Background

Colorectal cancer (CRC) is one of the most common malignant tumors, with an estimated 1.4 million new cases and nearly 700 thousand cancer deaths reported worldwide in the year 2012 [1]. The incidence and mortality rates of colorectal cancer rank fifth in China [2], and the incidence is still rising [3]. The prognosis of metastatic CRC (mCRC) is poor with a 5-year overall survival (OS) rate < 15% [4, 5]. Patients with unresectable mCRC receiving supportive care alone have been shown to have a poor prognosis, with a median OS of 5 months [6]. By contrast, the 5-year OS rate of patients with unresectable mCRC who received first-line chemotherapy with 5-fluorouracil (5-FU)/leucovorin (LV) plus oxaliplatin (FOLFOX) was 10% [7].

The FOLFOX regimen has become the standard first-line therapy for the treatment of unresectable mCRC [8,9,10,11]. Other recommended first-line chemotherapy regimens include capecitabine plus oxaliplatin (CapeOX), FOLFOX plus bevacizumab, CapeOX plus bevacizumab, and FOLFOX plus cetuximab (KRAS/NRAS wild type only) [12, 13]. However, after the failure with first-line therapy, which combination chemotherapy regimen would be the optimal second or third-line treatment is yet to be confirmed, as such strategic trials investigating these are urgently needed.

CMAB009, a recombinant, human/mouse chimeric monoclonal antibody (mAb) specifically targeting the human epidermal growth factor receptor (EGFR), competitively inhibits ligand-binding and interrelated downstream signaling. It has the same amino acid sequence as ERBITUX® (cetuximab), but slightly different abilities for glycosylation and other post-translational modifications (PTMs). CMAB009 is expressed by the Chinese hamster ovary (CHO) cells while cetuximab is expressed by the mouse cell line SP2/0 which also expresses the gene for α-1,3-galactosyltransferase [14]. In most patients who have developed a hypersensitivity reaction to cetuximab, IgE antibodies against the cetuximab were found to be already present in their serum prior to the start of the therapy [14]. These antibodies were found to be specific for galactose-α-1,3-galactose (Gal (α 1-3) Gal). Since CHO cells do not produce α-1,3-galactosyltransferase, they have a pattern of glycosylation that differs from that of SP2/0 [14,15,16], as such, CMAB009 expressed in CHO cells has a lower level of Gal (α 1-3) Gal-containing glycans [15]. This suggests that CMAB009 might have lower immunogenicity and reduced hypersensitivity reactions as compared to cetuximab.

In our previous retrospective study, we have shown that CMAB009 demonstrated good efficacy and acceptable tolerance in patients with chemotherapy-resistant advanced CRC [17]. In this study, we aimed to prospectively determine the clinical efficacy and safety of CMAB009 plus irinotecan as compared to that of irinotecan-only in KRAS wild-type mCRC patients who had treatment failure with first-line FOLFOX regimen.

Patients and methods

Patient selection

This prospective, open-label, randomized, phase III trial was conducted at 38 centers in China (Table 1). Patients were eligible if they had previous documented treatment failure (disease progression or discontinuation due to toxicity) with FOLFOX regimen for histologically confirmed mCRC and had wild-type KRAS mutation. Other inclusion criteria were: age between 18 and 70 years, an Eastern Cooperative Oncology Group (ECOG) performance status (PS) score of 0 or 1, a life expectancy of more than 3 months starting from the time of enrollment, no other malignant tumors, except for patients who had been cured for cervical carcinoma in situ, skin basal carcinoma, or squamous cell carcinomas. The exclusion criteria were: chemotherapy within 4 weeks prior to enrollment, abnormal serum hematologic function [hemoglobin (Hb) < 90 g/L; platelet count (PLT) < 100 × 109/L; absolute neutrophil count (ANC) < 1.5 × 109/L; or white blood cell count (WBC) < 4.0 × 109/L), abnormal hepatorenal function (total bilirubin (TBIL), more than onefold higher than the upper limit of the normal range; blood urea estrogen (BUN) and creatinine (Cr), more than 1.5-fold higher than the upper limit of the normal range; or alanine aminotransferase (ALT) and aspartate aminotransferase (AST), more than fivefold higher than the upper limit of the normal range with hepatic metastases or more than 2.5-fold higher than the upper limit of the normal range without hepatic metastases], serious cardiac insufficiency, known history of brain metastases, and prior therapy with EGFR-targeting agents. Women who were pregnant or breastfeeding were also excluded.

