CXCL13 and CCL21 Induce Tertiary Lymphoid Structures and Enhance the Efficacy of Immunotherapy for Melanoma

Abstract

ABSTRACT

Tertiary lymphoid structures (TLS) are acquired ectopic lymph follicle‐like structures observed inside and around tumors, in which clusters of CD20‐positive B lymphocytes are surrounded by CD3‐positive T lymphocytes. In many cancers, the existence of TLS is a useful biomarker for better prognosis and better response to immune checkpoint inhibitors (ICI) and plays important roles in activating anti‐tumor immunity. In order to induce TLS and enhance the therapeutic effect of ICI, we attempted to induce TLS using multiple chemokines in malignant melanoma, for which there have been no reports of TLS induction previously. Immunohistochemical analysis of tumor samples from 41 melanoma patients treated with ICI revealed TLS in 63.4% of cases. Patients with ≥ 5 TLS exhibited significantly improved disease‐specific survival compared to those with fewer or no TLS. Plasma chemokine profiling in 46 samples from 18 melanoma patients showed elevated CC motif chemokine ligand 21 (CCL21) in TLS‐positive samples before and after ICI treatment and CXC motif chemokine ligand 13 (CXCL13) significantly increased pre‐ to post‐ICI treatment in paired samples from TLS‐positive patients. In a mouse melanoma model, co‐administration of CXCL13 and CCL21 alongside anti‐programmed death ligand‐1 (PD‐L1) antibody therapy significantly increased TLS formation and improved tumor growth suppression. Gene expression analysis of human melanoma samples demonstrated that high CXCL13 and CCL21 expression correlated with upregulation of immune response, particularly B cell activation. These findings highlight the potential of chemokine‐based therapies. TLS induction using CXCL13 and CCL21 in combination may be useful for enhancing the effects of ICI therapy in melanoma.

Abbreviations

• CCL19
• CC motif chemokine ligand 19
• CCL21
• CC motif chemokine ligand 21
• CCL22
• CC motif chemokine ligand 22
• CCL24
• CC motif chemokine ligand 24
• CCL26
• CC motif chemokine ligand 26
• CCR7
• CC motif chemokine receptor 7
• CXCL12
• CXC motif chemokine ligand 12
• CXCL13
• CXC motif chemokine ligand 13
• CXCL16
• CXC motif chemokine ligand 16
• CXCL5
• CXC motif chemokine ligand 5
• CXCR5
• CXC motif chemokine receptor 5
• CXC3L1
• CX3C motif chemokine ligand 1
• FDC
• follicular dendritic cells
• FFPE
• formalin fixed paraffin embedded
• GC
• germinal center
• GEO
• gene expression omnibus
• GSEA
• gene-set enrichment analysis
• GO
• gene ontology
• ICI
• immune checkpoint inhibitors
• CSD
• cumulative sun damage
• IRB
• institutional review board
• PD-L1
• programmed death ligand-1
• TLS
• tertiary lymphoid structures

Introduction

Tertiary lymphoid structures (TLS) are acquired ectopic lymph follicle-like structures, defined as a CD20+ B cell cluster surrounded by CD3+ T cells [1]. The major subset of T cells within the TLS is CD4+ follicular helper T cells, accompanied by CD8+ cytotoxic T cells, CD4+ helper T1 (Th1) cells, and regulatory T cells (Treg). Other cells include macrophages and stromal cells [1]. Peripheral node addressin (PNAd)–positive high endothelial venules (HEVs) provide the specialized vasculature associated with TLSs and mediate lymphocyte recruitment [2]. Mature TLS, which are characterized by the presence of CD21+/CD23+ follicular dendritic cells (FDC) within CD20+ B cell clusters and the formation of germinal centers [3], are known to enhance immune function [1, 4]. TLS form in chronic inflammatory microenvironments such as malignant tumors [1, 5], autoimmune diseases [6], and chronic inflammations [7]. Regarding cutaneous malignancies, TLS have been reported to form in malignant melanoma [8–10], squamous cell carcinoma [11], basal cell carcinoma [12], Merkel cell carcinoma [13], angiosarcoma [14, 15], and extramammary Paget's disease [16]. In the tumor microenvironment of various skin cancers, TLS form within and around the tumor and function as sites of antigen presentation and immune activity, which are associated with a favorable prognosis. In addition, TLS activate anti-tumor immunity and enhance the therapeutic effect of immune checkpoint inhibitors (ICI) [9]. Therefore, TLS have been considered a prognostic marker and a predictive marker of response to ICI therapy in many cancers [1].

