Practical Molecular Testing in a Clinical Genitourinary Service

Renal cell carcinoma (RCC) is the most common primary kidney malignancy, and, as our understanding of the molecular underpinnings of RCC are better understood, many types of RCC are associated with, and even defined by, recurrent genomic abnormalities.

The most common type of RCC is clear cell RCC (CCRCC), which accounts for approximately 65% to 70% of all renal cancers.¹  Classically, CCRCC is composed of neoplastic cells with clear to eosinophilic cytoplasm arranged in nests that are surrounded by a delicately anastomosing vasculature with nuclei of variable grade. CCRCC usually occurs sporadically but may occasionally occur in association with a familial syndrome.^2–4  Sporadic CCRCC harbors genomic deletions and loss of heterozygosity on chromosome arm 3p, on which the VHL tumor suppressor gene is located.⁵  Although CCRCC can usually be diagnosed with routine histologic evaluation, possibly combined with immunohistochemistry (IHC), a fluorescence in situ hybridization (FISH) assay for 3p deletions is clinically available, which may be useful, albeit rarely, in challenging cases such as a biopsy with scant material.⁶  Notably, FISH analysis is unable to detect some cases of VHL loss, and molecular analysis is needed to identify copy number deletion, missense, and truncating mutations in VHL, which commonly occur in CCRCC.⁷ 

The second most common type of RCC is papillary RCC (PRCC), accounting for nearly 20% of RCCs.⁸  Papillary RCC is also further subdivided into type 1 and type 2 PRCC. Type 1 PRCC is characterized by fibrovascular cores lined by a single layer of neoplastic epithelium containing scant, basophilic cytoplasm, and usually possessing low-grade nuclei. In contrast, type 2 PRCC is characterized by pseudostratified layers of neoplastic cells lining fibrovascular cores, often with more abundant cytoplasm and higher-grade nuclei.¹  The difference between type 1 and type 2 PRCC is not only histologic, but also prognostic and genomic, as patients with type 2 PRCC typically have worse outcomes than patients with type 1 PRCC. Furthermore, type 1 PRCC is frequently associated with trisomy of chromosomes 7 and 17 and loss of chromosome Y,^9–12  whereas recent studies have shown that type 2 PRCC is characterized by other genetic alterations, including alterations in different chromosomes than observed in type 1 PRCC.^13,14  Indeed, type 2 PRCC may not even be a truly distinct entity, evidenced by the identification of other types of RCC in tumors previously diagnosed as type 2 PRCC. Nonetheless, there is currently no clinically available molecular assay to aid in the diagnosis of type 2 PRCC, but FISH for trisomy 7 and/or 17 could be used in cases in which type 1 PRCC is a consideration.¹⁵  Notably, type 1 PRCC can have considerable morphologic overlap with mucinous tubular and spindle cell carcinoma and clear cell papillary RCC, both of which lack trisomy 7 and 17; thus, FISH for trisomy 7 and/or 17 may be especially useful in this differential diagnosis.^16–18 

Translocation-associated RCC (tRCC) is a subtype of RCC defined by a translocation involving the TFE3 gene on chromosome band Xp11 with multiple gene partners (Figure 1, A and B) or the translocation t(6;11) creating the Alpha-TFEB gene fusion (Figure 2, A and B).^19–21  Morphologic features such as solid and papillary growth, voluminous cytoplasm, psammomatous calcifications, and extracellular basement membrane material are morphologic features that can aid in the diagnosis of tRCC.^19,20  Even so, there is considerable morphologic overlap between tRCC and other RCC subtypes, such as CCRCC and PRCC.^22,23  Immunohistochemistry is often useful to help differentiate tRCC from other subtypes of RCC, such as melanocytic markers or TFE3, but IHC is not entirely sensitive or specific for tRCC.^21,24,25  FISH analysis for TFE3 and TFEB gene rearrangements is highly sensitive and specific for tRCC and is extremely useful to confirm the diagnosis of tRCC.^23,25–27 