Table 1 Eligible KRAS wild-type patients were identified at 38 hospitals in China

The protocol of this study was approved by the ethics committee board at each center and all patients provided signed informed consent before participation.

Study design

Patients were randomly assigned in a 2:1 ratio to receive either CMAB009 (Shanghai Zhangjiang Biotech Co., Shanghai, China) plus irinotecan (Qilu Pharma, Jinan, Shandong, China) or irinotecan-only, respectively. Patients receiving irinotecan-only could switch to CMAB009 therapy (labeled as the sequential-CMAB009 arm) upon diagnosis of disease progression.

The primary endpoints were overall response rate (ORR) and median progression-free survival (PFS). ORR was defined as the proportion of patients with a confirmed complete response (CR) or partial response (PR) according to response evaluation criteria in solid tumor (RECIST) version, 1.0. PFS was defined as the time from the date of entry into the trial to the date of first observed treatment failure (local and/or regional persistence/recurrence or distant metastasis) or death from any cause. The secondary endpoints were median OS (time from the date of entry into the trial to the date of death or the date the patient was last known to be alive), disease control rate [DCR, the duration of CR, PR, and stable disease (SD)], clinical benefit rate (CBR, defined as the sum of the number of patients who achieved CR, PR, and SD, and remained stable for a more than 24 weeks) and duration of response (DOR, time from the date of first evidence of CR or PR to the date of objective progression or the date of death due to any cause), and treatment safety.

Treatment

Patients assigned to the CMAB009 plus irinotecan arm received an initial dose of CMAB009 at 400 mg/m2 intravenously over 2 h on day 1, and then 250 mg/m2 over 1 h weekly. Irinotecan, at a dosage of 180 mg/m2 intravenously, or 125–135 mg/m2 intravenously for those with prior pelvic/abdominal irradiation, was given over 90 min and was administered every 2 weeks in both treatment arms; starting more than 1 h after the CMAB009-infusion completion for patients in the CMAB009 plus irinotecan arm. Each treatment cycle lasted 2 weeks. The dosage of CMAB009 for patients in the sequential-CMAB009 arm was similar to that of the CMAB009 plus irinotecan arm. The treatments were continued until disease progression, unacceptable toxicity, or the patient withdrew consent.

The National Cancer Institute Common Terminology Criteria for Adverse Event Criteria (NCI CTCAE) version 3.0 was used to assess adverse events. The definition and grading of hypersensitivity reactions were based on documented symptoms list in the criteria, the characteristics of grade 1 reaction were transient flushing or rash; drug fever < 38 °C; those of a grade 2 reaction were rash; flushing; urticaria; dyspnea; drug fever ≥ 38 °C; and those of a grade 3 reaction were symptomatic bronchospasm, with or without urticaria; parenteral medication(s) indicated; allergy-related edema/angioedema; hypotension. Anaphylaxis and death were considered as a grade 4 and 5 reaction, respectively. CMAB009 was discontinued upon the occurrence of grade 3/4 hypersensitivity, after which the dose of irinotecan was to be reduced to 125–135 mg/m2 when grade 3/4 neutropenia, febrile neutropenia, thrombocytopenia, and leucopenia occurred. In the event of grade 4 nonhematologic toxicities (excluding diarrhea), both agents were discontinued.

Assay to detect mutant KRAS

The tissue specimens (surgery or biopsy from the primary or metastatic tumor) of mCRC patients were evaluated at the central laboratory of the Chinese National Human Genome Center in Shanghai, and only patients with available KRAS mutational status at codon 12, 13 were included. Formalin-fixed, paraffin-embedded tumor sections were deparaffinized and air dried, and DNA was extracted using standard Proteinase K digestion and a DNeasy minispin column (Qiagen, Valencia, CA, USA). Mutant KRAS was detected using a validated DNA sequencing method that identifies seven somatic mutations located in codons 12 and 13 (Gly12Asp, Gly12Ala, Gly12Val, Gly12Ser, Gly12Arg, Gly12Cys, and Gly13Asp) using allele-specific real-time polymerase chain reaction at the central laboratory of The Chinese National Human Genome Center, Shanghai, China [18, 19].

