A total of 78 nations actively contribute to TILs research. The USA dominates with the highest number of publications (753), citations (55,331), and total link strength (519) ( Figures 2A, B ). Other key contributors include China (481), and Japan (288). Although South Korea ranked sixth in publications (145), it placed twelfth in citations (5,565), likely due to its limited collaboration (total link strength=70). The proportion of international co-authorships reflects relatively few joint publications (18.61%). In international collaborations, USA-centric partnerships occupied nine of the top 15 positions, predominantly with China (55 publications) ( Supplementary Material 2 ).

An examination of countries’ keywords highlights the diverse research emphases across different nations (refer to “3.3 Analysis of keywords” for further details). The USA has significantly advanced various research areas within TILs, particularly focusing on adoptive cell therapy ( Figure 2C ). Conversely, China’s research has concentrated on the therapeutic and prognostic significance of TILs, and PD-1/PD-L1 immune checkpoint inhibitors ( Figure 2D ), Japan and Italy have also made important contributions to this area ( Figures 2E, F ).

The National Cancer Institute(NCI) leads in publications(104) and citations (15,872) ( Tables 1 , 2 ). Figure 3A shows that NCI focuses more on adoptive cell therapy. Furthermore, institutions such as the Jules Bordet Institute, the German Cancer Consortium, the National Institutes of Health, and Heidelberg University, despite having fewer publications, are distinguished by their high citations. Conversely, Sun Yat-sen University, the University of Ulsan, and Fudan University exhibit high publications but lower citations. Geographically proximate institutions show greater ease of collaboration. The center of the institutional collaboration network is characterized by numerous closely interconnected nodes, primarily comprising leading institutions in the USA and Europe. In contrast, Asian institutions, such as Sun Yat-sen University, Fudan University, the University of Ulsan, and Asan Medical Center, are comparatively isolated ( Figure 3B ). The burst analysis identifies several emerging institutions, including the University of Toronto and the University of Helsinki ( Figure 3C ). Figure 3D indicates the persistent contributors (NCI, University of Texas System, Harvard University, and the National Institutes of Health) and several recent contributors (UNICANCER, Sun Yat-sen University).

Prominent figures include Steven A. Rosenberg from the NCI, who leads in publications, local citations, total citations, h-index, and m-index ( Table 3 ). His early contributions have been substantial ( Figures 4A, B ), especially in adoptive cell therapy and metastatic melanoma. Conversely, although Lee HJ, Whiteside TL, Hwu P, and Gong G are among the top ten publications, their local citations are comparatively lower ( Supplementary Material 3 ). Figure 4A shows early contributors (e.g., Whiteside TL) and recent contributors (e.g., Lee HJ and Loi S). The bursts analysis reveals emerging scholars such as Roberto Salgado, Giuseppe Floris, Yoshinao Oda, and Marco Donia ( Figure 4B ). Roberto Salgado and Giuseppe Floris have close associations and substantial influence in breast cancer and TILs research ( Figure 4C ). Yoshinao Oda primarily investigates PD-L1 and TILs, whereas Marco Donia focuses on adoptive cell therapy. Figure 4C depicts the rise of numerous teams over time, including some emerging smaller teams, such as Ock Chan-Young’s, which are focusing on artificial intelligence and TIL evaluation.

Cancer Immunology Immunotherapy leads in publications(103) ( Figure 5A ). The Journal of Immunology leads in citations (7,856) and h-index (47) ( Table 4 ) ( Supplementary Material 4 ). According to Bradford’s Law, there are 18 core journals ( Figure 5B ). Figure 5C delineates early contributors (Journal of Immunology, Cancer Research, International Journal of Cancer), persistent contributors (Cancer Immunology Immunotherapy), and recent contributors (Frontiers in Immunology, Cancers, Frontiers in Oncology, Journal for Immunotherapy of Cancer, Scientific Reports). Figure 5D highlights the theme of journals, with recent contributor journals emphasizing PD-L1, prognosis, breast cancer, neoadjuvant chemotherapy, biomarkers, and adjuvant chemotherapy. Co-citation analysis identifies the Journal of Immunology (4,655) being the most frequently co-cited ( Figure 5E ). Overlay maps of journals reveal that those focusing on molecular biology and genetics are frequently cited by journals in molecular biology and immunology, highlighting the interdisciplinary nature of TILs research ( Figure 5F ).