An exceptionally rare subgroup of RCC, which is characterized by ALK gene rearrangement, is included in the World Health Organization classification of kidney tumors as a provisional entity. These tumors, so-called ALK rearrangement–associated RCC, are exceptionally rare and may occur in children or adults.^28,29  They are often located in the renal medulla, and they are morphologically characterized by large polygonal or spindle cells with abundant eosinophilic cytoplasm and frequent intracytoplasmic lumina.²⁸  In these tumors, the ALK gene can have a number of fusion partners, and a FISH break-apart assay or a molecular assay can aid in the diagnosis.^30–33 

Another type of RCC that is defined by a specific genetic abnormality is succinate dehydrogenase (SDH)–deficient RCC, in which there is a double-hit inactivation of 1 of the SDH genes, most commonly SDHB, but also including SDHA, SDHC, or SDHD.^34–36  Morphologically, SDH-deficient RCC is characterized by a well-circumscribed or pushing border that may entrap benign tubules with scattered cysts containing eosinophilic material. The neoplastic cells have smooth nuclear contours with fine chromatin, lacking nucleoli. Most characteristically, the cytoplasm contains vacuoles or flocculent inclusions that contain eosinophilic or pale material (Figure 3, A).^34,37,38  Nevertheless, the classic morphologic features may be only focally present, and the diagnosis must be confirmed with IHC. Immunohistochemistry for SDHB is available for clinical use, and loss of SDHB expression as assessed by IHC confirms the diagnosis of SDH-deficient RCC (Figure 3, B). There are 2 important things to note in regard to SDHB IHC. First, SDHB IHC is not only lost in patients with loss of heterozygosity of SDHB but also in SDHC and SDHD, though some mutations in SDHD may result in weak expression of SDHB IHC, and mutations in SDHA may result in a positive SDHB IHC finding.^39–46  The risk of a positive SDHB IHC result in a patient with loss of heterozygosity of SDHA is mitigated by the relative rarity of SDHA mutations.⁴⁵  Thus, loss of SDHB expression by IHC confirms inactivation of an SDH gene but is not necessarily diagnostic of inactivation of the SDHB gene. Furthermore, this also means that SDHB IHC has a relatively high sensitivity for detection of SDH-deficient RCC. The second important thing to note is that in tumors with abundant clear cytoplasm, SDHB IHC expression may be markedly decreased, but it is not negative, and care should be taken to not misdiagnose these cases as SDH-deficient RCC.³⁸  Most patients with SDH-deficient RCC have a germline mutation in an SDH gene, and the second hit inactivating the SDH gene results in neoplasia.^34,36,47  Indeed, an autosomal dominant syndrome consisting of paraganglioma/pheochromocytoma, gastrointestinal stromal tumor, SDH-deficient RCC, and pituitary adenoma is described, and patients should be offered genetic testing if an SDH-deficient RCC is diagnosed.⁴⁸ 

Hereditary leiomyomatosis and RCC (HLRCC) syndrome is characterized by hereditary leiomyomatosis and a unique subtype of RCC. HLRCC-associated RCC typically possesses neoplastic cells arranged in tubular, tubulocystic, papillary (Figure 4, A), and solid growth patterns with large nuclei containing prominent “viral-inclusion–like” nucleoli (Figure 4, B), which may be focal.⁴⁹  HLRCC is caused by germline mutations in the fumarate hydratase (FH) gene.⁵⁰  Loss of heterozygosity in the wild-type FH gene locus eliminates FH protein function in the neoplastic cells, resulting in a metabolic shift from oxidative phosphorylation to aerobic glycolysis.^51,52  Immunohistochemistry for FH is available for clinical use and can aid in diagnosis of HLRCC-associated RCC.⁵³  Immunohistochemistry for 2-succino-cysteine, which accumulates in the cytoplasm of HLRCC-associated RCC, is more sensitive and specific than FH IHC, but this antibody is not currently commercially available.^54,55 

