Abstract
Renal cell carcinoma (RCC) is linked to signature molecular alterations that influence their morphological characteristics and associated clinicopathologic features. This has been fueled by an incremental understanding of the molecular underpinnings of RCC. The current 2022 WHO classification has now incorporated novel diagnostic entities as well as a new category of “molecularly defined renal carcinomas”. This review delves into our current understanding of molecular correlates with the evolving taxonomic classification of RCC. This review’s contents consist of molecularly defined entities with morphological annotations that correspond to distinct morphologies such as clear cell, papillary, and eosinophilic. Thus, we highlight the morphological context, diagnostic algorithm, and the available biomarkers that can assist practicing pathologists in these respective categories and thereby correctly diagnosing these unique tumor entities. Wherever possible a brief summary of the current therapeutics that align with morphological and molecular correlates have been providedas well.
Keywords
INTRODUCTION
Renal cell carcinoma (RCC) classification has evolved over the last several decades, from just four original RCC subtypes recognized in the 1997 Heidelberg Classification to nearly two dozen in the 2022 5th edition of the WHO Classification of Urinary and Male Genital Tumors [1]. The current WHO classification introduces several novel diagnostic entities as well as a new category of “molecularly defined renal carcinomas”, with seven RCCs being placed under this heading, including TFE3-rearranged RCC, TFEB-altered RCC, ELOC-mutated RCC, fumarate hydratase-deficient RCC, succinate dehydrogenase-deficient RCC, ALK-rearranged RCC, and SMARCB1-deficient renal medullary carcinoma [2]. In parallel, our insight into the molecular mechanisms of disease in established renal tumor subtypes continues to deepen. Amongst several pathways considered to be pivotal towards fueling the development of diverse renal neoplasia are those involving aberrations in HIF/VHL, MET, MTOR/TSC, MiTF, and Hippo pathway genes.
In this review, we summarize recent updates to the field of renal neoplasia with defined molecular alterations and associated morphologic perspectives (Table 1). We present emerging concepts in the field of kidney cancer, where increasingly molecular correlates are being described to understand the phenotypic heterogeneity that surgical pathologists have been observing under the microscope.
Summary of biomarkers (proteins, genes, chromosomes) aligning with existing and newly defined emerging RCCs.
MOLECULAR AND TAXONOMY-BASED EVOLUTION OF CLEAR CELL RENAL CELL CARCINOMA
VHL-deficient clear cell renal cell carcinoma: morphologic and genomic correlates
Clear cell (also known as conventional) renal cell carcinoma (CCRCC) most often occurs sporadically, featuring cells with cytoplasmic ‘clearing’ arranged in a nested to alveolar architecture as well as accentuated vasculature. CCRCC also occurs within the spectrum of VHL syndrome, with a clinicopathologic phenotype ranging from simple to complex, and the kidneys typically house numerous cystic and solid lesions. Notably, patients with VHL disease have a 70% lifetime risk of developing RCC by the age of 60 years [3]. In CCRCC (sporadic or familial), activation of hypoxia-inducible factor (HIF) signaling is caused by biallelic inactivation of the VHL gene, resulting in HIF1 signaling deregulation [4]. VHL loss is currently considered an early truncal event and may happen due to various mechanisms such as 3p25 loss, point mutation/indel, and epigenetic promoter methylation, among others [5]. VHL silencing, in turn, leads to upregulation of HIF target genes, resulting in promotion of angiogenesis, glycolysis, and apoptosis, with the downstream accentuated vasculature and cytoplasmic clearing readily evident to pathologists under themicroscope.
BAP1, PBRM1, SETD2, and UCHL1 alterations in CCRCC
Loss of chr3p impacts chromatin remodeling genes located in this region including PBRM1, SETD2, and BAP1 (Table 1); BAP1 deficiency in particular is associated with disease progression and a more aggressive phenotype in CCRCC [6]. PBRM1 or BAP1 mutation is observed in roughly 40% and 10–15% of CCRCC respectively; however, mutations in either gene are largely mutually exclusive [7, 8]. BAP1 is a two-hit tumor suppressor gene encoding a nuclear protein impacting chromatin remodeling and downstream cell cycling signaling [8, 9]. BAP1 loss results in cell cycle dysregulation, with subsequent cellular proliferation and tumorigenesis. [6, 8, 10]. The presence of a germline BAP1 mutation is the hallmark of the autosomal dominant BAP1 tumor predisposition syndrome, a condition where up to 85% of affected individuals develop associated lesions, including CCRCC [11].
BAP1-mutated CCRCC demonstrates an abrupt demarcation of BAP1-mutated and conventional clear cell components, imparting a biphasic morphology. BAP1-mutated tumor areas have been reported to exhibit papillary architecture, voluminous eosinophilic cytoplasm, large eosinophilic cytoplasmic globules and granules, and high-grade nuclei spanning WHO/ISUP grades 3 to 4 (including sarcomatoid and rhabdoid features) [9, 12]. Most cases of BAP1-mutated CCRCC demonstrate loss of BAP1 expression in tumor cells by immunohistochemistry (IHC) with retention of expression in background stromal/inflammatory/endothelial cells. Additionally, AMACR expression by IHC has been reported in BAP1-mutated CCRCC [9]. Importantly, apart from the above morphologic correlates, multivariable analyses of a large cohort of small renal masses found BAP1 status to be independently associated with time to metastases after controlling for TNM stage [13]. These findings are in tune with the initial observations from The Cancer Genome Atlas Program (TCGA)-based interrogations of CCRCC. Utilization of BAP IHC in clinical practice may be useful in small renal masses that may otherwise be amenable to active surveillance [13].
