Bibliometric analysis of research trends and hotspots in immunotherapy biomarkers for non-small cell lung cancer from 2015 to 2024

Abstract

Background

Immunotherapy has revolutionized the treatment of non-small cell lung cancer (NSCLC), offering promising alternatives to traditional therapies. This study aimed to conduct a bibliometric analysis to identify research hotspots and emerging themes in immunotherapy biomarkers for NSCLC from 2015 to 2024.

Methods

Articles and reviews from the Web of Science Core Collection (January 1, 2015–December 31, 2024) were retrieved on June 23, 2025. Networks of countries, institutions, authors, journals, co-cited references, and keywords were constructed using CiteSpace and VOSviewer, with visualizations created using Scimago Graphica.

Results

A total of 2134 publications by 14,723 researchers from 73 countries and 3670 institutions, across 427 journals, were included. The number of annual publications increased steadily, peaking at 404 in 2024. China led in publication volume, while the U.S. had the highest citation impact. Tongji University was the most productive institution, and Caicun Zhou was the most prolific author. Among the journals, Cancers had the highest publication volume, while Lung Cancer had the highest citation frequency. The key topics identified included immune checkpoint inhibitors (ICIs), programmed death-ligand 1 (PD-L1), tumor mutational burden (TMB), liquid biopsy, and combination therapy. Emerging fields included the tumor microenvironment (TME), radiomics, proliferative signaling, and the gut microbiota.

Conclusion

This bibliometric analysis provides a comprehensive overview of NSCLC immunotherapy biomarker research, highlighting key trends and emerging directions. These insights are crucial for advancing precision immunotherapies and biomarker-driven strategies.

Graphical Abstract

Introduction

Non-small cell lung cancer (NSCLC) remains the leading cause of cancer-related morbidity and mortality worldwide, accounting for approximately 85% of all lung cancer cases [1–3]. Although conventional treatment modalities—including surgery, chemotherapy, and radiotherapy—provide clinical benefit for selected patients, their impact on overall survival (OS) in advanced disease remains limited. In metastatic NSCLC, marked therapeutic heterogeneity and the development of intrinsic or acquired resistance further complicate disease management, underscoring the urgent need for personalized treatment strategies and reliable predictive biomarkers [4, 5].

In recent years, immunotherapy has emerged as a transformative therapeutic approach for patients with advanced NSCLC. Among these strategies, immune checkpoint inhibitors (ICIs) targeting programmed cell death protein 1 (PD-1) and its ligand programmed death-ligand 1 (PD-L1) have fundamentally reshaped treatment paradigms. The KEYNOTE-024 trial demonstrated that pembrolizumab significantly prolonged progression-free survival (PFS) compared with platinum-based chemotherapy in treatment-naïve patients with a PD-L1 tumor proportion score (TPS) ≥ 50% (10.3 vs. 6.0 months), while also reducing the incidence of grade 3–5 adverse events [6]. Similarly, KEYNOTE-042 reported an OS benefit in patients with TPS ≥ 1% (16.7 vs. 12.1 months), with the greatest benefit observed in those with TPS ≥ 50%, together with a favorable safety profile [7]. In previously treated nonsquamous NSCLC, the CheckMate-057 trial further confirmed that nivolumab significantly improved long-term survival compared with docetaxel (5-year OS: 13.4% vs. 2.6%), while reducing severe treatment-related toxicity [8]. Beyond the KEYNOTE and CheckMate studies, the EMPOWER-Lung 1 trial demonstrated that, in patients with advanced NSCLC with PD-L1 expression ≥ 50%, cemiplimab monotherapy significantly improved OS and PFS compared with chemotherapy (median OS: 26.1 vs. 13.3 months; median PFS: 8.1 vs. 5.3 months), achieving a 5-year OS rate of 29.0% [9]. Collectively, these landmark studies have established ICIs as a standard-of-care option for selected patients with advanced NSCLC.

Importantly, however, the clinical benefit of ICIs is not uniform across all molecular subtypes of NSCLC. Accumulating evidence indicates that patients harboring oncogenic driver alterations—particularly epidermal growth factor receptor (EGFR) mutations and anaplastic lymphoma kinase (ALK) rearrangements—derive limited benefit from PD-1/PD-L1 blockade compared with patients with driver-negative tumors [10, 11]. Consistent with these observations, patients with EGFR or ALK alterations were excluded from pivotal trials such as KEYNOTE-024. Given the relatively high prevalence of these driver mutations in NSCLC, EGFR and ALK alterations are widely regarded as negative predictive biomarkers for immunotherapy efficacy, highlighting the complexity of biomarker-guided treatment selection.

Beyond NSCLC, immunotherapy has also demonstrated clinically meaningful efficacy in other thoracic malignancies. ICIs have become integral components of treatment for extensive-stage small cell lung cancer (SCLC) and malignant pleural mesothelioma, where they have shown survival benefits either as monotherapy or in combination with chemotherapy [12, 13]. These advances underscore the broader relevance of immunotherapy across thoracic oncology and reinforce the importance of identifying robust and reproducible biomarkers to optimize patient selection and therapeutic outcomes.

As the clinical application of immunotherapy continues to expand, marked inter-patient variability in therapeutic response and immune-related adverse events (irAEs) has emphasized the critical role of predictive biomarkers. Such biomarkers not only help identify patients most likely to benefit from treatment but also aid in recognizing individuals at increased risk of toxicity, thereby minimizing unnecessary exposure and improving clinical outcomes. Current evidence supports the predictive value of PD-L1 expression, tumor mutational burden (TMB), and microsatellite instability (MSI) [14, 15]. In parallel, emerging technologies such as liquid biopsy provide non-invasive approaches for dynamic disease monitoring and real-time assessment of treatment response, further broadening the scope of biomarker-driven immunotherapy [16–18].

Despite these advances, substantial challenges remain. Tumor heterogeneity, the complexity of the tumor microenvironment (TME), and the emergence of primary or acquired resistance continue to limit the durability of immunotherapy responses [19–22]. Moreover, the clinical translation of both established and emerging biomarkers is hindered by assay variability, limited standardization, insufficient prospective validation, and issues related to cost and accessibility.

