Comparative genomic characterization of small cell carcinoma of the bladder compared to urothelial carcinoma and small cell lung carcinoma

Abstract

Purpose:

Small cell carcinoma of the bladder (SCBC) is a rare histologic variant of bladder cancer with an aggressive disease course and poor outcomes. Given its uncommon nature, there is a paucity of high-quality data characterizing genomic drivers of this disease, and most patients are treated with approaches mirroring small cell lung cancer (SCLC). Leveraging the Tempus LENS de-identified clinically annotated genomic dataset, we sought to evaluate the mutational landscape of SCBC relative to urothelial carcinoma (UC) and SCLC.

Methods:

Somatic pathogenic genomic alterations in patients with SCBC, UC, and SCLC of any stage who underwent blood or tissue-based genomic profiling through the Tempus assay were catalogued. Baseline clinical and demographic features were compared across histologic groups. Alterations were collated and summarized using descriptive statistics. Pairwise comparisons were performed to assess differences in mutation frequency across pathologic cohorts.

Results:

In total, 149 SCBC, 4350 UC, and 1697 SCLC cases were included in the study. The most common genomic alterations in SCBC were in TP53 (87%), TERT (75%), and RB1 (70%). Among SCBC patients with TP53 mutations, RB1 co-mutations were observed in 77% of cases. Compared to UC, SCBC cases were significantly enriched for TP53, RB1, KMT2D, and KDM6A mutations. Compared to SCLC, SCBC cases were enriched for TERT, ARID1A, and CREBBP mutations, among others (p<0.05). Multiple clinically targetable mutations were observed in SCBC, including PIK3CA (19%), ERBB2/3 (13%), and ALK (10%). Limitations of this study include its retrospective nature

Conclusion:

This study represents one of the most extensive efforts to characterize SCBC to date, providing a novel understanding of the genomic alterations underlying the disease and revealing actionable mutations that could serve as potential targets for improved clinical outcomes.

INTRODUCTION

Small cell bladder cancer (SCBC) is a rare histologic variant that accounts for less than one percent of all bladder cancer cases diagnosed annually ⁽¹⁾. Histologically, SCBC is a subset of neuroendocrine carcinoma and is often characterized by the expression of neuron-specific enolase, chromogranin, and synaptophysin ⁽²⁾. The clonal origin of SCBC remains unclear, with previous work demonstrating concordant allele loss between urothelial carcinoma (UC) and SCBC samples in patients with both tumor types, suggestive of a common cell of origin ^(2–4). Next-generation sequencing and tumor amplification of SCBC samples has yielded loss of RRM1, MGMT, and TS, as well as high expression of ERCC1 and MRP1 as molecular signatures ^(5,6). Additional studies have suggested that Kulchistky-type neuroendocrine cells present in bladder tissue could be the cellular predecessors to SCBC ⁽⁷⁾.

Even when diagnosed at a clinically localized stage, patients with SCBC tend to have a significantly worse prognosis compared to that of UC, with a reported 5-year survival as low as 14% versus 71%, respectively ^(1,8). Genetic alterations in TP53, cMET, RB1, FBXW7, and PTEN could partly explain its aggressive course ⁽⁹⁾. The treatment paradigm for UC has changed dramatically in recent years, with regulatory approvals of antibody-drug conjugates, immune checkpoint inhibitors (ICI), and targeted agents ^(10–16). In contrast, the treatment of SCBC is limited to cytotoxic chemotherapy regimens with limited efficacy ⁽¹⁷⁾. For patients with localized SCBC, concurrent chemoradiotherapy and/or perioperative chemotherapy remains the standard of care ⁽¹⁷⁾. Despite curative intent therapy, recurrences are common, and systemic cytotoxic chemotherapy extrapolating from small cell lung cancer (SCLC) studies, given the clinical and histologic commonalities of these neuroendocrine tumor types, continues to be the cornerstone of therapy ⁽¹⁸⁾.