PBRM1 loss, unlike BAP1 loss, is not associated with higher tumor grade in CCRCC or non-CCRCC tumors [6, 8, 10]. Rarely, concomitant BAP1 and PBRM1 mutations occur within the same tumor and are associated with rhabdoid morphology [4, 5, 8, 14]. SETD2 mutations are present in approximately 10% of CCRCC [6, 8, 10]. SETD2 mutations are equally distributed between PBRM1-deficient and PBRM1-wildtype tumors [8]. Interestingly, when present, PBRM1 mutations appear to precede SETD2 mutations in the clonal evolution of CCRCC [4, 5] (Table 1).
A recent publication from the Clinical Proteomic Tumor Analysis Consortium (CPTAC) discusses the role of overexpression of the deubiquitinase UCHL1 as a prognostic marker in CCRCC [12]. High UCHL1 expression is associated with worse prognosis and survival, BAP1 mutation status, intratumoral heterogeneity, genomic instability, and a specific methylation profile. Primary tumors with strong UCHL1 expression by IHC appear more likely to develop metastatic disease. UCHL1 small-molecule inhibitors are being studied in triple-negative breast cancer and neuroendocrine carcinoma of the lung and could show promise in CCRCC [12]. A follow-up study supports UCHL1 expression with BAP1 loss and nodular tumor heterogeneity within CCRCC [15] (Table 1). The role of BAP1 and UCHL1 immunostains in RCC diagnosis is an active area of investigation.
While CCRCC continues to be a relatively straightforward pathologic diagnosis, confounding factors include the morphologic deviations seen within the spectrum of CCRCC as well as other (non-CCRCC) renal neoplasms that might present with cytoplasmic ’clearing’. CCRCC, especially when high-grade, may lack clear cytoplasmic, and present with eosinophilic features or other uncommon morphologies with associated genomic correlates such as sarcomatoid or rhabdoid features. A subset of CCRCC with eosinophilic features have been shown to be enriched for p53 mutations [16]. The differential diagnosis of apparently higher-grade CCRCC (for example, with BAP1, p53 alterations, and others) includes MiTF RCC, papillary RCC (PRCC), and other molecularly defined entities. CCRCC shows diffuse CAIX expression by IHC [17]. MiTF RCC will show no or patchy CAIX expression and can be ruled out by TFE3/TFEB fluorescence in situ hybridization (FISH) and/or TRIM63 RNA in situ hybridization (RNA ISH) (discussed in more detail below). CCRCC with BAP1 mutation shows BAP1 loss by IHC in most cases and may show strong AMACR expression; thus a papillary RCC with clear cell change needs to be kept in perspective when evaluating AMACR expression in renal tumors.
The concept of upregulated HIF target gene markers in CCRCC has been exploited by clinicians (for example, in establishing VEGF-TKI inhibitors). Indeed, the HIF2 inhibitor belzutifan has already been approved to treat adults who have several tumors associated with VHL disease and is currently in clinical trials for sporadic tumors (ClinicalTrials.gov Identifier: NCT05468697, NCT05239728, others) [14, 18].
MET PATHWAY AND CIMP ALTERATIONS IN PAPILLARY RENAL CELL CARCINOMA (PRCC)
Canonical MET aberrations and expanding the spectrum of PRCC
PRCC, classically characterized by papillary structures lined by cells with pale to eosinophilic cytoplasm and variable nuclear grade (historically categorized as type 1) are associated with gains of chr7 (encoding MET), chr17, and activating somatic mutations of MET [19, 20]. MET upregulation or gain of chr7 is seen in ∼80% of PRCC, a small subset of which are germline; the latter represents the defining molecular event in hereditary PRCC [19, 21]. From a functional perspective, MET is a proto-oncogene involved in tumor proliferation and invasiveness, as well as playing a role in driving resistance to anti-tumor therapies [22] (Table 1).
Previously PRCC was divided into subtypes 1 and 2; however, there are many reasons why this is no longer recommended. Above all, this distinction shows poor interobserver reproducibility in the literature with no demonstrated clinical importance [20]. It is also now well established that the category of PRCC incorporates numerous unique and well-defined carcinoma subtypes with papillary architecture, including fumarate hydratase (FH)-deficient RCC, acquired cystic disease-associated RCC, and MiTF-altered RCC, amongst others; many of these tumors, including FH-deficient RCC, are described in more detail in this edition in the review on tumor with papillary architecture. Furthermore, a recent study has shown PRCC with mixed so-called type 1 and type 2 features to be more akin to so-called type 1 PRCC at the clinicopathologic and molecular level [23].
