Background
Colorectal cancer (CRC) is one of the most common human malignancies with more than 300,000 cases both in the United States and in the European Union each year. In the past decade, significant improvements have been performed in response rates, progression-free survival (PFS) and overall survival (OS) [
1‐
4]. Despite these improvements, mainly due to the development of new combinations of standard chemotherapy including 5 fluorouracil, irinotecan and oxaliplatin, nearly all patients with metastatic CRC (mCRC) will die from their disease. Recently, new therapeutic agents targeting molecular events involved in colorectal carcinogenesis have been developed, including bevacizumab, a recombinant humanized monoclonal antibody, which binds to the vascular endothelial growth factor (VEGF) with a high specificity and prevents its interaction with receptors on endothelial cells. VEGF plays a key role in angiogenesis, which is involved in the development of carcinogenesis, tumor growth and malignant dissemination. Therefore, bevacizumab inhibits the activation of VEGF-receptor-mediated signaling pathways and resultant biological effects [
5]. This antiangiogenic agent, added to a 5-fluorouracil (5FU) ± irinotecan-based chemotherapy as first-line treatment, has been shown to improve response rates and survival of mCRC patients when compared to the chemotherapy alone [
6‐
8]. An improvement of PFS was also shown in first-line with the addition of bevacizumab to oxaliplatin-based chemotherapy [
9]. A randomized phase III study also reported a clinical efficacy of the association of bevacizumab and FOLFOX4 as second-line in mCRC patients previously treated with a fluoropyrimidine and irinotecan, with a significant improvement in response rates, PFS and OS when compared to FOLFOX4 alone [
10].
In Europe, bevacizumab was approved by the EMEA (European Medicines Evaluation Agency) in the beginning of 2005 and many CRC patients could not have received it in first or second-line treatment before this date. Moreover, it was reported to have no efficacy in 3rd or later-line, alone or with 5FU [
10,
11]. The aim of this retrospective study was to evaluate the activity of bevacizumab combined with a chemotherapy with 5FU/LV and irinotecan or oxaliplatin in mCRC who have failed prior chemotherapy with fluoropyrimidine plus irinotecan and/or oxaliplatin.
Methods
Patient characteristics
In this retrospective study, we included all the patients prospectively registered in two centers (Hôpital Ambroise Paré, Boulogne Billancourt and Centre Régional de Lutte contre le Cancer Val d'Aurelle, Montpellier, France) with histologically proven mCRC who had been previously treated with a fluoropyrimidine (e.g., fluorouracil or capecticabine) plus irinotecan and/or oxaliplatin with no response to treatment (as defined by tumor progression according to the RECIST criteria [
12] or unacceptable toxicity) and who then received bevacizumab in combination with FOLFIRI or simplified FOLFOX4 between May 2005 and October 2006. The decision of treatment was always taken during multidisciplinary staff for patients who had not the opportunity to receive bevacizumab at an earlier line of chemotherapy. The usual exclusion criteria were a history of major surgery within 28 days, a thrombotic or bleeding event within 6 months, a hypertension, a clinically significant cardiovascular disease, a hypertension, a therapeutic anticoagulation and the presence of brain metastases. All medical files of the patients were registered prospectively in a computerised database (after national registry council (CNIL) authorization). The following data were collected and analyzed: age and performance status (according to WHO criteria) at the time of the first cycle of bevacizumab, gender, primary tumor site (colon or rectum), number and localization of metastatic sites, previous anticancer drugs received and tumor response to them. This retrospective study was proposed in January 2007 and approved by the local scientific and ethical committee.
Treatment protocols
Patients were treated with bevacizumab given at a dose of 5 mg/kg on day 1 every two weeks, followed by a 2-hour infusion of 400 mg/m2 of leucovorin, given simultaneously with a 2-hour infusion of 85 mg/m2 of oxaliplatin, followed by a bolus of 400 mg/m2 of 5FU and then a 46-hour infusion of 2400 mg/m2 of 5FU (simplified FOLFOX4) or the same regimen with a 90 minutes (min) infusion of 180 mg/m2 of irinotecan instead of the infusion of oxaliplatin (FOLFIRI) (1 cycle = 14 days). The first infusion of bevacizumab was given over 90 min, the second over 60 min and the following over 30 min when previous infusions were well tolerated.
