Tailoring Oncofertility to Breast Cancer Subtype: A Systematic Review of Fertility Preservation Strategies in Premenopausal Women

This systematic review was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) and the PRISMA-P guidelines for protocols [ 16 ]. Eligible studies included women with breast cancer who were younger than 40 years and/or explicitly described as premenopausal or adolescent and young adult. Breast cancer needed to be stratified by type, including ER+/PR+ (hormone receptor-positive), HER2-positive, triple-negative breast cancer (TNBC), or BRCA1/2 mutation carriers. The intervention of interest was any fertility preservation method, initiated before or during systemic therapy, including oocyte and/or embryo cryopreservation, ovarian tissue cryopreservation, GnRH agonists administered during chemotherapy, and controlled ovarian stimulation (COS), with or without protocol modifications. Eligible study designs included randomized controlled trials, cohort studies, and case–control studies. Studies were excluded if (1) they included male patients or any animal or preclinical laboratory studies, (2) did not specify the subtype, or reported results from non-cancer populations, (3) did not report on fertility outcomes, (4) included patients with significant reproductive or endocrine comorbidities that independently affect ovarian reserve or fertility, such as PCOS, endometriosis, and uncontrolled diabetes. Ineligible study designs included case reports/series, reviews, editorials, conference abstracts without full-text articles, and opinion pieces, as well as studies older than 20 years and not published in English. Primary fertility outcomes were post-treatment ovarian reserve measures (Anti-Müllerian hormone [AMH], follicle-stimulating hormone [FSH], and antral follicle count [AFC]) relative to baseline. Secondary outcomes included FP uptake and completion rates, ovarian stimulation response (e.g., oocyte/embryo yield, cycle cancellation), FP-related adverse events, fertility outcomes specifically in BRCA carriers, and, when available, pregnancy and live birth. A comprehensive literature search was conducted across 3 major databases, including Web of Science, Scopus, and PubMed, restricted to English-language studies published over the last 20 years (2004–2024). Search strategies were customized to the indexing system and syntax of each database. The search was restricted to peer-reviewed literature indexed in these three databases; gray literature, conference abstracts without full-text availability, and clinical trial registries were not separately searched, and this is acknowledged as a limitation (see Section 4.7 ). Manual reference list screening of all included full-text articles was performed as a supplementary step to identify any relevant studies not captured by the database search. The following search formulas were applied: PubMed: (“Breast Neoplasms”[MeSH] OR “breast cancer” OR “breast carcinoma”) AND (“Fertility Preservation”[MeSH] OR “fertility preservation” OR “oocyte cryopreservation” OR “embryo cryopreservation” OR “ovarian tissue cryopreservation” OR “GnRH agonist*” OR “gonadotropin-releasing hormone agonist*”) AND (“Pregnancy”[MeSH] OR “Live Birth”[MeSH] OR “Fertility”[MeSH] OR “ovarian reserve” OR “AMH” OR “Anti-Müllerian Hormone” OR “antral follicle count” OR “recurrence” OR “survival”) AND (“ER-positive” OR “PR-positive” OR “HER2-positive” OR “triple-negative” OR “TNBC”) AND (“young women” OR “premenopausal” OR “adolescent and young adult”[MeSH]). Scopus: (TITLE-ABS-KEY (“fertility preservation” OR “oocyte cryopreservation” OR “embryo cryopreservation” OR “ovarian tissue cryopreservation” OR “GnRH agonists” OR “in vitro maturation”) AND TITLE-ABS-KEY (“breast cancer” OR “triple-negative” OR “HER2 positive” OR “BRCA mutation” OR “ER positive” OR “PR positive”) AND TITLE-ABS-KEY (young OR premenopausal OR “reproductive age”)) AND PUBYEAR > 1999 AND PUBYEAR < 2026 AND (LIMIT-TO (DOCTYPE, “ar”) OR LIMIT-TO (DOCTYPE, “re”)) AND (LIMIT-TO (LANGUAGE, “English”)). Web of Science: ([mh “Breast Neoplasms”] OR “breast cancer” OR “breast carcinoma”) AND ([mh “Fertility Preservation”] OR “fertility preservation” OR “oocyte cryopreservation” OR “embryo cryopreservation” OR “ovarian tissue cryopreservation” OR “GnRH agonist*” OR “gonadotropin-releasing hormone agonist*”) AND ([mh “Pregnancy”] OR [mh “Live Birth”] OR “fertility” OR “ovarian reserve” OR “AMH” OR “Anti-Müllerian Hormone” OR “antral follicle count” OR “recurrence” OR “survival”) AND (“ER-positive” OR “PR-positive” OR “HER2-positive” OR “triple-negative” OR “TNBC”) AND (“young women” OR “premenopausal” OR “adolescent and young adult”). All records identified in the database searches were imported into Rayyan AI (Qatar Computing Research Institute, https://www.rayyan.ai , accessed on 7 June 2026), a reference management software, for initial organization and duplicate removal. A two-stage approach was followed: in the first stage, two independent reviewers screened the titles and abstracts of all records against the eligibility criteria. Records deemed potentially relevant by either reviewer were forwarded for full-text assessment. In the second stage, full-text articles were independently evaluated to ensure complete relevance. Any disagreements between the two primary reviewers at either stage were resolved through discussion, and if consensus could not be reached, a third reviewer would make the final decision. The search results were documented and summarized in a PRISMA flow diagram. Data extraction was performed by two independent reviewers using a standardized extraction template to report details and variables needed for subtype-stratified analysis. It captured publication details, sample size, population characteristics, breast cancer subtype, fertility preservation intervention, primary and secondary fertility outcome measures, oncological outcome measures, and key findings. When data were missing or unreported, the reviewers noted ‘NR’ in the table. Risk of bias assessment was conducted at the study level, independently by two reviewers, using validated tools tailored to specific study designs. Randomized controlled trials were assessed using the Cochrane Risk of Bias tool version 2 (RoB 2), which evaluates domains including the randomization process, deviations from intended interventions, missing outcome data, outcome measurement, and the selection of reported results. For non-randomized observational studies (cohort and case–control designs), they were assessed using the Risk Of Bias In Non-randomized Studies of Interventions tool version 1 (ROBINS-I), which evaluates bias in relation to a hypothetical target randomized trial across seven domains: bias due to confounding, selection of participants, classification of interventions, deviations from intended interventions, missing data, measurement of outcomes, and selection of the reported result. The assessment focused on the representativeness of breast cancer cohorts, the reliable and valid use of statistical controls, and the reliability of the outcome measures. The assessment results were used to interpret the strength of the evidence and incorporated into the final qualitative synthesis. Given the expected clinical and methodological heterogeneity across studies, we planned a primarily narrative synthesis. Formal meta-analysis was not performed because the conditions required for valid quantitative pooling were not met: (1) clinical heterogeneity—stimulation protocols, patient age, chemotherapy regimens, and endocrine therapy varied substantially; (2) methodological heterogeneity—14 of 19 studies were retrospective cohorts with differing designs and comparison groups; and (3) outcome heterogeneity—ovarian reserve was measured using non-standardized combinations of AMH, FSH, and AFC at varying post-treatment timepoints, and live birth rates were reported in only a minority of studies. These same factors precluded quantitative subgroup analyses by breast cancer subtype; fewer than three studies per subtype reported comparable outcomes amenable to pooling. All evidence synthesis is therefore narrative. Studies were grouped by FP modality (COS with or without letrozole/tamoxifen; GnRH agonists; IVM/OTC) and, where data allowed, by tumor subtype. Subgroup analyses by tumor subtype and FP modality were planned but were ultimately not feasible due to the heterogeneity described above. Each included study represented a unique, non-overlapping patient cohort; no instances of overlapping populations from the same institution were identified. Given the absence of prespecified quantitative thresholds applicable to all included study designs, subtype-specific effectiveness was assessed narratively across three domains: ovarian stimulation response, post-treatment ovarian reserve, and, where reported, live birth rates. Confidence in subtype-specific conclusions was graded qualitatively based on the number, consistency, and methodological quality of contributing studies. Evidence was considered more robust for HR+ patients (the largest number of consistent studies) and more limited for TNBC, HER2+, and BRCA1-specific subgroups.