Although PRCC isn’t recommended to be regularly subtyped into type 1 and type 2 categories, a number of morphologic patterns have been described in recent years including “biphasic (alveolar and squamoid) PRCC,” “Warthin-like PRCC,” and “solid PRCC.” A recent study reported that biphasic PRCC, including “glomeruloid/alveolar” architecture, may harbor MET alterations in 60% of cases [24, 25]. In these tumors, the outer cell layer is composed of smaller cells, while the inner layer includes larger, non-cohesive cells, often involved by emperipolesis (or cytophagocytosis). Apart from enrichment for MET mutations, a subset of these cases show aggressive behavior [24]. Warthin-like papillary RCC contains eosinophilic tumor cells admixed with lymphocytes, and solid papillary RCC can be diagnostically ambiguous due to compression of tubular and papillary structures [25].
Finally, mounting evidence supports classifying the so-called papillary renal neoplasm with reverse polarity (PRNRP) as a separate entity from PRCC. PRNRP is composed of cells with low-grade nuclei and voluminous eosinophilic cytoplasm arranged in papillae with hyalinized cores, with characteristic arrangements of nuclei at the apical pole of tumor cells [20, 23]. While KRAS aberrations are distinctly uncommonly seen within the TCGA interrogation of RCC, KRAS alterations have been consistently reported in the majority of PRNRP [24, 25] (Table 1).
The majority of PRCCs tend to show diffuse AMACR and CK7 expression by IHC. Biphasic PRCC tumors show strong cyclin D1 expression in the larger squamoid cells. PRNRP is uniquely immunoreactive with GATA3 and L1CAM [26].
CIMP-RCC dichotomy in the PRCC spectrum
Another biologically important consideration that has been gaining traction in recent years is the so-called CpG island methylator phenotype (CIMP)-RCC, first characterized by the TCGA consortium [19]. These cases demonstrate worse overall survival compared with other PRCC cases (and other kidney tumors in general) [19]. Further research revealed “CIMP-RCC” may represent a more heterogeneous group of tumors as it does not appear to harbor specific copy number alterations, though 60% of cases showed increased loss of chr22 (where NF2 and SMARCB1 are located) as well as loss of chr13q (where RB1 and BRCA2 are located) [6] (Table 1).
A recent genome-wide methylation study investigated the methylation profile of metabolic RCC such as FH-deficient RCC, succinate hydrogenase B (SDHB)-deficient RCC, and remaining non-FH and SDHB-deficient CIMP-RCC. The results showed that FH-deficient RCC and SDHB-deficient RCC tumors included in the study also harbored distinct CpG island methylator phenotypes [27, 28]. While hypermethylation was extensive and uniformly identified in FH-deficient RCC, it was variable and less prominent in SDHB-deficient RCC. Nevertheless, CIMP-RCC is becoming an area of intense interrogation by multiple groups, and future studies may reveal prognostic and possible therapeutic connotations. Tissue-based biomarkers to label CIMP-RCC (especially the ones without FH or SDH deficiency) can help identify aggressive renal tumors. Meanwhile, FH- and SDHB-deficient RCCs are diagnosed with great certainty using morphological correlates and IHC assays, with loss of FH expression and positive 2-succinocystine (2SC) IHC noted in most FH-deficient RCCs and loss of SDHB expression by IHC in SDHB-deficient RCCs [27].
HIPPO PATHWAY ALTERATIONS IN RENAL NEOPLASIA
Mucinous tubular and spindle cell carcinoma (MTSCC)
MTSCC is a relatively indolent neoplasm which is known to demonstrate biallelic alterations of the Hippo pathway genes [29, 30]. Classically this tumor is described as demonstrating bland spindle cells and tubules in a mucinous/myxoid background (hence its nomenclature); however, aggressive cases with sarcomatoid transformation and distant metastasis have been reported [31]. Despite varied morphology, MTSCC has distinct chromosomal aberrations and gene signatures that help to distinguish it from PRCC and other such lesions with overlapping morphology. Accordingly, MTSCC consistently shows loss of heterozygosity of chromosomes 1, 4, 6, 8, 9, 13, 14, 15, and 22 with the absence of gains of chromosomes 7, and 17 (the latter are seen in PRCC) [29, 30, 32]. Biallelic alteration of Hippo signaling pathway genes (PTPN14, NF2, SAV1) result in increased YAP1 nuclear expression by IHC; however, this is not specific to MTSCC, nor is it used diagnostically [29, 30] (Table 1). In a similar space, VSTM2A overexpression detected by RNA ISH technology has emerged as a cancer-specific biomarker for MTSCC [33]. Recent work has also shown that a gain of 1q at the location of some oncogenes was the most common chromosomal aberration to differentiate locally advanced/metastatic MTSCC from indolent cases [33, 34]. Such tumors were seen at metastatic sites including lymph nodes, bone, and retroperitoneum, and were enriched for CDKN2A/B deletion and additional complex genomic abnormalities.
