Association between endocrine therapy and survival outcomes in estrogen receptor–low breast cancer: a systematic review and meta-analysis

Abstract

Background

The clinical significance of endocrine therapy (ET) in patients with estrogen receptor–low (ER-low, 1%-10%) breast cancer remains uncertain. Increasing evidence suggests that these tumors differ biologically from typical hormone receptor–positive cancers, raising questions about the true benefit of ET in this subgroup.

Methods

This systematic review and meta-analysis was conducted in accordance with the PRISMA 2020 guidelines. PubMed, Embase, the Cochrane Library, and Web of Science were searched from inception to July 1, 2025. Hazard ratios (HRs) with corresponding 95% CIs for time-to-event outcomes were extracted directly or estimated from Kaplan–Meier curves when not explicitly reported. Pooled analyses were performed using random-effects models to account for between-study heterogeneity. The protocol was registered in PROSPERO (CRD420251055521).

Results

A total of 6 retrospective observational studies involving 8765 patients with estrogen receptor–low breast cancer were included. Endocrine therapy was associated with improved overall survival (OS) (HR 0.81, 95% CI 0.69-0.95; P = .01; I² = 0%) and disease-free survival (DFS) (HR 0.64, 95% CI 0.45-0.90; P = .01; I² = 0%).

Conclusions

Endocrine therapy was associated with improved OS and DFS in patients with estrogen receptor–low breast cancer. These findings suggest that estrogen receptor–low breast cancer should not be uniformly considered endocrine-insensitive and that routine omission of endocrine therapy should be approached cautiously. Given that the available evidence is derived from retrospective observational studies, further prospective studies are needed to identify patients most likely to benefit from endocrine therapy.

Implications for Practice.

Estrogen receptor–low breast cancer is biologically heterogeneous and often overlaps with triple-negative breast cancer, making the value of endocrine therapy uncertain. In this meta-analysis, endocrine therapy was associated with improved overall survival and disease-free survival after harmonization of effect directions across studies. These findings suggest that endocrine therapy should not be routinely omitted solely because estrogen receptor expression is low. Treatment decisions should integrate estrogen receptor percentage, clinicopathologic features, response to neoadjuvant therapy, expected toxicity, and patient preferences. Prospective studies are needed to better define which patients are most likely to benefit.

Introduction

Breast cancer is the most commonly diagnosed cancer among women worldwide and remains a major global health burden.¹ Endocrine therapy is a cornerstone of adjuvant treatment for hormone receptor–positive breast cancer and substantially reduces the risk of recurrence and breast cancer–related mortality by targeting estrogen-driven signaling pathways.^2,3 However, the magnitude of benefit from endocrine therapy is closely related to the level of estrogen receptor expression, which serves as both a predictive and prognostic biomarker in clinical decision-making.^4,5

According to the American Society of Clinical Oncology and the College of American Pathologists (ASCO/CAP) guideline, breast tumors with nuclear estrogen receptor staining in at least 1% of tumor cells are considered estrogen receptor–positive.⁶ This threshold has led to the recognition of a distinct subgroup, commonly referred to as estrogen receptor–low (ER-low) breast cancer, defined as estrogen receptor expression of 1%-10%. Increasing evidence suggests that these tumors differ biologically from typical estrogen receptor–positive cancers, often showing higher histologic grade, higher Ki-67 expression, lower progesterone receptor expression, and more aggressive clinicopathologic features.^7,8 Molecular studies have further suggested partial overlap with triple-negative breast cancer, raising concerns about attenuated endocrine sensitivity in this subgroup.^9,10

Although estrogen receptor–low breast cancer is formally classified as estrogen receptor–positive, the clinical value of endocrine therapy in this subgroup remains uncertain. Some studies have reported favorable survival outcomes among patients who received endocrine therapy,^11–15 whereas others have shown limited or no significant association with improved outcomes.^16–23 These inconsistent findings have contributed to ongoing uncertainty regarding whether endocrine therapy should be routinely recommended, selectively used, or omitted in patients with estrogen receptor–low breast cancer.

Therefore, a systematic synthesis of the available evidence is warranted. We conducted this systematic review and meta-analysis to evaluate the association of endocrine therapy with survival outcomes in patients with estrogen receptor–low breast cancer. We also explored whether this association varied according to estrogen receptor expression level, duration of endocrine therapy, and clinical treatment context.