To comprehensively characterize the evolution of this rapidly expanding field, the present study employs a bibliometric analysis to systematically evaluate NSCLC immunotherapy biomarker research published between 2015 and 2024. By mapping publication trends, identifying influential authors and institutions, and highlighting emerging research hotspots—including biomarker development, resistance mechanisms, and the tumor microenvironment—this study aims to provide an integrated overview of the knowledge landscape and to inform future research directions in precision immunotherapy for NSCLC.

Materials and methods

Data source and search strategy

This study utilized the Web of Science Core Collection (WoSCC) [23], a globally recognized and authoritative database that is widely used in bibliometric research and indexes a large number of peer-reviewed articles [24]. To improve both the precision and comprehensiveness of literature retrieval, Medical Subject Headings (MeSH) terms and their related entry terms were incorporated into the search strategy.

The focus of this study was research on immunotherapy biomarkers in NSCLC. Relevant keywords, including “non-small cell lung cancer,” “immunotherapy,” and “biomarker,” were used to construct a comprehensive search strategy based on MeSH terms and the study objectives (Fig. 1). The literature search was conducted on June 23, 2025, and was restricted to English-language publications published between January 1, 2015, and December 31, 2024. This time frame was selected because 2015 marked the approval of pembrolizumab and nivolumab by the U.S. Food and Drug Administration (FDA) for the treatment of NSCLC, signaling the integration of immunotherapy into routine clinical practice and triggering rapid growth in related research output.

Fig. 1.

Flowchart of the search strategy and screening process

Only peer-reviewed original research articles and review articles were included in this analysis. Conference abstracts, editorials, book chapters, and other non-article document types were excluded. Conference abstracts, including late-breaking results presented at major international meetings such as the American Society of Clinical Oncology (ASCO) and the European Society for Medical Oncology (ESMO), were excluded for the following reasons:

1. Data integrity: Conference abstracts often lack complete reference lists, detailed methodological descriptions, and long-term follow-up data, which are essential for reliable citation and co-citation analyses using bibliometric tools such as CiteSpace and VOSviewer.

2. Reproducibility: Abstracts frequently report preliminary findings that may be modified or withdrawn prior to full publication, thereby limiting transparency and reproducibility.

Based on these inclusion and exclusion criteria, 2134 publications were ultimately included for analysis from an initial set of 2566 records. The detailed process of literature retrieval and screening is illustrated in Fig. 1.

Data analysis

A combination of bibliometric tools and statistical methods was employed to systematically analyze the literature on NSCLC immunotherapy biomarkers published between 2015 and 2024, with the aim of identifying research trends, collaboration patterns, and emerging research hotspots. Data processing and visualization were primarily performed using four software platforms: CiteSpace, VOSviewer, Scimago Graphica, and Microsoft Excel. Microsoft Excel was used to visualize overall publication trends and to evaluate the goodness of fit (R²) of regression models.

VOSviewer was mainly applied to construct and visualize bibliometric networks, including co-citation networks, keyword co-occurrence networks, and collaboration networks [25]. This software facilitated the identification of core research topics as well as collaboration relationships among countries, institutions, and authors. Specific analyses included:

1. Collaboration networks among countries and institutions;

2. Journals and co-cited journals;

3. Authors and co-cited authors; and

4. Clustering analysis of keyword co-occurrence.

To ensure clarity and interpretability of the generated network visualizations, parameters such as resolution and minimum cluster size were adjusted when necessary.

CiteSpace was used for in-depth exploration of knowledge structure evolution and temporal trends [26]. By selecting the “Full Record with Cited References” option, time-sliced analyses were conducted to examine changes in research themes across different periods. Cosine similarity was used to calculate the strength of connections between nodes, and the top 10% most influential nodes in each time slice were visualized. In addition, the Pathfinder pruning algorithm was applied to remove redundant or weak links, thereby optimizing network structure and highlighting the academic influence and developmental trajectories of key publications.

Scimago Graphica, an emerging interactive visualization tool, was used to present complex bibliometric networks. Its advantages in graphical representation, interactivity, and visual aesthetics enhanced the presentation of results, particularly in displaying multi-level and multidimensional network structures.

Quality control and visualization

To ensure the quality and readability of all visual outputs, parameters such as scale, labeling, and line thickness were dynamically adjusted. All figures and network maps were carefully reviewed to confirm clarity, consistency, and interpretability. These refinements in visualization helped to effectively convey the major trends, relationships, and structural characteristics identified within the dataset.

Results

Database search and data overview

A total of 2134 publications related to NSCLC immunotherapy biomarkers were retrieved, authored by 14,723 researchers from 73 countries and affiliated with 3670 institutions. These publications were distributed across 427 journals and cited literature from 4764 unique journals, yielding a total of 49,084 references. This comprehensive dataset provides a robust foundation for mapping the evolving global research landscape of immunotherapy biomarkers in NSCLC.

Analysis of publication trends

As shown in Fig. 2, the annual number of publications on NSCLC immunotherapy biomarkers increased markedly from 2015 to 2024. The year 2015 represented a pivotal milestone, coinciding with the FDA approvals of nivolumab and pembrolizumab for NSCLC, which facilitated the clinical adoption of immunotherapy and stimulated a sharp increase in related research output.

Fig. 2.

Annual publication trends of immunotherapy biomarkers for NSCLC (2015–2024)

Following this milestone, publication output exhibited a sustained upward trajectory, peaking in 2021, experiencing modest fluctuations thereafter, and ultimately reaching a new record high of 404 publications in 2024. The fitted exponential growth curve demonstrated excellent goodness of fit (R² = 0.9984), indicating a continuous and intensifying scholarly interest in NSCLC immunotherapy biomarker research.

Analysis of contributions by countries

Research on NSCLC immunotherapy biomarkers demonstrated broad international participation, with contributions from 73 countries. As illustrated in Fig. 3A, the United States, China, and the European Union (EU) emerged as the leading contributors. The United States recorded the highest total citation count (31,523 citations), reflecting strong academic influence, whereas China produced the largest number of publications (770 articles), indicating high research productivity. When aggregated, the EU ranked prominently in both publication volume (1009 articles) and total citations (47,845 citations), suggesting substantial cumulative impact. Because the EU represents multiple member states, direct comparisons with individual countries should be interpreted with caution.

