Impact of the interruption of tamoxifen for pregnancy on the recurrence and survival outcomes among young women with breast cancer

Abstract

Background

Young survivors of breast cancer often consider pregnancy after completing or interrupting endocrine therapy. However, concerns remain regarding the safety of tamoxifen interruption for pregnancy and its impact on survival outcomes. We aimed to evaluate the impact of interrupting and resuming tamoxifen for pregnancy on recurrence and mortality in patients with breast cancer, using real-world data.

Methods

We conducted a retrospective cohort study using data from the Korean National Health Insurance Service National Health Information Database. Among 32,378 women aged 18–45 years diagnosed with invasive breast cancer and who underwent surgery between 2009 and 2014, patients were categorized into groups based on tamoxifen interruption and pregnancy: Group 1 (interruption and resumption, n = 126), Group 2 (interruption without resumption, n = 261), Group 3 (initiation of tamoxifen after childbirth, n = 41), and Control (no interruption nor pregnancy, n =428). The control group was age-matched to the interruption groups in a 1:1 ratio. Clinical outcomes, including recurrence and mortality, were compared among groups.

Results

Over a median follow-up of 8.5 years, Groups 1 and 2 showed significantly lower risks of recurrence than the control group did (hazard ratio [HR] 0.41, 95 % confidence interval [CI]: 0.22–0.76, P = 0.005; HR 0.30, 95 % CI: 0.18–0.50, P < 0.001). In multivariate analysis, Group 2 also showed better survival outcomes (HR 0.18, 95% CI: 0.08–0.41, P < 0.001). Pregnancy outcomes differed across groups: Groups 1 and 2 had higher rates of full-term pregnancies, whereas Group 3 had a significantly higher abortion rate (23.8 %, 23.4 % vs. 56.1 %, respectively).

Conclusion

These findings suggest that temporary tamoxifen interruption for pregnancy may be a viable option for young survivors of breast cancer. Further studies are warranted to clarify the long-term impact of tamoxifen interruption on prognosis.

Highlights

• •Temporary interruption of tamoxifen for pregnancy did not increase recurrence or mortality in young breast cancer survivors in this nationwide cohort study.
• •Women who interrupted tamoxifen and achieved pregnancy had favorable oncologic and reproductive outcomes compared with controls.
• •These findings support individualized oncofertility counseling and safe integration of pregnancy into survivorship care for hormone receptor–positive breast cancer.

1. Background

Breast cancer is the most commonly diagnosed malignancy in women worldwide, with an increasing incidence among young women of reproductive age [1,2]. Advances in early detection and treatment have significantly improved survival rates, leading to a growing population of young survivors of breast cancer who may desire pregnancy after treatment [3,4]. Given the potential gonadotoxicity of chemotherapy and the prolonged duration of endocrine therapy (ET), typically at least 5 years, fertility preservation and post-treatment pregnancy have become major concerns for young survivors [[5], [6], [7], [8]]. Historically, pregnancy after breast cancer was discouraged owing to concerns that hormonal stimulation might increase the risk of recurrence. However, emerging evidence suggests that pregnancy after breast cancer does not compromise survival outcomes, even in patients with the hormone receptor-positive (HR+) subtype [9,10].

To address these concerns among young women who need ET and desire pregnancy, a recent prospective study was conducted to investigate the feasibility and safety of temporarily interrupting ET to allow pregnancy attempts. The POSITIVE trial (Pregnancy Outcome and Safety of Interrupting Therapy for Women with Endocrine-Responsive Breast Cancer) is a landmark study being conducted to investigate the feasibility and oncologic safety of temporary tamoxifen discontinuation for pregnancy in young women with HR + breast cancer [11]. Preliminary findings suggest that carefully planned ET interruption may be a viable option for selected patients, though long-term follow-up is needed to confirm safety and optimal management strategies. At a 71-month median follow-up, there was no significant differences in breast cancer free interval (BCFI) events between women who had a pregnancy and those who did not [12].

Given the growing interest in fertility preservation and pregnancy among young survivors of breast cancer, further research is essential to establish evidence-based guidelines for safely interrupting ET while minimizing recurrence risk. A deeper understanding of the oncologic safety and reproductive outcomes associated with ET interruption in young patients is crucial for informed decision-making by both patients and healthcare providers [13]. Therefore, we aimed to investigate the long-term recurrence and survival outcomes of women with HR + breast cancer after interrupting tamoxifen therapy to allow for pregnancy using real-world data. Specifically, we compared different subgroups of patients who interrupted tamoxifen for pregnancy, those who did not resume therapy post-pregnancy, and those who initiated tamoxifen after pregnancy and childbirth with patients who continued ET without pregnancy. By comparing overall mortality and pregnancy-related outcomes, we sought to provide further insights into the long-term oncologic safety of reproductive decisions in young survivors of breast cancer.

