Fertility concerns and treatment decision‐making among national sample of young women with breast cancer

Abstract Background Diagnosis of breast cancer in young women has been shown to affect their decision‐making with regard to fertility and family planning. Limited data are available from populations across the U.S. regarding this issue; thus, we sought to describe fertility concerns and efforts to preserve fertility in a national clinical trial population of young breast cancer patients. Methods The young and strong study was a cluster‐randomized controlled trial testing an intervention program for young women with breast cancer. Patients were surveyed within 3 months after diagnosis and at 3, 6, and 12 months after. Surveys asked about sociodemographics, psychosocial domains, fertility concerns, and fertility preservation strategies. Univariable and multivariable models were used to investigate sociodemographic, clinical, and psychosocial predictors of fertility concerns. Results Of 467 women from 54 clinical sites across the U.S. (14 academic, 40 community), 419 were evaluable regarding fertility concerns. Median age was 40 years (range 22–45), 11% were Black, 6% Hispanic, and 75% had children. Tumor stage was I (35%), II (51%), or III (14%); 82% received chemotherapy. At time of the treatment decision, 133 (32%) participants had fertility concerns, among whom 47% indicated this affected their treatment decisions. Sixty percent of participants reported having discussed fertility with their physician. Twenty percent of those with fertility concerns used fertility preservation strategies. History of difficulty becoming pregnant and younger age were associated with higher odds of fertility concerns in multivariable modeling. Conclusion Many young women with newly diagnosed breast cancer are concerned about fertility in a way that impacts their treatment decisions. Concerns were discussed, but few used fertility preservation strategies. These findings have implications for counseling young patients.


| INTRODUCTION
Approximately 10% of new breast cancer diagnoses in the U.S. affect women age 45 years and younger. 1 Tumors in this population are more likely to be aggressive, and intensive therapies including multiagent chemotherapy are frequently prescribed.Fertility can be affected by the direct gonadotoxic effects of chemotherapy or by the requisite delay of conception until the completion of adjuvant therapy, which is generally 5-10 years with standard endocrine therapy. 2 As postponement of pregnancy to the 30s and 40s is becoming more common in the general population, 3 it is more likely that a young woman who is diagnosed with breast cancer may not have started or completed her family and may harbor concerns about treatment-related infertility.][9][10][11][12] Even when fertility is addressed, a patient's options and access to, as well as effectiveness of fertility preservation strategies may be limited. 13][16] Some prior studies have examined fertility concerns in young breast cancer patients.In one European study led by Ruggieri et al., among 297 patients surveyed when they were making treatment decisions, 34% reported no concerns about becoming infertile. 15While Ruddy et al. 14 described the association of race with fertility concerns among 620 participants, this population was geographically restricted, and only 3% of participants were Black.We sought to better understand the prevalence of fertility concerns, factors associated with these concerns, and how fertility concerns affect treatment decisions and fertility preservation strategies at the time of decision-making in a more diverse national population of young women with breast cancer from across the U.S.

| Study design
The present study is a secondary data analysis from a prospective clinical trial, the Young and Strong study (NCT01647607 17 ).This cluster-randomized trial tested the effect of an educational and supportive care intervention for young women with breast cancer (diagnosed age ≤45 years) and their oncologists designed to improve attention to fertility.The study was conducted across the U.S. at 14 academic and 40 community practices which were randomized to either the young women's intervention (YWI) or to an attention control physical activity intervention (PAI).Participants provided written informed consent prior to enrollment.Institutional review board approval for the study was obtained from the Dana-Farber/Harvard Cancer Center and other participating institutions.
Between July 2012 and December 2013, 467 Englishspeaking women aged 18 to 45 years with newly diagnosed breast cancer were enrolled within 3 months of diagnosis.The primary results of the young and strong study are reported in detail elsewhere. 12In brief, the study did not demonstrate differences in attention to fertility, as assessed by medical record review, between the two intervention groups, with similar rates in both arms (55% for the YWI group vs. 58% for the PAI group, p = 0.88).
Thus, for the current analysis, in which we sought to describe the actual fertility concerns of the participants and related issues, the two arms were combined, and we used data from surveys completed by study participants at baseline (within 3 months of diagnosis, mean of 46 days), 3, 6, and 12 months after enrollment.