Fig. 3.

Global publications and collaborations on NSCLC immunotherapy biomarkers. A Geographic map of international cooperation (circle size = publication volume; colors = collaboration clusters; line thickness = strength of cooperation). B Timeline of the top 10 countries (circle size = annual publications; color intensity = citation counts). C Chord diagram of international collaborations (ribbon width = co-publication strength). D Network of the top 27 countries (node size = publication output; color = regional clusters; line thickness = collaboration intensity). E Overlay map showing temporal evolution (node color = average publication year; blue/purple = earlier; yellow/green = more recent)

Temporal trends shown in Fig. 3B indicate that the United States and China consistently maintained leading positions throughout the study period, while several European countries and Japan exhibited steady growth. Since 2020, global research activity has expanded markedly, reflecting increasing worldwide attention to immunotherapy biomarkers in NSCLC.

International collaboration networks presented in Figs. 3C and D reveal strong cooperative relationships, with thick connecting lines highlighting close collaborations among the United States, China, and European countries. The top 27 contributing countries and regions formed extensive collaborative structures, particularly emphasizing robust connections between East Asia and Europe. Figure 3E further illustrates China’s growing academic influence, largely driven by advances in molecular diagnostics and translational research.

Overall, the global research landscape of NSCLC immunotherapy biomarkers is characterized by multi-center leadership and extensive international collaboration. The United States and China remain dominant contributors, the EU demonstrates substantial collective influence, and other countries continue to strengthen their research presence in this rapidly evolving field.

Analysis of contributions by institutions

The institutional collaboration network, comprising 49 major institutions, is shown in Fig. 4A, where node size represents publication volume and line thickness indicates collaboration strength. Two major collaboration clusters were identified: one cluster located on the left side of the network (red) and another on the right side (green).

Fig. 4.

Institutional analysis in the field of immunotherapy biomarkers for NSCLC. A Network visualization map. B Overlay visualization map

Institutions with high productivity and extensive collaborative connections included Tongji University, Shanghai Jiao Tong University, and the Chinese Academy of Medical Sciences & Peking Union Medical College in China, as well as the University of Texas MD Anderson Cancer Center in the United States. These institutions functioned as central hubs within their respective collaboration networks, reflecting both strong research capacity and high levels of inter-institutional cooperation. As shown in Fig. 4B, the influence of Chinese institutions in this research field has increased steadily over time.

Analysis of contributions by prolific and co-cited authors

A total of 14,723 authors contributed to research on NSCLC immunotherapy biomarkers. The top 10 most prolific authors are listed in Table 1. Caicun Zhou ranked first in publication output, with 31 publications, indicating sustained productivity in this field. Luis Paz-Ares ranked first in total citations (1413 citations), reflecting the high academic influence of his work, particularly in shaping clinical practice. Mark Awad also demonstrated notable impact, with the highest average citation rate (76.06 citations per publication) among the top contributors.

Table 1.

Top 10 most prolific authors in the study of NSCLC immunotherapy biomarkers

Rank	Author	Affiliations	Documents	Citations	Average citations	H-index
1	Zhou, Caicun	Department of Oncology, Tongji University, Shanghai Pulmonary Hospital, China	31	1292	41.68	18
2	Zhang, Li	National Cancer Center / Chinese Academy of Medical Sciences, China	24	540	22.50	13
3	Garassino, Marina Chiara	Department of Oncology, University of Milan, Italy	21	789	37.57	11
4	Paz-Ares, Luis	Hospital Universitario 12 de Octubre, Universidad Complutense, Spain	20	1413	70.65	11
5	Besse, Benjamin	Gustave Roussy Cancer Campus, Villejuif, France	20	1255	62.75	14
6	Ferrara, Roberto	European Institute of Oncology (IEO), Milan, Italy	20	613	30.65	11
7	Zhao, Jing	Cancer Hospital, Chinese Academy of Medical Sciences, China	19	647	34.05	11
8	Hofman, Paul	Université Côte d’Azur, Nice University Hospital, France	18	521	28.94	10
9	Prelaj, Arsela	Department of Oncology, University of Milan, Italy	18	462	25.67	10
10	Awad, Mark	Dana-Farber Cancer Institute, Harvard Medical School, USA	17	1293	76.06	13

As shown in Fig. 5A, Caicun Zhou remained highly active through 2023, followed by Arsela Prelaj, who made substantial contributions in 2022. Author collaboration networks presented in Figs. 5B and C reveal increasingly close cooperative relationships among researchers from multiple countries, underscoring the growing importance of international collaboration in advancing NSCLC immunotherapy biomarker research.

Fig. 5.

Author analysis in the field of immunotherapy biomarkers for NSCLC. A Temporal trend of publications by the top 10 authors. B Network visualization map. C Overlay visualization map

Analysis of contributions by journals

To characterize the publication landscape and disciplinary distribution of research on immunotherapy biomarkers in NSCLC, journal contributions from 2015 to 2024 were analyzed. During this period, relevant studies were published in a total of 427 journals, highlighting the broad and highly interdisciplinary nature of this research field. The top 10 journals, ranked by publication volume and co-citation frequency, are summarized in Table 2, reflecting both research productivity and academic influence.

Table 2.