Induction of TLS may be a good strategy to enhance the efficacy of ICI and improve the prognosis. Many chemokines, including CXCL13 (CXC motif chemokine ligand 13), are involved in the formation of TLS [1, 3]. Past studies have reported in which chemokines were administered to mouse models of ovarian cancer [17] and pancreatic cancer [18] to induce TLS, and the relationship with therapeutic efficacy was evaluated. On the other hand, there are no reports of the induction of TLS by chemokine administration to mouse models of skin cancers, including malignant melanoma, or examination of the relationship between TLS and ICI effects. In this study, we hypothesized that the administration of chemokines could induce TLS in and around skin cancers and that the induction of TLS could enhance responsiveness to ICI, and verified this through experiments using human melanoma cohorts and a mouse melanoma model.

Materials and Methods

Melanoma Patient Cohorts

All experiments were initiated after receiving institutional review board (IRB) approval (Nagoya City University Clinical Trial Management Center, No. 60-22-0141, 60-23-0016). Patients who received ICI treatment at Nagoya City University between September 2014 and March 2024 for malignant melanoma and were followed for at least 6 months were collected. Among these, a cohort of 41 cases (25 males and 16 females) with a median age of 69.4 in which formalin-fixed paraffin-embedded (FFPE) samples were stored was used for the prognostic analysis based on the results of immunohistochemical staining of TLS (Table 1). A cohort of 18 cases (46 samples, 9 males and 9 females) with a median age of 66.1 in which blood samples were stored at multiple time points was used for comprehensive chemokine analysis (Table 2).

TABLE 1 Characteristics of 41 melanoma patients for immunohistochemistry of TLS and survival analysis.

Characteristics	Value
Cases/samples	41
Age (range)	69.4 (37–86)
Sex
Male	25 (61.0%)
Female	16 (39.0%)
Race
Asian (Japanese)	41 (100%)
Primary site
Head and neck	6 (14.6%)
Trunk	11 (26.8%)
Limbs	23 (56.1%)
Primary unknown	1 (2.4%)
Classification
Low-CSD	11 (26.8%)
High-CSD	5 (12.2%)
Acral	20 (48.8%)
Mucosal	4 (9.8%)
Primary unknown	1 (2.4%)
BRAF mutation
Wild type	29 (70.7%)
Mutant	8 (19.5%)
Unexamined	4 (9.8%)
Stage at collection
I	2 (4.9%)
II	12 (29.3%)
III	22 (53.7%)
IV	5 (12.2%)
Patients treated with ICI	41 (100%)

TABLE 2 Characteristics of 18 melanoma patients for chemokine analysis.

Characteristics	Value
Cases/samples	18
Age (range)	66.1 (37–86)
Sex
Male	9 (50.0%)
Female	9 (50.0%)
Race
Asian (Japanese)	18 (100%)
Primary site
Head and neck	3 (16.7%)
Trunk	5 (27.8%)
Limbs	9 (50.0%)
Primary unknown	1 (5.6%)
Classification
Low-CSD	6 (33.3%)
High-CSD	3 (16.7%)
Acral	6 (33.3%)
Mucosal	2 (11.1%)
Primary unknown	1 (5.6%)
BRAF mutation
Wild type	13 (72.2%)
Mutant	5 (27.8%)
Unexamined	0 (0%)
Stage at collection
I	0 (0%)
II	0 (0%)
III	4 (22.2%)
IV	14 (77.8%)
Patients treated with ICI	18 (100%)
Types of relevant ICI
Ipilimumab+Nivolumab	4 (22.2%)
Nivolumab	9 (50.0%)
Pembrolizumab	5 (27.8%)

