Precision Oncology and Systemic Targeted Therapy in Pseudomyxoma Peritonei

Abstract

Purpose: Pseudomyxoma peritonei (PMP) is a rare and poorly understood malignant condition characterized by the accumulation of intra-abdominal mucin produced from peritoneal metastases. Currently, cytoreductive surgery remains the mainstay of treatment but disease recurrence and death after relapse frequently occur in patients with PMP. New therapeutic strategies are therefore urgently needed for these patients.

Experimental design: A total of 120 PMP samples from 50 patients were processed to generate a collection of 50 patient-derived organoid (PDO) and xenograft (PDX) models. Whole exome sequencing, immunohistochemistry analyses, and in vitro and in vivo drug efficacy studies were performed.

Results: In this study, we have generated a collection of PMP preclinical models and identified druggable targets, including BRAFV600E, KRASG12C, and KRASG12D, that could also be detected in intra-abdominal mucin biopsies of patients with PMP using droplet digital PCR. Preclinical models preserved the histopathological markers from the original patient sample. The BRAFV600E inhibitor encorafenib reduced cell viability of BRAFV600E PMP-PDO models. Proof-of-concept in vivo experiments showed that a systemic treatment with encorafenib significantly reduced tumor growth and prolonged survival in subcutaneous and orthotopic BRAFV600E-PMP-PDX mouse models.

Conclusions: Our study demonstrates for the first time that systemic targeted therapies can effectively control PMP tumors. BRAF signaling pathway inhibition represents a new therapeutic opportunity for patients with BRAFV600E PMP who have a poor prognosis. Importantly, our present data and collection of preclinical models pave the way for evaluating the efficacy of other systemic targeted therapies toward extending the promise of precision oncology to patients with PMP.

Figure 1.

Generation of a collection of PMP-PDO, PDX, and PDXO models. A, Design showing the generation of PDO from tumor tissues of patients with PMP. B, Pictures of PMP-PDO models after 7 days in culture (PMP18.3, left image and PMP21.5, right image). Scale bar, 500 μm. C, Expansion of PMP-PDO models after dissociation with TrypLE Express reagent at day 2 (left) and day 7 (right; PMP5.2). Scale bar, 500 μm D, Design showing the generation of PDX and PDXO from tumor tissues of patients with PMP. PM, Peritoneal metastases. E–G, Macroscopic images from an orthotopic PMP-PDX (oPDX) models in mice derived from low-grade G1 (E and F; PMP3.1) or high-grade G3 (G; PMP5.1) samples from patients with mucinous carcinoma peritonei. Tumor mass is indicated by red arrows. H and I, H&E staining from low-grade G1 (H; PMP3.1) and high-grade G3 (I; PMP5.1) samples from patients with mucinous carcinoma peritonei and associated oPDXs. Scale bar, 2.5 mm and 100 μm. J, H&E staining from a subcutaneous PMP-PDX (sPDX) model derived from the sample from patient with high-grade G3 mucinous carcinoma peritonei (PMP5.1). Scale bar, 2.5 mm and 100 μm. (A and D, Created with BioRender.com.). PMP, Pseudomyxoma peritonei.

Figure 2.

Genomic characterization of preclinical models and intra-abdominal mucin reveal druggable targets in PMP. A, Oncoplot from exome sequencing analysis from 28 PMP-PDO or PMP-PDXO models derived from 20 patients. The most common mutated genes in all the models are presented and compared with the cohort of patients with colorectal cancer from the TCGA COADREAD, a cohort of patients with appendiceal tumors from the study of Ang and colleagues and a collection of PMP paraffin samples from the study of Alakus and colleagues. Patient and sample number, primary vs. peritoneal disease, and grade are indicated in the top. Genomic alterations are shown in different colors in the oncoplot. The percentage of PMP preclinical models with the specific mutations is indicated on the right site of the oncoplot. Finally, specific mutations in the KRAS gene are shown in the bottom site of the panel. B, ddPCR for KRAS^G12C using genomic DNA extracted from intra-abdominal mucin biopsy of PMP29 (sample PMP29.1; left), negative control (middle), and positive control (right). The study was performed in duplicates. Blue spots and green spots represent positive events for KRAS^G12C (top graph) and KRAS^WT genotype (bottom graph), respectively. The threshold line was determined by the control samples to separate the two clusters of negative and positive droplets. PMP, Pseudomyxoma peritonei; CRC, Colorectal cancer; PDO, Patient-derived organoid; PDXO, Patient-derived xenografts organoid; LAMN, low-grade appendiceal mucinous neoplasm; HAMN, high-grade appendiceal mucinous neoplasm; Grade 1, low-grade mucinous carcinoma peritonei; Grade 2, high-grade mucinous carcinoma peritonei; Grade 3, high-grade mucinous carcinoma peritonei with or without signet ring cells.

Figure 3.

