Cabergoline as a Novel Strategy for Post-Pregnancy Breast Cancer Prevention in Mice and Human

Post-pregnancy breast cancer often carries a poor prognosis, posing a major clinical challenge. The increasing trend of later-life pregnancies exacerbates this risk, highlighting the need for effective chemoprevention strategies. Current options, limited to selective estrogen receptor modulators, aromatase inhibitors, or surgical procedures, offer limited efficacy and considerable side effects. Here, we report that cabergoline, a dopaminergic agonist, reduces the risk of breast cancer post-pregnancy in a Brca1/P53-deficient mouse model, with implications for human breast cancer prevention. We show that a single dose of cabergoline administered post-pregnancy significantly delayed the onset and reduced the incidence of breast cancer in Brca1/P53-deficient mice. Histological analysis revealed a notable acceleration in post-lactational involution over the short term, characterized by increased apoptosis and altered gene expression related to ion transport. Over the long term, histological changes in the mammary gland included a reduction in the ductal component, decreased epithelial proliferation, and a lower presence of recombinant Brca1/P53 target cells, which are precursors of tumors. These changes serve as indicators of reduced breast cancer susceptibility. Additionally, RNA sequencing identified gene expression alterations associated with decreased proliferation and mammary gland branching. Our findings highlight a mechanism wherein cabergoline enhances the protective effect of pregnancy against breast cancer by potentiating postlactational involution. Notably, a retrospective cohort study in women demonstrated a markedly lower incidence of post-pregnancy breast cancer in those treated with cabergoline compared to a control group. Our work underscores the importance of enhancing postlactational involution as a strategy for breast cancer prevention, and identifies cabergoline as a promising, low-risk option in breast cancer chemoprevention. This strategy has the potential to revolutionize breast cancer prevention approaches, particularly for women at increased risk due to genetic factors or delayed childbirth, and has wider implications beyond hereditary breast cancer cases.

postlactational involution as a strategy for breast cancer prevention, and identi es cabergoline as a promising, low-risk option in breast cancer chemoprevention.This strategy has the potential to revolutionize breast cancer prevention approaches, particularly for women at increased risk due to genetic factors or delayed childbirth, and has wider implications beyond hereditary breast cancer cases.

MAIN
Breast cancer stands as the foremost cancer in women globally, leading in cancer-related deaths (1).This prevalence is in uenced by factors such as increased life expectancy and a trend towards pregnancies after age 30, factors that have been implicated in the elevated incidence rates in Western societies (2)(3)(4).
For instance, the United States has observed an annual increase in incidence rates by approximately 0.5% (5).
A critical epidemiological aspect is the heightened risk of breast cancer shortly after pregnancy, lasting up to a decade, especially in women who bear children after age 35 (4,6).This risk is underscored by a demographic shift towards later maternal age.In the U.S., the proportion of births to women over 35 increased from 8% in 1990 to 18% in 2018 (7).Alarmingly, an increasing incidence of breast cancer is noted among younger women (15 to 44 years), with post-pregnancy breast cancer accounting for a substantial portion of these cases.Notably, this cancer subtype, regardless of estrogen receptor (ER) status, is often marked by a more aggressive course, with increased risks of metastasis and mortality(8).
Current chemopreventive strategies, such as selective estrogen receptor modulators (SERMs) and aromatase inhibitors (AIs), are hampered by adverse effects including endometrial cancer and thrombosis (9).Surgical options like mastectomy and oophorectomy, though effective in drastically reducing risk, particularly in hereditary breast cancer, are associated with signi cant physical and psychological repercussions (10)(11)(12).Chemoprevention is typically reserved for women at very high risk (Gail score >3%), yet its utilization is impeded by side effects, leading to a 40% discontinuation rate in the prescribed 5-year regimen (13,14) .Moreover, these interventions offer temporary e cacy and do not entirely eliminate risk, as evidenced by the modest reductions in invasive breast cancer incidences achieved by tamoxifen and raloxifene (9).Critically, there are no speci c preventive strategies for postpregnancy and hormone receptor-negative breast cancers, frequently observed in BRCA1 mutation carriers (15).
Early childbearing, particularly before the age of 30, signi cantly reduces the long-term risk of breast cancer.This protective effect is inversely proportional to maternal age at rst childbirth, with earlier ages conferring greater bene ts (16).Early pregnancy offers partial protection against both ER-positive and ERnegative breast cancers in women with BRCA1 mutations (17)(18)(19), with each subsequent pregnancy further enhancing this protective effect (20).
The protective in uence of pregnancy against breast cancer is a universal phenomenon, observed across diverse ethnic groups and in mammalian models (21)(22)(23).Various cellular subpopulations and signaling pathways are implicated in this protection (24)(25)(26)), yet the precise mechanisms remain elusive.
Our study proposes that enhancing postlactational involution could lower breast cancer risk by eliminating damaged epithelial cells more effectively.We explored the use of cabergoline, a dopaminergic agonist known for prolactin inhibition, previously applied in dairy cows for faster involution (40)(41)(42) and to suppress lactation in non-breastfeeding women (43,44).
We demonstrate that cabergoline, by intensifying postlactational involution in mice, ampli es the pregnancy's protective effect against breast cancer.This effect involves histological and molecular alterations in the mammary glands, reducing their susceptibility to malignancy.Our ndings also show that cabergoline may decrease the risk of post-pregnancy breast cancer in women.

