NEW YORK (GenomeWeb) — Mayo Clinic researchers have demonstrated in a new study that measuring CYP2D6 status from breast tumor tissue can be inaccurate due to loss of heterozygosity in the CYP2D6 gene region. The authors argue that this may be the reason why some large studies have failed to show a connection between CYP2D6 genotype and response to tamoxifen.
The group published the study online this month in the Journal of the National Cancer Institute.
Biological evidence indicates that CYP2D6 variants influence the metabolism of tamoxifen, and some research by the Mayo group and others has shown that women carrying these variants benefit less from the drug and have higher rates of recurrence. However, based on two large clinical trials that found no link between CYP2D6 genotype and tamoxifen effectiveness, clinical recommendations have discouraged testing to guide treatment.
This discordance and resulting clinical controversy over the value of CYP2D6 genotyping spurred the Mayo team, led by Matthew Goetz, to further investigate a hypothesis that the use of tumor tissue for CYP2D6 testing in these large trials may have resulted in some patients being mischaracterized as homozygous when they were in fact heterozygous, leading to a false-negative result.
He and his colleagues took two routes to test this theory. First, they analyzed retrospective data from two breast cancer genomics datasets — more than 600 samples from the TCGA database and 360 samples from Foundation Medicine.
In the TCGA samples, the group observed loss of heterozygosity in about 41 percent of the ER-positive tumors and 35 percent of the ER-negative tumors, but only 15 percent of HER2-positive cases. In the Foundation Medicine group, which included mostly ER-positive cancers, the rate of LOH was similar, about 41 percent in these tumors.
To further explore the prevalence and potential impact of CYP2D6 LOH, the researchers then set out to directly compare CYP2D6 genotypes in tumor and in normal tissue from the same patient. For this study, they evaluated previously genotyped tumors from patients enrolled in the NCCTG 89-30-52 tamoxifen trial, and compared the results to the CYP2D6 genotypes derived from the same subjects' normal, nonmalignant FFPE sections, as well as buccal samples.
In these patients, the group showed that there was a perfect agreement between CYP2D6 genotypes derived from non-cancerous tissue and those derived from cheek swabs. In contrast, 20 percent of patients showed CYP2D6 alleles in their tumor tissue that were discordant from their normal buccal and tissue genotype.
Overall, the study results suggest that the use of tumor DNA in establishing a patient's CYP2D6 status can lead to a mischaracterization of her true germline CYP2D6 genotype in a significant proportion of cases, Goetz said.
Goetz told GenomeWeb that this provides at least one piece of evidence for why two central studies, of the Breast International Group 1-98 and ATAC clinical trials, found a lack of a connection between CYP2D6 genotype and tamoxifen response.
At the same time, Goetz said, the study does not prove that tamoxifen response is mediated by CYP2D6 variants. "We don't know this is the only problem. This issue of genotyping error could be fixed and there could still be no evidence for 2D6."
However, James Ingle, another co-author of the study, said in a statement that it is "sobering to consider that a decade of work by the oncology community has not produced a clear resolution of the value of CYP2D6 genotype for predicting tamoxifen efficacy."
According to Ingle the potential impact of this delay is quite large. "Over half a million women worldwide would have been expected to be poor metabolizers [during this delay]," he said. If the group's results are reflective of the true connection between CYP2D6 and tamoxifen benefit, it implies that "these women would have been given less-than-optimal adjuvant therapy had they received tamoxifen."
Goetz said that though the group's study results do not end this controversy, they are strong enough to suggest that "guidelines that recommend or do not recommend 2D6 testing should not be based on studies that used tumor tissue to derive genotype."
"guidelines that recommend or do not recommend 2D6 testing should not be based on studies that used tumor tissue to derive genotype."
Mayo clinic is one of a few centers that have continued to test CYP2D6 in breast cancer patients despite a lack of recommendations in clinical guidelines. Goetz said the Mayo group's rationale for testing is based on results they have seen in their own studies of prospective tamoxifen trials, which have conflicted with what was seen in analysis of BIG 1-98 and ATAC.
For example, in a study published in 2009 in JAMA, Goetz and his colleagues analyzed a large cohort of German and US patients and found that "the presence of 2 functional CYP2D6 alleles was associated with better clinical outcomes and the presence of nonfunctional or reduced-function alleles with worse outcomes."
Goetz said the Mayo group has used blood when available or the non-malignant tissue adjacent to tumor tissue as a way to avoid potential confounding of results due to CYP2D6 loss of heterozygosity.
Further analyses will be required to confirm both the influence of tumor LOH on the accuracy of CYP2D6 genotyping, and to resolve the discordant findings different studies have had in regard to the link between genotype and response to tamoxifen.
Luckily, Goetz said, ongoing trials in new cohorts are expected to be complete and published within the next few years. Careful analysis of genotype in these patients, with attention to the potential of tumor tissue to affect the analytical validity of CYP2D6 testing, will provide important answers.
Goetz and his colleagues are also planning a study using data from the International Breast Cancer Study Group's SOFT clinical trial, which has been collecting blood samples in addition to tumor tissue by design.
"We were not able to get blood samples from everyone," Goetz said, "but the hope is that we'll have enough to answer this question very soon."