Funding support is evident, comprising 44.9% (1,159/2,584). The top two funders are the National Institutes of Health(310) and the United States Department of Health and Human Services (310) ( Figure 6A ). Since 2009, over 60% of publications have received financial support ( Figure 6B ), emphasizing funding’s pivotal role in driving research advancement. Moreover, articles from the USA received support at 61.18%, while those from China received a higher rate of 78.91%. Articles from Japan received support at 38.82%, from Italy at 49.07%, and from Germany at 51.45%, illustrating various national support in TILs research.

Co-citation reflects the referencing of two or more papers by other scholarly works, showcasing their interconnectedness within the research domain (33). Through co-citation analysis, it elucidates foundational knowledge and frontiers of academic research, enhancing our understanding of the field’s structure and evolution. CiteSpace was employed for co-citation analysis, resulting in 8 principal clusters(size>100) ( Figure 8A ) ( Supplementary Material 7 ). The clustering results (Q = 0.8215, S = 0.9271) highlight reliability. The largest cluster is #0 breast cancer. Currently, breast cancer (#0), PD-L1 expression (#1), and adoptive cell therapy (#5) have progressively gained prominence ( Figure 8B ). Table 6 details key references from these clusters.

The central components of breast cancer (#0) primarily address the application of TILs in the treatment and prognosis of breast cancer, predominantly featured in high-impact journals such as Annals of Oncology, Lancet Oncology, and Journal of Clinical Oncology. Several studies have revealed the prognostic and predictive value of TILs across various breast cancer subtypes (12, 34–36), particularly TNBC (37–40), encompassing both chemotherapy-treated and untreated patients (38). Furthermore, numerous studies have investigated the role of TILs as a significant predictor of neoadjuvant chemotherapy response to breast cancer (9, 10). Elevated TIL levels in TNBC are generally associated with improved responses to neoadjuvant chemotherapy and enhanced survival benefits (41). Additionally, the presence of TILs in residual disease following neoadjuvant chemotherapy also correlates with a better prognosis (9).

The principal themes of clusters #5 and #7 revolve around adoptive cell therapy, with cluster #7 being established earlier, having an average emergence year of 2009, while cluster #5 emerged around 2018. Analyzing the clusters reveals a trajectory of TIL therapy evolving from initial exploration to personalized treatment approaches. Seminal studies conducted by Rosenberg SA and his team at the National Institutes of Health in 1986 and 1988 illustrated the antitumor potential of TILs combined with IL-2 (3, 32). Subsequent advancements—such as the optimization of TIL isolation and expansion, chemotherapeutic preconditioning (e.g., lymphocyte-depleting chemotherapy), and advancements in gene engineering—have markedly improved the clinical effectiveness of TIL therapy (42, 43). Over the past three decades, numerous clinical centers have validated the feasibility of this therapy (44–47). As research has advanced, the notion of personalized treatment has become increasingly prominent. Research by Goff SL has suggested the need for personalized lymphodepletion regimens tailored to individual patient profiles (47).

Articles on PD-L1 expression (#1) reveal pivotal insights into the relationship between TILs and PD-L1, predominantly featuring in prestigious journals such as the New England Journal of Medicine. Over the past decade, the introduction of immune checkpoint inhibitors, particularly PD-1/PD-L1 inhibitors, has significantly reshaped the landscape of cancer immunotherapy. Since 2015, there has been an explosion of research on the expression of PD-1 and PD-L1 in tumors and their interactions with TILs. Research indicates that tumors with high TILs frequently also show elevated PD-L1 expression (48–50). Extensive studies have established that elevated levels of both TILs and PD-L1 expression correlate with a favorable prognosis in cancers(e.g., breast cancer), and with the efficacy of neoadjuvant chemotherapy (11, 49, 50). Additionally, PD-1 expression helps identify tumor-reactive TILs (51), and the presence of TILs is closely linked to the effectiveness of immune checkpoint inhibitor therapies. These developments have accelerated research on TILs as biomarkers for predicting immunotherapy outcomes (52–54) and combined treatment approaches (21). In particular, some “cold” tumors, such as non-small-cell lung cancer, can be transformed into “hot” tumors through combination therapies, enhancing their sensitivity to immune checkpoint inhibitors (21). Moreover, by blocking the PD-L1/PD-1 pathway, TIL suppression can be mitigated, leading to increased TIL numbers in the TME and a restoration of their tumor-fighting capabilities (22).

