Predictive genomic biomarkers of therapeutic effects in renal cell carcinoma

Abstract

Background

In recent years, there have been great improvements in the therapy of renal cell carcinoma. Nevertheless, the therapeutic effect varies significantly from person to person. To discern the effective treatment for different populations, predictive molecular biomarkers in response to target, immunological, and combined therapies are widely studied.

Conclusion

This review summarized those studies from three perspectives (SNPs, mutation, and expression level) and listed the relationship between biomarkers and therapeutic effect, highlighting the great potential of predictive molecular biomarkers in metastatic RCC therapy. However, due to a series of reasons, most of these findings require further validation.

Introduction

Renal cell carcinoma (RCC) refers to cancer in the renal epithelium and is more prevalent in men than in women [1]. RCC accounts for more than 90% of cancers in the kidney and is more commonly diagnosed in the elderly [2, 3].

Histology of RCC are complicated. Clear cell RCC (ccRCC) makes up about 70% of all cancers of the kidney [4]. Histologically, they are defined by clear cytoplasm, with nested clusters of cells surrounded by a dense endothelial network. Papillary RCC (pRCC) is the second most common tumor with papillary architecture. The cells are typically foamy histiocytes with a basophilic cytoplasm. Chromophobe RCC cells have an empty cytoplasm with perinuclear clearing and a low mitotic rate [5]. On the contrary, tumors of the nephron and collecting system are rare (less than 5% of cases) [6]. In addition, there are some more rare subtypes based on genetic alternation classified by WHO these years, like Fumarate hydratase deficient renal cell carcinoma(FHRCC), ALK-rearranged RCC and so on [7]. Patients with different pathological subtypes can have distinct prognosis and disparate responses to treatments, therefore, accurate pathological diagnosis is important for patients’ disease management.

In ccRCC, the most common genetic feature is the loss or mutation of the von Hippel-Lindau (VHL) tumor suppressor gene, which is involved in angiogenesis, cell migration, and metabolism [6, 8]. Additionally, other inactivated or mutated genes were also detected in RCC, such as PBRM1, SETD2, BAP1, KDM5a, ARID1a, UTX, MTOR, PTEN, and PIK3CA [9, 10]. Mutations in the MET proto-oncogene and fumarate hydratase (FH) gene increase the risk of papillary RCC [11, 12].

Primary RCC is normally treated with surgical resection to prevent local recurrence in the primary lesion. In the past 12 years, the medical treatment of RCC has gradually evolved from non-specific immune approaches and targeted therapies to the current use of novel immunotherapy agents [2, 5, 13]. These main Therapeutic regimens for mRCC are introduced as followed.

Cytokines

Cytokine-related therapies proved to be effective in treating RCC patients. Over a series of phase II studies, most positive responses were reported with high doses of IL-2 [14], but the selection of these candidates was strict [15, 16]. In two randomized prospective phase III studies, patients who received interferon, compared to placebo, reported a modest survival advantage. However, when compared with newer therapeutic agents(TKI, mTOR inhibitors or immunotherapy), they displayed poorer progression-free survival (PFS) and faster drug resistance [17, 18].

VEGF/VEGFR inhibitor

As mentioned, studies have shown that ccRCC are characterized by loss of chromosome 3p and inactivated mutation of VHL gene, which results in an overexpression of its downstream-regulated genes, VEGF. In 2002, Judae Folkman et al. found that tumors that overexpress VEGF tend to have more blood supply. This revealed a relationship between VEGF and angiogenesis, invasive tumor growth, and metastasis [19], which later lead to the application of VEGF and VEGFR inhibitors in the treatment of tumors. In 2009, FDA first approved Bevacizumb, a VEGF antibody, for first line therapy of metastatic RCC(mRCC). Afterwards, other VEGFR inhibitors, such as Sunitinib, Sorafinib, Axitinib etc. came in use of systemic therapy of mRCC. Although significantly higher response rates and longer PFS and OS were observed in many phase II and III clinical trials [8, 20, 21], there is an obvious heterogeneity in treatment efficacy among patients. Reasons to explain this heterogeneity and predictive biomarkers are frequently researched [9].

mTOR inhibitors

Recent research identified a mutation of mTOR in the PI3K pathway in RCC, indicating the relevance of the PI3K signaling pathway in renal carcinogenesis [10]. Hence, molecular agents targeting mTOR in mRCC are currently being studied, such as temsirolimus and everolimus. In a multicenter phase 3 trial involving 626 patients with poor-prognosis mRCC, patients who received temsirolimus alone exhibit longer overall survival, median overall survival times and progression-free survival than those who received interferon alone [11]. Tesirolimus has been approved by the FDA for the first-line treatment of advanced RCC. In another study conducted by Xinan Sheng et al., for patients those who have received VEGFR-TKI therapy but experienced cancer progression, the addition of everolimus to vorolanib as second-line treatment leads to a better objective response rate, PFS and inhibition of tumor progression [12]. However, there is currently no solution to screen patients who respond to mTOR inhibitor therapy well.

Immune Checkpoint Blockade (ICB)

Overexpression of immune checkpoint genes (CTLA4, PD-1, PD-L1 etc.) is an important mechanism for tumor to escape anti-tumor immunity. Monoclonal antibodies, such as Ipilimumab and Nivolumab, target immune checkpoint genes, including CTLA-4 (cytotoxic T lymphocyte antigen 4) and PD-1 (programmed cell death 1) to reactivate immune response [6]. The approval of ICB in RCC stands for another revolution in mRCC treatment. Present first-line therapies based on immune checkpoint inhibitors include dual-targeted therapy (ipilimumab + nivolumab) and combinated therapy (nivolumab + cabozantinib, pembrolizumab + lenvatinib, pembrolizumab + axitinib, etc.). Immune checkpoint inhibitors once again significantly improve the response rate and prognosis of mRCC patients, yet great heterogeneity of response among patients also exists with no response to ICB drugs in part of patients, which possibly means that diverse immune microenvironments can influence the efficacy of immunotherapy.

Others

Other agents, such as TKI Sitravatinib, HIF-2αinhibitor, MK-6482 are promising and currently under development [6, 22, 23]. How to choose optimal populations for these drugs are yet to be tested in clinical trials.

