KRAS mutation and primary tumor location do not affect efficacy of bevacizumab-containing chemotherapy in stagae IV colorectal cancer patients

This study aims to investigate the efficacy of bevacizumab-combined chemotherapy (BCC) in Chinese stage IV colorectal cancer (CRC), and analyze the relationship between clinicopathological features with survival. Patients with stage IV CRC treated with BCC were analyzed retrospectively. 217 metastatic CRC (mCRC) patients were collected, out of which79 were right-sided CRCs and 138 were left-sided ones. Patients with Eastern Cooperative Oncology Group (ECOG) performance status ≤2, single agent chemotherapy, poor/mucous/signet ring cell component, second-and further-line of bevacizumab administration, multiple metastasis sites had comparatively worse survival. Among 141 patients with known KRAS status, 55 patients harbored KRAS mutation and 86 had wild type KRAS. The ORR and DCR were 41.9% and 78.9%, respectively, in patients with wild type KRAS, while ORR and DCR was 38.7% and 77.9%, respectively, in patients with KRAS mutation. The median PFS of patients with wild type and mutant KRAS were 8.38, and9.59 months, respectively; whereas the OS was 23.00 and 21.26 months, respectively for mCRC patients with wild-type and mutant KRAS. Cumulatively, our study indicated that BCC was effective and beneficial for Chinese stage IV CRC patients. KRAS mutation status and tumor location were not a prognostic factor for survival.

While the benefits of treatment with bevacizumab are well studied in American and European patients with stage IV CRC, the effect and safety of treatment with bevacizumab combined chemotherapy in Chinese patients, and whether the KRAS mutation status and primary tumor site could affect the prognosis of Chinese stage IV CRC have not been demonstrated clearly. This retrospective study aimed at investigating the efficacy and safety profile of combination treatment with bevacizumab in Chinese stage IV CRC patients and analyzing prognostic factors for predicting patients' survival.

Methods
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. The study protocol was approved by the Institutional Review Board of Chinese PLA General Hospital.
Study population. This retrospective study included 217 patients with stage IV CRC who had been treated with bevacizumab-containing chemotherapy between May 1, 2011 and August 1, 2015 in Chinese PLA General Hospital. Patients who met the following criterions were included in this study: (1) histologically confirmed colorectal adenocarcinoma with clinical and/or histological evidences of distant metastasis cancer; (2) ECOG performance status (PS) ≤2; (3) life expectancy >3 months; (4) measurable disease consistent with the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1, (5) adequate organ function, including liver and kidney (total bilirubin ≤1.5-times the institutional upper normal limit, aspartate aminotransferase and alanine aminotransferase ≤2.5-times the institutional upper normal limit, and serum creatinine ≤institutional upper normal limit or creatinine clearance (CCr, calculated using the Cockcroft-Gault formula) ≥50 ml/min); adequate bone marrow function (leucocyte count ≥3000/mm3,neutrophil count ≥1500/mm 3 , platelet count ≥100,000/mm 3 , and haemoglobin ≥9.0 g/dl); and, (6) provided signed informed consent. The key exclusion criteria were as follows: no pathological diagnosis; history of malignancy other than CRC; less than 4 cycles of bevacizumab-containing chemotherapy, thus the tumor response could be evaluated at least once; the presence of clinically significant cardiovascular disease; uncontrolled hypertension; bleeding diathesis or coagulopathy; central nervous system metastasis; use of full-dose anticoagulants or thrombolytics; pregnancy or lactation; non-healing wounds; inability to take therapy on time. Patients with no completed clinicopathological and survival data were also excluded.
Treatment. All of the 217 patients included were treated intravenously with 5 mg/kg bevacizumab (Avastin; Genentech, San Francisco, CA, USA) every 2 weeks or 7.5 mg/kg every 3 weeks according to different chemotherapy regimens, prior to the chemotherapy. Bevacizumab was administered initially over 90 minutes, and if the first infusion was well tolerated, the second was delivered no less than 60 minutes, and if well tolerated, the subsequent administration was over 30 minutes. Bevacizumab was temporarily or permanently reduced or forbidden in case of serious bevacizumab-related toxicity. Among the 217 patients included in this retrospective study, 75 patients received bevacizumab combined with XELOX chemotherapy, 41 patients received FOLFOX chemotherapy, 67 patients received FOLFIRI chemotherapy; 10 patients received oxaliplatin only; 8 patients received irinotecan alone; 10 received raltitrexed chemotherapy; 6 patients received 5-fluoropyrimidine or capecitabine chemotherapy. The chemotherapy regimens are detailed in Table 1.
Clinical Outcome Assessments. The long-term effectiveness measures included PFS, which is defined as the duration from the start of the initial bevacizumab-containing therapy to the first recorded occurrence of disease progression or death; overall survival (OS), which is calculated as the duration from the initiation of the bevacizumab-containing therapy to death or censoring. Patients without an event who still remained in follow-up were censored on the last follow-up date, July 31, 2016. Short-term effective objectives included overall response rate (ORR) and disease-control rate (DCR). Baseline tumor statuses of targeted lesions were evaluated using computer tomography (CT) scan of the chest, abdominal and pelvis. Tumor responses were evaluated at the completion of each 6-week cycle according to RECIST.