Methods

Search strategy

This systematic review and meta-analysis was conducted in accordance with the PRISMA 2020 (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines.²⁴ PubMed, Embase, the Cochrane Library, and Web of Science were searched using a combination of free-text terms and Medical Subject Headings (MeSH) related to “breast cancer,” “estrogen receptor,” “ER-low,” “endocrine therapy,” “hormonal therapy,” and relevant ER expression cutoffs (1%-10%). Boolean operators (“AND,” “OR”) were applied as appropriate. Reference lists of included studies and relevant reviews were manually screened to identify additional eligible articles. The detailed search strategies for each database are provided in Table S1.

Eligibility criteria

Studies were eligible if they met the following criteria (1) included patients with primary breast cancer showing estrogen receptor expression of 1%-10% by immunohistochemistry; (2) compared endocrine therapy with no endocrine therapy; and (3) reported time-to-event outcomes, such as disease-free survival (DFS) or overall survival (OS), with hazard ratios (HRs) or sufficient data to estimate HRs. Studies were excluded if they were case reports, review articles, editorials, or conference abstracts; involved metastatic disease at diagnosis or male breast cancer; or lacked sufficient data to estimate HRs for survival outcomes. Two investigators (F.X. and M.X.) independently screened titles and abstracts, followed by full-text review to assess eligibility. Disagreements were resolved through discussion or consultation with a third reviewer (X.D.P.).

Data extraction

Data were independently extracted by 2 investigators (F.X. and M.X.) using a standardized data collection form. Extracted information included first author, publication year, study region, study design, sample size, estrogen receptor definition, endocrine therapy comparison, survival outcomes, reported or estimated HRs with corresponding 95% confidence intervals, and adjustment status. When available, key clinicopathologic and treatment characteristics were also extracted, including age, menopausal status, tumor stage, nodal status, progesterone receptor status, HER2 status, endocrine therapy regimen or duration, chemotherapy, radiotherapy, anti-HER2 therapy, and response to neoadjuvant therapy.

When multiple HRs were reported for the same outcome, multivariable-adjusted estimates were preferentially extracted. When adjustment status was not explicitly reported, estimates were conservatively treated as unadjusted. The direction of all effect estimates was checked and harmonized before pooling.

If HRs and 95% CIs were not directly provided, survival data were extracted from published Kaplan–Meier curves using WebPlotDigitizer version 5.1, and HRs with corresponding 95% CIs were estimated according to the method described by Tierney et al.²⁵ Disagreements were resolved through discussion or consultation with a third reviewer (X.D.P.).

Literature quality assessment

The methodological quality of the included studies was assessed using the Newcastle–Ottawa Scale (NOS),²⁶ which evaluates 3 domains: selection of study groups, comparability between groups, and ascertainment of outcomes. Studies scoring ≥7 points were considered to be of high quality. Quality assessment was performed independently by 2 investigators (L.Z. and X.B.H.), with discrepancies resolved through discussion or consultation with a third reviewer (F.X.).

Statistical analysis

All statistical analyses were performed using Stata version 17.0. Pooled HRs and corresponding 95% CIs were calculated to assess the association between endocrine therapy and survival outcomes. Random-effects models were applied for all pooled analyses to account for potential between-study heterogeneity. Statistical heterogeneity was evaluated using Cochran’s Q test and the I² statistic. Sensitivity analyses were conducted using a leave-one-out approach to examine the robustness of the pooled estimates. Potential publication bias was assessed by visual inspection of funnel plots and Egger’s regression test, where appropriate.

Results

Study selection and characteristics

The literature search identified 3783 records. After removal of 448 duplicates, 3335 records were screened based on titles and abstracts. Twenty-seven full-text articles were assessed for eligibility, of which 6 studies met the inclusion criteria and were included in the meta-analysis (Figure 1).^{11,14,22,23,27,28} These studies comprised a total of 17 059 patients, including 8765 patients with ER-low (1%-10%) breast cancer who were included in the quantitative analyses. Most studies were retrospective cohort studies conducted in the United States and China, with sample sizes ranging from 407 to 7018 and median follow-up durations between 30 and 74 months. Key characteristics of the included studies are summarized in Table 1, and available clinicopathologic and treatment characteristics are provided in Table S2. The methodological quality of the included studies was generally high according to the NOS. Detailed assessments are provided in Table S3.

Figure 1.

PRISMA 2020 flow diagram of the study inclusion process. Abbreviation: PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

Table 1.

Summary of study and patient characteristics.