Tumor evaluation
All the patients must have received at least four cycles of chemotherapy with bevacizumab to be evaluable, but two patients had evidence of a clinical progression before the radiologic evaluation of tumor response and received only three cycles of bevacizumab.
Tumour response was prospectively assessed every four cycles according to RECIST criteria by computerized tomography-scan. Treatment was repeated until the occurrence of disease progression or unacceptable toxicity.
Toxicity was assessed according to the Common Toxicity Criteria version 2.0[
13].
Statistical analysis
Fisher's exact test was used to compare response and stabilization rates between groups. The PFS was calculated as the period from the first day of bevacizumab treatment to the date of tumor progression, to death from any cause or to the date of the last follow-up at which data point was censored. The OS time was calculated as the period from the first day of bevacizumab treatment until death of any cause or until the date of the last follow-up, at which data point was censored. Both PFS and OS were estimated by the Kaplan-Meier method and compared using the log-rank test. Analysis was carried out using the STATA software package (College Station, Texas). The level of significance was set at p = 0.05.
Discussion
This study reports an activity of bevacizumab at the dose of 5 mg/kg combined with FOLFOX or FOLFIRI regimen in patients who have failed prior chemotherapy with fluoropyrimidine, irinotecan and/or oxaliplatin. Indeed, 32% of the patients could achieved an objective response and one patient could secondary have a curative resection of its liver metastases. The disease control rate was superior to 70%. The addition of bevacizumab to FOLFIRI or FOLFOX was also associated with a PFS and a OS of 9.7 months and 18.4 months respectively. These response and survival rates are higher than those reported by Giantonio et al. with the combination of FOLFOX plus bevacizumab in second line treatment in patients previously treated by fluoropyrimidine and irinotecan, but this might be certainly explained by the small size and the retrospective nature of our study. Furthermore, patients who received bevacizumab during the period of the study were highly selected because bevacizumab was not registered in this setting but only in first line treatment. For these reasons, our results have to be taken with caution. However, they are nonetheless interesting in third-line treatment where few molecules have been shown to be effective. In patients pretreated with fluroropyrimidine, irinotecan and/or oxaliplatin, several studies have reported poor response and survival rates with standard chemotherapy such as the association of mitomycine C and capecitabine [
15‐
18]. The antibodies against the Epidermal Growth Factor Receptor (EGFR) are the only therapy to have shown their efficacy in chemorefractory CRC patients. Cetuximab was the first to show a significant response rate in association with irinotecan in irinotecan-refractory mCRC in the phase II BOND study, in which PFS and OS were of 4.1 and 8.6 months respectively [
19]. The superiority of cetuximab monotherapy over best supportive care (BSC) in terms of survival was demonstrate more recently [
20]. The fully human anti-EGFR antibody panitumumab was also associated with a longer PFS when compared to BSC in a recent randomized phase III trial [
21]. However, it is now clearly demonstrated that patients with a tumor
KRAS mutation are resistant to anti-EGFR antibodies and do not benefit from this targeted therapy [
22‐
26]. Bevacizumab should therefore be considered for these patients after failure of conventional chemotherapy. In our study, 61% of the patients had previously received cetuximab and were so resistant to all potentially efficient therapies approved in the treatment of mCRC.
