Prognostic impact and landscape of cellular CXCR5 chemokine receptor expression in clear-cell renal cell carcinoma

Abstract

CXCR5 is a chemokine receptor that promotes B cell follicular formation and antibody production. Indeed, CXCR5 has been found to be expressed in a variety of cancers; however, the role of CXCR5 expression in clear-cell renal cell carcinoma (ccRCC) remains unclear. We aimed to determine the impact of cellular CXCR5 expression on cancer outcomes, the PD-1/PD-L1 axis, and genetic states in patients with ccRCC. First, multiplex immunofluorescence staining for CXCR5, CD4, CD8, and AE1/AE3, along with automated single-cell counting, was performed to assess cellular CXCR5 expression in ccRCC and its association with prognosis. Second, the tumour microenvironment (TME) was analysed, with a focus on the relationship between the PD-1/PD-L1 axis and CXCR5 expression. Finally, an integrated analysis of CXCR5 expression and genomic mutation information was conducted to reveal the genetic background underlying CXCR5 expression. A total of 105 ccRCC patients were included. Among the 696,964 cells analysed, the distribution of CXCR5-expressing cells was as follows: 30% CXCR5⁺CD4⁺ cells, 9% CXCR5⁺CD8⁺ cells, and 26% CXCR5⁺AE1/AE3⁺ cells. Survival analysis revealed that tumours with low-CXCR5⁺CD8⁺ cells had a poor prognosis; TME analysis revealed a relationship between low-CXCR5⁺CD8⁺ status and a highly suppressive PD-L1-positive immune environment. Genomic analysis revealed a correlation between low-CXCR5⁺CD8⁺ status and high rates of alterations in chromatin remodelling genes, including PBRM1. This study highlights the significance of CXCR5⁺CD8⁺ cells in ccRCC, demonstrating their clinical implications and revealing the immunogenomic landscape underlying CXCR5 expression.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00262-025-04020-4.

Introduction

Chemokines are a large family of small, functionally diverse chemotactic cytokines that are expressed by cells in the body during homeostasis and during inflammatory responses [1]. Chemokines facilitate the migration of immune cells to sites of inflammation, e.g. sites of infection or tumour development, where cytokines may bind to specific chemokine receptors on the cell surface and activate downstream signalling pathways. Notably, in cancer patients, chemokines play important roles in differentially regulating the infiltration of various immune cell subsets within the tumour microenvironment (TME), inducing T cell migratory behaviour and affecting therapeutic outcomes [2, 3].

CXCR5 is one of the chemokine receptors that promotes B cell follicle formation and antibody production during inflammation through interactions with its ligand CXCL13 [4]. The clinical role of CXCR5 expression in tumours is controversial. In some tumours, CXCR5 expression has been associated with tumour progression and metastasis [5–11]. On the other hand, CXCR5 has also been reported to exert antitumour effects because the CXCR5 axis promotes the migration of antitumour immune cells into tumours. For example, the CXCR5 axis is required for B cell entry and organization into tertiary lymphoid structures, which promote efficient antigen presentation and the activation of antitumour T cell and B cell responses [12, 13]. However, the role of CXCR5 expression in renal cell carcinoma (RCC) remains unclear.

RCC is one of the most common malignancies of the urinary system in adults, accounting for approximately 2.2% of all tumours worldwide [14]. The clear-cell subtype, i.e. clear-cell renal cell carcinoma (ccRCC), accounts for approximately 75% of RCCs [15]; ccRCC exhibits high tumour neoantigen levels on the basis of insertions and deletions [16]. Immune checkpoint inhibitors (ICIs) targeting the PD-1/PD-L1 axis and CTLA-4 are the standard of care for advanced ccRCC; however, response rates remain insufficient [17, 18]. Furthermore, a high number of CD8⁺ T cells has been associated with a poor ccRCC prognosis [19]. Recent studies have revealed that CD8⁺ T cells in the ccRCC TME are exhausted, exhibit heterogeneous phenotypes, and express immune-evasive molecules (PD-1, PD-L1, PD-L2, and CTLA-4) [20, 21]. A more detailed understanding of the role of T cell subpopulations in RCC is essential for developing future biomarkers and novel therapeutic targets. In this context, CXCR5 is a candidate chemokine receptor that has the potential to play a significant role.

