Triple-negative breast cancer (TNBC) is a highly aggressive breast cancer subtype that has heterogeneous clinical and molecular characteristics (1, 2). Approximately 12% to 17% of breast cancer cases are TNBC, defined as tumors that lack expression of estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor type 2 (HER2) (3). The lack ER and PR expression makes it difficult for TNBC patients to benefit from endocrine therapy, while the negative expression of HER2 prevents effective targeted therapy. This results in worse prognosis for TNBC patients compared with other breast cancer subtypes. Therefore, identifying and validating novel targeted therapies and prognostic indicators for TNBC has become an urgent need for breast cancer research.

The prognostic significance of neuroendocrine (NE) differentiation in breast cancer has been controversial, partially because the classification criteria for neuroendocrine neoplasms (NENs) have evolved over the years or the limited numbers of NEN cases in breast cancer. Neuroendocrine carcinoma (NEC) of the breast was formally proposed as a distinct entity in 2003, defined as a subtype of invasive mammary carcinoma in which >50% of the tumor cells express NE markers (4, 5). The most commonly used NE markers in the breast include synaptophysin (Syn) or chromogranin A (CgA) (6, 7). The 2012 WHO Classification later stated that the 50% threshold for NE marker positivity was no longer required because of its arbitrary nature as a criterion (6). The newly published WHO Classification in 2019 emphasized that the key features of NENs are neuroendocrine histological features and expression of NE markers (8). Some cohort studies have explored the prognostic significance of NE markers in breast cancer. The majority of the studies have shown that breast cancer patients with NE features have worse prognoses compared with those without NE features (4, 9–11), while some studies have reported better prognoses (12–14) or no significant difference (9, 15–17). Currently, there were no reports about the prognosis of TNBC patients with NE marker expression.

The first-line therapy for NENs in breast cancer is mostly surgery combined with or without adjuvant radiotherapy, chemotherapy, or neoadjuvant chemotherapy. This is based on the therapeutic principles of other types of breast cancer, and the cases were treated with endocrine or targeted therapy according to the ER/PR and HER2 status. However, the relevant studies for the therapy of NENs in breast cancer are mainly case reports or case series (8, 18–22). In recent years, several studies have identified immunohistochemistry (IHC) markers that have prognostic or therapeutic significance for NENs, including ATRX, MGMT, SSTR2, and PD-L1 (23–30). However, there are no studies on therapeutic regimen for NENs in TNBC. It remains to be explored whether TNBCs with positive expression of NE markers can be treated as NENs at any other anatomical site.

Thus, we utilized two most common used NE markers, Syn and CgA, to evaluate their expression levels in TNBC patients and observe their correlations with clinicopathological parameters. Additionally, we sought to determine the prognostic significance of NE marker expression in TNBC patients and explore the potential treatment options for such individuals.

Because of the highly malignant nature, poor prognosis, and high heterogeneity associated with TNBC, researchers have been committed to identifying prognostic biomarkers and therapeutic targets specific to this disease. A large retrospective cohort study that collected data from nearly 160,000 breast cancer patients from the National Cancer Institute in 2017 showed that TNBC patients had a worse prognosis than non-TNBC patients of the same stage, adjusted for age, race, tumor stage, surgery, and radiation. These findings confirmed the importance of introducing biomarkers to the traditional TNM staging to help predict breast cancer patient prognosis, as well as the necessity of further exploring TNBC-related issues (32).

To our knowledge, our study is the first and the largest retrospective cohort study to explore the prognostic significance of NE markers in TNBC. We aimed to provide new biomarkers that can help predict TNBC patient survival and assist clinicians with treatment decisions.

Among the 396 TNBC patients included in this study, positive NE marker expression was observed in 30 cases (7.6%). 20 cases had <10% of tumor cells positive for NE markers. Eight cases had 10%–89% of tumor cells positive for NE markers. Two cases had ≥90% of tumor cells positive for NE markers. The WHO Classification of Tumours of the Breast in 2019 suggested that cancers with more than 90% NEN pattern should be classified as neuroendocrine tumor or NEC, while cancers with less than 10% should be classified as invasive NST or other types. Cancers with 10% to 90% NEN pattern should be classified as mixed invasive (NST or other special type) and NEC, and the percentage of NE component should be reported (8). According to the WHO Classification in 2019, two of these 30 cases were classified as NENs. Seven cases were classified as mixed invasive NST and NEC and one case were classified as mixed invasive micropapillary carcinoma and NEC. According to the 2012 WHO Classification of Tumours of the Breast, NED can be detected in approximately 30% of IBC-NST cases and in some special types (especially mucinous carcinoma) by IHC staining (6). However, because of different definitions, the incidence of NENs in all types of breast cancer has been previously reported to be as low as < 0.1% to as high as nearly 20% (9, 11, 13, 15, 16). For the TNBC subtype in our cohort, the percentage of cases with positive NE marker expression (7.6%) was lower than the incidence of NED in all breast cancers (30%) indicated by the 2012 WHO Classification of Tumours. Additionally, the incidence of NENs was very low according to the 2019 WHO Classification of Tumours definition. Only two cases had more than 90% of NE marker-positive tumor cells among the 396 TNBC patients. These results are also consistent with the previous findings that NED is significantly correlated with ER and PR expression (13, 33).