Top 10 publishing journals and top 10 co-cited journals in NSCLC immunotherapy biomarker research

Rank	Journals	Documents	Country	IF(JCR2024)	Cited journals	Citations	Country	IF(JCR2024)
1	Cancers	143	Switzerland	4.4	The New England Journal of Medicine	7425	USA	78.5
2	Frontiers in Oncology	97	Switzerland	3.3	Journal of Clinical Oncology	6009	USA	41.9
3	Frontiers in Immunology	91	Switzerland	5.9	Journal of Thoracic Oncology	5533	USA	20.8
4	Translational Lung Cancer Research	83	China	3.5	Clinical Cancer Research	3927	USA	10.2
5	Lung Cancer	76	Netherlands	4.4	Annals of Oncology	3783	Netherlands	65.4
6	Journal for Immunotherapy of Cancer	70	United Kingdom	10.6	The Lancet Oncology	2654	United Kingdom	35.9
7	Clinical Lung Cancer	53	USA	3.3	Lung Cancer	2587	Netherlands	4.4
8	Thoracic Cancer	46	China	2.3	The Lancet	2155	United Kingdom	88.5
9	Journal of Thoracic Oncology	41	USA	20.8	Nature	2078	United Kingdom	48.5
10	Cancer Immunology, Immunotherapy	35	USA	5.1	Science	1880	USA	45.8

Cancers ranked first in publication output, with 143 articles (impact factor: 4.4), underscoring its role as a major outlet for studies in this domain. Although the Journal of Thoracic Oncology exhibited the highest impact factor among the top 10 journals (20.8), it published a comparatively smaller number of articles (41), suggesting a focus on highly selective, high-impact contributions. The New England Journal of Medicine (NEJM) ranked first in total citations (7425 citations; impact factor: 78.5), highlighting its central role in disseminating landmark clinical evidence related to lung cancer immunotherapy.

To further explore intellectual flows and interdisciplinary interactions, a dual-map overlay visualization was applied (Fig. 6A). Three major citation pathways were identified. One pathway links molecular biology and genetics journals to molecular biology and immunology journals, reflecting the translational progression from basic research to immune mechanisms. Two additional pathways connect clinical and medical journals, emphasizing patient-centered innovation and therapeutic application. Collectively, these citation trajectories demonstrate the increasing integration of clinical medicine, molecular biology, and immunology in NSCLC immunotherapy research.

Fig. 6.

Journal analysis. A Dual-map overlay of journals. B Journal co-citation network. C Journal citation density visualization

The journal co-citation network shown in Fig. 6B is divided into three major clusters corresponding to clinical medicine, immunology, and basic science, which together constitute the foundational pillars of contemporary cancer research. This network highlights the central role of high-impact generalist journals and illustrates how NSCLC research has expanded beyond clinical treatment to encompass immune regulation and molecular mechanisms. Figure 6C further visualizes citation relationships among major journals using a heatmap, revealing strong citation interactions involving journals such as Lung Cancer, Clinical Cancer Research, and the Journal of Clinical Oncology. The identified research hotspots—immunotherapy, cell biology, and clinical research—reflect the rapid evolution and multidisciplinary convergence of lung cancer research.

Overall, journal-level analysis reveals a highly interconnected and translational publication ecosystem, providing a solid structural foundation for subsequent reference- and keyword-based analyses.

Analysis of references and co-cited references

To identify the intellectual foundations and developmental trajectory of research on immunotherapy biomarkers in NSCLC, a reference co-citation analysis was performed. This approach identifies publications that are frequently cited together, thereby revealing the core knowledge base and conceptual structure of the research field.

The top 10 most-cited references are presented in Table 3, the majority of which consist of landmark clinical trials published in high-impact journals. Among these, “Pembrolizumab versus chemotherapy for PD-L1–positive NSCLC (KEYNOTE-024)” and “Nivolumab versus docetaxel in advanced nonsquamous NSCLC (CheckMate-057)”, both published in The New England Journal of Medicine, ranked first and second, respectively. The prominence of these studies underscores their pivotal role in establishing immune checkpoint inhibitors as a cornerstone of NSCLC treatment.

Table 3.

The top 10 co-cited references in NSCLC immunotherapy biomarker research

Rank	Co-cited reference	Citations	Journal	Types
1	Pembrolizumab versus Chemotherapy for PD-L1-Positive Non-Small-Cell Lung Cancer. N Engl J Med. 2016 Nov 10;375(19):1823–1833. https://doi.org/10.1056/NEJMoa1606774	736	New England Journal of Medicine	Clinical trial
2	Nivolumab versus Docetaxel in Advanced Nonsquamous Non-Small-Cell Lung Cancer. N Engl J Med. 2015 Oct 22;373(17):1627–39. https://doi.org/10.1056/NEJMoa1507643	647	New England Journal of Medicine	Clinical trial
3	Pembrolizumab plus Chemotherapy in Metastatic Non-Small-Cell Lung Cancer. N Engl J Med. 2018 May 31;378(22):2078–2092. https://doi.org/10.1056/NEJMoa1801005	458	New England Journal of Medicine	Clinical trial
4	Nivolumab versus Docetaxel in Advanced Squamous-Cell Non-Small-Cell Lung Cancer. N Engl J Med. 2015 Jul 9;373(2):123–35. https://doi.org/10.1056/NEJMoa1504627	448	New England Journal of Medicine	Clinical trial
5	Mutational landscape determines sensitivity to PD-1 blockade in non–small cell lung cancer.Science348,124–128(2015). https://doi.org/10.1126/science.aaa1348	444	Science	Report
6	Atezolizumab versus docetaxel in patients with previously treated non-small-cell lung cancer (OAK): a phase 3, open-label, multicentre randomised controlled trial. Lancet. 2017 Jan 21;389(10,066):255–265. https://doi.org/10.1016/S0140-6736(16)32517-X	434	Lancet	Clinical trial
7	Pembrolizumab versus docetaxel for previously treated, PD-L1-positive, advanced non-small-cell lung cancer (KEYNOTE-010): a randomised controlled trial. Lancet. 2016 Apr 9;387(10,027):1540–1550. https://doi.org/10.1016/S0140-6736(15)01281-7	411	Lancet	Randomized controlled trial
8	KEYNOTE-001 Investigators. Pembrolizumab for the treatment of non-small-cell lung cancer. N Engl J Med. 2015 May 21;372(21):2018–28. https://doi.org/10.1056/NEJMoa1501824	379	New England Journal of Medicine	Clinical trial
9	Pembrolizumab versus chemotherapy for previously untreated, PD-L1-expressing, locally advanced or metastatic non-small-cell lung cancer (KEYNOTE-042): a randomised, open-label, controlled, phase 3 trial. Lancet. 2019 May 4;393(10,183):1819–1830. https://doi.org/10.1016/S0140-6736(18)32409-7	288	Lancet	Clinical trial
10	Pembrolizumab plus Chemotherapy for Squamous Non-Small-Cell Lung Cancer. N Engl J Med. 2018 Nov 22;379(21):2040–2051. https://doi.org/10.1056/NEJMoa1810865	278	New England Journal of Medicine	Clinical trial