Response assessment

Measurable lesions were obtained at baseline (within the 4 weeks prior to the start of treatment) and evaluated every 6 weeks by computed tomography (CT) scans. Tumor response was assessed by local investigators based on the RECIST criteria version 1.0, until disease progression. After treatment completion, a follow-up assessment was conducted every 4 weeks, for up to 5 years after the last dose or until the patient succumbed or the last date of follow-up (July 23, 2015).

Immunogenicity assessment

Blood samples were taken at week 0 (before CMAB009 infusion) and at 6, 12, 18, and 30 weeks after the first infusion, to determine the presence of ADA, which was analyzed using a competitive inhibition assay by the Surface Plasmon Resonance (SPR) (Shanghai Zhangjiang Biotechnology, Shanghai, P. R. China) while presence of neutralizing ADAs (NAb) were analyzed by competitive enzyme-linked immunosorbent assay (ELISA) (Shanghai Zhangjiang Biotechnology, Shanghai, P. R. China).

Statistical analysis

At least 333 patients (CMAB009 plus irinotecan arm: 222 patients; irinotecan-only arm: 111 patients) were required to obtain a 90% power to detect an absolute difference in ORR. This in turn meant that after accounting for a typical study dropout rate of 10%, 495 patients (CMAB009 plus irinotecan: irinotecan-only, 330:165) were to be enrolled in the study to meet statistical requirements by satisfying the minimum number of patients outlined by the China Food and Drug Administration (CFDA, http://www.nmpa.gov.cn/WS04/CL2174/300629_9.html). An O’Brien and Fleming type α spending function was used to ensure an overall, two-sided, type I error rate of 5%. The ORR was compared between the treatment arms using a Cochran–Mantel–Haenszel test stratified by ECOG PS score (0 vs. 1). DCR, CBR, and DOR were assessed according to RECIST criteria, version 1.0. PFS and OS were analyzed by the Kaplan–Meier method. Primary comparisons between the treatment arms were made using a two-sided log-rank test stratified by ECOG PS. Hazard ratios with 95% confidence intervals (CI) were calculated from stratified Cox regression models with gender, age, and ECOG PS score.

Results

Patient characteristics

Between May 31, 2009, and September 23, 2011, a total of 1077 patients were assessed for eligibility. There were 35 patients with insufficient or poor-quality DNA samples. The observed KRAS mutation (codons 12 and 13) rate was 32.3% (337/1042). After exclusion of non-eligible patients, 512 KRAS wild-type patients, 342 in the CMAB0009 plus irinotecan arm and 170 in the irinotecan-only arm, were enrolled from 38 sites in China (Fig. 1, Table 1). The study arms were well balanced for clinical characteristics (Table 2).

Fig. 1
figure 1

Flow chart illustrating the trial enrollment and patient outcomes

Table 2 Baseline demographic and clinical characteristics of the 501 patients before the start of treatment

Treatment exposure

The median number of treatment cycles was 8 (1–80 cycles) for the CMAB009 plus irinotecan arm, 5 (1–27 cycles) for the irinotecan-only arm, and 4 (1–36 cycles) for the sequential-CMAB009 arm. The median irinotecan treatment duration was longer for the CMAB009 plus irinotecan arm (14.0 weeks; range 2.0 to 102.6 weeks) as compared to the irinotecan-only arm (10.0 weeks; range 2.0 to 53.2 weeks). In the CMAB009 plus irinotecan arm, the median CMAB009 treatment duration was 16.3 weeks (range 1.0 to 159.7 weeks). There were 115 patients who switched to CMAB009 treatment from the irinotecan-only therapy (sequential-CMAB009 arm) and the median treatment duration of CMAB009 was 7.4 weeks (range 1.0 to 72.0 weeks). The median dose intensity of irinotecan was higher in the irinotecan-only arm (97.6 mg/m2/week) than in the CMAB009 plus irinotecan arm (92.6 mg/m2/week). In the CMAB009 plus irinotecan arm, the median CMAB009 dose intensity was 263.3 mg/m2/week. For the sequential-CMAB009 arm, the median dose intensity was 286.4 mg/m2/week.

