ISSN 1933-0863 (PRINT)     ISSN 1933-0871 (ONLINE)

Home               Archives               About Us             Submission Guidelines              Editorial Board              Advertising Information               Contact Us

Next Generation Sequencing in Renal Cell Carcinoma:
Towards Precision Medicine

 

Departments of Internal Medicine1 Kidney Cancer Program2, and Department of Urology3,
Simmons Comprehensive Cancer Center, University of Texas Southwestern Medical Center,
Dallas TX, 75390

 

 

Keywords: Kidney cancer, molecularly targeted therapies, immune checkpoint inhibitor,
next generation sequencing, genomics, renal cell carcinoma, BAP1, PBRM1

 

Corresponding Author: James Brugarolas, University of Texas Southwestern Medical Center.
5323 Harry Hines Blvd., Dallas, TX, USA, 75390-8852. Phone: 214-648-4059;
Fax: 214-648-1955.  E-mail: james.brugarolas@utsouthwestern.edu

 

 

 

Introduction

Worldwide, there were over 400,000 new cases and 175,000 deaths attributable to renal cell carcinoma (RCC) in 2018.1 In the United States alone, there are predicted to be over 73,000 new cases of RCC, accounting for nearly 15,000 deaths.2 17% of patients present with metastatic disease with only 12% of patients surviving 5 years.3 Fortunately, outcomes are improving. The number of efficacious systemic therapies for RCC has increased over the past decade and there are now over a dozen FDA approved agents and combinations for use in metastatic RCC.4-6 The treatment landscape has changed from one comprised exclusively of recombinant cytokines to one which includes angiogenesis inhibitors (mostly tyrosine kinase inhibitors, TKI), mammalian target of rapamycin (mTOR) inhibitors, and most recently, the immune checkpoint inhibitors (ICI).4,7 Despite this progress, precision medicine has advanced little and there are no biomarker tests approved by the FDA to guide treatment selection.

     Classically, RCC is subdivided histologically into clear cell RCC (ccRCC) accounting for 75% of cases, type I and II papillary RCC (pRCC) accounting for 10% of cases, chromophobe RCC (chRCC) accounting for 5% of cases, and other less frequent subtypes.8 RCC is now recognized to be a diverse group of diseases with updated society guidelines incorporating molecular and genomic data along with histologic information when defining RCC subtypes.9,10

 

Discovery of the von Hippel-Lindau Tumor Suppressor Gene

The first tumor suppressor discovered in ccRCC was the von Hippel-Lindau (VHL) gene. VHL syndrome is an autosomal dominant disease caused by germline mutations in the VHL gene. This syndrome results in development of numerous ccRCCs among other manifestations.11 Genetic linkage analysis of affected kindreds located the responsible gene to the chromosome region 3p25-26 in the late 1980s.12,13 Latif and colleagues were the first to identify the VHL gene,14 and the following year VHL was also shown to be mutated in sporadic ccRCC.15 Nearly two decades later, Nickerson and colleagues evaluated a cohort of 205 ccRCC samples, the largest cohort at that time, for aberrations in the VHL gene through targeted sequencing. They identified non-silent somatic mutations with a prevalence of 82% and VHL promoter hypermethylation in an additional 8.3% of tumors, for a total of 90% of tumors.16 In sporadic ccRCC, the initiating event is thought to be loss of 3p through a chromothripsis event.17 This is likely followed by mutation or epigenetic silencing of the second allele. While VHL loss is nearly universal in ccRCC, VHL inactivation has also been shown in preneoplastic cysts, and mice with VHL disruption in the kidneys do not develop ccRCC.18-20 Thus, while VHL loss is an important initiating step in the development of ccRCC, additional driver mutations were suspected.

     The VHL protein forms a complex with Elongin B, Elongin C, CUL2, and RBX1 that serves as an E3 ubiquitin ligase which, at normal oxygen tension, acts on the alpha unit of hypoxia inducible factor (HIF) transcription factors, leading to their degradation.21 Loss of VHL in ccRCC results in constitutive activation of HIF and expression of its target genes.21,22 Interestingly, mutations in TCEB1 (encoding Elongin C) have been reported in up to 5% of ccRCC cases and are associated with loss of heterozygosity of chromosome 8, which contains TCEB1.23 In addition, mutations in CUL2 were found in up to 1% of ccRCCs.24 Mutations in TCEB1 and CUL2 tend to be mutually exclusive with VHL mutations and likely represent other mechanisms to disrupt VHL function.25

 

Identification of the PBRM1 and BAP1 Tumor Suppressor Genes

The first large scale sequencing reports in ccRCC came out in 2009, and demonstrated frequent mutations in the chromatin remodeling genes SETD2, KDM6A (also known as UTX), KDM5C (also known as JARID1C), and MLL2.26,27 These reports were limited to sequencing a panel of ~3,500 genes, a small subset of the entire human exome, and the genes identified were mutated in up to 15% of tumors. In the subsequent two years, whole exome sequencing (WES) allowed the discovery of two major drivers of ccRCC; polybromo 1 (PBRM1) and the BRCA1 associated protein-1 (BAP1). Varela and colleagues performed WES of seven ccRCC and matching normal samples and identified truncating mutations in PBRM1 in four tumors.28 Non-silent mutations in PBRM1 were subsequently identified in 88 of 221 (39.8%) ccRCC cases, making it the second most common mutation in ccRCC.28 The following year, WES of paired tumor and patient-derived xenografts (PDX, also called tumorgrafts) identified BAP1 as an important driver of ccRCC.29 By incorporating tumorgrafts, our studies allowed for accurate calls of mutant allele frequencies (MAF) thereby confidently nominating putative two-hit tumor suppressor genes. In tumorgraft models, BAP1 was the only gene in addition to VHL and PBRM1 to demonstrate a MAF of ~1 (signifying absence of the wild-type allele). Subsequent BAP1 targeted sequencing studies identified mutations in 24 out of 176 (14%) largely primary ccRCC samples.29

     PBRM1 and BAP1 are both involved in chromatin modification and epigenetic regulation of gene expression. PBRM1 encodes BAF180, a component of the switching defective/sucrose nonfermenting (SWI/SNF) family of nucleosome remodeling complexes, which is thought to mediate recruitment of the complex to specific nucleosomes through recognition of acetylation patterns.25,30 The BAP1 protein is a nuclear-localized deubiquitinase which acts to deubiquitylate H2AK119ub1 to reverse polycomb-mediated gene repression.30 These processes likely regulate different gene subsets, and we discovered that BAP1 and PBRM1-deficient tumors can be distinguished by their gene expression signature.31

     Interestingly, both BAP1 and PBRM1 are located on chromosome 3p, within a 43 Mb region that is frequently lost in ccRCC, which also includes VHL and SETD2.17,32 Notably, BAP1 and PBRM1 mutations were found to anticorrelate in ccRCC and the prevalence of combined BAP1/ PBRM1 deficient ccRCC is ~1-2%, less than the ~5% rate expected given the rates of PBRM1 (~45%) and BAP1 (~12%) mutations in ccRCC.29,31-34 While mutual exclusivity often indicates that the encoded proteins are in the same pathway (and reflects the low selective pressure to have them simultaneously mutated), the finding that loss of BAP1 and PBRM1 results in distinct gene expression signatures, suggested that this was not the case for these genes.29,31 Furthermore, BAP1 mutant tumors exhibited higher nucleolar grade, more aggressive histology, and demonstrated worsened RCC-specific survival.31,35-38 This finding was subsequently confirmed in metastatic RCC, as BAP1 mutations were independently associated with worsened overall survival.39,40 Together these findings led us to propose a model where following inactivation of VHL and 3p loss, inactivation of the remaining copy of BAP1 caused aggressive ccRCC and inactivation of PBRM1, less aggressive ccRCC.25 Using methods we developed,29 this model was subsequently revised by the TRACERx consortium to show that the first event in sporadic tumors is likely the loss of 3p.17

     To determine the role of BAP1 and PBRM1 in ccRCC development, we inactivated them in nephron progenitor cells in the mouse and assessed their impact on RCC development. By simultaneously targeting Bap1 and Vhl, we developed the first mouse model of ccRCC thereby over- coming a decade-long struggle.19 We showed that ccRCC development required not only Vhl inactivation, but also the inactivation of Bap1 (or Pbrm1). As for Bap1, Pbrm1 loss was not sufficient to induce RCC. However, the simultaneous inactivation of Vhl and Pbrm1 caused ccRCC.37 Similar observations were made by others.41,42 We also found, that as in humans, Bap1-deficient tumors were of high grade, whereas Pbrm1-deficient tumors were of low grade. In addition, Pbrm1-deficient tumors developed after a longer latency period.37 Interestingly, targeting one allele of Tsc1, which encodes a negative regulator of mTOR complex 1, in a Vhl/Pbrm1-deficient background reduced the latency period and increased the frequency of higher grade tumors37 (Figure 1).

 

Figure 1. PBRM1, BAP1 and TSC1 are drivers of ccRCC development and tumor grade. According to PBRM1 and BAP1, ccRCC can be subclassified into 4 different subtypes. Double mutant tumors are under-represented suggesting that simultaneous mutations are mutually exclusive in tumors. PBRM1- and BAP1-mutant tumors are associated with different biology (gene expression), pathological features, and outcomes in patients. Modeling studies in the mice show that (i) Vhl inactivation is insufficient for ccRCC development; (ii) the combination of Vhl and Pbrm1 mutations results (as in humans) in low grade tumors; (iii) the combination of Vhl and Bap1 mutations results (as in humans) in higher grade tumors; and (iv) Tsc1 mutations increase the grade of Vhl/Pbrm1-mutant tumors.

 

 

     These discoveries explain why germline VHL mutations cause kidney cancer in humans, but not in mice. As it turns out, we found that in mice the Bap1 and Pbrm1 genes are on a different chromosome than Vhl and as such, loss of the chromosome arm containing the Vhl gene in the mouse would have no effect on Bap1 or Pbrm1 genes. In contrast, loss of 3p in humans results not only in the loss of a VHL allele, but also one allele of both BAP1 and PBRM1. Therefore, the arrangement of tumor suppressor genes in the genome likely accounts for differences in tumor predisposition across species.19,25,37

     In summary, somatic mutations in either PBRM1 or BAP1 tend to be mutually exclusive and activate distinct gene expression programs in tumors leading to differentiated pathological features, ultimately causing divergent clinical outcomes. As such, these discoveries resulted in the first molecular classification of ccRCC. Several unanswered questions remain, however. For instance, what other events cooperate with VHL in tumors that are seemingly wild-type for BAP1 and PBRM1? Or how do mutations in TCEB1 and CUL2, which are located on chromosome 8 and 10 respectively, lead to ccRCC formation, and what are the specific cooperating events?

 

Next Generation Sequencing Supports Comprehensive Integrated Analysis of RCC

The development of NGS and improved bioinformatics tools allowed the collection and integration of data from WES, copy-number analyses, DNA-methylation analyses, and messenger RNA and microRNA sequencing, from individual samples at a large scale.23,33 Through integrated analyses of ~400 ccRCC samples, The Cancer Genome Atlas (TCGA) consortium not only confirmed previous findings, but significantly expanded on them through the identification of altered subnetworks. Initial genome-wide copy number analysis, using next-generation sequencing technologies, identified 3p and 14q loss, as well as 5q gain as the three most common somatic copy number alterations (SCNA) in ccRCC.43,44 These were verified in a larger TCGA cohort - 3p loss (91%), 14q loss (45%) and 5q gain (67%).33 The most frequently mutated genes involved chromatin remodeling pathways. These included genes encoding the SWI/SNF family proteins PBRM1, SMARCA4 and ARID1A (47.1%), the histone methyltransferases SETD2 and MLL (23.8%), and the histone deubiquitinase BAP1 (12.1%). Alterations in PTEN, TSC1, and other components of the PI3K/AKT/mTOR pathway were observed in 28% of ccRCCs.33,45,46 Other findings of significance included loss of one CDKN2A allele in 16.2% of ccRCCs (mostly through deletion of 9p21.3), and mutation of TP53 in 2.6% of cases.24,46

     In a similar analysis of non-ccRCC, we found distinct mutated genes and SCNA alterations highlighting the diverse molecular landscape of RCC.47 In pRCC, we identified 10 significantly mutated genes including MET, SETD2, and NF2. Activating MET mutations were identified in 15% (10/65) of pRCC analyzed, including 4 previously unreported mutations. Furthermore, 70% of pRCC samples had amplification of chromosome 7, containing MET, and these samples had higher levels of MET expression, consistent with the role of MET in type I pRCC.47 These findings were confirmed in a subsequent analysis of the pRCC TCGA cohort, where activating MET mutations were identified in 17% (14/75) of type I pRCC samples, which exhibited near universal gain of chromosomes 7 and 17.48 In contrast, copy number analysis of type II pRCC revealed two distinct subtypes, one with relatively few SCNA and another with a high degree of aneuploidy and frequent chromosome 9p loss.48 DNA methylation studies revealed a subset of pRCC (9/160) referred to as the CpG Island Methylator Phenotype (CIMP), eight of which were categorized as type II pRCC histologically. CIMP tumors were characterized by universal hypermethylation of CDKN2A, frequent fumarate hydratase (FH) mutations (including somatic mutations; 57%), and worse survival relative to non-CIMP pRCC.48 Loss of FH and the subsequent accumulation of fumarate has been demonstrated to result in deregulation of the nuclear factor erythroid 2-related factor (NRF2)/antioxidant response element (ARE) pathway through direct effects of fumarate on KEAP1, an inhibitor of NRF2.49 This was first demonstrated in hereditary leiomyomatosis and renal cell cancer (HLRCC), an inherited form of type II pRCC arising from germline inactivation of FH. Consistent with the notion that FH loss leads to deregulation of the NRF2/ARE pathway, expression of NQO1, a canonical transcriptional target of NRF2, was highest in CIMP tumors.24,48 Targeted sequencing in sporadic type II pRCC revealed activating mutations in the NRF2/ARE pathway in four of five cases.50 Genomic analysis of the TCGA cohort supported this finding with activating mutations in the NRF2/ARE pathway in 16.7% (10/60) of cases.48 An NRF2/ARE gene transcription program is a distinguishing feature of type II pRCC.48

