Researchers from the ATAC and BIG 1-98 trials reported that CYP2D6 testing is not ready for prime time for gauging response to tamoxifen. While of the leading researchers in the field of CYP2D6 genotyping posed some key questions that need to be answered before CYP2D6 can be officially ruled out, or embraced, in clinical practice, these study results strongly suggest that CYPD2D6 testing should not be part of the standard of care, said Claudine Isaacs, MD.
SAN ANTONIO-What is known about the cytochrome CYP2D6† is that it’s intricately involved in the metabolism of tamoxifen and the latter’s bioactive form, endoxifen. Similarly, UGT2B7† is necessary for the catabolism of endoxifen to an inactive form, which may also play a role in tamoxifen metabolism. But restrospective analyses from two, large randomized trials strongly suggest that CYP2D6 levels do not serve as a hallmark of the benefits of adjuvant therapy. Researchers from these two adjuvant treatment trials, ATAC and BIG 1-98, sought to shed some light on this issue by testing if variations in CYP2D6 levels were associated with tamoxifen’s efficacy in hormone-receptor-positive women.
Dr. Isaac’s perspective:
“If CYP2D6 scoring was important for outcomes, we would have expected that those patients with a low score, randomized to tamoxifen, would have had a poorer outcome, whereas there would have been no difference in outcome in those allocated to anastrozole.”
James Rae, PhD, and Brian Leyland-Jones, MD, PhD, researchers from the ATAC and BIG 1-98 trials, respectively, reported that CYP2D6 testing is not ready for prime time in this patient population for gauging response to tamoxifen. However, one of the leading researchers in the field of CYP2D6 genotyping posed some key questions that need to be answered before CYP2D6 can be officially ruled out, or embraced, in clinical practice.
“There has been a lot of interest in the role of CYP2D6 because, previously, we’ve shown that CYP2D6 is genetically polymorphic, with about 10% [of Caucasian women] who do not make the enzyme,” said Dr. Rae, who is from the department of internal medicine and pharmacology at the University of Michigan in Ann Arbor. In addition, CYP2D6 can be inhibited by concomitant medicines, particularly selective serotonin reuptake inhibitors (SSRIs), he said.
“[Patients with] low CYP2D6 activity have low levels of endoxifen and may not respond as well to tamoxifen. Does the CYP2D6 genotype predict tamoxifen response? We thought it would be important to test this hypothesis in the ATAC trial,” he said.
ATAC† was a prospective, randomized, double-blind, placebo-controlled trial designed to compare the adjuvant use of anastrozole (Arimidex) against tamoxifen for five years in 9,366 women.
CYP2D6 genotype data for the seven most common alleles, including CYP2D6*4, were used to assign each patient a CYP2D6 score based on predicted allele activities from 0 (no activity) to 2 (high activity). Comprehensive CYP2D6 genotype data were obtained on 588 patients randomized to receive tamoxifen and 615 who received anastrozole.
The CYP2D6 scoring system separates patients into poor endoxifen metabolizers, intermediate metabolizers, or extensive metabolizers. Dr. Rae explained. “What we were trying to look at was an activity-dose response relationship.”
The study’s primary endpoint was any recurrence or local recurrence and contralateral cancer. After a median of 10 years of follow-up, there were 115 recurrences in the tamoxifen cohort, but the ATAC trialists found no associations between any of the CYP2D6 scores and rates of recurrence in this patient population. For patients with a score of 0 (poor metabolizers), the hazard ratio [HR] was 1.06 (P = .873). For those with a score of 2 (extensive metabolizers), the HR came in at 1 (no P-value). For intermediate metabolizers, the HR was 0.92 (P = .734).
In the anastrozole cohort, there were 92 recurrences, and, again, there was no evidence of an association between CYP2D6 score and rates of recurrence. For poor metabolizers, the HR was 0.55 (P = .243) while for intermediate metabolizers, the HR was 0.70 (P = .184), and for the extensive metabolizers, the HR was 1 (no P-value) (SABCS 2010 abstract S1-7).
Likewise, there was no detectable association in either the tamoxifen or anastrozole group with UGT2B7*2 genotype alone or in combination with CYP2D6 score. Finally, testing with the CYP2D6*4 variant did not predict recurrence in the tamoxifen or the anastrozole arm (overall P-value for trend = .688 for tamoxifen and .22 for anastrozole).
Dr. Rae said his group also analyzed if SSRI use had an effect on CYP2D6 levels. “We controlled for potent CYP2D6 inhibitors, and we did not see an influence on outcomes,” he said.
Dr. Rae concluded that, based on this study in the ATAC population, there is insufficient evidence to recommend CYP2D6 genotyping. In addition, there is no evidence that avoiding CYP2D6 inhibitors is necessary in tamoxifen-treated patients.