RCC with sarcomatoid features
Sarcomatoid differentiation can arise from any RCC and does not represent a distinct RCC subtype. Morphologically, sarcomatoid features includes atypical spindle cells and high cellularity, together with the relative loss of typical epithelioid phenotype. It is frequently associated with necrosis and is known to be associated with aggressive biologic behavior. It is thought that 5–8% of CCRCC, 8–9% of ChRCC, and 2–3% of PRCC can exhibit these characteristics [35]. Metastatic disease occurs in ∼75% of individuals with RCC with sarcomatoid differentiation regardless of underlying RCC subtype. The mechanism of sarcomatoid differentiation in RCC is not clearly understood, however, evidence suggests that the sarcomatoid component may originate from a common cell of origin, resulting in cells that lose their epithelial properties and acquire mesenchymal properties through a process called epithelial-mesenchymal transition (EMT). EMT can occur via a number of mechanisms, including alterations in TNF, TGFβ, Wnt, MAPK, and PI3K/AKT signaling [36].
In CCRCC with sarcomatoid differentiation, a subset of tumors are enriched for actionable driver Hippo pathway alterations. Mutations in Hippo pathway genes (most often NF2, FAT1, and LATS2) are more often present in the sarcomatoid component of a tumor than in the underlying non-sarcomatoid (epithelial) component of the same RCC [37, 38]. Additionally, RCC with sarcomatoid differentiation has been shown to have higher expression of PD-1 and PD-L1 than other RCC subtypes. This is an alternative perspective that is being explored using newer combinations of immune checkpoint inhibitor immunotherapies that may provide better responses and outcomes [36]. Amplification of JAK2, PDL1, and PDL2 at 9p24.1 is significantly enriched in approximately 6% of RCCs with sarcomatoid differentiation [39] (Table 1).
NF2-mutated RCC: an expanding category
Biphasic hyalinizing psammomatous RCC (BHP RCC) is an emerging RCC subtype which is characterized by a dual population of smaller hyperchromatic cells arranged in nests or pseudo-rosettes surrounding eosinophilic basement membrane material and larger cells arranged in tubules, papillae, or acini, while also harboring an NF2 mutation [40, 41]. Recently, RCC with sex-cord/gonadoblastoma-like features harboring NF2 mutations have been reported [42]; however, it is likely that these tumors are also BHP RCC. Regardless, it has been shown that NF2 encodes a tumor suppressor called merlin, which effects a tumor suppressive response by upstream regulating Hippo signaling through its action on YAP and TAZ [43].Recent focused research on NF2 mutations in RCC has also led to an observation that NF2 mutations are not specific to BHP RCC and can be present as secondary events in other types of RCC, including MTSCC, RCCs with sarcomatoid differentiation, FH-deficient RCC, amongst others [19, 29, 37, 43–46], suggesting a shared homology with the Hippo pathway. Additionally, an in vitro study demonstrated targeting of YAP1 partner by dasatinib and saracatinib in NF2-mutated cell lines led to repression of Hippo pathway transcriptional products, suggesting a potential therapeutic role [47]. In their seminal work, Chen et al. found NF2 mutations commonly occur in RCC, not otherwise specified or unclassified [43] (Table 1).
Merlin loss by IHC correlates well with the presence of an NF2 mutation in RCC [48] (Fig. 1A & B). A study by Collins et al. found merlin loss by IHC was ∼92% sensitive and ∼94% specific to distinguish BHP RCC from other RCC subtypes (including PRCC, TFE3-rearranged RCC and TFEB-altered RCC [48]. Of note, cases with morphology compatible with BHP RCC but lacking NF2 alterations have been reported [49]. Some cases lacking NF2 mutations demonstrate NF2 promoter methylation, showing that epigenetic silencing can lead to NF2 inactivation in a subset of cases [50]. Merlin IHC may be useful in these cases. Regardless, further studies are required to better define the clinicopathologic and molecular features of such entities (Table 1).
A-B) Merlin deficient renal cell carcinoma. H&E image demonstrating nested to trabecular features with eosinophilic cells and a background myxoid stroma; merlin immunohistochemistry demonstrates loss of protein expression in tumor cells but retained staining in admixed/background stromal, inflammatory, and endothelial cells. C-D) TFEB amplified renal cell carcinoma. H&E image demonstrating papillary architecture including cells with eosinophilic to clear cytoplasm and subtle linear arrangement of nuclei away from the base; RNA in situ hybridization demonstrating TRIM63 overexpression (brown, granular) in this tumor (inset demonstrates TFEB gene amplification by dual-color, break-apart FISH). E-F) ELOC (formerly TCEB1)-mutated renal cell carcinoma. H&E images demonstrating multinodular tumor with prominent fibromuscular stroma, papillary to alveolar features, and cells with voluminous clear and/or eosinophilic cytoplasm; these tumors have intact chr3 and loss of chr8. G-H) Hybrid oncocytic tumor from a Birt-Hogg-Dube syndrome patient. Combination of L1CAM IHC and H&E performed on the same section (G) shows L1CAM (brown, membranous) to be preferentially expressed in neoplastic cells with clear cytoplasm; combination of LINC01187 RNA ISH and H&E performed on the same section (H) shows LINC01187 (brown, nuclear) to be preferentially expressed in neoplastic cells nucleus with eosinophilic cytoplasm.
