Identification of molecular and functional subtypes using chronic pancreatitis patient-derived organoid models

Abstract Background and Aims Chronic pancreatitis (CP) affects more than 200,000 Americans and 1 million individuals world wide, but treatment generally focuses on supportive care like pain management. 28-80% of CP cases are idiopathic and 10-15% are hereditary with mutations in PRSS1, SPINK1, CFTR, and other genes. Developing human models of pancreatitis to understand the drivers of the disease and targeting key players in this disease will provide novel therapeutic approaches for these patients who have limited options. Methods Utilizing patient samples from surgical resections or TPIAT procedures we have generated a first-of-its-kind human CP patient derived organoid (PDO) biobank. This biobank is representative of hereditary mutations found across patients, includes idiopathic cases, and histologically recapitulates features of this disease. Genetic concordance was determined between PDO and primary specimens using whole genome sequencing. RNAseq and cytokine analysis was conducted on PDO to determine inflammatory features of this model. Forskolin induced swelling assays were conducted to determine the function of CFTR in PDO models. Results There is a 98% concordance between PDO and primary tissue specimens showing the recapitulation of patient genotypes in our model. RNA-seq revealed that CP organoids are transcriptionally distinct from normal pancreas organoids and identified three distinct subtypes of CP that are driven by transcriptional features. Utilizing the PDO model, we identified a high prevalence of ductal CFTR dysfunction, which could be restored using a combination of CFTR correctors and potentiators. Conclusions There are three subtypes of CP that require further study to employ targeted therapeutics against these drivers. hCP PDO are a model that can be used to assess CFTR function and correction providing evidence for these drugs to be used in CP patients.