Prognostic value of B7-H3 expression in metastatic renal cell carcinoma and its impact on immunotherapy response

Abstract

Background

Renal cell carcinoma (RCC) is characterised by its immunogenic and proangiogenic nature and its resistance to conventional therapies. The advent of immune checkpoint inhibitors (ICIs) and tyrosine kinase inhibitors (TKIs) has significantly improved patient survival, but resistance to these treatments remains a challenge. B7-H3, a potential immune checkpoint, has been implicated in modulating the tumour microenvironment and immune escape mechanisms in RCC.

Methods

Immunohistochemical analysis of B7-H3 expression was performed in 84 metastatic RCC patients. Tissue microarrays and separate sections of formalin-fixed paraffin-embedded tissue were used for immunohistochemical staining. Membranous staining of the tumor cells was scored and statistical analyses were performed to assess the correlation between B7-H3 expression and treatment outcome.

Results

B7-H3 expression was absent in 31% of patients, while 33.3% had a score of 1+, 31% had 2+, and 4.8% had 3+. High B7-H3 expression correlated with poorer OS (20 months vs. 45 months, p = 0.012). In patients receiving nivolumab, those with high B7-H3 expression had shorter PFS (2 months vs. 8 months, p = 0.037) and OS (17 months vs. 51 months, p = 0.01). B7-H3 expression was the only factor significantly affecting PFS and OS in multivariate analysis.

Conclusion

High B7-H3 expression is associated with poorer survival outcomes and reduced response to nivolumab in metastatic RCC patients. B7-H3 may serve as a predictive biomarker for immunotherapy response. Future studies should explore targeting B7-H3 in combination with existing therapies to enhance treatment efficacy.

Introduction

Renal cell carcinoma (RCC) is the most frequent type of renal neoplasm, representing for approximately 80% of cases. Unfortunately, 20 to 30% of people with RCC present with de novo metastatic disease. Additionally, approximately 20% of patients relapse after surgery, increasing the number of patients requiring systemic treatment [1]. The development of systemic therapy takes into account the immunogenic and proangiogenic nature of RCC, as well as its resistance to radiotherapy and chemotherapy.

Following the discovery of immune regulatory mechanisms known as immune checkpoints, the treatment approach for RCC has shifted towards combining two different immune checkpoint inhibitors (ICI) or tyrosine kinase inhibitors (TKI) with ICI, resulting in significant improvements in patient survival. However, many tumors may develop secondary resistance to antiangiogenesis due to mutation of the target protein or compensatory changes in the target pathway that bypass the inhibition site. Additionally, resistance to immunotherapy arises from a myriad of factors, with only a minority of patients achieving five-year survival post-diagnosis [2, 3]. Therefore, it is crucial to identify biomarkers that can detect tumour-specific escape and resistance mechanisms and predict treatment response. While various factors have been evaluated, such as tumour mutation burden, PD-L1 expression, cytotoxic T cells, cytokines, and angiogenesis factors, they have not been sufficient in predicting ICI response in RCC [4]. mRNA-based genomic signatures and various immune checkpoint molecules are promising in this regard, and the search for optimal biomarkers is ongoing [3].

Immune checkpoint molecule expression suggests a potential role as a biomarker reflecting negative feedback mechanisms that reduce sustained T cell activation [5]. In addition, it is hypothesised that selective targeting of different immune checkpoints may improve therapeutic outcomes by regulating the cascade of T regulatory cells (Tregs) and effector T cells Novel immunotherapeutic approaches have been explored, including IDO-1 inhibitors, CSFIR inhibitors, STING agonists and RIG-1 agonists, as well as those targeting immune checkpoints such as VISTA, TIM-3 and LAG3. These approaches aim to disrupt the tumour microenvironment and target alternative immune regulatory mechanisms that contribute to immune escape, potentially circumventing ICI resistance [3, 6–8]. Clinical trials investigating the inhibition of additional checkpoint molecules have been conducted in certain tumour types, most notably melanoma, and have shown positive results [9, 10].