In patients who were previously treated with the same chemotherapy regimen (90% of all patients included), the addition of bevacizumab to this regimen allowed a response rate of 35.7% and a stabilization of 39.3%, which suggests that the antiangiogenic therapy may circumvent resistance to conventional chemotherapy by allowing a more efficient delivery of chemotherapeutic agents. Tumor vasculature is structurally and functionally abnormal, which results in an heterogeneity in tumor blood flow with interstitial hypertension, hypoxia and acidosis [
27]. Hypoxia could therefore renders tumor cells resistant to several cytotoxic drugs by interfering with the penetration of these drugs throughout the tumor. The concept of "normalization" of tumor vasculature induced by bevacizumab, developed by Jain [
28] who demonstrated with some colleagues this effect of VEGF blockade in rectal carcinoma [
29], may therefore explain the circumvention of resistance to conventional therapies observed in our study. This phenomenon do not seem to be observed with the combination of an anti-EGFR antibody with bevacizumab considering the disappointing results of two recent randomized studies in which the addition of cetuximab or panitumumab to bevacizumab in combination with a conventional chemotherapy was not associated with any benefit but on the contrary perhaps with a deleterious effect [
30,
31], which remains to be elucidated.
A first randomized phase II trial of bevacizumab plus bolus 5FU/LV compared to 5FU/LV alone in untreated mCRC suggested an improvement of response rate, PFS and OS with the combined treatment [
7]. Then, the large randomized phase III study reported by Hurwitz et al. [
6] demonstrated that the addition of bevacizumab to irinotecan and bolus 5FU/LV (IFL regimen) in first-line treatment was associated with a significant prolongation of PFS (6.2 months versus 10.6 months) and OS (15.6 months versus 20.3 months), as a better response rate (34.8% versus 44.8%). The results of that pivotal study led to the registration of bevacizumab in the fist-line setting in the United States and in Europe. The association of bevacizumab to 5FU/LV was also shown to provide a statistically significant and clinically relevant benefit to patients with previously untreated mCRC [
6‐
8,
32]. More recent studies showed that bevacizumab with FOLFIRI or FOLFOX4 regimen was superior to the chemotherapy alone in this setting [
9,
33]. In second-line, the ECOG E3200 phase III trial showed that bevacizumab improved response rate, PFS and OS when added to the FOLFOX4 regimen in patients pre-treated with a fluoropyrimidine and irinotecan [
10]. Bevacizumab in combination with 5FU/LV plus irinotecan or oxaliplatin is therefore proved to be beneficial in first and 2
nd-line setting. In the absence of cross-over after progression in the chemotherapy arm of the studies evaluating the addition of bevacizumab to a combined chemotherapy [
6,
10], the effect of the antiangiogenic agent associated with a bichemotherapy after the 2
nd-line of treatment is not known. However, it was associated with a poor response rate (1%) in combination with a bolus regimen of 5FU/LV in chemotherapy-refractory mCRC [
11]. In this large multicenter phase II study, PFS and OS were 3.7 months and 9.1 months respectively. This result, adding to the lack of efficacy of bevacizumab in monotherapy[
10] led us to test it in this setting in combination with the more active FOLFOX4 or FOLFIRI regimens. Bevacizumab was administrated at the dose of 5 mg/kg every 2 weeks as it was reported in first-line studies with 5FU/LV/irinotecan or oxaliplatin [
6,
9,
33] and since the 10 mg/kg dose was not previously shown to be superior [
7].
Our results are consistent with those of a recent small study conducted in 14 mCRC patients that progressed after oxaliplatin and irinotecan and for whom bevacizumab plus infusional 5FU/LV and irinotecan allowed a 28.5% response rate with 57% of stabilized patients [
34]. Median PFS was 3.9 months and OS was 10.9 months.
In our study, bevacizumab was well tolerated, with only one interruption due to a reversible posterior leukoencephalopathy syndrome associated with hypertension [
14]. The other adverse events were easily manageable. As concerns chemotherapy-induced toxicity, it was relatively low because most of the patients previously received the same regimen as that associated with bevacizumab. Therefore required dose modifications were already performed before starting bevacizumab therapy
Competing interests
AL and MY: received honoraria from Roche France
PR: received research grant from Roche France
Authors' contributions
AL, JBB, ES, CBG collected the clinical data. AL, ES, EM, EA, CL, JBB, FP, JNV, MY and PR treated and followed the patients included in the study. PR, EM, MY and AL were involved in the conception of the study. EM performed the statistical analysis. AL wrote the manuscript. ES, EM, PR and MY were involved in the interpretation of data and critically revised the manuscript. All the authors read and approved the final manuscript.