Breast cancer is the most prevalent malignancy among women worldwide. Annually, around 2.3 million new cases occur, with considerable regional and socioeconomic variation in incidence, mortality, and survival rates [ 1 , 2 ]. The incidence rates among individuals of reproductive age are steadily rising, exacerbating challenges with fertility preservation and long-term quality of life [ 3 ]. Multimodal therapeutic advances, including neoadjuvant chemotherapy, targeted therapies, CDK4/6 inhibitors, immunotherapy, and precision oncology, have improved 5-year survival rates to over 90%, markedly increasing the number of young survivors [ 4 , 5 ]. However, these therapies (particularly gonadotoxic chemotherapy with alkylating agents and extended endocrine therapy) threaten ovarian reserve and future fertility, making preservation strategies an urgent clinical imperative [ 6 , 7 ]. For reproductive-age survivors, fertility preservation has become crucial for maintaining quality of life, addressing profound psychosocial distress caused by treatment-induced infertility and premature ovarian insufficiency [ 6 ]. To mitigate the gonadotoxic effects of breast cancer therapies, established fertility preservation strategies include embryo and oocyte cryopreservation, which typically require 10–14 days of controlled ovarian stimulation (COS) with gonadotropins, often combined with aromatase inhibitors like letrozole to minimize estrogen exposure in hormone receptor-positive disease [ 8 ]. Other options include ovarian tissue cryopreservation, a surgical technique that enables immediate treatment initiation without a delay for stimulation. Investigational and adjunctive approaches, such as in vitro maturation of oocytes, ovarian transposition, and modified stimulation protocols, can expand options for patients requiring urgent oncologic treatment or those with contraindications to standard methods [ 9 , 10 ]. The choice of fertility preservation should be personalized, primarily considering the patient’s age, urgency, the planned treatment’s gonadotoxicity, the nature of the disease, and the time available before therapy starts [ 8 ]. Both ASCO and ESMO guidelines recommend urgent, multidisciplinary referral to a fertility specialist and individualized counseling before gonadotoxic therapy begins [ 11 , 12 ]. However, it is important to note the distinction between gamete/embryo cryopreservation, which directly preserves reproductive material, and GnRH agonist use, which protects ovarian function but does not constitute an equivalent FP method; evidence supporting the latter, particularly for live birth outcomes, remains limited. Breast cancer subtypes differ substantially in biology, prognosis, and treatment. Hormone receptor–positive (ER+/PR+) tumors are typically luminal, often indolent but prone to late recurrences, and managed with long-course endocrine therapy (5–10 years) and CDK4/6 inhibitors in selected high-risk cases [ 13 ]. HER2-positive disease is more aggressive but now highly treatable with chemotherapy combined with HER2-targeted therapy, most commonly trastuzumab with or without pertuzumab and standard anthracycline/taxane-based backbones [ 13 ]. Triple-negative breast cancer (TNBC) often presents with high-grade, basal-like biology and an early peak in relapse risk, treated with anthracycline–taxane regimens, frequently incorporating platinum and immunotherapy in the neo/adjuvant setting [ 13 , 14 ]. Germline BRCA-associated breast cancer carries a higher lifetime risk and a younger age of onset; systemic therapy commonly includes platinum agents and PARP inhibitors [ 14 , 15 ]. These differences necessitate tailored FP strategies: for ER+/PR+ disease, tamoxifen/letrozole-based COS protocols minimize estrogen exposure, while extended endocrine therapy delays childbearing; HER2+ patients can typically undergo COS before neoadjuvant therapy, with pregnancy safety after trastuzumab washout (approximately 6 months) (Hong et al., 2023) [ 8 ]; TNBC’s aggressive biology and high chemotherapeutic intensity often favor ovarian tissue cryopreservation over time-intensive COS [ 8 ]; and BRCA1/2-associated cancers present unique challenges, including potentially diminished ovarian reserve, timing of risk-reducing salpingo-oophorectomy, and theoretical malignancy risks with ovarian tissue reimplantation. Despite this, existing FP guidelines remain largely uniform and do not provide subtype-specific guidance on uptake rates, stimulation outcomes, reproductive success, or oncologic safety—a critical knowledge gap this review addresses. This review is, to our knowledge, the first systematic review to use tumor subtype stratification as the primary analytical framework, explicitly comparing FP outcomes across HR+, HER2+, TNBC, and BRCA-associated subtypes, and examining how subtype-specific biology and treatment context modulate FP outcomes. Prior systematic reviews have addressed FP safety or efficacy in breast cancer broadly without this subtype-stratified focus. Specifically, this review aims to: (1) compare FP methods across breast cancer subtypes; (2) identify subtype-specific challenges; and (3) propose pathways toward precision oncofertility care.