Genomic profiling of SCLC tumoral samples has yielded a widely heterogeneous landscape, with some overlaps in the molecular aberrations described for SCBC. Mutations in TP53, RB1, ATM, cMET, PTEN, BRAF, SMAD4, and KRAS have been described ⁽¹⁹⁾. Although the treatment of SCBC has traditionally mirrored that of SCLC, an evaluation of the comparative biological similarities between both entities remains largely unexplored. Further, given the rarity of SCBC, it has historically been excluded from concerted large-scale studies aimed at genomic characterization of bladder carcinoma ^(20,21). Leveraging an extensive repository of molecular data from SCBC specimens, this retrospective study herein catalogues clinical and molecular data from patients with SCBC, UC, and SCLC to provide comparative genomic characterization and evaluate the prevalence of potentially actionable targets for therapeutic intervention in this rare tumor subtype.

METHODS

The de-identified Tempus Lens dataset was queried (September 2023) using filters for patients diagnosed with UC, small cell carcinoma of the renal pelvis, bladder or ureter, or SCLC of any stage (T1–4, N0–2, M0–1) with clinical metadata and somatic mutation analysis with or without germline sequencing data. Somatic genomic alterations identified by tissue (xTT, xO, and xE) from primary or metastatic site or blood-based (xF) next-generation sequencing assay, ⁽²²⁾ along with baseline demographic characteristics such as gender, age, race, ethnicity, and smoking status were catalogued from the Tempus Lens dataset. Patients were grouped into three groups based on underlying pathology: UC, SCBC, and SCLC. Clinical variables were summarized using descriptive statistics and compared between the groups (UC versus SCBC and SCBC versus SCLC) using a Fisher’s exact test with an alpha level of 0.05 used for significance. Genomic alterations in each group were summarized using descriptive statistics. Most frequently altered genes in SCBC were selected and compared pairwise with UC and SCLC using a two proportions Z-test with a significance level of 0.05 after correction for multiple hypothesis testing to control for false discovery rate. The most frequently altered genes in SCBC were analyzed for co-mutations with other frequently altered genes using descriptive statistics. All statistical analyses were performed in R Statistical Software, version 4.2.1.

Institutional Review Board approval

Institutional review board (IRB) approval was in place for this study (IRB No. 17–001313), and patient informed consent has been waived because of the retrospective nature of the study.

RESULTS

A total of 5,479,928 patients were included in the Tempus Lens database at the time of query (August 2023), including 149 patients with SCBC, 4,350 with UC, and 1,697 with SCLC who met inclusion criteria. Baseline demographic variables of the analyzed cohort are provided in Table 1. No significant differences were observed with regards to gender between UC and SCBC, with 72.7% (n=3,164) and 75.2% (n=112) of patients being male, respectively (P=0.57). In contrast, SCBC had a significantly lower proportion of females (24.8%) compared to SCLC (50.7%; P<0.00001). The median age of patients in the SCBC, UC, and SCLC cohorts were 71, 71, and 67 years, respectively (p>0.05 for all comparisons). Additionally, no significant differences were detected regarding race, ethnicity, or smoking status between SCBC, UC, and SCLC.

Table 1: Demographic data of patients with urothelial carcinoma (UC), small cell carcinoma of the bladder (SCBC), and small cell carcinoma of the lung (SCLC).

Demographic Variables; n (%)	UC	SCBC	SCLC
Gender
Male	3,164 (72.7%)	112 (75.2%)	837 (49.3%)
Female	1,185 (27.2%)	37 (24.8%)	860 (50.7%)
Unknown	1 (0.0%)	0 (0.0%)	0 (0.0%)
P Value	0.57 (vs. SCBC)	Reference	<0.00001 (vs. SCBC)
Age (years)
≤60	748 (17.0%)	28 (18.8%)	409 (24.1%)
>60	3,133 (72.0%)	106 (71.1%)	1,241 (73.1%)
Unknown	469 (10.8%)	15 (10.1%)	47 (2.8%)
Median Age (Range)	71 (15 – 90+)	71 (31 – 90+)	67 (16 – 90+)
P value	0.65 (vs. SCBC)	Reference	0.34 (vs. SCBC)
Race
White	2,253 (51.8%)	75 (50.3%)	984 (58.0%)
Black/African American	202 (4.6%)	5 (3.4%)	117 (6.9%)
Asian	100 (2.3%)	3 (2.0%)	15 (0.9%)
American Indian or Alaska Native	8 (0.2%)	0 (0.0%)	5 (0.3%)
Native Hawaiian/Pacific Islander	3 (0.1%)	0 (0.0%)	0 (0.0%)
Race not stated/unknown	1,784 (41.0%)	66 (44.3%)	576 (33.9%)
P value	0.94 (vs. SCBC)	Reference	0.19 (vs. SCBC)
Ethnicity
Hispanic or Latino	119 (2.7%)	6 (4.0%)	33 (1.9%)
Not Hispanic or Latino	1,408 (32.4%)	47 (31.5%)	638 (37.6%)
Unknown	2,823 (64.9%)	96 (64.4%)	1,026 (60.5%)
P value	0.03 (vs. SCBC)	Reference	0.05 (vs. SCBC)
Smoking Status at Diagnosis
Ex-smoker	1827 (42.00%)	61 (40.94%)	884 (52.09%)
Never smoker	1053 (24.21%)	36 (24.16%)	510 (30.05%)
Current smoker	582 (13.38%)	11 (7.38%)	110 (6.48%)
Unknown	888*(20.40%)	41 (27.50%)	193 (11.40%)
P value	0.18 (vs. SCC-B)	Reference	0.50 (vs. SCC-B)
Total	4,350	149	1,697

