Supplementary Table 2. Representativeness of Study Participants

Supplementary Table 3. Genetic variant detected in the custom gene panel through whole exome sequencing.

Supplementary Table 4. Preclinical model recommendation

PURPOSE:Pancreatic ductal adenocarcinoma (PDAC) has limited treatment options. We compared the efficacy of comprehensive precision medicine against that of the conventional treatment in PDAC. PATIENTS AND METHODS:We report a phase III trial of advanced PDAC in which patients were randomized (1:2) to a conventional treatment treated at physician's discretion (arm A) or to precision medicine (arm B). Subjects randomized to arm B underwent a tumor biopsy for whole-exome sequencing and to generate avatar mouse models and patient-derived organoids for phenotypic drug screening, with final treatment recommended by the molecular tumor board. The primary objective was median overall survival (OS). RESULTS:A total of 137 patients were enrolled with 125 randomized, 44 to arm A and 81 to arm B. Whole-exome sequencing was performed in 80.3% (65/81) patients of arm B, with potentially actionable mutations detected in 21.5% (14/65). Experimental models were generated in 16/81 patients (19.8%). Second-line treatment was administered to 39 patients in the experimental arm, but only four (10.2%) received personalized treatment, whereas 35 could not receive matched therapy because of rapid clinical deterioration, delays in obtaining study results, or the absence of actionable targets. The median OS was 8.7 and 8.6 months (P = 0.849) and the median progression-free survival was 3.8 and 4.3 months (P = 0.563) for the conventional and experimental arms, respectively. Notably, the four patients who received personalized treatment had a median OS of 19.3 months. CONCLUSIONS:Personalized medicine was challenging to implement in most patients with PDAC, limiting the interpretation of intention-to-treat analysis. Survival was improved in the subset of patients who did receive matched therapy.

Supplementary Table 5. Complete Efficacy Data

Supplementary Table 1. Custom virtual gene panel included in the whole exome sequencing analysis

Pancreatic cancer has the worst prognosis of all common tumors. Earlier cancer diagnosis could increase survival rates and better assessment of metastatic disease could improve patient care. As such, there is an urgent need to develop biomarkers to diagnose this deadly malignancy. Analyzing circulating extracellular vesicles (cEVs) using ‘liquid biopsies’ offers an attractive approach to diagnose and monitor disease status. However, it is important to differentiate EV-associated proteins enriched in patients with pancreatic ductal adenocarcinoma (PDAC) from those with benign pancreatic diseases such as chronic pancreatitis and intraductal papillary mucinous neoplasm (IPMN). To meet this need, we combined the novel EVtrap method for highly efficient isolation of EVs from plasma and conducted proteomics analysis of samples from 124 individuals, including patients with PDAC, benign pancreatic diseases and controls. On average, 912 EV proteins were identified per 100 µL of plasma. EVs containing high levels of PDCD6IP, SERPINA12, and RUVBL2 were associated with PDAC compared to the benign diseases in both discovery and validation cohorts. EVs with PSMB4, RUVBL2, and ANKAR were associated with metastasis, and those with CRP, RALB, and CD55 correlated with poor clinical prognosis. Finally, we validated a seven EV protein PDAC signature against a background of benign pancreatic diseases that yielded an 89% prediction accuracy for the diagnosis of PDAC. To our knowledge, our study represents the largest proteomics profiling of circulating EVs ever conducted in pancreatic cancer and provides a valuable open-source atlas to the scientific community with a comprehensive catalogue of novel cEVs that may assist in the development of biomarkers and improve the outcomes of patients with PDAC.

Figure S2: Read alignments from WES-cfDNA, showing somatic mutations in APC and TP53 genes.

Table S5: List of KDR/VEGFR2 somatic mutations from the Cosmic, TCGA, GENIE and PCA-WGS databases.

Table S3: Treatment regimens of the Avatar patient derived xenograft models carrying the VEGFR2 WT and L840F genotypes.

Figure S5: A) Levels of phosphorylated ERK following treatment of the MDST8 colorectal cell line, expressing R1032Q VEGFR2, with different kinase inhibitors in the presence. B) Proliferation assays of Colo320 cell lines, stably expressing WT and R1032Q VEGFR2, upon treatment with TKIs. The expression of the VEGFR2 R1032Q hot-spot mutant in Colo320 cell lines (WT to KRAS/NRAS/BRAF/PIK3CA and mutated to TP53 and APC) increased sensitivity to cabozantinib.

Legends for Supplementary Figures 1-6

Additional Materials and Methods information not included in the article text, and detailed description of the patient tumors from which xenograft models have been derived from

Table S2: Custom primers and probes designed based on the Thermofisher online design tool for TaqManÃ,® genotyping assay for KDR c.2518C>T (VEGFR2 p.L840L). TaqManÃ,® MGB (minor groove binder) probes incorporate a 5' reporter (VIC or FAM) and a 3' nonfluorescent quencher (NFQ).

Table S1: Mutation analysis concordance between plasma cfDNA exome sequencing and tumor exome sequencing.

S1-S7 all supplementary figures with their legends: S1:Figure S1. OMTX705 binding and internalization in HT1080-WT, -FAP and CAFs S2:Weight change of mice bearing Panc 007, Lung 024 and Breast 014 tumors. S3: OMTX705 Activity in the Immunodeficient Patient Derived Xenografts (PDX) Models of Ovarian Cancer. S4: FAP immunostaining in frozen tumor samples from humanized mice bearing CTG-0860 NSCLC PDX tumor. S5: Separate fluorescent staining of human IgG, Rab7 and nuclei in OMTX705 and OMTX005 treated HT1080-FAP cells (complements the merged image of figure 4C) S6: Immunohistochemical analysis of payload, OMTX005 and OMTX705 distribution in tumors from Panc 007 model. S7: Schematic summary of findings and proposed mechanism of action of OMTX705.

Figure S4: A) Molecular dynamics simulations with L840F VEGFR2 indicates that most of the adopted F840 conformations during the simulation are not compatible with inhibitor binding. Clashes between VEGFR2 and the sorafenib area are shown as red discs. B) Kinase assays showing impaired kinase activities of VEGFR2 L840F and R1032Q compared to wild-type. C) Effects of the L840F mutation on Y1175 VEGFR2 phosphorylation. HEK293 cells were transiently transfected with increasing levels of a plasmid encoding for WT or L840F VEGFR2. Cells at 70% confluence were starved in 1% BSA/DMEM for 4 h, and then incubated in the absence or presence of 60 ng/ml VEGF165 for 10 min at 37 {degree sign}C. Whole cell lysates were analyzed by western blotting, using antibodies against phosphoY1175 and total VEGFR2. Representative results are shown. D) PAE cells stably expressing WT or L840F VEGFR2 were generated from a PAE cell line that does not normally express VEGFR2 (empty). Cells at 70% confluence were starved in 1% BSA/DMEM for 4h, and then incubated in the absence or presence of 60 ng/ml VEGF165 for 10 min at 37 {degree sign}C. Whole cell lysates were then analyzed by western blotting, using antibodies against phosphoY1175 and total VEGFR2. Representative results are shown, highlighting the significant decrease in VEGF-induced Y1175 VEGFR2 phosphorylation in the presence of the L840F mutation.

Supplementary Data from Organoid Sensitivity Correlates with Therapeutic Response in Patients with Pancreatic Cancer

Table S4: Primers used for site-directed mutagenesis. The mutated nucleotides are indicated in bold and underlined in each mutant.