A total of 2453 records were initially identified: PubMed ( n = 109), Scopus ( n = 759), and Web of Science ( n = 1585). After removing 616 duplicates, 1837 records were screened. Of these, 1721 were excluded. From the remaining 116 articles assessed in full text, 97 were excluded (wrong design n = 55, no fertility outcomes n = 22, no subtype stratification n = 12, non-breast cancer n = 3, irrelevant n = 3, male-only n = 1, animal study n = 1). Ultimately, 19 studies met inclusion criteria ( Figure 1 ). The final studies included 2 randomized controlled trials and 17 cohort studies from Europe ( n = 8), North America ( n = 6), Asia ( n = 3), and multicenter/international cohorts ( n = 2). Most were retrospective ( n = 14) or prospective ( n = 3) cohorts. Interventions comprised COS with letrozole/tamoxifen-modified protocols ( n = 12), GnRH agonists during chemotherapy ( n = 4), and IVM/OTC ( n = 3). Thirteen of 19 studies stratified outcomes by tumor biology or treatment context; this subtype-stratified evidence base is the critical focus of this review. HR+ and BRCA-associated disease were the best-represented subtypes, while TNBC and HER2+ were underrepresented, with fewer than five studies reporting subtype-specific outcomes for these groups—representing a key evidence gap. Premenopausal status was confirmed by self-report, menstrual history, or hormonal assessment (FSH/estradiol) in most studies; some cohorts included women up to age 44–45 who may have had limited ovarian reserve, introducing potential heterogeneity in baseline fertility status. All studies evaluated fertility preservation strategies in premenopausal women diagnosed with breast cancer. The characteristics of all included studies are summarized in Table 1 . Among the 2 randomized controlled trials, both demonstrated low risk of bias across all Cochrane RoB 2 domains. Among the 17 non-randomized studies assessed with ROBINS-I, 11 received an overall low risk of bias rating, and 6 received a moderate rating [ 20 , 23 , 25 , 27 , 29 , 30 ], primarily in Domain 1 (bias due to confounding), reflecting insufficiently controlled variation in age, baseline ovarian reserve, disease stage, and treatment intensity. These figures are summarized in Figure 2 and Figure 3 . It should be noted that moderate confounding in observational cohorts (for example, healthier patients with better baseline ovarian reserve being preferentially selected for GnRH agonist co-administration) may bias associations toward apparent benefit for ovarian protection endpoints. This interpretive point is discussed further in Section 4.1 . In HR+ patients undergoing COS with letrozole or tamoxifen (the subgroup with the most consistent evidence) FP was feasible and produced acceptable ovarian stimulation responses. Evidence of feasibility was weaker for TNBC and HER2+ patients, where fewer subtype-specific data are available. Notably, controlled ovarian stimulation protocols using letrozole or tamoxifen were consistently associated with lower estradiol exposure, with most data relating to ovarian response endpoints; robust data on live births and long-term oncologic outcomes following modified COS remain limited. Among BRCA mutation carriers, several studies found outcomes comparable to non-carriers, although results are inconsistent between studies and between BRCA1 and BRCA2 carriers, with small sample sizes limiting firm subtype-specific conclusions. Some studies, particularly [ 21 ], reported lower mature oocyte yields and AMH levels in BRCA1-positive patients specifically. Studies assessing GnRH agonists during chemotherapy suggested a protective effect on ovarian function, with lower ovarian failure rates and better post-treatment menstrual or AMH recovery. However, GnRH agonist use should be understood as ovarian function preservation (reducing chemotherapy-induced ovarian failure) rather than as an equivalent to oocyte or embryo cryopreservation, which directly preserves gametes. Pregnancy and live birth data for GnRHa alone remain limited across the included studies. In terms of long-term follow-up, data across included studies with medium-term oncologic follow-up (predominantly 3–5.5 years) showed no signal of increased recurrence or mortality attributable to FP. This should be interpreted as reassuring medium-term safety, not a definitive oncologic guarantee, given that most studies were underpowered for rare or late events and follow-up beyond 5 years is sparse. Overall, the evidence suggests that FP is feasible across breast cancer subtypes, but conclusions regarding comparative effectiveness and long-term reproductive outcomes remain constrained, particularly for TNBC, HER2+, and BRCA1-specific subgroups.