The B7 family includes some of the integral membrane proteins present on activated antigen-presenting cells, as well as structurally similar cell surface protein ligands that bind to lymphocyte receptors. When B7-H3 (also known as CD276) first appeared in the literature, it was thought to be required for T cell costimulation, activation of autoimmune diseases and even induction of acute renal rejection [11, 12]. More recently, B7-H3 has been found to prevent the development and proliferation of both CD4 + and CD8 + T cells. A study found that B7-H3 inhibited T cell activity by upregulating protein 1 signaling pathways, nuclear factor of activated T cells, and NF-κB [13]. In another study, the B7-H3 deficient led to a significant rise in the synthesis of granzyme B, Ki-67, IFN-γ, and TNF-α proliferation indicators in CD8 + T cells and a notable drop in the levels of other co-inhibitory molecules, such as PD-1. These results imply that B7-H3 is a key player in the control of CD8 + T cell depletion. Within the same model, it was discovered that CD4 + T cells and NK cells entered an active state of TNF-α and IFN-γ production [14].

The high expression of B7-H3 has been demonstrated to be significantly associated with the density of FOXP3 + regulatory T cells (Treg) in RCC [15] and a positive correlation between the number of FOXP3 + Treg cells and B7-H3 expression was found in nonsmall cell lung cancer (NSCLC) [16]. Furthermore, a study showed a significant correlation between B7-H3 expression in NSCLC and lack of response to anti-PD-1 immunotherapy. In addition, a mouse model with dual blockade of PD-L1 and B7-H3 showed improved anti-tumour activity, suggesting that B7-H3 may be a promising option in combination with anti-PD-1 treatment [17]. A meta-analysis shows that increased B7-H3 expression is associated with aggressive disease and poorer overall survival in several solid tumours [18].

The aim of this study was to investigate and define the role of B7-H3 expression as a novel marker to predict response to immunotherapy in metastatic RCC.

Materials and methods

Patient characteristics and study design

This is a retrospective cohort study (8160-GOA) that has been approved by the Dokuz Eylül University Ethics Committee. All participants/guardians have provided written informed consent. Renal cell carcinoma patients undergoing follow-up in the Medical Oncology Clinic of Dokuz Eylül University Hospital between 2012 and 2023 were screened. A total of 167 patients with clear cell, papillary, chromophobe, translocation-associated and unclassified RCC were included in the study. As a total of 84 patients, 49 denovo metastatic patients and 35 metastatic relapsed patients were included in the study. Patients whose pathology specimens were not available (n = 49), patients undergoing diagnostic biopsy of metastases (n = 14) and patients with early stage tumours without metastases undergoing curative surgery (n = 18) were excluded. Patients diagnosed from metastasis were excluded due to the possibility of B7-H3 expression discordance between primary and metastatic sites. Metastatic patients under active surveillance (n = 5) were included; oligometastatic patients undergoing nephrectomy and metastasectomy (n = 2) were excluded because they were tumour free (Fig. 1). 84 patients’ clinicopathological and treatment-related characteristics, including age at diagnosis, performance status, tumour histology, de novo metastatic/recurrent disease status, International Metastatic RCC Database Consortium (IMDC) scores and treatment lines were reviewed from patient files and electronic records. All patients underwent computed tomography scans every 3 months to assess treatment response. Response evaluation was performed according to RECIST (Response Evaluation Criteria in Solid Tumours) version 1.1 [19].

Fig. 1.

Patient flow chart

Immunohistochemical analysis

Tissue microarrays (TMAs) were created using 3 mm punches of tumor tissue from radical or partial nephrectomy specimens. In addition to the TMAs, separate sections were collected from formalin-fixed paraffin-embedded tissues of biopsy or cell-block specimens. Immunohistochemical staining was performed on the Ventana BenchMark XT using the monoclonal B7-H3 (anti-CD276) antibody. The analysis focused on the membranous staining of the tumor cells. The B7-H3 expression was interpreted by an experienced pathologist (K.Y.), who was blinded to patient treatment, survival, and demographic data. A score of 3 + was given for complete and intense membrane staining in at least 10% of tumour cells, a score of 2 + was given for weak to moderate intensity, incomplete membranous staining and/or faint intensity was recorded as 1+, and 0 points were recorded if there was no staining (Fig. 2), as described by Inamura et al. [20]. After 10 patients were evaluated in the scoring, the same group was scored a second time and the agreement between them was evaluated. Since the kappa value was found to be high (> 0.61), general scoring was performed.

