Breast cancer currently accounts for 30% of newly diagnosed malignant tumors in women and causes 15% of women to die from cancer (1). For the first time, Perou described the intrinsic molecular subtypes of breast cancer and described Triple-negative breast cancer (TNBC) in 2000 using complementary DNA microarray technology (2). Furthermore, TNBC is the most aggressive subtype of breast cancer, accounting for about 10–20% of breast cancer cases (3, 4). TNBC is still unsatisfactory in diagnosis and treatment.

Bibliometrics is a quantitative analysis method of academic publications, which can discover the progress of discipline research from a macro perspective and provide support for future research directions (5). TNBC-related literature information analysis is scarce. Teles et al. (6) conducted a bibliometric study of 1,932 publications in 2018 to study nanomedicine research’s global trend on TNBC. However, the inclusion criteria of this study are too broad, and the analysis methods are insufficient to analyze the status quo of the TNBC study. Unfortunately, bibliometric studies on TNBC remain insufficient due to the lack of practical language analysis tools to integrate metatext data.

Natural Language Processing (NLP) is a computing technology used to analyze human language, a part of machine learning (7). Various algorithms have been successfully applied to deal with medical information (8). Latent Dirichlet Allocation (LDA) is bibliometrics’s most classical topic modeling method to present many unstructured texts and information (9, 10). LDA can perform topic analysis on texts (5). We recently constructed LDA and NLP methods to analyze more than 23,000 rectal cancer-related publications between 1994 and 2018. We have found the research deficiencies in the last 25 years and predicted the future research focus (11). Therefore, through the use of mature LDA methods and machine learning techniques to discover the current research from a macro perspective, at the same time discover the missing research topics in the past, and predict potential research breakthroughs in the future.

We analyzed all past TNBC publications indexed by PubMed under Triple-negative breast cancer in the present study. We improved our algorithm based on previous research and conducted a more detailed analysis of all TNBC publications with more visual expression to highlight current research focus in TNBC, research deficiencies, and specific areas with future opportunities.

We analyzed 16,826 publications in the field of TNBC from 2005 to 2022 using machine learning and NLP. Furthermore, we visualize and analyze the results from a macro perspective. Over the past 17 years, we found that TNBC-related publications have increased from none to 16,826 in 2021, with more extensive research content. TNBC research focuses on Therapeutic target research, Prognostic research, and Mechanism research. Research topics have changed over the years, and the current research focus is expected to be Therapeutic target research and Nanoparticle research, according to our LDA results.

Bibliometrics is a compelling analysis method to obtain information from massive texts quantitatively, and there are very few bibliometrics analyses on TNBC such as VOSviewer, Bibliographic Items Co-occurrence Matrix Builder (BICOMB), and CiteSpace. However, with the development of the publishing industry, these tools have difficulty applying to massive publication analysis due to their architecture, insufficient computer memory, and sharing protocols. Therefore, our research uses the LDA algorithm based on Python, an unsupervised topic model. Furthermore, our topic model is based on the publication’s abstract, not on the keywords. It is easy to use with negligible memory consumption and can analyze massive publications.

We found that Therapeutic target research has always been research-focused because TNBC lacks effective therapeutic targets and has high heterogeneity (24, 25). Our research found that this part contains a variety of attempts, DNA repair research, immune checkpoint research, and protein expression. We only found 137 publications related to immune checkpoint research, and immunotherapy research is not closely related to the prognosis and mechanism research of TNBC. Several clinical studies are being carried out, including IMpassion130, KEYNOTE-355, and Impassion 131 (26–28). Some positive results can reduce the risk of death by up to 35%. However, more important is the research on the underlying mechanism and the exploration of various influencing factors, especially the extracellular matrix, hypoxia, and immune cell infiltration (29). In addition, immune checkpoint research has just started for five years, according to our results, and several medications have already been applied in the clinic. This research trend will continue, and immunotherapy will become a safe and effective treatment option.

The research scope of the TNBC mechanism is pervasive, covering the immune microenvironment and subtypes of TNBC. The successful subtyping provides a solid theoretical basis for the precision therapy of TNBC (30). Gene sequencing technology allows us to fully understand the mutation rate of TNBC, which is about 1.68 bp/Mb (31). Mutations occur in genes in multiple key signaling pathways such as PI3K/Akt/mTOR pathway, RAS/RAF/MEK pathway, JAK/STAT pathway, DNA repair pathway, and cell cycle checkpoint (32–34). Therefore, various treatments targeting the signal pathways are currently undergoing clinical trials. Some inhibitors have been used as potential medications for TNBC treatment, including PI3K, MEK, PARP, EGFR, VEGF, and AR inhibitors (32).