     Analysis of chRCC revealed distinct molecular alterations defining this subgroup. chRCC can be classified histologically into classical and eosinophilic subtypes, the latter characterized by abundant eosinophilic cytoplasm and densely packed mitochondria.51 Analysis of the TCGA cohort of 66 chRCC identified molecular distinctions between these two subgroups.51,52 All 47 cases of classical chRCC in the TCGA cohort demonstrated characteristic chromosomal losses of chromosomes 1, 2, 6, 10, 13, and 17 whereas only half (10/19) of the eosinophil variants did.47,52 Studies from TCGA and our own group found that TP53 mutations were significantly enriched (P = 2.3E-5) in the classic chRCC subtype.47,52 The two most frequently mutated genes across chRCC variants are TP53 (31.1%) and PTEN (9%).47 Mutually exclusive mutations in genes of the PI3K/AKT/mTOR pathway were observed in 23% of cases.47,52 Interestingly, analysis of mitochondrial DNA in chRCC revealed recurring mutations in genes encoding components of complex 1 of the electron transport chain (ETC), although these mutations were not associated with changes in the expression of genes implicated in oxidative phosphorylation.52 chRCC are particularly enriched in mutations involving metabolic genes, including deleterious mutations in PDHB (which encodes a critical component of the pyruvate dehydrogenase complex) and PRKAG2 (encoding one of three subunits of AMP-Kinase (AMPK)). These findings reinforce the long-standing implication that metabolic derangements in RCC can contribute to oncogenesis.

     Integrated transcriptome and protein expression analysis (using reverse protein phase arrays) in TCGA revealed distinct metabolic patterns both across and within histological subtypes of RCC.24,5 ccRCC was characterized by overexpression of glycolytic and fatty acid synthesis enzymes as well as suppression of Krebs and ETC programs. This contrasts with pRCC and chRCC, which generally expressed intermediate and high levels of Krebs and ETC proteins, respectively.24 Decreased expression of AMPK (which inhibits fatty acid synthesis and mTOR), and increased ribose sugar metabolism, correlated with higher stage and worse prognosis in ccRCC.33 The CIMP subtype of pRCC demonstrated the highest level of ribose sugar metabolism across all RCC subtypes. A metabolically divergent subset of chRCC was identified, which demonstrated decreased levels of AMPK and ETC genes, increased ribose sugar metabolism, and demonstrated a profoundly worse outcome than metabolically traditional chRCC.24 This metabolically divergent subtype of chRCC lacked the typical chromosomal aberrations of chRCC, and four of the six displayed sarcomatoid de-differentiation. Taken as a whole, these landmark studies contributed tremendously to our understanding of the molecular biology of RCC, and the findings are driving the development of molecularly-based prognostic models across histological subtypes in RCC

 

Identification of Prognostic and Predictive Biomarkers

Successful prognostication of patients is critical to both the clinical practice of oncology and research, as it influences treatment approaches and may be used to stratify patients in clinical trials. Classically, the most robust prognostic variables were related to histology (subtype, grade, tumor size), clinical features (performance status and pace of disease), or laboratory parameters. For localized disease treated with curative nephrectomy, the two main prognostic models are the UCLA Integrated Scoring System (UISS) and the Mayo Clinic Stage, Size, Grade, and Necrosis (SSIGN) score.54,55 The UISS score integrates pathological (tumor size and nuclear grade) with clinical (ECOG performance status) variables, whereas the SSIGN score utilizes strictly pathological variables. Despite the frequent use of the aforementioned models in patient prognostication, a recent analysis has called into question the predictive power of these nomograms.56 Utilizing prospectively collected data from the ASSURE trial,57 a randomized phase III placebo-controlled trial of sunitinib and sorafenib in the adjuvant setting, Correa and colleagues analyzed 8 recurrence prediction models including the SSIGN and UISS tools. They reported that the predictive power for each model’s predefined outcome, as measured by the co-occurrence index (C-index), fell considerably for all models in the dataset. However, among these models, the SSIGN score performed best with a C-index of 0.688 (95% CI: 0.686 – 0.689). Most other nomograms provided only marginal improvement relative to the TNM staging system.56 Integrating underlying genomic information with clinical and histological variables is one strategy to improve these models. We previously demonstrated that BAP1 and PBRM1 expression by immunohistochemistry (IHC) are highly correlated with mutation status (P = 3E-58 and 4E-23 respectively34 and that BAP1/PBRM1 expression by IHC are independent predictors of both disease-free survival and overall survival in the localized disease setting.31,34 PBRM1 and BAP1 expression status did not, however, add independent prognostic information to the SSIGN score in a multivariate analysis of nearly 1,500 cases of localized RCC.34 One possible explanation is that BAP1-mutant tumors are significantly more likely to demonstrate higher nuclear grade, stage, and necrosis than PBRM1-mutant tumors.34-37 Thus, despite progress, additional strategies to stratify patients in the localized disease setting remain needed.

     For advanced disease, the most commonly used prognostic models are the Memorial Sloan Kettering Cancer Center (MSKCC) risk model58 and the International Metastatic RCC Database Consortium (IMDC) risk model.59 Both models integrate clinical variables (time from diagnosis to systemic treatment and performance status) and laboratory variables including calcium and hemoglobin (as well as lactate dehydrogenase in the MSKCC model vs platelet and neutrophil counts in the IMDC model) but are agnostic to the molecular biology of the underlying disease. Many of the key driver mutations in RCC have been shown to have negative (BAP1, SETD2, PTEN, and TP53) and positive (PBRM1) prognostic significance,24,34,53,60 although it is important to interpret the prognostic implications of these mutations in the context of histology. For example, PBRM1 mutations which tend to associate with favorable prognosis in ccRCC were strongly associated with reduced survival in type I pRCC (p<0.0001).24 Regardless, retrospective analysis of two phase III trials in metastatic RCC, COMPARZ (first-line sunitinib vs pazopanib) and RECORD-3 (first-line sunitinib vs everolimus) identified PBRM1, BAP1, and TP53 as having independent prognostic value.40 Their prognostic significance persisted after multivariate analysis with the traditional variables in the original MSKCC model. This resulted in the development of a genomically-annotated MSKCC risk model which stratifies patients into four risk groups (favorable, good, intermediate, and poor) versus three in the original. The genomically-annotated model had a more balanced distribution across these groups and an improved ability to predict overall survival. The C-index of the original model was 0.567 (95% CI: 0.529 – 0.604) vs 0.637 (95% CI: 0.529 – 0.604) with the new model.40 While promising, this model will need to be validated in the prospective setting.

     Prognostic models have also been developed using gene expression data. One of the first such attempts in ccRCC was by Brannon and colleagues who collected transcriptomic data in 47 ccRCC primary tumor samples. Unsupervised clustering analyses revealed two subgroups, referred to as clear cell types A and B (ccA/B), which had distinctly different clinical outcomes with a median survival of 59 vs 36 months respectively in a validation cohort of 177 patients (P=0.004).61 A 34 gene panel (Clearcode34), which could delineate ccA and ccB subtypes, was found to better predict relapse-free survival and cancer-specific survival than both the UISS and SSIGN models in a cohort of nearly 500 patients with localized RCC.62 Similar findings were seen in a smaller metastatic cohort (54 patients), however, only ccB classification added prognostic value when incorporating the IMDC and MSKCC risk models.63

     More recently, cluster of cluster analysis including DNA copy number, mRNA, microRNA, DNA methylation, and protein expression data from the TCGA cohort revealed 9 unique clusters across all histological subtypes,53 three of which were enriched in histologically defined ccRCC (e.1, e.2, and e.3). These demonstrated significant differences in survival, with e.3 demonstrating the poorest prognosis and e.2 demonstrating the most favorable. Tumors in the e.3 cluster tended to have frequent loss of the CDKN2A gene, frequent BAP1 mutations, and overexpression of cell cycle and hypoxia-related genes.53 Additionally, e.3 was enriched in inflammatory gene expression relative to e.2, and expression of PDCD1 (encoding for PD1) and CTLA4 were independent predictors of poor prognosis across the ccRCC subtype.53 The e.2 subgroup, on the other hand, was found to demonstrate significant overlap with the previously described ccA subtype,53,61 which is distinguished by overexpression of angiogenesis-related genes such as such as FLT4, FLT1, and VEGFB.61,62,64 Thus, given that the most widely utilized treatments in metastatic RCC are immune checkpoint inhibitors, angiogenesis inhibitors, or combinations of the two,4,7,65,66 transcriptomic information regarding inflammation or angiogenesis may also have predictive potential.

     Unlike prognostic biomarkers which inform on the spontaneous trajectory of the disease, predictive biomarkers have the potential to inform on treatment selection, which may be tailored to the particular patient. Beyond the obvious benefit of selecting the most efficacious treatment, predictive biomarkers have the potential to reduce adverse events as well as cost from exposures to non-beneficial drugs. The use of gene expression profiles to predict treatment response was considered in an exploratory analysis of the IMmotion150 trial, a phase 2 study comparing the combination of atezolizumab (a PD-L1 inhibitor) and bevacizumab (a neutralizing VEGF antibody) (atezo+bev) to atezolizumab (atezo) or sunitinib monotherapy.67 McDermott and colleagues showed that predefined gene signatures of angiogenesis (Angio), immune T cell infiltration (Teff), and myeloid inflammation (Myeloid) may have predictive potential. They found that patients with the AngioHigh gene signature had an improved progression-free survival (PFS) compared to those with AngioLow with the angiogenesis inhibitor, sunitinib.67 Interestingly, PBRM1 mutations were found to be enriched in the AngioHigh subgroup. In line with these findings, a similar analysis of the COMPARZ trial evaluating two anti-angiogenic agents, sunitinib and pazopanib, found that high expression of angiogenesis genes was associated with improved response rates.64 Furthermore, tumors with PBRM1 mutations demonstrated significantly higher angiogenesis gene expression scores than those with BAP1 mutations.64 In the IMmotion150 trial, patients with the TeffHigh gene signature had improved PFS with atezo+bev compared to the sunitinib and the atezo arms. Intriguingly, when examining the effect of myeloid signature on outcomes within the TeffHigh population, a PFS advantage of atezo+bev over atezo was observed in the TeffHigh /MyeloidHigh subgroup (n=66) but not in the TeffHigh /MyeloidLow subgroup (n=66). This finding may identify a particular population of patients which would benefit from combination ICI and TKI therapy. However, this is a relatively small study and has not yet been reproduced in larger datasets.