RESULTS
Enhanced breast cancer protection in Brca1/P53-de cient mice post-cabergoline treatment in pregnancy context The well-established protective effect of pregnancy against breast cancer, documented across diverse mammalian models including mice and rats (16, [21][22][23], was further explored in this study using the Brca1/P53-de cient mouse model.These mice, predisposed to developing triple-negative basal breast cancer (45), were utilized to determine if pregnancy's protective effects could be potentiated in a genetically susceptible context.
In this comprehensive longitudinal study, female Brca1/P53-de cient mice underwent two full reproductive cycles, including pregnancy and lactation.For the purposes of this manuscript, these mice will be referred to interchangeably as either 'parous and breastfeeding mice' or 'multiparous mice' (Fig. 1a).This cohort was compared to nulliparous counterparts to assess the protective effect of pregnancy.The results indicated a signi cant delay in breast cancer onset in multiparous mice (median onset shifted from 175 to 224 days; P = 0.0002) and a notable reduction in tumor incidence (from 88.6% to 59.4%; P = 0.0061) (Fig. 1b, c and Extended Data Fig. 1a, b).
Given the known relationship between postlactational involution and breast cancer risk in several genetically modi ed mouse models, where delayed involution is associated with increased risk and accelerated involution with decreased risk (27), we explored the impact of cabergoline, a drug recognized for hastening postlactational involution in dairy cows (41,42).Our focus was on its e cacy in altering breast cancer development post-pregnancy.
Administering a single dose of cabergoline at the conclusion of the second lactation period resulted in a marked extension of tumor latency in multiparous mice, compared to both untreated multiparous and nulliparous groups (extension to 259 days; P = 0.0119 and P < 0.0001, respectively) (Fig. 1b and Extended Data Fig. 1c, d).Moreover, cabergoline treatment notably decreased tumor incidence in multiparous mice, with a reduction to 27.8%, in contrast to 88.6% in nulliparous and 59.4% in untreated multiparous counterparts.These differences were statistically signi cant (P < 0.0001 and P < 0.0065, respectively) (Fig. 1c and Extended Data Fig. 1e, f).
Further exploration of cabergoline's potential protective effect in nulliparous Brca1/P53-de cient mice revealed a marginal and statistically non-signi cant decrease in tumor latency (P = 0.0771), with no observable change in tumor incidence (P = 0.3460) (Fig. 1d, f).A biweekly cabergoline regimen in a separate nulliparous cohort showed no signi cant alterations in tumor incidence or latency (Extended Data Fig. 1g, h), indicating the absence of a dose-dependent protective effect of cabergoline in the context of non-pregnancy.These ndings collectively underscore the signi cance of post-pregnancy cabergoline treatment in substantially reducing breast cancer risk in a genetically susceptible model, an effect not replicated in the absence of pregnancy.
Cabergoline's role in enhancing postlactational involution in Brca1/P53-de cient mice Building on the premise that accelerated postlactational involution is inversely correlated with breast cancer risk (27), this study delved into the mechanisms by which cabergoline augments pregnancyrelated breast cancer protection.A key focus was to determine if cabergoline indeed ampli es postlactational involution in a Brca1/P53-de cient mouse model.
In our detailed investigations, mammary glands of cabergoline-treated Brca1/P53-de cient mice were examined at intervals post-lactation.The ndings revealed an accelerated involution process, characterized by an expanded adipocyte area at three and seven days post-treatment (P < 0.0001), suggesting an early onset of involution (Fig. 2a-c).A corresponding increase in apoptotic epithelial cells at the 3-day mark (P < 0.0001) further con rmed this observation (Fig. 2d-f).These trends were also evident in wild-type mice, reinforcing the role of cabergoline in involution acceleration (Fig. 2a-f).
We then explored the molecular underpinnings of this accelerated involution.The focus was on pSTAT3 and pSTAT5, molecules pivotal in initiating postlactational involution (38,46).In cabergoline-treated mice, an upregulation of pSTAT3 was observed (Fig. 2g), aligning with the drug's involution-enhancing effects.Subsequent RNA-seq analysis of mammary glands 24 hours post-lactation identi ed 14 differentially expressed genes (DEGs) (P Adj ≤ 0.05) between untreated and cabergoline-treated groups, although initial enrichment analysis did not directly link these changes to involution (Supplementary Tables S1, S2).
To elucidate the underlying mechanisms of our observed increase in cell death during post-lactation involution, we re ned our analytical approach.Initially, we identi ed differentially expressed genes (DEGs) using a threshold of P ≤ 0.05.Subsequent enrichment analysis of these DEGs enabled us to focus on speci c biological processes relevant to our study.This approach, in line with established methodologies (47)(48)(49)(50)(51), enabled the identi cation of 263 overexpressed and 347 underexpressed transcripts in treated mice, facilitating distinct clustering of mammary gland samples from treated and untreated mice (Extended Data Fig.S2a and Supplementary Table S3).
The data collectively indicate that cabergoline enhances post-lactational involution, marked by increased apoptosis and accelerated adipocyte expansion (Fig. 2a-f).The drug appears to facilitate involution by regulating ion channel homeostasis, particularly impacting calcium ion transport, thus enhancing cell death and promoting involution.
Long-term impact of cabergoline on mammary gland ductal structure and proliferation dynamics in Brca1/P53-de cient mice Given the established correlation between mammographic breast density and breast cancer risk (59,60), and the inverse relationship between glandular fraction and mammary involution (61), we focused on the long-term effects of cabergoline on mammary gland composition as an indirect measure of breast cancer susceptibility (62, 63).
This study employed a cross-sectional and longitudinal approach to assess the glandular tissue evolution in Brca1/P53-de cient mice post-cabergoline treatment (Fig. 3).
In the cross-sectional analysis, we quanti ed ductal tissue in Brca1/P53-de cient mice post-involution.The comparison involved multiparous mice, both untreated and treated with cabergoline, at 30-and 60days post-involution, alongside nulliparous, age-matched controls.Nulliparous mice exhibited an increased glandular component over time (P < 0.0001), while multiparous mice showed a decrease (P = 0.0038).Notably, cabergoline-treated mice consistently demonstrated the least glandular component, with minimal variation between the two time points (Fig. 3a).
In the longitudinal analysis, which included mice previously assessed for breast cancer susceptibility (Fig. 1), we observed a progressive increase in ductal area in nulliparous mice (P < 0.0001; r = 0.7842) (Fig. 3b).This pattern was not evident in multiparous mice (Fig. 3c).Intriguingly, cabergoline-treated mice maintained a smaller initial ductal area that increased minimally with age (Fig. 3d), but consistently less than in untreated mice.The minimal ductal area in the cabergoline-treated group was con rmed at 60 days post-involution (Fig. 3e, f), aligning with our hypothesis of cabergoline's role in reducing breast cancer susceptibility.Comparing cabergoline's effect on multiparous wild-type mice, we observed a consistent reduction in ductal area (Extended Data Fig. 3a-c).
To further explore the impact of cabergoline on mammary gland dynamics, we assessed apoptosis and epithelial proliferation in Brca1/P53-de cient mice.The cabergoline-treated group showed a signi cantly higher apoptotic rate than untreated multiparous mice (P = 0.0268) (Extended Data Fig. 3d, e), albeit minimal across all groups.Ki-67 expression, a marker for basal proliferation and a recognized breast cancer risk factor (64-66), was employed to quantify glandular tissue proliferation.Thus, we used this marker to assess glandular tissue proliferation as an indicator of breast cancer susceptibility.We found a signi cant increase in epithelial cell proliferation from 30 to 60 days in nulliparous mice (P = 0.0006), a trend absent in multiparous mice.Cabergoline-treated mice exhibited markedly lower proliferation rates at both time points (Fig. 3g).
Comparisons in wild-type mice at 30 and 60 days post-involution, both treated and untreated with cabergoline, corroborated these ndings (Extended Data Fig. 3f-h).Notably, the proliferation rate in nulliparous wild-type mice was lower than in their Brca1/P53-de cient counterparts (Extended Data Fig. 3i), emphasizing the potential role of basal breast tissue proliferation in breast cancer susceptibility (64-68).
Lastly, we linked the diminished proliferation in cabergoline-treated Brca1/P53-de cient mice to reduced pAKT and pERK levels in breast organoids, as determined by Western blot analysis (Fig. 3m, n).
These ndings suggest that extended pregnancy, followed by cabergoline treatment, leads to a reduction in glandular components and suppresses ductal proliferation, thereby aligning with the reduced breast cancer susceptibility observed in Brca1/P53-de cient mice (Fig. 1b, c and Extended Data Fig. 1a-f).
Deciphering the molecular mechanisms underlying decreased breast cancer risk in multiparous Brca1/P53-de cient mice through RNA-seq analysis In an effort to unravel the molecular basis for the observed reduction in breast cancer risk in multiparous Brca1/P53-de cient mice, we carried out a comprehensive RNA-seq study.We analyzed non-tumoral breast tissues from nulliparous and multiparous mice 60 days post-lactational involution.This comparative approach identi ed distinct sets of differentially expressed genes (DEGs) (P Adj ≤ 0.05) between these groups, providing insights into the molecular divergence contributing to the altered cancer susceptibility (Extended Data Fig. 4a and Supplementary Table S6).
Utilizing Gene Ontology (GO) enrichment, Reactome, and WikiPathways analyses, we categorized these DEGs into functionally relevant biological processes and pathways.The enrichment analysis highlighted a spectrum of biological processes and pathways associated with both overexpressed and underexpressed genes in the multiparous group compared to their nulliparous counterparts (Supplementary Table S7).Key biological processes are depicted in Extended Data Fig. 4b, c.
Correlating our ndings with the progressive increase in ductal area over time in nulliparous mice (Fig. 3b,   c), and the attenuated proliferation observed in multiparous mice (Fig. 3h, i, k), we focused on clusters related to tubular growth, breast development, and cellular proliferation (Fig 4a and Supplementary Tables S8, S9).Pregnancy consistently resulted in lower gene expression within these clusters in multiparous mice.In contrast, nulliparous mice showed a higher and broader range of transcriptomic variability in non-tumoral mammary glands (Fig. 4a, b), aligning with the observed differences in ductal area and proliferation rates.
The underexpression of genes associated with cell proliferation, mammary gland development, and branching in multiparous mice's breasts may be a key factor in their reduced cancer susceptibility.Critical genes in this context include Pgr, Sox9, Fgfr2, Erbb4, and the Hgf/Met signaling axis, among others (Extended Data Fig. 4d).The altered expression of Pgr suggests shifts in hormonal signaling pivotal to mammary tissue development and proliferation (69, 70).Sox9, essential in stem cell differentiation (71), alongside Fgfr2 and Erbb4, crucial for cell survival and signaling (72)(73)(74)(75)(76)(77), indicate a protective shift against cancer cell proliferation and transformation.The role of the Hgf/Met pathway in cell proliferation and mammary morphogenesis further underscores its potential contribution to mitigating cancer initiation and progression (78, 79).
In summary, these molecular patterns suggest a link between diminished activity in speci c pathways and the reduced glandular development and proliferation observed in multiparous mice.This molecular landscape potentially underlies the lowered breast cancer susceptibility in this speci c mouse model, providing valuable insights into the interplay between genetic factors, pregnancy, and cancer risk.

