MOSAIC clinical study showed significant DFS and OS benefits with the FOLFOX regimen. Based on this result, the FOLFOX regimen became the standard of care for postoperative patients with stage III colon cancer (NCCN guideline level 1 evidence, priority recommendation). Given the greater toxicity of intravenous push FU compared to infusion, the 2010 NCCN guidelines recommended the full replacement of the FOLFOX4 regimen with mFOLFOX6. There is now a global consensus on this point.
(2) XELOX (capecitabine/oxaliplatin): The XELOXA study also confirmed that XELOX significantly prolongs DFS and OS compared to infusion of 5-FU/LV (Mayo), and therefore, XELOX has become the standard of care for postoperative patients with stage III colon cancer (NCCN guideline level 1 evidence, priority recommendation). Based on the results of the NSABP C-07 study in the United States, the FLOX regimen (push-injected 5-FU/LV/oxaliplatin) is also recommended by the NCCN guidelines for adjuvant chemotherapy, but given the little use of this type of push-injected 5-FU/LV in China and the high incidence of diarrhea, the FLOX regimen is not currently recommended for adjuvant chemotherapy in Chinese patients.
(iii) Timing and duration of chemotherapy.
A systematic review and meta-analysis published in 2011 meta-analyzed 10 studies including more than 15,000 cases, focusing on the impact of the timing of adjuvant chemotherapy on outcome after radical surgery. The results of this analysis showed that a 4-week delay in adjuvant chemotherapy decreased overall survival by 14%, suggesting that postoperative adjuvant chemotherapy should be started as soon as medically feasible for the patient. The current recommended standard course of adjuvant chemotherapy is 6 months. Given the neurotoxicity of oxaliplatin, a shorter course of adjuvant chemotherapy, such as 3 months, is currently being investigated. However, results are not yet available. In conclusion, the consensus of adjuvant chemotherapy for colon cancer is that chemotherapy is not required for stage I, stage III is the absolute indication for chemotherapy, and stage II is still controversial; the only drugs that can be used as adjuvant chemotherapy are fluorouracil and oxaliplatin, and the standard course of adjuvant chemotherapy is 6 months.
Based on the above basic consensus, how to individualize the implementation of adjuvant chemotherapy? In other words, we need to solve the problem of individualized medical treatment for the right group.
This process is based on two main principles.
① Find those groups that need adjuvant chemotherapy. It is generally accepted that those with a poor prognosis need adjuvant chemotherapy, and this is achieved through prognostic factors.
When we find those who need to receive adjuvant chemotherapy from a prognostic point of view, we give standard chemotherapy, but obviously not all of them will eventually benefit from chemotherapy, so how to identify them? This relies on predictive factors.
This process is particularly evident in the exploration of adjuvant chemotherapy for stage II colon cancer over the years.
(a) Prognostic predictors of stage II colon cancer
The aim is to identify patients who are at high risk of recurrence and metastasis, hence the term “high-risk factors”. Currently, at least one of the following factors is recognized by major institutions and societies worldwide as “high-risk stage II”: T4 tumor, tumor perforation, intestinal obstruction, poor histologic differentiation (except MSI-H), vascular/neural infiltration, and less than 10 lymph nodes sent for examination. Some studies have suggested that elevated preoperative blood CEA levels should also be considered a “high risk factor”, but there is no consensus on this point. Patients with these high-risk factors have a high risk of recurrence and are strong evidence for chemotherapy recommendation in adjuvant chemotherapy decisions.
2. MMR status and MSI DNA mismatch repair (MMR) gene mutations or modifications (e.g., methylation) result in MMR protein deletion and microsatellite instability (MSI), which is the result of MMR deletion due to insertion or deletion of DNA repeat units. Patients with MMR deletion (dMMR) belong biologically to the same group as MSI-H (high microsatellite instability). dMMR accounts for about 15% to 20% of patients with stage II. There has been substantial evidence that in patients with stage II colon cancer, deletion of MMR protein expression or MSI-H is a marker of good prognosis, but the prognostic value is not as pronounced in stage III as in stage II, whereas in stage IV The reason why dMMR was found to have a worse prognosis in stage IV colon cancer is unclear. However, the prognostic value of MMR in stage II bowel cancer patients is well recognized, so the prognosis can be screened by detecting MMR protein expression by immunohistochemistry (IHC), and dMMR is a “low risk factor” for recurrence of metastasis in stage II colon cancer.