Statistical analyses.
For survival analyses, the Kaplan-Meier method was used to estimate the correlation between PFS, OS rates and clinicopathological variables, at 95% CI. The log-rank test was used to compare survival curves. All the statistical analysis was conducted by SPSS 19.0 software package and a P < 0.05 was considered as statistically significant.

Results
Patient characteristics. 217 stage IV CRC patients (120 men, 97 women, median age 58 years old) treated with bevacizumab-containing chemotherapy between May 1, 2011 and August 1, 2015 in Chinese PLA General Hospital were collected and retrospectively analyzed. Baseline demographics and clinical characteristics are summarized in Table 1. 169 patients had an ECOG PS scored 0-1 and 48 patients scored 2 at the initial bevacizumab administration. The number of metastasis in no more than 2 organs was discovered in 156 patients till the last follow-up date. The most common metastatic organ was liver (158 patients), 106 of whom suffered from synchronous liver metastases. One hundred fifty three patients received primary tumor surgery.
Short-term effect. Among 217 patients, 193 patients had progressive disease, out of whom 118 patients expired by the last follow-up date. The total ORR was 38.3% and DCR was 87.1%. The ORR and DCR were 51.5% and 96.2% when bevacizumab was administered in first-line therapy and 25.5% and 78.2% when administered in second-line therapy. However, ORR and DCR were 10.0% and 53.3% when bevacizumab was given just in thirdand forth-line, respectively (Table 2).  Table 1). Stage IV CRC patients with ECOG 2 had worse PFS and OS than patients with ECOG 0-1 (P < 0.05) ( Fig. 2A,B; Table 1).
As for chemotherapy regimens, single agent such as oxaliplatin, irinotecan, 5-fluoropyrimidine, capecitabine and raltitrexed did not prolong the patients' survival compared with combinatorial chemotherapeutic regimens. There was no statistically significant difference of PFS and OS between oxaliplatin-based combined chemotherapies and irinotecan-based ones (Fig. 2C,D; Table 1).
Poor/mucous/signet ring cell component had different effort on PFS and OS. Poor differentiation, resembling advanced metastatic stage, significantly shortened the OS (P < 0.05) (Fig. 3A) instead of PFS (P > 0.05) ( Fig. 3B; Table 1). Similar observations were made for number of metastatic sites with patients with more than 2 sites having a significantly lower OS and PFS (P < 0.05 in each case) (Fig. 3C,D; Table 1).
We divided time of bevacizumab regimen to two groups -first line and ≥2 (Table 1). The PFS and OS were directly dependent on how quickly bevacizumab was administered (P < 0.05 in each case) (Fig. 3E,F; Table 1). The other clinicopathological features did not present differential long-term effect.  [15][16][17] , and referring to stage IV CRC, right-sided ones had significantly worse survival than left-sided ones 15 . We were interested in the effect of primary tumor location on bevacizumab-related survival. Among 217 enrolled patients, 79 were right-sided CRCs and 138 were left-sided ones. Left-and right-sided CRC had no difference in overall survival following treatment with bevacizumab-containing chemotherapy (Fig. 4A,B; Tables 2,3). Furthermore we analyzed the differences of survival of right-and left-sided CRC in every line of bevacizumab, and there were still no significant differences (data not shown).
The other commonly used monoclonal antibody in mCRC is cetuximab, which is an anti-EGFR monoclonal antibody 18 . However, KRAS mutation is a negative predictive marker for anti-EGFR treatment. For this reason, only mCRC with wild type KRAS could be treated with cetuximab 19 . We wondered if KRAS mutation status would affect bevacizumab efficacy, so patients who had the KRAS status information (n = 141) were divided in to KRAS mutation group (n = 55) and wild type group (n = 86). The ORR and DCR were 41.9% and 87.2% in patients with wild type KRAS, while 38.2% and 85.5% in patients with mutant KRAS. The median PFS and OS of KRAS wild type and mutation type were 8.44 vs 9.43 months and 24.61 vs 19.09 months, but the differences were not significant (P > 0.5) (Fig. 4C,D; Tables 1,2).

