A new study suggests that a pregnancy after treatment for early-stage, estrogen receptor (ER)-positive breast cancer does not affect the recurrence rate (J Clin Oncol. 2013;31:73-79). A previous meta-analysis had suggested a lower risk of breast cancer recurrence in patients who experienced a subsequent pregnancy, but this was believed to be possibly due to selection bias, because patients who become pregnant are usually those without recurrence (Eur J Cancer. 2011;47:74-83). In addition, none of the previous studies addressed the effect of pregnancy in patients with ER-positive disease specifically.
Hatem A. Azim Jr, MD, MSc, associate scientific director of the BrEAST Data Center at the Institut Jules Bordet in Brussels, Belgium, and colleagues had the goal of addressing previous limitations of the data with a study to correct for selection bias as much as possible. Their primary objective was comparing disease-free survival (DFS) in patients with ER-positive disease both with and without a subsequent pregnancy. Studying the effects of pregnancy on DFS in patients with ER-negative disease, DFS in the entire cohort, and overall survival were secondary objectives. “I believe that this study addresses the methodological limitations of older studies on this subject and clearly demonstrates the lack of a detrimental effect of pregnancy on prognosis. Hence, I believe that this data are strong enough to avoid discouraging women who completed their adjuvant therapy and want to become subsequently pregnant,” says Dr. Azim.
Investigators conducted a multicenter, retrospective cohort study that matched patients with known ER status who became pregnant any time after the diagnosis of breast cancer at a ratio of 1:3 with those who did not become pregnant based on ER status, lymph node status, adjuvant therapy, age, and year of diagnosis. A retrospective search of the databases of the 5 participating institutions was performed to identify patients aged younger than 50 years with a diagnosis of nonmetastatic breast cancer with a known ER status. Patients who developed breast cancer while pregnant or who experienced recurrence before a subsequent pregnancy were excluded. Each nonpregnant patient in the study had to have a disease-free interval that was at least as long as the interval between the breast cancer diagnosis and conception of the matched patient with a subsequent pregnancy.
Risk Not Increased
Ultimately, 333 patients with a subsequent pregnancy were matched to 874 women without a pregnancy. Patients with a pregnancy were significantly younger (median age, 31 years vs 34 years) and were more likely to have received breast-conserving therapy (50% vs. 42%)..
No difference in DFS was observed between the pregnant and nonpregnant groups in the ER-positive cohort, the ER-negative cohort, or within the entire cohort analyzed together. Overall survival was significantly better in the pregnancy group (hazards ratio, 0.72; P = .03) without respect to ER status.
No differences in DFS were detected between the pregnancy and nonpregnancy groups when restricting analyses to those who completed the pregnancy to term and their matched controls, with the same result noted for those who had a miscarriage or abortion. Furthermore, there was no difference in DFS noted between the 193 patients who became pregnant 2 or more years after their diagnosis of breast cancer and their matched controls. By contrast, the 140 patients who became pregnant fewer than 2 years after diagnosis had a better DFS than their matched controls (hazards ratio, 0.56; P = .02).
The authors noted that the better outcome noted for the patients with early pregnancy may have been due to selection bias. Even though the selection of nonpregnant patients was random, the group matched to the patients with an early pregnancy had a significantly shorter DFS than the control group matched to the patients with a later pregnancy, they wrote.
“This study provides robust data that further supports the standard of advising women who desire a biological child after breast cancer that the pregnancy will not adversely affect their prognosis. However, a person with high-risk breast cancer is still at high risk for recurrence, and that concern may be cause for advising against pregnancy, but not because the pregnancy increases their risk. It is a complex and personal decision,” says Ann Partridge, MD, MPH, director of the Program for Young Women with Breast Cancer at the Dana-Farber Cancer Institute in Boston, Massachusetts.
Although Dr. Azim admits that weaknesses of the study include the fact that human epidermal growth factor receptor-2 (HER2) status was unknown for 80% of pregnant and 82% of nonpregnant patients, that not all patients were treated with hormonal therapy, and that the study was retrospective in nature, he notes that a randomized prospective trial is impossible in this setting, and that the lack of hormonal therapy was evenly distributed between the groups.
“Study strengths are that for the first time this question was addressed in ER-status-known patients only, had controls who were free of relapse at the time of pregnancy of the matched cases (to reduce selection bias or what is referred to as the healthy-mother effect), and had a sample size that was prospectively planned according to ER status, to allow having enough [statistical] power to examine the impact of pregnancy on outcome,” he states.
“ It does not eliminate the possibility that pregnancy has a favorable effect on prognosis as previous data suggest, but it does add to the data that we can use to reassure patients,” says Dr. Partridge.
The authors conclude that pregnancy after a diagnosis of breast cancer, regardless of ER status, does not protect against recurrence, but can be considered safe. The timing of pregnancy does not seem to matter either because no difference was noted if pregnancy occurred within or after 2 years from diagnosis, but this study was not powered to definitively answer that question.
“A main question remaining is how to advise patients with ER-positive breast cancer who are willing to become pregnant before completing the classic 5 years of hormonal therapy,” says Dr. Azim. Currently, the Endocrine Working Group of the Breast International Group (BIG) and the North American Breast Cancer Group (NABCG) are working on a study to address the safety and efficacy of taking a break from tamoxifen.
“The BIG-NABCG Endocrine Working Group is developing a randomized trial, still in the feasibility stage, for women aged 37 years old or younger at diagnosis of ER-positive breast cancer to assess pregnancy success following tamoxifen interruption after 18 months versus 36 months of treatment,” says Dr. Partridge.
Although selection bias cannot be completely eliminated in a study of pregnancy after breast cancer because researchers obviously cannot randomize patients to get pregnant or not, a unique ongoing study is currently being performed to help assess the healthy-mother bias (J Clin Oncol. 2011;29(suppl). Abstract 6025). The study is collecting prospective data from young women with newly diagnosed breast cancer and asking them about their fertility concerns; the impact on treatment decisions, anxiety, and depression; and changes over time.
“This study will quantify the differences between the people who got pregnant and those who did not in terms of what they were thinking at different points in time to help assess the healthy-mother effect. This will help us relate disease outcomes to fertility outcomes, and the potential impact of patient attitudes and thought processes, and foster a deeper understanding of this complicated issue,” says Dr. Partridge.