Immunohistochemical Staining

FFPE tissue sections of 41 human melanoma samples and mouse samples were processed for indirect immunofluorescence to detect the expression of signal transduction proteins. The primary and secondary antibodies used are listed in Table S1. 4′,6-diamidino-2-phenylindole (DAPI, Vector Laboratories, Burlingame, CA) was used as a nuclear counterstain. The red fluorescence produced by Alexa 594, green fluorescence produced by Alexa 488, and blue fluorescence produced by DAPI were observed and captured using a fluorescence microscope BZ-X810 (Keyence, Osaka, Japan). The human TLS were identified as clusters of 10 or more CD20-positive cells surrounded by more than half a circle of CD3-positive cells as previously reported [13, 14]. Among them, the mature TLS were identified as CD21-positive FDC and Bcl-6-positive germinal center (GC) B cells that were observed inside the CD20-positive cell clusters, and those that could not be observed were defined as immature TLS. Regardless of whether it was mature or immature, if there was even one TLS in the tissue, the case was considered TLS-positive. The mouse TLS were identified as clusters of 5 or more CD20-positive cells surrounded by CD3-positive cells approximately half the circumference.

Multiplex Immunoassay for Chemokines

Forty chemokines were analyzed by the Bio-Plex suspension array system (Bio-Rad Laboratories, Hercules, CA) using the Bio-Plex Pro Human Chemokine Panel 40-plex (171AK99MR2, Bio-Rad Laboratories, Hercules, CA) in 46 plasma samples from 18 melanoma patients. Two chemokines (CXC motif chemokine ligand 5; CXCL5, CX3C motif chemokine ligand 1; CX3CL1) were excluded from the analysis because no values were obtained for all samples. The values are shown after standardization.

Melanoma Mouse Models

All animal experiments were approved by the University's Ethics Committee (No. 21-024H03). 1 × 10⁵ B16F0 (ATCC-CRL-6322) melanoma cells (American Type Culture Collection, Manassas, VA) were injected subcutaneously into the back of 6–8 week-old C57BL/6 female mice. Starting 3 days after melanoma cell injection, 200 μg/mouse anti-programmed death ligand-1 (PD-L1) antibody (Leinco Technologies, St. Louis, MO) was administrated intraperitoneally twice weekly. Injections of chemokines were started three times a week at the same time. 1 μg/mouse rmCXCL13/BLC/BCA-1 (R&D Systems, Minneapolis, MN) or PBS as control were subcutaneously injected around the back tumors, and 1 μg/mouse rmCCL21/6Ckine (R&D Systems) and 1 μg/mouse rmCXCL16 (R&D Systems) were injected intraperitoneally. After five times of anti-PD-L1 antibody injections and seven times of chemokines or PBS injections, mice were sacrificed 19 days after tumor injection. The tumor volume of the mice was measured by caliper, and the number of TLS was counted by immunohistochemistry. If some mice died during the experiment, all tumor size measurements for those mice were excluded. Each group consisted of five mice, and all experiments were performed in triplicate.

Clustering and Gene Set Enrichment Analysis

The gene set enrichment analysis was performed using the profiled data set containing 83 melanoma samples containing 31 primary and 52 metastatic lesions with comprehensive gene expression analysis for 13,488 genes accessible at the gene expression omnibus (GEO) database (GDS3966) [19]. A clustered heatmap of all samples was generated using the online tool iDEP.96 () [20]. The same tool was used for generally applicable gene-set enrichment analysis (GSEA); GO (gene ontology) biological process was selected for gene sets, and pathway significance cutoff (false discovery rate; FDR) was set to 0.2.

Statistical Analysis

Statistical analyses were performed using Graph Pad Prism 9 (Graph Pad Software, San Diego, CA). Probability values of less than 0.05 were considered statistically significant.