Despite the benefits of newly developed therapies and the improved five-year survival rate at diagnosis, treatment efficacy varies greatly among patients. Many patients have little to no therapeutic response, while others respond selectively to specific therapies. Likewise, all patients eventually develop various degrees of resistance to different kinds of drugs, whose mechanisms remains unclear. The FDA has approved several biomarkers for precise treatment of tumors, such as anti-PD-1 antibodies, pembrolizumab and nivolumab for head and neck squamous cell carcinoma (HNSCC) [24], anti-CTLA4 therapy for advanced melanoma [25] and Pemigatinib, a FGFR1-3 inhibitor for biliary tract cancers etc. [26]. This means the identification of genomic biomarkers can also play an important role in optimizing therapeutic options for different groups of mRCC patients, and allows doctors to suggest the optimal treatment strategy, thereby reducing the toxicity and expenses of inappropriate therapies for patients.

This review aims to summarize current evidence of genetic biomarkers to predict the prognosis of patients treated with immune, targeted, and combined therapies from three different perspectives, SNPs, mutation, and expression level. Despite most clinical trials and researches are restricted to ccRCC, there are also unique biomarkers for other RCC pathological subtypes. Biomarkers introduced in the review are first summarized according to pathological subtypes of RCC involved, including ccRCC, and nccRCC(pRCC, Chromophobe RCC and MiT family translocation RCC). Lastly, we also concluded biomarkers for overall RCC regardless of histology classification。

ccRCC

Single nucleotide polymorphism

Single nucleotide polymorphism (SNP) refers to an alteration in the DNA at a single base position and is the most common type of genetic variation in the population (i.e., five million SNPs in a human genome). Although most SNPs do not affect health and development, some of them do contribute to complex phenotypes by regulating function of genes, usually when they locate within or near a gene [27]. Hence, SNPs can be regarded as biomarkers, and the association between SNPs and the prognosis of RCC is described below.

IL-8

In 1992, Koch et al. established IL8 to be pro-angiogenic and IL-8 protein involves in mitogenic and angiogenic process [28]. Later, Huang et al. reported the elevation of IL-8 expression to be an important compensatory mechanism in sunitinib-resistant patients, which was possibly due to the anti-angiogenic condition [29]. One step further, Hacking et al. discovered the association between the IL8 polymorphism rs1126647 and an increased expression of the IL8 gene [30], which is related to disease progression, tumour stage and cancer recurrence [31], comprising worse prognosis in RCC [32]. Accordingly, in an observational study conducted by Xu et al., a significant correlation between the expression of variant alleles of IL8 polymorphism rs1126647 and poorer OS was found (N = 88; 50 events; P = 0.034; HR = 1.62; 95% CI: 1.04–2.54) [33].

ITGAV

Similar to IL-8, ITGAV encodes the αV subunit of integrin, heterodimerizing with multiple β subunits and the integrin αvβ3 is often upregulated in tumor and endothelial cells as well, indicated to regulate angiogenesis [34, 35]. The ITGAV polymorphism rs3816375 AA/AG enhanced angiogenesis and reduced the efficacy of sorafenib. In combined sorafenib and placebo arms, ITGAV rs3816375 AA/AG patients exhibited shorter OS as compared with ITGAV rs3816375 GG patients at 387 and 371 days, respectively (HR = 5.9; 95% CI: 2.1–16.4; P = 4.9 × 10⁻⁴) [36]. In the same article, WWOX rs8047917 (T > A) was also revealed to be associated with shorter OS (HR = 4.1; 95% CI: 1.9–8.8; P = 3.3 × 10^–4). Through the increased luciferase assay, the rs8047917 (T > A) polymorphism was attributed to the tumor suppressor properties of WWOX [36, 37].

VEGFR

Another important biomarker vascular endothelial growth factor (VEGF), plays a key role in physiologic and pathologic angiogenesis and lymphangiogenesis [38, 39], which has been proven to be a target in developing anticancer therapeutics. For example, VEGF-A acts as a key driver of the growth of neovessel needed to support solid tumor progression [40]. VEGF receptor VEGFR2, the primary signaling receptor for VEGF-A, once activated by VEGF-A on endothelial tip and stalk cells, can direct the migration and extension of sprouting vessels [41]. While VEGFR3, encoded by FLT4 [40], is the receptor for VEGF-C and VEGF-D [42] whose expression is greatly constricted to lymphatic endothelium [38]. It also has been recently found to express on the endothelial tip cells of angiogenic tumor vasculature, while blood vessel endothelial cell will sprout and vessel will branch once the receptors in tumor models are blocked [43]. In view of this point, drugs inhibiting VEGF signaling pathway are actively developed and registered for therapy including bevacizumab [44], sorafenib, sunitinib, Cediranib, pazopanib etc. [45]. Based on the findings of an observational, prospective study, previously untreated ccRCC adults (≥ 18 years) who were given sunitinib reported an association between two VEGFR3 missense polymorphisms and reduced PFS: rs307826 (HR per allele = 3.57, 1.75–7.30; p (unadjusted) = 0.00049, p (adjusted) = 0.0079) and rs307821 (HR per allele = 3.31, 1.64–6.68; p (unadjusted) = 0.00085, p (adjusted) = 0.014) [46]. Interestingly, Jit et al. revealed that in primary ccRCC patients treated with sunitinib, polymorphism rs307826 which down regulates expression of VEGFR3 was associated with more drug resistance (P = 0.002) and shorter progression times (HR = 0.40 ( 0.20–0.82)) [47]. The rs307826 (A > G) in FLT4 was revealed to substitute threonine with alanine in VEGFR-3, which increased VEGFR-3 membrane trafficking and VEGFR-3 phosphorylation [48]. Consequently, the genetic variation of FLT4 (A > G) is associated with shorter PFS (HR = 2.92, 95% CI: 1.19–7.19; P = 0.019) and OS (HR = 13.8; 95% CI: 3.0–62.6; P = 1.2 × 10^–4) [36].

Other subtypes in the same family are also studied and summrized.

