Breast cancer has the highest incidence rate and is the leading cause of cancer-related death among women worldwide (1). Among the breast cancer subtypes, luminal A and luminal B express estrogen receptor (ER) and/or progesterone receptor, and certain luminal B tumors express HER2 and highly express the proliferation marker Ki-67 (MKI67) (2,3). Additionally, luminal B breast cancer has a poorer prognosis among the breast cancer subtypes (2-10). The main treatments for luminal B breast cancer entail surgery, radiotherapy and drug therapy, which include endocrine therapy, HER2-targeted therapy, and chemotherapy (2,3,11). However, since the luminal B subtype still exhibits a poorer prognosis, further stratification and the development of prognostic markers and therapeutic targets are required for improving the quality of life of patients. It is also important to identify effective prognostic biomarkers to assess the effectiveness of treatments for luminal B breast cancer.

Cancer stem cells (CSCs) have stem-like properties such as self-renewal, multi-differentiation and tumorigenicity (12,13). Since CSCs are resistant to conventional drug therapy and radiotherapy, the development of targeted therapies against CSCs is necessary to improve the clinical outcomes of patients (12-15). Aldehyde dehydrogenase 1 (ALDH1) is a detoxifying enzyme that converts aldehydes into carboxylic acids and is known as a CSC marker (16). Among the ALDH1 genes, ALDH1A3 is known to be a CSC marker in several cancer types (17,18). ALDH1A3 expression is correlated with tumor grade, metastasis and the clinical outcome of breast cancer (19-21), and it is a significant contributory factor to ALDH1 activity in this disease (20,22). ALDH1A3 expression is also associated with chemoresistance and radioresistance in several cancer types (17,22,23). However, the association between ALDH1A3 expression and the effectiveness of endocrine therapy in luminal B breast cancer remains to be elucidated.

p62 is a multifunctional scaffold protein that is involved in various biological and cellular activities, including the NF-κB signaling pathway, antioxidant response and autophagy (24-34). p62 is highly expressed in a number of cancer types (35-45), and high p62 expression is corelated with a poor clinical outcome in several cancer types (38,41,43,46-49). The p62 gene and protein are also highly expressed in breast cancer (50-52), and high p62 expression is corelated with a poor clinical outcome (51-55). Furthermore, high expression of p62 and ALDH1A3 indicates a poor clinical outcome in luminal B breast cancer, and p62 is involved in the cancerous progression of ALDH1-positive luminal B breast CSCs (55). Moreover, co-expression of p62 and ALDH1A3 in luminal B breast cancer reduces the effectiveness of radiotherapy (56). However, the association between endocrine therapy and high p62 and ALDH1A3 expression, in luminal B breast cancer remains unclear.

The results of the present study revealed that p62^high ALDH1A3^high luminal B tumors are associated with a reduced effectiveness of endocrine therapy, especially aromatase inhibitor therapy.

Datasets. The Molecular Taxonomy of Breast Cancer International Consortium (METABRIC; n=2,509) (57,58) and The Cancer Genome Atlas (TCGA; n=1,084) (59) datasets were downloaded from cBioportal (60,61) on August 1, 2024. The details of the data are shown in Figure 1 and have also been reported previously (62).

Statistical analysis. The details of the statistical analyses have been reported previously (62). The optimal cutoff determined by the Youden’s index following receiver operating characteristic analysis was used to divide each group into the p62 or/and ALDH1A3 high and low gene expression groups for disease-specific survival (DSS) analysis (Table I). Kaplan-Meier curves for DSS were plotted and then compared using the log-rank (Cochran-Mantel-Haenszel) test. Multiplicity was adjusted using the Holm test as a post hoc test. A multivariate Cox regression analysis for DSS was performed to evaluate the influence of gene expression and to estimate the adjusted hazard ratios (HRs) with the age at diagnosis as a confounding factor in Table II, Table III, Table IV and Table V. In Table II, chemotherapy and radiotherapy were also included as additional confounding factors. Two-sided p<0.05 was considered to indicate a statistically significant difference. All statistical analyses were conducted using R software (version 4.3.0) and BellCurve for Excel (version 4.05; Social Survey Research Information Co., Ltd., Tokyo, Japan).