To capture temporal shifts in research focus, a citation burst analysis was conducted. As shown in Fig. 7A, a total of 25 references exhibited strong citation bursts, indicating periods of intensified scholarly attention. Based on burst timing and thematic characteristics, the evolution of the field can be broadly divided into three phases. The initial phase (≤ 2015–2017) focused on PD-1/PD-L1 biological mechanisms and early clinical trials, establishing the clinical efficacy of immune checkpoint inhibition and the predictive value of PD-L1 expression and TMB. The consolidation phase (2016–2017) emphasized first-line immunotherapy strategies, during which pembrolizumab was established as a standard first-line treatment for patients with high PD-L1 expression. The development phase (2018–present) expanded toward resistance mechanisms, tumor microenvironment regulation, emerging biomarkers, combination immunotherapy, and liquid biopsy technologies, reflecting increasing mechanistic depth and clinical refinement.

Fig. 7.

Analysis of co-cited references and their visualizations. A Top 25 references with the strongest citation bursts. B Co-citation network of references. C Reference clustering analysis

Bibliometric network analysis further revealed a well-defined knowledge structure within this research domain (Fig. 7B). Distinct co-citation clusters were formed around core publications, particularly pivotal clinical trials involving nivolumab and pembrolizumab. Highly co-cited studies by Rizvi et al., Topalian et al., Brahmer et al., and Garon et al. collectively laid the theoretical and clinical foundation for immune checkpoint inhibitors as transformative therapies in NSCLC.

The reference clustering analysis generated by CiteSpace is shown in Fig. 7C. Cluster #0 (Nivolumab, Pembrolizumab) represents the largest cluster, indicating the sustained dominance of immune checkpoint inhibitor–related research. Additional clusters correspond to in-depth investigations into immune mechanisms, molecular subtyping, therapeutic strategies, metastatic processes, liquid biopsy technologies, and emerging biomarkers such as the gut microbiota. Notably, several clusters reflect expanding research interest toward new cancer indications, real-world evidence, and novel detection approaches.

Overall, the reference-based knowledge map demonstrates that NSCLC immunotherapy research has evolved from an early focus on pivotal clinical trials to a more diversified and precision-oriented framework integrating biomarker discovery, mechanistic insights, and expanded clinical applications.

Keyword clustering and co-occurrence analysis

To further elucidate the thematic structure, research hotspots, and emerging frontiers of immunotherapy biomarker research in NSCLC, keyword co-occurrence, clustering, burst, and timeline analyses were systematically conducted.

The keyword co-occurrence network is shown in Fig. 8A, where node size represents keyword frequency and edge thickness indicates co-occurrence strength. The network is centered on core terms such as “non-small cell lung cancer,” “immune checkpoint inhibitors,” and “targeted therapy,” and is surrounded by closely related keywords including “efficacy,” “safety,” “biomarkers,” and “resistance.” This structure reflects the close integration of therapeutic strategies, clinical outcomes, and predictive factors within NSCLC immunotherapy research. Figure 8B highlights emerging high-frequency keywords, particularly tumor microenvironment, radiomics, and tumor cell proliferation, suggesting increasing research attention to mechanistic interpretation and quantitative imaging approaches.

Fig. 8.

Visualization of networks, overlays, and keyword analyses of immunotherapy biomarkers for NSCLC. A Keyword co-occurrence network visualization. B Overlay visualization map. C Keyword clustering analysis. D Top 25 keywords with the strongest citation bursts and their active periods. E Keyword timeline and clustering visualization showing temporal evolution of research topics

Keyword clustering analysis generated by CiteSpace is presented in Fig. 8C, revealing multiple well-defined thematic clusters. Major clusters include targeted therapy, immunotherapy, liquid biopsy, tumor mutational burden, PD-L1 expression, and first-line treatment. Among these, targeted therapy and immunotherapy represent the two dominant therapeutic paradigms, reflecting the widespread clinical adoption of targeted agents and immune checkpoint inhibitors. Clusters related to liquid biopsy and tumor mutational burden highlight the growing importance of precision oncology and genomic profiling in patient stratification and treatment evaluation, whereas biomarkers and PD-L1 expression remain central to diagnostic assessment and therapeutic decision-making.

The keyword burst analysis illustrated in Fig. 8D depicts the temporal evolution of research hotspots. During the early period (2015–2017), research attention focused primarily on safety, antibodies, and early clinical activity. Between 2017 and 2020, emphasis shifted toward first-line treatment, immunohistochemistry standardization, and combination therapy. From 2020 to 2024, burst keywords were dominated by tumor mutational burden, liquid biopsy, radiomics, and resistance mechanisms, indicating a transition toward dynamic disease monitoring and refined therapeutic stratification. The keyword timeline visualization (Fig. 8E) further corroborates this progressive shift in research focus.

Collectively, the integration of multiple keyword-based analyses demonstrates that NSCLC immunotherapy biomarker research has evolved into a multidimensional and stratified knowledge system. Over time, research emphasis has shifted from early concerns regarding safety and feasibility toward precision stratification, dynamic monitoring, and mechanistic understanding. Current frontier topics include liquid biopsy, radiomics, resistance mechanisms, and the tumor microenvironment, reflecting the increasing convergence of technological innovation and clinical application in this field.

Discussion

Research background and overview

NSCLC remains one of the most prevalent and lethal malignancies worldwide. Although conventional therapeutic modalities such as chemotherapy and radiotherapy have achieved incremental progress, their clinical efficacy—particularly in patients with advanced-stage disease—remains limited. In recent years, immunotherapy has emerged as a transformative treatment strategy and has gradually become a central component of NSCLC management. ICIs, especially PD-1/PD-L1 inhibitors, have demonstrated substantial survival benefits in selected patient populations. Nevertheless, immunotherapy continues to face major challenges, including intrinsic and acquired resistance, irAEs [27], and pronounced inter-patient heterogeneity in treatment response. Despite these limitations, the introduction of immunotherapy has markedly improved survival outcomes in advanced NSCLC and has provided valuable paradigms for immunotherapeutic strategies in other malignancies.