We herein applied a recent multiplexed imaging technique to examine the relationship between CXCR5 expression and oncological outcomes in individual subsets of T cells and cancer cells in ccRCC. We used a novel tyramide signal amplification method in combination with multispectral image analysis software [22, 23], allowing automatic single-cell stratification and revealing protein coexpression. The aim of the present study was to clarify the clinical significance and immune correlates of CXCR5 in individual subsets of T cells and cancer cells in ccRCC and provide important insights that will inform future practice.

Methods

Human tumour samples

After Institutional Review Board approval, formalin-fixed and paraffin-embedded (FFPE) tumour specimens obtained at Keio University Hospital (Tokyo, Japan) between 1999 and 2017 were randomly collected and grouped on the basis of the histological type, pathological T stage, and systematic treatment used. The UICC TNM system was used for tumour staging, and nuclear grading was performed according to the WHO/International Urologic Pathology Society grading system. The cohort details were identical to those used in our previous study [24–26]; primary ccRCC tumours were treated surgically (n = 105, Table 1). Statistical methods were not used to predetermine the sample size.

Table 1.

Characteristics of 105 patients with clear-cell renal cell carcinoma

	All (n = 105)	CXCR5+ CD4+T cell		p vale	CXCR5+CD8+T cell		p vale	CXCR5+AE1/AE3+T cell		p vale
		High (n = 52)	Low (n = 53)		high (n = 51)	low (n = 54)		high (n = 51)	low (n = 54)
Sex, n(%)				0.043			0.203			0.203
Male	76 (72)	33 (63)	43 (81)		34 (67)	42 (78)		34 (67)	42 (78)
Female	29 (28)	19 (37)	10 (19)		17 (33)	12 (22)		17 (33)	12 (22)
Age, n (%)				0.907			0.964			0.658
< 65 years	62 (59)	31 (60)	31 (58)		30 (59)	32 (59)		29 (57)	33 (61)
≥ 65 years	43 (41)	21 (40)	22 (42)		21 (41)	22 (41)		22 (43)	21 (39)
Tumour size, n (%)				0.923			0.623			0.623
< 60 mm	53 (50)	26 (50)	27 (51)		27 (53)	26 (48)		27 (53)	26 (48)
≥ 60 mm	52 (50)	26 (50)	26 (49)		24 (47)	28 (52)		24 (47)	28 (52)
Tumour grade, n (%)				0.452			0.307			0.541
G1 + G2	69 (66)	36 (69)	33 (62)		36 (71)	33 (61)		35 (69)	34 (63)
G3 + G4	36 (34)	16 (31)	20 (38)		15 (29)	21 (39)		16 (31)	20 (38)
Pathological T stage, n (%)				0.602			0.444			0.973
pT1 + pT2	64 (61)	33 (63)	31 (58)		33 (65)	31 (57)		31 (61)	33 (33)
pT3 + pT4	41 (39)	19 (37)	22 (42)		18 (35)	23 (43)		20 (39)	21 (21)
Venous invasion, n (%)				0.951			0.537			0.537
Yes	30 (29)	15 (29)	15 (28)		16 (31)	14 (26)		16 (31)	14 (26)
No	75 (71)	37 (71)	38 (72)		35 (69)	40 (74)		35 (69)	40 (74)

The tissue microarray (TMA) blocks used in this study were prepared as follows. An experienced pathologist (S.M.), board certified in urogenital malignancies, evaluated the suitability of tissue sections for TMA construction. The tumour centre region was selected, and 3-mm tumour cores were punched out from this region. All samples were deidentified by assigning a number to avoid investigator bias during tissue preparation and data analysis. All procedures were performed in accordance with the 1964 Declaration of Helsinki, current ethical standards, and hospital ethical guidelines.

