Prevalence and Co-Occurrence of Actionable Genomic Alterations in High-Grade Bladder Cancer

Authors

Iyer G

Al-Ahmadie H

Schultz N

Hanrahan AJ

Ostrovnaya I

Balar AV

Kim PH

Lin O

Weinhold N

Sander C

Zabor EC

Janakiraman M

Garcia-Grossman IR

Heguy A

Viale A

Bochner BH

Reuter VE

Bajorin DF

Milowsky MI

Taylor BS

Solit DB

Doi

10.1200/JCO.2012.46.5740

PMID: 23897969 · DOI: 10.1200/JCO.2012.46.5740 · Journal: Journal of Clinical Oncology (2013)

TL;DR

An integrated genomic analysis of 97 high-grade urothelial bladder tumors profiled by array comparative genomic hybridization (Agilent 1M aCGH), mass-spectrometry–based Sequenom genotyping, Sanger sequencing of 15 selected oncogenes/tumor suppressors, and Illumina HT-12 expression arrays. 61% of tumors harbored a potentially actionable genomic alteration. Unsupervised clustering of CNAs split bladder tumors into a high-CNA-burden and low-CNA-burden class; TP53 and RB1 alterations were significantly enriched in the high-burden class (P<.001 and P<.003). Mutations in the RTK–RAS–RAF and PI3K/AKT/mTOR pathways and in regulators of G1-S cell-cycle progression occurred in a largely mutually exclusive distribution, and the AKT inhibitor MK-2206 selectively inhibited bladder cancer cell lines as a function of their underlying pathway lesion (PIK3CA/AKT1 sensitive; TSC1-null resistant). The data underwrite the blca_mskcc_solit_2012 cBioPortal study.

Cohort & data

  • 97 high-grade urothelial carcinomas of the bladder (BLCA); cystectomy 94, nephroureterectomy 1, TURBT 2; macrodissected to ≥60% tumor content; median age 73 (range 42–89); 72 male / 25 female.
  • Stage distribution (95 tumors): pTa/pTis 4, I 11, II 15, III 33, IV 32; 34 (35%) received neoadjuvant chemotherapy.
  • Histology: 57 (59%) transitional cell carcinoma NOS; 30 (31%) TCC with minor variant components; 10 (10%) predominantly neuroendocrine.
  • Comparator cohort: 285 consecutively treated age-, sex-, stage-matched bladder cancer patients (cystectomy at MSKCC 2007–2010), with comparable median overall survival (38 vs 35.2 months; P=.19).
  • Dataset: blca_mskcc_solit_2012 — publicly available via cBioPortal.
  • Reference genome: NCBI build 36.1 (hg18).
  • Assays: array-cgh-agilent-1m (Agilent 1M human oligonucleotide aCGH, cohybridized with reference normal DNA; segmented by circular binary segmentation; analyzed with the RAE algorithm at FDR <1%); sequenom-genotyping (mass-spectrometry iPLEX hotspot panel); sanger-sequencing of all coding exons of 15 selected oncogenes/tumor suppressors; illumina-microarray (Illumina HumanHT-12 expression BeadChip, quantile-normalized).