In the S-TRAC trial conducted by George et al., 286 high-risk RCC patients were treated with adjuvant sunitinib and genotyped. Patients with VEGFR1 rs9582036 C/A and C/C reported poorer disease-free survival (DFS) in sunitinib, placebo, and combined therapies (P ≤ 0.05) than VEGFR1 rs9582036 A/A and C/C with sunitinib (P = 0.022) [49]. However, in another independent cohort of mRCC patients, Kaplan–Meier analysis revealed that AA/AC-alleles of the VEGFR1 SNP rs9582036 led to better median overall survival (OS) in comparison with CC-WT allele in mRCC patients with sunitinib, which was also validated by multivariable Cox proportional hazard regression analyses [50]. The conflicting results might be due to the low sample size. Hence, further validation in a larger sample cohort is required to confirm its predictive value. The same study also indicated that the WT allele of the VEGFA SNP rs699947 displayed poorer PFS as compared to other variant alleles when treated with sunitinib [50]. Gene VEGFA encodes VEGF-A, the most potent pro-angiogenic ligand [51]. Crona et al. genotyped germline variants from 56 genes and identified that poorer OS was related to VEGFA rs1885657 (T > C, HR = 17.3; 95% CI: 5.7–52.7; P = 1.4 × 10^–4), rs6719561 (3’ of UGT1A9, C > T, HR = 3.8; 95% CI: 1.6–9.5; P = 3.3 × 10^–4), rs200809375 (3’ of NRP1, ATG insertion, HR = 6.8; 95% CI: 2.6–17.5; P = 2.7 × 10^–4), and rs3024987 (C > T, HR = 3.0; 95% CI: 1.7–5.4; P = 8.3 × 10⁻⁵). Similarly, shorter PFS is associated with genetic variations of VEGFA rs1885657 (HR = 4.00, 95% CI: 1.57–10.21; P = 0.004) and rs6719561 (3’ of UGT1A9, HR = 1.94, CI: 1.00–3.76; P = 0.05) [36].

ABCB1

Next comes to the gene ABCB1, who acts as an efflux transporter for drug absorption, excretion, cellular accumulation, and resistance [52] while polymorphisms in ABCB1 correlate to the absorption and excretion of sunitinib via the association of the TCG copy in the ABCB1 haplotype. As a result, these polymorphisms could lead to increased systemic exposure to sunitinib by reducing the efflux transport of sunitinib into the gastrointestinal lumen and bile [53]. Accordingly, Diekstra et al. observed that patients with heterozygote genotypes and some wild-types (rs1128503 and rs2032582) in ABCB1 exhibited reduced clearance of sunitinib and its active metabolite, SU12662 [54].,Also, the CGT copy present in the ABCB1 haplotypes rs2032582 and rs1128503 was linked to the increase in PFS (HR = 1.9; 95% CI: 1.3–2.6; p < 0.001) [55]. Similarly, Veldt et al. found that patients with clear-cell mRCC that were treated with sunitinib experienced improved PFS and OS. In addition, patients being carriers of a TCG copy present in the ABCB1 haplotype (3435C/T, 1236C/T, 2677G/T; HR = 0.52; P = 0.033) show improved conditions (median PFS = 19.9 months, median OS = 12.3 months) as compared to non-carriers (median PFS = 13.1, median OS = 7.5 months) [53].

CYP3A5

Another gene discussed is CYP3A5, which encodes enzymes that metabolize several tyrosine kinase inhibitors in oxidation, such as gefitinib, erlotinib, imatinib, and sunitinib [56]. Multivariate analysis by Veldt et al. revealed that patients with clear-cell mRCC that were treated with sunitinib and carriers of A-allele in CYP3A5 6986A/G reported better PFS. Furthermore, in TKI metabolism, polymorphisms in NR1I2 and NR1I3 can regulate the enzymatic expression of CYP3A4 [57]. Veldt et al. revealed that the presence of the C-allele in NR1I2 8055C/T, the C-allele in NR1I2 -25385C/T, and the absence of a CAT copy in the NR1I3 haplotype resulted in longer PFS. In addition, prolonged OS is related to the presence of the C-allele in NR1I2 -25385C/T [53].

The relationship between SNPs and the prognosis of patients is summarized in Table 1.

Table 1.

Relationship between SNPs and the prognosis of patients

Author (Year)	Tumor type	Biomarker(s)	SNPs	Therapy	N pts (total)	OS			PFS			Response
						mOS (mo)	HR (95% CI)	P value	mPFS (mo)	HR (95% CI)	P value	ORR (95% CI)	P value
Xu et al. (2015) [33]	ccRCC	IL8	rs1126647 2767A > T	sunitinib	88	-	1.62 (1.04–2.54)	0.034	-	-	-	-	-
Crona et al. (2019) [36]	ccRCC	ITGAV rs3816375	A > G	sorafenib	295	-	5.90 (2.10–16.40)	0.001	-	-	-	-	-
		VEGFA rs1885657	T > C				17.30 (5.70–52.70)	0.0001	-	4.00 (1.57–10.21)	0.004	-	-
		VEGFA rs6719561	C > T				3.80 (1.60–9.50)	0.0003	-	1.94 (1.00–3.76)	0.050	-	-
		VEGFA rs200809375	ATG insertion				6.80 (2.60–17.50)	0.0003	-	-	-	-	-
		FLT4 rs307826 (A > G,	A > G				13.80 (3.00–62.60)	0.0001	-	2.92 (1.19–7.19)	0.019	-	-
		VEGFA rs3024987	C > T				3.00 (1.70–5.40)	0.0001	-	-	-	-	-
Garcia-Donas et al. (2011) [46]	ccRCC	VEGFR3	rs307826 A > G	sunitinib	89	-	1.17 (0.65–4.84)	0.260	-	3.57 (1.75–7.30)	0.008	8.79 (1.92–40.28)	0.082
			rs307821 G > T			-	1.24 (0.41–3.75)	0.710	-	3.31 (1.64–6.68)	0.014	7.14 (1.46–35.00)	0.250
Garcia-Donas et al. [47]	ccRCC	VEGFR3 rs307826	IHC scoring high	sunitinib	64	-	4.29 (1.43–12.8)	0.009	15.3	0.40 (0.20–0.82)	0.012	-	-
		HIF2A	IHC scoring low		60	18.4	0.39 (0.15–0.99)	0.048	-	0.11 (0.017–0.75)	0.024	0.11 (0.02– 0.75)	0.024
		PDGFRB	IHC scoring high		62	-	-	-	-	-	-	0.04 (0.002–0.68)	0.026
		VEGFA	IHC scoring high		62	9.5	4.29 (1.43–12.8)	0.009	-	2.56 (0.99–6.63)	0.053	-	-
Dornbusch et al. (2016) [50]	ccRCC	VEGFA	rs699947 A > C	sunitinib	117	-	0.63 (0.26–1.53)	0.304	16.0	1.03 (0.50–2.14)	0.939	-	-
		VEGFR1	rs9582036 C > A		116	42.0	0.29 (0.09–0.94)	0.039	-	0.55 (0.20–1.53)	0.253	-	-
Diekstra et al. (2015) [55]	ccRCC	ABCB1	CGT	sunitinib	333	-	-	-	16.0	1.90 (1.30–2.60)	< 0.001	-	-
van der Veldt et al. (2011) [53]	ccRCC	ABCB1	3435C/T, 1236C/T, 2677G/T	sunitinib	136	12.3	0.59 (0.33–1.06)	0.078	19.9	0.52 (0.29–0.95)	0.072	-	-