Although not yet codified into the World Health Organization classification of kidney tumors, a recently described group of RCCs harbors recurrent TCEB1 gene mutations.⁵⁶  These tumors characteristically are composed of neoplastic cells possessing voluminous clear cytoplasm with nested, tubular, and/or papillary architecture with admixed thick, fibromuscular bands transecting the tumor. Previously, these tumors were likely diagnosed as CCRCC or clear cell papillary RCC, both of which are the main differential diagnoses. To resolve this, IHC for cytokeratin 7 (CK7), CA-IX, and CD10 may be useful, as TCEB1-mutated RCC is expected to be positive for CA-IX, CK7, and CD10. Currently, there are no specific clinical assays available for confirmation of the diagnosis of TCEB1-mutated RCC.

In addition to HLRCC and SDH-deficient RCC, there are other familial RCC syndromes. These include von Hippel–Lindau syndrome (VHL gene on chromosome arm 3p), hereditary papillary RCC (MET gene on chromosome arm 7p), Birt-Hogg-Dubé syndrome (FCLN on chromosome arm 17p), and tuberous sclerosis (TSC1 gene on chromosome arm 9q and TSC2 gene on chromosome arm 16p). Patients with von Hippel–Lindau syndrome have an increased risk of developing CCRCC.^4,57  Patients with hereditary papillary RCC often present with multiple, bilateral PRCCs.⁵⁸  Patients with Birt-Hogg-Dubé syndrome usually develop multiple, bilateral kidney tumors, including a characteristic hybrid oncocytic tumor.⁵⁹  Patients with tuberous sclerosis often develop angiomyolipomas, but they can also develop morphologically unique RCCs.¹²  Although familial RCC syndromes are much less common than sporadic RCC, when recognized by the pathologist, patients may be referred for genetic testing and counseling.^5,60 

Urothelial carcinoma (UC) accounts for nearly all of the cancers of the urinary bladder, ureter, and renal pelvis. Although UC accounts for such a high proportion of neoplasia of the urinary tract, clinical outcome among patients with UC is highly variable, ranging from noninvasive low-grade papillary UC to deeply invasive and lethal high-grade UC. Accurate prognostic stratification and treatment decisions rely on pathologic examination of biopsy, resection, and cytology specimens. Nonetheless, molecular testing may occasionally aid morphologic evaluation and guide the therapeutic decision process.

Molecular oncology data suggest there are 2 separate molecular routes for UC, one resulting in superficial disease and the second resulting in deeply invasive disease.^61,62  The first route, which is associated with noninvasive or superficially invasive (ie, invasion into lamina propria) papillary UC, is composed of early, activating point mutations in FGFR3 or RAS family genes. A small subset of tumors in this group (∼15%) subsequently develops loss of function of 1 or more tumor suppressor genes, such as TP53, RB1, and PTEN. Acquisition of mutations in these tumor suppressor genes results in the potential for progression to deeply invasive disease. Alternatively, TP53, RB1, and PTEN mutations can occur in the absence of FGFR3 or RAS family genes, and this second route is typically associated with flat UC.

Although a few molecular diagnostic assays are available for prognosticating UC, these are not currently widely used in clinical practice. Rather, the most widely used molecular assay in regard to UC is likely UroVysion (Vysis Inc). UroVysion is a FISH assay performed on exfoliated cells in urine (Figure 5, A) that assesses for aneuploidy of chromosomes 3, 7, and 17 and loss of chromosome band 9p21, all of which are abnormalities characteristic of UC (Figure 5, B). UroVysion has a relatively high sensitivity and specificity for the presence of UC, and it has been implemented into many bladder cancer screening programs.⁶³  Additionally, UroVysion can be used to monitor patients with a history of UC or for stratification of patients with an abnormal cytology result and no clinical or cystoscopic evidence of a bladder tumor. However, more recent studies demonstrate that the sensitivity and positive predictive value of UroVysion, particularly for low-grade urothelial carcinoma, may not be as optimal as initially thought.^64–67 

