Role of Kras Status in Patients with Metastatic Colorectal Cancer Receiving First-Line Chemotherapy plus Bevacizumab: A TTD Group Cooperative Study

Background In the MACRO study, patients with metastatic colorectal cancer (mCRC) were randomised to first-line treatment with 6 cycles of capecitabine and oxaliplatin (XELOX) plus bevacizumab followed by either single-agent bevacizumab or XELOX plus bevacizumab until disease progression. An additional retrospective analysis was performed to define the prognostic value of tumour KRAS status on progression-free survival (PFS), overall survival (OS) and response rates. Methodology/Principal Findings KRAS data (tumour KRAS status and type of mutation) were collected by questionnaire from participating centres that performed KRAS analyses. These data were then cross-referenced with efficacy data for relevant patients in the MACRO study database. KRAS status was analysed in 394 of the 480 patients (82.1%) in the MACRO study. Wild-type (WT) KRAS tumours were found in 219 patients (56%) and mutant (MT) KRAS in 175 patients (44%). Median PFS was 10.9 months for patients with WT KRAS and 9.4 months for patients with MT KRAS tumours (p = 0.0038; HR: 1.40; 95% CI:1.12–1.77). The difference in OS was also significant: 26.7 months versus 18.0 months for WT versus MT KRAS, respectively (p = 0.0002; HR: 1.55; 95% CI: 1.23–1.96). Univariate and multivariate analyses showed that KRAS was an independent variable for both PFS and OS. Responses were observed in 126 patients (57.5%) with WT KRAS tumours and 76 patients (43.4%) with MT KRAS tumours (p = 0.0054; OR: 1.77; 95% CI: 1.18–2.64). Conclusions/Significance This analysis of the MACRO study suggests a prognostic role for tumour KRAS status in patients with mCRC treated with XELOX plus bevacizumab. For both PFS and OS, KRAS status was an independent factor in univariate and multivariate analyses.


Introduction
At present, standard first-line treatment for patients with metastatic colorectal cancer (mCRC) includes combination chemotherapy in conjunction with either an anti-epidermal growth factor receptor (EGFR) agent such as cetuximab [1,2] or panitumumab [3], or an antiangiogenic agent, such as bevacizumab [4][5][6]. One critical issue is the selection of patients who will benefit from treatment with these biological agents. In the case of anti-EGFR therapies, the presence of a KRAS mutation is a negative predictive factor for response to treatment [7][8][9] and determination of KRAS status is now required by American and European authorities before these agents can be administered [10][11][12][13].
Recent studies of chemotherapy regimens, with or without cetuximab, in the first-line treatment of patients with mCRC have sparked new interest in this issue [7,[25][26][27].
The interaction of EGFR and vascular endothelial growth factor (VEGF) is well known [28,29], although the potential role of KRAS mutation status in patients undergoing treatment with bevacizumab remains of great interest. Retrospective analyses have shown that bevacizumab in combination with irinotecan/5fluorouracil (5-FU)/leucovorin chemotherapy provides a significant clinical benefit for patients with mutant (MT) and wild-type (WT) KRAS tumours [30,31]. The authors also noted that the benefit of treatment was greater in patients with WT compared with MT KRAS tumours. Other studies have shown no prognostic effect of tumour KRAS status on survival in patients receiving combination chemotherapy with bevacizumab [32][33][34][35].
We undertook an analysis of data from the MACRO study to evaluate the prognostic value of tumour KRAS status in patients receiving combination therapy with capecitabine plus oxaliplatin (XELOX) and bevacizumab. Correlations between KRAS status and response rate, progression-free survival (PFS) and overall survival (OS) were analysed.

Ethics Statement
The Institutional Review Board and Ethic Committee of Hospital Clinico San Carlos, Madrid as Reference Ethics Committee, as well as the Spanish Medicine Agency, approved the study protocol (Study TTD-05-02; EudraCT: 2005-003325-67; clinicaltrials.gov identifier NCT00335595). Study procedures were carried out in accordance with the Declaration of Helsinki and its subsequent amendments and Good Clinical Practice guidelines. Written informed consent was obtained from all patients before enrolment.