RNA-Seq analysis of cabergoline-induced gene expression changes in mammary glands of multiparous mice
Later, we expanded our investigation to elucidate the impact of cabergoline treatment on gene expression in the mammary glands of multiparous mice.Through comparative RNA-Seq analysis, distinct differentially expressed genes (DEGs) (P Adj ≤ 0.05) were identi ed between cabergoline-treated and untreated groups, as detailed in Supplementary Table S10.Heatmap analyses effectively demarcated the gene expression pro les of these cohorts (Extended Data Fig. 5a).
Utilizing Gene Ontology (GO), Reactome, and Wikipathways analyses, we categorized these DEGs into signi cant biological pathways impacted by cabergoline treatment in multiparous mice.These pathways, crucially altered by cabergoline, might elucidate the molecular underpinnings of its role in augmenting breast cancer protection post-pregnancy (Supplementary Table S11).Noteworthy among these pathways are those associated with attenuated in ammatory responses and modulated TNFα and insulin/IGF1 sensitivity (Supplementary Table S12), key elements in breast cancer susceptibility (80, 81).
Following the observation of a reduced glandular component and lower proliferation rates in Brca1/P53de cient mice's mammary glands post-cabergoline treatment (Fig. 3), we honed in on pertinent DEGs related to proliferation and development (Fig. 5a, b; Supplementary Tables S12, S13).Among others, cabergoline was found to upregulate genes such as Krt14, Serpinb5, and Trim29 (Extended Data Fig. 5b), potentially augmenting cellular differentiation, DNA stability, and immune responses.Of particular interest, Trim29 may offer protective roles against DNA damage and in ammation-associated cancer, impacting estrogen receptor (ER) signaling (82, 83), while Serpinb5 (Maspin) is implicated in tumor suppression in mammary cells (84).Decreased expression of genes like Scd1, Fasn, Sox2, and Lep suggests a reduction in fat production and pluripotency (85, 86); additionally, elevated Leptin levels have been linked to mammary gland hyperplasia (87) (Extended Data Fig. 5b).These genetic alterations might underpin the observed reduced proliferation and diminished breast cancer risk in cabergoline-treated mice.
To further dissect cabergoline's speci c effects within the context of parity, we compared gene expression pro les between untreated and cabergoline-treated multiparous mice with those of nulliparous mice.DEGs (P Adj ≤ 0.05) unique to cabergoline-treated multiparous mice were examined (Fig. 5c; Extended Data Fig. 5c; Supplementary Table S14), revealing pathways that might underlie cabergoline's enhanced protective effect against breast cancer, particularly those involving underexpressed DEGs related to in ammation and glandular development (Supplementary Table S15).
Our focus extended to genes implicated in proliferation and development, informed by literature and our enrichment study (Fig. 5d, e; Supplementary Table S16).In cabergoline-treated multiparous mice, key genes such as ErbB3, Klf5, Igfbp1, Fgf17, Msx2, and Sox10 (88-92), among others, were notably underexpressed compared to nulliparous counterparts (Extended Data Fig. 5d).This underexpression suggests a plausible mechanism for the observed reduction in breast cancer risk, potentially due to constrained mammary gland growth and regulated cell proliferation, particularly in these pivotal genes governing mammary gland development and cell signaling (88-92).
In sum, these results provide a detailed molecular landscape that might elucidate how cabergoline administration in multiparous mice might confer a reduced risk of breast cancer.
In uence of cabergoline on reducing Brca1/P53-de cient cell populations in mammary glands Informed by the concept of " eld cancerization," which links an increased number of pre-malignant cells in a tissue to heightened cancer risk (93,94), our study examined the long-term impact of cabergoline on the prevalence of Brca1/P53-de cient cells in the mammary gland.This analysis is pivotal given that in Brca1/P53-de cient mice, these cells, generated through Cre-recombinase action on K14+ cells, are the potential tumor progenitors (45).We sought to ascertain whether pregnancy and cabergoline treatment would alter the ratio of these recombinant cells, using this as a barometer for breast cancer risk.
Quantifying the fraction of recombinant cells within the mammary gland of Brca1/P53-de cient mice, where 10-40% of cells are recombinant and predisposed to tumorigenesis (45), we utilized qPCR, an established method for cellular chimerism detection (95,96).Our research reveals that the proportion of recombinant cells was signi cantly lower in cabergoline-treated multiparous mice compared to the other groups of mice in the longitudinal cohort (P < 0.001) (Fig. 6a).
An upward trend in recombinant cells de cient in Brca1/ P53 was observed over time in nulliparous mice (P = 0.0365; r = 0.6333.)(Fig. 6b).However, this increase was not evident in either untreated or cabergoline-treated multiparous mice (P = 0.2589 and P = 0.7092, respectively) (Fig. 6c, d).At 60 days post-weaning, the cross-sectional cohort revealed no signi cant differences in the proportion of recombinant cells (P = 0.3416) (Fig. 6e), suggesting that the impact of cabergoline on recombinant cell populations becomes more pronounced over time.Notably, after adjusting for the cabergoline-induced reduction in ductal area, a substantial decrease in P53 recombinant alleles was evident in treated multiparous mice (P = 0.0344) (Fig. 6f).
Further validation was pursued using a second mouse model with Red Fluorescent Protein (RFP)+ Brca1/P53-de cient cells (Fig. 6g, h and Supplementary Figure S1) (97).Flow cytometry quanti cation of these cells post-lactational involution showed no signi cant differences at 60 days (P = 0.3855), but a marked reduction in ductal area was observed in cabergoline-treated mice (P < 0.0001) (Fig. 6i, j), accompanied by a decrease in recombinant RFP+ cells after adjusting for ductal area (P = 0.0006) (Fig. 6k).
Flow cytometry was also employed to dissect changes in various RFP+ recombinant epithelial cell subpopulations (Fig. 6l).Pregnancy and cabergoline treatment led to a reduced proportion of luminal cells (characterized by low CD49f and high EpCAM), with this effect pronounced in untreated multiparous mice compared to nulliparous counterparts (Fig. 6m, left).No signi cant difference was observed between cabergoline-treated and untreated multiparous groups across all cell subpopulations (Fig. 6m).Yet, accounting for the cabergoline-induced reduction in ductal area, all recombinant cell subpopulations exhibited a decline in the treated group (Fig. 6n).
Given the hypothesis that BRCA1-de cient breast cancer originates from luminal precursors (98, 99), the observed decrease in luminal cells implies that both pregnancy and cabergoline may diminish the pool of potential tumor-initiating cells.This effect was also re ected in RFP-negative cells, indicating a broader impact of the interventions (Extended Data Fig. 6a, b).
In summary, cabergoline treatment in parous mice signi cantly lowers the count of Brca1/P53-de cient cells, chie y by reducing the glandular component.This consequential decrease narrows the ' eld cancerization' zone, potentially enhancing post-pregnancy protective effects against breast cancer development.