For the CYP2D6 study based on the BIG 1-98† trial, “our research questions were: For postmenopausal, endocrine-responsive, early breast cancer patients treated with tamoxifen, are CYP2D6 phenotypes or reduced enzyme activity associated with reduced disease control? Second, are they associated with reduced hot flushes?” Dr. Leyland-Jones said.
BIG 1-98 was a double-blind, randomized trial comparing five years of letrozole (Femara), tamoxifen, and sequences of letrozole and tamoxifen as adjuvant endocrine therapy for postmenopausal women with endocrine-responsive breast cancer. BIG 1-98 enrolled 8,010 patients between 1998 and 2003.
Of the total patient population, 4,628 were successfully genotyped for CYP2D6, Dr. Leyland-Jones reported. “They received either tamoxifen or letrozole monotherapy, 98% were Caucasian, 57% had node-negative disease, all had estrogen-receptor-positive tumors, 77% did not receive neoadjuvant or adjuvant chemotherapy.”
As with the previous study, a scoring system was used that divided patients into poor metabolizers, intermediate metabolizers, or extensive metabolizers. The primary endpoint was breast-cancer-free interval.
Again, Dr. Leyland-Jones and colleagues found no association between CYP2D6 phenotype and breast-cancer-free interval. The poor metabolizers and the intermediate group did not have worse disease control than the extensive metabolizers, he reported.
In poor metabolizers who received tamoxifen and no chemotherapy, the unadjusted HR was 0.71. In the intermediate metabolizers, the HR came in at 1.22, and in the extensive metabolizers, it was 1 (P = .032). In the adjusted (for nodal involvement, tumor size, local therapy, race, and other factors) analysis, the HR in poor metabolizers was 0.58. The HR was 0.95 in intermediate metabolizers and 1 in extensive metabolizers (P = .055).
With regard to hot flushes, in women who received tamoxifen and no chemotherapy, the percentage of patients who experienced hot flushes was similar: 48% for poor metabolizers, 49% for intermediate metabolizers, and 42% for extensive metabolizers (SABCS 2010 abstract S1-8).
“Our results showed the lowest hot flush incidence in extensive metabolizers, and do not support the hypothesis that poor [endoxifen] metabolism is associated with reduced hot flushes,” Dr. Leyland-Jones said. “Consistent with COBRA†, the intermediate group had significantly more hot flushes compared with the extensive metabolizers.”
He concluded that CYP2D6 testing is not justified to determine whether tamoxifen should be given and that the incidence of hot flushes should not be used to determine tamoxifen efficacy. Dr. Leyland-Jones added that the BIG 1-98 group continues to look at the pharmacogenetics in other genetic lines in pre- and postmenopausal women.
Vantage Point: Need for scoring standards
Since 2003, 178 publications have been listed in PubMed on CYP2D6 genotyping and endocrine therapy. Of those, 14 have reported positive associations while 15 have reported no association, explained Matthew P. Goetz, MD, from the Mayo Clinic in Rochester, Minn. The data presented from ATAC and BIG 1-98 represent “data derived from large, prospective adjuvant hormonal studies that really define the standard of care for early-stage breast cancer,” he said.
In addition, the two studies managed to circumvent some of the limitations associated with CYP2D6 testing based on tumor-derived DNA. “The number of CYP2D6 alleles was adequate and was about as good as you can get from tumor-derived DNA,” Dr. Goetz said. Also, he said, the researchers “controlled for factors such as estrogen-receptor expression and chemotherapy administration, factors that have been problematic with other studies.”
One major issue with CYP2D6 phenotyping is the lack of a standardized scoring system, particularly when it comes to deciding which patients are intermediate metabolizers vs extensive metabolizers, he said.
Other confounding issues include the retrospective nature of these studies given that CYP2D6 only partially explains the wide range of variability in endoxifen pharmacokinetics. Instead, prospective studies need to be done to study the role of tamoxifen and its metabolites, which would answer key questions about the critical threshold for endoxifen concentrations and tamoxifen efficacy as well as the interaction between tumor subtypes and the pharmacokinetics of tamoxifen, Dr. Goetz said.
Dr. Goetz summarized what he considered the take-home message from these studies: “Should women be informed of the controversy regarding the pharmacogenetics of tamoxifen? Yes. Should caregivers exert caution and avoid potent endoxifen inhibitors in women treated with tamoxifen? Yes. Should caregivers recommend routine CYP2D6 testing in postmenopausal women, to preferentially select [them for] either tamoxifen or aromatase inhibitors? Absolutely not. The data are still too controversial.”
However, there may be some patients in whom testing could be considered, such as postmenopausal women with high-risk disease, Dr. Goetz said.
†CYP2D6 =cytochrome P450 2D6
†UGT2B7 = UDP-glucuronosyltransferase-2B7
†ATAC = Arimidex, Tamoxifen, Alone or in Combination †BIG = Breast International Group
†COBRA = Consortium on Breast Cancer Pharmacogenomics