Fig. 2.

B7-H3 negative staining (A); B7-H3 + 1 staining (B); B7-H3 + 2 staining (C); B7-H3 + 3 staining (D)

Statistical analysis

Descriptive statistics were utilized to analyze the clinicopathological and treatment-related features of the patients. Categorical variables were presented as percentages, while continuous variables were presented as medians and ranges. The duration from the initiation of nivolumab treatment until disease progression was defined as progression-free survival with nivolumab (PFSnivo), while overall survival (OS) was defined as the duration from treatment initiation to death from any cause. Survival was assessed using the Kaplan-Meier method and compared between the groups using the log-rank test. The median follow-up period was determined using the reverse Kaplan-Meier method. Unadjusted hazard ratios (HRs) for PFSnivo and OS were calculated using Cox proportional hazard regression models. To address potential confounding factors, adjusted HRs were determined using multivariate regression analysis. The statistical analyses were conducted using the SPSS Statistics 25.0 for iOS software program (SPSS, Inc., Chicago, IL, USA), and a significance level of P ≤ 0.05 was applied. Figure 2: B7-H3 negative staining (A); B7-H3 + 1 staining (B); B7-H3 + 2 staining (C); B7-H3 + 3 staining (D).

Results

The study included a group of 84 patients with metastatic kidney cancer, 24 (28%) women and 60 (72%) men, with a median age of 64 ± 10 years. The clinicopathological and treatment-related features are summarised in Table 1. It was observed that B7-H3 expression was not detected in 31% (n = 26) of the patients. Among the samples, B7-H3 expression score was + 1 in 33.3% (n = 28), + 2 in 31% (n = 26), and + 3 in 4.8% (n = 4). While patients in the IMDC poor risk group and non-clear-cell carcinoma were more prevalent among high B7-H3 expressers, no statistically significant difference was found between the groups.

Table 1.

Patients characteristics

	All cohort (n = 84)	CD276 −/+ (n = 54)	CD276 ++/+++ (n = 30)
Age (mean ± SD)	64 ± 10	65 ± 10	63 ± 9
Gender, n(%)
Male  Female	60(71.4) 24(28.6)	38(70.4) 16(29.6)	22(73.3) 8(26.7)
Karnofsky, n(%)
<70%  >70%	25(29.8) 59(70.2)	17(31.5) 38(68.5)	8(26.7) 22(73.3)
Histology, n(%)
Clear cell  Non-clear cell	59(70.2) 25(29.8)	41(75.9) 13(24.1)	18(60) 12(40)
Nephrectomy, n (%)	56(66.7)	38(70.4)	18(60)
Stage at diagnosis, n (%)
Stage I  Stage II  Stage III  Stage IV	3(3.6) 10(11.9) 22(26.2) 49(58.3)	2(3.7) 8(14.8) 13(24.1) 31(57.4)	1(3.3) 2(6.7) 9(30) 18(60)
IMDC, n (%)
Favorable  Intermediate  Poor	8(9.5) 49(58.3) 27(32.1)	7(13) 32(59.3) 15(27.8)	1(3.3) 17(56.7) 12(40)
Firstline treatment, n (%)
No treatment  IFN  Sunitinib  Pazopanib  Cabozantinib	5(6) 5(6) 38(45.2) 34(40.5) 2(2.4)	3(5.6) 4(7.4) 26(48.1) 21(38.9) 0(0)	2(6.7) 1(3.3) 12(40) 13(43.3) 2(6.7)
Line of Immunotherapy, n(%)
Second line  Third line	34(40.5) 30(35) 4 (4.7)	26(48.1) 23 3	8(26.7) 7 1

The median follow-up period was 64 (CI: 33.8–94.1) months, median PFS was 10.1 (CI: 7.9–12.3) and median OS was 40 (CI: 29.9–50) months. There was no statistical difference in PFS between low and high B7-H3 expression groups (12 months vs. 7.9 months, p = 0.461) (Fig. 3). Median OS of patients with B7-H3 low expression was 45 months (CI: 33.7–56. 2), while the mOS of those with B7-H3 high expression was calculated as 20 months (CI: 14.0-25.9) (p = 0.012) (Fig. 4). 3-year and 5-year overall survival rates were 50% and 30%, respectively, in the low B7-H3 group, while 30% and 15% in the high B7-H3 group.