Oncotype DX (Genomic Health, Inc) genetic testing technology and ColoPrint (Agendia) genetic testing technology are the two most mature systems for colon cancer. The ColoPrint genetic test quantifies the expression of 18 prognosis-related genes and classifies them into low- and high-risk groups, which were also found to be effective in screening different recurrence risks in the validation trial. The HR for recurrence rate was 3.34 (p=0.017) for stage II patients. In conclusion, multigene testing has shown encouraging results in evaluating prognosis, and it is believed that there will be further progress in this area with the opening of the era of precision medicine.
(ii) Predictors of the efficacy of adjuvant chemotherapy for colorectal cancer Will those “likely to need adjuvant chemotherapy” identified by the above prognostic predictors actually benefit from adjuvant chemotherapy in the end? Can we have similar markers to predict this benefit? That is, are there predictive marekrs for the efficacy of adjuvant chemotherapy? In fact, clinicopathological stage is a predictive marekrs of adjuvant chemotherapy: stage I is not beneficial, stage III is definitely beneficial, stage II is not necessarily beneficial; then, what other marekrs have the function of predicting efficacy?
1. Can clinical high-risk factors predict the efficacy of phase II adjuvant chemotherapy? There are no randomized controlled studies predicting the efficacy of fluorouracil monotherapy adjuvant chemotherapy based on clinical high-risk factors, and the MOSAIC study (n=2246, Tournigand et al. JCO 2012) (FOLFOX4 versus FU/LV biweekly infusion regimen for adjuvant therapy II/III) was the first RCT to prospectively include stratification of high-risk factors (as detailed above) in the study and showed that for all stage II patients, the 5-year DFS of FOLFOX4 versus 5-FU/LV adjuvant chemotherapy was 83.7% and 79.9%, respectively, with an absolute benefit of 3.8%, HR=0.84, P=0.258; however, for the “high-risk stage II” group with high-risk factors (n=569), it was 82.3% and 74.6% HR=0.72, P=0.062, respectively, with an absolute benefit of 7.7%, similar to the absolute DFS benefit for stage III colon cancer ( The findings suggest that the addition of oxaliplatin to adjuvant chemotherapy does not result in significant DFS benefit for overall stage II patients, but does result in DFS benefit for high-risk stage II patients similar to that of stage III, and it is on the basis of this study that adjuvant chemotherapy for high-risk stage II is considered to be equivalent to stage III, and in the NCCN and ESMO guidelines The NCCN and ESMO guidelines also include oxaliplatin-containing regimens as an option for adjuvant chemotherapy in high-risk stage II colon cancer, but do not recommend oxaliplatin-containing regimens for adjuvant chemotherapy in patients with low-risk or low-risk stage II without “high-risk factors”. Therefore, based on the current evidence, we can say that clinicopathological high-risk factors can be used as predictors of the efficacy of oxaliplatin in adjuvant chemotherapy for stage II colon cancer.