Key findings

  • 61% (59/97) of high-grade bladder tumors carried at least one potentially actionable alteration (defined as a clinically validated drug target in another tumor type or a target with an inhibitor in clinical investigation).
  • Unsupervised hierarchical clustering of aCGH data identified two distinct subsets — high-CNA-burden vs low-CNA-burden — whose CNA load differed but which were not clinically distinct (recurrence-free, OS, cancer-specific survival P=.98, .75, .49). The high-burden bladder subset’s structural-aberration load was second only to serous ovarian among 14 reference tumor types (n=5,135).
  • TP53 mutations (34% overall) and RB1 alterations (15% overall) were significantly enriched in high-CNA-burden tumors (P<.001 and P<.003).
  • 65% (63/97) of tumors harbored at least one somatic mutation among the panel-tested genes; most prevalent: TP53 34%, PIK3CA 18%, FGFR3 13%.
  • MAPK pathway: altered in 35% of samples; the most commonly altered RTK/RAS/RAF genes were FGFR1, FGFR3, ERBB2, MET, NF1, KRAS, BRAF, each in 2%–13% of cases, and arose in a predominantly mutually exclusive pattern. ERBB2 was focally amplified in 6 of 97 (6.2%) tumors with concordant mRNA up-regulation and 3+ HER2 IHC overexpression.
  • TP53 mutation and MDM2 amplification (5%) occurred in a nonoverlapping distribution; combined alterations of these two genes affected 38% of samples.
  • G1–S cell-cycle regulators were altered in 61% (59/97); focal amplifications of CCND1 (14%) and CCNE1 (5%) were mutually exclusive and inversely correlated with RB1 mutations/deletions. E2F3 amplifications occurred in 21% of bladder tumors versus 4.9% of 1,932 non-urothelial epithelial tumors and were associated with stage III–IV disease (28% vs 6.7% in stage I–II; P=.028).
  • An E2F3-target expression signature (23-gene score) was highest in tumors with E2F3 amplification or RB1 mutation/homozygous deletion; intermediate in CCND1/CCNE1-amplified tumors; lowest in CDKN2A-altered or all-pathway–wild-type tumors.
  • CDKN2A/CDKN2B homozygous deletion at 9p21.3 was the most common deletion event (CDKN2A altered in 24–25% of samples).
  • PI3K/AKT/mTOR pathway altered in 30% (29/97): PIK3CA mutation 18%, TSC1 mutation 7%, PTEN deletion 4%, AKT1 mutation 2%.
  • Tumors with neuroendocrine differentiation were enriched for E2F3 amplification (50% vs 17%; P=.03), RB1 deletion/mutation (50% vs 13%; P=.01), TP53 alteration, and higher overall CNA burden.
  • Bladder cancer cell line panel showed only modest overlap with primary-tumor genomics — notably, no cell line had focal ERBB2 amplification.
  • MK-2206 (allosteric AKT1/2/3 inhibitor) drug-sensitivity assay: PIK3CA- and AKT1-mutant lines were highly sensitive; FGFR3- and HRAS-mutant lines less so; TSC1-null lines were resistant. Pharmacodynamic immunoblot in MGH-U4 (PIK3CA H1047R) showed downregulation of pAKT, pPRAS40, pFOX-01/03, pS6, p-4EBP1; in HCV-29 (TSC1 Q55*) AKT and proximal substrates were inhibited but S6 and 4EBP1 were not — analogous to KRAS-mutant resistance to EGFR-directed therapy.
  • Most common arm-length changes: 20q gain (41.2%) and 11p loss (36.1%); recurrent amplifications include 11q13.2–13.3 (CCND1) and 17q12 (ERBB2).

Genes & alterations

  • TP53 — somatic mutation in 34% of tumors; strongly associated with high-CNA-burden cluster (P<.001); mutually exclusive with MDM2 amplification; enriched in neuroendocrine-differentiated tumors.
  • RB1 — mutation/deletion in 15%; enriched in high-CNA-burden (P<.003) and neuroendocrine subsets (50% vs 13%; P=.01); inversely correlated with CCND1/CCNE1 amplification.
  • FGFR3 — somatic hotspot mutation in 13% of high-grade bladder tumors.
  • FGFR1 — recurrent amplification; member of the MAPK-pathway alteration set (35% combined).
  • PIK3CA — mutation in 18%; predicts sensitivity to MK-2206 in cell lines.
  • ERBB2 — focal high-level amplification in 6.2% (6/97), with corresponding mRNA up-regulation and 3+ HER2 IHC; framed as a validated actionable target (analogous to breast/gastroesophageal disease).
  • MDM2 — amplification in 5%; nonoverlapping with TP53 mutation.
  • CDKN2A / CDKN2B — 9p21.3 homozygous deletion; CDKN2A altered in 24–25% of samples.
  • CCND1 — focal amplification at 11q13.2–13.3 in 14%; mutually exclusive with CCNE1 amplification and inversely correlated with RB1 loss.
  • CCNE1 — focal amplification in 5%.
  • E2F3 — amplification in 21% of bladder tumors (vs 4.9% in 1,932 non-urothelial epithelial tumors); associated with advanced stage and high E2F3-target signature; enriched in neuroendocrine-differentiated tumors (50% vs 17%; P=.03); proposed as a bladder-lineage–specific driver, analogous to MITF in melanoma.
  • KRAS — mutation in ~4% of bladder tumors as part of the MAPK alteration set.
  • BRAF — mutation in ~2% of bladder tumors; cited as a validated drug target in melanoma.
  • HRAS — mutation present in a subset of cell lines and primary tumors within the MAPK pathway set.
  • NF1 — somatic mutation in ~2% of tumors.
  • MET — alteration in ~2% of tumors.
  • RAF1 — amplification in ~6% of tumors.
  • AKT1 — mutation in 2%; predicts sensitivity to MK-2206.
  • PTEN — deletion in 4%; deleted samples showed reduced PTEN mRNA expression.
  • TSC1 — mutation in 7%; tumors/cell lines with TSC1 inactivation were resistant to MK-2206 at the level of S6/4EBP1 despite AKT inhibition.
  • MTOR — pathway downstream of PI3K/AKT in the analyzed PI3K/AKT/mTOR signaling map; pathway altered in 30% of tumors overall.