Multiplex immunofluorescence staining and cellular image analysis

Multiplex immunofluorescence staining was performed using the Opal Polaris seven-colour Manual IHC Kit (Akoya Biosciences, Marlborough, Massachusetts, USA). TMA samples were cut into 5-μm-thick sections and placed on silane-coated glass slides. Each slide was baked in an oven at 65 °C for 1 h, deparaffinized with xylene, and rehydrated with a graded ethanol solution. They were then fixed in 0.3% hydrogen peroxide buffer (25 mM Tris-HCl (pH 7.5), 150 mM NaCl, and 0.05% Tween 20). After blocking, the sections were incubated with primary antibodies for 20 h and with the polymer HRP Ms + Rb secondary antibody at room temperature for 10 min. Opal fluorescent material was pipetted onto each slide, and after standing at room temperature for 10 min, the slides were heated in a microwave oven to detach the primary and secondary antibodies. The same protocol was then repeated with the next primary antibody target. Only the last opal fluorophore was pipetted in two steps. DAPI was pipetted onto each slide, which was subsequently incubated at room temperature for 5 min. The slides were covered with ProLong Diamond Antifade Mountant (Thermo Fisher Scientific). The primary antibodies used were as follows: CXCR5 (1:50, 51,505, R&D Systems, Minneapolis, MN, USA), CD4 (Diluted, 4B12, Nichirei, Tokyo, Japan), CD8 (Diluted, C8/144B, Nichirei, Tokyo, Japan), and AE1/AE3 (1:2, IR0530, Dako, Santa Clara, CA, USA). Separately, multiplex immunofluorescence staining with PD-1, PD-L1, and DAPI staining was also performed on the same TMA samples using the method described above. The incubation time with the primary antibodies was 4 h, and the primary antibodies used were as follows: anti-PD-1 (1:200, #43,248; Cell Signaling Technology, Danvers, MA, USA) and anti-PD-L1 (1:100, #13,684; Cell Signaling Technology, Danvers, MA, USA) antibodies.

TMA samples were scanned with a Vectra Polaris digital slide scanner (Akoya Biosciences) and analysed using inForm software ver. 2.6 (Akoya Biosciences), allowing for the separation and measurement of spectrally overlapping markers in current multiplex assays at the single-cell level. Cell segmentation was performed by locating individual cell nuclei separately using the DAPI-based counterstain-based approach. A multicolour phenotype was applied to 4 antigens via intensity-based thresholds selected manually on inForm images. We classified the samples into two groups (high expression and low expression) using the median of the density of each marker as the cut-off value.

DNA extraction and sequencing

Genomic DNA was extracted from 43 fresh-frozen ccRCC samples that matched the TMA samples with the DNeasy Blood & Tissue Kit (Qiagen) according to the manufacturer’s protocol. The DNA integrity number was 4.0, which was calculated using an Agilent 2000 TapeStation (Agilent Technologies, Waldbronn, Germany). A genomic DNA library was constructed via the GeneRead DNAseq Targeted Panel V2 (Human Comprehensive Cancer Panel) by QIAGEN, covering more than 95% of the total exon region in 160 cancer-related genes [27]. The library was amplified via the GeneRead DNA I Amp Kit (Qiagen) and sequenced via MiSeq (Illumina). FastQ files obtained from MiSeq (Illumina) were analysed via an original bioinformatics pipeline called GenomeJack (Mitsubishi Space Software, Tokyo, Japan). In brief, protein damage, such as “loss of function” and “gain of function”, was assessed by JAX CKB (https://ckb.jax.org/), and “pathogenic/likely pathogenic” single nucleotide variants were annotated by ClinVar or CiVIC, respectively.