Patients and Study Design
The design of the MACRO study has been reported previously [36]. In brief, patients aged$18 years, with histologically confirmed mCRC, Eastern Cooperative Oncology Group performance status (ECOG PS)#2, measurable disease, no previous chemotherapy for advanced disease, adequate hepatic and renal function, and no contraindications to bevacizumab therapy were included.
The primary endpoint of the MACRO study was PFS; secondary endpoints included OS, objective response rate (ORR), toxicity and several translational research assessments. Between July 2006 and September 2008, 480 patients were entered into the study; 239 were randomized to maintenance XELOX plus bevacizumab after induction XELOX plus bevacizumab and 241 were randomized to single-agent bevacizumab after induction XELOX plus bevacizumab. Induction XELOX consisted of 6 cycles of bevacizumab (7.5 mg/kg intravenously [iv] d1), capecitabine (1000 mg/m 2 orally bid d1-14) and oxaliplatin (130 mg/m 2 iv d1) every 3 weeks followed by XELOX plus bevacizumab or bevacizumab alone until progression.

KRAS Mutation Analysis
Evaluation of KRAS status was performed retrospectively. As KRAS analysis is standard practice in Spain, sample analysis was performed either at the treating centre or centrally using existing platforms. Participating centres sent KRAS findings to the data collection centre. Data were obtained on KRAS status (WT or MT), the type of mutation found (12 Ala, 12 Arg, 12 Asp, 12 Cys, 12 Ser, 12 Val or 13 Asp) and the methodology used (method DxS, StripAssay and sequencing). These findings were correlated with patient's existing data, including response rate, PFS, OS, rescue surgery and second-line therapy.

Statistical Analysis
The primary objective of this ancillary analysis of the MACRO study was to evaluate the utility of tumour KRAS status as a prognostic factor in patients with mCRC who were treated with chemotherapy and bevacizumab. The MACRO study demonstrated that single-agent bevacizumab was not statistically significantly inferior to XELOX plus bevacizumab as maintenance therapy [36]; therefore tumour KRAS mutation data from patients in the two treatment groups were combined for the purposes of the present analysis. PFS and OS curves were calculated according to KRAS tumour mutation status using the Kaplan-Meier method. The prognostic value of the biological marker was determined using the log-rank test. Univariate and multivariate Cox proportional hazards models were built with the following variables: ECOG PS 0-1 versus 2; age ,70 versus $70 years; number of metastatic sites 1 versus $2; lactate dehydrogenase (LDH) high levels versus within the normal range; alkaline phosphatase high levels versus within the normal range; male versus female sex; KRAS MT versus WT status; maintenance treatment (XELOX-bev vs bev), prior adjuvant chemotherapy and radiotherapy and surgical removal of metastases prior to the study entry.

Patient Characteristics
The intent-to-treat (ITT) population of the MACRO study comprised 480 patients, 394 (82.1%) of whom had information on KRAS status and were included in this biomarker sub-study ( Figure 1).
Patient characteristics according to KRAS status are shown in Table 1. Significant differences were observed between the two groups in prior adjuvant chemotherapy, prior radiotherapy and surgical removal of metastases prior to entry into the study.

Prognostic Value of KRAS
Median OS was 28.0 months in patients with WT KRAS tumours who received post-study anti-EGFR therapy versus 20.2 months in those with MT KRAS tumours (HR: 1.68; 95% CI: 1.25-2.26; p = 0.0006). OS was longer in patients with WT KRAS tumours who did not receive anti-EGFR therapy than in patients with MT KRAS tumours (26.9 months versus 20.2 months; HR: 1.48; 95% CI: 1.02-2.16; p = 0.0379). There was no difference in OS between patients with WT KRAS tumours who did not receive anti-EGFR therapy and those who did (28.

Univariate and Multivariate Analysis
Results of the univariate and multivariate analyses for PFS are shown in Table 3