Fig. 3.

PFS according to B7-H3 expression

Fig. 4.

OS according to B7-H3 expression

40.5% (n = 34) of the patients received nivolumab as 2nd or 3rd line treatment. When the variation of immunotherapy responses according to B7-H3 expression level in patients receiving nivolumab was analysed, it was observed that the objective response rate (CR + PR) was higher in patients with low expression (46.2% vs. 25%, p = 0.42). Duration of response was longer in the group with low B7-H3 expression (21.7 vs. 4.9 months, n = 12 vs. 2). PFS with nivolumab (PFSnivo) was median 7 months (CI: 5–9) in the whole group. When we divided the group receiving nivolumab according to B7-H3 expression, PFSnivo was median 8 [5–10] months in the low B7-H3 expression group and 2 months (CI: 0–4) in the high expression group (p = 0.037) (Fig. 5). There was also a statistically significant difference between low and high expression groups in terms of OS (51 months vs. 17 months, p = 0.01) (Fig. 6).

Fig. 5.

PFS with nivolumab treatment

Fig. 6.

OS from nivolumab treatment

Table 2.

Response to Nivolumab therapy in each subgroup with the B7-H3 staining

B7-H3 staining score	Number of response	Number of non response	Response rate
negative	7	6	53.8%
+	5	8	38.5%
++	2	6	25%
+++	0	2	0%

Univariate and multivariate analysis were used to examine the factors influencing PFSnivo. There was no association between PFSnivo and age, gender, IMDC risk score, histological type, presence of de novo metastasis, or nephrectomy (Table 3). It was discovered that the only factor affecting PFSnivo and OS was B7-H3 expression (Table 4).

Table 3.

Univariate and multivariate analysis for PFSnivo

	Univariate analysis		Multivariate analysis
	HR(95%CI)	p- value*	HR(95%CI)	p- value*
Age (< 70 vs. > 70)	1.2(0.56–2.90)	0.55	1.35(0.51–3.60)	0.53
Gender (woman vs. man)	2.44(0.94–6.34)	0.06	0.71(0.20–2.52)	0.60
De-novo vs. relapse	0.90(0.41–1.98)	0.80	0.92(0.33-2.57)	0.88
IMDC risk group (Favorable-intermediate vs. poor)	0.67(0.27–1.6)	0.39	0.34(0.09–1.22)	0.10
Nephrectomy (yes or no)	0.67(0.30–1.48)	0.32	0.44(0.12–1.58)	0.21
Histology (clear cell vs. non clear cell)	1.85(0.79–4.33)	0.15	1.89(0.66–5.41)	0.23
B7-H3 expression (low vs. high)	2.40(0.97–5.93)	0.05	3.95(1.40–11.1)	0.009

*p ≤ 0.05

Table 4.

Univariate and multivariate analysis for OS in Nivolumab receiving group

	Univariate analysis		Multivariate analysis
	HR(95%CI)	p- value*	HR(95%CI)	p- value*
Age (< 70 vs. > 70)	1.14(0.36–3.57)	0.81	0.61(0.11–3.43)	0.58
Gender (woman vs. man)	0.40(0.12–1.27)	0.12	0.65(0.15–2.89)	0.58
De-novo vs. relapse	1.55(0.50–4.83)	0.44	1.73(0.40–7.52)	0.46
IMDC risk group (Favorable-intermediate vs. poor)	1.32(0.41–4.21)	0.63	0.46(0.09–2.22)	0.33
Nephrectomy (yes or no)	0.49(0.18–1.33)	0.16	0.44(0.10–1.88)	0.27
Histology (clear cell vs. non clear cell)	2.37(0.85–6.59)	0.09	3.27(0.84–12.7)	0.08
B7-H3 expression (low vs. high)	6.48(1.90-22.07)	0.003	12.3(2.2–66.3)	0.003

*p ≤ 0.05

Discussion

In the treatment of advanced RCC, the combination of IO-TKI or IO-IO is considered the standard of care. It is important to note that while these treatments have shown impressive results in some cases, not all patients respond equally well in terms of response rate and/or duration of response. Ongoing research is focused on identifying biomarkers that can predict clinical outcomes [21].