Results

The Higher Number of Tertiary Lymphoid Structures Correlate With a Better Prognosis

Immunohistochemical staining of tumor samples from 41 malignant melanoma patients treated with ICI in our institution (Table 1) revealed findings of TLS as clusters of CD20-positive B cells (green, Figure 1A) surrounded by CD3-positive T cells (red, Figure 1A) in and around the tumors. TLS were observed in 63.4% of cases. In mature TLS, CD21-positive FDC (red, arrowhead, Figure 1B) and GC B cells (yellow, arrowhead, Figure 1C) were observed inside the CD20-positive B cell clusters. Although some mature TLS were identifiable in hematoxylin and eosin (HE) stained images (Figure 1D), there was no GC formation that could be distinguished by HE staining. Patients were classified according to the presence or absence of TLS and the number of TLS, and the disease-specific survival was compared. Comparison based on the presence or absence of mature TLS revealed no significant difference between cases with mature TLS (n = 15) and those with only immature TLS (n = 11) or no TLS (n = 15) (p = 0.1266, Log-lank test, Figure 1E). Comparison based on the number of TLS regardless of whether it was mature or immature showed that the prognosis was significantly better for cases with TLS ≥ 5 (n = 16) than for cases with 1 ≤ TLS ≤ 4 (n = 10) or TLS = 0 (n = 15) (p = 0.0189, Log-lank test, Figure 1F). In the comparison between the TLS-positive and TLS-negative groups, there was no difference in age or gender ratio, and the averages of 5-SCD and LDH were lower in the TLS-positive group (not significant, unpaired t-test, Table 3).

[IMAGE OMITTED. SEE PDF]

TABLE 3 Comparison between the TLS-positive and TLS-negative groups.

Characteristics	TLS positive	TLS negative
Cases/samples	26	15
Age (range)	69.2 (37–86)	69.8 (39–84)
Sex
Male	16	9
Female	10	6
Primary site
Head and neck	3	3
Trunk	7	4
Limbs	16	7
Primary unknown	0	1
Classificasion
Low-CSD	8	3
High-CSD	3	2
Acral	14	6
Mucosal	1	3
Primary unknown	0	1
Tumor Thickness (mm) (range)	5.79 (0.6–20)	5.01 (1.2–15)
BRAF mutation
Wild type	18	11
Mutant	6	2
Unexamined	2	2
Stage at collection
I	0	2
II	6	6
III	17	5
IV	3	2
5-SCD at collection (nmol/L)	12.8	19.8
LDH at collection (U/L)	198.6	231.6

Plasma CCL21 Levels Were Significantly Increased in TLS-Positive Samples Before and After ICI Administration

We collected 46 plasma samples from 18 melanoma patients (Table 2) treated with ICI and comprehensively measured 40 chemokines. Two chemokines (CXCL5, CX3CL1) were excluded from the analysis because no values were obtained for all samples. A heatmap showed the expression levels of 38 chemokines across all plasma samples (Figure 2A). Clustering analysis revealed that some chemokines showed similar patterns, such as CC motif chemokine ligand 21 (CCL21), CXC motif chemokine ligand 16 (CXCL16), and CXC motif chemokine ligand 12 (CXCL12). In the samples taken before ICI administration, CXCL16 (p = 0.0013) and CCL21 (p = 0.0179) were significantly elevated in the TLS-positive cases (Figure 2B). On the other hand, after ICI administration, CCL21 (p = 0.0228) was the only chemokine that was significantly elevated in the TLS-positive samples (Figure 2C).

[IMAGE OMITTED. SEE PDF]

Plasma CXCL13 Levels Increase After ICI Administration in the TLS-Positive Same Patients

In TLS-positive patients with blood samples taken before and after ICI administration (n = 8), changes in all measured chemokines due to treatment were analyzed. Among the measured chemokines, only CXCL13 levels were significantly elevated after ICI administration in TLS-positive cases (p = 0.0156, paired t-test, Figure 3A). No significant differences were observed before and after ICI administration for other chemokines, including CCL21 (p = 0.6406, paired t-test, Figure 3B), CXCL16 (p = 0.8438, paired t-test, Figure 3C), CXCL12 (p = 0.8438, paired t-test, Figure 3D), CC motif chemokine ligand 19 (CCL19, p = 0.2188, paired t-test, Figure 3E), CC motif chemokine ligand 22 (CCL22, p > 0.9999, paired t-test, Figure 3F), CC motif chemokine ligand 24 (CCL24, p = 0.8438, paired t-test, Figure 3G), or CC motif chemokine ligand 26 (CCL26, p = 0.4609, paired t-test, Figure 3H).