     These studies highlight the potential role of the tumor microenvironment (TME) on clinical outcomes. Bioinformatics techniques can be leveraged to gather information about the TME from RNA extracted from tumors. One such approach is to utilize single sample gene set enrichment analysis (ssGSEA) in which signature gene panels attributed to particular cell types are utilized to disentangle a heterogeneous tumor.68 Our group expanded upon this method by exploiting patient-derived tumorgrafts (patient tumors implanted in mice), where the human TME is ultimately replaced by the host.69 By focusing on human RNA and subtracting from the patient tumor the transcriptome of the tumorgraft, one is left with the human TME transcriptome. Utilizing this dissection algorithm and a cutoff of 20-fold to distinguish TME vs tumor genes, an empirically derived TME (eTME) signature was obtained. Clustering analyses of the ccRCC TCGA (KIRC) tumors according to the eTME revealed two subgroups, an inflamed subtype (IS) and a non-inflamed subtype (NIS). Interestingly, the IS was enriched for BAP1 mutations (P = 7.7E-5) and demonstrated a worse prognosis compared to NIS. Furthermore, the eTME-IS subtype correlated with systemic inflammatory markers such as elevated platelet counts and decreased hemoglobin levels, as well as worse prognosis in three distinct cohorts totaling approximately 1,000 patients. This correlation draws a link between inflammatory subtypes of RCC and key prognostic variables in IMDC or MSKCC models. Notably, the presence of such variables is associated with intermediate/poor risk disease, which suggests that inflammatory tumors are particularly aggressive in patients. Interestingly, in the Checkmate 214 trial comparing the combination of nivolumab and ipilimumab to sunitinib, the intermediate/poor risk groups appeared to derive the most benefit from ICI,65 which is consistent with the predictive potential of the IS subtype. Thus, the combined use of NGS and bioinformatics has potential to predict responses to therapy and is an active area of investigation.70

 

Intratumoral Heterogeneity

One important disadvantage of the aforementioned studies is the use of a limited number of tumor samples per patient (typically just one), as this may fail to capture intratumoral heterogeneity (ITH). In one of the earliest attempts to measure ITH in ccRCC, Gerlinger and colleagues performed multiregional WES of two primary tumors and demonstrated that only 31% of the somatic mutations were ubiquitous amongst all sampled regions.71 Furthermore, the previously described “cc-A” and “cc-B” gene expression patterns could be identified in spatially distinct areas of an individual patient’s tumor, highlighting how ITH can confound efforts to establish effective prognostic models based on analyses of a single sample.71,72  This laid the groundwork for the TRAcking Cancer Evolution through therapy (Rx) (TRACERx) consortium, which prospectively collects tumor samples and performs multiregional sequencing, when possible, over time.73 In a recently published report, Turajlic and colleagues sequenced 1,206 primary tumor samples from 101 patients with the use of a 110 gene panel, allowing for an unprecedented view of the molecular diversity within a single tumor.74 Multiregional sampling allowed for the detection of clonal and subclonal somatic mutations. Thus, the prevalence of PBRM1 (55%), SETD2 (25%), BAP1 (19%) and other driver mutations could be more accurately calculated.74 Interestingly, while BAP1 and PBRM1 mutations could be identified in the same tumor, they were typically located in spatially distinct regions, consistent with previous reports that these mutations anticorrelate with one another and are found in different areas of the same tumors.29,32,34,74

 

Figure 2. An integrated molecular genetic classification of ccRCC. The initiating event in most  ccRCC is thought to be chromothripsis with loss of 3p, followed by loss of the remaining VHL allele, represented by the first layer in the pie chart. PBRM1 mutations tend to be an early truncal event, and in the absence of other driver mutations, result in tumors of low aggressivity. Further acquisi-tion of driver somatic copy number alterations (SCNA), mutations in mTOR pathway proteins (TSC1, TSC2, MTOR, RHEB, PTEN), or in SETD2 increases tumor aggressiveness. When a substantial compet-itive advantage is gained by a particular clone, it becomes the dominant clone reducing thereby intra-tumoral heterogeneity and showcasing increased aggressiveness. The natural history of BAP1/PBRM1-deficient tumors is one of high aggressiveness, but inasmuch as PBRM1 mutations may sensitize to anti-angiogenic agents, the natural history is altered, and outcomes may be improved. Tumors with mutations in VHL and BAP1 are more aggressive than those with VHL and PBRM1 mutations as a whole. However, VHL/BAP1 tumors may be inflamed, and inflammation  may impact response to immune therapies. A subset of tumors are characterized by wild-type  VHL protein, which herein refers to tumors with no mutations in any VHL complex subunit, and these tumors tend to be heterogeneous and aggressive.

 

     Genomic data obtained from spatially distinct regions provided the ability to assess the timing of mutations, and thus patterns of tumor evolution could be inferred (Figure 2). In the TRACERx studies, seven distinct patterns could be identified utilizing rule-based clustering. However, 36.6% (37/101) of cases could not be assigned an evolutionary subtype. Subtypes were assessed for nuclear grade, stage, microvascular invasion, genomic instability, and degree of ITH. The most aggressive subtype, based on the aforementioned parameters, was the “multiple clonal drivers” subtype, which contained truncal aberrations in two or more of the following: BAP1, PBRM1, SETD2, or PTEN.74 It is important to note that in the “multiple driver mutations” subtype, the temporal relationships of driver mutations were indistinguishable, which separates this group from subsets defined by subclonal (sequential) aberrations in driver genes. The “multiple clonal drivers” subtype includes BAP1/PBRM1-deficient tumors previously described which we and others noted to portend a poor prognosis.29,34,40 The “BAP1 driven” evolutionary subtype, characterized by truncal VHL and BAP1 mutations was also found to have aggressive histological characteristics, with decreased DFS and OS. Three PBRM1-driven subtypes were noted to have sequential loss of PBRM1 followed by either loss of SETD2, activation of PI3K signaling, or distinct SCNAs.74,75 Consistent with this, in 11/101 cases, PBRM1 loss was noted to precede SETD2 loss, but the converse was not seen.75 The least aggressive subtype was the VHL monodriver subtype, which likely represents sampling early in the disease course. When compared with “multiple clonal driver” and “BAP1 driven” subtypes, the three PBRM1 subtypes tended to have increased ITH, less aggressive biology, and demonstrated a more attenuated disease course.74,75 Conversely, the “multiple clonal driver,” “BAP-1 driven” and, perhaps unexpectedly, “VHL wild-type” evolutionary subtypes tended to demonstrate rapid progression to metastases.75 Decreasing ITH tended to correlate with a more aggressive disease course, consistent with the notion that an aggressive clone would “outcompete” other clones. On the other hand, evolutionary subtypes with greater degrees of ITH demonstrated a more attenuated disease course. Tumors with a high degree of ITH may, however, also harbor a diverse reservoir of cancer cells which are resistant to therapy, providing a potential explanation for the mixed responses sometimes exhibited by patients.74 Capturing the full degree of ITH may be challenging, as on average at least 7 biopsies are required to detect over 75% of variants.74 Though challenging to implement clinically, techniques are evolving that allow the dissociation of the whole tumor followed by single-cell sequencing to provide molecular information at the cellular level.76 Other approaches include lysing large quantities of tumor material with deep sequencing in an attempt to comprehensively catalogue all mutations. Future prognostic and predictive models will likely need to incorporate methods to capture ITH to faithfully predict patient outcomes.

 

Leveraging Next Generation Sequencing for Therapeutic Gain

A deeper understanding of the biological underpinnings of RCC has led to novel therapeutic opportunities. This has resulted in a dramatic shift in the treatment landscape over the past decade. Immunotherapy and ICI+TKI combinations are now frontline therapies and objective response rates as high as 60% are seen.4,65,66 Future directions will focus on identifying the correct agent for the correct patient, as well as developing novel therapies. NGS has allowed the identification of activating mutations in oncogenes, and these have been very effective targets in cancers such as melanoma and non-small cell lung carcinoma. Oncogenes are not, however, commonly mutated in RCC and thus have not been an area of therapeutic gain. One exception is the MET oncogene in which activating mutations have been observed in nearly 20% of type I pRCC.47,48,77,78 Cabozantinib is a TKI with activity not only against the VEGFR, but also MET. However, how much MET targeting contributes to its activity against ccRCC is unclear. In a phase III trial comparing second line cabozantinib to everolimus in ccRCC, cabozantinib was found to result in an improved OS rate (HR 0.66 [95% CI: 0.53 – 0.83]; p= 2.6E-4), as well as improved PFS (HR 0.51 [95% CI: 0.41 – 0.62]; p< 1E-4). MET overexpression by IHC did not predict, however, PFS (HR 0.41 [95% CI: 0.53 – 0.83] vs HR 0.58 [95% CI: 0.43 – 0.79]).79 Retrospective analysis of 112 patients with non-ccRCC treated with cabozantinib demonstrated efficacy with 27% (30/112) of patients achieving an objective response.80 For the small subset of patients with genomic data, 40% (4/10) of pRCC patients with MET mutations demonstrated partial responses.80 While these data did not reach statistical significance, prospective trials investigating cabozantinib (as well as other MET inhibitors) in MET-driven pRCC are ongoing (i.e. NCT03091192).

     While targeting driver mutations may provide benefit to a subset of RCC patients, techniques leveraging tumor suppressor genes therapeutically are needed to benefit the larger population. One strategy to tackle loss-of-function mutations in tumor suppressor proteins has been to inhibit downstream effector pathways. In the setting of VHL loss, HIF-2aα accumulates and binds HIF-1b, and the heterodimer upregulates the expression of hun- dreds of genes important to tumor growth including VEGF.21 Given the key role HIF-2 mediated transcription in ccRCC development, direct inhibition of HIF-2α has substantial potential. Transcription factors such as HIF-2α have classically been regarded as “undrugable,” as they lack catalytic pockets suitable for targeting by small molecules. However, characterization of the atomic structure of HIF-2α identified a highly structured pocket that could be bound by small molecule inhibitors.81,82 Compounds with improved pharmacological properties were subsequently developed through iterative structure-based design.83 This led to the development of PT2385 and PT2399, which were shown to be potent and highly selective inhibitors leading to the dissociation of HIF-2α complexes.22,83,84 Preclinical testing of PT2399 in our laboratory demonstrated decreased tumor growth across ~50% of ccRCC tumorgrafts analyzed (P<0.0001), including in sunitinib resistant tumors.22 However, prolonged therapy with PT2399 led to the development of acquired resistance in tumorgraft models.22 Sequencing of tumorgrafts with acquired resistance to PT2399 led to the identification of point mutations which restored dimerization in the presence of inhibitors,22 one of which was subsequently identified in patient tumors that developed resistance to HIF-2α inhibition.85 In a phase I trial, PT2385 demonstrated a favorable safety profile and disease control lasting greater than 4 months in 40% (21/52) of patients, despite heavy pretreatment with a median of 4 prior therapies.86 PT2977, a second generation inhibitor with more consistent drug circulating levels, demonstrated a similar safety profile to PT2385 in recently reported results of a phase I trial in ccRCC.87 Anemia, which is thought to be an on-target effect through suppression of erythropoietin, was the most common adverse event, and only 4% (2/55) of patients stopped therapy due to adverse events. The patients in this trial were heavily pretreated, 62% (34/55) had greater than three lines of therapy, including TKI and ICI therapy. In spite of this, a promising efficacy signal was seen; the median PFS was 11 months (95% CI: 6 – 17), 24% (13/55) of patients experienced a partial response and 56% (31/55) demonstrated stable disease (NCT02974738).87 A phase II trial of (now MK-6482) in combination with cabozantinib is ongoing (NCT03634540), as well as phase II trials of both agents in VHL syndrome related ccRCC (NCT03108066, NCT03401788).

     An alternative approach to targeting tumor suppressor genes leverages “synthetic lethality,” where loss of two genes results in cell death whereas loss of either gene does not88. Since VHL is lost in nearly all ccRCC, identifying molecular targets that exhibit synthetic lethality with VHL loss is an attractive strategy. Several groups, including our own, have developed high-throughput screening platforms of chemical libraries that are capable of identifying compounds which exhibit selective killing of VHL deficient cells.89-94 In one of the first such studies, Turcotte and colleagues screened a panel of ~64,000 small molecules in parallel on VHL deficient RCC4 cells and RCC4 cells with re-introduced VHL.89 They found that STF-62247 was able to selectively induce apoptosis in VHL deficient cells, likely through inhibition of protein trafficking.89 Utilizing the same screen, STF-31 which acts through inhibition of GLUT1, was also found to have preferential toxicity among VHL deficient cells.90 Employing a strategy where differentially labeled VHL deficient and reconstituted RCC cell lines were co-cultured, we identified homoharringtonine (HHT) as a hit compound.93 Furthermore, HHT demonstrated efficacy in ~30% of tested tumorgrafts.93 While these screens have the potential to identify promising compounds, other strategies utilize short hairpin RNA (shRNA) libraries to identify gene combinations which exhibit synthetic lethality.

     One initial report utilizing a shRNA library directed against 88 kinases in VHL-deficient RCC cell lines identified CDK6, MET, and MAP2K1 as potential targets.95 More recently, an expanded shRNA library targeting ~1000 genes identified EZH1 depletion to be synthetically lethal with VHL loss.96 EZH1/2 are histone methyltransferases which canonically act to trimethylate lysine residue 27 on histone 3 (H3K27). Interestingly, constitutive HIF signaling mediates relative H3K27 hypomethylation, potentially explaining EZH1 and VHL synthetic lethality.96 Pharmacological inhibition of EZH1/2 in VHL- deficient RCC cell lines recapitulated these findings, however, the compounds were toxic in mice models.96 Whereas EZH2 was not identified as exhibiting synthetic lethality in the aforementioned screen, there is preclinical evidence that EZH2 inhibitors may be effective in the setting of BAP1 deficiency. Mice with isolated BAP1 deficiency in hematopoietic precursors develop myelodysplastic syndrome.97,98 In these models, BAP1 deficiency results in increased EZH2 expression and methylation of H3K27. Interestingly, EZH2 depletion by both genetic and pharmacologic methods abrogated the oncogenic effect of BAP1 loss.97 In ccRCC, increased levels of EZH2 expression by IHC are associated with higher grade and worse outcomes.99 In addition, RCC-derived cell lines deficient in BAP1 overexpress EZH2, and are sensitive to EZH2 inhibitors in vitro.100 Furthermore, in a sunitinib-resistant xenograft model of RCC, the EZH2 inhibitor EPZ011929 demonstrated rescue of sunitinib sensitivity through epigenetic reprograming (BAP1 status was not reported in this study).101 At the time of preparation of this manuscript there are no RCC specific clinical trials involving EZH2 inhibitors.

