Younger women with early breast cancer may benefit from chemo when clinical risk is high but genomic risk is low
December 13th 2019
By Wayne Kuznar for oncoXchange
Subgroup analysis of the MINDACT study suggests that in women ≤50 years old with early-stage breast cancer, tamoxifen alone may be suboptimal when the women are at high clinical-pathological risk but deemed low risk by a 70-gene signature (MammaPrint).
Distant metastasis-free survival (DMFS) at 5 years in the women ≤50 years at high clinical risk but low genomic risk was 96.1% (95% CI, 91.9-98.2%) in the women who received chemotherapy versus 93.1% (95% CI, 88.6-95.8%) in those who did not, reported Fatima Cardoso, MD, at the 2019 San Antonio Breast Cancer Symposium.
Because the confidence intervals are wide “cautious interpretation is needed,” said Dr. Cardoso, director of the Breast Unit of the Champalimaud Clinical Center in Lisbon, Portugal.
The interaction between age, risk, and effects of chemotherapy in MINDACT are similar to those seen in the TAILORx study, which used a 21-gene expression signature to guide the chemotherapy decision, although the absolute difference in the rates of DMFS and distant metastasis free interval (DMFI) in younger women between the chemotherapy and no-chemotherapy groups are smaller in MINDACT compared with TAILORx, she said.
The MINDACT investigators sought to determine the added value of the genomic signature to the clinical traditional pathological way of assessing risk in women with early breast cancer. All enrolled patients had their risk of breast cancer recurrence assessed clinically/pathologically through use of a modified version of Adjuvant Online. All patients also had their risk assessed by the 70-gene signature. One of three situations could occur: both methods would define that the risk of recurrence was high, and chemotherapy was proposed; both would indicate that the risk was low and no chemotherapy would be proposed; and discordant cases in which one method would indicate high risk and the other low risk, and these patients were randomized to receipt of chemotherapy or no chemotherapy.
“MINDACT is also a de-escalation study and as many de-escalation studies; the primary endpoint is not a comparison, it’s rather a threshold that was decided,” said Dr. Cardoso. “This threshold was for the very relevant group of patients classified as high risk but considered low risk by MammaPrint to follow the genomic risk and not provide chemotherapy.”
Five- year DMFS was chosen as the primary endpoint because the effect of chemotherapy on risk recurrence is mostly observed in the first 5 years of therapy, she said. The primary test for significance in the 5-year DMFS rate between the chemotherapy and no-chemotherapy groups was set at a two-sided 95% confidence interval that exceeded 92%. The observed 5-year DMFS was 94.7%. The 95% CI exceeded 92%, at 92.5-96.2%, so “MINDACT is also a positive study,” she said. “In addition, MINDACT also proved the clinical utility of MammaPrint.”
There was no statistical difference in the 5-year rate of DFMS between the chemotherapy and no chemotherapy arms in the discordant clinical high/genomic low risk group, but a numeric difference of 1.5%.
The intent-to-treat population for this analysis of MINDACT consisted 1,317 patients, all of whom were estrogen receptor-positive/HER2-negative, and all of whom were clinical high and genomic low risk. There were 452 patients ≥50 years old and 865 patients >50 years who were randomized to receive chemotherapy or not. Only 8% of women ≤50 years received ovarian function suppression.
There was a 3% numeric difference in the rates of DMFS at 5 years between the chemotherapy and no chemotherapy groups among the women ≤50 years with discordant risk, and no difference (95.2% vs. 95.4%) in the group of women >50 years.
DMFS included distant recurrence and death by any cause, but almost all of the events were distant recurrence, said Dr. Cardoso.
The DMFI endpoint included only distant recurrence, closer to the endpoint used in TAILORx, and there was a numeric 2.5% difference between chemotherapy and no chemotherapy groups for women ≤50 years in MINDACT, and again no difference in the women >50 years.
By comparison, the absolute difference in distant recurrence at 9 years between the chemotherapy and no chemotherapy groups in the women ≤50 years in TAILORx with a recurrence score of 21 to 25 (not stratified by clinical risk) was 6.5%.
“It is very much possible that this age-dependent effect is due to the chemotherapy-induced ovarian function suppression and not to a direct cytotoxic effect of chemotherapy,” Dr. Cardoso postulated.