Abbreviation: N pts Number of patients; ccRCC Clear-cell renal cell carcinoma; Mo Months; mOS Median overall survival; mPFS Median progression-free survival; OR Odds ratio; HR Hazard ratio; CI Confidence interval; ORR Overall response rate

Mutation

Mutation refers to changes in the DNA sequence of an organism, which can be caused by infection, problematic cell division, toxins, and mutagens. Uncontrolled mutation can lead to abnormal cellular functions, diseases, and diverse responses to therapy [58]. The influence of mutation on RCC therapy is discussed as follows.

VHL

The most common mutation in ccRCC is located on von Hippel–Lindau(VHL) gene, which can be found in nearly 70% of ccRCC cases and causes abnormal vascularization by regulating the hypoxia pathway via the hypoxia-inducible factors (HIFs) [59–61]. Drugs targeted agents modulating the VHL–HIF pathway like sunitinib, sorafenib and temsirolimus have shown superiority to previous cytokine therapies [11, 62, 63]. Also, VHL mutation relates to therapeutic effect. For example, Gossage et al. have found that the loss-of-function VHL mutation (nonsense, frameshift, deletion, insertion, etc.) led to a 52% response rate as compared to a 31% response rate for wild-type VHL (p = 0.04) [64].In another study, 35 patients with VHL mutations experience PFS of 18 months compared with 10 months for the nonmutant groups. Thus, VHL mutations are identified as prognostic biomarkers of a longer PFS (p = 0.017, 0.04, 0.014) [65].

PBRM1

The PBRM1 gene is the second most commonly mutated gene after VHL [66], and, in ccRCC, acts as the most frequently mutated chromatin-modifying gene. Bromodomains that it contains interact with lysine-acetylated histones, regulating gene expression [67], so its recurrent mutation can be found in novel tumor suppressors [68]. Consequently, Fay et al. identified the PBRM1 mutations led to better responses towards VEGF-targeted therapy in mRCC [69]. Likewise, Voss et al. revealed that mutated PBRM1 resulted in better prognostic of OS, PFS, and the overall response [70]. Patients with mutated PBRM1 exhibited a lower risk of progression from first-line everolimus therapy (HR = 0.53 (0.3–0.8)) [71]. Likewise, Carloa et al. observed that the time to treatment failure was shorter for those with PBRM1 mutation than wild-type controls [72]. In the CheckMate 025 trial, WTS data from 177 tumors were obtained for further analysis. The low proportion of PBRM1 mutation in ccRCC might be a predictor for poor response to anti-PD-1 therapy and poor survival. In contrast, increased PBRM1 mutation and GATM (glycine amidinotransferase) expression led to better response and prolonged survival [73]. In one more randomized phase III clinical trial, 823 aRCC patients, receiving either combined atezolizumab and bevacizumab or sunitinib, were analyzed to correlate between gene mutation and treatment efficacy. Mutated PBRM1 resulted in more angiogenesis and higher expressions of AMPK and mutated KDM5C. Patients without PBRM1 mutation reported improved PFS and ORR when treated with a combination of atezolizumab and bevacizumab as compared to sunitinib only (HR = 0.74, 95% CI: 0.58–0.94, mPFS atezolizumab + bevacizumab: 9.9 months; mPFS sunitinib: 6.9 months) [74].

BAP1

The BAP1 protein is a nuclear deubiquitinase that binds to HCF-1 in tumor grafts. Both the BAP1 and VHL genes are located on chromosome 3p, and a mutation may suppress cell proliferation, thereby sensitizing RCC cells to genotoxic stress [67]. In the COMPARZ study, 57 aRCC and mRCC patients were treated with first-line TKIs and evaluated for genomics. It was revealed that mutations of BAP1 correlated with poor OS but no correlation with PFS [70]. For metastatic ccRCC patients treated with everolimus, which binds to the FK506 binding protein-12 (FKBP-12) with high affinity, BAP1 mutation was associated with shorter PFS (4.9 months) and a higher risk of progression when treated by first-line everolimus or sunitinib [71]. In another retrospective review of 105 metastatic ccRCC patients treated with VEGF-targeted therapy, the time to treatment failure was shorter for those with mutated BAP1 [72].

KDM5C

Besides, the KDM5C gene is located on the X chromosome and is regarded as an oncogene [75]. In the study by Hsieh et al., patients treated with sunitinib with mutated KDM5C exhibited a lower risk of progression, while no difference was found in patients treated with everolimus [71].

TERT

Similarly, telomerase reverse transcriptase (TERT) acts as a catalytic subunit of telomerase that facilitates cancer formation, while somatic mutations of the TERT gene can conserve the telomere complex [76]. Carloa et al. observed that shorter OS was associated with TERT mutations (median OS: 29.6 months for mutated-type vs 52.6 months for wild-type) [72]. Likewise, Due to the regulation of P53 to produce senescence-associated secretory phenotype (SASP) under physiological circumstances [77], mutated P53 in tumor cells may lead to longer uncontrolled SASP response, which can lead to IL-6 and IL-8 secretions and increased angiogenesis, resulting in tumor progression [78–80]. In the research conducted by Voss et al., patients with mutations in TP53 have reduced OS than those with wild-type TP53 (13.3 months vs 30.6 months) [70].

ARID1A and KTM2C

Studies were also conducted to evaluate the gene expression signatures associated with angiogenesis and effector T cell response to predict outcomes. Tripathi et al. found that for those patients who were treated with sunitinib, the expression of proangiogenic gene signature led to a better response rate and prolonged PFS. In immunotherapy, patients with high effector T cell response and anti-angiogenic gene signature tend to have significantly longer PFS with combined atezolizumab and bevacizumab (HR = 0.55; 95% CI: 0.32–0.95) [81].

Another two biomarkers involved are ARID1A and KTM2C. The former comprises the chromatin remodeling SWI/SNF complex, while the latter acts as a histone methyl transferase that participates in DDR deficiency and epigenetic dysregulation [82, 83]. Consequently, patients exhibiting loss-of-function mutations of ARID1A and/or KMT2C were revealed to have better PFS when treated with combined atezolizumab and bevacizumab than sunitinib only (ARID1A: 20.7 vs 6.8 months; KMT2C: 13.8 months vs. 7.0 months) [74].