Also commonly present in UC are mutations in the TERT promoter gene, with 60% to 80% of UCs harboring a point mutation (Figure 6, A and B).^68–70  These mutations appear to be early events in oncogenesis of UC, and they are present in both low-grade papillary UC and high-grade UC. However, TERT promoter mutations are not entirely specific for UC, as they are also present in benign urothelial neoplasms, such as urothelial papilloma, inverted papilloma, and papillary urothelial neoplasm of low malignant potential.^71,72  Furthermore, they are also frequently present in neoplasms of other organs, such as glioblastoma and melanoma, but they are notably absent in prostate cancer.^73–76  Thus, the practical implications of TERT promoter mutations in UC are likely in 3 situations: (1) the differential diagnosis of UC versus nonneoplastic urothelium (eg, cystitis glandularis); (2) the differential diagnosis of UC versus prostatic cancer, particularly in the setting of divergent differentiation; and (3) screening urine for recurrent UC. Although histomorphologic features of UC are usually diagnostic, some cases may include reactive urothelium in the differential; in this setting, a positive TERT promoter mutation test finding would weigh in favor of a neoplastic process over a reactive process, as reactive urothelium lacks TERT promoter mutations.⁷¹  A negative TERT promoter test result, however, does not exclude the possibility of an urothelial neoplasm, as not all cases of UC harbor TERT promoter mutations. Because prostate cancer lacks TERT promoter mutations, a positive TERT promoter test result would weigh in favor of UC over prostate cancer; an example of this is small cell carcinoma of the bladder, which has been demonstrated to harbor TERT promoter mutations in approximately half of cases, as it is clonally related to UC.^73,77,78  Finally, TERT promoter mutations have been identified in urine specimens, and it has been proposed that testing urine of patients with a history of UC for TERT promoter mutations may be an effective method of screening for recurrent UC.^79,80 

As UC becomes better understood, immunotherapy is likely to increasingly become a treatment option, and IHC performed on biopsy or resection material may guide therapeutic decision making. An example is the use of programmed death receptor-1 (PD-1) and programmed death ligand-1 (PD-L1) checkpoint inhibitors; indeed, atezolizumab, a PD-L1 inhibitor, is US Food and Drug Administration (FDA) approved for treatment of patients with chemotherapy-resistant advanced or metastatic UC.⁸¹  Trials are underway for similar drugs, including durvalumab, pembrolizumab, avelumab, nivolumab, and ipilimumab.⁸²  Thus, it is likely that testing for PD-L1 expression in biopsies and resections for UC may become more routine in the near future. Indeed, the FDA has already approved the use of several of these agents with IHC as a companion diagnostic test in select patients with UC.⁸³ 

Prostatic adenocarcinoma is the most common malignancy in men, excluding skin cancer, and the clinical course is highly variable, varying from indolent to aggressive, lethal disease in a relatively large subset of men.⁸⁴  Clearly, it is necessary to conduct patient stratification on the basis of the expected clinicopathologic behavior of these tumors, and this is done by using pathologic grade (ie, Gleason score and, more recently, Grade Groups) and stage (ie, AJCC [American Joint Committee on Cancer] Cancer Staging Manual, 8th edition).^85,86  Additionally, as the molecular underpinnings of prostatic adenocarcinoma are elucidated, molecular testing may play a role in the risk stratification and treatment of some of these patients.