     An alternative downstream target of HIF-2, CCND1 (encoding Cyclin D1) is also overexpressed in VHL-deficient RCC. Cyclin D1 binds the CDK4/6 kinases resulting in phosphorylation and inactivation of the retinoblastoma (RB) protein, with subsequent progression through the cell cycle.102 Of note, RB loss is an uncommon event in ccRCC, and thus cell cycle progression likely remains CDK4/6 dependent. The CDK4/6 inhibitor palbociclib demonstrated a G0/G1 cell-cycle arrest, induction of late apoptosis, and blockade of RB phosphorylation in multiple RCC cell lines.103 Abemaciclib, another inhibitor of the CDK4/6 enzyme, was shown to diminish tumor growth in combination with sunitinib in mouse tumorgraft models,104 and a phase I trial of the combination is now actively recruiting (NCT03 905889). Recently, Nicholson and colleagues demonstrated synthetic lethality between CKD4/6 and VHL in ccRCC cell lines as well as in a Drosophila model, suggesting a fundamental dependency between these two gene products.105 Furthermore, the anti-proliferative effects of CDK4/6 inhibition were synergistic with HIF-2 inhibition in mouse xenograft models of disease,105 suggesting the combination of CKD4/6 inhibitors with the HIF-2 antagonists described above may also have therapeutic potential.

     Another emerging strategy is to target the NRF2 pathway. As previously alluded to, NRF2 is negatively regulated by KEAP1, and under conditions of oxidative stress KEAP1 is bound by p62 releasing NRF2 to localize in the nucleus and bind ARE. The NRF2/ARE pathway regulates a number of genes involved in oxidative stress regulation, drug metabolism, and cell proliferation.106 Among other functions, NRF2 plays a vital role in overcoming oxidative stress and treatment resistance. Accordingly, overexpression of NRF2 has been implicated as a negative prognostic marker in several tumor types.107 Overexpression of NRF2/ARE-controlled genes are a distinguishing feature of type II pRCC,48 and elevated expression of NQO1 is associated with worsened outcomes.48 Activating mutations in ccRCC are less frequent than type II pRCC, however, emerging evidence suggests epigenetic silencing of KEAP1 may contribute to NRF2/ARE deregulation in ccRCC.108 Consistent with a potential role of NRF2/ARE in ccRCC, NRF2 depletion via shRNA was recently shown to decrease proliferation and increase sensitivity to sunitinib in the 786-O ccRCC cell line.109 Several flavonoids have been demonstrated to have non-specific NRF2 inhibition, possibly through stimulating polyubiquitination of NRF2.107 Direct inhibitors of NRF2 via the KEAP1 binding domains are also in development.106,110

 

Conclusion

Next generation sequencing has provided unprecedented insight into the biology of RCC. The field has moved dramatically from the discovery of the VHL gene by genetic linkage analysis to the simultaneous and comprehensive analysis of genomic, transcriptomic, metabolomics, and proteomic data across multiple regions of a single tumor. Novel bioinformatic strategies and insightful experimental designs are revealing new molecular profiles of tumors and the tumor microenvironment. The culmination of these technologies has already resulted in refined prognostic opportunities based on molecular, pathological, and clinical variables. We are now rapidly shifting towards developing and validating predictive models with the ultimate goal to deliver on precision medicine.

 

References

1. Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2018;68(6):394-424.

2. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2019. CA Cancer J Clin. 2019;69(1):7-34.

3. Howlader N, Noone A, Krapcho M, Miller D, Brest A, Yu M, et al. SEER Cancer Statistics Review, 1975-2016, National Cancer Institute. Bethesda, MD. National Cancer Institute. Bethesda, MD; 2019.

4. Choueiri TK, Motzer RJ. Systemic Therapy for Metastatic Renal-Cell Carcinoma. N Engl J Med. 2017;376(4):354-66.

5. Barata PC, Rini BI. Treatment of renal cell carcinoma: Current status and future directions. CA Cancer J Clin. 2017;67(6):507-24

6. Network NCC. Kidney Cancer Version 1.2020 2019 [updated 6/7/2019] Available from: https://www.nccn.org/professionals/physician_gls/pdf/kidney.pdf.

7. Sheng IY, Rini BI. Immunotherapy for renal cell carcinoma. Expert Opin Biol Ther. 2019:1-9.

8. Lopez-Beltran A, Scarpelli M, Montironi R, Kirkali Z. 2004 WHO Classification of the Renal Tumors of the Adults. European Urology. 2006;49(5):798-805.

9. Srigley JR, Delahunt B, Eble JN, Egevad L, Epstein JI, Grignon D, et al. The International Society of Urological Pathology (ISUP) Vancouver Classification of Renal Neoplasia. Am J Surg Pathol. 2013;37(10):1469-89.

10. Moch H, Cubilla AL, Humphrey PA, Reuter VE, Ulbright TM. The 2016 WHO classification of tumours of the urinary system and male genital organs—part A: renal, penile, and testicular tumours. European Urology. 2016;70(1):93-105.

11. Nielsen SM, Rhodes L, Blanco I, Chung WK, Eng C, Maher ER, et al. Von Hippel-Lindau Disease: Genetics and Role of Genetic Counseling in a Multiple Neoplasia Syndrome. J Clin Oncol. 2016;34(18):2172-81.

12. Seizinger BR, Rouleau GA, Ozelius LJ, Lane AH, Farmer GE, Lamiell JM, et al. Von Hippel-Lindau disease maps to the region of chromosome 3 associated with renal cell carcinoma. Nature. 1988;332(6161): 268-9.

13. Hosoe S, Brauch H, Latif F, Glenn G, Daniel G, Bale S, et al. Localization of the von Hippel-Lindau disease gene to a small region of chromosome 3. Genomics. 1990;8(4):634-40.

14. Latif F, Tory K, Gnarra J, Yao M, Duh FM, Orcutt ML, et al. Identification of the von Hippel-Lindau disease tumor suppressor gene. Science. 1993;260(5112):1317-20.

15. Gnarra JR, Tory K, Weng Y, Schmidt L, Wei MH, Li H, et al. Mutations of the VHL tumour suppressor gene in renal carcinoma. Nat Genet. 1994;7(1):85-90.

16. Nickerson ML, Jaeger E, Shi Y, Durocher JA, Mahurkar S, Zaridze D, et al. Improved identification of von Hippel-Lindau gene alterations in clear cell renal tumors. Clinical Cancer Research. 2008;14(15):4726-34.

17. Mitchell TJ, Turajlic S, Rowan A, Nicol D, Farmery JHR, O’Brien T, et al. Timing the Landmark Events in the Evolution of Clear Cell Renal Cell Cancer: TRACERx Renal. Cell. 2018;173(3):611-23 e17.

18. Haase VH, Glickman JN, Socolovsky M, Jaenisch R. Vascular tumors in livers with targeted inactivation of the von Hippel-Lindau tumor suppressor. Proc Natl Acad Sci U S A. 2001;98(4):1583-8.

19. Wang SS, Gu YF, Wolff N, Stefanius K, Christie A, Dey A, et al. Bap1 is essential for kidney function and cooperates with Vhl in renal tumorigenesis. Proc Natl Acad Sci U S A. 2014;111(46):16538-43.

20. Mandriota SJ, Turner KJ, Davies DR, Murray PG, Morgan NV, Sowter HM, et al. HIF activation identifies early lesions in VHL kidneys: evidence for site-specific tumor suppressor function in the nephron. Cancer Cell. 2002;1(5):459-68.

21. Shen C, Kaelin WG, Jr. The VHL/HIF axis in clear cell renal carcinoma. Semin Cancer Biol. 2013;23(1):18-25.

22. Chen W, Hill H, Christie A, Kim MS, Holloman E, Pavia-Jimenez A, et al. Targeting renal cell carcinoma with a HIF-2 antagonist. Nature. 2016;539(7627):112-7.

23. Sato Y, Yoshizato T, Shiraishi Y, Maekawa S, Okuno Y, Kamura T, et al. Integrated molecular analysis of clear-cell renal cell carcinoma. Nat Genet. 2013;45(8):860-7.

24. Ricketts CJ, De Cubas AA, Fan H, Smith CC, Lang M, Reznik E, et al. The Cancer Genome Atlas Comprehensive Molecular Characterization of Renal Cell Carcinoma. Cell Reports. 2018;23(1):313-26.e5.

25. Brugarolas J. Molecular genetics of clear-cell renal cell carcinoma. J Clin Oncol. 2014;32(18):1968-76.

26. van Haaften G, Dalgliesh GL, Davies H, Chen L, Bignell G, Greenman C, et al. Somatic mutations of the histone H3K27 demethylase gene UTX in human cancer. Nat Genet. 2009;41(5):521-3.

27. Dalgliesh GL, Furge K, Greenman C, Chen L, Bignell G, Butler A, et al. Systematic sequencing of renal carcinoma reveals inactivation of histone modifying genes. Nature. 2010;463(7279):360-3.

28. Varela I, Tarpey P, Raine K, Huang D, Ong CK, Stephens P, et al. Exome sequencing identifies frequent mutation of the SWI/SNF complex gene PBRM1 in renal carcinoma. Nature. 2011;469(7331):539-42.

29. Peña-Llopis S, Vega-Rubín-de-Celis S, Liao A, Leng N, Pavía-Jiménez A, Wang S, et al. BAP1 loss defines a new class of renal cell carcinoma. Nature Genetics. 2012;44:751.

30. de Cubas AA, Rathmell WK. Epigenetic modifiers: activities in renal cell carcinoma. Nature Reviews Urology. 2018;15(10):599-614.

31. Kapur P, Peña-Llopis S, Christie A, Zhrebker L, Pavía-Jiménez A, Rathmell WK, et al. Effects on survival of BAP1 and PBRM1 mutations in sporadic clear-cell renal-cell carcinoma: a retrospective analysis with independent validation. The Lancet Oncology. 2013;14(2):159-67.

32. Pena-Llopis S, Christie A, Xie XJ, Brugarolas J. Cooperation and antagonism among cancer genes: the renal cancer paradigm. Cancer Res. 2013;73(14):4173-9.

33. Cancer Genome Atlas Research N. Comprehensive molecular characterization of clear cell renal cell carcinoma. Nature. 2013;499(7456): 43-9.

34. Joseph RW, Kapur P, Serie DJ, Parasramka M, Ho TH, Cheville JC, et al. Clear Cell Renal Cell Carcinoma Subtypes Identified by BAP1 and PBRM1 Expression. J Urol. 2016;195(1):180-7.

35. Joseph RW, Kapur P, Serie DJ, Eckel-Passow JE, Parasramka M, Ho T, et al. Loss of BAP1 protein expression is an independent marker of poor prognosis in patients with low-risk clear cell renal cell carcinoma. Cancer. 2014;120(7):1059-67.

36. Kapur P, Christie A, Raman JD, Then MT, Nuhn P, Buchner A, et al. BAP1 immunohistochemistry predicts outcomes in a multi-institutional cohort with clear cell renal cell carcinoma. J Urol. 2014;191(3): 603-10.

37. Gu YF, Cohn S, Christie A, McKenzie T, Wolff N, Do QN, et al. Modeling Renal Cell Carcinoma in Mice: Bap1 and Pbrm1 Inactivation Drive Tumor Grade. Cancer Discov. 2017;7(8):900-17.

38. Hakimi AA, Chen YB, Wren J, Gonen M, Abdel-Wahab O, Heguy A, et al. Clinical and pathologic impact of select chromatin-modulating tumor suppressors in clear cell renal cell carcinoma. Eur Urol. 2013;63 (5):848-54.

39. Carlo MI, Manley B, Patil S, Woo KM, Coskey DT, Redzematovic A, et al. Genomic Alterations and Outcomes with VEGF-Targeted Therapy in Patients with Clear Cell Renal Cell Carcinoma. Kidney Cancer. 2017;1(1):49-56.

40. Voss MH, Reising A, Cheng Y, Patel P, Marker M, Kuo F, et al. Genomically annotated risk model for advanced renal-cell carcinoma: a retrospective cohort study. The Lancet Oncology. 2018;19(12):1688-98.

41. Espana-Agusti J, Warren A, Chew SK, Adams DJ, Matakidou A. Loss of PBRM1 rescues VHL dependent replication stress to promote renal carcinogenesis. Nat Commun. 2017;8(1):2026.

42. Nargund AM, Pham CG, Dong Y, Wang PI, Osmangeyoglu HU, Xie Y, et al. The SWI/SNF Protein PBRM1 Restrains VHL-Loss-Driven Clear Cell Renal Cell Carcinoma. Cell Rep. 2017;18(12):2893-906.

43. Beroukhim R, Brunet JP, Di Napoli A, Mertz KD, Seeley A, Pires MM, et al. Patterns of gene expression and copy-number alterations in von-hippel lindau disease-associated and sporadic clear cell carcinoma of the kidney. Cancer Res. 2009;69(11):4674-81.

44. Chen M, Ye Y, Yang H, Tamboli P, Matin S, Tannir NM, et al. Genome-wide profiling of chromosomal alterations in renal cell carcinoma using high-density single nucleotide polymorphism arrays. Int J Cancer. 2009;125(10):2342-8.

45. Kucejova B, Pena-Llopis S, Yamasaki T, Sivanand S, Tran TA, Alexander S, et al. Interplay between pVHL and mTORC1 pathways in clear-cell renal cell carcinoma. Molecular Cancer Research: 2011;9(9): 1255-65.

46. Linehan WM, Ricketts CJ. The Cancer Genome Atlas of renal cell carcinoma: findings and clinical implications. Nature Reviews Urology. 2019.

47. Durinck S, Stawiski EW, Pavia-Jimenez A, Modrusan Z, Kapur P, Jaiswal BS, et al. Spectrum of diverse genomic alterations define non-clear cell renal carcinoma subtypes. Nat Genet. 2015;47(1):13-21.