The most common genomic alterations observed in SCBC group were TP53 (87.3%), TERT (75.2%), RB1 (69.8%), ARID1A (34.2%), KMT2D (24.2%), KMT2C (22.8%), CREBBP (20.1%), SYNE1 (20.1%), KDM6A (20.1%), and PIK3CA (18.8%; Table 2). Patients with these mutations were analyzed for the presence of the other most frequently occurring mutations (Figure 1). The most frequently co-occurring somatic mutations with TP53 include TERT (83.0%), RB1 (76.9%), and ARID1A (38.5%). Other than TP53, the most frequently co-occurring somatic mutations with TERT included RB1 (79.4%) and ARID1A (40.2%). Notable co-occurring mutations with RB1 included CREBBP (83.3%) and KDM6A (93.3%). The prevalence of clinically significant gene alterations was comparable between tissue- and blood-based sequencing with TP53 (95.9%), TERT (81.0%), RB1 (74.3%), ARID1A (21.6%), and CDKN1A (17.5%) being the most frequently identified. No notable discrepancies in mutation prevalence were observed between tissue- and blood-derived samples.

Table 2: Comparison of most common genomic alterations in patients with urothelial carcinoma (UC), small cell carcinoma of the bladder (SCBC), and small cell carcinoma of the lung (SCLC).

Somatic Variant Genes	UC		SCBC		SCLC		SCBC vs. UC	SCBC vs. SCLC
	N	%	N	%	N	%	P Value	P Value
TP53	2624	60.32%	130	87.25%	1499	88.33%	<0.00001	0.79
TERT	3201	73.59%	112	75.17%	105	6.19%	0.73	<0.00001
RB1	988	22.71%	104	69.80%	1140	67.18%	<0.00001	0.57
ARID1A	1220	28.05%	51	34.23%	209	12.32%	0.11	<0.00001
KMT2D	1425	32.76%	36	24.16%	394	23.22%	0.03	0.87
CREBBP	622	14.30%	30	20.13%	183	10.78%	0.06	0.0009
KDM6A	1265	29.08%	30	20.13%	79	4.66%	0.02	<0.00001
PIK3CA	894	20.55%	28	18.79%	112	6.60%	0.67	<0.00001
LRP1B	699	16.07%	25	16.78%	639	37.65%	0.90	<0.00001
ERCC2	425	9.77%	22	14.77%	42	2.47%	0.06	<0.00001
EP300	545	12.53%	22	14.77%	94	5.54%	0.49	0.00001
CDKN1A	388	8.92%	22	14.77%	9	0.53%	0.02	<0.00001
KMT2A	522	12.00%	21	14.09%	133	7.84%	0.51	0.01
ZFHX3	429	9.86%	21	14.09%	143	8.43%	0.12	0.02
ERBB3	474	10.90%	20	13.42%	47	2.77%	0.40	<0.00001
ERBB2	802	18.44%	19	12.75%	68	4.01%	0.09	<0.00001
MYCL	216	4.97%	17	11.41%	115	6.78%	0.0009	0.05
PTEN	256	5.89%	17	11.41%	206	12.14%	0.009	0.89
GRIN2A	245	5.63%	15	10.07%	145	8.54%	0.03	0.63
CDKN2A	1395	32.07%	14	9.40%	171	10.08%	<0.00001	0.90
MKI67	292	6.71%	14	9.40%	85	5.01%	0.26	0.03

Figure 1: Co-occurring somatic gene variants in small cell carcinoma of the bladder (SCBC).