B7-H3 plays a dual role in the regulation of adaptive immunity. It inhibits T cell activation and effector functions, while stimulating major histocompatibility complex T cell receptor signalling. This highlights the complex nature of its role in immune modulation [22, 23]. This study is the first to evaluate the potential use of B7-H3 expression as a biomarker for predicting response to immunotherapy in RCC.

The rate of B7-H3 expression in our study was 69%. The expression rate varies according to tumour type and staining used (vascular, membranous, nuclear, cytoplasmic) as reported in the literature. In one study, up to 96% of RCC samples showed membranous staining [24].

Our study population had a similar distribution to the literature, with the clear cell carcinoma subtype being the most common at a rate of 70.2% [25, 26] and a male/female ratio of 2.5:1 [27, 28]. Median overall survival was 40 months, which is slightly shorter than reported in the literature. However, it is important to note that the study group had a high rate of denovo metastatic patients (58.3%) and the patients did not receive up-to-date treatments as they were recruited prior to 2012. A comparative study by Demasure et al. showed that new treatment modalities have led to an improvement in overall survival over the last 20 years [29]. Nevertheless, the PFSnivo (7 vs. 4.6 months) and OSnivo (38 vs. 25.8 months) values observed in our study group were higher than those reported in the Checkmate 025 study [30]. In studies utilising real-world data in which nivolumab was employed in a second-line setting, PFSnivo ranged between 5 and 10 months, while OS ranged between 24 and 44 months [31–33]. This may be primarily as a result of heterogeneity in the proportion of patients with poor prognostic metastatic site involvement, the proportion of patients in the IMDC poor risk group, the proportion of patients receiving nivolumab after the second line, and the proportion of patients with non-clear cell subtype. Furthermore, the evaluation of potential immunotherapy biomarkers, including PDL1, TMB, sarcomatoid differentiation, PBRM1, BAP1, multi-gene signature and others, has not been performed comprehensively [34]. Different rates of immunotherapy resistance-associated biomarker distribution may also cause survival time differences between studies. Therefore, a more appropriate approach would be to evaluate a broader panel of biomarkers in a larger population in combination with current treatment combinations.

Previous studies of B7-H3 in relation to RCC have mainly investigated immunohistochemical expression levels and the relationship with tumour aggressiveness. These studies have consistently shown that B7-H3 expression is associated with metastatic disease, nodal involvement [35–37], poor progression free survival, and cancer-specific mortality [24, 38, 39]. One of the strengths of our study is that we have shown that overall survival decreases with increasing B7-H3 expression, supporting the finding of Mischinger et al. [40].

Studies have shown that B7-H3 plays a critical role in inducing angiogenesis, inhibiting ferroptosis and promoting immune escape in various tumours [41, 42]. It has been observed that treatment response can be improved by reducing immune escape through the combination of B7-H3 blockade and an immune checkpoint inhibitor. Yonesaka et al. demonstrated a negative relationship between B7-H3 expression and patient response to immunotherapy in patients receiving pembrolizumab and nivolumab for NSCLC and showed that the combination of anti-PD1 and anti-B7-H3 resulted in a better response compared to anti-PD1 monotherapy in a mouse model of NSCLC [17]. In their TNBC mouse experiment, Mei et al. also showed that the combination of anti-PD1 and anti-B7-H3 increased the tumour response rate in their TNBC mouse experiment. They also found that the response to immunotherapy decreased as B7-H3 expression increased [43]. Similarly, we found that PFS achieved with nivolumab worsened with increasing B7-H3 expression, although the relationship between B7-H3 expression level and ORR in this group was not statistically significant. Duration of response was numerically longer in the low B7-H3 expression group. We believe that the patient population should be expanded to demonstrate the relationship between ORR and DOR with nivolumab and B7-H3. We did not perform PDL-1 staining in this study. Lee et al. found a positive correlation between PDL-1 and B7-H3 expression in RCC [37]. Considering that B7-H3 is used as an immune escape mechanism, when the negative correlation between B7-H3 expression and OS and PFSnivo and the positive correlation between PDL-1 and B7-H3 level obtained in RCC patients receiving nivolumab are evaluated together, the combined use of antiB7-H3 therapy and nivolumab treatment may provide additional benefit. Aggarwal et al. combined pembrolizumab with anti-B7-H3 monoclonal antibody in head and neck SCC and NSCLC patients and observed an ORR of 33–36% [44].