48. Linehan WM, Spellman PT, Ricketts CJ, Creighton CJ, Fei SS, Davis C, et al. Comprehensive Molecular Characterization of Papillary Renal-Cell Carcinoma. N Engl J Med. 2016;374(2):135-45.

49. Ooi A, Wong J-C, Petillo D, Roossien D, Perrier-Trudova V, Whitten D, et al. An antioxidant response phenotype shared between hereditary and sporadic type 2 papillary renal cell carcinoma. Cancer Cell. 2011; 20(4):511-23.

50. Ooi A, Dykema K, Ansari A, Petillo D, Snider J, Kahnoski R, et al. CUL3 and NRF2 mutations confer an NRF2 activation phenotype in a sporadic form of papillary renal cell carcinoma. Cancer Research. 2013;73(7):2044-51.

51. Brunelli M, Eble JN, Zhang S, Martignoni G, Delahunt B, Cheng L. Eosinophilic and classic chromophobe renal cell carcinomas have similar frequent losses of multiple chromosomes from among chromosomes 1, 2, 6, 10, and 17, and this pattern of genetic abnormality is not present in renal oncocytoma. Mod Pathol. 2005;18(2):161-9.

52. Davis CF, Ricketts CJ, Wang M, Yang L, Cherniack AD, Shen H, et al. The somatic genomic landscape of chromophobe renal cell carcinoma. Cancer Cell. 2014;26(3):319-30.

53. Chen F, Zhang Y, Senbabaoglu Y, Ciriello G, Yang L, Reznik E, et al. Multilevel Genomics-Based Taxonomy of Renal Cell Carcinoma. Cell Rep. 2016;14(10):2476-89.

54. Frank I, Blute ML, Cheville JC, Lohse CM, Weaver AL, Zincke H. An outcome prediction model for patients with clear cell renal cell carcinoma treated with radical nephrectomy based on tumor stage, size, grade and necrosis: the SSIGN score. J Urol. 2002;168(6):2395-400.

55. Zisman A, Pantuck AJ, Wieder J, Chao DH, Dorey F, Said JW, et al. Risk group assessment and clinical outcome algorithm to predict the natural history of patients with surgically resected renal cell carcinoma. J Clin Oncol. 2002;20(23):4559-66.

56. Correa AF, Jegede O, Haas NB, Flaherty KT, Pins MR, Messing EM, et al. Predicting Renal Cancer Recurrence: Defining Limitations of Existing Prognostic Models With Prospective Trial-Based Validation. J Clin Oncol. 2019;37(23):2062-71.

57. Haas NB, Manola J, Uzzo RG, Flaherty KT, Wood CG, Kane C, et al. Adjuvant sunitinib or sorafenib for high-risk, non-metastatic renal-cell carcinoma (ECOG-ACRIN E2805): a double-blind, placebo-controlled, randomised, phase 3 trial. The Lancet. 2016;387(10032):2008-16.

58. Motzer RJ, Bacik J, Murphy BA, Russo P, Mazumdar M. Interferon-alfa as a comparative treatment for clinical trials of new therapies against advanced renal cell carcinoma. J Clin Oncol. 2002;20(1):289-96.

59. Heng DYC, Xie W, Regan MM, Warren MA, Golshayan AR, Sahi C, et al. Prognostic Factors for Overall Survival in Patients With Metastatic Renal Cell Carcinoma Treated With Vascular Endothelial Growth Factor–Targeted Agents: Results From a Large, Multicenter Study. J Clin Oncol. 2009;27(34):5794-9.

60. Ho TH, Kapur P, Joseph RW, Serie DJ, Eckel-Passow JE, Tong P, et al. Loss of histone H3 lysine 36 trimethylation is associated with an increased risk of renal cell carcinoma-specific death. Mod Pathol. 2016; 29(1):34-42.

61. Brannon AR, Reddy A, Seiler M, Arreola A, Moore DT, Pruthi RS, et al. Molecular Stratification of Clear Cell Renal Cell Carcinoma by Consensus Clustering Reveals Distinct Subtypes and Survival Patterns. Genes Cancer. 2010;1(2):152-63.

62. Brooks SA, Brannon AR, Parker JS, Fisher JC, Sen O, Kattan MW, et al. ClearCode34: A prognostic risk predictor for localized clear cell renal cell carcinoma. European Urology. 2014;66(1):77-84.

63. de Velasco G, Culhane AC, Fay AP, Hakimi AA, Voss MH, Tannir NM, et al. Molecular Subtypes Improve Prognostic Value of International Metastatic Renal Cell Carcinoma Database Consortium Prognostic Model. The Oncologist. 2017;22(3):286-92.

64. Hakimi AA, Voss MH, Kuo F, Sanchez A, Liu M, Nixon BG, et al. Transcriptomic Profiling of the Tumor Microenvironment Reveals Distinct Subgroups of Clear Cell Renal Cell Cancer: Data from a Randomized Phase III Trial. Cancer Discov. 2019;9(4):510-25.

65. Motzer RJ, Tannir NM, McDermott DF, Arén Frontera O, Melichar B, Choueiri TK, et al. Nivolumab plus Ipilimumab versus Sunitinib in Advanced Renal-Cell Carcinoma. N Engl J Med. 2018;378(14):1277-90.

66. Rini BI, Plimack ER, Stus V, Gafanov R, Hawkins R, Nosov D, et al. Pembrolizumab plus Axitinib versus Sunitinib for Advanced Renal-Cell Carcinoma. N Engl J Med. 2019;380(12):1116-27.

67. McDermott DF, Huseni MA, Atkins MB, Motzer RJ, Rini BI, Escudier B, et al. Clinical activity and molecular correlates of response to atezolizumab alone or in combination with bevacizumab versus sunitinib in renal cell carcinoma. Nature Medicine. 2018;24(6):749-57.

68. Barbie DA, Tamayo P, Boehm JS, Kim SY, Moody SE, Dunn IF, et al. Systematic RNA interference reveals that oncogenic KRAS-driven cancers require TBK1. Nature. 2009;462(7269):108.

69. Wang T, Lu R, Kapur P, Jaiswal BS, Hannan R, Zhang Z, et al.

An Empirical Approach Leveraging Tumorgrafts to Dissect the Tumor Microenvironment in Renal Cell Carcinoma Identifies Missing Link to Prognostic Inflammatory Factors. Cancer Discov. 2018;8(9): 1142-55.

70. Havel JJ, Chowell D, Chan TA. The evolving landscape of biomarkers for checkpoint inhibitor immunotherapy. Nature Reviews Cancer. 2019;19(3):133-50.

71. Gerlinger M, Rowan AJ, Horswell S, Math M, Larkin J, Endesfelder D, et al. Intratumor heterogeneity and branched evolution revealed by multiregion sequencing. N Engl J Med. 2012; 366(10):883-92.

72. Gerlinger M, Horswell S, Larkin J, Rowan AJ, Salm MP, Varela I, et al. Genomic architecture and evolution of clear cell renal cell carcinomas defined by multiregion sequencing. Nat Genet. 2014;46(3):225-33.

73. TRACERx Renal: tracking renal cancer evolution through therapy. Nature Reviews Urology. 2017;14(10):575-6.

74. Turajlic S, Xu H, Litchfield K, Rowan A, Horswell S, Chambers T, et al. Deterministic Evolutionary Trajectories Influence Primary Tumor Growth: TRACERx Renal. Cell. 2018;173(3):595-610 e11.

75. Turajlic S, Xu H, Litchfield K, Rowan A, Chambers T, Lopez JI, et al. Tracking Cancer Evolution Reveals Constrained Routes to Metastases: TRACERx Renal. Cell. 2018;173(3):581-94 e12.

76. Sierant MC, Choi J. Single-Cell Ssequencing in Cancer: Recent Applications to Immunogenomics and Multiomics Tools. Genomics Inform. 2018;16(4):e17-e.

77. Schmidt L, Duh F-M, Chen F, Kishida T, Glenn G, Choyke P, et al. Germline and somatic mutations in the tyrosine kinase domain of the MET proto-oncogene in papillary renal carcinomas. Nature Genetics. 1997;16(1):68.

78. Schmidt L, Junker K, Nakaigawa N, Kinjerski T, Weirich G, Miller M, et al. Novel mutations of the MET proto-oncogene in papillary renal carcinomas. Oncogene. 1999;18(14):2343.

79. Choueiri TK, Escudier B, Powles T, Tannir NM, Mainwaring PN, Rini BI, et al. Cabozantinib versus everolimus in advanced renal cell carcinoma (METEOR): final results from a randomised, open-label, phase 3 trial. The Lancet Oncology. 2016;17(7):917-27.

80. Martínez Chanzá N, Xie W, Asim Bilen M, Dzimitrowicz H, Burkart J, Geynisman DM, et al. Cabozantinib in advanced non-clear-cell renal cell carcinoma: a multicentre, retrospective, cohort study. The Lancet Oncology. 2019;20(4):581-90.

81. Scheuermann TH, Tomchick DR, Machius M, Guo Y, Bruick RK, Gardner KH. Artificial ligand binding within the HIF2α PAS-B domain of the HIF2 transcription factor. Proceedings of the National Academy of Sciences. 2009;106(2):450-5.

82. Scheuermann TH, Li Q, Ma HW, Key J, Zhang L, Chen R, et al. Allosteric inhibition of hypoxia inducible factor-2 with small molecules. Nat Chem Biol. 2013;9(4):271-6.

83. Wallace EM, Rizzi JP, Han G, Wehn PM, Cao Z, Du X, et al. A Small-Molecule Antagonist of HIF2alpha Is Efficacious in Preclinical Models of Renal Cell Carcinoma. Cancer Res. 2016;76(18):5491-500.

84. Cho H, Du X, Rizzi JP, Liberzon E, Chakraborty AA, Gao W, et al. On-target efficacy of a HIF-2alpha antagonist in preclinical kidney cancer models. Nature. 2016;539(7627):107-11.

85. Courtney KD, Ma Y, Diaz LA, Christie A, Xie Z, Woolford L, et al. HIF-2 Complex Dissociation, Target Inhibition, and Acquired Resistance with First-in-cClass HIF-2 Inhibitor in Patients. Clinical Cancer Research. [published online ahead of print November 19, 2019]

86. Courtney KD, Infante JR, Lam ET, Figlin RA, Rini BI, Brugarolas J, et al. Phase I Dose-Escalation Trial of PT2385, a First-in-Class Hypoxia-Inducible Factor-2alpha Antagonist in Patients With Previously Treated Advanced Clear Cell Renal Cell Carcinoma. J Clin Oncol. 2018;36(9): 867-74.

87. Jonasch E, Plimack ER, Bauer T, Merchan JR, Papadopoulos KP, McDermott DF, et al. 911PDA first-in-human phase I/II trial of the oral HIF-2a inhibitor PT2977 in patients with advanced RCC. Annals of Oncology. 2019;30(Supplement_5).

88. Kaelin WG, Jr. Synthetic lethality: a framework for the development of wiser cancer therapeutics. Genome Med. 2009;1(10):99.

89. Turcotte S, Chan DA, Sutphin PD, Hay MP, Denny WA, Giaccia AJ. A Molecule Targeting VHL-Deficient Renal Cell Carcinoma that Induces Autophagy. Cancer Cell. 2008;14(1):90-102.

90. Chan DA, Sutphin PD, Nguyen P, Turcotte S, Lai EW, Banh A, et al. Targeting GLUT1 and the Warburg Effect in Renal Cell Carcinoma by Chemical Synthetic Lethality. Science Translational Medicine. 2011;3(94): 94ra70-94ra70.

91. Woldemichael GM, Turbyville TJ, Vasselli JR, Linehan WM, McMahon JB. Lack of a functional VHL gene product sensitizes renal cell carcinoma cells to the apoptotic effects of the protein synthesis inhibitor verrucarin A. Neoplasia. 2012;14(8):771-7.

92.Woldemichael GM, Turbyville TJ, Linehan WM, McMahon JB. Carminomycin I Is an Apoptosis Inducer That Targets the Golgi Complex in Clear Cell Renal Carcinoma Cells. Cancer Research. 2011; 71(1):134-42.

93. Wolff NC, Pavía-Jiménez A, Tcheuyap VT, Alexander S, Vishwanath M, Christie A, et al. High-throughput simultaneous screen and counterscreen identifies homoharringtonine as synthetic lethal with von Hippel-Lindau loss in renal cell carcinoma. Oncotarget. 2015;6(19): 16951-62.

94. Thompson JM, Nguyen QH, Singh M, Pavesic MW, Nesterenko I, Nelson LJ, et al. Rho-associated kinase 1 inhibition is synthetically lethal with von Hippel-Lindau deficiency in clear cell renal cell carcinoma. Oncogene. 2017;36(8):1080-9.

95. Bommi-Reddy A, Almeciga I, Sawyer J, Geisen C, Li W, Harlow E, et al. Kinase requirements in human cells: III. Altered kinase requirements in VHL-/- cancer cells detected in a pilot synthetic lethal screen. Proc Natl Acad Sci USA. 2008;105(43):16484-9.

96. Chakraborty AA, Nakamura E, Qi J, Creech A, Jaffe JD, Paulk J, et al. HIF activation causes synthetic lethality between the VHL tumor suppressor and the EZH1 histone methyltransferase. Science Translational Medicine. 2017;9(398).

