Molecular Characterization and Therapeutic Opportunities in KRAS Wildtype Pancreatic Ductal Adenocarcinoma

Abstract

Purpose: To investigate the molecular characteristics of and potential for precision medicine in KRAS wildtype pancreatic ductal adenocarcinoma (PDAC).

Patients and methods: We investigated 27 patients with KRAS^WT PDAC at our institution. Clinical data were obtained via chart review. Tumor specimens for each subject were interrogated for somatic single nucleotide variants, insertion and deletions, and copy number variants by DNA sequencing. Gene fusions were detected from RNA-seq. A patient-derived organoid (PDO) was developed from a patient with a MET translocation and expanded ex vivo to predict therapeutic sensitivity prior to enrollment in a phase 2 clinical trial.

Results: Transcriptomic analysis showed our cohort may be stratified by the relative gene expression of the KRAS signaling cascade. The PDO derived from our patient harboring a TFG-MET rearrangement was found to have in vitro sensitivity to the multi-tyrosine kinase inhibitor crizotinib. The patient was enrolled in the phase 2 SPARTA clinical trial and received monotherapy with vebrelitinib, a c-MET inhibitor, and achieved a partial and durable response.

Conclusions: KRAS^WT PDAC is molecularly distinct from KRAS^MUT and enriched with potentially actionable genetic variants. In our study, transcriptomic profiling revealed that the KRAS signaling cascade may play a key role in KRAS^WT PDAC. Our report of a KRAS^WT PDAC patient with TFG-MET rearrangement who responded to a cMET inhibitor further supports the pursuit of precision oncology in this sub-population. Identification of targetable mutations, perhaps through approaches like RNA-seq, can help enable precision-driven approaches to select optimal treatment based on tumor characteristics.

Keywords: APL-101; KRAS wildtype; MET fusion; PLB1001; pancreatic ductal adenocarcinoma; vebreltinib.

Figure 1

Tumor genotyping of KRAS^WT PDAC reveals potentially therapeutically actionable mutations. Depicted is an Oncoprint demonstrating the presence and type of mutation detected on NGS of the KRAS^WT PDAC cohort. Each column represents an individual subject. General groups of genes are separated and labeled on the left. The colored subject numbers are indicative of the patient’s transcriptomic subgroup with red representing Group 1 and blue representing Group 2 (see Figure 2). Black means that RNA-sequencing data was not available or not performed.

Figure 2

Transcriptomic characterization of KRAS^WT PDAC reveals clustered subtypes distinguished by differential expression of genes regulated by KRAS activity. (A) RNA-sequencing data represented as a heat map of the top 1000 most differentially expressed genes. The dendrogram (top) represents unsupervised non-hierarchical clustering of the subjects’ gene expression. (B) Gene set enrichment analysis comparing Group 1 and Group 2 showing differential enrichment of the HALLMARK_KRAS_SIGNALING_UP gene set. (C) Heatmap comparing individual relative expression of genes included in the KRAS signaling gene set. Each column represents data from an individual subject.

Figure 3

A PDO harboring a TFG-MET translocation serves as a successful patient avatar predicting sensitivity to muti-tyrosine kinase inhibitors. (A) Diagram illustrating the translocation breakpoints between the TFG and MET loci. Colored areas are representative of predicted functional domains. The purple C-terminal domain at the MET locus encodes the protein’s catalytic domain. The translocation involves TFG at exon 6 (3:100455560) and MET at exon 15 (7:116414935). AA = amino acid. (B) Representative images from dose–response experiments of a PDO harboring the TFG-MET translocation. Phase contrast images (40× magnification) were taken on days 1, 5 and 7 of treatment with the indicated increasing doses of crizotinib (above) and gemcitabine (below). Immunofluorescence images taken on Day 7 were after live–dead staining in which live cells have green fluorescence while dead cells fluoresce red. (C) Dose–response curve comparing percentage cell viability between crizotinib and gemcitabine (n = 3). Data are normalized to vehicle control. (D) Coronal computed tomography images of the KRAS^WT pancreatic head mass harboring the TFG-MET translocation. The red line indicates the largest diameter of the tumor. Images depict the tumor at baseline (above) and two months after initiation of vebreltinib (below).