2. Methods

2.1. Study design and population

Data from the National Health Insurance Service National Health Information Database (NHIS-NHID) were used in this study. As a single compulsory health insurance system in Korea, the NHIS-NHID includes information on demographics, healthcare use, vital statistics, and national cancer screening results for the entire Korean population [14].

We included data from women aged 18–45 years who were diagnosed with invasive breast cancer and underwent curative surgery between January 1, 2009, and December 31, 2014. Breast cancer was defined as any breast cancer event according to the International Classification of Diseases, Tenth Edition (ICD-10) codes for invasive breast cancer (C50.0-C50.9) in combination with the rare and intractable disease registration code. Exclusion criteria included: (1) receiving surgery after >1 year from initial diagnosis (n = 5327); (2) having a history of other cancer prior to breast cancer diagnosis (n = 1904); (3) not receiving tamoxifen until the end of the study period (n = 7028); (4) receiving tamoxifen pre-surgery (n = 533); and (5) a duration of tamoxifen prescription <180 days during the 1st year after tamoxifen initiation (n = 452) (Fig. 1).

Fig. 1.

Study schema.

Patients were categorized into the following subgroups (Fig. 2): (1) those who received adjuvant tamoxifen therapy, interrupted it for pregnancy, and resumed tamoxifen after childbirth (Group 1); (2) those who received tamoxifen, interrupted it for pregnancy, but did not resume treatment (Group 2); and (3) those who initiated tamoxifen after pregnancy (Group 3). The control group included patients who continued tamoxifen therapy without interruption and did not experience pregnancy after breast cancer diagnosis through December 2020. After selecting Groups 1–3, exact age-matching was performed at a 1:1 ratio to identify the matched controls. Therefore, all patients included in this study received adjuvant tamoxifen following curative surgery. Tamoxifen use was defined as ≥ 180 days of use within 2 years post-surgery, or ≥180 days of use within 2 years after childbirth. An “interruption” was defined as a gap of >1 year in tamoxifen prescriptions.

Fig. 2.

Consort diagram for the analysis population.

2.2. Assessment of main exposure

The primary exposure was the “interruption” of tamoxifen for pregnancy. Childbirth-related events, including full-term delivery, preterm delivery, and abortion, were extracted with the following ICD-10 codes using the medical record information: full-term (O80, O81, O82, O83, O84), preterm (O601, O603), and abortion (O010, O011, O019, O020, O021, O039, O064, O065, O069, O200).

2.3. Assessments of outcomes

The primary outcomes were recurrence and all-cause mortality. Disease recurrence was defined as either (1) the prescription of recurrence-related chemotherapy or endocrine therapy (such as exemestane, fulvestrant) after the last tamoxifen prescription or (2) initiation of chemotherapy >2 years after the last tamoxifen prescription. Because the NHIS database does not include direct recurrence codes, we used this operational definition which has been applied in prior study using Korean claims data as validated surrogates [15,16].

Secondary outcomes included the pregnancy type and birth outcomes. Ascertainment time for childbirth events and the two main outcomes was up to December 2020.

2.4. Statistical analysis

Descriptive statistics were used to compare baseline characteristics, such as age, type of adjuvant therapy, comorbidities index, and level of income, among subgroups, with the differences compared using the χ2 test. Continuous variables are presented as means with standard deviations (SDs), and categorical variables are presented as numbers with percentages. Cox proportional hazards regression analyses were performed to compare the recurrence and survival outcomes across the three study groups versus the control group. The regression model was adjusted for age at diagnosis, chemotherapy, radiotherapy, ET, targeted therapy, Charlson Comorbidity Index (CCI), and income level. We tested the proportional hazards assumption using Kaplan–Meier curves and identified the parallel lines of the log–log survival distribution function. Hazard ratios (HRs) for recurrence and mortality risk for each group were presented. Additionally, the duration of tamoxifen use between Groups 1 and 2 and pregnancy outcomes across all three groups were presented as means with SDs or as numbers with proportions. A two-sided P < 0.05 was considered statistically significant. All statistical analyses were performed using SAS version 9.4 (SAS Institute).