Discussion
The present analysis of the MACRO study indicates that tumour KRAS status is a prognostic factor in patients with mCRC receiving bevacizumab in combination with capecitabine plus oxaliplatin. Patients with WT KRAS tumours had a significantly greater clinical benefit than those with MT KRAS tumours in terms of ORR (57.5% versus 43.4%), PFS (10.9 versus 9.4 months) and OS (26.7 versus 18.0 months). The same trend was observed when both treatment arms were combined and when analysed separately. Univariate and multivariate analyses showed the independent role of KRAS status for both PFS and OS. This indicates that in the MACRO study, KRAS was a prognostic factor in patients with mCRC receiving bevacizumab in combination with chemotherapy.
The prognostic value of KRAS was initially suggested by the RASCAL I study, which included 2721 patients [16]. Multivariate analysis established that the effect of KRAS was independent of other variable factors such as sex, tumour site or Dukes stage. The    RASCAL study suggested that KRAS was important not only for carcinogenesis of CRC but also for prognosis in patients with all stages of the disease. Other studies have confirmed the prognostic value of KRAS status [14,15,[17][18][19]25,26]. Several studies have evaluated the role of KRAS status in patients with mCRC receiving first-line treatment with bevacizumab (  [30]. In the subsequent analysis by Hurwitz et al., response rates (60% versus 43%) and PFS (13.5 versus 9.3 months) were numerically greater for bevacizumab-treated patients with WT versus MT KRAS tumours, although the difference in PFS was not statistically significant (HR 0.66; p = 0.09) [31]. Other studies have also reported that KRAS mutation status is not a prognostic factor for patient outcome [32][33][34][35]. This variability in results could be a result of several factors, including the number of patients included in the different trials, the proportions of patients tested for KRAS mutation status and the technology used, the chemotherapy regime under investigation, and subsequent secondand third-line therapies.
Our study did not aim to determine whether KRAS status was predictive in patients receiving bevacizumab, as all patients were treated with bevacizumab. Results from the studies by Hurwitz et al. [31] and Ince et al. [30] appear to show that KRAS status is not predictive and that all patients can benefit from bevacizumab treatment. Our study did not include a bevacizumab-free control arm and therefore we cannot state with certainty that patients with MT KRAS tumours benefit from bevacizumab treatment;  however, bevacizumab does not seem to have an adverse effect on survival, as has been seen in some studies of the anti-EGFR agents [3,35]. The benefit of treatment with bevacizumab appears to be greater in patients with WT KRAS tumours. Our study has several limitations. Firstly, the KRAS mutation analysis was retrospective, which is common in randomized studies examining the role of biomarkers [7,8,31]. Secondly, KRAS mutation analyses were performed either at the treatment centre or at a Spanish reference centre, with the result that different technologies were used, potentially introducing a bias. However, the percentage and type of mutations found in MACRO was consistent with reports from other studies [7,8], as well as the Determina KRAS project, which has analysed samples from 12 262 patients in five Spanish reference centres [37]. In addition, KRAS data were not available for 18% of the population, potentially introducing selection bias. Another possible bias in our study is the imbalance observed in some parameters between patients with WT and MT KRAS tumours (prior adjuvant therapy and surgery for metastatic disease prior to the study entry). Whether this is a chance finding, or because patients with MT KRAS tumours have a different natural history with more aggressive disease, remains to be resolved. None of these clinical features had a significant value in the univariate and multivariate analyses.
It is possible that the difference observed in OS may be a result of patients with WT KRAS tumours receiving anti-EGFR therapy after the treatment of the study. In our study, 49.8% of patients with WT KRAS tumours received an anti-EGFR agent after bevacizumab treatment. The difference was apparent in the survival curves for patients with WT KRAS tumours who received anti-EGFR, those with WT KRAS tumours who did not receive an anti-EGFR agent and patients with MT KRAS tumours. Patients with WT KRAS tumours who received an anti-EGFR therapy had significantly longer OS than patients with MT KRAS tumours (28.0 versus 20.2 months; HR: 1.68; p = 0.0006), yet there was no significant difference in patients with WT KRAS tumours who were treated with an anti-EGFR versus those who were not (28.0 versus 26.9 months; HR: 1.13; p = 0.5373). Later lines of therapy might influence OS, but response rates and PFS, which are not influenced by second-and third-line therapy, were also better in patients with WT KRAS tumours and may be a more appropriate marker for the success of treatment than OS.
In conclusion, this analysis of the MACRO study highlights the prognostic role of tumour KRAS mutation status in patients receiving chemotherapy in combination with bevacizumab, which is consistent with some literature reports but not others. The reasons for the discrepancy between studies are not yet apparent, and sufficiently powered, prospective studies will be required to answer this question. The importance of KRAS mutation type, as suggested by the RASCAL studies (Andreyev et al 2001;Andreyev et al 1998) and the correlation of KRAS status with microsatellite instability suggested by the PETACC-3 study [24] and other biomarkers such as BRAF, PTEN and PIK3CA [38], remain to be explored.