The burst of citations has provided a valuable means for observing the evolution of research focal points ( Figure 8C ). A total of 499 references exhibited citation bursts ( Supplementary Material 7 ). As of 2024, 82 references are still experiencing citation bursts ( Figure 8D ), with 35 in cluster #0 (breast cancer), 34 in cluster #5 (adoptive cell therapy), and 4 in cluster #1 (PD-L1 expression). The primary topics include the prognostic significance of TILs in melanoma, breast cancer, colorectal cancer, and head and neck squamous cell carcinoma; the role of TILs in predicting the efficacy of neoadjuvant chemotherapy; the relationship between TILs and PD-L1; TIL assessment; TIL-related biomarkers; TILs in the TME; and tertiary lymphoid structures. These topics represent key current and emerging focuses in TILs research.

The prognostic value of TILs is unequivocal, particularly in TNBC (37–40). TILs also serve as a predictive marker for positive chemotherapy response in these patient populations (9, 10). Furthermore, the prognostic role of TILs in other solid tumors, such as head and neck squamous cell carcinoma, oropharyngeal cancer, and laryngeal cancer, is becoming increasingly well-understood and will become a research hotspot in the future (70, 71). For instance, the Society for Immunotherapy of Cancer utilizes the consensus Immunoscore assay to evaluate tumor-infiltrating T-cell counts to predict recurrence risk in patients with stage I-III colon cancer (72).

With the emergence of immune checkpoint inhibitors such as PD-1/PD-L1 inhibitors, integrating PD-L1 expression and TIL characteristics for disease prognosis has gained considerable attention currently (11, 73). Several studies have delved into combining these markers with relevant gene expression levels to predict overall survival in cancer patients (74, 75). Moreover, numerous studies have shown that the presence and activity of TILs are closely linked to the efficacy of PD-1/PD-L1 checkpoint inhibitors. High-density TILs may indicate a more favorable response to ICI therapy (19).

Additionally, aside from PD-L1, LAG3+ (a negative immune regulator) is also highly expressed in TILs (76, 77), with ongoing research examining its potential as a biomarker for predicting immune therapy responses and prognosis (26). Furthermore, serving as local immune centers, numerous studies have highlighted tertiary lymphoid structures as a favorable prognostic factor for cancer (78), particularly breast cancer (8, 79). The interplay between tertiary lymphoid structures, TILs, and clinical outcomes is being thoroughly investigated across different tumor types (80, 81).

Consequently, as key biomarkers for prognosis and therapeutic response in certain cancers, TIL-based patient stratification is currently a significant area of research, aiming to reduce overtreatment, minimize potential toxicity, and address cost considerations (82). For example, identifying TNBC patients who might benefit from less aggressive chemotherapy is essential (83, 84). Some studies have proposed TIL cutoff values to stratify chemotherapy strategies for stage I TNBC patients (85); however, controversy remains regarding the definition of these cutoff values.

Moreover, using TILs, PD-L1, and other biomarkers to identify patients likely to benefit from ICI therapy is being hotly investigated (23–25, 82). For instance, Eftychia Chatziioannou’s team discovered that in untreated metastatic melanoma, an electronic TIL score ≤12.2% could predict poor survival outcomes in patients undergoing anti-PD-1 therapy (86). Similarly, Carmine Valenza’s team identified that PD-L1 CPS positivity and/or TIL scores ≥1% could forecast skin metastases in 57% of patients with chest wall breast cancer (87).

In summary, the prognostic value of TILs in specific cancers is well established, and the concept of utilizing TILs for patient stratification is increasingly acknowledged. Developing models that integrate TILs and other biomarkers to predict treatment response and long-term prognosis remains a.

The significant role of TILs in prognosis and therapy has led to increased focus on TIL assessment. Recent advancements in digital pathology and artificial intelligence have facilitated the automation of TIL assessment, improving its efficiency, objectivity, and consistency (15, 16). These technological innovations also facilitate the detailed evaluation of the functional characteristics and spatial distribution of TILs (15, 17–19), such as the AI-driven TIL spatial analysis developed by Sehhoon Park’s team (18), positioning it as a current and future research hotspot. For example, Joel Saltz’s team employed convolutional neural networks to analyze H&E-stained digital slides (17). Additionally, the integration of multidimensional data (e.g., genomics, transcriptomics, proteomics) creates a panorama of TILs within the TME. This approach elucidates their complex interaction networks and supports the identification of novel biomarkers and therapeutic targets (20), marking a significant area of future research.