Parallel to these therapeutic advances, biomarker research has attracted increasing attention, particularly in areas such as PD-L1 expression, TMB, and tumor-infiltrating lymphocytes (TILs) [28]. Biomarker-driven studies not only facilitate the identification of patients most likely to benefit from immunotherapy but also accelerate the transition toward precision oncology. Based on bibliometric data from 2015 to 2024, the present study demonstrates the rapid expansion of immunotherapy biomarker research in NSCLC and delineates its major developmental trajectories. A total of 2134 publications were retrieved from the Web of Science Core Collection, involving 73 countries, 3670 institutions, and 14,723 authors. Publication output increased significantly after 2015, with a pronounced acceleration in recent years.

The year 2015 represents a critical inflection point in NSCLC immunotherapy research, coinciding with the FDA approvals of nivolumab and pembrolizumab for lung cancer indications. This milestone marked a paradigm shift from predominantly experimental investigation to widespread clinical application. Since then, publication output has exhibited sustained growth, peaking in 2021, followed by modest fluctuations, and ultimately reaching a new maximum in 2024. The fitted exponential growth curve (R² = 0.9984) reflects continuously intensifying scholarly interest in this field. This expansion has been driven by multiple factors, including the broadening of clinical indications and treatment guidelines, advances in next-generation sequencing (NGS), single-cell and spatial omics technologies, the maturation of liquid biopsy approaches, progress in perioperative and combination immunotherapy strategies, the integration of artificial intelligence (AI) into multi-omics analysis, and the establishment of large multicenter cohorts and real-world data (RWD) platforms. In addition, post-pandemic recovery of academic activity and the publication of long-term follow-up results from pivotal clinical trials have further stimulated research output.

At the national level, China ranks first globally in publication volume, followed by the United States. However, the United States demonstrates nearly twice the total citation count of China, reflecting greater academic influence. This discrepancy likely arises from long-term investment in biomedical research, advanced infrastructure, and extensive international collaboration networks. Although China has rapidly enhanced its global research visibility, continued efforts are required to further improve research quality and citation impact. With sustained governmental support and expanding international collaboration, this gap is expected to narrow progressively. Notably, approvals of immunotherapies by the China National Medical Products Administration (NMPA) have significantly stimulated domestic research activity. NMPA-driven clinical trials, such as the CameL study [29], have not only confirmed therapeutic efficacy in Chinese populations but also facilitated the exploration of population-specific biomarkers, thereby contributing valuable evidence to the global NSCLC immunotherapy knowledge base. Meanwhile, countries with comparatively smaller research budgets, including the United Kingdom, Germany, and Japan, continue to exert strong academic influence, underscoring the importance of long-standing research traditions and stable collaborative networks.

Institutional analysis further highlights the growing research capacity of Chinese centers. Among the world’s top ten most productive institutions, eight are based in China and two in the United States, reflecting China’s expanding institutional strength and international engagement. High-impact studies from institutions such as Shanghai Pulmonary Hospital and Shanghai Jiao Tong University have applied multi-omics approaches to elucidate mechanisms of immunotherapy response and resistance in NSCLC [30, 31] , providing mechanistic insights that support biomarker discovery and rational combination strategies.

At the author level, Caicun Zhou of Tongji University Affiliated Hospital emerges as one of the most influential contributors in this field. His research has focused on large-scale clinical trial datasets and translational biomarker analyses, identifying multiple predictive factors—including DNA damage repair pathways, spatial genomic features, immune microenvironment characteristics, TMB, and telomerase reverse transcriptase alterations—that are associated with immunotherapy response [32–35]. Marcin Kowanetz of F. Hoffmann–La Roche AG is among the most highly cited authors, with substantial contributions to biomarker research, particularly in elucidating the predictive value of TMB for response to atezolizumab therapy [36].

From a journal perspective, Cancers and the Journal of Thoracic Oncology are leading outlets in terms of publication volume and disciplinary influence, whereas The New England Journal of Medicine demonstrates unparalleled citation impact by disseminating landmark clinical evidence. These findings suggest that journal selection should consider not only impact factor but also thematic alignment and target readership to maximize academic dissemination.

Finally, analysis of the most highly cited references underscores the foundational role of landmark clinical trials in shaping the field. Seminal studies such as “Pembrolizumab versus Chemotherapy for PD-L1–Positive Non–Small-Cell Lung Cancer” and “Nivolumab versus Docetaxel in Advanced Nonsquamous Non–Small-Cell Lung Cancer” established the clinical efficacy and safety of ICIs and laid the evidence base for their widespread adoption [6, 8]. Building upon these pivotal trials, future research is expected to increasingly focus on combination regimens, novel immunotherapeutic agents, and refined biomarker-guided patient stratification to further enhance therapeutic outcomes in NSCLC.

The evolution of research trends and emerging themes

Literature co-citation and cluster analysis

In this study, bibliometric clustering, co-citation analysis, keyword co-occurrence, and timeline visualization were integrated to characterize the evolution of research trends and hotspots in immunotherapy biomarker research for NSCLC. Co-citation and clustering analyses performed using CiteSpace revealed a structured and progressively expanding knowledge framework encompassing immunotherapy, TME, and tumor genetic alterations.

Based on the co-citation network shown in Fig. 7B, ICIs occupied a dominant position in NSCLC research, particularly during the period from 2015 to 2020. During this phase, citation frequencies of PD-1/PD-L1 inhibitors such as nivolumab and pembrolizumab increased sharply, reflecting their rapid establishment as key therapeutic options in clinical practice. The widespread adoption of ICIs was associated with substantial survival improvements in patients with advanced disease and stimulated intensive investigation into predictive biomarkers, including PD-L1 expression and TMB, thereby enhancing the precision of immunotherapy-based treatment strategies.

Beyond immunotherapy, the evolving literature network reflects diversification of therapeutic paradigms in NSCLC. The increasing prominence of targeted therapy and liquid biopsy technologies indicates a shift toward more personalized and adaptive treatment approaches. Targeted therapy, guided by specific molecular alterations, selectively inhibits oncogenic signaling pathways while limiting damage to normal tissues and has become a major treatment strategy in biomarker-driven oncology. In parallel, liquid biopsy—particularly through analysis of circulating tumor DNA (ctDNA)—has emerged as a minimally invasive tool for early detection, treatment monitoring, and investigation of resistance mechanisms. Together, these approaches underscore the growing emphasis on dynamic genomic profiling and longitudinal biomarker assessment in NSCLC management.