A dose modification of 35.0% (118/337) was recorded for irinotecan in the CMAB009 plus irinotecan arm and 20.1% (33/164) in the irinotecan-only arm. Dose modification for CMAB009 was 18.1% (61/337) in CMAB009 plus irinotecan arm and 12.2% (14/115) in the sequential-CMAB009 arm.

The safety analysis population consisted of patients who received at least one dose of the study drug and had at least one safety assessment after treatment administration (338 in the CMAB0009 plus irinotecan arm and 165 in the irinotecan-only arm). The efficacy analysis was performed in patients with at least one dose of the study drug and had complete baseline data (337 in the CMAB0009 plus irinotecan arm and 164 in the irinotecan-only arm) (Fig. 1).

Treatment efficacy

Primary endpoints

The tumor response was evaluated in 501 investigated patients. The ORR was 33.2% (112/337) and 12.8% (21/164) in the CMAB009 plus irinotecan and irinotecan-only arms, respectively (P < 0.001, Table 3). For the sequential-CMAB009 arm, 13.9% (16/115) of the patients achieved PR and 49.6% (57/115) demonstrated SD.

Table 3 Therapeutic efficacies of CMAB009 plus irinotecan treatment versus irinotecan-only treatment

The median PFS was significantly longer in the CMAB009 plus irinotecan arm than in the irinotecan-only arm (169 vs. 95 days; HR, 0.50; 95% CI 0.40 to 0.63; P < 0.001) (Fig. 2). In the sequential-CMAB009 arm, the median PFS was 84 days (95% CI 65–113 days).

Fig. 2
figure 2

Kaplan-Meier analysis comparing the progression-free survival of patients from the CMAB009 plus irinotecan arm to those in the irinotecan-only arm only. PFS, progression free survival; HR, hazard ratio; CI, confidence intervals

Secondary endpoints

Median OS was 425 days in the CMAB009 plus irinotecan arm and 401 days in the sequential-CMAB009 arm (HR, 1.02; 95% CI 0.82 to 1.28; P = 0.940) (Fig. 3). The DCR and CBR were higher for patients in the CMAB009 plus irinotecan arm as compared to the irinotecan-only arm (both P  < 0.001, Table 3). In the sequential-CMAB009 arm, the DCR and CBR were 63.5% and 23.1%, respectively.

Fig. 3
figure 3

Kaplan-Meier analysis comparing the overall survival of patients from the CMAB009 plus irinotecan arm to those in the sequential-CMAB009 arm. OS, overall survival; HR, hazard ratio; CI, confidence intervals

DOR in the CMAB009 plus irinotecan arm was almost twice of that in the irinotecan-only arm (210 vs. 109 days, HR, 0.39; 95% CI 0.22 to 0.66; P < 0.001; Fig. 4). For the sequential-CMAB009 arm, the DOR was 148 days (95% CI 59 to 230 days).

Fig. 4
figure 4

Kaplan–Meier analysis for the duration of treatment response. Median DOR was significantly longer in patients who received CMAB009 plus irinotecan compared with those who received irinotecan-only. DOR, duration of response

Treatment safety

Over the course of the study, 96.2% (484/503) of patients experienced at least one adverse event. There were 55.3% (187/338) of patients in the CMAB009 plus irinotecan arm and 37.6% (62/165) of patients in the irinotecan-only arm who experienced at least one grade ≥ 3 adverse event. Although the frequency and severity of some adverse events were greater in the CMAB009 plus irinotecan arm, however, CMAB009 plus irinotecan was generally well-tolerated. The most common adverse events (Table 4) consisted of diarrhea (39.6% vs. 35.8%), emesis (18.9% vs. 37.0%), leucopenia (50.0% vs. 39.4%), neutropenia (30.5% vs. 19.4%), and fatigue (22.2% vs. 16.4%) in CMAB009 plus irinotecan and irinotecan-only arms, respectively. Hypersensitivity reactions were experienced in 0.9% (4/453) of patients in this study, three (one grade 1, one grade 2 and one grade 3) in the CMAB009 plus irinotecan arm and one (grade 2) in the sequential-CMAB009 arm (Table 4). As expected, the addition of CMAB009 to irinotecan significantly increased the risk of skin rash (66.9% vs. 5.5%, P < 0.001) and paronychia (9.8% vs. 0.0%, P < 0.001). Additionally, one patient in the sequential-CMAB009 arm developed grade 4 neutropenia but was determined unrelated to the CMAB009.