[IMAGE OMITTED. SEE PDF]

CXCL13 and CCL21 Administration Induced TLS and Enhanced the Efficacy of ICI in Mouse Melanoma Model

A mouse melanoma model was established by injecting B16 melanoma cells into the backs of mice on Day 0. All mice received 200 μg of anti-PD-L1 antibody twice a week for a total of five doses. Along with anti-PD-L1 antibody treatment, they received either CXCL13, CCL21, CXCL16, CXCL13 and CCL21, or CXCL13 and CXCL16 administration. Chemokines were administered at a dose of 1 μg each, and PBS was injected as a control. On Day 19, the mice were sacrificed, and immunohistochemical staining for CD20 and CD3 was performed on tumor sections, revealing the presence of TLS within the tumors (Figure 4A,B). The number of TLS in tumors was counted and averaged for each group (n = 5, Figure 4C). Mice that were administered CXCL13 and CCL21 during anti-PD-L1 antibody treatment exhibited a higher number of TLS compared to other mice (*p = 0.0342, One-way ANOVA). No TLS was observed in mice treated only with the anti-PD-L1 antibody or combination with chemokines other than CXCL13. Mice that received CXCL13 and CCL21 during anti-PD-L1 antibody treatment had significantly suppressed tumor growth compared to mice that received only CXCL13 during anti-PD-L1 antibody treatment (p = 0.0284, Tukey's multiple comparisons test, Figure 4D,E). Mice treated with anti-PD-L1 antibody in combination with only CCL21 and mice treated with anti-PD-L1 antibody in combination with CXCL13 and CXCL16 showed more tumor growth than mice treated with anti-PD-L1 only (no significant difference, Figure 4F). In the comparison of mice treated with anti-PD-L1 antibody in combination with only CCL21 and mice treated with anti-PD-L1 antibody in combination with CXCL13 and CCL21, the average tumor size on Day 19 was smaller in the PD-L1 + CXCL13 + CCL21 group than in the PD-L1 + CCL21 group (no significant difference, Figure S1).

[IMAGE OMITTED. SEE PDF]

Melanoma Samples With High CXCL13 and CCL21 Expression Significantly Express Gene Sets Associated With Activation of the Immune Response, Especially B Cell Related

The analysis was conducted using a profiled dataset of 83 human melanoma samples including 31 primary and 52 metastatic lesions, with comprehensive gene expression data for 13,488 genes available in the GEO database (GDS3966). Samples with CXCL13 expression above the overall average were classified as CXCL13 high, while those below the average were classified as CXCL13 low. The same classification criteria were applied to CCL21. The 83 melanoma samples were subsequently grouped into four categories: 10 samples in the CXCL13 high, CCL21 high group, 10 samples in the CXCL13 high, CCL21 low group, three samples in the CXCL13 low, CCL21 high group, and 60 samples in the CXCL13 low, CCL21 low group. Clustering analysis identified genes that were upregulated or downregulated in each group, and all CXCL13 high and CCL21 high samples were clustered close together (Figure 5A). Volcano plots highlight genes with significantly higher expression in the CXCL13 high, CCL21 high group compared to other groups (Figure 5B). GSEA further revealed that the genes significantly increased in expression in cases of the CXCL13 high, CCL21 high group were related to immune responses, particularly B cell activation (Figure 5C).

[IMAGE OMITTED. SEE PDF]

Discussion

In this study, we successfully induced TLS by co-administering CXCL13 and CCL21 in a mouse melanoma model, clarifying that this could enhance the therapeutic effect of ICI. Multiple chemokines are involved in the formation of TLS. Among them, CXCL13 is known to play important roles especially. By binding to its receptor CXC motif chemokine receptor 5 (CXCR5), CXCL13 promotes the mobilization and differentiation of B cells and promotes the formation of germinal centers. CXCR5 is also expressed on T cells, including follicular helper T cells, and CXCL13 is also involved in the mobilization of T cells to the TLS. In this way, the CXCL13-CXCR5 axis plays a central role in the formation and maturation of TLS [21]. It is known that CD4+ T cells produce CXCL13 in the early stages of TLS formation, and that CD21+ FDC produce CXCL13 in the stages where the TLS mature and follicles are formed. As the TLS mature, the main producer of CXCL13 shifts from CD4+ T cells to CD21+ FDC [17]. CXCL13 is involved in the formation of TLS in many cancer types, and the expression of CXCL13 is positively correlated with treatment response to ICI [8, 22–24].