Statistical analysis

Values are presented as the means with standard errors or medians with interquartile ranges for continuous variables and as frequencies with percentages for categorical variables. Differences in variables were evaluated via the chi-square test and the Mann‒Whitney U test, as appropriate. Univariate and multivariate Cox regression models with stepwise selection were used to evaluate variables related to disease recurrence and overall mortality. Survival curves were estimated using the Kaplan‒Meier method and compared via the log-rank test. The significance of differences was set as p < 0.05. All analyses were performed using SPSS version 27.0 (IBM-SPSS Inc., Tokyo, Japan) and JMP version 17.0 (SAS Institute Inc., Cary, NC, USA).

Results

Multiplexed immunostaining and cellular phenotyping

Tumour samples obtained from 105 primary ccRCC patients who did not receive neoadjuvant or adjuvant therapy were retrospectively analysed. CXCR5 expression and its colocalization with immune and cancer cell markers (CD4, CD8, and AE1/AE3) were assessed at single-cell resolution via quantitative multiplex immunofluorescence staining with a tyramide signal amplification system (Fig. 1a–d). Automated cell segmentation and single-cell counting were performed using the inForm software system, and nuclear staining with DAPI enabled the accurate identification of live cells and debris.

Fig. 1.

CXCR5 expression and colocalization with immune and cancer cell lineage markers (CD4, CD8, and AE1/AE3) in the TME of ccRCC. a A representative image of ccRCC displaying multispectral five-colour fluorescence signals for CXCR5, CD4, CD8, and AE1/AE3 together with DAPI. Scale bar, 200 μm. b–d A zoomed-in image of the indicated boxed region. Scale bar, 20 μm. e. The distribution of individual CXCR5-positive cells

The fluorescence intensity of each cell was recorded individually, allowing for the phenotypic classification of each cell. The present study outlined the CXCR5 expression patterns of CD4⁺ and CD8⁺ CTLs and AE1/AE3⁺ cells. Individual CXCR5-positive cells were counted from 696,964 cells, 3% of which were positive for CXCR5. Among the CXCR5⁺ cells, 30% were CXCR5⁺CD4⁺, 9% were CXCR5⁺CD8⁺, and 26% were CXCR5⁺AE1/AE3⁺ (Fig. 1e).

Survival analysis

The clinicopathological characteristics of 105 patients with ccRCC in our cohort are shown in Table 1: 28% were female, 41% were older than 65 years, 50% had tumours larger than 60 mm, 39% had stage pT3/4, 34% had tumour grade G3/G4, and 29% had positive venous invasion (Table 1). To assess the prognostic impact of CXCR5 expression, we initially stratified the cohort according to the CXCR5 expression patterns of CD4⁺, CD8⁺ tumour-infiltrating lymphocytes (TILs), and AE1/AE3⁺ cancer cells using the percentage of cells with positive expression of each molecule among the total number of cells in the centre of the tumour. The actual high and low cut-offs for the CXCR5⁺, CXCR5⁺CD4⁺, CXCR5⁺CD8⁺, and CXCR5⁺AE1/AE3⁺ groups were set to median values of 1.71%, 0.44%, 0.11%, and 0.16%, respectively. Among the patient characteristics, high-CXCR5⁺CD4⁺ expression was related to female sex (p = 0.043), with no other clinical indices showing correlations (Table 1).

During follow-up, 59 cases of metastasis were observed, with overlapping metastatic sites: lung in 36 cases, liver in 8 cases, brain in 4 cases, bone in 17 cases, and lymph nodes in 18 cases. The initial treatment regimens included tyrosine kinase inhibitors (TKIs) in 55 patients, ICIs and TKIs in 3 patients, and interferon (IFN) alfa in 1 patient. We then evaluated the relationships between overall survival (OS) and recurrence-free survival (RFS) and the percentage of cells positive for each molecule (CXCR5⁺ alone, colocalized CXCR5⁺CD4⁺, CXCR5⁺CD8⁺, and CXCR5⁺AE1/AE3⁺).