Clinically actionable mutations were noted with varying frequency in SCBC cases. FGFR3 alterations were noted in only 6% of cases, but 14% of patients had an alteration in FGFR1–4. DNA damage repair alterations were seen in a significant proportion of patients, which included BRCA1/2 (15.4%), ATM (12.0%), BRIP1 (8.7%), CDK12 (4.7%), PALB2 (4.0%), and RAD51 (5%). Mismatch repair gene alterations (MSH2/6 and PMS2) were registered in 9.4% of patients. Additional targetable alterations identified amongst the SCBC population included PI3KCA (18.9%), ERBB-2 (12.6%), ERBB-3 (13.4%), and PTEN (11.4%). With several novel therapeutic agents targeting these genomic alterations currently in development (Table 3), these targetable alterations in SCBC provide novel therapeutic opportunities.

Table 3. Select clinical trials evaluating novel therapies targeting clinically actionable mutations.

Alteration	Drug	Study phase	Name/NCT	Disease setting
FGFR1–4	Erdafitinib	Phase 3	THOR (NCT03390504)	mUC
		Phase 2	THOR-2 (NCT04172675)	High-risk NMIBC
	Futibatinib	Phase 1/2	NCT02052778	Advanced solid tumors
	LOXO-435	Phase 1	NCT05614739	Advanced solid tumors
DLL/NOTCH 1–4	Rovalpituzumab tesirine*	Phase 1	NCT02819999	SCLC
	MK-6070*	Phase 1/2	NCT04471727	Advanced solid tumors expressing DLL3
	Enoticumab**	Phase 1	NCT00871559	Advanced solid tumors
	Tarextumab	Phase 1/2	PINNACLE (NCT01859741)	SCLC
	Brontictuzumab	Phase 1	NCT01778439	Relapsed or refractory solid tumors
HER(ERBB)-2	Disitimab vedotin	Phase 2	KEYNOTE-D78 (NCT04879329)	HER-2 expressing UC
		Phase 3	DV-001 (NCT05911295)	Locally advanced or mUC
		Phase 2	DV-005 (NCT06003231)	Solid tumors that express HER-2
	Trastuzumab emtansine	Phase 2	TAPISTRY (NCT04589845)	(Cohort F) Locally advanced or metastatic solid tumors with HER-2 mutations
		Phase 2	KAMELEON (NCT02999672)	(Cohort 1) HER-2 positive urothelial bladder cancer
		Phase 2	NCT02675829	Solid tumors with HER-2 mutations
PI3KCA	Buparlisib	Phase 2	NCT01551030	Platinum refractory mUC
	Alpelisib	Phase 1	NCT01928459	Solid tumors with PIK3CA mutation
	Inavolisib	Phase 2	TAPISTRY (NCT04589845)	(Cohort H) Locally advanced or metastatic solid tumors with PIK3CA mutations
PTEN	Ipatasertib	Phase 2	NCT05554380	Solid tumors with PTEN/AKT mutations

^*Targets DLL3, which is part of the NOTCH pathway.

^**Targets DLL4, which is part of the NOTCH pathway.

mUC, metastatic urothelial carcinoma; NMIBC, non-muscle invasive bladder cancer; SCLC, small-cell lung cancer; UC, urothelial carcinoma.

A comparison of somatic variants between SCBC and UC demonstrated significant differences between the two groups for multiple genes of interest (Figure 2). Compared to UC, the prevalence of alterations in TP53 (87.3% vs. 60.3%, P<0.00001), RB1 (69.8% vs. 22.7%, P<0.00001), CDKN1A (14.8% vs. 8.9%, P=0.02), MYCL (11.4% vs. 4.9%, P=0.0009), PTEN (11.4% vs. 5.9%, P=0.009), and GRIN2A (10.0% vs. 5.6%, P=0.03) was significantly higher in the SCBC group. In contrast, alterations in KDM6A (20.1% vs. 29.0%, P=0.02), KMT2D (24.2% vs. 32.7%, P=0.03), and CDKN2A (9.4% vs. 32.0%, P<0.00001) were less frequent in the SCBC group compared to UC. Interestingly, both groups also demonstrated shared genomic characteristics, and several genomic alterations known to be frequently present in UC were also prevalent in SCBC, including TERT (75.1% vs. 73.6%), ARID1A (34.2 vs. 28.0%), and ERCC2 (14.8% vs. 9.8%).