Discussion
After bevacizumab was initially approved by the FDA for stage IV colorectal cancer in 2004 based on the results of the AVF2107g RCT, it became one of the standards for first-line, second-line and cross-line therapeutic regimen 20 . The addition of bevacizumab to chemotherapy had been shown to prolong stage IV CRC patients' survival   Table 4).
The BEAT study established the combination chemotherapy plus bevacizumab other than monotherapy as the standard therapy for mCRC first-line treatment 11 . However, all RCTs did not get equivalent results. In the N016966 study, even though bevacizumab administration resulted in longer PFS, it did not have the same effect on OS 12 . The CARIO3 and MACRO phase III studies proved the benefit bevacizumab when used only as maintenance therapy 21,22 . The E3200 study, where mCRC patients treated with FOLFIRI were enrolled, revealed that  the addition of bevacizumab significantly improved survival after first progression 13 . The ML18147 and BEBYP studies revealed the benefits of continuation of bevacizumab even after initial chemotherapeutic resistance 14,23 .
In China where people are mostly of Mongoloid decent, the evidence of RCT study in bevacizumab was not as sufficient as described above. The ARTIST study, a phase III RCT study, demonstrated that in Chinese population bevacizumab could increase mCRC survival significantly 24 . Oncologists in China mostly decide therapeutic regimens following international guidelines and additional experiences of bevacizumab administration are required. In the present study, we obtained similar results as RCT studies discussed above. The subtle differences in duration of survival and response rate may due to the different regimens and population.
The poor survival and the prognostic impact of KRAS mutation in the CRC have been previously reported 25 . It was shown that KRAS mutation status did not have a significant effect on the PFS in first-line treatment of mCRC 26 . Hurwitz et al. (2009) reported that the clinical benefit of bevacizumab in mCRC was independent of KRAS mutation status; bevacizumab provided significant clinical benefit in patients with mCRC expressing either mutant or wild-type KRAS 27 . We explored the efficacy of bevacizumab-contained chemotherapy for stage IV CRC patients according to different KRAS mutation status. Our result revealed that patients with KRAS mutation had a shorter survival than patients with wild type KRAS wild type; however, the difference was not statistically significant.
Current evidence indicates that right-and left-sided CRC respond differently to treatment. Proximal and distal colon show different pathways to develop and tumorigenesis. The proximal colon originating from the embryonic midgut is perfused by the superior mesenteric artery, however the distal colon deriving from the hindgut is supplied by the inferior mesenteric artery 28 . KRAS mutations were more frequently found in the right than left Dukes' C colon cancer 29 . The patients with right-sided CRC were older female, poorly differentiated and suffered from poorer survival 30 . The more active EGFR signaling in distal CRC meant that left-sided cancers benefited significantly more from cetuximab 31,32 . In addition, it was shown using data obtained from PROVETTA, AVF2107g and NO16966 trials, that even though efficacy of bevacizumab was independent of tumor location, patients with left-sided tumors had significantly better outcomes 33 .
In the current study, we analyzed the short and long term efficacy in different anatomic sites of mCRC treated with bevacizumab. We divided mCRC into right-and left-sided groups and found that tumor location was not a prognostic factor for survival. Although the median PFS, OS and response rate of right-sided mCRC were all worse than left-sided cancers, none of the differences achieved statistical significance, which is contradictory with two previous reports 33,34 . The difference is perhaps because of the different population and line of bevacizumab administration. This highlights the importance of similar studies in the context of different population and analytic setting.    Statement of Ethics. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.