Endocrine therapy is insufficient for treating p62^high luminal A breast cancer but sufficient for p62^high luminal B breast cancer in the METABRIC dataset. High p62 expression is associated with poorer clinical outcomes in luminal A and luminal B breast cancer (55). However, the association between endocrine therapy and high p62 expression in the luminal A and luminal B breast cancer subtypes remains unclear. Therefore, to examine the clinical outcome of patients with p62^high luminal A and luminal B breast cancer treated with endocrine therapy, Kaplan-Meier and multivariate analyses for DSS were performed. The overall workflow of the present study is shown in Figure 1. We first analyzed the gene expression data from patients with breast cancer in the METABRIC dataset. We performed Kaplan-Meier analysis to compare the DSS of patients with p62^high or p62^low luminal A and luminal B breast cancer who did or did not receive endocrine therapy. Patients with p62^high luminal A breast cancer who did not receive endocrine therapy exhibited a good prognosis and those who received endocrine therapy displayed a poorer prognosis (p=0.0035) (Figure 2A). Although patients with p62^high luminal B breast cancer who did not receive endocrine therapy exhibited a poorer prognosis (p<0.001), those who did receive endocrine therapy showed an improved prognosis (Figure 2B).

Next, multivariate Cox regression analysis of DSS was performed with age, chemotherapy and radiotherapy as confounding factors. As with the Kaplan-Meier analysis, patients with p62^high luminal A breast cancer who received endocrine therapy exhibited poorer clinical outcomes [without endocrine therapy: hazard ratio (HR)=1.66, 95% confidence interval (CI)=0.85-3.25, p=0.142; with endocrine therapy: HR=1.57, 95%CI=1.06-2.33, p=0.024] (Table II). Patients with p62^high luminal B breast cancer who did not receive endocrine therapy exhibited poorer clinical outcomes (HR=3.84, 95%CI=1.64-8.99, p=0.002) and those who received endocrine therapy exhibited improved clinical outcomes (HR=1.41, 95%CI=0.90-2.19, p=0.132) (Table II). These results suggest that endocrine therapy may be insufficient for treating p62^high luminal A breast cancer.

Endocrine therapy is insufficient for treating p62^highALDH1A3^high luminal B breast cancer in the METABRIC dataset. We have previously shown that high p62 and ALDH1A3 expression is associated with poorer clinical outcomes in luminal B breast cancer (55). Moreover, the effectiveness of radiotherapy is reduced in luminal B breast cancer with p62 and ALDH1A3 co-expression (56). However, the association between endocrine therapy and high p62 and ALDH1A3 expression, in luminal B breast cancer remains unclear. Therefore, we next performed Kaplan-Meier analysis for patients with p62^highALDH1A3^high luminal A and luminal B breast cancer who did or did not receive endocrine therapy. First, the effect of high ALDH1A3 expression in luminal A and luminal B breast cancer was examined. Kaplan-Meier analyses showed that patients with ALDH1A3^high luminal A and luminal B breast cancer treated without endocrine therapy did not have poor clinical outcomes compared with ALDH1A3^low patients (Figure 2C and D). Additionally, multivariate Cox regression analysis showed that patients with ALDH1A3^high luminal A and luminal B breast cancer treated without endocrine therapy did not have poorer clinical outcomes compared with ALDH1A3^low patients (Table II). However, ALDH1A3^high luminal B breast cancer treated with endocrine therapy was predictive of poorer clinical outcomes in Kaplan-Meier analysis (p=0.032) and poorer tendency in multivariate Cox regression analysis (HR=1.75, 95%CI=0.99-3.09, p=0.055) compared with ALDH1A3^low (Figure 2C and D and Table II).