| Outcome measures
The primary outcome of this analysis was concerns about fertility when making breast cancer treatment decisions as assessed by one question on the 3-month survey.Women were asked "When you were making breast cancer decisions about your treatment, how concerned were you about the possibility of becoming infertile?"and categorized as concerned if they responded "a little," "somewhat," or "very," or not concerned if they responded "not at all" to the question.
Secondary outcomes, using items from a refined version of the fertility issues survey (FIS), 18 included: (1) use of fertility preservation prior to therapy ("no" or "unsure" vs. "yes" on the 3-month survey in response to "Prior to cancer treatment, did you take any special steps to lessen the chance that you would become infertile with cancer treatment?")and ( 2) impact of fertility concerns on breast cancer, decision-making, fertility, fertility concerns and preservation strategies, young women treatment decisions ("a little", "somewhat," or "very" vs. "not at all" on the 3-, 6-, or 12-month survey in response to "How much did your concern about becoming infertile after your cancer treatment affect your treatment decisions?").

| Statistical analysis
Demographic and clinical characteristics were analyzed using descriptive statistics.Stacked bar charts were plotted, and McNemar's test was used to show the change in fertility concerns over time.Univariate logistic regression was used to assess the association between fertility concerns and sociodemographics, tumor and treatment characteristics, psychosocial measures, and fertility-related variables.Parameter estimates are reported as odds ratios (ORs) with 95% confidence intervals (CI) derived from robust standard errors.Variables where the p-value was ≤0.20 in univariate analyses were evaluated in a multivariate logistic regression model using backward stepwise selection, and variables achieving significance at p ≤ 0.05 were included in the final model.Additional subgroup analysis was conducted among only those participants who were concerned about fertility assessed the association between patient/disease characteristics and treatment decisions affected by fertility concerns, as well as use of fertility preservation.Statistical analyses were performed in SAS 9.4 (SAS Institute Inc. Cary, NC).

| RESULTS
Of the 467 women enrolled, 48 (10%) did not respond to the 3-month survey or to the questions regarding fertility at the time of treatment decision-making and thus were excluded (Figure 1).Table 1 presents the patient and disease characteristics in the total analytic cohort (n = 419) and among the subset who indicated any fertility concerns (n = 133, including those reporting they were "a little" concerned [N = 46, 11%], "somewhat" concerned [N = 28, 6.7%], and "very" concerned [N = 59, 14.1%]).The median age of participants at enrollment was 40 years (range, 22-45).The majority of women were white (79%), 11% were Black, and 6% Hispanic.Most women were married or living with a partner (80%), had children (75%), and were college educated (88%).Chemotherapy was administered to the majority (82%).Thirty-five percent of tumors were stage I, 51% stage II, and 14% stage III.Approximately one-third of women reported high levels of anxiety, and two-thirds reported moderate or high levels of stress.
Among women who indicated any fertility concerns (n = 133), the median age was 35 (range, 22-45).Other demographics, tumor/treatment characteristics, and psychosocial measures were comparable to the full study sample.Half of the concerned group had one or more children already at the time of the cancer diagnosis (51%), and 26% had experienced some difficulty becoming pregnant previously.