The discovery of recurrent gene fusions involving the ETS family of transcription factors in nearly half of prostatic cancers was the first major breakthrough in understanding the molecular basis of these tumors.^87–89  TMPRSS2-ERG is the dominant ETS gene fusion found in prostatic adenocarcinoma, the result of which brings ERG expression under androgen control on the basis of androgen receptor (AR)–mediated TMPRSS2 transcriptional regulation. Thus, ETS gene fusions, particularly TMPRSS2-ERG, are the basis for the development of many diagnostic, prognostic, and therapeutic-driven assays. Most ETS prostatic adenocarcinoma rearrangements can be detected by FISH,^90,91  and IHC using an anti-ERG antibody, which detects the ERG gene fusion product, is also highly sensitive and specific for ERG aberrations.⁹²  Therefore, FISH or IHC for ERG may be occasionally useful in the diagnosis of prostatic adenocarcinoma versus a benign process (Figure 7, A through C). However, it should be noted that IHC for ERG may be positive in high-grade prostatic intraepithelial neoplasia (HGPIN), and overexpression of ERG is not pathognomonic of malignancy.⁹³  Another complication of ERG IHC is that in the castration-resistant state, ERG protein may not be overexpressed, even in the presence of ERG aberrations at the genomic level. In this setting, ERG FISH may be more helpful, such as in prostatic small cell carcinoma, which typically harbors an ERG aberration but often lacks detectable ERG protein expression.⁹⁴  Furthermore, it is also notable that approximately half of prostatic cancers lack an ERG rearrangement, and a negative ERG FISH or IHC result does not exclude the diagnosis of prostatic adenocarcinoma. Recently, Kunju et al⁹⁵  described an RNA in situ hybridization strategy for identifying ETS fusions involving genes other than ERG, such as ETV1, ETV4, and ETV5. The molecular underpinnings of ETS fusion-negative prostatic adenocarcinoma are also being studied, and abnormalities in SPINK1,⁹⁶  PTEN,⁹⁷  AURKA, MYCN,⁹⁸  SPOP, HOXB13, CHD1, and ADRB2^99–102  have all been implicated in the pathogenesis of prostate cancer. At present, however, these abnormalities hold little clinical relevance.

AR signaling status is an important factor in determining treatment for castration-resistant patients, as patients with active androgen signaling (ie, retained expression of AR and or AR-regulated genes, including ERG) are candidates for more aggressive androgen deprivation therapy, regardless of the genomic status of AR and TMPRSS2-ERG. Because tissue biopsies are often impractical to perform routinely for patients with castration-resistant prostate cancer, there has been a focus on the development of minimally invasive biomarkers from blood and urine.^103–106  The profiling of circulating tumor cells and cell-free DNA has revealed truncated AR splice variants, AR copy number gain, and AR point mutations, which are linked to castration resistance.^{103,104,107,108}  Specifically, AR splice variant 7 messenger RNA (AR-V7) in circulating tumor cells and copy number increase and point mutations in circulating cell-free DNA are associated with resistance to enzalutamide and abiraterone.¹⁰⁴ 

Gene expression profiling using DNA microarrays is difficult to perform on formalin-fixed, paraffin-embedded tissue and remains a more useful research tool than it is useful clinically. Nonetheless, several clinical assays have been developed, including Myriad Diagnostic's Polaris, Genomic Health's Oncotype DX Prostate Cancer Assay, and GenomeDx's Decipher. These assays may have some usefulness in prognostication of patients, but their clinical utility is not well defined at this time, and the clinical standard of prognostication remains standard clinicopathologic data.^109–111 

While FISH and IHC for ERG may be relevant to tissue biopsy or resection, there is also interest in the development of better and more specific biomarkers for prostate cancer surveillance. While serum protein-specific antigen (PSA), the current standard for prostate cancer screening, is sensitive for prostate cancer, it is not specific, as benign processes (eg, prostatitis) may increase serum PSA, and it may be increased in the setting of low-grade, indolent tumors. Because of this, assays that detect additional biomarkers associated with prostatic adenocarcinoma that are present in urine have been developed. As already discussed, TMPRSS2-ERG is present in a large subset of prostate cancers. Additionally, PCA3 is a noncoding RNA that is expressed in prostate cancer and HGPIN.^112–117  Several molecular assays for use on urine specimens—using some combination of TMPRSS2-ERG, PCA3, and PSA—are available for clinical use, including Mi-Prostate Score, which provides a likelihood of a patient having prostate cancer as well as the chance of that cancer being aggressive,¹¹⁸  and Progensa, which has a high negative predictive value.¹¹⁹  However, as the use of MRI-guided prostatic biopsy increases, the utility of prebiopsy molecular assays is likely decreasing. In addition to urine tests, there is also continued interest in developing better serum biomarkers. One such assay, which is commercially available, is 4Kscore (OPKO Health Inc), which measures 4 PSA-related molecules—total PSA, free PSA, intact PSA, and kallikrein-related peptidase 2—and is reported to predict risk of cancer more accurately than PSA alone.¹²⁰ 