Next, Kaplan-Meier analyses showed that patients with p62^high ALDH1A3^high luminal A and luminal B breast cancer treated without endocrine therapy did not have poorer clinical outcomes than the other patient groups (p62^high ALDH1A3^low, p62^low ALDH1A3^high and p62^low ALDH1A3^low) with the luminal A and luminal B subtypes (Figure 2E and F). Notably, in patients treated with endocrine therapy, there was a significant difference between the survival rates of those with p62^high ALDH1A3^high luminal B breast cancer and the other patients with the luminal B subtype (p=0.030) (Figure 2F). The comparison among the four groups indicated that p62^high ALDH1A3^high did not have the poorest survival of the four groups of patients with the luminal A breast cancer subtype, whereas p62^high ALDH1A3^high had the poorest survival among the four groups of patients with the luminal B breast cancer subtype (p=0.067) (Figure 2G and H). Next, a multivariate Cox regression analysis for DSS was performed with age at diagnosis as a confounding factor (Table II). Multivariate Cox regression analysis also showed that p62^high ALDH1A3^high was not predictive of poorer clinical outcomes of patients with luminal A breast cancer treated without endocrine therapy compared with the other patient groups (p62^high ALDH1A3^low, p62^low ALDH1A3^high and p62^low ALDH1A3^low) with the luminal A subtype (Table II). However, p62^high ALDH1A3^high was predictive of poorer clinical outcomes of patients with luminal B breast cancer treated with endocrine therapy compared with other patient groups with the luminal B subtype (HR=1.52, 95%CI=1.05-2.20, p=0.028) (Table II). These results suggest that p62 may be involved in cancer progression and contribute to the poor prognosis of patients with ALDH1A3-positive CSC-enriched luminal B breast cancer. These results also suggest that endocrine therapy may be insufficient for treating p62^high ALDH1A3^high luminal B breast cancer.

Endocrine therapy is insufficient for treating p62^highALDH1A3^high luminal B breast cancer in TCGA dataset. To confirm the above results from the METABRIC dataset analysis, we next examined TCGA dataset as another cohort. Of the p62^high groups of patients with luminal A and luminal B breast cancer who did or did not receive endocrine therapy, those with the luminal B subtype treated with endocrine therapy showed a poor prognosis (p=0.0021) (Figure 3A and B). The ALDH1A3^high groups of patients with luminal A and luminal B breast cancer who did or did not receive endocrine therapy did not show a poor prognosis (Figure 3C and D). Next, Kaplan-Meier analyses showed that patients with p62^high ALDH1A3^high luminal A and luminal B breast cancer treated without endocrine therapy did not have poorer clinical outcomes than the other patient groups (p62^high ALDH1A3^low, p62^low ALDH1A3^high and p62^low ALDH1A3^low) with the luminal A and luminal B subtypes (Figure 3E and F). For patients treated with endocrine therapy, there was a significant difference between the survival rates of those with p62^high ALDH1A3^high luminal B breast cancer and the other patient groups with the luminal B subtype (luminal B: p<0.001) (Figure 3F). The comparison among the four groups indicated that the p62^high ALDH1A3^high group did not have the poorest survival of the four groups of patients with the luminal A subtype (Figure 3G), whereas the p62^high ALDH1A3^high group had the poorest survival of the four groups of patients with the luminal B subtype (p=0.0013) (Figure 3H). Subsequently, multivariate Cox regression analysis for DSS was performed with age as a confounding factor, and similar results were obtained as those in the Kaplan-Meier analysis (Table III). These results also suggest that endocrine therapy may be an insufficient treatment option for patients with p62^high ALDH1A3^high luminal B breast cancer.

Tamoxifen is sufficient for treating p62^high luminal A and luminal B breast cancer in TCGA dataset. Next, the endocrine therapy group was divided into two groups, those treated with tamoxifen and those treated with aromatase inhibitors, to examine in more detail the relationship between high p62 and ALDH1A3 expression, and endocrine therapy. Of the p62^high groups of patients with luminal A and luminal B breast cancer who did or did not receive tamoxifen treatment as endocrine therapy, those with the luminal B subtype treated without tamoxifen showed a poor prognosis (p=0.0045) (Figure 4A, B and Table IV). The ALDH1A3^high groups of patients with luminal A and luminal B breast cancer who did or did not receive tamoxifen treatment as endocrine therapy did not show a significant difference (Figure 4C, D and Table IV).