In previous studies, CXCL13 administration induced TLS in a mouse ovarian cancer model. However, the combination of anti-PD-L1 antibody and CXCL13 administration did not improve the prognosis of mice [17]. In another report, the combination of CXCL13 and CCL21 administration induced TLS in a mouse pancreatic cancer model; induced TLS enhanced the anti-tumor effect of the chemotherapy (gemcitabine) [18]. The induction of TLS by chemokine administration has not been reported in mouse models of skin cancer, including melanoma, and there have been no reports that the induced TLS increased the efficacy of ICI. In the present study, we considered the importance of CCL21 in the formation of TLS based on the measurement of chemokines in the blood samples obtained from malignant melanoma patients who had received ICI treatment. In a mouse melanoma model, we administered a combination of CXCL13 and CCL21 together with anti-PD-L1 antibody and successfully induced TLS and enhanced the efficacy of ICI. While inducing TLS, CXCL13 also has a tumor-promoting effect [25, 26]. The mechanisms are not fully understood, but the following is thought to be the case: regulatory B cells induced by CXCL13 release the immunosuppressive cytokines IL-10, TGF-β, and IL-35, which suppress the activity of dendritic cells and T cells [26, 27], or B cells with activated STAT3 promote tumor angiogenesis [28]. This adverse effect of CXCL13 is a potential barrier to the application of chemokine-induced TLS induction in the treatment of cancer. In the present study, the addition of CCL21 counteracted the tumor-promoting effects of CXCL13. CCL21 is known to be involved in the recruitment of T cells, dendritic cells, and natural killer cells in the tumor microenvironment via its receptor, CC motif chemokine receptor 7 (CCR7) [29]. Previous studies in a mouse model of pancreatic cancer have shown that CCL21 increased PD-L1 expression on the surface of tumor cells and increased T-cell infiltration around the tumor, suggesting that the combination of anti-PD-L1 antibody and CCL21 may be useful for enhancing anti-tumor immunity [30]. Furthermore, CCR7 is also a receptor for CCL19 [29]. Similar to CCL21, CCL19 is involved in the recruitment of dendritic cells and T cells in the tumor microenvironment, and it has been reported that the intra-tumoral administration of CCL19 in mouse lung cancer models increased the infiltration of dendritic cells and T cells into the tumor [31].

Melanoma is a highly aggressive malignant tumor, and there is an urgent need to develop effective and diverse treatment options. In recent years, the success of ICI therapy has greatly improved treatment outcomes, but the issue remains that there are still many cases of non-response to ICI treatment. The present study suggests that the induction of TLS by the combination of CXCL13 and CCL21 in ICI treatment for malignant melanoma may be a promising new treatment strategy, although more selective administration or control of specific expression of chemokines should be necessary for application to humans. This method can also be applied to other cancers and has the potential to be an effective way to optimize immunotherapy.

Author Contributions

Maki Yoshimitsu: data curation, formal analysis, investigation, visualization, writing – original draft. Motoki Nakamura: conceptualization, funding acquisition, methodology, project administration, supervision, validation, visualization, writing – original draft, writing – review and editing. Shinji Kano: data curation, formal analysis. Tetsuya Magara: data curation, formal analysis, investigation, visualization. Hiroshi Kato: data curation, writing – review and editing. Aiko Sakai: data curation, formal analysis, investigation, writing – review and editing. Masaya Sugiyama: data curation, investigation, supervision, writing – review and editing. Masashi Mizokami: supervision, writing – review and editing. Akimichi Morita: funding acquisition, project administration, supervision, writing – review and editing.

Acknowledgments

We thank Ms. Kasuya and Ms. Nishioka for their technical assistance.

Ethics Statement

All human experiments were initiated after receiving IRB approval (Nagoya City University Clinical Trial Management Center, No. 60-22-0141, 60-23-0016). All animal experiments were approved by the Nagoya City University's Ethics Committee (No. 21-024H03).