Figure 2: Prevalence of the 30 most common genomic alterations amongst small cell carcinoma of the bladder (SCBC), small-cell carcinoma of the lung (SCLC), and urothelial carcinoma (UC).

When comparing genomic alterations between SCBC and SCLC, the prevalence of TP53 (87.3% vs. 88.3%; P=0.79), RB1 (69.8% vs. 67.18%; P=0.57), and PTEN (11.4% vs 12.1%; P=0.89) alterations was similar between the two groups. In contrast, mutations in TERT were almost exclusive to the SCBC group (75.2% vs. 6.2%, P<0.0001). Similarly, ARID1A (34.3% vs. 12.3%, P<0.00001), CREBBP (20.1% vs. 10.8%, P=0.001), KDM6A (20.1% vs. 4.7%, P<0.00001), PIK3CA (18.8% vs. 6.6%, P<0.00001) and ERCC2 (14.8% vs. 2.5%, P<0.00001) were significantly higher in the SCBC group compared to SCLC. NOTCH1 mutations in SCLC were detected at a significantly higher rate of 15.7% vs. 6.0% in SCBC (P=0.002), without significant differences in the prevalence of NOTCH2–4 mutations. All detected genomic variants in SCBC, UC, and SCLC are shown in Figure 2.

DISCUSSION

This study represents one of the most extensive efforts to characterize SCBC to date, providing a genomic comparison with UC, given their shared organ of origin, and SCLC, given their similar treatment strategies and presumed biological similarity. Our analysis revealed both significant genomic overlap alongside potentially clinically meaningful as well as notable genomic differences, which could contribute to their divergent clinical behaviors. A significant proportion of SCBC tumors also had potentially actionable mutations that could inform future therapeutic strategies.

It is notable that the distribution of age, race, ethnicity, and smoking status was not significantly different between SCBC, UC, and SCLC groups. Most patients in all three groups were older, of male gender, and had either a prior or current smoking history, which likely reflects relatively similar occupational and environmental exposures linked to these diseases. However, gender was statistically significant between groups, with female gender depicting a higher prevalence of SCLC than UC and SCBC. Both SCBC and SCLC had a high frequency of co-occurring mutations in TP53 and RB1, which are a signature of poorly differentiated neuroendocrine carcinomas, including those arising from the bladder ⁽²³⁾. This prevalence of co-occurring mutations has been described in SCLC cohorts, as well as prior smaller studies in SCBC, and likely reflects the underlying loss of G1-to-S phase cell cycle regulation, which has been described in small cell carcinoma irrespective of origin ^(24,25). Interestingly, co-occurring mutations in TP53 and RB1 were also enriched in the “neuronal” or “neuroendocrine-like” molecular subtypes within UC and correlated with aggressive disease course and worse prognosis, but also simultaneously higher responsiveness to ICI ⁽²⁶⁾. In SCLC, ICI in combination with chemotherapy has already demonstrated improvement in overall survival in the front-line setting ^(27,28). Given these biological similarities seen between SCBC and SCLC in ours and other studies, clinical trials such as the South-west Oncology Group (SWOG) S2012 (NCT05058651) investigating combination chemo-immunotherapy in extrapulmonary neuroendocrine carcinoma, including SCBC, will play a crucial role in testing the hypothesis that these ICI-based combination strategies can improve treatment outcomes for patients diagnosed with this rare disease.

In addition to TP53 and RB1, several additional genomic alterations (e.g. TERT, KMT2D, ARID1A, CREBBP, PIK3CA, EP300) were noted in SCBC, which are also commonly found in UC (21). Both SCBC and UC had a similar high prevalence of TERT mutations (>70% in both), which was not observed in the SCLC cohort (6.1%). The prevalence of TERT alterations seen in our study is in line with that reported previously in UC, SCLC, and SCBC ^(23,29,30). Considering TERT alterations are early genomic events observed in both low- and high- grade bladder tumors, their concordant prevalence in UC and SCBC suggests similar clonal cells of origin for both histologic subtypes, as has previously been reported ⁽³¹⁾. The prevalence of clinically significant gene alterations was comparable between tissue- and blood-derived samples, suggesting that sample type does not introduce a confounding variable in detecting genomic alterations within our cohort.