Tamoxifen is insufficient for treating p62^high ALDH1A3^high luminal A breast cancer in TCGA dataset. In patients treated with tamoxifen, there was a significant difference between the survival rates of those with p62^high ALDH1A3^high luminal A breast cancer and the other patient groups with the luminal A subtype, and patients with p62^high ALDH1A3^high luminal B breast cancer treated with tamoxifen tended to have a poor outcome (luminal A with tamoxifen: p=0.013; luminal B with tamoxifen therapy: p=0.054) (Figure 4E and F). The comparison among the four groups showed that the p62^high ALDH1A3^high group had the poorest survival of the four groups of patients with the luminal A and luminal B subtypes, although there was not a significant difference (Figure 4G and H).

Aromatase inhibitor therapy is insufficient for treating p62^high luminal B breast cancer in TCGA dataset. Of the p62^high groups of patients with luminal A and luminal B breast cancer who did or did not receive aromatase inhibitor treatment as endocrine therapy, those with the luminal B subtype showed a poor prognosis (p<0.001) (Figure 5A and B). The ALDH1A3^high groups of patients with luminal A and luminal B breast cancer who did or did not receive aromatase inhibitor treatment as endocrine therapy did not show a poor prognosis (Figure 5C and D).

Aromatase inhibitor therapy is insufficient for treating p62^high ALDH1A3^high luminal B breast cancer in TCGA dataset. For patients treated with aromatase inhibitors, there was no significant difference between the survival rates of those with p62^high ALDH1A3^high luminal A breast cancer and other patients with the luminal A subtype (Figure 5E). However, patients with p62^high ALDH1A3^high luminal B breast cancer treated with aromatase inhibitors had a poor clinical outcome (p<0.001) (Figure 5F). The comparison among the four expression groups showed that p62^high ALDH1A3^high luminal A breast cancer did not show a poor prognosis (Figure 5G), but p62^high ALDH1A3^high indicated a poorer prognosis among the four groups of patients with the luminal B subtype (p<0.001) (Figure 5H). These results also suggest that aromatase inhibitor as endocrine therapy may be an insufficient treatment option for patients with p62^high ALDH1A3^high luminal B breast cancer.

The present study demonstrated that the p62^high ALDH1A3^high luminal B subtype tumors treated with endocrine therapy, especially those treated with aromatase inhibitors, exhibited poor clinical outcomes. High p62 expression in cancer tissue at the gene and protein level contributes to cancer progression and metastasis (52,63-68) and is associated with a poor clinical outcome in patients with breast cancer, including the luminal B subtype (55). Therefore, p62 is expected to predict prognosis after medical treatment for several cancer types. A series of experiments at the cellular level have shown that p62 is also involved in cell proliferation, survival, antioxidant response, autophagy and tumor formation (24,26,31,55,69). In addition, p62 is involved in chemoresistance and radioresistance in cancer cells by enhancing the level of NF-κB signaling, mTOR signaling and autophagy flux (70,71). In breast cancer cell lines, p62 binds to vimentin and induces invasion and metastasis (52). In endometrial cancer cell lines, 17β-estradiol induces p62 phosphorylation and ERα expression via the PI3K/Akt/mTOR signaling pathway, and p62 phosphorylation releases ERα from the p62-KEAP1-ERα complex (72). Endocrine therapy activates the integrated stress response through NF-κB signaling in a subpopulation of ER-positive breast cancer cells (73). However, the mechanism linking high p62 expression to cancer progression and to the reduced effectiveness of endocrine therapy in luminal B breast cancer is unclear. Because aromatase inhibitors block estrogen production in postmenopausal adipose tissue, the estrogen-independent and p62-dependent signaling pathways that remain active in luminal B breast cancer cells may be responsible for the reduced therapeutic efficacy of aromatase inhibitors. This is supported by the results of the present study in which tamoxifen, which acts directly on the ER, showed a tendency towards a reduced effectiveness in p62^high ALDH1A3^high luminal B breast cancer (Figure 4). Thus, it will be important to clarify the detailed molecular mechanism in the future.