Assessment of cabergoline's post-pregnancy protective effect against breast cancer in women
Our study embarked on a retrospective dynamic cohort analysis (2005-2021) to explore the potential protective impact of cabergoline administered post-pregnancy against breast cancer.Utilizing data from the Information System for the Development of Research in Primary Care (SIDIAP) database (100,101) (Extended Data Fig. 7), we scrutinized the breast cancer incidence among women who had their rst child at age 30 or older.The cohort comprised 14,810 women, with 780 receiving cabergoline within one month of childbirth and 13,898 serving as the control group.The epidemiological characteristics of these groups are detailed in Supplementary Table S17.
In the control group, 132 breast cancer cases were documented, contrasting with only 2 cases in the cabergoline-treated group.The annual cumulative incidence rates per 1,000 population were 1.17 for the control group and 0.3 for the treated group, with average follow-up durations of 8.03 and 8.56 years, respectively (Fig. 7a).Bivariate Cox regression yielded a Hazard Ratio (HR) of 0.247 (95% CI: 0.061 -0.998, P = 0.050) for the cabergoline-treated group (Fig. 7b), with additional bivariate analysis details in Supplementary Table S18.Further, a multivariate Cox regression, adjusting for signi cant epidemiological breast cancer risk factors, substantiated the diminished risk in the cabergoline-treated group (HR = 0.239; 95% CI: 0.059 -0.968) (Fig. 7c).
Expanding the scope, we examined breast cancer incidence over a decade following the rst pregnancy, irrespective of maternal age.This included 27,467 women, with 1,604 in the cabergoline-treated group and 25,863 controls.The control group reported 125 breast cancer cases, compared to just 2 in the treated group.Detailed epidemiological pro les are available in Supplementary Table S19.The annual cumulative incidence rates per 1,000 population were 0.66 for the control group and 0.17 for the treated group (Fig. 7d).Bivariate Cox regression analysis indicated an HR of 0.239 (95% CI: 0.059 -0.968, P = 0.045) for the cabergoline group (Fig. 7e), with further information in Supplementary Table S20.The multivariate Cox regression analysis rea rmed a consistently lower breast cancer risk in the cabergolinetreated group (adjusted P = 0.056) (Fig. 7f).
In summary, these results collectively suggest that administering cabergoline postpartum may signi cantly reduce the risk of breast cancer following pregnancy, highlighting its potential as a protective intervention in this speci c context.