This systematic review synthesized the findings of 19 studies examining fertility preservation outcomes in women with breast cancer. The findings suggest that FP is procedurally feasible in young women with breast cancer across most subtypes, with ovarian stimulation responses generally within acceptable ranges in the subgroups where evidence is most robust—primarily HR+ patients. These conclusions must be interpreted within the context of a predominantly retrospective evidence base, heterogeneous outcome definitions, and follow-up periods that are insufficient to exclude late oncologic risks or to characterize long-term reproductive success. Substantial heterogeneity in study design, patient populations, tumor characteristics, and outcome reporting limited direct comparison across studies and precluded quantitative pooling. A key method investigated was ovarian function preservation through the use of gonadotropin-releasing hormone (GnRH) agonists during chemotherapy. Overall, GnRHa use in the included studies was associated with a reduction in chemotherapy-induced ovarian failure and improved ovarian recovery [ 19 , 31 , 32 ]. For example, a randomized controlled trial ( n = 98) reported that the rate of one-year ovarian failure was significantly reduced from 80.6% in the chemotherapy-alone group to 44.7% in the GnRHa group ( p = 0.002), with benefit observed across HR+ and HR− subgroups, though statistical significance was reached only in the HR+ subgroup ( p = 0.029); the HR− subgroup showed a numerical trend that did not reach significance ( p = 0.111) [ 32 ]. Likewise, another observational study reported the resumption of menses in 97% of patients and anti-Müllerian hormone levels greater than 1 ng/mL in 70% at 12 months [ 31 ]. Across studies, younger age and higher baseline ovarian reserve were associated with improved outcomes, whereas exposure to alkylating chemotherapy was consistently linked to increased gonadotoxicity [ 31 , 33 , 35 ]. Notably, the effect of tumor subtype on ovarian recovery was less consistent across investigations [ 27 ]. Importantly, GnRH agonist co-administration during chemotherapy should be understood as ovarian function preservation—reducing the rate of chemotherapy-induced ovarian failure and supporting menstrual recovery—rather than as an equivalent to oocyte or embryo cryopreservation, which directly preserve gametes for future use. This distinction is clinically important: women who receive only GnRHa during chemotherapy retain no cryopreserved material and rely entirely on spontaneous ovarian recovery to conceive. Pregnancy and live birth data specifically attributable to GnRHa use (as opposed to spontaneous recovery) remain limited in the included studies. The interpretive caution noted in Section 3.3 regarding confounding also applies here: healthier patients with better baseline reserve may be preferentially selected for GnRHa, potentially inflating apparent effectiveness. Further, the majority of included studies evaluated controlled ovarian stimulation for oocyte or embryo cryopreservation. These studies consistently confirmed that fertility preservation is feasible prior to chemotherapy, with high procedural success rates. A procedure success rate of 89.5% has been reported, with a mean retrieval of 12.8 oocytes and 9.8 mature oocytes vitrified [ 17 ]. Results on oocyte yield with modified protocols are not entirely consistent. Most studies, including two randomized controlled trials [ 17 , 22 ], found no statistically significant difference in mature oocyte yield between letrozole- or tamoxifen-modified COS and standard gonadotropin stimulation. However, Revelli A et al. [ 23 ] reported 1–2 fewer oocytes per cycle with letrozole-gonadotropin in HR+ patients (6.6 ± 3.5 vs. 8.0 ± 5.0 oocytes available for cryostorage; p = 0.038) compared with gonadotropin alone in ER-negative patients, suggesting a possible modest trade-off in yield. This comparison involved different patient subgroups and was not a protocol-matched randomized analysis; the observed difference should therefore be interpreted cautiously. These protocols are associated with markedly lower peak estradiol levels, supporting their use in HR+ disease. The clinical significance of any yield reduction and its impact on cumulative live birth probability remain unstudied. Furthermore, most data relate to ovarian response endpoints; evidence on live births and long-term oncologic outcomes following modified COS is limited, and conclusions about true effectiveness for childbearing must be tempered accordingly. The oncologic safety of COS in HR+ patients is supported by several lines of evidence. Vriens et al. reported no compromise of oncologic outcomes at median 52-month follow-up in a prospective cohort that included a majority of HR+ patients [ 29 ]. Azim Jr et al., in a secondary analysis of the POSITIVE trial ( n = 516, all HR+), found no significant increase in 3-year recurrence risk associated with prior ovarian stimulation at diagnosis [ 30 ]. Dezellus et al. reported 5-year disease-free survival of 82% and overall survival of 90% in their tamoxifen-COS cohort [ 17 ]. The mechanistic basis for safety is supported by letrozole or tamoxifen co-administration during stimulation, which suppresses peak estradiol, mitigating the theoretical concern of estrogen exposure in ER+ disease. These data are reassuring but should be interpreted as medium-term safety signals rather than definitive long-term guarantees. In terms of genetic risk, the effect of BRCA mutation status on fertility preservation outcomes was evaluated