ESMO 2014: Results From the CALGB/SWOG 80405 and FIRE-3 (AIO KRK-0306) Studies In All RAS Wild Type Population

Optimal treatment strategy with anti-EGFR or anti-VEGF treatments in first-line RAS wild type metastatic colorectal cancer patients

During the special session at ESMO 2014 Congress (Madrid, Spain) on defining the optimal treatment strategy with anti-EGFR or anti-VEGF treatments in first-line RAS wild type metastatic colorectal cancer patients, the CALGB/SWOG 80405 researchers presented long awaited results of overall survival in all RAS wild type population, while FIRE-3 study researchers  presented independent radiological evaluation of objective response rate, early tumour shrinkage, and depth of response in the final RAS evaluable population.

The session was moderated by Dr Dirk Arnold and the data were discussed by Doctors Andres Cervantes, Alberto Sobrero, and Fortunato Ciardiello.

CALGB/SWOG 80405: Phase III trial of FOLFIRI or mFOLFOX6 with bevacizumab or cetuximab for patients with expanded RAS analyses in untreated metastatic adenocarcinoma of the colon or rectum

Overall survival in all RAS wild type patients in CALGB/SWOG 80405 study

Overall survival in all RAS wild type patients in CALGB/SWOG 80405 study. © Heinz-Josef Lenz

Dr Heinz-Josef Lenz of the Division Of Medical Oncology, University of Southern California Norris Comprehensive Cancer Center, Los Angeles, USA presented results of expanded RAS analysis in the CALGB/SWOG 80405 study. Earlier this year, the study researchers presented the findings with a conclusion that FOLFIRI/cetuximab and mFOLFOX6/bevacizumab are equivalent in terms of overall survival (OS) in patients with previously untreated KRAS wild type (codons 12 and 13) metastatic colorectal cancer and that either regimen is appropriate in first-line treatment. The OS longer than 29 months and 8% of long-term survivors confirmed the progress in this setting. However, the oncology community expected that expanded RAS and other molecular and clinical analyses may identify subsets of patients who get more or less benefit from specific regimens.

The original CALGB/SWOG 80405 study included unselected patients with metastatic colorectal cancer who received treatment according to physician-selected chemotherapy (FOLFIRI or mFOLFOX6) and were randomised to cetuximab, bevacizumab or both (the third arm was subsequently closed). After 1420 patients accrued the study was amended as following: only patients with wild type KRAS tumours (codon 12 and 13) were included. Accrual goal was 1142 patients.

Between November 2005 and March 2012, 3058 unselected patients were enrolled and 2334 KRAS wild type patients randomised. The final number included 1137 patients (333 pre-amend eligible retrospective KRAS test, and 804 post-amend).

Expanded RAS was tested in all wild type RAS exon 2 using beaming technology including KRAS exon 3,4 and NRAS exon 2, 3 and 4 with a detection sensitivity of 0.01%. The primary endpoint was OS.

In expanded RAS wild type population, the median OS was pushed beyond 30 months. However, there was no significant difference between the cetuximab and bevacizumab in combination with chemotherapy (32 months vs 31.2 months). There was no difference in the progression-free survival. However, there was higher response acheived in the cetuximab arm in the expanded RAS population, 68.6% vs 53.6% (p < 0.01).

In a separate analysis from the study of KRAS wild type patients undergoing surgery as a part of treatment strategy, which goal was to determine the characteristics and the long-term outcome of patients enrolled in the trial, Dr Alan Venook of the Division Of Medical Oncology, University of California, San Francisco, USA reported that 130 patients enrolled reached a stage of non evidence of disease (NED) after chemotherapy and surgery. The median OS in these patients was 60 months although many have recurred.

However, at the time of the abstract submission, the study researchers anticipated evaluation of all RAS status and planned an analysis in the subset of patients who underwent surgery to identify possible predictive characteristics but also to determine if there is an explanation for the fact that more patients on cetuximab went to surgery than patients on bevacizumab did. During the session, Dr Venook reported that patients were likelier to reach NED stage on cetuximab but the ultimate outcome seem to be similar.

Independent radiological evaluation of objective response rate, early tumour shrinkage, and depth of response in the FIRE-3 study: Analysis in the final RAS evaluable population

Based on an independent radiological review, FOLFIRI plus cetuximab induced a significantly higher overall response rate (ORR), a greater rate of early tumour shrinkage (ETS), and an increased depth of response (DpR) compared to FOLFIRI plus bevacizumab. These response-related outcomes may in part explain the significant overall survival (OS) advantage of FOLFIRI plus cetuximab treatment observed in the extended RAS wild type study population. The findings were reported by Dr Sebastian Stintzing of the University of Munich, Department of Hematology and Oncology, at ESMO 2014 Congress during the special session on defining the optimal treatment strategy with anti-EGFR or anti-VEGF treatments in first-line RAS wild type metastatic colorectal cancer patients.

Overall survival in the final RAS wild type population in the FIRE-3 study

Overall survival in the final RAS wild type population in the FIRE-3 study. © Sebastian Stintzing

The FIRE-3 study, performed in 150 centres in Germany and Austria, compared first-line therapy with FOLFIRI plus either cetuximab or bevacizumab (1:1) and was amended in October 2008 to include only KRAS wild-type patients. The study was conducted in 592 KRAS exon 2 wild-type metastatic colorectal cancer patients. Extended RAS analysis was carried out in KRAS and NRAS exon 2, 3 and 4 (codons 12, 13, 59, 61, 117 and 146) using pyro-sequencing technique.