DDR

Last genetic biomarker of interest is in the DNA damage and repair (DDR) pathway. DDR includes 7 major pathways, comprising homologous recombination (HR), mismatch repair (MMR), base excision repair (BER), nucleotide excision repair, nonhomologous end-joining (NHEJ), checkpoint factors (CPF) and Fanconi anemia genes, participating in cellular processes that sense DNA damage and initiate signaling pathways, which leads to apoptosis, activation of cell-cycle checkpoints, and coordination of DNA repair [84].. In a retrospective study by Zhou et al., patients with metastatic ccRCC, DDR mutations reported longer PFS when treated with VEGF-TKI systemic therapy, including sunitinib, pazopanib, axitinib, and sorafenib [65]. In another study with 225 metastatic ccRCC patients, 37 (16%) of them exhibited loss-of-function in DDR genes and reported prolonged OS when treated with immune-oncology therapeutics (p = 0.04) [85].

Additional details on the association of mutation status with therapeutic effect are described in Table 2.

Table 2.

The association of mutation status with therapeutic effect

Author (Year)	Tumor type	Biomarker(s)	Therapy	N pts (total)	OS			PFS
					mOS (mo)	HR (95% CI)	P value	mPFS (mo)	HR (95% CI)	P value
Zhou et al. (2022) [65]	ccRCC	DDR	VEGF-TKI systemic therapy (sunitinib, pazopanib, axitinib, sorafenib)	7	-	-	-	18.0	-	0.017
		VHL		10	-	-	-	14.0	-	0.040
Hsieh et al. (2016) [71]	ccRCC	PBRM1	everolimus or sunitinib	109	-	-	-	12.8	0.53 (0.3–0.8)	0.002
		BAP1			-	-	-	4.9	1.84 (1.1–3.2)	0.020
Carloa et al. (2017) [72]	ccRCC	TERT	everolimus or sunitinib	105	29.6	(18.40-NA)	0.030	-	-	-
		BAP1			28.7	(23.90-NA)	0.020	-	-	-

Abbreviation: N pts Number of patients; ccRCC Clear-cell renal cell carcinoma; Mo Months; mOS Median overall survival; mPFS Median progression-free survival; HR Hazard ratio; CI Confidence interval

Expression level

Gene transcription and translation are elaborate processes that result in intercellular variations via mRNA and proteins and these variations play important roles in regulating cell functions [86].

PD-1, PD-L1

Aberrant cells, such as cancer cells are detected and destroyed by T cell-based immune system through peptide-major histocompatibility complexes on target cells, which is controlled by co-stimulatory and co-inhibitory receptors and their ligands (refer to immune checkpoints) to a great extent [87].

Programmed cell death protein 1 (PD-1; also called CD279) belongs to co-inhibitory receptors [88]., interacting with PD-L1 (CD274) which expresses on the surface of hematopoietic cells such as T cells, B cells, macrophages and non-hematopoietic healthy cells. They can be detected on tumor cells and tumor stroma [89]. PD-1 and PD-L1 are both transmembrane proteins which belong to the immunoglobulin (Ig) superfamily with extracellular domain, transmembrane domain, and a cytoplasmic domain. A conformational change in PD-1 can be induced by interactions of PD-L1 and PD-1 between their extracellular domains, which leads to the phosphorylated immunoreceptor tyrosine-based switch motif (ITSM) and the cytoplasmic immunoreceptor tyrosine-based inhibitory motif (ITIM) [90–92]. Subsequently, T cell-activating signals are attenuated through recruiting the SHP-2 and SHP-1 protein tyrosine phosphatases and then, activity of T cells are inhibited [93–99], losing functions progressively and turning into an exhausted state [100] in both cancer [101] and chronic infection [102]. Under these circumstances, pro-survival signals to tumor cells are delivered, leading to resistance in apoptosis induced by Fas or staurosporine [103]. Additionally, it can protect tumor cells from cytotoxic T lymphocyte (CTL)-mediated cytolysis and the cytotoxic effects of type I and type II interferons [103, 104]. Although expression of PD-L1/PD-1 are reported to be related to immunological therapy in multiple solid tumors, such as non small-cell lung cancer or melanoma [105]. The relationship between these two genes and response to PD-1/PD-L1 inhibitor in renal cell carcinoma are controversial. Carretero-Gonzalez et al. evaluated the relationship between PD-L1 expression and the therapeutic effect of ICIs. 4635 metastatic RCC patients in six studies were involved (2367 patients with ICIs and 2268 patients with standard care), and a meta-analysis from five randomized clinical trials was used. Longer PFS was associated with higher expression of PD-L1 (p < 0.0001), but such association was not observed in OS (p = 0.63) [106]. However, PD-L1 expression status in cancer cells was not associated with prognosis(PFS, HR:1.00, 95% CI: 0.80–1.26) in mRCC patients treated with nivolumab + ipilimumab in CheckMate214. Meanwhile, objective response rate was found to be higher in PD-L1 positive (if PD-L1 expression was detected in more than 1% of cancer cells) patients than negative ones, the ORR in PD-L1 positive group was 58% vs 22% in immunotherapy arm and sunitinib arm respectively (p < 0.001) and corresponding ORR was 37% vs 28% in PD-L1 negative group (p = 0.03) [107]. Predictive effect of PD-1/PD-L1 expression in combination therapy of VEGFR TKI and ICB is also unconfirmed. In a phase III clinical trial JAVELIN Renal 101, no association between PD-L1 status and PFS was observed in avelumab + axitinib, suprisingly higher PD-L1 expression correlated with worse prognosis in terms of PFS in sunitinib arm (HR: 1.57, 95% CI: 1.15–2.14) [108]. The controversial relationship of PD-1/PD-L1 concluded by different research might owes to inconsistent criteria of defining PD-1/PD-L1 status (ie. some define 1% PD-1/PD-L1 expression as positive, while some set 5%) and heterogeneity between primary and metastatic lesions [109].

Genes contributed to drug resistence

Despite the progression of different therapies, some patients might develop resistance to treatment options. For example, 70% of patients treated with VEGF-TKIs eventually developed resistance [110]. The gene RCAN1.4 is located in the region of 21q22.12 on chromosome 21 and functions to regulate HIF2α expression and epithelial-mesenchymal transition. In the findings, cells that were resistant to sunitinib had significantly lower levels of RCAN1.4 mRNA and proteins than non-resistant cells [111].