Table 4 Treatment-emergent AE occurring during the study in the safety analysis set

Immunogenicity

A total of 1115 serum samples were obtained from 310 patients (249 patients from the CMAB009 plus irinotecan arm and 61 patients from the sequential-CMAB009 arm) and were analyzed for the presence of ADA, which was detected in only 3.6% (11/310) of patients. Excluding 4 patients with pre-existing ADA, the incidence of ADA in this study cohort was found to be 2.3% (7/310) (Table 5). All the ADA positive patients were in the CMAB009 plus irinotecan arm and none in the sequential-CMAB009 arm. Samples positive for binding antibodies in the confirmatory ADA assay were further evaluated for the presence of NAb to the CMAB009, of which 1.3% (4/310) patients were found to be positive (Table 5). No significant differences in clinical safety were found between ADA positive patients and ADA negative patients.

Table 5 Summary of the detection of ADA and NAb in the enrolled patients

Discussion

This is the first prospective, open-label, randomized, phase III study comparing the clinical efficacy and safety of an anti-EGFR monoclonal antibody in KRAS wild-type mCRC patients with documented previous failure with FOLFOX regimen. In this study, we have found that patients treated with CMAB009 plus irinotecan demonstrated significantly better ORR and prolonged PFS as compared with those having irinotecan-only. In addition, the combination treatment with CMAB009 was generally well-tolerated and manageable. Therefore, this regimen could be considered as a new standard of treatment in the second-line setting for KRAS wild-type mCRC patients after failure with the FOLFOX regimen.

From the knowledge of the predictive value of KRAS mutation (codons 12 and 13) status for the efficacy of cetuximab, wild-type KRAS is required for evaluating cetuximab efficacy in mCRC patients [19,20,21,22,23]. KRAS mutations have been reported in 30% to 50% of CRC tumors and are also common in other tumor types [19]. In the present study, of the 1077 mCRC initially assessed in the KRAS analyses, of which the KRAS status in 35 patients could not be determined due to insufficient or poor-quality DNA samples, only 512 patients were found to have KRAS wild-type mCRC, demonstrating an observed KRAS mutations incidence of 32.3%. This was within the expected range of previous studies which reported mutation rates of approximately 36% [24,25,26].

Treatment with CMAB009 plus irinotecan, as compared to irinotecan-only, demonstrated significantly improved ORR, reduced the risk of disease progression by nearly 50% (PD, 13.9% vs. 26.8%, respectively), improved clinical efficacy (DOR, 210 days vs. 109 days; DCR, 80.1% vs. 65.2%; CBR, 30.0% vs. 14.6%, respectively), and had an acceptable safety profile. Further, the findings of this study showed that when CMAB009 was used as a third-line treatment in the sequential-CMAB009 arm, 13.9% (16/115) of the patients achieved PR with a PFS of 84 days, and 49.6% (57/115) of the patients achieved SD, comparable with the results of the CO.17 Trial [27] from the National Cancer Institute of Canada Clinical Trials Group (NCIC CTG), in which cetuximab was found to significantly improve PR (8%), SD (31.4%), and quality of life compared to the best supportive care in CRC patients in whom other treatments had failed. Several studies have reported the combination of irinotecan with other drugs, such as modified XELIRI (mXELIRI, capecitabine plus irinotecan) regimen and FOLFIRI (leucovorin, fluorouracil, and irinotecan), with or without bevacizumab regimens. The median OS for mXELIRI with or without bevacizumab was found to be non-inferior (16.8 months vs. 15.4 months) to FOLFIRI with or without bevacizumab regimens for mCRC [28]. A small-scale retrospective study reported that a re-challenge strategy with cetuximab and irinotecan may be active in patients with RAS and BRAF wild-type mCRC with acquired resistance to first-line irinotecan- and cetuximab-based therapy [29]. These results showed that CMAB009/cetuximab plus irinotecan might have good clinical efficacy in KRAS wild-type mCRC patients.