DISCUSSION
The escalating incidence of breast cancer, particularly in the context of delayed childbearing and the speci c challenges of post-pregnancy breast cancer, underscores the pressing need for novel chemopreventive strategies (1-4).Our study breaks new ground by proposing cabergoline, a dopaminergic agonist, as a potential pharmacological agent for breast cancer prevention, notably in the post-pregnancy period.This research demonstrates the e cacy of cabergoline in enhancing postlactational involution, subsequently reducing breast cancer risk in Brca1/P53-de cient mice.The signi cance of this nding is ampli ed by the increasing trend of delayed childbirth (7) and its correlation with heightened breast cancer risk, particularly among women aged over 30-35 years (4).The search for new pharmacological alternatives is further emphasized by the limitations and adverse effects associated with current chemopreventive methods such as selective estrogen receptor modulators (SERMs), aromatase inhibitors (AIs), and surgical options (9)(10)(11).Cabergoline emerges as a promising candidate, offering potential for wider application with fewer side effects (43,44).
Our ndings are particularly relevant considering that approximately 80% of mammary epithelial cells undergo apoptosis during postlactational involution, including those with genetic mutations (102,103).
The age-related accumulation of premalignant lesions, if not cleared, may lead to cancer initiation (104,105).Cabergoline-induced intensi cation of postlactational involution aligns with a decrease in long-term tumor susceptibility, a crucial consideration given the pro-in ammatory environment of postlactational involution, which could otherwise promote tumorigenesis (106).
Importantly, our study demonstrates that cabergoline not only induces a hypoproliferative state in mammary epithelial tissue but also mitigates the glandular component, which correlates positively with mammographic density, a well-documented risk factor for breast cancer (59,61,94).RNA sequencing and enrichment analyses have further substantiated cabergoline's role in activating antiproliferative pathways, underscoring its preventive potential.Moreover, we highlight the impact of cabergoline on reducing the fraction of Brca1/P53-de cient recombinant cells in the mammary gland.Drawing on the concept of " eld cancerization" (94), our ndings indicate that cabergoline-treated mice exhibit fewer pre-malignant cells.This reduction, paired with the diminished glandular component, suggests that cabergoline narrows the ' eld cancerization' zone, amplifying the protective shield against breast cancer established post-pregnancy.
Extending to human applications, our epidemiological study suggests cabergoline's potential in reducing post-pregnancy breast cancer risk.This is especially pivotal given the current absence of targeted chemopreventive strategies for this subtype of cancer (8).While our data do not distinguish between ERnegative and ER-positive tumors, they hint at cabergoline's potential e cacy against ER-negative breast cancer, typically more common in younger women (107).
However, our study has limitations.The epidemiological component con rmed breast cancer prevention only in post-pregnancy cases.Further analysis, possibly including cohorts treated with cabergoline prior to 2005, is needed to ascertain cabergoline's long-term effects on postmenopausal breast cancer.
Additional studies are required to con rm the long-term implications of cabergoline on breast cancer risk.
Cabergoline, typically used in a single dose for lactation cessation, is known for its tolerability (43,44).Its minimal side effects and convenient dosing present an advantageous alternative to current chemopreventive methods that often require daily intake over extended periods (9,14,108).
In conclusion, our study introduces cabergoline as a novel and potentially effective strategy for breast cancer prevention, particularly in the post-pregnancy context.Its application post-breastfeeding in women at high risk of breast cancer could enhance the natural protective effect of pregnancy.Cabergoline's potential to prevent both hormone receptor-negative and post-pregnancy breast cancer could represent a signi cant advancement in breast cancer prevention strategies.Further research is warranted to explore its long-term effects and suitability for a broader population.