97. LaFave LM, Beguelin W, Koche R, Teater M, Spitzer B, Chramiec A, et al. Loss of BAP1 function leads to EZH2-dependent transformation. Nat Med. 2015;21(11):1344-9.

98. Dey A, Seshasayee D, Noubade R, French DM, Liu J, Chaurushiya MS, et al. Loss of the Tumor Suppressor BAP1 Causes Myeloid Transformation. Science. 2012;337(6101):1541-6.

99. Ho TH, Kapur P, Eckel-Passow JE, Christie A, Joseph RW, Serie DJ, et al. Multicenter Validation of Enhancer of Zeste Homolog 2 Expression as an Independent Prognostic Marker in Localized Clear Cell Renal Cell Carcinoma. J Clin Oncol. 2017;35(32):3706-13.

100. Sun C, Zhao C, Li S, Wang J, Zhou Q, Sun J, et al. EZH2 Expression is increased in BAP1-mutant renal clear cell carcinoma and is related to poor prognosis. J Cancer. 2018;9(20):3787-96.

101. Adelaiye-Ogala R, Budka J, Damayanti NP, Arrington J, Ferris M, Hsu CC, et al. EZH2 Modifies Sunitinib Resistance in Renal Cell Carcinoma by Kinome Reprogramming. Cancer Res. 2017;77(23):6651-66.

102. O’Leary B, Finn RS, Turner NC. Treating cancer with selective CDK4/6 inhibitors. Nature Reviews Clinical Oncology. 2016;13:417.

103. Logan JE, Mostofizadeh N, Desai AJ, E VONE, Conklin D, Konkankit V, et al. PD-0332991, a potent and selective inhibitor of cyclin-dependent kinase 4/6, demonstrates inhibition of proliferation in renal cell carcinoma at nanomolar concentrations and molecular markers predict for sensitivity. Anticancer Res. 2013;33(8):2997-3004.

104. Small J, Washburn E, Millington K, Zhu J, Holder SL. The addition of abemaciclib to sunitinib induces regression of renal cell carcinoma xenograft tumors. Oncotarget. 2017;8(56):95116-34.

105. Nicholson HE, Tariq Z, Housden BE, Jennings RB, Stransky LA, Perrimon N, et al. HIF-independent synthetic lethality between CDK4/6 inhibition and VHL loss across species. Sci Signal. 2019;12(601).

106. JSingh, A., et al. (2016). "Small Molecule Inhibitor of NRF2 Selectively Intervenes Therapeutic Resistance in KEAP1-Deficient NSCLC Tumors." ACS Chemical Biology. 11(11): 3214-3225.

107. Telkoparan-Akillilar P, Suzen S, Saso L. Pharmacological Applications of Nrf2 Inhibitors as Potential Antineoplastic Drugs. International Journal of Molecular Sciences. 2019;20(8):2025.

108. Fabrizio FP, Costantini M, Copetti M, la Torre A, Sparaneo A, Fontana A, et al. Keap1/Nrf2 pathway in kidney cancer: frequent methylation of KEAP1 gene promoter in clear renal cell carcinoma. Oncotarget. 2017;8(7):11187-98.

109. Ji S, Xiong Y, Zhao X, Liu Y, Yu LQ. Effect of the Nrf2-ARE signaling pathway on biological characteristics and sensitivity to sunitinib in renal cell carcinoma. Oncol Lett. 2019;17(6):5175-86.

110. Jiang Z-Y, Lu M-C, You Q-D. Nuclear Factor Erythroid 2-Related Factor 2 (Nrf2) Inhibition: An Emerging Strategy in Cancer Therapy. Journal of Medicinal Chemistry. 2019;62(8):3840-56  KCJ

 

Departments of Internal Medicine1 Kidney Cancer Program2, and Department of Urology3,
Simmons Comprehensive Cancer Center, University of Texas Southwestern Medical Center,
Dallas TX, 75390

The first tumor suppressor discovered in ccRCC was the von Hippel-Lindau (VHL) gene. VHL syndrome is an autosomal dominant disease caused by germline mutations in the VHL gene. This syndrome results in development of numerous ccRCCs among other manifestations.11 Genetic linkage analysis of affected kindreds located the responsible gene to the chromosome region 3p25-26 in the late 1980s.12,13 Latif and colleagues were the first to identify the VHL gene,14 and the following year VHL was also shown to be mutated in sporadic ccRCC.15 Nearly two decades later, Nickerson and colleagues evaluated a cohort of 205 ccRCC samples, the largest cohort at that time, for aberrations in the VHL gene through targeted sequencing. They identified non-silent somatic mutations with a prevalence of 82% and VHL promoter hypermethylation in an additional 8.3% of tumors, for a total of 90% of tumors.16 In sporadic ccRCC, the initiating event is thought to be loss of 3p through a chromothripsis event.17 This is likely followed by mutation or epigenetic silencing of the second allele. While VHL loss is nearly universal in ccRCC, VHL inactivation has also been shown in preneoplastic cysts, and mice with VHL disruption in the kidneys do not develop ccRCC.18-20 Thus, while VHL loss is an important initiating step in the development of ccRCC, additional driver mutations were suspected.

     The VHL protein forms a complex with Elongin B, Elongin C, CUL2, and RBX1 that serves as an E3 ubiquitin ligase which, at normal oxygen tension, acts on the alpha unit of hypoxia inducible factor (HIF) transcription factors, leading to their degradation.21 Loss of VHL in ccRCC results in constitutive activation of HIF and expression of its target genes.21,22 Interestingly, mutations in TCEB1 (encoding Elongin C) have been reported in up to 5% of ccRCC cases and are associated with loss of heterozygosity of chromosome 8, which contains TCEB1.23 In addition, mutations in CUL2 were found in up to 1% of ccRCCs.24 Mutations in TCEB1 and CUL2 tend to be mutually exclusive with VHL mutations and likely represent other mechanisms to disrupt VHL function.25

The first large scale sequencing reports in ccRCC came out in 2009, and demonstrated frequent mutations in the chromatin remodeling genes SETD2, KDM6A (also known as UTX), KDM5C (also known as JARID1C), and MLL2.26,27 These reports were limited to sequencing a panel of ~3,500 genes, a small subset of the entire human exome, and the genes identified were mutated in up to 15% of tumors. In the subsequent two years, whole exome sequencing (WES) allowed the discovery of two major drivers of ccRCC; polybromo 1 (PBRM1) and the BRCA1 associated protein-1 (BAP1). Varela and colleagues performed WES of seven ccRCC and matching normal samples and identified truncating mutations in PBRM1 in four tumors.28 Non-silent mutations in PBRM1 were subsequently identified in 88 of 221 (39.8%) ccRCC cases, making it the second most common mutation in ccRCC.28 The following year, WES of paired tumor and patient-derived xenografts (PDX, also called tumorgrafts) identified BAP1 as an important driver of ccRCC.29 By incorporating tumorgrafts, our studies allowed for accurate calls of mutant allele frequencies (MAF) thereby confidently nominating putative two-hit tumor suppressor genes. In tumorgraft models, BAP1 was the only gene in addition to VHL and PBRM1 to demonstrate a MAF of ~1 (signifying absence of the wild-type allele). Subsequent BAP1 targeted sequencing studies identified mutations in 24 out of 176 (14%) largely primary ccRCC samples.29

     PBRM1 and BAP1 are both involved in chromatin modification and epigenetic regulation of gene expression. PBRM1 encodes BAF180, a component of the switching defective/sucrose nonfermenting (SWI/SNF) family of nucleosome remodeling complexes, which is thought to mediate recruitment of the complex to specific nucleosomes through recognition of acetylation patterns.25,30 The BAP1 protein is a nuclear-localized deubiquitinase which acts to deubiquitylate H2AK119ub1 to reverse polycomb-mediated gene repression.30 These processes likely regulate different gene subsets, and we discovered that BAP1 and PBRM1-deficient tumors can be distinguished by their gene expression signature.31

     Interestingly, both BAP1 and PBRM1 are located on chromosome 3p, within a 43 Mb region that is frequently lost in ccRCC, which also includes VHL and SETD2.17,32 Notably, BAP1 and PBRM1 mutations were found to anticorrelate in ccRCC and the prevalence of combined BAP1/ PBRM1 deficient ccRCC is ~1-2%, less than the ~5% rate expected given the rates of PBRM1 (~45%) and BAP1 (~12%) mutations in ccRCC.29,31-34 While mutual exclusivity often indicates that the encoded proteins are in the same pathway (and reflects the low selective pressure to have them simultaneously mutated), the finding that loss of BAP1 and PBRM1 results in distinct gene expression signatures, suggested that this was not the case for these genes.29,31 Furthermore, BAP1 mutant tumors exhibited higher nucleolar grade, more aggressive histology, and demonstrated worsened RCC-specific survival.31,35-38 This finding was subsequently confirmed in metastatic RCC, as BAP1 mutations were independently associated with worsened overall survival.39,40 Together these findings led us to propose a model where following inactivation of VHL and 3p loss, inactivation of the remaining copy of BAP1 caused aggressive ccRCC and inactivation of PBRM1, less aggressive ccRCC.25 Using methods we developed,29 this model was subsequently revised by the TRACERx consortium to show that the first event in sporadic tumors is likely the loss of 3p.17

     These discoveries explain why germline VHL mutations cause kidney cancer in humans, but not in mice. As it turns out, we found that in mice the Bap1 and Pbrm1 genes are on a different chromosome than Vhl and as such, loss of the chromosome arm containing the Vhl gene in the mouse would have no effect on Bap1 or Pbrm1 genes. In contrast, loss of 3p in humans results not only in the loss of a VHL allele, but also one allele of both BAP1 and PBRM1. Therefore, the arrangement of tumor suppressor genes in the genome likely accounts for differences in tumor predisposition across species.19,25,37

The development of NGS and improved bioinformatics tools allowed the collection and integration of data from WES, copy-number analyses, DNA-methylation analyses, and messenger RNA and microRNA sequencing, from individual samples at a large scale.23,33 Through integrated analyses of ~400 ccRCC samples, The Cancer Genome Atlas (TCGA) consortium not only confirmed previous findings, but significantly expanded on them through the identification of altered subnetworks. Initial genome-wide copy number analysis, using next-generation sequencing technologies, identified 3p and 14q loss, as well as 5q gain as the three most common somatic copy number alterations (SCNA) in ccRCC.43,44 These were verified in a larger TCGA cohort - 3p loss (91%), 14q loss (45%) and 5q gain (67%).33 The most frequently mutated genes involved chromatin remodeling pathways. These included genes encoding the SWI/SNF family proteins PBRM1, SMARCA4 and ARID1A (47.1%), the histone methyltransferases SETD2 and MLL (23.8%), and the histone deubiquitinase BAP1 (12.1%). Alterations in PTEN, TSC1, and other components of the PI3K/AKT/mTOR pathway were observed in 28% of ccRCCs.33,45,46 Other findings of significance included loss of one CDKN2A allele in 16.2% of ccRCCs (mostly through deletion of 9p21.3), and mutation of TP53 in 2.6% of cases.24,46

     Analysis of chRCC revealed distinct molecular alterations defining this subgroup. chRCC can be classified histologically into classical and eosinophilic subtypes, the latter characterized by abundant eosinophilic cytoplasm and densely packed mitochondria.51 Analysis of the TCGA cohort of 66 chRCC identified molecular distinctions between these two subgroups.51,52 All 47 cases of classical chRCC in the TCGA cohort demonstrated characteristic chromosomal losses of chromosomes 1, 2, 6, 10, 13, and 17 whereas only half (10/19) of the eosinophil variants did.47,52 Studies from TCGA and our own group found that TP53 mutations were significantly enriched (P = 2.3E-5) in the classic chRCC subtype.47,52 The two most frequently mutated genes across chRCC variants are TP53 (31.1%) and PTEN (9%).47 Mutually exclusive mutations in genes of the PI3K/AKT/mTOR pathway were observed in 23% of cases.47,52 Interestingly, analysis of mitochondrial DNA in chRCC revealed recurring mutations in genes encoding components of complex 1 of the electron transport chain (ETC), although these mutations were not associated with changes in the expression of genes implicated in oxidative phosphorylation.52 chRCC are particularly enriched in mutations involving metabolic genes, including deleterious mutations in PDHB (which encodes a critical component of the pyruvate dehydrogenase complex) and PRKAG2 (encoding one of three subunits of AMP-Kinase (AMPK)). These findings reinforce the long-standing implication that metabolic derangements in RCC can contribute to oncogenesis.