Consent

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data generated in this study are available upon request from the corresponding author.

References

T. N. Schumacher and D. S. Thommen, “Tertiary Lymphoid Structures in Cancer,” Science 375, no. 6576 (2022): eabf9419.

G. Di Caro, F. Bergomas, F. Grizzi, et al., “Occurrence of Tertiary Lymphoid Tissue Is Associated With T‐Cell Infiltration and Predicts Better Prognosis in Early‐Stage Colorectal Cancers,” Clinical Cancer Research 20, no. 8 (2014): 2147–2158.

Y. Chen, Y. Wu, G. Yan, and G. Zhang, “Tertiary Lymphoid Structures in Cancer: Maturation and Induction,” Frontiers in Immunology 15 (2024): 1369626.

L. Vanhersecke, M. Brunet, J. P. Guégan, et al., “Mature Tertiary Lymphoid Structures Predict Immune Checkpoint Inhibitor Efficacy in Solid Tumors Independently of PD‐L1 Expression,” Nature Cancer 2, no. 8 (2021): 794–802.

C. Sautès‐Fridman, M. Lawand, N. A. Giraldo, et al., “Tertiary Lymphoid Structures in Cancers: Prognostic Value, Regulation, and Manipulation for Therapeutic Intervention,” Frontiers in Immunology 7 (2016): 407.

A. Manzo, M. Bombardieri, F. Humby, and C. Pitzalis, “Secondary and Ectopic Lymphoid Tissue Responses in Rheumatoid Arthritis: From Inflammation to Autoimmunity and Tissue Damage/Remodeling,” Immunological Reviews 233, no. 1 (2010): 267–285.

K. Neyt, F. Perros, C. H. GeurtsvanKessel, H. Hammad, and B. N. Lambrecht, “Tertiary Lymphoid Organs in Infection and Autoimmunity,” Trends in Immunology 33, no. 6 (2012): 297–305.

R. Cabrita, M. Lauss, A. Sanna, et al., “Tertiary Lymphoid Structures Improve Immunotherapy and Survival in Melanoma,” Nature 577, no. 7791 (2020): 561–565.

B. A. Helmink, S. M. Reddy, J. Gao, et al., “B Cells and Tertiary Lymphoid Structures Promote Immunotherapy Response,” Nature 577, no. 7791 (2020): 549–555.

K. T. Lynch, S. J. Young, M. O. Meneveau, et al., “Heterogeneity in Tertiary Lymphoid Structure B‐Cells Correlates With Patient Survival in Metastatic Melanoma,” Journal for Immunotherapy of Cancer 9, no. 6 (2021): e002273.

Y. H. Wu, F. Wu, G. R. Yan, et al., “Features and Clinical Significance of Tertiary Lymphoid Structure in Cutaneous Squamous Cell Carcinoma,” Journal of the European Academy of Dermatology and Venereology 36, no. 11 (2022): 2043–2050.

C. Byers, M. Gill, N. R. Kurtansky, et al., “Tertiary Lymphoid Structures Accompanied by Fibrillary Matrix Morphology Impact Anti‐Tumor Immunity in Basal Cell Carcinomas,” Frontiers in Medicine (Lausanne) 9 (2022): 981074.

M. Nakamura, T. Magara, S. Kano, A. Matsubara, H. Kato, and A. Morita, “Tertiary Lymphoid Structures and Chemokine Landscape in Virus‐Positive and Virus‐Negative Merkel Cell Carcinoma,” Frontiers in Oncology 12 (2022): 811586.

T. Magara, M. Nakamura, Y. Nojiri, et al., “Tertiary Lymphoid Structures Correlate With Better Prognosis in Cutaneous Angiosarcoma,” Journal of Dermatological Science 103, no. 1 (2021): 57–59.

T. Magara, M. Nakamura, Y. Nojiri, et al., “Tumor Immune Microenvironment of Cutaneous Angiosarcoma With Cancer Testis Antigens and the Formation of Tertiary Lymphoid Structures,” Frontiers in Oncology 13 (2023): 1106434.