Our analysis also revealed several additional genomic alterations of potential therapeutic significance. In contrast to prior studies, we found a significant proportion of SCBC tumors harboring actionable genomic alterations. Although FGFR-3 alterations were uncommon, any FGFR mutation was seen in 14% of patients. Considering the established efficacy of pan-FGFR inhibitors such as erdafitinib in UC and other tumor types, these findings warrant evaluation of FGFR inhibition in SCBC ⁽³²⁾. HER(ERBB)-2 alterations were noted in 12.6% of SCBC tumors. Antibody-drug conjugates (ADCs) targeting HER-2 have demonstrated encouraging efficacy in HER-2 over-expressing tumors ⁽³³⁾. Given the frequency of HER-2 alterations noted in our dataset, further evaluation of HER-2 expression and the role of HER-2 targeted ADCs in SCBC is warranted.

Additionally, a significant proportion of tumors had alterations in NOTCH 1–4. NOTCH signaling pathway is involved in tumor cell neuroendocrine differentiation, proliferation, and even chemoresistance ⁽³⁴⁾. In a prior study of 110 SCLC samples, NOTCH family mutations were shown to have tumor-suppressive properties in SCLC ⁽³⁵⁾. Lurbinectedin is approved after progression on first-line therapy for advanced SCLC, with a recent study showing that drug efficacy is in part driven by modulation of NOTCH activity ^(18,36). One case report of lurbinectedin in metastatic SCBC demonstrated a metabolic complete response after first-line platinum-based chemotherapy ⁽³⁷⁾. Delta-like ligand 3 (DLL3), a ligand of the NOTCH signaling pathway, is characteristically overexpressed on the surface of tumor cells, making it an attractive therapeutic target ⁽³⁸⁾. Several DLL3-targeting therapies, including antibody-drug conjugates, T-cell engagers, and chimeric antigen receptor (CAR)-T cells, are currently under evaluation ⁽³⁸⁾. These findings provide preliminary evidence of the NOTCH pathway as a therapeutic vulnerability for patients with SCBC harboring NOTCH1 mutations. Further investigation of these findings could yield clinical benefits in patients with SCBC and potentially other undifferentiated neuroendocrine carcinomas.

Limitations of this study include its retrospective nature. Furthermore, given that data is extracted from a large database of patients who have undergone genomic testing, there may be a bias towards those with disease presenting to centers where such testing is available and done routinely. Extensive clinical information such as ECOG performance status, clinical stage, and survival were limited to a small number of patients, preventing survival analysis adjusted for these clinical parameters. It is possible that amongst patients classified as having either UC or SCBC, an element of the other histology was present, given the high rates of co-occurrence of these diseases in patients, which could have influenced genomic findings ^(23,39). Lastly, the impact of these genomic alterations on oncologic outcomes with contemporary chemotherapy and/or immunotherapy could not be evaluated, which could add additional prognostic value, due to lack of data on treatment in the database.

CONCLUSION

This study yields the largest genomic characterization of SCBC to date, revealing a relatively unique genomic profile with significant overlap and differences between SCBC and UC. Importantly, these results indicate a significant proportion of SCBC tumors harbor clinically actionable alterations. Inclusion of SCBC in biomarker-selected tumor agnostic trials is imperative to further outcomes for this rare histologic variant of bladder cancer.

Context.

Key Objective:

To compare the genomic mutational landscape of small cell bladder cancer (SCBC) relative to urothelial carcinoma (UC) and small cell lung cancer (SCLC)

Knowledge Generated:

Our findings reveal a significant mutational overlap between tumor histologies, with SCBC and SCLC exhibiting frequent co-occurring TP53 and RB1 mutations—hallmarks of poorly differentiated neuroendocrine carcinomas. Additionally, we identified several clinically actionable mutations, including FGFR3, HER(ERBB)-2, NOTCH1–4, PTEN, and PIK3CA. These insights may inform future research on targeted therapeutic strategies for SCBC, a relatively underexplored area in oncology.

Relevance:

There is a significant overlap of genomic alterations between SCBC and SCLC tumor histologies. Actionable mutations in this space could inform future therapeutic efforts.