When patients were stratified by the percentage of CXCR5⁺ cells alone, Kaplan‒Meier analysis revealed that patients in the low-CXCR5 expression group had slightly shorter RFS (p = 0.011) and OS (p = 0.150) (Fig. 2a‒b). To further understand the clinical value of CXCR5 expression in the tumour space, we assessed which type of CXCR5-expressing cells was most closely associated with prognosis. Patients in the low-CXCR5⁺CD8⁺ expression group had shorter OS (p = 0.024) and RFS (p = 0.018) than did those in the high-CXCR5⁺CD8⁺ expression group (Fig. 2a–b). No significant differences were observed in OS or RFS between the other groups.

Fig. 2.

Survival analysis in ccRCC patients grouped according to CXCR5 expression levels. a–b Kaplan‒Meier survival analyses of postoperative overall (a) and recurrence-free survivals (b) in 105 ccRCC patients grouped according to high/low proportions of CXCR5⁺ cells alone or CXCR5⁺CD4⁺, CXCR5⁺CD8⁺, and CXCR5⁺AE1/AE3⁺ cells. p values < 0.05 were considered to indicate significance

In multivariate analyses, low-CXCR5⁺CD8⁺ expression was a significant predictor of OS (HR 2.223, p = 0.047), independent of large tumour size (HR 1.017, p = 0.005), and high tumour grade (HR 3.437, p = 0.003). Similarly, for RFS, low-CXCR5⁺CD8⁺ expression remained a significant factor (HR 1.722, p = 0.048), independent of large tumour size (HR 1.012, p = 0.010), high tumour grade (HR 2.537, p = 0.003), and advanced tumour stage (HR 1.910, p = 0.027) (Table 2).

Table 2.

Univariate and multivariate Cox regression analyses of 105 patients with clear-cell renal cell carcinoma following surgery

Variables	Overall survival			Recurrence-free survival
	Univariate	Multivariate	p value	Univariate	Multivariate	p value
	p value	HR (95% CI)		p value	HR (95% CI)
Sex (male vs female)	0.077			0.713
Age (≥ 65 years vs < 65 years)	0.023			0.139
Tumour size (mm)	< 0.001	1.017 ( 1.005–1.029)	0.005	< 0.001	1.012 ( 1.003–1.022)	0.010
Tumour grade (G3/4 vs G1/2)	< 0.001	3.437 ( 1.515–7.798)	0.003	< 0.001	2.537 ( 1.379–4.668)	0.003
Pathological T stage (pT3/4 vs pT1/2)	< 0.001			< 0.001	1.910 (1.077-3.387)	0.027
Venous invasion (positive vs negative)	0.195			0.059
CXCR5+ T cell (low vs high)	0.150			0.012
CXCR5+ CD4+ T cell (low vs high)	0.471			0.307
CXCR5+ CD8+ T cell (low vs high)	0.024	2.223 ( 1.009–4.897)	0.047	0.022	1.722 ( 1.004–2.953)	0.048
CXCR5+ AE1/AE3+ T cell (low vs high)	0.477			0.532

Therefore, cell stratification at the single-cell level revealed that CXCR5 expression in CD8⁺ TILs was the most prognostically relevant factor in patients with ccRCC.

CXCR5⁺CD8⁺expression, immune checkpoint molecules, and genomic landscapes

To understand the immunological background of the high incidence and poor prognosis of CXCR5⁺CD8⁺ TILs, we examined the relationship of the percentage of cells in this subset with the expression of PD-1/PD-L1, i.e. representative immune checkpoint molecules. The percentages of PD-1- and PD-L1-positive cells in the two groups of CXCR5⁺CD8⁺ cells were analysed in 90 ccRCC samples for which PD-L1 expression information was available in the current cohort.

PD-L1 expression was higher in the low-CXCR5⁺CD8⁺ expression group than in the high- CXCR5⁺CD8⁺ expression group (p = 0.040). However, there was no significant difference in the rate of PD-1 expression between the two CXCR5⁺CD8⁺ expression groups (Fig. 3a–b). We also assessed differences in PD-1/PD-L1 expression among the CXCR5⁺, CXCR5⁺CD4⁺, and CXCR5⁺AE1/AE3⁺ cell groups (Fig. 3a–b); PD-L1 expression was higher in the low-CXCR5⁺ (p < 0.001) and low-CXCR5⁺CD4⁺ cell groups (p = 0.005).