METHODS
Mice models and experimental procedures -Model development: Utilizing genetically modi ed mice, our study employed the K-14 Cre; Brca1 f/f /P53 f/f (Brca1/P53) double knockout model, expressing Cre recombinase under the cytokeratin 14 promoter for targeted Brca1 and P53 deletion in speci c cells (45).Additionally, we integrated the RFP; K-14 Cre; Brca1 f/f /P53 f/f model, established by crossing Brca1/P53 mice with C57BL/6-Gt(ROSA)26tm1Hjf mice (97).All models were housed at the University of Salamanca's Animal Research Facility, adhering to ethical and pathogen-free standards.In the different studies, female subjects started mating at the age of six weeks.
-Study design: The investigation was segmented into three core studies: Early postlactational analysis: Focusing on early post-weaning phases (1, 3, and 7 days), we evaluated mammary gland involution in Brca1/P53-de cient and wild-type mice, both treated and untreated with cabergoline.Each subgroup consisted of ve mice.
Longitudinal tumor susceptibility: This segment tracked breast tumor development in female Brca1/P53de cient mice, categorized into various groups based on parity and cabergoline treatment, totaling 169 subjects, segmented into 35 nulliparous, 32 multiparous, 37 cabergoline-treated multiparous, 37 cabergoline-treated nulliparous, and 28 bi-weekly treated nulliparous mice.We monitored them weekly for tumor emergence, applying humane euthanasia protocols for those meeting prede ned criteria.
Cross-sectional mammary assessment: We analyzed mammary gland status in both Brca1/P53-de cient and wild-type mice, post-lactational involution (30 and 60 days), and in age-matched nulliparous counterparts, with and without cabergoline treatment.
-Genotyping procedures: Utilizing PCR, we identi ed the RFP+ transgene and Brca/P53 deletion in tail DNA samples.The PCR protocol involved speci c reagents, cycling conditions, and oligonucleotide sequences previously reported (45,97), with nal products analyzed on 1% agarose gels.
-Cabergoline administration: Mimicking human dosage for lactation inhibition (109,110), cabergoline was prepared in a saline-methylcellulose solution and administered intraperitoneally at 0.25 mg/kg.Dosage adjustments were based on individual weights, with administration under iso uorane anesthesia.
-Tumor susceptibility monitoring: Weekly palpation sessions were conducted to detect primary mammary tumors, recording both tumor latency and incidence.Euthanasia was carried out upon signs of sickness, rapid tumor growth, or wound development, ensuring adherence to ethical guidelines.