EZH2, the histone methyl transferase, is frequently altered in various cancers including ccRCC, whose expression was also found to be negatively associated with OS of patients and positively correlated with sunitinib resistance in ccRCC patients. Patients with low EZH2 expression reported better OS as well [112]. From the studies, the phosphorylation of serine and tyrosine residues correlated with upregulated EZH2 expression in resistant samples, implying that EZH2 could potentially interact with kinases for stable expression and activity. Moreover, increased EZH2 expression changes gene expressions that are related to resistance, such as PTPN3 and DAB2IP. In resistant samples, the binding of EZH2 to the PTK2 gene, encoding Focal Adhesion Kinase (FAK), was found to decrease. FAK was reported to interact with EZH to promote tumor progression [113, 114].

Another related gene is ecotropic viral integration site 1 (EVI1), who encodes for a dual-domain zinc-finger transcription factor, which regulates the differentiation of myeloid progenitor cells and the renewal of hematopoietic stem cells [115]. Their interaction with polycomb-group proteins reduces the expression of PTEN and DNA methyltransferases, leading to a hypermethylated genomic signature [116, 117]. Palomero et al. revealed that high expression of the EVI1 isoform in RCC led to increased resistance to everolimus. The overexpression of the EVI1 rs1344555 genetic variant significantly correlated with PFS and OS in everolimus-treated metastatic ccRCC patients as well [118].

In the last study with aRCC patients, the expression of UDP-glucuronosyltransferase (UGT1A6) and its family members (i.e., UGT1A1 and UGT1A3) were found to be significantly upregulated and over-expressed in non-respondents to anti–PD-1 therapy. The UGT1A6 gene regulates the clearance of exogenous lipophilic chemicals [119], which were revealed to be in abundance in patients who were resistant to anti–PD-1 therapy. Likewise, the other members of the UGT1A family were also upregulated [120]. The expression of the contactin-associated protein 1 (CNTNAP1) gene is associated with regulatory T cells, which regulate activated macrophage (M2), myeloid-derived suppressor cells, and the infiltration levels of cancer-associated fibroblast. Li et al. revealed that higher expressions of CNTNAP1 reduced the sensitivity of nilotinib (COR = 0.381905509) and crizotinib (COR = 0.353582241) in ccRCC patients [121].

C-Myc

Metastasis can occur both before and after surgical treatment, reducing the survival time and prognosis of patients. Over the past two decades, various mRCC-targeted drugs have been developed [122], and biomarkers are extensively studied for their predictive potential for therapies. For example, the overexpression of the regulator gene C-Myc empowers cells with uncontrolled proliferation and division [123]. In a prospective observational study, 80 mRCC patients were treated with sunitinib. Those who were c-Myc-positive reported worse PFS as compared with those who were c-Myc-negative, with a median PFS of 5.4 months and 11.4 months, respectively [124].

DDX39

Last biomarker discussed is DDX39 gene. CCK-8 and colony formation experiments revealed that the increased expression of DDX39 subsequently enhanced cell proliferation, migration, and invasion. ssGSEA analysis indicated that the expression of DXX39 is associated with immune cell infiltration, antigen function, and inhibitory or co-inhibitory immune genes. This demonstrated that poor immune function in ccRCC patients could be inferred by high DDX39 expression. Accordingly, Bao et al. revealed that increased expression of DDX39 correlated with the adverse response of immune checkpoint therapy (28.7% vs. 39.5%, p = 0.011) [125].

The expression levels of different biomarkers and their effect on the prognosis of patients are listed in Table 3.

Table 3.

The expression level of biomarkers and their effects on patient prognosis

Author (Year)	Tumor type	Biomarker(s)	Expression level	therapy	N pts (total)	OS
						mOS (mo)	HR (95% CI)	P value
Adelaiye-Ogala et al. (2017) [112]	ccRCC	EZH2	H-score > median	sunitinib	33	23	5.60 (1.70- 8.70)	0.005
Palomero et al. (2020) [118]	ccRCC	EVI1 rs1344555(CC vs CT/TT)	overexpression	everolimus	39	-	2.09 (1.08–4.08)	0.029
Garcia-Donas et al. (2013) [51]	ccRCC	VEGFR3 rs307826	IHC scoring high	sunitinib	64	-	4.29 (1.43–12.8)	0.009
		HIF2A	IHC scoring low		60	18.4	0.39 (0.15–0.99)	0.048
		PDGFRB	IHC scoring high		62	-	-	-
		VEGFA	IHC scoring high		62	9.5	4.29 (1.43–12.8)	0.009
Li et al. (2022) [121]	ccRCC	CNTNAP1	high expression	nilotinib	533	-	-	-
				crizotinib		-	-	-

Author (Year)	PFS			Response		Correlation
	mPFS(mo)	HR (95% CI)	P value	ORR (95% CI)	P value	COR	P value
Adelaiye-Ogala et al. (2017) [112]	-	-	-	-	-	-	-
Palomero et al. (2020) [118]	-	1.96 (1.01–3.81)	0.047	-	-	-	-
Garcia-Donas et al. (2013) [51]	15.3	0.40 (0.20–0.82)	0.012	-	-	-	-
	-	0.11 (0.017–0.75)	0.024	0.11 (0.02–0.75)	0.024	-	-
	-	-	-	0.04 (0.002–0.68)	0.026	-	-
	-	2.56 (0.99–6.63)	0.053	-	-	-	-
Li et al. (2022) [121]	-	-	-	-	-	-0.382	0.003
	-	-	-	-	-	-0.354	0.003

Abbreviation: N pts Number of patients; ccRCC Clear-cell renal cell carcinoma; Mo Months; mPFS Median progression-free survival; HR Hazard ratio; CI Confidence interval; COR Correlation between drug and gene marker; ORR Overall response rate

nccRCC

Other researches involving nccRCC(pRCC, Chromophobe RCC and MiT family translocation RCC) have been collected and summarized..Additional details of genomic biomarkers and therapeutic effect are described in Table 4.

Table 4.