across 11 studies, with generally consistent findings showing that ovarian reserve and stimulation response were comparable between mutation carriers and non-carriers [ 19 , 24 , 26 , 34 ]. However, results are inconsistent across studies, and a possible differential impact for BRCA1 vs. BRCA2 carriers cannot be resolved with current data. Some studies reported lower mature oocyte yield and reduced maturation rates among BRCA1 mutation carriers specifically [ 21 , 25 ], while others found no significant differences [ 20 , 25 ]. Moujahed et al. [ 25 ]—the largest study assessing ovarian response to COS in BRCA carriers—found no statistically significant difference in ovarian reserve markers or response between BRCA1 and BRCA2 subgroups, consistent with Gunnala et al. [ 20 ]. In contrast, Porcu et al. [ 21 ] reported lower serum AMH and fewer mature oocytes specifically in BRCA1-positive patients. In another study, lower maturation rates (78.6% vs. 85.7%) and fewer mature oocytes were observed in patients with BRCA mutations [ 25 ] (El Moujahed et al., 2023). However, some studies reported lower mature oocyte yield and reduced maturation rates among BRCA1 mutation carriers [ 18 , 21 , 25 , 28 ]. Overall, these results suggest that while BRCA1 mutations may be associated with measurable deficits in ovarian reserve and oocyte quality, it remains unclear whether these factors significantly reduce the clinical success rates of fertility preservation procedures. The mechanistic basis for potential BRCA1-specific impairment merits elaboration. BRCA1 and BRCA2 proteins play essential roles in homologous recombination repair of double-strand DNA breaks. During oocyte meiosis (particularly prophase I), DNA strand breaks are physiologically generated and require efficient repair; defective homologous recombination in BRCA-mutant oocytes has been hypothesized to trigger accelerated apoptosis and primordial follicle depletion [ 25 ] (Moujahed et al., 2023). Consistent with this, a meta-analytic dataset cited by Moujahed et al. suggests AMH levels are approximately 1.0 ng/mL lower in BRCA-mutated patients undergoing FP for breast cancer compared to non-carriers [ 25 ]. Additionally, BRCA1 carriers more frequently present with TNBC and high-grade tumors, and tumor aggressivity itself may independently impair ovarian response—making it difficult to disentangle the respective contributions of germline mutation and tumor biology to observed yield differences. These findings should be interpreted as ‘no large, clearly established impairment of overall FP effectiveness attributable to BRCA mutation status’ rather than confirmation of fully preserved effectiveness. BRCA1 mutation carriers may warrant individualized counseling about potential variability in oocyte maturity and ovarian response, and earlier or repeated FP attempts may be advisable in selected cases. Turning attention to alternative techniques, such as in vitro fertilization, in vitro maturation, and ovarian tissue cryopreservation, IVM was identified as an alternative fertility preservation strategy, particularly for patients requiring urgent initiation of chemotherapy and therefore unable to undergo conventional ovarian stimulation. Among these approaches, Raad et al. reported a maturation rate of approximately 58%, with a mean of 5.8 mature oocytes vitrified per cycle, indicating lower efficiency than standard protocols [ 35 ]. Despite its feasibility, these cohorts were impacted by both patient and disease-related factors. Specifically, reduced ovarian reserve, as indicated by anti-Müllerian hormone levels below 1.5 ng/mL and a low antral follicle count, was associated with lower maturation rates [ 31 , 32 , 35 ]. Additionally, Raad et al. suggested that more aggressive tumor subtypes, including HER2-positive disease, high tumor grade, and triple-negative breast cancer, were linked to reduced success. IVM and OTC should be presented to patients as approaches that remain relatively experimental in the breast cancer setting, with lower established oocyte competence than conventional COS and with essentially no live birth outcome data available in this specific population. IVM is best considered in combination with OTC to maximize cumulative fertility potential, though this combined approach also lacks long-term efficacy data and should be framed accordingly in patient counseling. The observation by Raad et al. that more aggressive tumor subtypes (including TNBC, high-grade disease, and HER2-positive) may be associated with reduced IVM success is hypothesis-generating and derived from a single retrospective study with small subgroup numbers [ 35 ]. It should not be interpreted as a causal relationship or a contraindication to IVM in these subtypes, but rather as a signal warranting prospective investigation. Turning now to post-treatment fertility outcomes, reporting was sparse: only 6 studies provided long-term follow-up data, representing less than 20% of the overall cohorts. Among these, assisted reproductive outcomes were generally favorable. In a large cohort of hormone receptor-positive patients ( n = 518), the use of cryopreserved embryos was associated with a significant increase in pregnancy rates, with embryo transfer doubling the odds of conception (odds ratio 2.41) [ 30 ]. Vriens et al. reported a 5-year live birth rate of 27% (95% CI 17–38%) and a 5-year ovarian function recovery rate of 92% in their prospective cohort [ 29 ]. Live birth outcomes following controlled ovarian stimulation were also encouraging, with reported rates comparable between BRCA mutation