Histological and immunohistochemical analyses
-Histological analysis of mammary glands: Mammary glands were xed in 4% paraformaldehyde (PFA) and processed for para n embedding (Shandon Excelsior, Thermo).Tissue sections were stained with hematoxylin and eosin for morphological examination.Quantitative analysis of adipocyte and ductal areas was conducted on ve randomly selected images at 10x magni cation per sample using a Leica ICC50 HD camera and Leica Application Suite V3.7 software.ImageJ was employed to determine the relative ductal epithelial area as a percentage of the total eld area.
-Quanti cation of caspase-3 and Ki-67 positive cells: Five random images of the distal area of each mammary gland were captured at 20X magni cation (Leica ICC50 HD camera, Leica DM750 microscope).The Leica Application Suite V3.7 was used to identify and count positive and negative cells.
The proportion of positive cells was calculated as the number of positive cells divided by the total number of cells, multiplied by 100.

Immunohistochemical detection of RFP
For the detection of red uorescent protein (RFP), tissue sections (3 µm) were prepared on Superfrost® Plus slides and air-dried overnight.Immunohistochemistry was conducted using the automated Discovery ULTRA system (Ventana-Roche).Antigen retrieval was achieved using CC1 buffer (Ventana, Roche), and endogenous peroxidase activity was blocked using 3% hydrogen peroxide.

Organoid generation from mouse mammary glands
Organoids enriched in epithelial cells were generated following the protocol by Smalley (111).Mammary glands were mechanically fragmented and treated with DMEM/F-12, fetal bovine serum, and antibiotics, supplemented with collagenase (#C2674, Sigma-Aldrich) and trypsin (#15090046, Thermo Fisher Scienti c).Post-incubation, the epithelial cell-enriched pellet was washed with PBS containing fetal bovine serum and DNase (#DN25, Sigma-Aldrich), and subsequently aliquoted and stored at -80°C for RNA and DNA extractions.

Protein analysis in organoid samples
Protein extraction from organoids used RIPA buffer, with protease and phosphatase inhibitor cocktails (#P8340, Sigma Aldrich), was carried out as previously (28).Post-extraction, protein concentration was determined using the Bradford Protein Assay (#5000006, Bio-Rad).Electrophoretic separation utilized SDS-PAGE on 10% or 12% gels, followed by transfer to PVDF membranes (Immobilon-P, Millipore).The membranes were probed with primary antibodies: pSTAT5 (1:1000, #9351, Cell Signaling), STAT5 RNA extraction and library preparation for sequencing RNA was extracted from organoids using the Qiagen miRNeasy Mini Kit.Quality and integrity assessments involved agarose gel electrophoresis, Nanodrop, and an Agilent 5400 bioanalyzer.Library construction for RNA sequencing was conducted by Novogene, including rRNA removal, RNA fragmentation, cDNA synthesis, and adapter ligation, with sequencing on the Illumina NovaSeq 6000 PE150.Data is available under GSE250534 in the Gene Ontology database.