Co-citation cluster analysis further divided the research landscape into multiple thematic clusters, each representing a distinct research direction. Several clusters warrant particular attention.

Immunotherapy-related research (Cluster 0) remains centered on nivolumab and pembrolizumab, with studies focusing not only on therapeutic efficacy but also on patient stratification strategies and immune escape mechanisms.

TME and immune response research (Cluster 3) highlights the critical role of immune cell composition within the tumor microenvironment, particularly T-cell infiltration, in shaping immunotherapy response [37, 38].

Mutation burden and biomarker research (Cluster 6) emphasizes the association between TMB and immunotherapy efficacy, as tumors with higher mutational loads are more likely to generate neoantigens that enhance immune recognition [39].

Liquid biopsy and ctDNA research (Cluster 7) reflects growing interest in noninvasive biomarker assessment for early diagnosis, therapeutic monitoring, and resistance analysis in NSCLC [40–43].

Finally, adjuvant therapy and multicancer research (Clusters 9 and 12) highlight expanding efforts to combine ICIs with chemotherapy or targeted therapy and to extend immunotherapy applications beyond lung cancer, supporting its broader oncologic relevance [43, 44].

Collectively, co-citation network analysis demonstrates that NSCLC immunotherapy research has evolved from a singular focus on ICIs toward a multifaceted framework integrating immune context, genomic alterations, and dynamic biomarker technologies. This evolution reflects a shift from treatment-centric investigation to precision-oriented therapeutic optimization.

Implications of emerging themes beyond current guidelines

Bibliometric trends identified in this study indicate that immunotherapy biomarker research in NSCLC has increasingly extended beyond the scope of current clinical guidelines. Although international guidelines, such as those issued by the National Comprehensive Cancer Network (NCCN) and ESMO, continue to rely primarily on PD-L1 expression for treatment stratification, the rapid emergence of new research themes highlights persistent unmet clinical needs and the limitations of single-biomarker–based decision-making [45].

One prominent emerging theme involves the TME. Increasing attention to immune cell composition, T-cell infiltration, and immune regulatory pathways suggests growing recognition that responses to ICIs cannot be fully explained by tumor-intrinsic biomarkers alone. Instead, immunotherapy efficacy appears to be shaped by complex interactions between tumor cells and the surrounding immune milieu. Although TME-related biomarkers have not yet been incorporated into routine clinical guidelines—largely due to challenges in standardization, reproducibility, and prospective validation—their prominence in recent literature underscores their potential value in refining patient stratification and elucidating resistance mechanisms [46].

Another rapidly advancing research direction is liquid biopsy–based biomarker assessment, particularly ctDNA analysis. The increasing emphasis on ctDNA reflects a shift toward dynamic and minimally invasive monitoring of tumor burden, molecular evolution, and therapeutic response. Unlike tissue-based biomarkers, ctDNA enables longitudinal assessment throughout the treatment course, offering insights into early response, minimal residual disease, and emerging resistance. However, despite strong research momentum, ctDNA has not yet been adopted as a guideline-recommended biomarker for immunotherapy initiation in NSCLC, primarily due to variability in detection platforms, threshold definitions, and the absence of large-scale prospective validation [47].

In addition, TMB continues to attract sustained research interest as a potential predictive biomarker, particularly in patients with low or negative PD-L1 expression. Ongoing attention to TMB reflects efforts to identify complementary biomarkers that may enhance patient selection for immunotherapy. Nevertheless, heterogeneity in sequencing methodologies, inconsistent cutoff values, and variable clinical performance have limited its routine guideline incorporation. Bibliometric trends suggest that TMB is increasingly viewed not as a replacement for PD-L1, but as one component of a broader, multidimensional biomarker framework [48].

Importantly, the emergence of these research themes highlights a discernible gap between rapidly advancing scientific investigation and the conservative pace of guideline adoption. This gap reflects both stringent evidentiary requirements for clinical implementation and the inherent complexity of translating exploratory biomarkers into standardized tools. Rather than challenging existing guidelines, the expanding research focus identified in this analysis suggests an effort to complement and refine current biomarker strategies.

Overall, the evolution of research themes beyond established guidelines indicates a transition from static, single-marker paradigms toward integrative and dynamic biomarker models. Future advances in NSCLC immunotherapy are likely to depend on the systematic integration of emerging biomarkers—such as TME characteristics, ctDNA, and genomic indicators—with established clinical markers, ultimately enabling more precise and personalized immunotherapy strategies.

Clinical relevance, guideline alignment, and controversies of immunotherapy biomarkers in NSCLC

The clinical success of immunotherapy in NSCLC relies on accurate biomarkers to guide patient selection and treatment stratification. Current international guidelines, including those issued by NCCN and ESMO, emphasize a limited number of validated biomarkers for routine clinical decision-making. In contrast, the bibliometric patterns identified in this study reveal a rapidly expanding biomarker research landscape that extends well beyond existing guideline recommendations, highlighting both substantial progress and persistent challenges in precision immunotherapy.

Biomarkers supported by current international guidelines

Among the proposed biomarkers, PD-L1 expression remains the only biomarker consistently recommended by international guidelines for routine clinical use in NSCLC immunotherapy [49]. PD-L1 testing is widely applied to determine eligibility for ICI monotherapy or combination regimens. However, PD-L1 expression is subject to intratumoral heterogeneity, temporal variability, and inter-assay differences. Moreover, a subset of patients with low or negative PD-L1 expression still derive clinical benefit from ICIs, underscoring the limited predictive capacity of PD-L1 as a single biomarker.

Microsatellite instability–high (MSI-H) status, although primarily recognized as a pan-cancer biomarker, is also acknowledged in international guidelines as an indicator of potential responsiveness to immunotherapy [50]. Nevertheless, MSI-H tumors are relatively rare in NSCLC, which substantially limits their overall impact on routine clinical decision-making in this disease.

