Evolution of ER-cut-off definition in clinical trials for triple negative breast cancer: trends and patterns over 15 years Authors: Giorgio Bonomi, Davide Napetti, Gaia Griguolo, Francesca Zanghì, Fabio Girardi, Stefania Lando, Pietro Napolitano, Giusy Landa, Marina La Commare, Davide Massa, Valentina Guarneri, Maria Vittoria Dieci Published: 2026 ABSTRACT AbstractTriple negative breast cancer (TNBC) is historically defined as oestrogen receptor (ER) < 1% and HER2-negative. Increasing evidence supports that ER-low breast cancer (BC) (ER 1-10%) shares similar clinical behaviour and biology with classic TNBC. However, how ER cut-off definitions have been applied as eligibility criteria in clinical trials for TNBC over time remains unclear.In August 2024, the Clinicaltrials.gov database was systematically queried using the following string: “breast cancer” AND “interventional studies” AND “phase II-III-not applicable”. We included trials started from January, 1, 2010 to August, 7, 2024 restricted to TNBC, using either ER<1% or any other ER cut-off up to 10%. Differences across study descriptors were assessed using Mann–Whitney U and χ2 tests.A total of 6747 studies were retrieved. Of 620 eligible trials, 474 (76.5%) restricted the inclusion to TNBC and 146 (23.5%) included ER-low tumours. The percentage of trials allowing ER-low was lower for phase III versus phase I and phase II (16.8%, 17.5%, and 27.2%, respectively; p = 0.018); in industry versus investigator-driven studies (11.6% vs 29.1%, p < 0.001), and in metastatic versus early-stage settings (18.0% vs 33.5%, p < 0.001). ER-low inclusion increased over time, particularly in investigator-driven trials (since 2015), while remaining limited in industry-sponsored studies.We evaluated the ER cut-off adopted in trials testing therapies for TNBC. Although more inclusive criteria have been applied in recent years, patients with ER-low disease remain largely excluded from registrational trials, particularly in the metastatic setting, potentially limiting access to effective therapies. Broadening inclusion criteria could improve treatment opportunities for this high-risk population.Keywords: Breast cancer, TNBC, ER-low, Clinical trials Introduction Breast cancer (BC) is the most common malignancy among women and, despite diagnostic and therapeutic improvements, remains the leading cause of cancer-related death in women worldwide [1]. As BC represents a biologically and clinically heterogeneous disease, the identification and accurate assessment of clinically relevant biomarkers to dissect biological heterogeneity is essential to guide treatment choices. In this context, the assessment of oestrogen receptor (ER) status represents a key biomarker, acting both as a prognostic indicator and as a predictor of response to endocrine therapy. Indeed, ER+/HER2- BC generally presents a more favourable prognosis as compared to other BC subtypes and endocrine-based treatment options represent a cornerstone in the management of this disease, while ER–/HER2- BC (TNBC) is associated with an unfavourable prognosis, not only due to their inherently more aggressive clinical and biological behaviour, but also to the limited availability of targeted therapeutic options. However, the identification of a biologically sound and clinically meaningful threshold to discriminate between these two entities still presents significant challenges. Indeed, ER-negative BC is currently defined, according to the American Society of Clinical Oncology/College of American Pathologists (ASCO/CAP) guidelines, by the absence of oestrogen and progesterone receptors (ER or PgR <1%) [2]. The choice of this very low threshold was driven by the evidence of potential benefit from adjuvant endocrine therapy also in patients with low levels of ER expression [3]. However, increasing evidence has pointed out that HER2 negative BC with low levels of ER (1-10% of ER expression) presents clinical, pathological, biological and prognostic characteristics similar to TNBC [[4], [5], [6], [7], [8], [9]]. [4,10,11] Moreover, from a biological point of view, ER-low HER2- BC and TNBC have been reported to present similar gene-expression characteristics and immune features [10,12]. Recently, a retrospective multicentric study reported that ER-low HER2- BC and TNBC present comparable levels of tumour-infiltrating lymphocytes (TILs) density and similar CD8/FOXP3-positive cell ratios and the same study also confirmed a similar prognostic impact of TILs in both subgroups [10]. The 2020 ASCO/CAP guideline reflected this evidence by recognizing a new reporting entity called ER-low positive (ER-low), including BC cases with 1% to 10% of tumour cells with positive staining for ER by IHC [13]. However, there is still limited consensus on how this new entity, representing approximately 2-5% of all breast cancers, should be managed in clinical practice and specific guidelines are lacking [14]. As the development of novel treatment strategies for BC has been increasingly guided by BC subtype, the management of patients with ER-low BCs has