According to the K-M survival analysis for the TNBC patients stratified by stage in our study, we found that those with positive NE marker expression had higher DFS rates than those without NE marker expression in the same stage. To our knowledge, no previous cohort study has reported on the prognostic effect of NE markers in TNBC patients. In general, sufficient studies have confirmed that NED in breast cancer implies a poor prognosis (4, 9–11). Actually, our result does not conflict with the conclusions of previous studies. The objects of previous studies were breast cancer patients, which were dominated by hormone-positive cases, and the numbers of TNBC cases were small. While our study focused on TNBC, a breast cancer subtype with a worse prognosis relative to others. Therefore, we speculated that the influence of NE markers on prognosis of breast cancer patients might be related to their hormonal status.

In addition, our study showed that positive Syn expression has a better prognostic significance than CgA. Previous work has concluded that NED can be confirmed using IHC stains for several markers, including Syn, and CgA. Each marker has different degrees of specificity and sensitivity: Syn is highly sensitive, but not entirely specific. CgA is highly specific, but less sensitive than Syn (34). In our study, 6.1% and 1.9% of TNBC cases showed positive IHC staining for Syn and CgA, respectively. Positive Syn expression significantly prolongated the DFS in TNBC patients compared with Syn-negative TNBC patients at the same stage, but no significant difference in DFS was found for CgA expression. Our study concludes that Syn positivity in TNBC was higher than that of CgA, and the prognostic significance of Syn was greater than that of CgA for TNBC patients. The results of our study may provide guidance for clinical identification of TNBC patients with better prognoses.

Furthermore, we observed a significant negative correlation between NE marker expression and basal-like marker expression in our cases. Previous literature suggests that positive expression of basal-like markers can be used as adverse prognostic factors for TNBC (35–37). These conclusions are consistent with our results of positive NE marker expression being associated with a better prognosis in TNBC, as well as the observed negative correlation between NE marker and basal-like marker positivity.

After further exploring the 30 TNBC cases with positive NE marker expression, we found that 13.3% of these cases were negative for ATRX expression, indicating poor prognosis (23, 24). 26.7% of these cases were negative for MGMT expression, which may response to temozolomide (25–27). Additionally, 13.3% of these cases had a 3+ score for SSTR2 IHC staining, suggesting poor prognosis but also the potential to be treated with somatostatin analogues (28). The TPS of PD-L1 was higher than 10% in 10.0% of these 30 cases, meaning that a small percentage of the NE marker-positive TNBC patients may benefit from immunotherapy (29, 30). However, because of the low levels of NE marker expression in TNBC and the small number of cases included, further investigation is warranted regarding the positive rate of IHC staining for these markers and their prognostic and therapeutic implications. A meta-analysis involving 49,425 cancer patients suggested that immune checkpoint inhibitors (ICIs) may increase the possibility of complete remissions compared to control treatments (38). Nevertheless, only a subset of breast cancer patients benefit from ICIs, necessitating exploration into reliable predictors of response. A review discussing potential predictors of response to ICIs in TNBC indicated that PD-L1 could serve as an important marker for selecting TNBC patients who would derive greater clinical benefits from immunotherapy. However, relying solely on PD-L1 assessment might exclude certain TNBC patients who could potentially respond to immunotherapy. More efforts should be devoted to evaluating novel biomarkers for predicting response to ICIs in TNBC patients, such as PD-L1 expression, tumor mutational burden, microsatellite instability status, etc. (39) Recent clinical trials have demonstrated the benefits of various immune-based combinations for metastatic TNBC patients; thus chemoimmunotherapy has been approved as a new first-line treatment option for metastatic TNBC patients with elevated CPS or PD-L1 overexpression. Several clinical studies potentially modifying the landscape of first-line treatment options for TNBC are still ongoing (40).

This study has some limitations. Differences in the time from disease onset to when surgical treatment was administered, the surgical methods received, and the adjuvant therapies became confounding factors that affected our prognostic evaluation. We reduced the effect of the different treatment regimens on prognosis by excluding patients who received neoadjuvant therapy.

In conclusion, expression of NE markers, especially Syn, in TNBC indicated a better prognosis compared with cases negative for NE marker expression at the same stage. There was a significant negative correlation between NE marker expression and basal-like marker expression in TNBC. The use of NE markers as biomarkers for predicting TNBC patient prognosis requires further verification. Additionally, we found that a small number of TNBC cases with positive NE marker expression also showed positive IHC staining for PD-L1 and SSTR2, suggesting that these patients may potentially benefit from immunotherapy or somatostatin analogue treatment.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving humans were approved by the Ethics Committee of Peking Union Medical College Hospital. The studies were conducted in accordance with the local legislation and institutional requirements. The human samples used in this study were acquired from primarily isolated as part of your previous study for which ethical approval was obtained. Written informed consent for participation was not required from the participants or the participants’ legal guardians/next of kin in accordance with the national legislation and institutional requirements.

CX, ZL, and XR contributed to the conception and design of the study. SS, JP, LC, and JY acquired and organized the data for the work. CX performed the statistical analysis and wrote the first draft of the manuscript. XR wrote sections of the manuscript. All authors contributed to the article and approved the submitted version.