MITF-ASSOCIATED RCC, INCLUDING TFE3-REARRANGED AND TFEB-ALTERED RCC
MiTF RCC comprises approximately 40% of pediatric RCC and < 10% of adult RCC [1, 51]. The current WHO Classification separates MiTF RCC into TFE3-rearranged RCC (formerly Xp11.2 translocation RCC) and TFEB-altered RCC (including both TFEB-rearranged RCC and TFEB-amplified RCC) [1]. It is recognized that MiTF RCC has been miscategorized as so-called type 2 PRCC in the past [52]. MiTF RCC shows a wide spectrum of morphologic features, and often mimics high-grade CCRCC, PRCC, chromophobe RCC (ChRCC), other oncocytic tumors of the kidney, such as perivascular epithelioid cell neoplasm (PEComa), and tumors with uncommon clinicopathologic phenotypes [53–57]. The morphology and fusion partners involved in TFE3-rearranged RCC and TFEB-rearranged RCC have been described in several studies, with some carrying unique morphologic annotations [25, 58–60]. Notably, TFE3 and TFEB rearrangements do not co-occur within the same tumor [57]. Confirmation via FISH is considered the gold standard for diagnosis of these RCCs [61]. Issues with TFE3 and TFEB IHC have been previously detailed [61].
TFE3-rearranged RCC demonstrates gene fusions involving TFE3 with an increasing number of recognized partner genes, most frequently PRCC, ASPL, and SFPQ; several other fusion partners have been identified in recent years and have been described elsewhere [1, 25]. Fusion partners have been associated with specific morphology have been described at length [25, 58, 60]. While TFEB-rearranged RCC commonly demonstrates biphasic morphology, there is a considerable overlap with TFE3-rearranged RCC. TFE3- and TFEB-rearranged RCCs are more frequently diagnosed in children/young adults, and TFEB-rearranged RCC have a better prognosis. In contrast, TFEB-amplified RCC has emerged as distinct from TFEB-rearranged RCC with distinct clinical, behavioral, and morphological features [25, 56, 57]. TFEB-amplified RCC is seen as much as three times more frequently than TFEB-rearranged RCC, and reported cases tend to occur at an older age and are more aggressive when compared with TFE3-/TFEB-rearranged RCC [56, 57]. The morphology of TFEB-amplified RCC is heterogenous and can include high-grade nested features with solid to papillary architecture; high-grade tumors with oncocytic and papillary features including linearly arranged nuclei with alignment away from the basement membrane are seen to be enriched for TFEB amplification [57] (Fig. 1C).
TRIM63 overexpression by RNA ISH has emerged as a cancer-specific biomarker for MiTF RCC and can help distinguish TFE3-rearranged RCC, TFEB-rearranged RCC and TFEB-amplified RCC from other RCC subtypes [62] (Fig. 1D) (Table 1). Notably, initial studies indicate that TRIM63 overexpression can detect cases of RBM10::TFE3 inversion-based fusion that are difficult to diagnose on FISH or are falsely negative by FISH-based technology [62]. However, as TRIM63 overexpression is seen in all 3 different types of MiTF RCC, FISH, PCR, or NGS is necessary to distinguish between these tumor types.
In a significant development, recent artificial intelligence-based studies have manufactured deep-learning algorithms which, utilizing hematoxylin and eosin slides only, can reliably help distinguish TFE3-rearranged RCC from CCRCC [63]; such models can also help categorize renal tumors into different diagnostic classes [64]. Artificial intelligence in pathology is rapidly evolving and carries the potential to transform kidney cancer management from tumor biology, diagnostic, prognostic, and therapeutic perspectives [64–66].
MAMMALIAN TARGET OF RAPAMYCIN (MTOR)/TUBEROUS SCLEROSIS COMPLEX (TSC) PATHWAY ALTERATIONS IN RENAL NEOPLASIA
The mTOR/TSC pathway has been increasingly recognized as central to the oncogenesis of diverse renal neoplasms, often with oncocytic cytoplasm [67–70]. The mTOR protein interacts with other proteins to constitute mTOR complex 1 and 2 (mTOR1 and mTOR2), which in turn influence cellular proliferation and progression. Similarly, alterations in the TSC genes have been implicated in numerous renal lesions with somatic and germline mutations of TSC1 and TSC2. Hamartin and tuberin, the respective proteins of TSC1 and TSC2, regulate cellular division, progression, and, most importantly, cell size. Interactions between the mTOR and TSC pathways produce numerous downstream effects, with TSC exerting an inhibitory effect on mTOR pathway activation [69] (Table 1). Currently, the RCC seen within spectrum of TSC1/2/MTOR mutations is not listed as a formal renal tumor subtype in the 2022 WHO classification [1].