An interesting subtype of prostate cancer is basal cell carcinoma (BCC).¹  Previously, it has been referred to as adenoid cystic carcinoma, owing to its morphologic similarity to adenoid cystic carcinoma of the salivary gland and other sites.¹²¹  Basal cell carcinoma of the prostate typically is composed of basaloid cells forming infiltrative nests and tubules, often with cribriform architecture, extracellular hyaline material, and perineural invasion (Figure 8, A). Recently, it was demonstrated that a subset of prostatic BCCs also mimics adenoid cystic carcinoma on the molecular level, harboring rearrangement of the MYB gene.¹²²  A subsequent study confirmed the presence of MYB-NFIB gene fusion in approximately half of prostatic BCCs (Figure 8, B), while prostatic basal cell hyperplasia and adenoma were ubiquitously negative for MYB gene rearrangements (Figure 8, C and D).¹²³  Thus, although clinical validation is still necessary, it is possible that FISH analysis may soon aid in the diagnosis of prostatic BCC.

Several germline mutations have been found to increase a man's risk of developing prostate cancer. Germline mutations in BRCA1 and BRCA2 not only predispose women to breast and ovarian cancer, but they also confer an increased risk of developing prostate cancer.^124–129  Additionally, men who harbor pathogenic germline BRCA2 mutations may have earlier onset of disease and decreased survival.^130–133  In addition to BRCA1/2 mutations, a germline mutation in HOXB13, although rare, increases the risk of prostate cancer.¹⁰⁰  However, there are currently no recommendations specifically for prostate cancer screening in men with BRCA1/2 or HOXB13 mutations.^134–136  In the absence of formal societal recommendations, Cheng et al¹³⁷  proposed a screening algorithm for such men.

Tumors of the testis can be broadly split into 2 categories: germ cell tumors (GCTs) and sex cord stromal tumors. Germ cell tumors account for most testicular neoplasms, particularly in young adult men. Although the morphologic spectrum of GCTs is broad, most GCTs (approximately 80%) harbor a copy number gain of chromosome arm 12p, usually present as isochromosome 12p [i(12p)],¹³⁸  and the presence of 12p abnormalities can be identified with FISH for chromosome arm 12p or with next-generation sequencing.^139,140  Although most testis GCTs can be diagnosed on the basis only of morphology with judicious use of IHC, there are rare situations in which ancillary testing to demonstrate a chromosome arm 12p abnormality may be useful. Situations in which this testing may be informative include resolving the differential diagnosis of prepubertal-type versus postpubertal-type teratoma (Figure 9, A and B); identifying a somatic-type malignancy as GCT derived (ie, secondary somatic malignancy; Figure 9, C and D)^141–143 ; diagnosis of metastatic GCT¹³⁹ ; and early detection of germ cell neoplasia in semen samples.^138,144  While most teratomas occurring in adult men are malignant, a subset of teratomas in adult men have recently been described that appear to behave indolently and lack i(12p).¹⁴⁵  Thus, demonstration of i(12p) would preclude a diagnosis of prepubertal-type teratoma and would imply malignant potential. Secondary somatic-type malignancies occur in approximately 5% of testis GCTs and can manifest as sarcoma, carcinoma, primitive neuroectodermal tumor, or leukemia; demonstration of i(12p) in the somatic malignancy is indicative of GCT origin.^141–143 