This study was approved by the Institutional Review Board of Hanyang University Hospital (No. HYUH 2022-05-019). In addition, the National Health Insurance Sharing Service System (NHIS) approved the use of the NHIS database, which was constructed after individual identities were anonymized. The need for informed consent was thus waived owing to the use of secondary data.

3. Results

3.1. Baseline characteristics of study participants

A total of 856 patients were included in the analysis and categorized into three groups based on the pattern of tamoxifen interruption for pregnancy: Groups 1 (n = 126), 2 (n = 261), 3 (n = 41), and a matched control cohort (neither interrupted nor pregnant, n = 428). The median age at diagnosis varied significantly across groups (P < 0.001, Table 1). Notably, the proportion of younger patients (<35 years) was highest in Group 2 (77.4 %) and lowest in Group 3 (41.5 %). CCI scores also differed significantly (P = 0.001), with Group 3 exhibiting a higher proportion of patients with CCI scores of ≥3 than those in the other groups. . ET regimens, including tamoxifen alone or with ovarian suppression, were comparable across groups (P = 0.857).

Table 1.

Demographic and treatment variables of the tamoxifen-interruption subgroups and the matched control.

	Group 1 (n = 126)	Group 2 (n = 261)	Group 3 (n = 41)	Control (n = 428)	P
Age group, years					<0.001
19–34	84 (66.7)	202 (77.4)	17 (41.5)	303 (70.8)
35–39	24 (19.1)	45 (17.2)	14 (34.1)	83 (19.4)
40–45	18 (14.3)	14 (5.4)	10 (24.4)	42 (9.8)
CCIa score					0.002
None	66 (51.4)	93 (35.6)	17 (41.5)	156 (36.5)
1	17 (13.5)	22 (8.4)	6 (14.6)	31 (7.2)
2	23 (18.3)	90 (34.5)	8 (19.5)	133 (31.1)
≥3	20 (15.9)	56 (21.5)	10 (24.4)	108 (25.2)
Income					0.308
Level 1	34 (27.0)	76 (29.1)	6 (14.7)	94 (22.0)
Level 2	23 (18.3)	67 (25.7)	13 (31.7)	109 (25.5)
Level 3	34 (27.0)	42 (16.1)	10 (24.4)	102 (23.8)
Level 4	34 (27.0)	73 (28.0)	11 (26.8)	119 (27.8)
Missing	1 (0.8)	3 (1.2)	1 (2.4)	4 (0.9)
Chemotherapy					0.079
No	38 (30.2)	87 (33.3)	10 (24.4)	105 (24.5)
Yes	88 (69.8)	174 (66.7)	31 (75.6)	323 (75.5)
Radiation therapy					0.011
No	28 (22.2)	37 (14.2)	11 (26.8)	104 (24.3)
Yes	98 (77.8)	224 (85.8)	30 (73.2)	324 (75.7)
Anti-HER2b therapy					0.394
No	106 (84.1)	223 (85.4)	31 (75.6)	366 (85.5)
Yes	20 (15.9)	38 (14.6)	10 (24.4)	62 (14.5)
Endocrine therapy					0.857
TMXc alone	70 (55.6)	150 (57.5)	25 (61.0)	236 (55.1)
TMX + OFSd	56 (44.4)	111 (42.5)	16 (39.0)	192 (44.9)

Note.

Group 1 refers to those who received adjuvant tamoxifen therapy, interrupted it for pregnancy, and resumed tamoxifen after childbirth.

Group 2 refers to those who received adjuvant tamoxifen therapy, interrupted it for pregnancy, but did not resume treatment.

Group 3 refers to those who began adjuvant tamoxifen therapy after pregnancy.

^aCharlson Comorbidity Index.

^bHuman epidermal growth receptor 2.

^cTamoxifen.

^dOvarian function suppression.

3.2. Effects of tamoxifen interruption on survival probability

The median follow-up duration for overall survival in the total population was 11.5 years (interquartile range: 10.0–12.8). Specifically, the median durations for Groups 1–3 were 10.9, 12.1, and 10.6 years, respectively. The recurrence and mortality outcomes were evaluated using adjusted HRs in two models (Table 2). Groups 1 and 2 had a significantly lower recurrence risk (HR 0.41, 95 % confidence interval [CI]: 0.22–0.76, P = 0.005; HR 0.30, 95 % CI: 0.18–0.50, P < 0.001) than the control group did, whereas Group 3 showed no significant difference in recurrence risk (HR 1.56, 95 % CI: 0.82–2.94, P = 0.175). Similarly, in terms of overall mortality, Groups 2 (HR0.18, 95 % CI: 0.08–0.41, P < 0.001) demonstrated significantly lower mortality risk than the control group did, while Group 3 showed no significant difference (HR 1.02, 95 % CI: 0.40–2.57, P = 0.968). Kaplan–Meier survival analysis (Fig. 3) showed a statistically significant difference in recurrence-free and overall survival across groups (P < 0.001).