2. Predictive value of MMR/MSI on the efficacy of adjuvant chemotherapy This is the area that has received the most attention and made the most progress, and is also a breakthrough in individualized adjuvant chemotherapy for stage II colon cancer. (1) The relationship between MMR and the efficacy of 5-FU adjuvant chemotherapy: More than a decade ago, studies have been conducted on the predictive value of MSI status on the efficacy of 5-FU adjuvant chemotherapy for colon cancer. However, MSI-H patients did not benefit from postoperative 5-FU adjuvant chemotherapy and had a lower 5-year survival rate (HR=2.14, P=0.11). This study was the first to suggest that MSI may be a predictor of the efficacy of adjuvant chemotherapy and that MSI colon cancer does not benefit from 5-FU chemotherapy. However, it did not receive sufficient attention and concern in the industry until a later retrospective meta-analysis by Sargent et al. (JCO 2010), which combined data from several clinical studies, found similar results: 5-FU adjuvant chemotherapy appeared to result in impaired survival in stage II dMMR colon cancer (5-year DFS, chemotherapy 72% vs. surgery alone 80%, HR 2.80, P = 0.05) In stage II pMMR patients, this phenomenon was not present, and chemotherapy did not result in impaired survival (5-year DFS, chemotherapy 77% vs. surgery alone 72%, HR 0.84, P=0.38). In stage III colon cancer, dMMR tumors treated with adjuvant chemotherapy did not have impaired survival (5-year DFS, 67% with chemotherapy vs. 62% with surgery alone, HR 1.08, P=0.86, 5% chemotherapy benefit), but the chemotherapy benefit was significantly smaller than in pMMR (5-year DFS, 58% with chemotherapy vs. 41% with surgery alone, HR 0.64, P=0.01, 17% chemotherapy benefit). From Sargent’s meta-analysis, it is clear that dMMR colon cancer does not benefit from FU single-agent adjuvant chemotherapy, especially in stage II, with the potential for impaired survival; in contrast, pMMR patients benefit significantly from the same treatment. The value of MMR in predicting the efficacy of FU adjuvant chemotherapy has only been widely appreciated since then. In light of these findings, the NCCN guidelines from 2010 included MMR as a predictor of efficacy in adjuvant chemotherapy, recommending that all patients with stage II colon cancer should be tested for MMR and stating that patients with dMMR do not benefit from adjuvant chemotherapy with fluorouracil alone. However, in contrast to Sargent’s findings, the recent QUASAR study (Hutchins et al, 2011), which analyzed data from 1913 stage II colon cancer cases (half of the patients received adjuvant chemotherapy), showed that although dMMR was a prognostic predictor (11% vs. 26% recurrence rate for dMMR vs. pMMR), it was not predictive of benefit or compromise to chemotherapy. gain or impairment to chemotherapy and cannot be used as a predictor of efficacy. However, the authors explained in this study that the lack of predictive value of MMR for chemotherapy was mainly attributed to the small proportion of dMMR and the even smaller proportion of recurrences within 2 years in this group (15/208, 7.2%), but in terms of specific results, MMR showed some predictive trends: the recurrence rate within 2 years in the dMMR group was 6.1% for chemotherapy vs. 7.7% for surgery alone. OR 0.81 (95% CI 0.29 to 2.22); pMMR group chemotherapy 9.0% vs. surgery alone 15.0%, OR 0.59 (95% CI 0.45 to 0.77). These data suggest that dMMR still showed some predictive value for efficacy (no significant reduction in postoperative recurrence, suggesting no benefit from chemotherapy). However, no predictive value was found in the interaction test between MMR status and chemotherapy efficacy; therefore, the final conclusion of this study is that dMMR does not predict the lack of efficacy of adjuvant chemotherapy to 5-FU in patients. Similarly, the results of PETACC-3 (Tejpar et al, 2009), which found that patients with either MSI or MSS benefited from adjuvant chemotherapy, suggest that dMMR is only a prognostic, not an efficacy predictor. In light of these inconsistent data, ESMO guidelines have not yet considered the MMR as a predictor of efficacy for adjuvant chemotherapy in colon cancer. Why is a defective MMR function predictive of ineffectiveness to 5-FU chemotherapy? Current clinical data suggest that dMMR induces resistance to methylating agents, platinum and antimetabolites: cell lines with dMMR resist the cytotoxic effects of 5-FU, while normalization of MMR function restores the sensitivity of these cells to 5-FU. Possible mechanisms include: MMR deficiency may increase drug resistance; cells with MMR mutations are tolerant to DNA damage caused by chemotherapeutic agents because MMR proteins are involved in mediating activation of cell cycle checkpoints and apoptosis, whereas MMR-deficient cells exhibit longer G2 phase arrest; MMR mutations may also have an indirect effect by increasing mutations in the genome. It has been suggested that the predictive value of MMR for the efficacy of adjuvant chemotherapy with 5-FU may be related to the molecular mechanisms leading to defective MMR function. dMMR in sporadic intestinal cancer is mainly caused by promoter methylation of MLH1, whereas dMMR in Lynch syndrome is due to germline mutations, and these differences may lead to differential sensitivity to drugs. It has been found (Sinicrope et al, 2011) that patients with Lynch syndrome with a dMMR phenotype can significantly improve survival after 5-FU chemotherapy (DFS: HR 0.31, 95% CI 0.14-0.70, P=0.006), while sporadic patients still do not benefit from 5-FU treatment (P=0.15). Whether dMMR can predict the lack of efficacy of adjuvant chemotherapy against 5-FU is still controversial. The need to differentiate between sporadic and hereditary patients in the clinical selection of treatment for patients with dMMR phenotype is also inconclusive and deserves further study.