This study was also designed to explore whether CMAB009 plus irinotecan therapy would prolong OS. Our results showed that the median OS was similar between the CMAB009 plus irinotecan and sequential-CMAB009 arms (425 days vs. 401 days, P = 0.940), which we presume was possibly influenced by the sequential CMAB009 treatment in the patients after irinotecan failure, and thereby indicated that either CMAB009 plus irinotecan or sequential-CMAB009 may be considered as an effective treatment choice.

The safety profile of CMAB009 plus irinotecan in our study was comparable with that of other anti-EGFR monoclonal antibodies [27, 30,31,32]. Skin rash is associated with all EGFR inhibitors and is the most frequently associated with cetuximab/CMAB009. This adverse event seems to be closely linked to the biologic activity of cetuximab/CMAB009 as EGFR is expressed on the epidermal keratinocytes and hair follicles, and is thought to play a role in maintaining the skin integrity and follicular homeostasis [33]. Therefore blocking these effects may be responsible for the observed rashes. As such, in the present study, the most noticeable adverse event related with CMAB009 was skin rash (66.9%), which was similar to the cetuximab-related acneiform rash reported in 76.3% of patients in the EPIC study [20]. Notably, the CMAB009 plus irinotecan arm did not significantly increase gastrointestinal toxicity as compared with the irinotecan-only arm.

Humans have baseline levels of antibodies against certain non-human glycan motifs, including N-gl ycolylneuraminic acid (NGNA) and Gal (α 1-3) Gal, and severe hypersensitivity reactions occurring during the initial infusion of cetuximab are mediated by preexisting IgE antibodies against cetuximab [14, 34]. A high prevalence of severe hypersensitivity reactions of approximately 2% was reported in patients who had been injected with cetuximab because cetuximab is attached to N-linked oligosaccharide containing the Gal (α 1-3) Gal motif at the Fab region [16]. This non-human glycan may induce immunogenicity [35]. However, CMAB009 expressed in CHO and has a different glycosylation pattern not containing the NGNA or Gal (α 1-3) Gal motif at the Fab region [34, 36]. Therefore, CMAB009 has lower immunogenicity than cetuximab and the presence of ADA was found to be low at 3.6% (11/310). Only 0.9% (4/453) of patients experienced hypersensitivity reactions in our study, three in the CMAB009 plus irinotecan arm and one in the sequential-CMAB009 arm (Table 4).

Our study had several limitations worth noting. First, this study did not analyze other biomarkers such as NRAS. It was recently reported that NRAS was mutated in 6% of mCRC and were associated with a shorter OS compared to wild-type patients [31]. One meta-analysis showed that non-functional mutation or loss of NRAS, BRAF, PIK3CA, and PTEN predicted poor efficacy of cetuximab [31, 32]. Therefore, to demonstrate the predictive value of RAS and BRAF, we propose a prospective phase III study to explore the clinical efficacy and safety of CMAB009 plus FORFIRI as first-line chemotherapy in RAS/BRAF wild-type patients with mCRC in China. Second, there were some patients failing to provide serum samples and some samples could not be analyzed due to hemolysis. This is because the release of cellular material into the serum or plasma would have introduced additional confounding factors in the downstream analysis of such samples and were therefore excluded from the immunogenicity analysis. Third, the quality of life of the patients was not assessed.

Conclusions

Treatment with CMAB009 plus irinotecan, compared to irinotecan-only, demonstrated superior clinical efficiency and was well tolerated as a second-line of treatment in KRAS wild-type mCRC patients with documented previous failure with the FOLFOX regimen. Therefore, this regimen could be considered as an optimal second-line treatment of choice for such patients. Further, for those whose disease progressed after being treated with irinotecan-only, as a second-line of treatment, switching to CMAB009 can be considered as an effective third-line of treatment.