Historically, renal lesions recognized as part of the TSC complex included angiomyolipomas (AML), cysts, and RCCs. Our current understanding of TSC-associated RCC comes from two landmark studies conducted by Yang et al. and Guo et al. [71, 72]. Since then, several renal epithelial neoplasms with overlapping appearances have been detected in individuals with TSC syndrome and from patients without TSC syndrome (thus, sporadic) and marked by somatic mutations in the TSC1, TSC2, and MTOR genes. Accordingly, TSC-associated tumor/RCC does not equate to a single renal neoplasm subtype, but rather represents multiple tumors type which harbor overlapping mTOR pathway alterations. Some of these tumors are accepted as distinct entities in the WHO displaying relatively consistent morphologic features and IHC profiles, while others are still considered emerging entities. Examples of such entities include RCC with fibromyomatous or leiomyomatous stroma (RCCFMS), eosinophilic solid and cystic renal cell carcinoma (ESC RCC), low-grade oncocytic tumor (LOT), and eosinophilic vacuolated tumor (EVT) (each is described in more detail below) [67, 73]. In addition to these mentioned, the differential diagnosis of oncocytic tumors demonstrating mTOR pathway alterations also includes renal oncocytoma, ChRCC, CCRCC with eosinophilic features, FH-deficient RCC, SDHB-deficient RCC, and others. Although not routinely used biomarkers, nuclear expression of the recently characterized transcription factor FOXI1 by IHC and long noncoding RNA LINC01187 by RNA ISH are highly enriched in ChRCC and oncocytomas, and may be useful to detect ‘pink tumors’ originating from the intercalated cell [74]. (Table 1). CCRCC with eosinophilia tends to be labeled with CAIX; FH-deficient and SDHB-deficient renal neoplasms (lesions that can show oncocytic morphology) often show a lack of immunohistochemical expression of FH and SDHB proteins, respectively. Recently, a majority of TSC1/2/MTOR-altered renal tumors have been shown to express GPNMB (glycoprotein nonmetastatic B) protein by IHC; however, this protein may not be entirely specific as it tends to be overexpressed in MiTF RCC because of increased activity of TFE3/TFEB in both tumor subtypes (MiTF RCC and TSC1/2/MTOR alteration-associated renal tumors) [75, 76] (Table 1).
RCC with Fibromyomatous Stroma (RCCFMS)
The gross presentation of these tumors is often small, solid, tan-brown, and lobulated. Microscopically, they display an epithelial component consisting of tumor nodules with elongated and branched tubules lined by clear to eosinophilic cells and a prominent fibromuscular stroma (contributing to the multinodularity seen in these tumors). RCCFMS is characterized by diffuse CAIX and CK7 as well as focal CD10 IHC expression. The differential diagnosis for RCCFMS associated with TSC1/2/MTOR mutations includes CCRCC, clear cell papillary renal cell tumor with fibromyomatous stroma, and TCEB1/ELOC-mutated RCC. These tumors can have similar morphology and IHC characteristics, but show distinct genetic findings (chr3p loss, diploid, and TCEB1/ELOC mutation with monosomy 8, respectively). Promisingly, a recent study has shown GPNMB to be strong and diffusely positive in all tumors with TSC1/2/MTOR alterations, while negative in all TCEB1/ELOC-mutated RCC [77].
Eosinophilic solid and cystic RCC (ESC RCC)
Recognized as a distinct entity by WHO in 2022 classification [78], ESC RCC may be sporadic or demonstrate germline TSC1/2 alterations [77, 78]. Like other entities, ESC RCC may have been erroneously diagnosed as other renal tumor entities, such as eosinophilic ChRCC, in the past [52]. As suggested by the name, ESC RCC shows solid and cystic architecture, though some cases demonstrate entirely solid or even papillary growth patterns. Tumor cells show abundant eosinophilic cytoplasm plus prominent granular, basophilic stippling in the cytoplasm [78]. TSC loss is a clonal event in both sporadic and tuberous sclerosis-associated ESC RCC [79]. While the immunohistochemical profile for these tumors is relatively non-specific, CK20 positivity by IHC is frequently useful in separating ESC RCC from other oncocytic tumors of the kidney [73, 78].
Low-grade oncocytic tumor (LOT)
LOT is an emerging entity which shows morphologic similarity with renal oncocytoma and ChRCC, but was distinctly identified because of its CD117 (KIT) negative and diffuse CK7 positive immunophenotype with a “boats in the bay” (i.e., small nests of cells in edematous stroma) morphologic appearance [78, 80]. Most cases of LOT appear to harbor TSC1/2/MTOR mutations [46, 81–83], although a subset of LOTs have been reported in the setting of end-stage renal disease (ESRD) and tuberous sclerosis complex (TSC) [67, 79, 80, 84]. Additionally, LOT demonstrate GATA3 and L1CAM positivity by IHC, which can be useful in distinguishing LOT from oncocytoma or ChRCC, but not PRNRP[85] (Table 1).
Eosinophilic vacuolated tumor (EVT)
Yet another emerging entity, EVT is composed of large eosinophilic cells with prominent nucleoli, intracytoplasmic vacuolization and solid or nested architecture [78]. Recent cohorts characterizing EVT have demonstrated MTOR pathway aberrations in these tumors [86, 87]. Cathepsin K and CD117 (KIT) positivity along with cytoplasmic vacuolization and prominent nucleoli are useful in distinguishing EVT from renal oncocytoma and ChRCC [73].
Xanthomatous giant cell renal cell carcinoma
Recently Argani et al have described stand-alone cases of TSC2-mutated RCC with an unusual and distinct phenotype. A high degree of permeation within the adjoining kidney, big multinucleate tumor giant cells with tumor cells showing variable cytoplasmic eosinophilic to xanthomatous stippling or vacuolization, and positivity for CK20 are characteristic findings of this newly describe entity [88].