Relationship between genomic biomarkers and therapeutic effect in other types of RCC

Author (Year)	Biomarker type	Biomarker(s)	SNPs/expression level	Therapy	N pts (total)	OS			PFS			Response
						mOS (mo)	HR (95% CI)	P value	mPFS (mo)	HR (95% CI)	P value	ORR (95% CI)	P value
Xu et al. (2011) [126]	SNP	2767TT	2767TT	pazopanib	-	-	-	-	27.0	1.80 (1.20–2.70)	0.009	-	-
		251AA	251AA		-	-	-	-	32.0	1.70 (1.10–2.50)	0.010	-	-
Motzer et al. (2020) [74]	Mutation	ARID1A	-	atezolizumab + bevacizumab	823	20.7	0.62 (0.36–1.07)	0.083	-	-	-	-	-
				sunitinib		6.8	1.34 (0.89–2.02)	0.160	-	-
Qu et al. (2019) [127]	Expression level	PAK1	IRS score < 9.9	sunitinib or sorafenib	111	-	3.30 (2.58–10.89)	< 0.001	-	-	-	-	-
Merhautova et la. (2015) [128]	Expression leve	miR-155	high ≥ 0.2381	sunitinib	63	-	-	-	-	-	-	2.41 (1.24–4.68)	-
		miR-484	high ≥ 1.4408			-	-	-	-	-	-	2.62 (1.10–6.25)	-

Abbreviation: N pts Number of patients; ccRCC Clear-cell renal cell carcinoma; Mo Months; mOS Median overall survival; mPFS Median progression-free survival; OR Odds ratio; HR Hazard ratio; CI Confidence interval; ORR Overall response rate

PRCC

PTTG1

Papillary RCC is the second most common type of RCC. A high expression of PTTG1 is associated with the proliferation of T cells, which reinforces anti-tumor immunity. In the FUSCC cohort, a better response to immune checkpoint blockade (ICB) correlated with a higher expression of PTTG1 (p < 0.05), which is commonly implicated in lymphocytes, immunomodulators, and chemokine [129].

C-MET

The c-Met receptor tyrosine kinase, encoded by the MET proto-oncogene, is located at the cell surface for hepatocyte growth factor (HGF). The abnormal activation of c-Met/HGF axis existing in various cancers through MET transcriptional upregulation, genomic amplification and ligand-dependent mechanisms will lead to tumor angiogenesis, metastasis, progression, invasive growth and resistance to therapies [130]. C-MET is commonly expressed in papillary RCC, so accordingly, in one study conducted by Susan Halabi et al., they reveal that expression of C-MET is related to lower rates of radiographic response in sunitinib and everolimus in papillary RCC patients [131]

Chromophobe RCC

Chromophobe kidney cancer (ChRCC), as a subtype, belongs to RCC, arising from the nephron [132]. Despite the indolent pattern of local growth and more than 90% ten-year cancer-specific survival it often exhibits, aggressive features and metastasis still exist [133, 134].

C-kit

Receptor tyrosine kinase c-kit, is a defining tissue biomarker of mTOR pathway activation through folliculin mutations. Once activated, it starts autophosphorylation, initiates signal transduction. After recruiting signaling proteins by specific interaction domains (e.g., PTB and SH2) they will bind to phosphorylated tyrosine residues located in the intracellular region of c-Kit. Activated c-Kit signaling is indicated to regulate cell migration, proliferation and survival while deregulated c-Kit activity is found in various pathological conditions, including cancer [135]. In the same study by Susan Halabi et al. mentioned above, they found that expresssion of c-kit in chromophobe RCC is related to improved therapeutic outcomes with both everolimus and sunitinib. While, high level expression of c-kit is found to be associated with improved rPFS, ORR, and OS when treated with everolimus compared to sunitinib [131].

MiTF

MiT family translocation RCC is a rare and aggressive variety of RCC appearing in the pediatric and young adults population,characterized by various translocations [136].

BAP1

As aforementioned, BAP1 mutation disrupts the suppression of cell proliferation, sensitizing RCC cells to genotoxic stress. Consequently, resistance to immunotherapy was observed in patients with loss of BAP1, which upregulated immunosuppressive genes [137]. At the same time, when patients with Microphthalmia Transcription Factor (MiTF) family translocation RCC (MiTF) were treated with immune checkpoint inhibitors (ICIs), it was reported that mutations in PBRM1 correlated with clinical benefits [138]. Also, it was reported that a better response to anti-PD1 therapy nivolumab was associated with loss-of-function mutations of PBRM1 [139].

Overall RCC regardless of histology classification

Lastly, biomarkers for overall RCC regardless of histology classification are listed below.

IL8

In a study about aRCC patients, the median OS of wild-type IL8 2767AA was 29.6 months, while the median OS of the TT variant genotype was 14.8 months. In patients with heterozygous IL8 2767, the median OS was 23.9 months (n = 119, 53%). Similarly, poorer OS was detected in the VEGFA-1154 G > A variant AA genotype in comparison with the wild-type GG genotype [median OS: 16.7 vs. 25.3 months, HR (95% CI) = 2.2 (1.3, 3.6), p = 0.004].. So as NR1IR, NR1IR − 25385C > T was found to be associated with reduced OS (p < 0.05) [140]. What’s more, a study conducted by Xu et al. indicated that the IL8 2767TT and IL8 251AA variant genotypes led to poorer PFS as compared with the wild-type genotype [126, 141].

ABCB1

Sunitinib is a substrate of ABCB1, and ABCB1 knockout is found to be associated with increased accumulation of sunitinib in the brain [142]. ABCB1 1236C/T and 2677G/T were also related to the clearance of sunitinib [54]. Chu et al. reported that the presence of the ABCB1 3435, 1236, and 2677 TTT haplotypes could be associated with sunitinib resistance. Besides, these polymorphisms are correlated with inferior PFS and OS (PFS: HR = 5.5, P = 0.001; OS: HR = 5.0, P = 0.005) [143].

PD-L1

Shiuan et al. revealed that in RCC patients, respondents to anti-PD-1 or PD-L1 therapy exhibited a higher percentage of cells expressing PD-L1. In addition, high PD-L1 expression (> 10% cells) also correlated with improved PFS in both tumor cells and non-tumor cells [144]. Meanwhile, Taube et al. also indicated that PD-L1 expression was associated with patient response to anti-PD-1 therapy [145]. Likewise, in the JAVELIN-101 trial, PD-L1 expression (≥ 1% immune cells) was analogous to prolonged PFS in the combined therapy of avelumab and axitinib and a shorter PFS in the sunitinib subset [146].