An independent radiological review was performed to evaluate tumour response according to RECIST 1.1 and to define ETS and DpR. Reviewers were blinded to patient data.

ETS was defined as a reduction in tumour diameter of more than 20% at first-tumour assessment after baseline (week 6). DpR was defined as the maximal tumour shrinkage observed at the nadir compared with baseline.

Calculations were done for both, the intent to treat (ITT) and the extended RAS wild-type population. At ESMO 2014 Congress, Dr Stintzing presented data from the analysis in final RAS evaluable patients.

Independent evaluation in KRAS exon 2 wild type population showed an ORR of 66.5% in the cetuximab arm and 55.6% in the bevacizumab arm (p = 0.016). In the final RAS wild type population, the ORR was 72% in the cetuximab arm vs 56.1% in the bevacizumab arm (p = 0.003).

OS favoured the cetuximab arm, 33.1 months vs 25.0 months (HR 0.697, p = 0.0059). 

In the final RAS wild type population, progression-free survival (PFS) in patients with ETS in the cetuximab arm was 9.7 months vs 5.8 months in patients with no-ETS. In the bevacizumab arm the PFS in ETS patients was 11.7 months vs 8.3 months in non-ETS patients.

DpR correlated significantly with OS and PFS (p = 0.0003 in KRAS exon 2 wild-type patients and p < 0.0001 in the final RAS wild type population).

Median time to tumour nadir in cetuximab arm was 15.0 weeks and 15.7 weeks in the bevacizumab arm.

Dr Stintzing said that extended RAS testing was possible in > 80% of FIRE-3 ITT population. He concluded that median OS was markedly superior in all-RAS wild type patients receiving first-line therapy with cetuximab. The independent radiology review demonstrated a significantly higher ORR in FOLFIRI plus cetuximab treated patients compared to those receiving FOLFIRI plus bevacizumab. The ETS was significantly more frequent in the cetuximab arm, and it was significantly associated with prolonged survival independent of the treatment arm. Median DpR was significantly greater in the cetuximab arm and correlated with survival. 

Data Interpretation

Dr Andres Cervantes addressed the following two important questions:

Why the overall survival results are discordant in these two trials?

A detailed information on second- and further line therapies is needed. In particular, the proportion of patients randomised to chemotherapy plus bevacizumab who never got cetuximab could be of importance to interpret these results. In a setting where more than 80% of patients get second-line therapy, it could imply that the sequence of treatments would not be relevant for OS. However, if the proportion of patients missing cetuximab in second-line is higher, the sequence of treatments could be very relevant.

What is the relevance of having a higher response rate in cetuximab containing regimens?

Chemotherapy plus bevacizumab or plus cetuximab are two potential options to start treatment in all RAS wild type advanced colorectal cancer patients. But, how the fact that a higher response rate with cetuximab containing regimens can influence decision in selecting initial therapeutic approach? The CALGB 80405 data on response rate has to be further analysed in respect to depth of response and early tumour shrinkage. This could be an important point to communicate with patients when taking therapeutic decisions.

Dr Alberto Sobrero said that the FIRE-3 study shows internal consistency and strengthened plausibility while the CALGB researchers presented early data analysis from a trial performed over 10 years. Therefore, he would wait for complete data before conclusion.

Dr Dirk Arnold said that patients with RAS wild type do (slightly) better with anti-EGFR therapies. However, many open questions remain: why is this so, will further information of treatment characteristics (e.g. duration, second-line treatments,...) really change clinical view on the data.

RAS is important, but taken alone not a super positive predictive biomarker. Any of those combinations used is an excellent option and should be part of the first-line treatment as OS is longer than 31 months. Presumably, the first-line choice is not that important for most patients (maybe except need for early tumour shrinkage and deapth of response).

According to Dr Fortunato Ciardiello, who was one of discussants in the session, metastatic colorectal cancer is a heterogenous disease. A subset of metastatic colorectal cancers is highly dependent on EGFR signalling. KRAS and NRAS testing are the first step to identify those patients that could benefit from anti-EGFR monoclonal antibodies treatment.

All metastatic colorectal cancer patients should have extended RAS testing before first-line treatment choice to offer them all available therapeutic opportunities.

Patients with RAS wild type cancer have two good therapeutic options that should be offered sequentially in first- and second-lines (FOLFOX or FOLFIRI plus either anti-EGFR monoclonal antibodies or bevacizumab).

In his opinion, FOLFIRI or FOLFOX with an anti-EGFR antibody is the preferred first-line choice if tumour shrinkage is a relevant therapeutic goal (high tumour burden, symptomatic disease, potential conversion to surgical resection of liver metastases or, possibly, of the primary tumour).

Tolerabily, side effects and informed discussion with the patient could be important aspects for first-line choices.

For Dr Ciardiello, open issues for clinical and translational research on how to optimise therapeutic management of metastatic colorectal cancer patients with RAS wild type tumours are:

  • Role of maintenance; chemotherapy depotentiation; reintroduction of chemotherapy.
  • Role of appropriate sequencing of bevacizumab and anti-EGFR monoclonal antibodies.
  • Strategies to prevent/overcome acquired resistance to anti-EGFR monoclonal antibodies and to bevacizumab.


The trial lead organizations/sponsors of abstract 501O were the Cancer and Leukemia Group B (CALGB), USA National Cancer Institute, and Southwest Oncology Group (SWOG). FIRE-3 was an investigator sponsored study. The study collaborator was Merck KGaA.