(2) Relationship between MMR and the efficacy of adjuvant chemotherapy with oxaliplatin: There are few data on the relationship between MMR and the efficacy of adjuvant chemotherapy with FOLFOX (oxaliplatin/5-FU/LV), and there are no results from randomized controlled studies. One retrospective study (Zaanan et al, 2009) found a significant benefit from FOLFOX chemotherapy in stage III colon cancer patients with MSI-H compared to 5-FU/LV chemotherapy (3-year DFS 100% vs. 57.9%). Several other retrospective studies have also yielded similar results. The data observed so far suggest that, unlike that observed in 5-FU adjuvant chemotherapy alone, the benefit of oxaliplatin-containing adjuvant chemotherapy is independent of MSI status and provides a benefit compared to 5-FU/LV alone. In summary, the predictive value of the efficacy of MMR, although still controversial, is currently considered to be mainly limited to fluorouracil monotherapy, especially in stage II patients, where the available evidence suggests that dMMR/MSH-H is not beneficial and may even be harmful. In contrast, for oxaliplatin, MSI status is not associated with adjuvant chemotherapy efficacy, and both can benefit.
The predictive value of multigene testing systems for adjuvant chemotherapy has not been validated by ColoPrint in the RCT system, but the Oncotype DX 12 gene testing system was retrospectively analyzed for its predictive value for adjuvant chemotherapy in cases from the NSABP C-07 (pushing 5-FU/LV± oxaliplatin for adjuvant chemotherapy in stage II/III colon cancer) study (Yothers G et al. 2013). The results found that RS (risk of recurrence score) did not predict the efficacy of oxaliplatin and that the relative survival benefit from oxaliplatin, which was present across the RS distribution interval, was present regardless of low or high RS (p=0.48), but the absolute survival benefit of oxaliplatin correlated with the RS score, with the benefit increasing as the RS score increased. This study concludes that the Oncotype DX 12 genetic testing system cannot be used as a predictor of efficacy of oxaliplatin adjuvant chemotherapy.
(iii) Clinical value of prognostic indicators and efficacy predictors in decision making of adjuvant chemotherapy for stage II colon cancer Staging, clinicopathological high-risk factors and MMR are the most important factors to be considered in decision making. Of course, other factors such as patient function, co-morbidities, and other factors should be taken into account to fully communicate the benefit-risk profile of adjuvant chemotherapy to patients. For stage II colon cancer, there are two factors to consider from the prognostic point of view, namely MMR and clinical high-risk factors, while only MMR can provide a reference from the perspective of efficacy prediction. For those with dMMR, firstly, the prognosis is good, and secondly, dMMR predicts ineffectiveness to fluorouracil drug monotherapy, which is the standard adjuvant chemotherapy regimen for the majority of stage II colon cancers. In view of this, adjuvant chemotherapy is generally not recommended for those with dMMR, and observation alone is sufficient. In case of dMMR with clinical high-risk factors, it is still controversial because this group is very small (the proportion of T4 and dMMR is about 1%), and there is a lack of specific research data, which needs to be considered in combination with other factors. In other words, at this time, chemotherapy is recommended with T4b as the main reference, and all clinical high-risk factors other than this give way to dMMR in the treatment decision, i.e., a preference for no chemotherapy. However, these views have not yet reached a full consensus due to the lack of RCT data support, let alone in the guidelines. In conclusion, adjuvant chemotherapy decisions for stage II colon cancer with dMMR combined with clinical risk factors are controversial, but if chemotherapy is decided, I personally advocate an oxaliplatin-containing combination regimen to avoid possible dMMR resistance to 5-FU. In the case of pMMR, clinical decisions at this point are not influenced by MMR and follow clinical routine.