T. Magara, M. Nakamura, M. Yoshimitsu, et al., “Tertiary Lymphoid Structures Associated With Improved Prognosis and Reduced Invasion in Advanced Extramammary Paget's Disease,” Clinical and Experimental Dermatology 50 (2024): llae439.

M. Ukita, J. Hamanishi, H. Yoshitomi, et al., “CXCL13‐Producing CD4+ T Cells Accumulate in the Early Phase of Tertiary Lymphoid Structures in Ovarian Cancer,” JCI Insight 7, no. 12 (2022): e157215.

F. R. Delvecchio, R. E. A. Fincham, S. Spear, et al., “Pancreatic Cancer Chemotherapy Is Potentiated by Induction of Tertiary Lymphoid Structures in Mice,” Cellular and Molecular Gastroenterology and Hepatology 12, no. 5 (2021): 1543–1565.

L. Xu, S. S. Shen, Y. Hoshida, et al., “Gene Expression Changes in an Animal Melanoma Model Correlate With Aggressiveness of Human Melanoma Metastases,” Molecular Cancer Research 6 (2008): 760–769.

S. X. Ge, E. W. Son, and R. Yao, “iDEP: An Integrated Web Application for Differential Expression and Pathway Analysis of RNA‐Seq Data,” BMC Bioinformatics 19 (2018): 534.

M. Hoellwerth, P. Koelblinger, R. Lang, and A. Harrer, “Revisiting the Role of the CXCL13/CXCR5‐Associated Immune Axis in Melanoma: Potential Implications for Anti‐PD‐1‐Related Biomarker Research,” Life (Basel) 13, no. 2 (2023): 553.

M. Rouanne, N. Arpaia, and A. Marabelle, “CXCL13 Shapes Tertiary Lymphoid Structures and Promotes Response to Immunotherapy in Bladder Cancer,” European Journal of Cancer 151 (2021): 245–248.

C. S. Groeneveld, J. Fontugne, L. Cabel, et al., “Tertiary Lymphoid Structures Marker CXCL13 Is Associated With Better Survival for Patients With Advanced‐Stage Bladder Cancer Treated With Immunotherapy,” European Journal of Cancer 148 (2021): 181–189.

S. Park, H. Cha, H. S. Kim, et al., “Transcriptional Upregulation of CXCL13 Is Correlated With a Favorable Response to Immune Checkpoint Inhibitors in Lung Adenocarcinoma,” Cancer Medicine 12, no. 6 (2023): 7639–7650.

J. Ren, T. Lan, T. Liu, et al., “CXCL13 as a Novel Immune Checkpoint for Regulatory B Cells and Its Role in Tumor Metastasis,” Journal of Immunology 208, no. 10 (2022): 2425–2435.

A. J. Rubio, T. Porter, and X. Zhong, “Duality of B Cell‐CXCL13 Axis in Tumor Immunology,” Frontiers in Immunology 11 (2020): 521110.

J. M. Berthelot, C. Jamin, K. Amrouche, B. Le Goff, Y. Maugars, and P. Youinou, “Regulatory B Cells Play a Key Role in Immune System Balance,” Joint, Bone, Spine 80, no. 1 (2013): 18–22.

C. Yang, H. Lee, S. Pal, et al., “B Cells Promote Tumor Progression via STAT3 Regulated‐Angiogenesis,” PLoS One 8, no. 5 (2013): e64159.

A. Salem, M. Alotaibi, R. Mroueh, H. A. Basheer, and K. Afarinkia, “CCR7 as a Therapeutic Target in Cancer,” Biochimica Et Biophysica Acta. Reviews on Cancer 1875, no. 1 (2021): 188499.

Q. Chen, H. Yin, N. Pu, et al., “Chemokine C‐C Motif Ligand 21 Synergized With Programmed Death‐Ligand 1 Blockade Restrains Tumor Growth,” Cancer Science 112, no. 11 (2021): 4457–4469.

S. Hillinger, S. C. Yang, R. K. Batra, et al., “CCL19 Reduces Tumour Burden in a Model of Advanced Lung Cancer,” British Journal of Cancer 94, no. 7 (2006): 1029–1034.