Cyclophosphamide-Free Adjuvant Chemotherapy for Ovarian Protection in Young Women With Breast Cancer

A Randomized Phase 3 Trial

Ke-Da Yu, MD, PhD; Jing-Yu Ge, MD; Xi-Yu Liu, MD; Miao Mo, MD; Min He, MD; Zhi-Ming Shao, MD


J Natl Cancer Inst. 2021;113(10):1352-1359. 

In This Article


The Substitution of Paclitaxel for Cyclophosphamide on Survival Outcomes and Resumption of Menses in Young Women with ER-Positive Breast Cancer trial was designed to determine whether EP-wP is superior to standard EC-wP and whether the elimination of cyclophosphamide would result in a higher menstrual resumption rate at 12 months after chemotherapy in young patients with ER-positive and HER2-negative breast cancer. To the best of our knowledge, this is the first randomized trial to compare 2 adjuvant chemotherapy regimens specifically in young patients with ER-positive breast cancer.

Recent guidelines have recommended LHRHa given concurrently with chemotherapy as a strategy to reduce the risk of premature menopause.[4,5] For the first time, to our knowledge, we demonstrate that a cyclophosphamide-free regimen could increase the rate of menses recovery without comprising survival. Our results are consistent with the findings from the NSABPB-30 trial, in which 76.4% of patients in the doxorubicin-docetaxel group experienced amenorrhea for at least 6 months compared with 89.5% of patients in the doxorubicin-docetaxel-cyclophosphamide group.[22] It seems that the menstrual resumption rate in our study is somewhat lower than the data reported elsewhere. In the ZORO trial, the reappearance of menstruation at 6 months was 70% for the LHRHa group and 57% for the chemotherapy-only group.[11] Another study indicates that the rate of 12-month menstrual resumption in young breast cancer women was up to 85%.[23] A potential explanation is the common use of tamoxifen in our enrolled ER-positive patients. It has been reported that menstrual pattern changes, including amenorrhea, were more frequent in patients on tamoxifen treatment.[24]

In our trial, the primary endpoint was the menstrual resumption rate, and 2 other trials, POEMS and PROMISE-GIM6,[25,26] chose the ovarian failure rate and the early menopause incidence, respectively. Nevertheless, regular menses resumption is a clinically relevant outcome and is a requirement for subsequent fertility. Our trial's posthoc exploratory analysis indicated that pregnancy occurred in more women in the EP-wP group than in the EC-wP group.

In the era of molecular oncology, a better approach is to analyze patients with high-risk ER-positive breast cancer and subsequently perform translational research using tissue samples to provide more precise genetic or genomic information. In 2011 when the trial was initiated, gene expression arrays were not popular for subtype classification and optimal treatment determination. The results of a prospective trial of 21 genes(TAILORx) were published in 2015,[27] and those of 70 genes (MINDACT) were published in 2016.[28] We tried our best to restrict our study to high-risk patients. According to the inclusion criteria of our trial, all the patients enrolled were defined as clinically high-risk patients based on the modified version of Adjuvant! Online in the MINDACT design.[28] These young and clinically high-risk patients might not avoid adjuvant chemotherapy even when with the genomic low-risk disease according to the updated analysis of the MINDACT trial.[29]

Currently, young high-risk patients with ER-positive breast cancer are likely to receive 5-year ovarian suppression. A cyclophosphamide-free regimen, which is associated with ovarian protection, is not contradictory to a subsequent ovarian suppression strategy during endocrine therapy. In routine clinical practice, patients who have yet to complete their 5-year ovarian suppression period are not recommended candidates for pregnancy. However, it is noteworthy that our findings can be extrapolated to patients with other subtypes of breast cancer, such as triple negative or HER2 enriched, because the effect of paclitaxel and cyclophosphamide on menstrual resumption is not cancer subtype specific.

Another critical and controversial issue raised was when and how to conceive after breast cancer. The standard duration of endocrine therapy is at least 5 years, and all the patients who had a pregnancy within 48 months interrupted their endocrine therapy. According to the St. Gallen Breast Cancer Consensus, for women contemplating pregnancy after breast cancer, the optimal timing and impact of interrupting endocrine therapy is unknown. However, the panel recommended a minimum of 18 months following diagnosis before anticipated pregnancy.[30] Several clinical studies have investigated the optimal timing for pregnancy after breast cancer. In a matched case-control study, no DFS difference was observed between the pregnant and nonpregnant cohorts in ER-positive patients. In that study, approximately 60% of cases were pregnant after 2 years from surgery, though time to pregnancy had no statistically significant impact on survival.[31,32] Another study showed a trend toward increased risk of recurrence in patients who conceived within 24 months after diagnosis.[33] In our clinical practice, we asked women contemplating pregnancy to conduct restaging scans before attempting conception and suggested a minimum 24-month interval between diagnosis and anticipated pregnancy. The decision of pregnancy should be very discreet. Subsequent adequate follow-up among breast cancer survivors is crucial,[34] and women who temporarily interrupt endocrine therapy due to pregnancy should be reminded to resume endocrine therapy following attempted or successful pregnancy.

Limitations of the trial included that we did not consider the rates of pregnancy or successful delivery as secondary endpoints due to the high probability of confounding factors. But we reported the posthoc exploratory analysis of pregnancy outcomes. Moreover, early intervention with LHRHa makes it impossible to observe menses resumption. We thus performed sensitivity analyses as complementary results to augment the main findings. We also did not report on the pattern of menstruation, and irregular menses might also lead to infertility. The fertility function should be further evaluated by measuring the levels of other markers, such as anti-Müllerian hormone and inhibin. Anti-Müllerian hormone may help to separate patients with amenorrhea due to tamoxifen use from those with loss of ovarian reserve. In addition, contrasting pathological characteristics of breast cancer between Asian and American women suggest racial differences in biology,[35] and whether our findings can be extrapolated to other races is unknown. Furthermore, the recurrence of ER-positive breast cancer occurs at a steady rate during 5 to 20 years, and an adequate follow-up is needed for more reliable survival outcomes.[36] Finally, in the era of genetic testing, information on gene predisposition for early-onset patients might provide new insights. Although young women with triple-negative breast cancer have higher mutation rates in BRCA1/2, RAD51C, and RAD51D genes,[8,37] the germline mutation spectrum in young women with ER-positive or HER2-negative breast cancer as well as its effect on chemotherapy sensitivity are unknown and warrant further study.

In conclusion, among young women with ER-positive breast cancer, the cyclophosphamide-free regimen might be associated with a statistically significantly higher probability of menses resumption.