Table 2.

Hazard ratio estimates of subgroups for disease recurrence and overall mortality.

	Model 1		Model 2
	HR a (95 % CI b)	P	HR (95 b CI)	P
Recurrence
Control	Ref.		Ref.
Group 1	0.41 (0.22–0.76)	0.005	0.41 (0.22–0.76)	0.005
Group 2	0.25 (0.15–0.42)	<0.001	0.30 (0.18–0.50)	<0.001
Group 3	1.52 (0.81–2.86)	0.194	1.56 (0.82–2.94)	0.175
Death
Control	Ref.		Ref.
Group 1	0.44 (0.21–0.93)	0.030	0.55 (0.26–1.16)	0.116
Group 2	0.14 (0.06–0.33)	<0.001	0.18 (0.08–0.41)	<0.001
Group 3	0.93 (0.37–2.33)	0.877	1.02 (0.40–2.57)	0.968

Note.

Model 1: adjusted for age at diagnosis.

Model 2: adjusted for age at diagnosis, chemotherapy, radiotherapy, endocrine therapy, target therapy, Charlson Comorbidity Index, and income.

^aHazard ra^aio.

^bConfidenc^b interval.

Fig. 3.

Survival curves of breast cancer recurrence and overall mortality.

Blue line: Group 1, Green line: Group 2, Red line: Group 3, Black line: Controls.

3.3. Duration of tamoxifen and pregnancy outcomes

Tamoxifen duration pre-pregnancy differed significantly between Groups 1 and 2 (Supplementary Table 1). Most patients in Group 2 (72.0 %) received tamoxifen for >2 years pre-pregnancy, compared with those in Group 1 (62.7 %). However, subgroup analyses based on tamoxifen duration pre-pregnancy indicated no significant impact on recurrence in Group 1 (P = 0.818) or 2 (P = 0.097), and similarly, no significant difference in overall mortality (P = 0.520 and P = 0.136, respectively) on Kaplan–Meier analysis (Supplementary Fig. 1).

Pregnancy outcomes differed across subgroups (Supplementary Table 2). Groups 1 and 2 had higher rates of full-term pregnancies (75.4 % and 75.9 %, respectively) than Group 3 did (39.0 %). Additionally, the abortion rate was highest in Group 3 (56.1 %) compared with those in Groups 1 (23.8 %) and 2 (23.4 %). The mean duration of interruption for tamoxifen before pregnancy was 18.7 months and 32.9 months in Group 1 and Group 2, respectively.

The duration from cancer diagnosis to pregnancy was the longest in Group 2 (70.8 months) and shortest in Group 3 (11.9 months).

4. Discussion

The findings of this nationwide, real-world cohort study add to the growing body of evidence supporting the oncologic safety of temporary tamoxifen interruption for pregnancy in young survivors of breast cancer with HR + disease. In particular, our results suggest that patients who interrupted and subsequently resumed tamoxifen therapy (Group 1) had significantly lower recurrence and mortality risks than did those who never became pregnant or interrupted therapy (control group). This observation is consistent with the interim results of the POSITIVE trial, which reported that planned ET interruption for pregnancy did not significantly increase short-term recurrence rates.

Tamoxifen remains a cornerstone of ET for HR + breast cancer, with adherence to a 5-to-10-year regimen associated with improved survival [17]. However, for young women desiring pregnancy, the necessity of a prolonged treatment course presents a dilemma. Tamoxifen is known to be teratogenic and may pose a risk to the developing fetus; therefore, it is recommended that tamoxifen be discontinued at least 3 months before attempting conception to allow drug clearance and reduce potential fetal harm [18,19]. Our results showed that patients in Group 1, who resumed tamoxifen therapy post-pregnancy, exhibited lower recurrence and mortality risks than did those in the control group. Interestingly, patients in Group 2, who interrupted but did not resume tamoxifen, also exhibited reduced recurrence and mortality risks, which may reflect differences in favorable tumor biology or selection bias in those who opted not to resume therapy. Moreover, the longer overall duration of tamoxifen use in Group 2 than that in Group 1 might also have influenced the survival outcomes. While pregnancy itself does not appear to negatively impact prognosis even in HR + breast cancer, as consistently reported in meta-analyses and population-based studies, the individualized strategy for oncofertility services is essential for young women seeking pregnancy during ET [[20], [21], [22]]. Oncofertility counseling, ideally before cancer treatment initiation, enables women to make informed decisions about fertility preservation options and ET planning [8,23,24].