Reference	Country	Inclusion period	Study design	Total number of patients	Number of patients according to ER status	Survival outcome
Fujii et al. 22	United States	1982-2013	Retrospective	3055	<1%: 932 1%–9%: 171 ≥10%: 1952	TTR, OS
Fei et al. 23	United States	1998-2018	Retrospective	4179	<1%: 1100 1%–10%: 97 >10%: 2982	RFS
Cai et al. 11	China	2012-2017	Retrospective analysis of prospectively collected data	1013	1%–10%: 1013	DFS
Xie et al. 14	China	2010-2020	Prospective	407	1%-10%: 407	DFS
Wang et al. 27	China	2006-2016	Retrospective	1387	<1%: 422 1%–9%: 59 ≥10%: 906	DFS, OS
Choong et al. 28	United States	2018-2020	Retrospective	7018	1%–10%: 7018	OS

Endocrine therapy and OS

Three studies involving 7248 patients contributed data on OS. Before pooling, the direction of effect estimates was checked and harmonized. The HR reported by Choong et al.²⁸ was originally expressed as endocrine therapy omission vs endocrine therapy receipt and was therefore inverted for consistency with the other included studies.

After correction, the pooled analysis showed that endocrine therapy was associated with improved OS in patients with estrogen receptor–low breast cancer (HR = 0.81, 95% CI 0.69-0.95; P = .01), with no observed heterogeneity (I² = 0%; Figure 2). The leave-one-out sensitivity analysis showed that the direction of the pooled estimate remained generally consistent. When the largest study by Choong et al. was excluded, the pooled estimate still favored endocrine therapy but was no longer statistically significant (HR = 0.79, 95% CI 0.52-1.20; P = .27; Figure S1 [see online supplementary material for a color version of this figure], Table S4).

Figure 2.

Pooled hazard ratio for overall survival in patients with ER-low breast cancer. Abbreviation: REML, restricted maximum likelihood.

Endocrine therapy and DFS

Three studies comprising 1479 patients reported DFS outcomes. Pooled results demonstrated a significant reduction in recurrence risk with ET (HR = 0.64, 95% CI 0.45-0.90; P = .01) and no observed heterogeneity (I² = 0%; Figure 3). Given that time to recurrence and relapse-free survival (RFS) form components of DFS, studies reporting these endpoints were included in an exploratory pooled analysis, which showed comparable results (HR = 0.68, 95% CI 0.51-0.92; P = .01; I² = 0%; Figure S2 [see online supplementary material for a color version of this figure]).

Figure 3.

Pooled hazard ratio for disease-free survival in patients with ER-low breast cancer. Abbreviation: REML, restricted maximum likelihood.

ET in the ER 1%-5% subgroup

Two studies provided data specifically for patients with estrogen receptor expression of 1%-5%. After harmonizing the direction of effect estimates, endocrine therapy was not significantly associated with OS in this subgroup, although the point estimate favored endocrine therapy (HR = 0.87, 95% CI 0.70-1.08; P = .20), with no observed heterogeneity (I² = 0%; Figure 4).

Figure 4.

Pooled hazard ratio for overall survival in patients with ER 1%-5% breast cancer. Abbreviation: REML, restricted maximum likelihood.

Duration of endocrine therapy

Exploratory subgroup analyses comparing short-term (2-3 years) and long-term (≥3-5 years) endocrine therapy did not demonstrate a statistically consistent association with improved survival (HR = 0.45, 95% CI 0.13-1.50; P = .19; Figure S3 [see online supplementary material for a color version of this figure]). Given the limited number of contributing studies and the variability in definitions of “long-term” endocrine therapy across studies, these findings should be interpreted with caution and are presented in detail in the Supplementary Appendix.

Discussion

This systematic review and meta-analysis found that endocrine therapy was associated with improved OS and DFS in patients with estrogen receptor–low (ER-low, 1%-10%) breast cancer after harmonization of effect directions across studies. This finding suggests that, despite the biological overlap between estrogen receptor–low tumors and triple-negative breast cancer, this subgroup should not be uniformly considered endocrine-insensitive.