Although some progress has been made in the last ten years in identifying kidney tumors associated with TSC1/2/MTOR mutations in both hereditary and sporadic situations, there are still unanswered questions. It is crucial to determine if these tumors are signature entities with unique biological traits and prognoses, or if they are a range of morphologies connected to activation of the mTOR pathway, for which a common/unifying nomenclature would be enough. Of note, the established RCC subtypes (like CCRCC) demonstrate bi-allelic TSC loss only uncommonly. Currently, the RCC seen within spectrum of TSC1/2/MTOR mutations is not listed as a formal renal tumor subtype in the 2022 WHO classification [1].
ELOC (FORMERLY TCEB1)-MUTATED RCC AND OTHER MORPHOLOGICAL DIFFERENTIALS
Divergent VHL abnormalities in CCRCC and ELOC-mutated RCC
The biallelic loss of the chromosome 3 tumor suppressor gene, VHL, characterizes most CCRCCs. Biallelic inactivation of VHL can be caused by chr3p deletion in conjunction with a mutation in VHL, or hypermethylation of the VHL promoter. While > 90% of CCRCC have chr3p loss, only 60–70% of tumors exhibit VHL point mutations, and 5–10% have documented epigenetic silencing. Given the preponderance, most of the genomic research has been on CCRCC with inactivated VHL, essentially excluding subgroups of tumors that contain wildtype VHL, yet express clear cell features [89].
In 2013, Sato et al. identified TCEB1 inactivation in a fraction of renal tumors with clear cell features and intact VHL [90]. Since then, these tumors have been studied in more detail and are now designated as ELOC (formerly TCEB1)-mutated RCC, and are officially recognized as a distinct entity in the 2022 WHO classification under the category of “molecularly defined renal entities” [2]. ELOC-mutated RCC has thick fibromuscular bands, clear cell morphology in the form of voluminous cytoplasm, and solid acinar and papillary formations (Fig. 1E & F). ELOC-mutated RCC shows recurrent hotspot mutations of the ELOC (TCEB1) gene (8q21), encoding for elongin C, and this is considered an essential criterion to diagnose this entity [91–94]. Elongin C is also a component of the VHL complex, which not only activates HIF (hypoxia-inducible growth factor) but also drives proteasomal destruction (ubiquitination) of hydroxylated HIF. Hence, ELOC-mutated RCC harbor biallelic inactivation of the ELOC (TCEB1) gene and mutation with concurrent monosomy 8 or loss of heterozygosity at 8q21, and in clear distinction with conventional CCRCC, they lack VHL mutations, hypermethylation or loss of heterozygosity at 3p. These tumors can show variable clinical behavior, with predominantly an indolent to uncommonly aggressive phenotype [95].
Convergent morphological overlap including clear cell cytology and fibromyomatous stroma
Tumors with diverse genotypes have been reported to show overlapping histopathological findings including tumor cells with voluminous clear cytoplasm as well as fibromuscular stroma which divides the tumor tissues into nodular areas and branched acini or tubular epithelial structures. Due to the morphologic overlap, diverse renal tumor entities like CCRCC, CCPRT, RCC, NOS with fibromyomatous stroma and an associated TSC mutation, and ELOC-mutated RCC are the principal candidates that need to be distinguished in such instances [93, 94].
CAIX and CK7 are sensitive but not entirely specific in distinguishing ELOC-mutated RCC from other RCC with fibromyomatous stroma. However, as mentioned previously, GPNMB IHC has recently been shown to differentiate RCC, NOS with fibromyomatous stroma and an associated TSC mutation from ELOC-mutated RCC, CCPRT, and CCRCC with fibromyomatous stroma; accordingly strong, diffuse GPNMB staining in mTOR pathway-altered RCCs has been consistently observed [77]. However, GPNMB expression has also been described in a small subset of CCRCC, TFE3-rearranged RCC, TFEB-altered RCC, ESC RCC, and LOT [76] thereby limiting its potential and making its usage relatively context-dependent [69]. Finally, ELOC-mutated RCC, tends to have negative CK903 expression, which is predominantly positive in RCCFMS.
OTHER RENAL NEOPLASIA WITH DEFINED MOLECULAR ALTERATIONS
Birt-Hogg-Dube Syndrome associated hybrid oncocytic tumors (HOT)
An increased incidence of multifocal renal tumors, such as ChRCC, renal oncocytoma, and hybrid oncocytic tumor (HOT), is linked to Birt-Hogg-Dubé (BHD) syndrome. BHD syndrome is an autosomal dominant condition with characteristic germline pathogenic variants in the gene FLCN. Due to heterogeneous histologic characteristics that may overlap with renal oncocytoma and ChRCC, HOT and renal tumor types that resemble HOT can be difficult to diagnose. In a recent study, it was conclusively demonstrated that BHD-associated HOT expresses the principal cell lineage marker L1CAM and the intercalated cell lineage marker LINC01187 in a unique checkered pattern that is mutually exclusive [85] (Fig. 1G & H). These two lineage markers together represent the two distinct tumor epithelial populations that are observed to coexist morphologically in HOT. The usage of these two markers in tandem can distinguish HOT from its morphologic mimics due to the distinct checkered expression pattern of L1CAM and LINC01187 [85]. Importantly, the hybrid morphologic phenotype seen in HOT with a mixture of eosinophilic and relatively clear cells (often in abrupt juxtaposition) is reflective of the dual cellular populations seen in this tumor, with the eosinophilic cells simulating the intercalated cell (labeled by LINC01187) and the clear cell likely simulating the principal cell (labeled by L1CAM) (Table 1).