become challenging. Patients with ER-low BCs have often been excluded from clinical trials specifically designed for TNBC only, while at the same time being underrepresented in trials focusing on HR + HER2-disease on the assumption of scarce benefit from endocrine therapy. Therefore, this rare subgroup of patients has experienced an increasing therapeutic gap in recent years. Therefore, a better understanding of how specific ER expression thresholds were applied in clinical trials underpinning current TNBC treatment represents an opportunity to inform future clinical trial design. To assess this, we conducted a systematic review of ER expression cut-off values applied as inclusion criteria in clinical trials focusing on TNBC conducted between January 2010 and August 2024. Materials and methods Literature search This systematic review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [15]. A systematic literature review of ClinicalTrials.gov for studies started between January 1, 2010 and August 7, 2024 was conducted. The search strategy included the following combination of terms: “breast cancer” AND “interventional studies” AND “phase II-III-not applicable”. Evaluation of the identified records was carried out independently by three authors (GB, DN, and FZ), and any discrepancies were resolved through discussion with a third author (MVD and GG). As registration of a systematic review is not mandatory according to the latest PRISMA guidelines, this systematic review was not registered in public registries. Selection criteria and data extraction To be included in the present analysis, eligible trials had to satisfy all of the following inclusion criteria: i) studies included TNBC or ER low BC patients; ii) studies assessing pharmacological strategies. Exclusion criteria were: i) studies assessing locoregional strategies (surgery or radiotherapy); ii) studies only assessing quality of life outcomes. After trial identification, the following variables were extracted from all trials, when available: trial number, study title, year of study initiation, number of patients planned for enrolment, the clinical trial's phase and the clinical setting of the trial (metastatic versus neoadjuvant/adjuvant). The study phase was categorised: phase I and phase I/II as phase I, phase II as phase II, phase II/III and phase III as phase III; when the study phase was not available (i.e., not applicable), publications were reallocated to the relevant category based on the information provided by the summary. In addition, the type of therapy investigated in the trial was assessed and categorised in one of the following: chemotherapy-based studies, target therapy-based studies (studies included target therapies only or target therapies plus chemotherapy), immunotherapy-based trials (studies included immunotherapy only or immunotherapy plus chemotherapy), antibody-drug conjugate [ADC] based studies (which included trials assessing ADCs and trials assessing combinations of ADCs and immunotherapy), trials assessing the combination of target therapies and immunotherapy, and other treatments (treatments or combinations not included in the previous categories). Additionally, data on the trial's sponsor was extracted and classified as industry-driven versus investigator-driven. For all trials included, the ER cut-off value used in the inclusion criteria to define TNBC was extracted and categorised as follows: ER<1%, ER<10%, or other cut-off between 1 and 9%. Descriptive statistics were used to describe study and patients’ characteristics. For continuous variables, median and interquartile range values were computed; continuous variables were compared using the Mann–Whitney U test, whereas categorical variables were compared using the χ2 test. All reported p values are two-sided, and significance level was set at 5% (p < 0.05). Statistical analysis was conducted using IBM SPSS version 29.0. Results In total, 6.747 records were identified from ClinicalTrials.gov; of these, 4.935 were excluded by title and 1.812 b y summary review, as they did not meet the predefined inclusion criteria of our study. A total of 620 studies met the inclusion criteria and were included in the present analysis. The PRISMA flowchart is shown in Fig. 1. PRISMA flowchart. A complete list of trials included is presented in Supplementary Table 1. Trial characteristic in the overall study cohort The characteristics of the 620 studies (median patients planned per trial: 60; range 0-2400) included in the present analysis are summarized in Table 1. Table 1Clinical trial characteristics in the overall study cohort.Table 1N%Patients, median (IQR)60 (29-169)Study period 2010–201410316.6% 2015–201924339.2% 2020–202427444.2%Study phase I12620.3% II39363.4% III10116.3%Randomization Randomized24639.7% Non-randomized37460.3%Sponsorship Investigator-driven42268.1% Industry-sponsored19831.9%Setting Neoadjuvant/Adjuvant22135.6% Metastatic39964.4%Therapy CT only11718.9% Target ± CT18930.5% Immunotherapy ± CT16025.8% ADC ± immunotherapy416.6% Target + immunotherapy213.4% Other9214.8%CT: chemotherapy; ADC: Antibody drug conjugate. Clinical trial characteristics in the overall study cohort. CT: chemotherapy; ADC: Antibody drug conjugate. Of these, 103 (16.6%) studies were initiated from 2010 to 2014, 243 (39.2%) from 2015 to 2019, and 274 (44.2%) from 2020 to 2024. Overall, 126 (20.3%) were phase I, 393 (63.4%) were phase II, 101 (16.3%) were phase III trials. Coherently, only slightly more than one third of studies included were randomized (N = 246, 39.7%), while the majority of trials included were not (N = 374, 60.3%). In addition, around one third of the trials (N = 221, 35.6%) focused on neoadjuvant/adjuvant treatment for BC, while the majority (N = 399, 64.4%) enrolled patients with metastatic BC. The type of experimental treatment was chemotherapy based for 117 trials (18.9%), targeted therapy ± chemotherapy for 189 trials (30.5%), immunotherapy ± chemotherapy for 160 trials (25.8%), ADCs ± immunotherapy for 41 trials (6.6%), targeted therapy + immunotherapy for only 21 studies (3.4%), while 92 trials assessed other types of treatments or a combination of treatments not included in the previous categories (N = 92 studies, 14.8%) (Fig. 2). Type of treatment assessed in the included trials. The majority of trials included (N = 422, 68.1%) were investigator-driven, while 198 trials (31.9%) were sponsored by industry. Association between clinical trial characteristics and ER-expression cut-off adopted as inclusion criteria When assessing the cut-off for ER expression levels adopted as inclusion criteria in each study, we observed that 474 (76.5%) of trials strictly restricted enrolment to patients with TNBC, while 146 studies (23.5%) also included patients with ER-low expression. In particular, among ER-low trials, only 21 studies used a cut-off different from ER<10% as inclusion criteria: 17 studies used an ER cut-off of 5% (≤5% or <5%); 2 of these used an ER cut-off of 20% (≤20 % or <20%) and 2 used an ER cut-off of <2%. For the following analyses, we classified all studies including patients with ER between 1% and 10% as studies including patients with ER-low disease, independently from the specific cut-off used in the study. We then assessed potential associations between clinical trial characteristics and ER-expression cut-off adopted as inclusion criteria (TNBC only versus TNBC and ER-low disease) (Table 2). Table 2Clinical trial characteristics according to ER cut-off adopted as inclusion criteria.Table 2Patients, median (IQR)aER<1% (N = 474, 76.5%)ER-low (N = 146, 23.5%)p-valueN%N%63 (29-182)53.5 (26.7-133.7)0.314Study period0.121 2010–20148683.5%1716.5% 2015–201917873.3%6526.7% 2020–202421076.6%6423.4%Study phase0.018 I10482.5%2217.5% II28672.8%10727.2% III8483.2%1716.8%Study design0.689 Randomized18675.6%6024.4% Non-randomized28877.0%8623.0%Study Sponsor<0.001 Investigator-driven29970.9%12329.1% Industry-sponsored17588.4%2311.6%Clinical Setting<0.001 Neoadjuvant/Adjuvant14766.5%7433.5% Metastatic32782.0%7218.0%Therapy0.831 CT only9177.8%2622.2% Target ± CT14878.3%4121.7% Immunotherapy ± CT11873.8%4226.3% ADC ± immunotherapy2970.7%1229.3% Target + immunotherapy1781.0%419.0% Other7177.2%2122.8%CT: chemotherapy; ADC: Antibody drug conjugate.aContinuous variables are reported as median (range) and were compared using the Mann–Whitney U test; Categorical variables are reported as frequencies and percentages and were compared using the χ2 test. Clinical trial characteristics according to ER cut-off adopted as inclusion criteria. CT: chemotherapy; ADC: Antibody drug conjugate. Continuous variables are reported as median (range) and were compared using the Mann–Whitney U test; Categorical variables are reported as frequencies and percentages and were compared using the χ2 test. Studies allowing the inclusion of patients with ER-low BC were significantly more represented among investigator-driven trials (29.1%, 123/422) than among industry-sponsored studies (11.6%, 23/198, p < 0.001), although studies strictly including only patients with TNBC represented the majority of trials in both subgroups. Moreover, trials allowing the inclusion of patients with ER-low BC were significantly more represented among studies conducted in the neoadjuvant/adjuvant setting (33.5%, 74/221) than among studies conducted in the metastatic setting (18.0%, 72/399) (p < 0.001). When assessing the study design, we observed that studies allowing the inclusion of patients with ER-low BC were more frequent among phase II studies (27.2%, 107/393) than among phase I (17.5%, 22/126) and phase III trials (16.8%, 17/101) (p = 0.018). The median number of patients per trial planned did not differ significantly between trials strictly enrolling patients with TNBC (median 63 patients; interquartile range 29-182) and trials including patients with ER-low expression were (median 53.5 patients; interquartile range 26.7-133.7) (p = 0.314). We did not observe a significant association between other characteristics of the clinical trials (including year of study initiation) and the inclusion of patients with ER-low BC (p = 0.121 for year