carriers and non-carriers [ 35 ]. Nonetheless, utilization of cryopreserved material remained low, with return-to-use rates ranging from 5% to 10%. Notably, spontaneous pregnancies were also reported, occurring in approximately 22.1% of patients at five years in one cohort utilizing tamoxifen-based stimulation protocols [ 17 ]. Regarding treatment delay, in the cohort reported by Zhong et al. [ 32 ], FP was associated with a chemotherapy initiation delay of less than one week; this finding should be interpreted within the context of that specific study design rather than as a universal estimate applicable to all FP modalities. Regarding oncologic safety, no signal of increased recurrence was detected in any included study over the available follow-up periods (predominantly 3–5.5 years); this should be interpreted as ‘no detectable signal in medium-term observational and trial data’ rather than a definitive safety guarantee, given that most studies were underpowered for rare or late oncologic events. No increase in recurrence was observed after COS/FP (including HR+ cases) [ 29 , 30 ]. Overall, no clear superiority of one fertility preservation modality over another could be established, largely due to heterogeneity in follow-up duration, utilization of assisted reproductive technology, and reporting of partner status. Furthermore, key reproductive outcomes such as time to pregnancy and cumulative pregnancy rates were infrequently reported, further challenging comparison between modalities. These findings support the early, routine integration of oncofertility counseling into the multidisciplinary breast cancer care pathway for all premenopausal women. Fertility preservation (FP) should be discussed before systemic therapy begins, as evidence consistently indicates that FP does not substantially delay cancer treatment and is not associated with increased oncologic recurrence or mortality, even in aggressive subtypes. Based on currently available medium-term data, FP has not been associated with a statistically significant increase in recurrence or mortality in the included studies; this conclusion is most applicable to HR+ and BRCA-associated populations with 3–5 years of follow-up, and evidence for TNBC and HER2+ subtypes remains more limited. Tumor subtype should additionally inform counseling on expected outcomes: women with triple-negative breast cancer and high-grade disease may have lower baseline ovarian reserve and yield fewer oocytes regardless of the stimulation method employed, and this information should be used to set realistic expectations and prioritize earlier referral [ 35 ]. For HR+ patients, letrozole-based controlled ovarian stimulation (COS) is a preferred protocol to lessen potential risks associated with peak estradiol exposure [ 36 ]. Where letrozole is unavailable or not tolerated, tamoxifen-gonadotropin represents a well-supported alternative, having demonstrated equivalent mature oocyte yields in a randomized controlled trial and a large prospective cohort [ 17 , 22 ]. The adoption of random-start GnRH antagonist protocols is further recommended in practice, as oocyte yield does not differ significantly by menstrual cycle phase at stimulation onset [ 18 , 26 ]. Furthermore, BRCA1 mutation carriers may warrant individualized counseling regarding potential variations in oocyte maturity and ovarian response, and earlier or repeated FP attempts may be advisable in selected cases [ 21 , 37 ]. When COS is contraindicated or chemotherapy is critically urgent, in vitro maturation (IVM) of immature oocytes represents a viable alternative initiable within days of diagnosis; given the lower competence of IVM-derived oocytes relative to those from conventional COS, IVM is best considered in combination with ovarian tissue cryopreservation (OTC) to maximize cumulative fertility potential [ 34 , 35 ]. In settings where established FP services are unavailable, administration of a gonadotropin-releasing hormone agonist (GnRHa) may be considered to preserve ovarian function, particularly in hormone receptor-negative disease, though it should be clearly distinguished from gamete or embryo cryopreservation [ 38 ]. This review has several methodological limitations that should be considered when interpreting its findings. Regarding the search strategy, restricting the search to three databases and English-language publications may have introduced selection and language bias, potentially omitting relevant studies indexed in regional databases or published in other languages, including those from regions where breast cancer is highly prevalent. The exclusion of studies older than 20 years, while justified by the rapid evolution of FP techniques and treatment protocols, may have excluded early cohort data relevant to long-term reproductive outcomes. This review was not prospectively registered. At the study level, 14 of 19 included studies were retrospective in design, which carries an inherent risk of selection bias and unmeasured confounding. Women who pursued FP may differ systematically from those who did not in disease severity, socioeconomic status, and personal preferences, thereby limiting generalizability. Although RCTs demonstrated low risk of bias on the Cochrane RoB tool, cohort and non-randomized studies showed moderate risk of bias in domains related to confounding and intervention protocol variability. Outcome definitions were also heterogeneous across studies: ovarian reserve was assessed using AFC, AMH, and FSH at non-standardized time points, and the criteria used to define successful preservation