Biomarkers with controversial or conditional clinical evidence

TMB represents one of the most extensively studied yet controversial biomarkers in NSCLC immunotherapy. Biologically, high TMB is associated with increased neoantigen load and enhanced tumor immunogenicity [51]. Despite substantial research interest and encouraging evidence in selected patient cohorts, international guidelines remain cautious regarding routine TMB-based treatment decisions. This caution reflects unresolved challenges, including the lack of standardized sequencing platforms, variability in cutoff thresholds, and inconsistent predictive performance across clinical trials. The sustained prominence of TMB in the literature suggests that it may function as a complementary biomarker rather than a standalone determinant.

TILs and other immune-related features of the TME are also closely associated with immunotherapy outcomes. Increased T-cell infiltration and immune-active microenvironments are generally linked to favorable responses to ICIs [52]. However, limitations related to reproducibility, standardized quantification, and contextual interpretation currently restrict the clinical implementation of TME-related biomarkers, confining their use largely to research settings.

Emerging biomarkers and integrative precision strategies

Liquid biopsy–based biomarkers, particularly ctDNA, represent a rapidly evolving area of translational research. ctDNA enables minimally invasive and dynamic monitoring of tumor burden, molecular evolution, and treatment response during immunotherapy. Accumulating evidence suggests that ctDNA kinetics may predict early therapeutic response, minimal residual disease, and the emergence of resistance. Despite its considerable promise, ctDNA is not yet recommended by international guidelines for guiding immunotherapy initiation in NSCLC, largely due to methodological heterogeneity and limited prospective validation [41].

The expanding diversity of biomarker research identified in this study reflects a broader conceptual shift toward integrative precision strategies. Rather than relying on a single biomarker, future immunotherapy approaches are likely to incorporate multidimensional models that combine tumor-intrinsic features (e.g., PD-L1, TMB), immune contexture (e.g., TILs, TME composition), and dynamic biomarkers (e.g., ctDNA), thereby more accurately capturing the biological complexity underlying treatment response and resistance.

Contextual comment on progastrin-releasing peptide (ProGRP)

In addition to predictive biomarkers relevant to immunotherapy, certain diagnostic biomarkers warrant brief contextual clarification. Progastrin-releasing peptide (ProGRP) is a well-established diagnostic biomarker of neuroendocrine differentiation. As comprehensively reviewed by La Salvia and Fanciulli (2024) [53], ProGRP demonstrates high sensitivity and specificity in pulmonary and extrapulmonary neuroendocrine neoplasms, particularly small cell lung cancer, and plays an important role in tumor classification, differential diagnosis, and disease monitoring. However, ProGRP is not a predictive biomarker for immunotherapy response in NSCLC and is not incorporated into current international guidelines for immunotherapy decision-making. Its inclusion here serves solely to highlight the biological heterogeneity of lung cancer and the importance of accurate histopathological classification when interpreting immunotherapy outcomes.

Implications for clinical practice and future research

The divergence between guideline-endorsed biomarkers and emerging research directions reflects the necessary balance between scientific innovation and clinical standardization. While international guidelines appropriately prioritize robustness, reproducibility, and prospective validation, the rapidly evolving biomarker landscape revealed by this study highlights persistent unmet needs in patient stratification and treatment optimization. Future advances in NSCLC immunotherapy are likely to depend not on replacing established biomarkers, but on refining and integrating them into composite and dynamic biomarker frameworks that support more precise and personalized treatment strategies.

Limitations

This study has several limitations that should be acknowledged. First, the literature retrieval was restricted to WoSCC, which may introduce database-related bias. Although WoSCC is a widely recognized and authoritative source for bibliometric research, it is subject to indexing delays, particularly for preprints (e.g., bioRxiv and medRxiv) and newly launched journals. Consequently, some cutting-edge or rapidly emerging studies may not have been captured in the present analysis. In addition, non-English publications, regional journals, and conference abstracts—including those reporting late-breaking findings from major scientific meetings such as ASCO or ESMO—were excluded due to concerns regarding data completeness, methodological transparency, and reproducibility. As a result, the findings of this study should be interpreted as a snapshot of the research landscape as of the data collection date (June 23, 2025), rather than as an exhaustive representation of all ongoing research activity. Future bibliometric studies may benefit from integrating multiple bibliographic databases, such as Scopus and PubMed, as well as preprint platforms, to improve the comprehensiveness, timeliness, and robustness of analyses in rapidly evolving research fields.

Conclusion

Using CiteSpace and VOSviewer, this study systematically mapped the global research landscape of immunotherapy biomarkers in NSCLC over the past decade. The analysis revealed that China and the United States are the leading contributors in terms of publication output and academic impact, highlighting the importance of strengthening international and institutional collaboration to advance translational research.

Recent progress in NSCLC immunotherapy has been largely driven by the clinical application of immune checkpoint inhibitors, with biomarkers such as PD-L1 expression, tumor mutational burden, and liquid biopsy–based indicators playing key roles in patient stratification. However, the limited predictive power of single biomarkers underscores the need for integrated biomarker strategies.

Future research is expected to focus on multi-biomarker combinations and tumor immune microenvironment modulation to enhance therapeutic efficacy and overcome resistance. Continuous identification of novel biomarkers and technological innovation will further accelerate progress in this field.

Author contributions

Xiangnv Meng: Conceived the study, designed the methodology, and conducted the bibliometric analysis. Drafted and revised the manuscript. Conceptualized and supervised the entire study. Provided critical revisions to the manuscript and served as the corresponding author. Zhongting Lu: Provided critical input to the study design and supervised the research process. Contributed to the interpretation of the data and manuscript review. Fu Mi: Assisted in data collection and analysis. Provided technical support and contributed to manuscript editing. Participated in data visualization and provided intellectual contributions to the discussion and interpretation of findings.

Funding

This study was supported by the Cangzhou Key Research and Development Program (Project No. 222106036).

Data availability

All the data can be obtained from the open-source website we provide, and the conclusion can be drawn through the analysis of the relevant software.

Declarations

Ethics approval and consent to participate

All data used in this study were downloaded from public databases; therefore, no ethical approval or informed consent was required.

Patient consent

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.