Triple-negative breast cancer subtyping has always been the focus of research. There is no unified standard based on the TNBC genome and cell heterogeneity. The first classification was based on Lehmann’s gene expression analysis of breast cancer and constructed a “triple negative classification” and six subclassifications (24). In 2016, Lehmann’s further research found that immunomodulatory (IM) patients are more likely to benefit from checkpoint inhibitor therapy (35). With the advancement of technology, such as the emergence of single-cell RNA sequencing, spatial transcriptomics, and radionics, and the further expansion of data volume, new technologies have provided new insights into the typing of TNBC and proposed guidance for treatment. Xie’s research established a new prognostic model through the comprehensive analysis of multiple cell death patterns on more than 1,000 breast cancer patients, which can predict the clinical prognosis and drug sensitivity after TNBC surgery (36). In addition to technological progress, an in-depth understanding of the oncological course, mechanism of occurrence and development, and algorithm advances will provide a more detailed classification of TNBC.

On the other hand, studies on operations and radiotherapy were rarely reported, especially for re-operations related to local-regional recurrence risk or distant metastasis. Many studies suggest that surgery is essential in treating distant metastases of cancers, such as colorectal cancer (37). In addition, many studies on other cancers, including pancreatic and colorectal cancer, demonstrated that the tumor microenvironment, especially the extracellular matrix, has been found to play an essential role in cancer metastasis, local recurrence, and chemotherapeutic drug resistance (38, 39). Many potential drugs are used due to their ability to target the extracellular matrix, such as PEGPH20 (an enzyme that targets matrix hyaluronic acid), pegilodecakin (a PEGylated IL-10) (40, 41). However, the study on extracellular matrix in TNBC is insufficient so far.

Although the research on TNBC has made significant progress in many aspects, the present research also found some research deficiencies on TNBC. There is a lack of research on TNBC from patients’ perspectives, health economics, and hospice care. Although, at present, the 5 years overall survival rate of most tumors has been dramatically improved, helping tumor patients with psychological issues re-enter society will become a new important research topic (42). TNBC patients are more likely to relapse and metastasize than other breast cancer subtypes, resulting in more significant mental and economic pressure on patients and their families. Studies on patients with more prolonged survival can better understand TNBC and even other long-term survival tumors (43). In the future, we will face more challenges for patients with a long survival period of 5–10 years (44).

There are some limitations in the present study. Besides PubMed, several other databases, including Scopus, Web of Science, and Embase, could be used for bibliometric research. Although PubMed contains the highest quality peer-reviewed research and excludes irrelevant, non-peer-reviewed publications, the literature will provide detailed and comprehensive knowledge if other databases are explored simultaneously. Secondly, we considered that all publications publish more positive research results. Negative results and clinical participants’ perspectives are naturally more difficult to be published. With the development of complete medical record texts, publication databases, and improved algorithms, it is reasonable for machine learning to play a more active auxiliary role in future clinical practice. The data presented in this study will hopefully help scientists understand the current status of TNBC research and design more relevant basic and clinical research projects.

We analyzed 16,826 TNBC publications through the NLP Method. TNBC research shows insufficiencies, especially in long-term survival-related research, and a lack of research from patients’ perspectives. The publications mainly focused on three aspects: Therapeutic target research, Prognostic research, and Mechanism research. The research direction of TNBC may require the intervention of new technologies.

The original contributions presented in this study are included in the article/Supplementary material, further inquiries can be directed to the corresponding authors.

KW initiated the project, analyzed the data, constructed analytical methods, and wrote the primary manuscript draft. XD initiated and supervised all aspects of the project and wrote the primary manuscript draft. CZ performed statistical analyses and contributed to the manuscript writing. DD helped interpret results and contributed to the statistical analyses. LZ contributed to the manuscript’s revision in terms of writing and interpretation. ML contributed to the interpreting results and supervising statistical analyses. All authors contributed to the manuscript writing and read and approved the final version of the manuscript.