| Fertility concerns, decision-making, and fertility preservation use
Regarding the changes in fertility concerns over time, the level of concern at 1 year was significantly lower compared to the level of concern at baseline (32%, baseline vs. 23%, 1 year, p < 0.001 by McNemar's test).
Table 2 presents fertility concerns, decision-making, and strategies for preservation and communication among the whole cohort and those defined as concerned.In the population of patients with concerns, a substantial proportion were concerned that a pregnancy would increase the risk of recurrence, whether they wished to have a child or not (55% and 42%, respectively).Other concerns about having children were caring for them if cancer recurred (37% if they wished to have a child, and 21% if not) and the child having an increased risk of developing cancer (45% if they wished to have a child, and 11% if not).
Decision-making about treatments was reported as affected by fertility concerns to some degree in almost half of the patients with concerns (47%) and 14% of those without concerns.Over the year of follow-up, using data from the 3-, 6-, and 12-month follow-up surveys, participants  reported various impacts on treatment including 29% of those concerned considering taking fewer than 5 years of endocrine therapy.Standard strategies to preserve fertility were primarily pursued by women who were concerned about fertility, though only 20% of these participants did undergo preservation strategies: 14 underwent cryopreservation of embryos, 14 underwent cryopreservation of eggs, and 8 took a GnRH agonist through chemotherapy.Overall, 60% of participants discussed fertility concerns with a provider at the time of treatment decision-making, including 88% of those who indicated fertility concerns, and 87% who reported that this was addressed adequately.
In the multivariate model (Table 3), women with a history of difficulty becoming pregnant had a 3-fold greater odds of having fertility concerns and younger women had higher odds of fertility concerns.Those who had a child before diagnosis had 81% lower odds of fertility concerns compared with women who did not.While being highly educated was associated with greater fertility concerns and menstruating fewer than once every 12 months at diagnosis was associated with less fertility concerns in the univariate models, they were not statistically significant in the multivariate model.Factors not associated with fertility concerns in univariate models, and thus not included in multivariate models, included race, site, income, marital status, having a first-degree relative with breast or ovarian cancer, tumor biology (stage, ER/PR receptors, and HER2 expression), treatment, psychosocial measures and history of infertility treatments.5).