of study initiation) (Table 2). However, we observed a numerical increase over time in the fraction of trials designed for TNBC that also allowed the inclusion of patients with ER-low BC (16.5%, 26.7%, and 23.4% for the periods 2010–2014, 2015–2019, and 2020–2024, respectively, p = 0.121, Table 2). The trend towards an increased inclusion of patients with ER-low disease observed from 2015 was mainly for investigator-driven trials, while among industry-sponsored trials the proportion of trials allowing for ER-low BC was consistently lower than in investigator-driven trials. However, a slight, not significant, increase in the proportion of trials permitting ER-low disease was also observed among industry-sponsored trials (10.4%, 11.4 %, and 12.2 % for the periods 2010–2014, 2015–2019, and 2020–2024, respectively) (Fig. 3). Fraction of trials designed for TNBC allowing the inclusion of patients diagnosed with ER-low BC. Further details on ER cut-off values used, experimental treatments assessed and additional characteristics of the 146 ER-low trials are reported in Supplementary Table 1. Among studies including ER-low patients (N = 146), 66 trials (45.2%) had not yet been published, neither as full manuscripts nor as conference abstracts, at time of assessment. Among these studies, the recruitment status as reported in Clinicaltrials.gov was as follows: 11 active but not recruiting, 25 recruiting, 7 terminated, 11 completed, 6 unknown, 1 suspended, 3 withdrawn, and 2 not yet recruiting. Only 53 (36.3%) studies had been published as full peer-reviewed articles and an additional 11 (7.5%) had been reported as conference abstracts with available results, while 16 (11.0%) have been presented as abstracts without results (Supplementary Table 1). Among studies with available results (either published as full manuscripts or reported as conference abstracts), we identified only two studies which provided subgroup analyses comparing TNBC and ER-low populations. Detailed results for these two studies, overall and in the ER-low and TNBC subgroup, are reported in Supplementary Table 2. Discussion In the past decades, ER-positive BC has been defined by ER expression ≥1% [2], based on evidence that also patients with minimal ER positivity may benefit from endocrine therapy [16,17]. Based on this definition, only BCs with ER expression <1% have been classified as ER-negative. Over the years, however, many retrospective studies have reported that early BCs with ER <10% present biological features, pathological clinical response (pCR) rates and survival patterns [4,5,8,9,18], and and biological features, in terms of gene-expression and immune features [10,12], that closely resemble those of TNBCs. Now that targeted therapeutic options for BC patients extend beyond endocrine treatment, including immune-checkpoint inhibitors and antibody–drug conjugates (ADCs), these observations have questioned the exclusion of patients with ER-low tumours from the TNBC spectrum, as this may lead to missed opportunities for potentially effective treatments [19]. Our analyses showed that the majority of trials allowing for patients with ER-low tumours were investigator-driven rather than industry-sponsored. Despite a slight improvement over time in favour of patients with ER-low tumours, proportions remain largely suboptimal (Fig. 3). One possible explanation for the differences observed between investigator-driven and industry-driven trials is that investigator-driven trials reflect clinicians’ understanding of ER-low disease biology, while industry-sponsored trials are typically bound by stricter adherence to guideline definitions and regulatory directives. Coherently, we found that inclusion of patients with ER-low disease was more common in exploratory studies (phase II) than in phase III registrational trials. In this context, the exclusion of ER-low tumours from industry-sponsored trials may reflect the interest in maintaining cohort homogeneity, thereby optimizing the probability of demonstrating treatment efficacy. Indeed, some studies have raised potential doubt that ER-low tumours might constitute a more heterogeneous group of patients in terms of clinicopathologic features, molecular classification, use of adjuvant endocrine treatment, and prognosis, as compared to TNBC [20]. Therefore, there might be some potential worries that the inclusion of these patients may increase variability in treatment response and reduce the likelihood of detecting a statistically significant treatment effect. However, while excluding ER-low tumours might potentially increase internal validity by ensuring cohort homogeneity, it simultaneously results in low external validity, limiting the applicability of the trial's findings to the broader clinical reality [21]. In this regard, real-world evidence (RWE) will be essential to validate these findings in ER-low populations, capturing the biological and clinical complexity that is often underrepresented in randomized clinical trial. In addition, we observed a statistically significant association between the inclusion