Clinical implications

  • 61% of high-grade bladder cancers harbor a potentially actionable genomic alteration, providing a substrate for precision-oncology trial design.
  • ERBB2 focal amplification with HER2 overexpression in 6.2% of high-grade bladder tumors supports investigation of HER2-directed therapy (trastuzumab, lapatinib analogs) in molecularly selected bladder patients.
  • Mutual exclusivity of MAPK-pathway lesions (FGFR1/FGFR3/ERBB2/MET/NF1/KRAS/BRAF) suggests they confer overlapping phenotypic effects and may each represent independent therapeutic entry points.
  • PI3K/AKT/mTOR alterations (30%) are actionable but response to selective inhibitors such as MK-2206 is modulated by the underlying genetic determinant: PIK3CA/AKT1 mutations confer sensitivity, while downstream TSC1 loss confers resistance at the level of S6/4EBP1 — a KRAS-EGFR–analogous resistance pattern.
  • The authors argue that prior failures of target-directed agents in unselected bladder cohorts may reflect genomic heterogeneity, and call for prospective genomic characterization to enrich future trials for patients whose tumors carry the relevant lesion.

Limitations & open questions

  • Mutation profiling was limited to Sequenom hotspot panels plus Sanger sequencing of 15 preselected oncogenes/tumor suppressors — not whole-exome — so novel driver mutations outside these genes were not detectable.
  • Neuroendocrine-differentiated tumors (10% of the cohort) were absent from the comparator cohort and may have inflated the prevalence of TP53, RB1, and E2F3 alterations.
  • aCGH was performed on bulk tumors; intratumoral heterogeneity (~33% of tumors had minor variant histology components such as squamous or sarcomatoid differentiation) may not be captured.
  • Cell line panel only modestly recapitulated primary-tumor genomics (e.g., no ERBB2 amplification in any line), constraining preclinical drug screening.
  • Matched primary/metastatic sample analysis is needed to determine whether actionable-alteration prevalence differs between primary and metastatic disease.
  • The two CNA-burden classes did not segregate clinically (recurrence-free, OS, or cancer-specific survival), so their prognostic utility is unclear in this cohort.
  • Coordinates were aligned to hg18, predating contemporary builds.

Citations from this paper used in the wiki

  • “Sixty-one percent of tumors harbored potentially actionable genomic alterations.” (Abstract / Results)
  • “Mutations in TP53 and RB1 were significantly more common in tumors with a high CNA burden (P<.001 and P<.003, respectively).” (Abstract / Fig 1A)
  • “The most common arm-length gains and losses were 20q (41.2%) and 11p (36.1%).” (Results)
  • “Sixty-three samples (65%) harbored mutations in at least one gene, the most prevalent of which were TP53 (34%), FGFR3 (13%), and PIK3CA (18%).” (Results)
  • “Mutually exclusive focal amplifications of CCND1 and CCNE1 were present in 14% and 5% of samples, respectively, and were inversely correlated with the presence of RB1 mutations or deletions.” (Results)
  • “Amplification of this region [E2F3] rarely occurs in other epithelial tumor types (21% v 4.9% of 1,932 nonurothelial epithelial tumors) and was associated with more advanced stage (28% stage III to IV v 6.7% stage I to II; P=.028).” (Results)
  • “Both E2F3 amplification and RB1 deletion/mutation were more prevalent in the subset of tumors exhibiting neuroendocrine differentiation compared with those with predominantly urothelial morphology (E2F3, 50% v 17%; P=.03; RB1, 50% v 13%; P=.01).” (Results)
  • “Cell lines harboring PIK3CA and AKT1 mutations were highly sensitive to MK2206, those with alterations in upstream pathway components including FGFR-3 and HRAS less so, whereas cell lines with downstream TSC-1 alterations were resistant to AKT inhibition.” (Results)
  • “Approximately one third of the tumors in this study exhibited minor histologic components such as squamous or sarcomatoid differentiation.” (Discussion)

This page was processed by crosslinker on 2026-05-09.