A gradient in endocrine responsiveness across the estrogen receptor–low spectrum is biologically plausible. Estrogen receptor–low tumors are characterized by reduced ESR1 transcriptional activity, low progesterone receptor expression, high Ki-67 proliferation indices, and enrichment of basal-like or triple-negative–associated gene signatures.^8,10 These features suggest that endocrine sensitivity in this subgroup may be attenuated but not necessarily absent. Tumors with estrogen receptor expression closer to 10% may retain greater hormone dependence, whereas those with estrogen receptor expression of 1%-5% may share more biological and clinical features with estrogen receptor–negative or triple-negative breast cancer.^28,29 This gradient may help explain why endocrine therapy was associated with improved OS in the overall estrogen receptor 1%-10% population, whereas the association did not reach statistical significance in the estrogen receptor 1%-5% subgroup, although the point estimate still favored endocrine therapy. Together, these observations support the view that estrogen receptor–low breast cancer represents a biologically heterogeneous continuum rather than a uniform endocrine-responsive or endocrine-insensitive entity.

Our exploratory analysis on endocrine therapy (ET) duration did not identify a consistent survival advantage associated with prolonged treatment. Given the limited number of contributing studies and the heterogeneity in definitions of long-term ET, these findings should be interpreted with caution. Further studies are warranted to better clarify the optimal duration of endocrine therapy in patients with ER-low breast cancer.

From a clinical perspective, endocrine therapy remains an integral component of breast cancer management.³⁰ However, its use in patients with estrogen receptor–low breast cancer should be guided by comprehensive clinicopathologic evaluation rather than estrogen receptor status alone. Given the potential adverse effects associated with endocrine therapy, including bone loss, vasomotor symptoms, and thromboembolic complications,^31,32 optimizing the selection of patients most likely to benefit remains an important clinical goal. In this context, the findings of the present study support individualized treatment decision-making and highlight the need for prospective studies to clarify the optimal use and duration of endocrine therapy in this subgroup.

In addition, patients with estrogen receptor–low breast cancer may achieve relatively high pathological complete response rates after neoadjuvant chemotherapy, and attainment of pathological complete response is generally associated with favorable prognosis.^33,34 However, the clinical relevance of endocrine therapy may differ according to residual disease status after neoadjuvant treatment. In Choong et al.,²⁸ endocrine therapy omission was associated with worse OS among patients with residual disease after neoadjuvant chemotherapy, whereas this association was not observed among those who achieved pathological complete response. Although the underlying mechanisms remain incompletely understood, residual tumors in this setting may retain variable degrees of estrogen-related signaling activity, which could contribute to heterogeneity in endocrine responsiveness. These findings suggest that treatment decisions for estrogen receptor–low breast cancer should consider chemotherapy response and residual disease status, rather than uniformly withholding endocrine therapy in this subgroup.

Beyond endocrine therapy and chemotherapy, emerging evidence has highlighted the potential role of immunotherapy in estrogen receptor–low breast cancer. Several studies have suggested that estrogen receptor–low tumors may share immune microenvironment features with triple-negative breast cancer, including higher tumor-infiltrating lymphocyte levels and increased PD-L1 expression.^35,36 Recent real-world studies of pembrolizumab-based neoadjuvant chemoimmunotherapy have reported encouraging pathological response rates in patients with estrogen receptor–low/HER2-negative breast cancer, although these findings remain preliminary and require further validation.^37–39 Future research should clarify how endocrine therapy may be integrated with chemotherapy, immunotherapy, and risk-adapted treatment strategies for this biologically heterogeneous subgroup.

Several limitations of this study should be acknowledged. First, all included studies were retrospective and observational in design, which may introduce residual confounding, selection bias, and variability in treatment practices. Second, heterogeneity in endocrine therapy regimens—including differences in drug type, treatment duration, and adherence—could not be fully accounted for and may have influenced the observed outcomes. Third, the relatively small number of eligible studies limited the statistical power to conduct more detailed subgroup analyses.

In summary, endocrine therapy was associated with improved OS and DFS in patients with estrogen receptor–low breast cancer. These findings suggest that estrogen receptor–low disease should not be uniformly considered endocrine-insensitive, although its biological and clinical heterogeneity should be recognized in treatment decision-making. Further prospective studies are needed to identify patients most likely to benefit from endocrine therapy and to clarify the optimal treatment strategy in this subgroup.

Author contributions

Fan Xia (Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing—original draft, Writing—review & editing), Xiaobo Hu (Investigation, Validation, Writing—review & editing), Lei Zhang (Investigation, Validation, Writing—review & editing), Xiaodong Peng (Methodology, Supervision, Validation, Writing—review & editing), and Mei Xie (Methodology, Validation, Visualization, Writing—review & editing)

Supplementary material

Supplementary material is available at The Oncologist online.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Conflicts of interest

None declared.

Data availability

Data were derived from already published manuscripts and are available from the corresponding author upon reasonable request.