Statistical analysis of RNA-seq data
Analysis of RNA-Seq data by the Bioinformatics Department at Salamanca Cancer Research Center included quality control with RaNA-seq (112), gene expression quanti cation using Salmon (v0.9.1) (113), differential expression analysis with edgeR (v3.24) (114), and over-representation analysis using clusterPro ler (v4.6.0)(115,116).Gene set similarities were assessed using the Jaccard similarity index, and K-means clustering was applied for data categorization, using the elbow method for optimization.
-Normalization for ductal area: Each epithelial cell population's percentage was normalized against the mean ductal area speci c to each mouse, calculated from ve randomly selected elds.The normalized value for each population was computed by multiplying its percentage by the mean ductal area percentage and dividing by 100.
Epidemiological study design and data analysis in SIDIAP database -Study design and cohort selection: We conducted a retrospective cohort study using the SIDIAP (Information System for the Improvement of Research in Primary Care) (Catalunya, Spain) database, encompassing over 80% of Catalonia's population from 2005 onwards, to evaluate the impact of cabergoline on breast cancer prevention (100,101).The study, authorized by the SIDIAP Scienti c and Ethics Committees, included women who had pregnancies lasting between 35 and 50 weeks.This range was chosen based on existing evidence of pregnancy's protective effect against breast cancer starting at 35 weeks (117).
-Inclusion criteria and data collection: The cohort comprised women with detailed pregnancy records, including dates of the last menstrual period and pregnancy outcomes.From the 123,422 eligible pregnancies, 5,472 women who took cabergoline during or immediately after pregnancy formed the treatment group, while 102,959 women served as controls.Exclusions were made for cabergoline use unrelated to pregnancy and previous breast cancer diagnoses (Extended Data Fig. 7).
-Variables and sociodemographic factors: The study considered various factors, including age, socioeconomic status (MEDEA index), term and premature births, and pregnancy duration.Clinical variables encompassed health conditions, family cancer history, and lifestyle factors such as tobacco use and physical activity (Supplementary Tables S17, S19).NGS, RCC, and ABG were responsible for conducting mouse studies, including supervision, treatment, and histopathological analysis.They prepared and analyzed ow cytometry samples, as well as samples for RNA-seq and immunohistochemistry.They critically reviewed the manuscript.NGS and RCC contributed to the creation and layout of gures.OC conducted statistical and epidemiological studies using human data from the SIDIAP database.MMC participated in the aforementioned statistical and epidemiological studies and was instrumental in interpreting the epidemiological data from SIDIAP.SCL led the protein analysis throughout the study and played a key role in organizing the initial mouse cohorts.She also contributed signi cantly to the critical editing of the manuscript.DAL and JDLR focused on RNA-Seq analysis, including initial data processing, identi cation of differentially expressed genes, and creation of global heatmaps.JP expertly handled sample analysis using ow cytometry.AO supervised JP's cytometric work, contributed to designing the ow cytometry experiments, and critically edited the manuscript.LVJ and CPA were involved in designing the epidemiological project for SIDIAP approval.DP oversaw the development of immunohistochemistry protocols in ABG's experiments on non-tumoral breast tissue research and proposed the initial quanti cation protocols for recombinant cells in normal breast tissue.ARV provided supervision, mentorship, and logistical support for DP and ABG experiments done in his lab, and contributed to the critical editing of the manuscript.JHM and HC supervised statistical analyses and the generation of graphics from RNA-seq data, and critically edited the manuscript.MME also participated in supervising these statistical analyses and graphic generation.PGG, EC, and IP conducted immunohistochemistry and immuno uorescence studies for RFP detection.MBGC and FJGC provided logistical support for managing mouse cohorts and developed the initial protocol for breast gland surgery.JPL was involved in the global project design and supervision, resource acquisition, and assistance in generating and studying initial mouse cohorts.He also contributed to the design of the epidemiological study and the writing of the manuscript.Gene expression dynamics in cabergoline-treated Brca1/P53-de cient mice.a) Heatmap analysis: compares gene expression in cabergoline-treated (N = 5) versus untreated (N = 4) multiparous mice's mammary glands.Focuses on genes associated with proliferation and development.b) Gene network visualization: maps differentially expressed genes linked to proliferation and development, highlighting underexpressed and overexpressed genes in cabergoline-treated mice using the Kamada-Kawai layout.c) Venn diagram analysis: depicts the overlap and unique DEGs between multiparous (treated and untreated) and nulliparous mice, focusing on cabergoline-speci c gene expression changes (purple section).d, e) Targeted gene expression analysis: heatmap and network analysis of genes uniquely altered in cabergoline-treated mice, emphasizing their potential roles in reducing breast cancer risk through altered mammary gland development and proliferation.

Figure 1 Impact
Figure 1

Figure 7 Evaluation
Figure 7