     Integrated transcriptome and protein expression analysis (using reverse protein phase arrays) in TCGA revealed distinct metabolic patterns both across and within histological subtypes of RCC.24,5 ccRCC was characterized by overexpression of glycolytic and fatty acid synthesis enzymes as well as suppression of Krebs and ETC programs. This contrasts with pRCC and chRCC, which generally expressed intermediate and high levels of Krebs and ETC proteins, respectively.24 Decreased expression of AMPK (which inhibits fatty acid synthesis and mTOR), and increased ribose sugar metabolism, correlated with higher stage and worse prognosis in ccRCC.33 The CIMP subtype of pRCC demonstrated the highest level of ribose sugar metabolism across all RCC subtypes. A metabolically divergent subset of chRCC was identified, which demonstrated decreased levels of AMPK and ETC genes, increased ribose sugar metabolism, and demonstrated a profoundly worse outcome than metabolically traditional chRCC.24 This metabolically divergent subtype of chRCC lacked the typical chromosomal aberrations of chRCC, and four of the six displayed sarcomatoid de-differentiation. Taken as a whole, these landmark studies contributed tremendously to our understanding of the molecular biology of RCC, and the findings are driving the development of molecularly-based prognostic models across histological subtypes in RCC

Successful prognostication of patients is critical to both the clinical practice of oncology and research, as it influences treatment approaches and may be used to stratify patients in clinical trials. Classically, the most robust prognostic variables were related to histology (subtype, grade, tumor size), clinical features (performance status and pace of disease), or laboratory parameters. For localized disease treated with curative nephrectomy, the two main prognostic models are the UCLA Integrated Scoring System (UISS) and the Mayo Clinic Stage, Size, Grade, and Necrosis (SSIGN) score.54,55 The UISS score integrates pathological (tumor size and nuclear grade) with clinical (ECOG performance status) variables, whereas the SSIGN score utilizes strictly pathological variables. Despite the frequent use of the aforementioned models in patient prognostication, a recent analysis has called into question the predictive power of these nomograms.56 Utilizing prospectively collected data from the ASSURE trial,57 a randomized phase III placebo-controlled trial of sunitinib and sorafenib in the adjuvant setting, Correa and colleagues analyzed 8 recurrence prediction models including the SSIGN and UISS tools. They reported that the predictive power for each model’s predefined outcome, as measured by the co-occurrence index (C-index), fell considerably for all models in the dataset. However, among these models, the SSIGN score performed best with a C-index of 0.688 (95% CI: 0.686 – 0.689). Most other nomograms provided only marginal improvement relative to the TNM staging system.56 Integrating underlying genomic information with clinical and histological variables is one strategy to improve these models. We previously demonstrated that BAP1 and PBRM1 expression by immunohistochemistry (IHC) are highly correlated with mutation status (P = 3E-58 and 4E-23 respectively34 and that BAP1/PBRM1 expression by IHC are independent predictors of both disease-free survival and overall survival in the localized disease setting.31,34 PBRM1 and BAP1 expression status did not, however, add independent prognostic information to the SSIGN score in a multivariate analysis of nearly 1,500 cases of localized RCC.34 One possible explanation is that BAP1-mutant tumors are significantly more likely to demonstrate higher nuclear grade, stage, and necrosis than PBRM1-mutant tumors.34-37 Thus, despite progress, additional strategies to stratify patients in the localized disease setting remain needed.

     For advanced disease, the most commonly used prognostic models are the Memorial Sloan Kettering Cancer Center (MSKCC) risk model58 and the International Metastatic RCC Database Consortium (IMDC) risk model.59 Both models integrate clinical variables (time from diagnosis to systemic treatment and performance status) and laboratory variables including calcium and hemoglobin (as well as lactate dehydrogenase in the MSKCC model vs platelet and neutrophil counts in the IMDC model) but are agnostic to the molecular biology of the underlying disease. Many of the key driver mutations in RCC have been shown to have negative (BAP1, SETD2, PTEN, and TP53) and positive (PBRM1) prognostic significance,24,34,53,60 although it is important to interpret the prognostic implications of these mutations in the context of histology. For example, PBRM1 mutations which tend to associate with favorable prognosis in ccRCC were strongly associated with reduced survival in type I pRCC (p<0.0001).24 Regardless, retrospective analysis of two phase III trials in metastatic RCC, COMPARZ (first-line sunitinib vs pazopanib) and RECORD-3 (first-line sunitinib vs everolimus) identified PBRM1, BAP1, and TP53 as having independent prognostic value.40 Their prognostic significance persisted after multivariate analysis with the traditional variables in the original MSKCC model. This resulted in the development of a genomically-annotated MSKCC risk model which stratifies patients into four risk groups (favorable, good, intermediate, and poor) versus three in the original. The genomically-annotated model had a more balanced distribution across these groups and an improved ability to predict overall survival. The C-index of the original model was 0.567 (95% CI: 0.529 – 0.604) vs 0.637 (95% CI: 0.529 – 0.604) with the new model.40 While promising, this model will need to be validated in the prospective setting.

     Prognostic models have also been developed using gene expression data. One of the first such attempts in ccRCC was by Brannon and colleagues who collected transcriptomic data in 47 ccRCC primary tumor samples. Unsupervised clustering analyses revealed two subgroups, referred to as clear cell types A and B (ccA/B), which had distinctly different clinical outcomes with a median survival of 59 vs 36 months respectively in a validation cohort of 177 patients (P=0.004).61 A 34 gene panel (Clearcode34), which could delineate ccA and ccB subtypes, was found to better predict relapse-free survival and cancer-specific survival than both the UISS and SSIGN models in a cohort of nearly 500 patients with localized RCC.62 Similar findings were seen in a smaller metastatic cohort (54 patients), however, only ccB classification added prognostic value when incorporating the IMDC and MSKCC risk models.63

     More recently, cluster of cluster analysis including DNA copy number, mRNA, microRNA, DNA methylation, and protein expression data from the TCGA cohort revealed 9 unique clusters across all histological subtypes,53 three of which were enriched in histologically defined ccRCC (e.1, e.2, and e.3). These demonstrated significant differences in survival, with e.3 demonstrating the poorest prognosis and e.2 demonstrating the most favorable. Tumors in the e.3 cluster tended to have frequent loss of the CDKN2A gene, frequent BAP1 mutations, and overexpression of cell cycle and hypoxia-related genes.53 Additionally, e.3 was enriched in inflammatory gene expression relative to e.2, and expression of PDCD1 (encoding for PD1) and CTLA4 were independent predictors of poor prognosis across the ccRCC subtype.53 The e.2 subgroup, on the other hand, was found to demonstrate significant overlap with the previously described ccA subtype,53,61 which is distinguished by overexpression of angiogenesis-related genes such as such as FLT4, FLT1, and VEGFB.61,62,64 Thus, given that the most widely utilized treatments in metastatic RCC are immune checkpoint inhibitors, angiogenesis inhibitors, or combinations of the two,4,7,65,66 transcriptomic information regarding inflammation or angiogenesis may also have predictive potential.

     Unlike prognostic biomarkers which inform on the spontaneous trajectory of the disease, predictive biomarkers have the potential to inform on treatment selection, which may be tailored to the particular patient. Beyond the obvious benefit of selecting the most efficacious treatment, predictive biomarkers have the potential to reduce adverse events as well as cost from exposures to non-beneficial drugs. The use of gene expression profiles to predict treatment response was considered in an exploratory analysis of the IMmotion150 trial, a phase 2 study comparing the combination of atezolizumab (a PD-L1 inhibitor) and bevacizumab (a neutralizing VEGF antibody) (atezo+bev) to atezolizumab (atezo) or sunitinib monotherapy.67 McDermott and colleagues showed that predefined gene signatures of angiogenesis (Angio), immune T cell infiltration (Teff), and myeloid inflammation (Myeloid) may have predictive potential. They found that patients with the AngioHigh gene signature had an improved progression-free survival (PFS) compared to those with AngioLow with the angiogenesis inhibitor, sunitinib.67 Interestingly, PBRM1 mutations were found to be enriched in the AngioHigh subgroup. In line with these findings, a similar analysis of the COMPARZ trial evaluating two anti-angiogenic agents, sunitinib and pazopanib, found that high expression of angiogenesis genes was associated with improved response rates.64 Furthermore, tumors with PBRM1 mutations demonstrated significantly higher angiogenesis gene expression scores than those with BAP1 mutations.64 In the IMmotion150 trial, patients with the TeffHigh gene signature had improved PFS with atezo+bev compared to the sunitinib and the atezo arms. Intriguingly, when examining the effect of myeloid signature on outcomes within the TeffHigh population, a PFS advantage of atezo+bev over atezo was observed in the TeffHigh /MyeloidHigh subgroup (n=66) but not in the TeffHigh /MyeloidLow subgroup (n=66). This finding may identify a particular population of patients which would benefit from combination ICI and TKI therapy. However, this is a relatively small study and has not yet been reproduced in larger datasets.

     These studies highlight the potential role of the tumor microenvironment (TME) on clinical outcomes. Bioinformatics techniques can be leveraged to gather information about the TME from RNA extracted from tumors. One such approach is to utilize single sample gene set enrichment analysis (ssGSEA) in which signature gene panels attributed to particular cell types are utilized to disentangle a heterogeneous tumor.68 Our group expanded upon this method by exploiting patient-derived tumorgrafts (patient tumors implanted in mice), where the human TME is ultimately replaced by the host.69 By focusing on human RNA and subtracting from the patient tumor the transcriptome of the tumorgraft, one is left with the human TME transcriptome. Utilizing this dissection algorithm and a cutoff of 20-fold to distinguish TME vs tumor genes, an empirically derived TME (eTME) signature was obtained. Clustering analyses of the ccRCC TCGA (KIRC) tumors according to the eTME revealed two subgroups, an inflamed subtype (IS) and a non-inflamed subtype (NIS). Interestingly, the IS was enriched for BAP1 mutations (P = 7.7E-5) and demonstrated a worse prognosis compared to NIS. Furthermore, the eTME-IS subtype correlated with systemic inflammatory markers such as elevated platelet counts and decreased hemoglobin levels, as well as worse prognosis in three distinct cohorts totaling approximately 1,000 patients. This correlation draws a link between inflammatory subtypes of RCC and key prognostic variables in IMDC or MSKCC models. Notably, the presence of such variables is associated with intermediate/poor risk disease, which suggests that inflammatory tumors are particularly aggressive in patients. Interestingly, in the Checkmate 214 trial comparing the combination of nivolumab and ipilimumab to sunitinib, the intermediate/poor risk groups appeared to derive the most benefit from ICI,65 which is consistent with the predictive potential of the IS subtype. Thus, the combined use of NGS and bioinformatics has potential to predict responses to therapy and is an active area of investigation.70

One important disadvantage of the aforementioned studies is the use of a limited number of tumor samples per patient (typically just one), as this may fail to capture intratumoral heterogeneity (ITH). In one of the earliest attempts to measure ITH in ccRCC, Gerlinger and colleagues performed multiregional WES of two primary tumors and demonstrated that only 31% of the somatic mutations were ubiquitous amongst all sampled regions.71 Furthermore, the previously described “cc-A” and “cc-B” gene expression patterns could be identified in spatially distinct areas of an individual patient’s tumor, highlighting how ITH can confound efforts to establish effective prognostic models based on analyses of a single sample.71,72  This laid the groundwork for the TRAcking Cancer Evolution through therapy (Rx) (TRACERx) consortium, which prospectively collects tumor samples and performs multiregional sequencing, when possible, over time.73 In a recently published report, Turajlic and colleagues sequenced 1,206 primary tumor samples from 101 patients with the use of a 110 gene panel, allowing for an unprecedented view of the molecular diversity within a single tumor.74 Multiregional sampling allowed for the detection of clonal and subclonal somatic mutations. Thus, the prevalence of PBRM1 (55%), SETD2 (25%), BAP1 (19%) and other driver mutations could be more accurately calculated.74 Interestingly, while BAP1 and PBRM1 mutations could be identified in the same tumor, they were typically located in spatially distinct regions, consistent with previous reports that these mutations anticorrelate with one another and are found in different areas of the same tumors.29,32,34,74

     Genomic data obtained from spatially distinct regions provided the ability to assess the timing of mutations, and thus patterns of tumor evolution could be inferred (Figure 2). In the TRACERx studies, seven distinct patterns could be identified utilizing rule-based clustering. However, 36.6% (37/101) of cases could not be assigned an evolutionary subtype. Subtypes were assessed for nuclear grade, stage, microvascular invasion, genomic instability, and degree of ITH. The most aggressive subtype, based on the aforementioned parameters, was the “multiple clonal drivers” subtype, which contained truncal aberrations in two or more of the following: BAP1, PBRM1, SETD2, or PTEN.74 It is important to note that in the “multiple driver mutations” subtype, the temporal relationships of driver mutations were indistinguishable, which separates this group from subsets defined by subclonal (sequential) aberrations in driver genes. The “multiple clonal drivers” subtype includes BAP1/PBRM1-deficient tumors previously described which we and others noted to portend a poor prognosis.29,34,40 The “BAP1 driven” evolutionary subtype, characterized by truncal VHL and BAP1 mutations was also found to have aggressive histological characteristics, with decreased DFS and OS. Three PBRM1-driven subtypes were noted to have sequential loss of PBRM1 followed by either loss of SETD2, activation of PI3K signaling, or distinct SCNAs.74,75 Consistent with this, in 11/101 cases, PBRM1 loss was noted to precede SETD2 loss, but the converse was not seen.75 The least aggressive subtype was the VHL monodriver subtype, which likely represents sampling early in the disease course. When compared with “multiple clonal driver” and “BAP1 driven” subtypes, the three PBRM1 subtypes tended to have increased ITH, less aggressive biology, and demonstrated a more attenuated disease course.74,75 Conversely, the “multiple clonal driver,” “BAP-1 driven” and, perhaps unexpectedly, “VHL wild-type” evolutionary subtypes tended to demonstrate rapid progression to metastases.75 Decreasing ITH tended to correlate with a more aggressive disease course, consistent with the notion that an aggressive clone would “outcompete” other clones. On the other hand, evolutionary subtypes with greater degrees of ITH demonstrated a more attenuated disease course. Tumors with a high degree of ITH may, however, also harbor a diverse reservoir of cancer cells which are resistant to therapy, providing a potential explanation for the mixed responses sometimes exhibited by patients.74 Capturing the full degree of ITH may be challenging, as on average at least 7 biopsies are required to detect over 75% of variants.74 Though challenging to implement clinically, techniques are evolving that allow the dissociation of the whole tumor followed by single-cell sequencing to provide molecular information at the cellular level.76 Other approaches include lysing large quantities of tumor material with deep sequencing in an attempt to comprehensively catalogue all mutations. Future prognostic and predictive models will likely need to incorporate methods to capture ITH to faithfully predict patient outcomes.