of ovarian function varied substantially. This heterogeneity precluded formal meta-analysis and limits the precision of the review’s conclusions. Finally, long-term reproductive outcomes—including pregnancy rates, cumulative ART success, and live birth rates—were reported in fewer than one-third of included studies, and return-to-use rates for cryopreserved material were low across the literature. The predominance of surrogate ovarian endpoints (AMH, oocyte yield) rather than true fertility endpoints (live births) is a critical limitation that must temper conclusions about FP ‘effectiveness’ throughout this review. These gaps collectively reduce the certainty of the synthesized evidence and limit the ability to draw firm conclusions about the comparative effectiveness of different FP strategies across patient subgroups. Addressing inconsistent outcome reporting is a research imperative. We propose adoption of a minimum core outcome set for future oncofertility research, encompassing at minimum: number of mature oocytes retrieved, live birth rate, time to pregnancy, post-treatment AMH and menstrual recovery, and cancer-free survival. Adherence to CONSORT and STROBE reporting standards, with mandatory subgroup reporting by tumor subtype and the establishment of international prospective registries with standardized data collection, are also recommended. Future research should prioritize prospective, multicenter, subtype-stratified trials powered for live birth as a primary endpoint. Establishing robust evidence on live birth rates, cumulative ART success, and late-term oncologic safety is essential. Furthermore, comparative effectiveness research is required to define optimal FP strategies, such as COS versus ovarian tissue cryopreservation, across specific molecular subtypes. Additionally, research should examine the biological mechanisms underlying reduced oocyte maturation in BRCA1 carriers. Researchers suggested that BRCA1 has a role in meiotic spindle assembly and DNA damage response in oocytes, and that mutation carriers have significantly lower primordial follicle densities and higher rates of DNA double-strand breaks in oocytes [ 39 , 40 ]. These findings can inform strategies to improve gamete quality. A critical and urgent research gap concerns the reproductive safety of newer systemic therapies. CDK4/6 inhibitors target cell-cycle pathways vital for follicular development, raising theoretical concerns about ovarian toxicity that remain unconfirmed by current clinical data [ 41 ]. PARP inhibitors have been shown to cause primordial follicle depletion in preclinical models, but human data are virtually absent. Immune checkpoint inhibitors, such as pembrolizumab, have unknown implications for ovarian reserve. Definitive reassurance for these agents is currently impossible; prospective evaluation of ovarian reserve markers, menstrual function, and fertility outcomes in women exposed to these therapies is urgently required, and FP should ideally be completed before their initiation. Biomarker-guided FP strategies integrating AMH, AFC, and tumor biology to predict stimulation response and individualize protocols represent an important future direction, as does AI-assisted reproductive risk prediction incorporating tumor and treatment characteristics. Additionally, equity-focused research is imperative, as race, ethnicity, socioeconomic status, insurance coverage, and geography significantly limit equitable access to FP. Future studies must include underrepresented populations, pre-specified equity analyses, and prospective testing of structural interventions, including patient navigation and financial support.

This systematic review demonstrates that fertility preservation is procedurally feasible and associated with acceptable ovarian stimulation responses in young women with breast cancer when initiated before systemic therapy, with the strongest evidence in HR+ patients. Letrozole- and tamoxifen-modified COS protocols were associated with acceptable oocyte yields while limiting estradiol exposure. GnRH agonist co-administration during chemotherapy was associated with reduced rates of ovarian failure and improved recovery of ovarian function (most consistently in hormone receptor-negative disease), though it should be understood as ovarian function preservation rather than as an equivalent to gamete cryopreservation. No included study identified a statistically significant increase in cancer recurrence or mortality attributable to FP; however, this reassurance is explicitly limited by predominantly retrospective designs, follow-up durations of 3–5.5 years, and studies that were underpowered to detect rare or late oncologic events. Patient and tumor characteristics (including BRCA mutation status, tumor grade, and hormone receptor profile) influence stimulation outcomes to varying degrees and should inform individualized counseling rather than serve as grounds to withhold FP. Conclusions regarding BRCA1-specific outcomes, TNBC, HER2+, and IVM/OTC remain constrained by small sample sizes, inconsistent results, and the lack of live-birth data. These conclusions are tempered by the predominance of retrospective designs, heterogeneity in outcome definitions, and a paucity of long-term reproductive follow-up data. Future prospective, subtype-stratified studies with standardized outcomes (in particular live birth rates) and longer follow-up are urgently needed. Until such evidence is available, early, multidisciplinary oncofertility counseling remains the cornerstone of care for premenopausal women with breast cancer.