Anaplastic lymphoma kinase (ALK)-rearranged renal cell carcinoma (ALK-RCC)
ALK-RCC is now officially recognized in the most recent edition of the WHO classification of renal tumors and is characterized by rearrangements of ALK gene located at chromosome 2p23 [1]. Young patients with sickle cell trait have been identified as a group with increased frequency of ALK-RCC (specifically VCL::ALK fusion) that demonstrates consistent morphologic features [96, 97]. In contrast, adults with ALK-RCCs show variable morphology exhibiting polygonal, rhabdoid, signet ring, and spindled cells, mucin production, and cytoplasmic vacuolization and are less frequently associated with sickle cell trait. The latter are more frequently associated with fusion partner genes identified in recent years, including TPM3, HOOK1, STRN, EML4, CLIP1, KIF5B, and KIAA1217. ALK IHC (clones: 5A4 or D5F3) can be a useful screening tool when considering ALK-RCC; molecular studies (e.g., FISH, PCR or NGS) will help to confirm the gene rearrangement and characterize the specific fusion partner [96, 99] (Table 1).
Due to variable histopathological appearance, the differential diagnosis for ALK-rearranged RCC is broad and includes SMARCB1-deficient renal medullary carcinoma, FH-deficient RCC, MiTF RCC, and others. A correct diagnosis of ALK-RCC is critical as these patients may be tailored to targeted ALK inhibitors (alcetinib, crizotinib etc.), which are currently used in the treatment of ALK fusion lung carcinomas, thus another instance of targeted therapy being employed in renal tumors secondary to morphological and molecular correlations [100].
SMARCB1-deficient renal medullary carcinoma (and RCC with medullary phenotype)
SMARCB1 (INI1)-deficient renal medullary carcinoma (RMC) is a high-grade RCC arising in the distal nephron most frequently occurring in younger black patients with hemoglobinopathies (primarily sickle cell trait and hemoglobin SC disease). SMARCB1 encodes the tumor suppressor INI1, a member of the SWI/SNF chromatin remodeling complex. Cases morphologically compatible with SMARCB1-deficient RMC but lacking a history of hemoglobinopathy have been described; in this setting the terminology ‘unclassified RCC with medullary phenotype’ may be used [25, 101] (Table 1). SMARCB1-deficient RMCs have a diverse microscopic appearance with features ranging from reticular/yolk sac tumor-like and sieve-like/cribriform/microcystic/adenoid cystic-like patterns [102]. Thus, the differential diagnosis of SMARCB1-deficient RMC includes FH-deficient RCC, ALK-RCC, rhabdoid tumor, metastatic germ cell tumor, and high-grade urothelial carcinoma. Loss of nuclear expression of INI1 by IHC within a high-grade adenocarcinoma primary to the kidney is essential for diagnosis; clinical/laboratory evidence of hemoglobinopathy is seen in a majority of these cases [103].
RENAL NEOPLASIA WITH SPECIFIC MOLECULAR ABERRATIONS: POTENTIAL FUTURE IMPLICATIONS
RCC classification continues to evolve with the recognition of newer entities as well as a deeper understanding of the mechanisms underlying the biology of previously recognized tumor subtypes. New imaging technologies, integration of multi-omics approach, and artificial intelligence-assisted tumor interrogation continue to transform the landscape of renal medicine. Recently, the deconvolution of large organ systems (such as the kidney) into several distinct single-cell types, which can in turn be labeled by lineage-specific biomarkers, has been enabled by single-cell RNA-sequencing (ScRNA-seq) technology. With this technique, we may employ related biomarkers to map the whole nephron and describe the putative cell of origin for many of the renal tumor subtypes [38]. We can do this by leveraging molecular platforms used in diagnostic pathology, such as RNA ISH and next-generation IHC. Factors at interplay in RCC development and tumor biology include cell of origin, tumor microenvironment, copy number variations, signature genomic aberrations, metabolic perturbations, and others. Perhaps ‘oncogenic competence’ plays a critical role in determining which cells within the nephron are impacted (and to what extent) by aberrant genomic events, in the correct copy number variation background, within the correct intratumoral milieu [104]. As we continue to learn more about renal tumors, surgical pathologists will continue to play a central role in the diagnostic management of RCCs.
Footnotes
ACKNOWLEDGMENTS
The authors have no acknowledgments.
FUNDING
The authors report no funding.
AUTHOR CONTRIBUTIONS
Conceptualized by Rohit Mehra. Written by Rob Humble, Rahul Mannan and Rohit Mehra.
CONFLICTS OF INTEREST
RMH, RH and RM have no conflicts of interest to report.