Fig. 3.

Comparison of the density of PD-1- and PD-L1-positive cells across high- and low-expression groups of CXCR5⁺ alone, CXCR5⁺CD4⁺, CXCR5⁺CD8⁺, and CXCR5⁺AE1/AE3⁺ in ccRCC patients. a PD-1. b PD-L1. p values < 0.05 were considered to indicate significance

An analysis of genomic alterations in the high/low-CXCR5⁺CD8⁺ expression group revealed prominent features of ccRCC tumours that have potential as biomarkers for targeted cancer immunotherapy. We examined 43 ccRCC samples for which genomic alteration information was available in the current cohort, comprising high (n = 20) vs. low (n = 23) CXCR5⁺CD8⁺ expression (Fig. 4a).

Fig. 4.

Somatic genetic alterations related to the levels of CXCR5⁺CD8⁺ expression in ccRCC patients. a Genomic alterations in tumorigenic signalling pathways associated with ccRCC development stratified by CXCR5⁺CD8⁺ expression levels. b Percentage of samples with alterations in each of the selected signalling pathways. p values < 0.05 were considered to indicate significance

The assessment of 160 defined cancer-associated genes (see the Methods section) revealed that the most frequently altered genes were VHL, PBRM1, SETD2, and MTOR. Comparisons of typical cancer-associated pathway alterations revealed unique genetic features according to CXCR5⁺CD8⁺ expression profiles. Alterations in chromatin remodelling genes, e.g., PBRM1, were more frequent in the CXCR5⁺CD8⁺ low group (48% of patients) than in the high group (15%). However, there were no significant differences between the two groups in terms of the subsets of PI3K‒mTOR pathway genes, TP53/cell cycle genes, HIF1 signalling genes, or histone modification genes (Fig. 4b).

Therefore, ccRCC tumour samples with low-CXCR5⁺CD8⁺ expression were more likely to have high PD-L1 expression and more frequent alterations in chromatin remodelling genes.

Discussion

The clinical success of cancer immunotherapy has underscored the importance of immunosuppressive mechanisms within the TME. Chemokines are a family of small chemotactic molecules that regulate immune cell trafficking by binding to receptors on various cells, including immune cells, endothelial cells, mesenchymal cells, and cancer cells, thereby directly affecting their survival and proliferation [1–3]. Therefore, targeting the chemokine signalling pathway in combination with current ICIs has potential as an approach to improve therapeutic outcomes in cancers [2].

In the past decade, several chemokine receptors and ligands have been evaluated in relation to RCC. For example, elevated levels of the chemokine receptor ligand CXCL9 are associated with a poor ccRCC prognosis [28]. Similarly, high expression of the chemokine receptor CXCR2 has also been reported to be associated with a poor prognosis for nonmetastatic ccRCC [29]. In a mouse model, CXCR2 inhibitors potentiated the effects of anti-CTLA-4 and anti-PD-1 agents by inhibiting the recruitment of tumour-associated M2 macrophages and tumour-associated neutrophils; certain therapeutic effects may be observed in human RCC tumours when CXCR2 inhibitors are combined with anti-PD-1 therapy [30].

On the other hand, CXCR5 is a chemokine receptor that interacts with ligands to promote B cell folliculogenesis and antibody production during inflammation, and its expression in tumours has been confirmed in various cancers [4, 31, 32]. However, its role in renal cancer remains unclear. In the present study, we evaluated the relationship between oncological outcomes in patients with ccRCC and the expression density of CXCR5 in individual subsets of immune cells and cancer cells via multiple new staining methods and obtained important results that may have implications for future practice.