The duration of tamoxifen therapy before pregnancy varied among study participants. Most patients in Group 2 received >2 years of treatment. Prior studies have suggested a minimum of 18 months of tamoxifen therapy before interruption to optimize recurrence outcomes; however, our subgroup analysis did not show a statistically significant difference in recurrence or mortality risks between those who interrupted treatment before or after 2 years. This is consistent with data from the POSITIVE trial, where over 90 % of participants completed at least 18 months of prior ET. However, owing to borderline statistical significance in Group 2 (P = 0.058), further investigation with longer follow-up is necessary.

Pregnancy outcomes also differed across subgroups. Groups 1 and 2 had higher rates of full-term pregnancies and lower abortion rates than did Group 3, where tamoxifen was initiated only after pregnancy. Additionally, the time to conception was significantly longer in Group 2, potentially reflecting an underlying impact of prolonged ET on fertility. The higher abortion rate in Group 3 may reflect differences in baseline fertility status or older age at conception. These findings are consistent with those of a previous observational study by Labrosse et al., who also reported favorable maternal and neonatal outcomes in survivors of breast cancer [25]. The POSITIVE trial also reported that approximately 74 % of patients who attempted conception successfully achieved pregnancy, with no increase in adverse perinatal outcomes such as preterm birth, low birth weight, or congenital anomalies.

This study has several strengths. The use of national health insurance data allowed us to evaluate real-world outcomes in a diverse population. The inclusion of a matched control group further strengthens the validity of our findings. However, some limitations should be noted. The retrospective nature and residual confounding factors, such as missing data on tumor grade, hormone receptor status, proliferation index, or patient preference, introduce potential bias. These missing data may have influenced both treatment selection and clinical outcomes. Specifically, the absence of biomarker information raises the possibility of treatment selection bias, which may partially explain the higher chemotherapy use observed in Group 3. The higher abortion rate in Group 3 may also reflect differences in underlying tumor biology, higher biological age, or physician recommendations based on perceived risk factors that could not be fully captured within the claims database. These limitations should be considered when interpreting the comparison of oncologic and reproductive outcomes across the three groups. Selection bias may have influenced the decision to interrupt or resume therapy, and variations in ET adherence may have impacted clinical outcomes. Additionally, the study's observational design limits our ability to establish causal relationships between tamoxifen interruption and oncologic outcomes. Further studies with extended follow-up are warranted to explore the long-term impact of tamoxifen duration and reproductive decisions on breast cancer prognosis.

5. Conclusion

In conclusion, our findings support the growing evidence that temporary tamoxifen interruption for pregnancy is a viable option for young survivors of breast cancer. Our findings, along with those from the POSITIVE trial, reinforce that ET interruption for conception does not appear to significantly increase recurrence or mortality risks, particularly when treatment is resumed post-pregnancy. These results contribute to the evolving discussion on balancing fertility and oncologic safety in young women with HR + breast cancer, underscoring the need for individualized patient counseling and shared decision-making.

CRediT authorship contribution statement

Chihwan David Cha: Writing – review & editing, Writing – original draft, Visualization. Somin Jeon: Methodology, Formal analysis. Jihyun Keum: Validation, Conceptualization. Shin Jeong Pak: Software, Resources. Boyoung Park: Supervision, Project administration, Formal analysis. Min Sung Chung: Writing – review & editing, Writing – original draft, Conceptualization.

Ethics approval and consent to participate

This study was approved by the Institutional Review Board of Hanyang University Hospital (No. HYUH 2022-05-019) and followed the principles of the Declaration of Helsinki, which protects human subjects in medical research.

Role of the funder/sponsor

The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Data sharing statement

The datasets generated during and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Funding/support

This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF), funded by the Ministry of Education (RS-2023-00241942), and by the National Research Foundation of Korea (RS-2024-00462658 and RS-2025-00558296). This research was further supported by a Korea Basic Science Institute (National Research Facilities and Equipment Center) grant funded by the Ministry of Education. (2023R1A6C101A009).”

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article.