Most penile neoplasms are squamous cell carcinomas (SCCs), and, like SCC of the cervix and oropharynx, a subset of penile SCCs is associated with high-risk human papillomavirus (HPV) infection (Figure 10, A and B).¹⁴⁶  High-risk HPV DNA is integrated into the host cell genome, leading to production of the viral proteins E6 and E7, which inactivate TP53 and RB1 and increase expression of p16. Thus, HPV-associated SCC is typically p53⁻/p16⁺, whereas non-HPV–associated SCC is typically p53⁺/p16⁻, and IHC for p16 or in situ hybridization for high-risk HPV can be used to assess for HPV status (Figure 10, C).¹⁴⁷  While routine pathologic examination is the most reliable predictor of outcome,¹⁴⁸  confirming the presence or absence of high-risk HPV may be useful in select situations. Interestingly, while p16⁺ penile SCC has generally been shown to have better prognosis than p16⁻ penile SCC,¹⁴⁹  basaloid penile SCC is thought to be HPV driven and confers a worse prognosis than conventional penile SCC.¹⁵⁰ 

Nonetheless, relative to malignancies of other sites, the molecular underpinnings of penile SCC are poorly understood and are still to be elucidated. Recently, next-generation sequencing revealed frequent alterations in TP53, CDKN2A, PIK3CA, MYC, HRAS, and SOX2,¹⁵¹  alterations similar to those found in SCC of various other sites.¹⁵²  However, the clinical utility of these findings are currently limited. The penile SCC study found no significant associations between an individual gene's mutation status and tumor grade, stage, or histology.

The PD-1/PD-L1 immune checkpoint is a hot area of investigation as a potential target in malignancies of various sites (eg, lung carcinoma), and penile SCC is no exception. Indeed, it has been demonstrated that most non-HPV–related penile SCCs express PD-L1 as assessed with IHC, and this provides a rationale for targeted anti–PD-1 and anti–PD-L1 immunotherapy in penile SCC.¹⁵³ 

Immunotherapy has already been discussed in this review insofar that the PD-1/PD-L1 immune checkpoint may be clinically relevant in some genitourinary malignancies. Proper patient selection for immunotherapy is critical, as not all patients demonstrate response to these agents. Direct assessment for PD-1/PD-L1 expression via IHC for PD-L1 is discussed above and has been evaluated in many clinical trials.¹⁵⁴  Alternative methods for determining patient selection for immunotherapy include tumor mutational burden (TMB) and microsatellite instability (MSI). Indeed, TMB has been demonstrated to be more significantly associated with response to anti–PD-1 and anti–PD-L1 immunotherapy than IHC for PD-1/PD-L1 expression, but TMB analysis is not currently widely available for clinical use.^155,156  Microsatellite instability, on the other hand, is more readily available for clinical use than TMB, either through IHC for mismatch repair enzymes (ie, MLH1, PMS2, MSH2, and MSH6) or molecular analysis. Because MSI-high (MSI-H) tumors are unstable and hypermutational, they tend to express high levels of checkpoint proteins, such as PD-1 and PD-L1.^157,158  Clinical trials in a variety of tumors have demonstrated that MSI-H tumors are associated with response to immunotherapy, and, in May 2017, the FDA broadly approved pembrolizumab for use in all pediatric and adult patients with MSI-H or mismatch repair-deficient solid tumors.^157–159  Thus, although MSI-H is a relatively rare event in genitourinary malignancies, there are instances in which testing for MSI may be clinically beneficial.^155,160 

Clearly, molecular understanding of genitourinary neoplasms has advanced to the point of being able to incorporate some molecular assays into clinical use. Recurrent genomic abnormalities have been identified in many types of RCC, and some types of RCC are specifically defined by the molecular abnormality. Two major routes of developing UC have been molecularly described. Recurrent translocations involving ETS family genes are found in approximately half of prostate cancers. Testicular GCTs typically harbor i(12p). Penile neoplasms are often high risk and HPV driven. Nonetheless, even as genitourinary neoplasms are increasingly better understood at the molecular level, further research with eventual clinical validation is needed for optimal diagnosis, prognosis, and treatment of aggressive malignancies in the genitourinary tract.