A deeper understanding of the biological underpinnings of RCC has led to novel therapeutic opportunities. This has resulted in a dramatic shift in the treatment landscape over the past decade. Immunotherapy and ICI+TKI combinations are now frontline therapies and objective response rates as high as 60% are seen.4,65,66 Future directions will focus on identifying the correct agent for the correct patient, as well as developing novel therapies. NGS has allowed the identification of activating mutations in oncogenes, and these have been very effective targets in cancers such as melanoma and non-small cell lung carcinoma. Oncogenes are not, however, commonly mutated in RCC and thus have not been an area of therapeutic gain. One exception is the MET oncogene in which activating mutations have been observed in nearly 20% of type I pRCC.47,48,77,78 Cabozantinib is a TKI with activity not only against the VEGFR, but also MET. However, how much MET targeting contributes to its activity against ccRCC is unclear. In a phase III trial comparing second line cabozantinib to everolimus in ccRCC, cabozantinib was found to result in an improved OS rate (HR 0.66 [95% CI: 0.53 – 0.83]; p= 2.6E-4), as well as improved PFS (HR 0.51 [95% CI: 0.41 – 0.62]; p< 1E-4). MET overexpression by IHC did not predict, however, PFS (HR 0.41 [95% CI: 0.53 – 0.83] vs HR 0.58 [95% CI: 0.43 – 0.79]).79 Retrospective analysis of 112 patients with non-ccRCC treated with cabozantinib demonstrated efficacy with 27% (30/112) of patients achieving an objective response.80 For the small subset of patients with genomic data, 40% (4/10) of pRCC patients with MET mutations demonstrated partial responses.80 While these data did not reach statistical significance, prospective trials investigating cabozantinib (as well as other MET inhibitors) in MET-driven pRCC are ongoing (i.e. NCT03091192).

     While targeting driver mutations may provide benefit to a subset of RCC patients, techniques leveraging tumor suppressor genes therapeutically are needed to benefit the larger population. One strategy to tackle loss-of-function mutations in tumor suppressor proteins has been to inhibit downstream effector pathways. In the setting of VHL loss, HIF-2aα accumulates and binds HIF-1b, and the heterodimer upregulates the expression of hun- dreds of genes important to tumor growth including VEGF.21 Given the key role HIF-2 mediated transcription in ccRCC development, direct inhibition of HIF-2α has substantial potential. Transcription factors such as HIF-2α have classically been regarded as “undrugable,” as they lack catalytic pockets suitable for targeting by small molecules. However, characterization of the atomic structure of HIF-2α identified a highly structured pocket that could be bound by small molecule inhibitors.81,82 Compounds with improved pharmacological properties were subsequently developed through iterative structure-based design.83 This led to the development of PT2385 and PT2399, which were shown to be potent and highly selective inhibitors leading to the dissociation of HIF-2α complexes.22,83,84 Preclinical testing of PT2399 in our laboratory demonstrated decreased tumor growth across ~50% of ccRCC tumorgrafts analyzed (P<0.0001), including in sunitinib resistant tumors.22 However, prolonged therapy with PT2399 led to the development of acquired resistance in tumorgraft models.22 Sequencing of tumorgrafts with acquired resistance to PT2399 led to the identification of point mutations which restored dimerization in the presence of inhibitors,22 one of which was subsequently identified in patient tumors that developed resistance to HIF-2α inhibition.85 In a phase I trial, PT2385 demonstrated a favorable safety profile and disease control lasting greater than 4 months in 40% (21/52) of patients, despite heavy pretreatment with a median of 4 prior therapies.86 PT2977, a second generation inhibitor with more consistent drug circulating levels, demonstrated a similar safety profile to PT2385 in recently reported results of a phase I trial in ccRCC.87 Anemia, which is thought to be an on-target effect through suppression of erythropoietin, was the most common adverse event, and only 4% (2/55) of patients stopped therapy due to adverse events. The patients in this trial were heavily pretreated, 62% (34/55) had greater than three lines of therapy, including TKI and ICI therapy. In spite of this, a promising efficacy signal was seen; the median PFS was 11 months (95% CI: 6 – 17), 24% (13/55) of patients experienced a partial response and 56% (31/55) demonstrated stable disease (NCT02974738).87 A phase II trial of (now MK-6482) in combination with cabozantinib is ongoing (NCT03634540), as well as phase II trials of both agents in VHL syndrome related ccRCC (NCT03108066, NCT03401788).

     An alternative approach to targeting tumor suppressor genes leverages “synthetic lethality,” where loss of two genes results in cell death whereas loss of either gene does not88. Since VHL is lost in nearly all ccRCC, identifying molecular targets that exhibit synthetic lethality with VHL loss is an attractive strategy. Several groups, including our own, have developed high-throughput screening platforms of chemical libraries that are capable of identifying compounds which exhibit selective killing of VHL deficient cells.89-94 In one of the first such studies, Turcotte and colleagues screened a panel of ~64,000 small molecules in parallel on VHL deficient RCC4 cells and RCC4 cells with re-introduced VHL.89 They found that STF-62247 was able to selectively induce apoptosis in VHL deficient cells, likely through inhibition of protein trafficking.89 Utilizing the same screen, STF-31 which acts through inhibition of GLUT1, was also found to have preferential toxicity among VHL deficient cells.90 Employing a strategy where differentially labeled VHL deficient and reconstituted RCC cell lines were co-cultured, we identified homoharringtonine (HHT) as a hit compound.93 Furthermore, HHT demonstrated efficacy in ~30% of tested tumorgrafts.93 While these screens have the potential to identify promising compounds, other strategies utilize short hairpin RNA (shRNA) libraries to identify gene combinations which exhibit synthetic lethality.

     One initial report utilizing a shRNA library directed against 88 kinases in VHL-deficient RCC cell lines identified CDK6, MET, and MAP2K1 as potential targets.95 More recently, an expanded shRNA library targeting ~1000 genes identified EZH1 depletion to be synthetically lethal with VHL loss.96 EZH1/2 are histone methyltransferases which canonically act to trimethylate lysine residue 27 on histone 3 (H3K27). Interestingly, constitutive HIF signaling mediates relative H3K27 hypomethylation, potentially explaining EZH1 and VHL synthetic lethality.96 Pharmacological inhibition of EZH1/2 in VHL- deficient RCC cell lines recapitulated these findings, however, the compounds were toxic in mice models.96 Whereas EZH2 was not identified as exhibiting synthetic lethality in the aforementioned screen, there is preclinical evidence that EZH2 inhibitors may be effective in the setting of BAP1 deficiency. Mice with isolated BAP1 deficiency in hematopoietic precursors develop myelodysplastic syndrome.97,98 In these models, BAP1 deficiency results in increased EZH2 expression and methylation of H3K27. Interestingly, EZH2 depletion by both genetic and pharmacologic methods abrogated the oncogenic effect of BAP1 loss.97 In ccRCC, increased levels of EZH2 expression by IHC are associated with higher grade and worse outcomes.99 In addition, RCC-derived cell lines deficient in BAP1 overexpress EZH2, and are sensitive to EZH2 inhibitors in vitro.100 Furthermore, in a sunitinib-resistant xenograft model of RCC, the EZH2 inhibitor EPZ011929 demonstrated rescue of sunitinib sensitivity through epigenetic reprograming (BAP1 status was not reported in this study).101 At the time of preparation of this manuscript there are no RCC specific clinical trials involving EZH2 inhibitors.

     An alternative downstream target of HIF-2, CCND1 (encoding Cyclin D1) is also overexpressed in VHL-deficient RCC. Cyclin D1 binds the CDK4/6 kinases resulting in phosphorylation and inactivation of the retinoblastoma (RB) protein, with subsequent progression through the cell cycle.102 Of note, RB loss is an uncommon event in ccRCC, and thus cell cycle progression likely remains CDK4/6 dependent. The CDK4/6 inhibitor palbociclib demonstrated a G0/G1 cell-cycle arrest, induction of late apoptosis, and blockade of RB phosphorylation in multiple RCC cell lines.103 Abemaciclib, another inhibitor of the CDK4/6 enzyme, was shown to diminish tumor growth in combination with sunitinib in mouse tumorgraft models,104 and a phase I trial of the combination is now actively recruiting (NCT03 905889). Recently, Nicholson and colleagues demonstrated synthetic lethality between CKD4/6 and VHL in ccRCC cell lines as well as in a Drosophila model, suggesting a fundamental dependency between these two gene products.105 Furthermore, the anti-proliferative effects of CDK4/6 inhibition were synergistic with HIF-2 inhibition in mouse xenograft models of disease,105 suggesting the combination of CKD4/6 inhibitors with the HIF-2 antagonists described above may also have therapeutic potential.

     Another emerging strategy is to target the NRF2 pathway. As previously alluded to, NRF2 is negatively regulated by KEAP1, and under conditions of oxidative stress KEAP1 is bound by p62 releasing NRF2 to localize in the nucleus and bind ARE. The NRF2/ARE pathway regulates a number of genes involved in oxidative stress regulation, drug metabolism, and cell proliferation.106 Among other functions, NRF2 plays a vital role in overcoming oxidative stress and treatment resistance. Accordingly, overexpression of NRF2 has been implicated as a negative prognostic marker in several tumor types.107 Overexpression of NRF2/ARE-controlled genes are a distinguishing feature of type II pRCC,48 and elevated expression of NQO1 is associated with worsened outcomes.48 Activating mutations in ccRCC are less frequent than type II pRCC, however, emerging evidence suggests epigenetic silencing of KEAP1 may contribute to NRF2/ARE deregulation in ccRCC.108 Consistent with a potential role of NRF2/ARE in ccRCC, NRF2 depletion via shRNA was recently shown to decrease proliferation and increase sensitivity to sunitinib in the 786-O ccRCC cell line.109 Several flavonoids have been demonstrated to have non-specific NRF2 inhibition, possibly through stimulating polyubiquitination of NRF2.107 Direct inhibitors of NRF2 via the KEAP1 binding domains are also in development.106,110

The Official Journal of the Kidney Cancer Association

Home      Archives      About Us      Submission Guidelines         Editorial Board       Advertising Information      Contact Us

Worldwide, there were over 400,000 new cases and 175,000 deaths attributable to renal cell carcinoma (RCC) in 2018.1 In the United States alone, there are predicted to be over 73,000 new cases of RCC, accounting for nearly 15,000 deaths.2 17% of patients present with metastatic disease with only 12% of patients surviving 5 years.3 Fortunately, outcomes are improving. The number of efficacious systemic therapies for RCC has increased over the past decade and there are now over a dozen FDA approved agents and combinations for use in metastatic RCC.4-6 The treatment landscape has changed from one comprised exclusively of recombinant cytokines to one which includes angiogenesis inhibitors (mostly tyrosine kinase inhibitors, TKI), mammalian target of rapamycin (mTOR) inhibitors, and most recently, the immune checkpoint inhibitors (ICI).4,7 Despite this progress, precision medicine has advanced little and there are no biomarker tests approved by the FDA to guide treatment selection.

     Classically, RCC is subdivided histologically into clear cell RCC (ccRCC) accounting for 75% of cases, type I and II papillary RCC (pRCC) accounting for 10% of cases, chromophobe RCC (chRCC) accounting for 5% of cases, and other less frequent subtypes.8 RCC is now recognized to be a diverse group of diseases with updated society guidelines incorporating molecular and genomic data along with histologic information when defining RCC subtypes.9,10

     To determine the role of BAP1 and PBRM1 in ccRCC development, we inactivated them in nephron progenitor cells in the mouse and assessed their impact on RCC development. By simultaneously targeting Bap1 and Vhl, we developed the first mouse model of ccRCC thereby over- coming a decade-long struggle.19 We showed that ccRCC development required not only Vhl inactivation, but also the inactivation of Bap1 (or Pbrm1). As for Bap1, Pbrm1 loss was not sufficient to induce RCC. However, the simultaneous inactivation of Vhl and Pbrm1 caused ccRCC.37 Similar observations were made by others.41,42 We also found, that as in humans, Bap1-deficient tumors were of high grade, whereas Pbrm1-deficient tumors were of low grade. In addition, Pbrm1-deficient tumors developed after a longer latency period.37 Interestingly, targeting one allele of Tsc1, which encodes a negative regulator of mTOR complex 1, in a Vhl/Pbrm1-deficient background reduced the latency period and increased the frequency of higher grade tumors37 (Figure 1).

Vol 17, No 3    2019