MiR-155 and miR-484

What’s more, Merhautova et al. resesarched two miRNAs, miR-155 and miR-484. MiR-155 is an oncogene that promotes neovascularization, while the biological role of miR-484 remains to be clarified. Through the Kaplan–Meier analysis, patients on sunitinib treatment with the expression of both miRNAs reported increased time to progression (TTP). This finding indicated that miR-155 and miR-484 might be related to sunitinib resistance and poor response to therapy [128].

STAT3

Besides, the signal transducer and activator of transcription 3 (STAT3) act as a transcription factor that is activated via the phosphorylation of kinases through cytokines and growth factors [147]. STAT3 is an important target for cancer therapy because of its involvement in the apoptosis, migration, and proliferation of tumor cells [148–150]. 51 mRCC patients were treated with sunitinib. Those with phosphorylated p-STAT3-positive tumors experienced significantly poorer PFS in comparison with p-STAT3-negative tumors. Besides, patients with p-STAT3-negative tumors reported longer OS [151].

HIF-2α

Additionally, HIF-2α promotes aggressive tumor behaviors by regulating genes related to E2F and MYC transcriptional activities, ribosome biogenesis, and lipoprotein metabolism. This reduces the expression of antigens and the activation and infiltration of CD8 + T cells and interferon signaling [152]. The HIF2A gene was associated with the inhibition of disease progression (OR = 0.11 (0.02– 0.75)) and longer OS (P = 0.048).

CDKN2A/B

Interestingly, the CDKN2A/B gene acts as a tumor suppressor, and functional depletions can lead to increased cell-cycle, anabolic metabolism and high proliferation of tumor cells. In the randomized Phase III clinical trial by Motzer RJ et al., mutations in CDKN2A/B led to increased cell cycle and anabolic metabolism. Treatment with combined atezolizumab and bevacizumab or sunitinib reported the association between mutant CDKN2A/B and both prolonged PFS (mPFS: 8.3 months vs 4.1 months) and higher ORR (42% vs 20%) [74].

PAK1

PAK1, a member of the p21-activated kinase (PAK) family, was found to be activated by Rac, Cdc42, and Rho GTPases, functioning in steroid receptor and growth factor signaling, cell survival, gene transcription and oncogenic transformation. PAK1 is revealed to be up-regulated in many human cancers and is related to tumor progression and therapeutic resistance intimately [153]. Accordingly, a retrospective study with 138 mRCC patients who were treated with TKIs indicated that low PAK1 expression resulted in prolonged OS (3.301 (2.579 − 10.899), P < 0.001) and PFS (3.108 (1.795 − 5.381), P < 0.001), implying an improvement in the therapeutic effect of sunitinib [127].

Discussion

The biomarkers mentioned above mainly derive from three perspectives. Among these, IL8, VEGFR, ABCB1, VHL, BAP1, PBRM1 and PD-L1 are studied more often than others but no definitely prognostic and predictive biomarkers for response to therapy in patients with RCC are validated and ready to use in clinical practice.

In terms of SNP, biomarkers related to angiogenesis seem to have influence on therapeutic effect. Polymorphisms of IL8 decreased the therapeutic effect of drugs through anti-angiogenesis and, similarly, polymorphisms in the VEGFR family also play an important role in promoting vasculature growth and remodeling in tumor cells. Hence, drugs targeting angiogenesis, such as sorafenib, sunitinib, and axitinib, should be re-considered when dispensed. Polymorphism of ABCB1 reduces the clearance of sunitinib and its active metabolite SU12662, implying the importance of patient profile during treatment selection. Nevertheless, the strength of studies are hampered by small sample sizes and the single institution, which should be further validated in larger trials.

Mutations in RCC such as VHL, VEGF-TKI, BAP1 and PBRM1 are emerging as predictive factors but are far from being convincing. The most common mutation in ccRCC is exhibited in VHL, which can cause abnormal vascularization. Consequently, VEGF-TKI systemic therapy is not suitable for patients with VHL mutation. While the intratumor heterogeneity, baseline clinical differences between responders and nonresponders, single specimens which may not identify all subclone drivers during the evolution process and lack of evaluation of after-treatment specimens raise the need of larger studies with more statistical power. Likewise, the mutation of BAP1, also located on chromosome 3p, disrupts the cell proliferation process and sensitizes RCC cells to genotoxic stress. Mutated BAP1 correlates to poor prognosis of TKIs, mTOR inhibitors, and VEGF-targeted therapy, indicating that immunotherapy might be a better treatment option. However, the conclusion is limited by small sample size, an unavoidable risk of bias and the discrepancy between time of tumor sampling and anti-PD1 therapy. So as to mutated PBRM1, the second most frequently mutated gene after VHL, who is found in novel tumor suppressors and led to unfavorable therapeutic effects in targeted therapy and immunotherapy. On the contrary, a better response was observed for the anti-PD1 therapy nivolumab. Two paradoxical conclusions remind us that more research needs to be conducted to screen for better therapeutic options.

On the expression level, although the expression of PD-L1 plays an incredibly important role in the therapeutic effect of immunotherapy, the predictive value of PD-L1 in therapy of RCC still remains under-studied. Several studies have indicated that high expression of PD-L1 is associated with a better response from anti-PD-1 or PD-L1 therapy. But there lacks a unique standardized method to assess PD-L1 expression level in all tumor types and potential equivalence among difffferent techniques now, resulting in heterogeneous results of positivity, and thus, needs further confirmation of relative researches.

Conclusion

Despite the new drugs being developed to treat RCC, extensive heterogeneity in the clinical outcomes of patients still exist. Several genomics studies have revealed the prognostic potential of various genomic biomarkers to evaluate heterogeneity. The different responses to treatment and therapeutic resistance prompt the search for clinically reliable and useful biomarkers, which can enable better treatment decisions and personalized medicine. However, most of the biomarker studies summarized here have relatively small patient populations, inconclusive findings, poor analytical approaches, absence of further validation, unclear molecular mechanisms, biased retrospective studies, and lack of a standardized genetic evaluation method. Hence, these findings require further validation. Based on the great progress in molecular and genomics research, the potential of predictive and prognostic genomic biomarkers for treatment efficacy definitely holds exciting opportunities to pursue better therapeutic choices for RCC patients.

Author contributions

Weijie Yan wrote the main part of the article. Naiqiao Hou, Wei Zhai & Junhua Zheng guided writing, offering suggestions to revise, reviewed and agreed upon the final version of the manuscript. All authors read and approved the final manuscript.

Data availability

Not applicable.

Declarations

Ethics declarations

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.