Integrated molecular and pharmacological characterization of patient-derived xenografts from bladder and ureteral cancers identifies new potential therapies

Abstract

Background: Muscle-invasive bladder cancer (MIBC) and upper urinary tract urothelial carcinoma (UTUC) are molecularly heterogeneous. Despite chemotherapies, immunotherapies, or anti-fibroblast growth factor receptor (FGFR) treatments, these tumors are still of a poor outcome. Our objective was to develop a bank of patient-derived xenografts (PDXs) recapitulating the molecular heterogeneity of MIBC and UTUC, to facilitate the preclinical identification of therapies.

Methods: Fresh tumors were obtained from patients and subcutaneously engrafted into immune-compromised mice. Patient tumors and matched PDXs were compared regarding histopathology, transcriptomic (microarrays), and genomic profiles [targeted Next-Generation Sequencing (NGS)]. Several PDXs were treated with chemotherapy (cisplatin/gemcitabine) or targeted therapies [FGFR and epidermal growth factor (EGFR) inhibitors].

Results: A total of 31 PDXs were established from 1 non-MIBC, 25 MIBC, and 5 upper urinary tract tumors, including 28 urothelial (UC) and 3 squamous cell carcinomas (SCCs). Integrated genomic and transcriptomic profiling identified the PDXs of three different consensus molecular subtypes [basal/squamous (Ba/Sq), luminal papillary, and luminal unstable] and included FGFR3-mutated PDXs. High histological and genomic concordance was found between matched patient tumor/PDX. Discordance in molecular subtypes, such as a Ba/Sq patient tumor giving rise to a luminal papillary PDX, was observed (n=5) at molecular and histological levels. Ten models were treated with cisplatin-based chemotherapy, and we did not observe any association between subtypes and the response. Of the three Ba/Sq models treated with anti-EGFR therapy, two models were sensitive, and one model, of the sarcomatoid variant, was resistant. The treatment of three FGFR3-mutant PDXs with combined FGFR/EGFR inhibitors was more efficient than anti-FGFR3 treatment alone.

Conclusions: We developed preclinical PDX models that recapitulate the molecular heterogeneity of MIBCs and UTUC, including actionable mutations, which will represent an essential tool in therapy development. The pharmacological characterization of the PDXs suggested that the upper urinary tract and MIBCs, not only UC but also SCC, with similar molecular characteristics could benefit from the same treatments including anti-FGFR for FGFR3-mutated tumors and anti-EGFR for basal ones and showed a benefit for combined FGFR/EGFR inhibition in FGFR3-mutant PDXs, compared to FGFR inhibition alone.

Keywords: basal tumors; luminal tumors; molecular subtypes; squamous cell carcinoma; tumor heterogeneity; tyrosine kinase receptor; upper-urinary tract carcinoma; urothelial carcinoma.

Figure 1

Histological and genomic characteristics of patient tumors and paired derived patient-derived xenografts (PDXs). (A) Histology of bladder/ureteral patient tumors and corresponding PDX models demonstrated similar histological features as assessed by hematoxylin and eosin (H&E) staining. Scale bar corresponds to 50 µm. H&E slides of a section of each patient tumor and matching PDX were reviewed by a board-certified pathologist, and representative pictures for the different histologies or variants are shown. (B) Short tandem repeat signature of a patient specimen and PDX tumor, an example of the L987 case. (C) Somatic genomic landscape of 31 bladder and ureteral PDXs analyzed using an in-house targeted sequencing assay (571 cancer-related genes, Supplementary Tables 2, 3), the tumor mutational burden per megabase (TMB, indicated in the log2 scale for each sample), and microsatellite instability–high (MSI-H) versus a microsatellite-stable (MSS) status. (D) Concordance of genomic alterations in five pairs tumor/PDX.

Figure 2

Transcriptomic analysis of patient tumors and paired PDXs. (A) Tumors and PDXs were classified into six subtypes using transcriptomic Affymetrix U133plus2 array data according to the molecular consensus classifier developed for MIBC, the corresponding box colors indicated in the legend on the right. A total of 22 patient tumor–PDX pairs and 7 individual PDXs were analyzed, where tumors without transcriptomic data are indicated in gray. Urothelial carcinomas with divergent squamous differentiation are highlighted with *. (B) Upper panels: tumors and PDXs were classified using transcriptomic data as luminal and basal subtypes according to the BASE47 classifier (27). In patient tumors with missing transcriptomic data, we assessed the luminal (blue), basal (red), or heterogeneous (orange) subtype by immunohistochemistry (IHC; inset small boxes), as defined in methods. Lower panels: the intra-tumoral heterogeneity and proportion of luminal and basal subtype admixture as evaluated from transcriptomic profiles using the WISP (Weighted In Silico Pathology) algorithm. Based on the PDX WISP results, samples were molecularly classified as luminal, basal, or heterogeneous. (B) Heatmap of PDX samples based on the gene expression of selected luminal or basal markers. (C) Heatmap of PDX samples based on the regulon activity of the main regulators previously identified within the different molecular subtypes of MIBC (9, 10).

Figure 3

Intratumor heterogeneity in tumors and PDXs at the protein level by IHC. (A) GATA3 and KRT5/6 expression levels (normalized quick scores), grouped according to the PDX WISP molecular classification (luminal, heterogeneous, and Ba/Sq, Figure 2A). (B) Proportion of tumors with intratumoral heterogeneity (left) and GATA3 + KRT5/6 coexpression at the single-cell level (right), grouped according to the PDX WISP molecular classification (luminal, heterogeneous, and Ba/Sq, Figure 2A). (C) Patterns of dual IHC staining for GATA3 (brown, nuclear) and KRT5/6 (red, cytoplasmic) in the paired tumor/PDX of a luminal (BLCU-011), a heterogeneous (F659), and a Ba/Sq (C704) example.

Figure 4

Chemosensitivity of representative basal and luminal PDX models. Mice with established PDXs (67–270 mm³) were treated with cisplatin plus gemcitabine (green) as indicated, and control mice were treated with vehicle alone (black) (n = 7–10 animals per group). Tumor size was measured at the indicated time points. Data are presented as mean ± SEM. Results were compared using the Mann–Whitney test. n.s, non-significant.

Figure 5

Sensitivity to anti-EGFR or the combination of EGFR and FGFR inhibitors in Ba/Sq and FGFR3-mutated PDXs. (A) Mice with established basal/squamous PDXs (67–270 mm³) were treated with anti-EGFR (erlotinib 90 mg/kg, red) or control vehicle alone (black). (B) Mice with established FGFR3-mutated tumors and controls were treated with control vehicle (black), low-dose anti-EGFR (erlotinib 30 mg/kg, red), a pan-FGFR inhibitor (erdafitinib 10 mg/kg, blue), or the combination (purple), as indicated (n = 7–10 animals per group). Tumor size was measured at the indicated time points. Data are presented as mean ± SEM. Results were compared using the Mann–Whitney test.