Our multiplex immunofluorescence data revealed diversity among CXCR5-expressing cells, indicating that ccRCC tumours with a high proportion of CXCR5⁺CD8⁺ TILs had a better prognosis. CXCR5⁺CD8⁺ TILs, also known as T follicular cytotoxic cells, are a distinct subset of CD8⁺ T cells identified for their strong effector functions in infectious diseases [33, 34]. In various cancers, a high density of CXCR5⁺CD8⁺ TILs has been associated with better OS [13, 35–37]. CXCR5⁺CD8⁺ T cells increase the expression of effector molecules, including IFN-γ, TNF-α, and granzymes, and exert stronger antitumour effects than their counterparts [38, 39]. Treatment with PD-L1 was shown to significantly enhance the antitumour effects of CXCR5⁺CD8⁺ TILs in liver cancer [40].

Furthermore, an assessment of the potential for immune suppression may contribute to a more detailed understanding of the characteristics of environments with a low number of CXCR5⁺CD8⁺ T cells. We found that tumours with low-CXCR5⁺CD8⁺ T cells had a higher percentage of PD-L1-positive cells; PD-L1 expression has been associated with a poor prognosis and increased recurrence rates through its suppression of T cell activity and induction of T cell exhaustion [41–44]. This immunosuppressive environment may contribute to the poor prognosis observed in the low-CXCR5⁺CD8⁺ T cell group of ccRCC patients. Although the mechanisms underlying the interaction between CXCR5⁺CD8⁺ T cells and the PD-1/PD-L1 axis remain unclear, previous findings indicated that blockade of PD-L1 improved the cytotoxicity mediated by CXCR5⁺CD8⁺ T cells [40].

Moreover, a genetic alteration analysis based on the CXCR5⁺CD8⁺ status revealed a high prevalence of alterations in chromatin remodelling genes, including PBRM1 (40%), in the low-CXCR5⁺CD8⁺ group. The loss of PBRM1 was previously linked to a better ICI response in ccRCC [45, 46]. However, a recent study indicated that the loss of PBRM1 inhibited the JAK-STAT pathway, resulting in the reduced expression of IFN-γ target genes [47]; therefore, the loss of PBRM1 in ccRCC may be linked to an immunosuppressive microenvironment [47]. PBRM1 alterations have also been associated with tumour invasiveness, including advanced tumour stage and high nuclear grade, as well as poor OS in patients with ccRCC [48–50]. These findings suggest a genetic interaction in tumours with CXCR5⁺CD8⁺ cell infiltrates in ccRCC.

There are some limitations that need to be addressed. This was a retrospective study of a limited number of patients, and in this cohort, most patients with tumour metastasis were treated with molecularly targeted therapy rather than ICIs. In addition, genetic assessment was performed for a small number of patients, and the associations between CXCR5⁺CD8⁺ status and specific genetic mutations in this cohort could not be fully assessed. Furthermore, substantial data supporting the biological significance of CXCR5⁺CD8⁺ TILs in ccRCC are lacking. In addition, whether the status of CXCR5⁺CD8⁺ TILs is useful as a predictive biomarker needs to be evaluated in prospective studies.

In summary, analysis of CXCR5 dynamics in ccRCC revealed the prognostic impact of CXCR5⁺CD8⁺ T cells, immune microenvironment status, and genetic characteristics. The present results provide a more detailed understanding of the immunobiological mechanisms underlying resistance to immune-oncological and molecular-targeted therapies and may indicate new treatment targets or biomarkers in ccRCC.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

N.T. and M.O. designed the study. M.A., K.T., T.M., and S.M. performed the experiments. T.I. provided conceptual advice. K.N. and H.N. performed genetic experiments. M.A. and N.T. wrote the manuscript.

Funding

Funding was provided by SGH Foundation for Cancer Research, Princess Takamatsu Cancer Research Fund, Takeda Science Foundation, Foundation for Promotion of Cancer Research Japan, Keio Gijuku Academic Development Fund, KAKENHI: 21K19414, 23H03217, 24K21301 to N.T.; 24H00649 to M.O., Nippon Shinyaku.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Competing Interests

The authors declare no competing interests.