Treatment with BRAF inhibitor is enough to block RAS/ERK signaling pathway and reduce cell viability in PMP models. BRAF^WT (T70) or BRAF^V600E (CTAX34) colorectal cancer PDXO and BRAF^V600E PMP PDXO (PMP5.1) or PDO (PMP5.2) derived from two different peritoneal metastasis from PMP patient 5 treated with vehicle, mitomycin C 150 ng/mL, cetuximab 100 μg/mL, encorafenib 1 μmol/L, doublet (encorafenib + cetuximab). A and B, Cell viability (A) and caspase 3/7 activation (B) were measured after 5 days or 48 hours on treatment, respectively. Mean ± SD of triplicates is shown. Significant differences were assessed using one-way ANOVA and Dunnett’s multiple comparisons tests compared to control (*, P value < 0.05; **, P value < 0.01; ***, P value < 0.001; ****, P value < 0.0001). C and D, Western blot analysis of phospho-ERK, phospho-EGFR, ERK, and EGFR was performed after 30 minutes or 24 hours on treatment. Quantification of phospho-EGFR/EGFR (C) and phospho-ERK/ERK (D) ration in all the samples. r.u, relative units. PMP, Pseudomyxoma peritonei; CRC, Colorectal cancer; PDO, Patient-derived organoid; PDXO, Patient-derived xenografts organoid.

Figure 4.

BRAF inhibitor suppresses subcutaneous tumor growth in a BRAF^V600E PMP-PDX model. BRAF^V600E mutant PMP-PDX cells from PMP5.1 were subcutaneously implanted in mice and treated with vehicle, intraperitoneal cetuximab (20 mg/kg), oral encorafenib (20 mg/kg), or doublet (n = 10–12/group). A, Tumor growth curve of the treated mice overtime. Mean ± SEM is shown. B, Graphs show the Western blot quantification of pEGFR/EGFR (left) and pERK/ERK (right) ratio from samples of representative excised tumors at day 4 after treatment. Mean ± SD is shown. C, Representation of the fold change (FC) in tumor volume for each tumor at day 44 after treatment. Mean ± SEM is shown. Significant differences were assessed using one-way ANOVA and Dunnett’s multiple comparisons tests compared to vehicle group (*, P value < 0.05; ****, P value < 0.0001; A and C). D, Image of representative tumors from each group at the end of the experiment. E, Survival curves represent PFS percentage of each tumor. Mice survival was evaluated by the Kaplan–Meier survival curve and log-rank test group (****, P value < 0.0001). F, Ki-67 IHC was performed from all tumor samples at the end of the experiment. Images from a representative tumor from vehicle, cetuximab, encorafenib and doublet groups. Scale bar, 100 μm. PMP, Pseudomyxoma peritonei; PDX, Patient-derived xenografts.

Figure 5.

BRAF inhibitor reduces tumor growth in an orthotopic BRAF^V600E PMP-PDX model. BRAF^V600E mutant PMP-PDX mucinous tumor tissue from PMP5.1 was implanted intraperitoneally in mice. Animals were scanned using microCT at day 3 and day 6 after tumor implantation. A, Representation of paired tumor volume of each mouse at day 3 and 6 after tumor implantation. Significant differences were assessed using Unpaired t test (****, P < 0.0001). B–H, Based on the microCT images, at day 6 animals were divided in four groups and treated with vehicle, intraperitoneal cetuximab (20 mg/kg), oral encorafenib (20 mg/kg), or doublet (n = 10–11/group). B, Representation of the FC in percentage of body weight the treated mice overtime. Mean ± SEM is shown. Significant differences were assessed using one-way ANOVA and Dunnett’s multiple comparisons tests compared to the vehicle group. Unpaired t test to assess significant differences in encorafenib and doublet group was used at the end of experiment. (**, P value < 0.01; ***, P value < 0.001; ****, P value < 0.0001). C–F, At day 37, animals were monitored with microCT and a representative animal of each group was euthanized. C, FC representation of the tumor volume from microCT images from representative mice from each group (n = 2–4). Mean ± SEM is shown. Significant differences were assessed using one-way ANOVA and Dunnett’s multiple comparisons tests compared to vehicle group (*, P value < 0.05). D and E, Axial, coronal, and sagittal microCT images (D) and tridimensional rendering of microCT studies (E) from a representative mouse from each group. E, Pallid blue mucinous tissue volume and pallid yellow bone structures/tissue are shown. F, Postmortem images of mice (top) and tumor lesions in the peritoneal cavity (bottom) are shown. G and H, Animals were euthanized when they reached the end point criteria of 50% of increase in body weight. G, Quantification of the weight increase (left) and the volume of mucinous tumor (right) isolated from the intra-abdominal cavity when mice reached the end point criteria. Mean ± SEM is shown. Significant differences were assessed using one-way ANOVA and Tukey’s multiple comparisons tests (**, P value < 0.01; ***, P value < 0.001; ****, P value < 0.0001). H, Survival curves from each group. Overall mice survival was evaluated by the Kaplan–Meier survival curve and log-rank test group (****, P value < 0.0001). PMP, Pseudomyxoma peritonei; PDX, Patient-derived xenografts.

Figure 6.

Schematic representation of the results of the study and the potential clinical translation. Design showing the generation of a unique collection of preclinical models derived from PMP tumor samples used in this study. Genomic characterization of these models revealed druggable targets that were validated in vitro and in vivo. In a clinical scenario, ddPCR from intra-abdominal mucin biopsies of patients with PMP could be used as a diagnostic tool to detect these druggable mutations and propose an individual matched systemic targeted therapy for each patient (Created with BioRender.com.). PMP, Pseudomyxoma peritonei; PDXO, Patient-derived xenografts organoid.