NEW YORK (GenomeWeb) – A new study has confirmed that a biomarker suspected from earlier studies to be associated with response to the EGFR inhibitor cetuximab (Erbitux) does indeed appear to predict which KRAS wild-type colorectal cancer patients are more or less likely to respond to the drug.
In the study, published today in Clinical Cancer Research, investigators analyzed samples retrospectively from a phase III clinical trial, confirming that patients with normal KRAS who were also homozygous for a specific polymorphism of the FCGR gene had a statistically significant increase in survival with added cetuximab treatment compared to those who not.
The link between FCGR2A and cetuximab response is not novel, and has been suggested by numerous previous studies and also contradicted by others.
The new results however, are the first to definitely demonstrate a link between FCGR2A status and benefit from the drug in samples from a randomized controlled trial, Geoffrey Liu, molecular genomics chair at the Princess Margaret Cancer Centre in Toronto and lead author of the study, told GenomeWeb in an email.
According to Liu, if he and his colleagues' findings can be further validated in prospective trials, it would imply that clinicians should test their KRAS wild-type CRC patients for this polymorphism, and consider avoiding cetuximab in favor of other therapies in the 25 percent who carry the FCGR2A R/R genotype, thereby avoiding potential harmful toxicities of a drug unlikely to benefit them.
The first positive trials of cetuximab revealed that adding the drug to standard of care increased the median overall survival of metastatic colorectal cancer patients, but only in those with KRAS-wildtype disease.
However, only half of patients with normal KRAS actually derive a significant benefit from cetuximab, leading researchers to search for biomarkers that can better predict, or subdivide within the KRAS wild-type population, who is or is not likely to benefit from the drug.
In their new study, Liu and his colleagues in the Canadian Cancer Trials Group and the Australasian Gastro-Intestinal Trials Group analyzed archived tumor and normal tissue samples from patients enrolled in the phase III NCIC-CTG CO.17 clinical trial, which evaluated the addition of cetuximab to best supportive care for patients with metastatic CRC.
Interestingly the study involved a triplicate testing strategy in which Liu, his Australian coauthor Alexander Dobrovic, and commercial testing lab Transgenomic all tested the same samples for their FCGR genotype.
In an email to GenomeWeb Liu explained that he, Dobrovic, and Transgenomic all independently proposed similar studies of FCGR2A in the CO.17 trial cohort, and were asked by the trial director to work together.
"Transgenomic was interested in testing their platform," Liu said, while Dobrovic had been working on a technique to accurately genotype the FCGR3A polymorphism, because "he suspected that others in the field may have been genotyping a sequence in the pseudogene in addition to the actual polymorphism."
Liu's laboratory became the reference laboratory, using a classical Sanger method that used primers to avoid a potential overlap with the pseudogene. "The triplicate comparison was the result," Liu said.
Overall, the investigators were able to analyze the FCGR2A status of 286 patients from the overall NCIC-CTG CO.17 trial, 148 of whom had KRAS-wildtype tumors.
Median overall survival among FCGR2A H/H and KRAS-normal patients was 5.5 months longer for those who received cetuximab in addition to best supportive care versus standard of care alone (10.35 months versus 4.83 months).
Among patients with KRAS-wildtype disease who had one copy of the variant (FCGR2A H/R) or no copies (FCGR2A R/R), the increases in median overall survival were not statistically significant, the authors reported.
Moving forward, Liu said that he and his colleagues have received permission to perform an analysis of both RAS mutations and FCGR polymorphisms together in CO.20, a second Phase III randomized clinical trial in which patients were randomized to receive cetuximab either alone or in combination second drug, brivanib.
The overall results of that trial were negative, Liu said, but it will be a good source for further validation of the team's findings in CO.17.
Ultimately, the authors wrote, a prospective validation is necessary to confirm whether this marker should be integrated into clinicians' therapeutic decision-making around cetuximab.
As genomically guided basket or umbrella trials proliferate, Liu said he and his team anticipate that there will be an opportunity to use FCGR2A to identify and recruit patients to treatment with cetuximab, with the results providing more solid evidence of the clinical utility of this strategy.
"We need to complete assessment in CO.20 first … [but afterward] we may start the process of working with one or more of these large molecularly informed umbrella trials to add a FCGR/RAS-cetuximab arm," Liu said.
According to Liu, test platforms that can be used to genotype FCGR are already available, used mainly in the context of other types of targets like genetic hypercoagulation markers.
Liu said that Transgenomic has previously developed a test for a different FCGR polymorphism, FCGR3A, in the context of the drug rituximab in patients with follicular non-Hodgkin’s lymphoma. However, in the group's analysis of CO.17, only FCGR2A was found to be associated with cetuximab outcome.
Data on FCGR2A, or other FCGR polymorphisms, has been mixed in the context of other monoclonal antibody-based drugs and in other tumor types. For example, a study published earlier this year — in which researchers genotyped 321 follicular NHL patients from a single-agent rituximab trial — found that hat neither FCGR3A nor FCGR2A polymorphisms predicted response to FCGR 3A and 2A polymorphisms.