| DISCUSSION
Our findings confirm and expand upon the previous literature regarding fertility concerns and their consequences among young women with breast Abbreviations: CI, confidence interval; ER, estrogen receptor; Her2, human epidermal growth factor receptor 2; OR, odds ratio; PR, progesterone receptor.
cancer. 11,14,15,18,24,25The fact that 32% of the patients had some fertility concerns at the time of treatment decision-making is somewhat lower than noted in prior studies.In a cross-sectional analysis of 657 breast cancer survivors (average age at diagnosis was 33 years), 73% of respondents reported at least some degree of fertility concern. 18In an analysis of 724 women enrolled in a prospective cohort study, 51% of women (median age of 37 years) reported at least some degree of concerns. 14n a prospective study conducted in Europe (n = 297), 32% of whom were under 35 years of age, and 64% had at least some fertility concerns. 15These collective findings suggest that the lower prevalence observed in our study may in part be attributable to the older age of our population (45% of the participants were aged between 41 and 45 years), with these women more likely to have finished their desired childbearing at the time of diagnosis.It is also possible that this difference may be attributable to the large proportion of patients enrolled from community sites in our study; patients seen at academic sites may have more concerns.
Our finding that experiencing difficulty becoming pregnant in the past and not having had a child were associated with greater concerns about fertility was not unexpected.However, the fact that the youngest participants had the highest odds of concerns might suggest that there is need for additional patient education regarding fertility-related risks.Current evidence suggests that about 88% of the youngest women will not experience any treatment-related amenorrhea, 26 and that only 5% of those diagnosed at age 30 are likely to be infertile after resumption of menses. 27Furthermore, our finding in the concerned population that about half were concerned that a pregnancy would increase risk of recurrence, underscores the importance of ensuring patients understand these risks.It should be communicated to patients who are interested in future childbearing that the currently available collective evidence indicates that pregnancy is safe in breast cancer survivors, including those with ERpositive disease. 28,291][32] Providers generally appear to be following guidelines recommending that fertility be routinely addressed with young cancer patients.Oncofertility counseling to inform patients about the risk of infertility and strategies for preservation of ovarian function and/or fertility is increasingly considered standard of care at the time of diagnosis in all young women with breast cancer.It is notable and reassuring that none of the psychosocial factors was associated with fertility concerns, which is consistent with prior research.In a previous study by Partridge et al., 18 nearly 30% of women interested in future fertility reported consideration of taking hormonal medication for fewer than 5 years.However, in a prospective study of 384 women with stage I to III breast cancer diagnosed at age 40 years and younger, there was no significant difference at 30 months regarding the proportion of "more concerned about fertility" patients between adherers and non-adherers of hormonal therapy (39% of the adherers were more concerned versus 43% for the non-adherers, p = 0.44). 33From this same prospective cohort, Sella et al. 34 reported that concern about fertility was a contributor to adjuvant endocrine therapy decisions among a substantial proportion of young breast cancer survivors.It is also possible that feelings about fertility and potential impact on treatment decisions may change over time, which we were unable to observe in our study given that we only followed patients out 12 months.Nevertheless, recent data from a prospective clinical trial of temporary interruption of endocrine therapy for pregnancy (the POSITIVE Trial) demonstrated short-term safety from this approach, which should help to address the concerns and inform the decisions regarding endocrine therapy. 29t is recommended that providers be prepared to discuss fertility preservation options and/or to refer all potential patients to appropriate reproductive specialists, 6 yet only one-fifth of those concerned about fertility underwent fertility preservation.This low rate is consistent with prior reports including that of Lambertini et al. 35 that demonstrated that only a minority of young breast cancer patients (≤40 years of age) accessed a fertility unit to discuss preservation procedures (29%), and even fewer (12%) ultimately decided to undergo one of the available cryopreservation strategies.This was comparable with the data from an earlier prospective cohort study 14 : Despite the fact that half of the cohort presented some fertility concerns, only 8.5% accessed one of the available cryopreservation procedures.Race, income, education, community versus academic site, and undergoing neoadjuvant chemotherapy did not predict that utilization of fertility preservation strategies.It is important to highlight that these data were collected back in 2012-2013.Since then, access and options to fertility preservation have evolved favorably in the U.S., including the emergence of state-based fertility preservation insurance mandates. 36However, a significant number of barriers to accessing oncofertility care remain, including lack of financial or geographic access, 37 lack of information and referral, and personal fear or uncertainty regarding fertility presentation.While we were unable to evaluate specific reasons for non-utilization in our analysis, we believe that medical insurance coverage and cost of the procedures were factors that may have discouraged patients from pursuing these procedures, as demonstrated in prior studies. 38All of the patients in this study had insurance, and we did not assess other barriers to the use of fertility preservation.
The present study has several important limitations.While the trial was conducted at sites across the U.S. and clinics were asked to recruit all eligible patients systematically at their first visit, patients who enrolled may not be fully representative of all patients with premenopausal breast cancer.For example, the proportion with college education was very high (88%), limiting generalizability.In addition, the relatively small sample of "concerned" patients may hinder the detection of weak associations.However, strengths of the present study include the enrollment and baseline assessment of fertility concerns within 3 months following diagnosis, T A B L E 5 Logistic regression evaluating associations between patient/disease characteristics and any fertility preservation strategy in the concerned cohort (n = 133).minimizing the recall bias that impacted prior studies of this issue among longer-term survivors.Moreover, our population included patients from 14 academic and 40 community practices geographically distributed across the U.S. Further, the 90% survey response rate was quite high for our main outcome.

| CONCLUSION
This large, prospective, geographically, and racially/ ethnically diverse study confirms and expands upon existing knowledge regarding fertility concerns and their consequences for treatment decision-making and fertility preservation strategies among young women with breast cancer.As a significant proportion of women are concerned about fertility at diagnosis that may impact their treatment decisions, it is critical to ensure that these issues are adequately addressed.

T A B L E 1
Patient demographic and clinical characteristics.

Factors associated with having a treatment decision affected by fertility concerns and utilization of fertility preservation among patients with fertility concerns
Fertility concerns, decision-making, and strategies.
T A B L E 2a Some patients indicated >1.T A B L E 2 (Continued) 3.2 | 1.66-9.50)wasassociatedwith having a treatment decision affected by fertility concerns, whereas having children was associated with a lower likelihood of having a treatment decision affected (OR = 0.33, 95% CI: 0.15-0.74) in the final multivariate model (Table4).Other demographic, treatment, and psychosocial characteristics Logistic regression evaluating associations between patient/disease characteristics and fertility concerns.
T A B L E 3 14 Logistic regression evaluating associations between patient/disease characteristics and treatment decisions being affected by fertility concerns in the concerned cohort (n = 133).