The luminal B breast cancer subtype expresses HER2 and is treated with HER2-targeted therapy (2,3). Therefore, we investigated the prognosis (DSS) of patients with p62^high luminal B breast cancer who received HER2-targeted therapy, but we were unable to perform a prognostic analysis as no patients who received HER2 targeted therapy died in the dataset. Additionally, since the luminal B type has a high proliferation rate, chemotherapy is usually also performed. It is important to examine the effects of HER2-targeted therapy and chemotherapy in patients with high p62 expression in luminal B subtype in the future.

p62 is involved in the maintenance of cancer stem-like properties by stabilizing MYC mRNA and NRF2 activation (51,74) in several cancer types. p62 is also involved in the stem-like properties of ALDH1-positive luminal B breast CSCs (55). Furthermore, p62^high ALDH1A3^high luminal B breast cancer exhibits reduced response to radiotherapy, and X-ray irradiation suppresses the tumor formation of p62-deficient, ALDH1-positive luminal B breast CSCs (56). These findings raise the possibility that the reduced effectiveness of endocrine therapy in p62^high ALDH1A3^high luminal B breast cancer results from the stem-like properties of ALDH1A3-positive luminal B breast CSCs. Therefore, it is necessary to reveal the mechanism by which p62 is involved in these properties. p62 binds to atypical protein kinase C such as PKCζ and PKCλ/ι via a PB1-PB1 domain interaction (25). PKCζ^high and PKCζ^high ALDH1A3^high luminal B breast cancers exhibit reduced response to endocrine therapy and aromatase inhibitor treatment (62,75). Thus, p62 and PKCζ may be cooperatively involved in the reduced effectiveness of endocrine therapy in ALDH1A3-positive luminal B breast cancer.

The present study revealed that endocrine therapy, especially aromatase inhibitor therapy, exhibits reduced effectiveness in p62^high ALDH1A3^high luminal B subtype tumors for. Therefore, p62^high and ALDH1A3^high may be used as a prognostic biomarker for predicting the efficacy of endocrine therapy for luminal B breast cancer.

The Authors declare that they have no competing interests in relation to this study.

Conceptualization: YM, AO, KA; Formal analysis: YM, AO, RC, KN; Founding Acquisition: AO, RC, ST, SO, KS, KA; Investigation: YM, AO, RC, ME, KN, KS, KA; Methodology: YM, AO, KN, KA; Project Administration: KA; Supervision: KA; Validation: RC, ME, YT, HI, KN, TK, RO, YH, ST, KS; Visualization: YM, AO, ME; Writing - Original Draft Preparation: YM, AO, YN, HI, KA; Writing - Review & Editing: YM, AO, RC, ME, YT, YN, HI, KN, TK, RO, YH, ST, SO, KS, KA.

The present study was supported by the Tokyo University of Science Grant for President’s Research Promotion, JST Moonshot R&D (grant no. JPMJMS2022-21), Grant-in-Aid for Scientific Research (C) (grant no. 20K08936, 24K12517), Grant-in-Aid for Research Activity Start-up (grant no. 21K20732), Grant-in-Aid for Early-Career Scientists (grant no. 23K14352), JST SPRING (grant no. JPMJSP2151), Grant-in-Aid for Special Research in Subsidies for ordinary expenses of private schools from The Promotion and Mutual Aid Corporation for Private Schools of Japan, Grant from Institute for Environmental & Gender-specific Medicine, Juntendo University.

No artificial intelligence (AI) tools, including large language models or machine learning software, were used in the preparation, analysis, or presentation of this manuscript.