of patients with ER-low disease and the clinical setting in which they were conducted, with a higher proportion of studies conducted in the early setting than in the metastatic setting. Indeed, within the period 2010-2024, 33.5% of studies conducted in the neoadjuvant/adjuvant setting allowed for the inclusion of patients with ER-low disease, while only 18% of studies conducted in the metastatic disease setting allowed for the inclusion of patients with ER-low disease (p < 0.001). Achieving pCR after neoadjuvant therapy represents an important prognostic factor for patients with TNBC as it is associated with improved long-term outcome, while patients with residual disease are more likely to experience disease recurrence and a worse prognosis [22,23]. As several studies reported that ER-low/HER2-negative BC present similar pCR rates to those observed in ER-negative/HER2-negative BC [4], the intensification of efforts from the scientific community to promote the inclusion of ER-low BC patients in neoadjuvant trials is warranted. On the other hand, emerging evidence has highlighted the potential clinical relevance of ER-low BC in the metastatic setting. Recently, a large multicentric study assessing the prognostic impact of conversion to ER-low BC at relapse reported that patients diagnosed with HR+/HER2- BC and relapsing with an ER-low disease experienced the poorest survival, with outcomes that were worse than when converting to a full TNBC phenotype [24]. This observation may be related to the limited access for these patients to more recent therapeutic options specifically developed for TNBC. These observations underscore the need for strengthening the representation of ER-low patients in trials dedicated to metastatic TNBC. We therefore advocate for a commitment to promote dedicated research efforts for this subgroup of patients in the metastatic setting. Conclusions Based on our findings, between 2% and 5% of patients with BC may not be treated according to the most appropriate therapeutic algorithm [14]. To close this gap, we believe that the scientific community should make greater efforts to promote the inclusion of patients with ER-low BC in clinical trials designed for TNBC, including industry-sponsored phase III trials. CRediT authorship contribution statement Giorgio Bonomi: Writing – review & editing, Writing – original draft, Formal analysis, Data curation. Davide Napetti: Writing – review & editing, Writing – original draft, Formal analysis, Data curation. Gaia Griguolo: Writing – review & editing, Supervision, Methodology, Funding acquisition, Conceptualization. Francesca Zanghì: Writing – original draft, Data curation. Fabio Girardi: Writing – review & editing, Methodology, Formal analysis. Stefania Lando: Writing – review & editing, Supervision, Data curation. Pietro Napolitano: Data curation. Giusy Landa: Data curation. Marina La Commare: Data curation. Davide Massa: Data curation. Valentina Guarneri: Writing – review & editing, Supervision, Methodology, Funding acquisition, Conceptualization. Maria Vittoria Dieci: Writing – review & editing, Supervision, Methodology, Funding acquisition, Conceptualization. Funding This work was supported by University of Padova, Department of Surgery, Oncology and Gastroenterology DOR 2024-2025 (to VG, MVD, GG; grant number not applicable), Ricerca Corrente funding from the Italian Ministry of Health (grant number not applicable). Declaration of competing interest MVD reports personal fees for consultancy/advisory role from Eli Lilly, Pfizer, Novartis, Seagen, Gilead, MSD, Exact Sciences, AstraZeneca, Roche, Daiichi-Sankyo, Roche. VG reports personal fees for advisory board participation for AstraZeneca, Daiichi-Sankyo, Eli Lilly, Gilead, MSD, Novartis, Pfizer, Pierre Fabre, Roche, Menarini Stemline, Exact Sciences; personal fees as an invited speaker for AstraZeneca, Daiichi-Sankyo, Eli Lilly, Exact Sciences, Gilead, GSK, Novartis, Roche, Zentiva, Menarini Stemline; personal fees for expert testimony for Eli Lilly, patents for HER2DX (Institution). DM reports fees travel expenses form: Eli Lilly, Pfizer. FG reports honoraria for lectures from Eli Lilly, Gilead; consulting fees from the World Health Organization, support for attending meetings from Eli Lilly, Gilead, Novartis. GG reports fees for advisory role from Gilead, Seagen, Menarini, personal fees as an invited speaker from Eli Lilly, Novartis, MSD. All remaining authors have declared no conflicts of interest. Footnotes Supplementary data to this article can be found online at https://doi.org/10.1016/j.breast.2026.104819. Supplementary data The following are the Supplementary data to this article: Multimedia component 1 Multimedia component 2 Data availability All data supporting the findings of this study are available within the article and its Supplementary Material. References Associated Data Supplementary Materials Multimedia component 1 Multimedia component 2 Data Availability Statement All data supporting the findings of this study are available within the article and its Supplementary Material.