In our study population, IMDC score was not associated with PFSnivo and OS. Ernst et al. reported a correlation between OS and PFSnivo with IMDC score in patients receiving first-line IO + IO therapy [45]. We believe that the IMDC score had no impact on the outcome of our study as immunotherapy was given as monotherapy in the second and third lines.

Our study has several limitations. The first is the limited sample size with retrospectively assessed oncological outcomes. The small sample size, which limits the statistical power, should be taken into account, especially regarding the role of B7-H3 in the survival analysis. The second is the assessment of response to nivolumab, not the current combination of IO + IO or IO + TKI. The third limitation is that the study did not include biomarkers that may further predict response to immunotherapy, such as sarcomatoid differentiation, PD-L1 expression, TMB level, etc. and metastatic distribution, which may affect survival data.

Conclusions

Expression of B7-H3 in RCC was associated with poor OS and PFSnivo. B7-H3 appears to be a promising predictive biomarker for response to immunotherapy in mRCC. In addition, although there are many phase 1/2 trials involving therapies targeting B7-H3, clinical trials examining the combination of ICI/ADC targeting B7-H3 with anti-PD1/PD-L1 therapy in metastatic RCC should be considered in the near future.

Abbreviations

ADC

Antibody-Drug Conjugate

B7-H3

CD276 Molecule

CNS

Central Nervous System

CR

Complete Response

CTL

Cytotoxic T Lymphocyte

FOXP3

Forkhead Box P3

HR

Hazard Ratio

ICIs

Immune Checkpoint Inhibitors

ICOS

Inducible T-Cell Costimulator

IFN-γ

Interferon Gamma

IMDC

International Metastatic RCC Database Consortium

IO

Immuno-Oncology

LAG3

Lymphocyte Activation Gene 3

mOS

Median Overall Survival

NF-κB

Nuclear Factor Kappa-light-chain-enhancer of Activated B Cells

NSCLC

Non-Small Cell Lung Cancer

ORR

Objective Response Rate

OS

Overall Survival

PD-1

Programmed Cell Death Protein 1

PD-L1

Programmed Death Ligand 1

PFS

Progression-Free Survival

PR

Partial Response

RCC

Renal Cell Carcinoma

RECIST

Response Evaluation Criteria in Solid Tumours

SCC

Squamous Cell Carcinoma

TILs

Tumor-Infiltrating Lymphocytes

TMA

Tissue Microarray

TMB

Tumor Mutational Burden

TKIs

Tyrosine Kinase Inhibitors

TNF-α

Tumor Necrosis Factor Alpha

Tregs

Regulatory T Cells

Author contributions

Faruk Recep Özalp contributed to the conceptualization of the study, data curation, funding acquisition, investigation, methodology, project administration, validation, writing of the original draft, and the review and editing process.Kutsal Yörükoğlu contributed to data curation and formal analysis, and played a supporting role in the investigation, methodology, and project administration. He also contributed to the supervision and supported the writing and editing process.Eda Çalışkan Yıldırım contributed to the conceptualization and formal analysis of the study, supported the investigation and methodology, and contributed to software analysis and visualization.Mehmet Uzun contributed to data curation and resource acquisition and supported the writing and editing process.Hüseyin Salih Semiz contributed to the conceptualization, data curation, investigation, methodology, project administration, supervision, and the writing and editing process.

Funding

This research was supported Turkish Society of Medical Oncology.

Data availability

The data that support the findings of this study are available on request from the corresponding author.

Declarations

Ethics approval and consent to participate

The study was approved by Ethics Board of Dokuz Eylul University (Decision number 8160-GOA). All participants/guardians have provided written informed consent.

Consent for publication

Not Applicable.

Conflict of interest

None.

Competing interests

The authors declare no competing interests.