Gene Dysregulation and Islet Changes in PDAC-Associated Type 3c Diabetes

Abstract

Pancreatic ductal adenocarcinoma (PDAC) is a highly lethal malignancy, often associated with new-onset diabetes. The relationship between PDAC and diabetes, particularly type 3c diabetes, remains poorly understood. This study investigates whether PDAC-associated diabetes represents a distinct subtype by integrating transcriptomic and histological analyses. Whole-tumour RNA sequencing data from The Cancer Genome Atlas (TCGA) were analysed to compare gene expression profiles between PDAC patients with and without diabetes. Cell-type Identification By Estimating Relative Subsets Of RNA Transcripts (CIBERSORT) deconvolution was employed to assess immune cell populations. Histopathological evaluations of pancreatic tissues were conducted to assess fibrosis and islet morphology. Histological analysis revealed perivascular fibrosis and islet basement membrane thickening in both PDAC cohorts. Transcriptomic data indicated significant downregulation of islet hormone genes insulin (INS) and glucagon (GCG) but not somatostatin (SST) in PDAC-associated diabetes, consistent with a type 3c diabetes phenotype. Contrary to previous reports, no distinct immunogenic signature was identified in PDAC with diabetes, as key immune checkpoint genes (Programmed Cell Death Protein 1 (PDCD1), Cytotoxic T-Lymphocyte Associated Protein 4 (CTLA4), Programmed Death-Ligand 1(PD-L1)) were not differentially expressed. The findings suggest that PDAC-associated diabetes arises through neoplastic alterations in islet physiology rather than immune-mediated mechanisms. The observed reductions in endocrine markers reinforce the concept of PDAC-driven β-cell dysfunction as a potential early indicator of malignancy. Given the poor response of PDAC to PD-L1 checkpoint inhibitors, further research is needed to elucidate alternative therapeutic strategies targeting tumour-islet interactions.

Keywords: INS; PD-L1; islet basement membrane; pancreatic ductal adenocarcinoma; type 3 diabetes.

Figure 1

The selection process for the final study cohort. The total TCGA dataset initially included 183 samples, of which 69 were excluded due to non-PDAC cancers. A total of 114 PDAC patients were identified, which were further stratified into two groups: PDAC patients without diabetes (n = 84) and PDAC patients with diabetes (n = 30). Among the latter group, 8 patients with a history of pancreatitis were excluded, resulting in the final PDAC with diabetes cohort (n = 22). The final PDAC without diabetes cohort included 84 patients.

Figure 2

Islet pathology in PDAC. (A) Islets from individuals with PDAC are situated within tissue showing high levels of fibrosis (indicated by black arrow) and some demonstrate thickened basal membranes (indicated by white arrow). Representative islets are shown from the patients with no diabetes (AAB0 and AAB1) and those with diabetes (A77Q, A6UF). All samples were taken from PDAC tumours graded stage II (B) and located at the head of the pancreas. Pancreatic tissue from an individual with PDAC and diabetes showing high levels of fibrosis and alteration of the exocrine tissue (indicated by black arrow). Pancreatic tissue from non-PDAC/non-diabetic patient is shown for comparison. The scale bar represents 20 µm.

Figure 3

Volcano Plot of Differential Gene Expression in PDAC associated Diabetes vs. Non-Diabetes. Volcano plot showing the relationship between gene expression changes and statistical significance in PDAC patients with diabetes compared to those without diabetes. The horizontal blue dashed line indicates the significance threshold (p = 0.05). Grey dots represent individual genes, while red dots highlight key islet-related genes (p < 0.05) involved in glucose homeostasis and pancreatic function.

Figure 4

Gene expression analysis of endocrine markers across cohorts. (A) Endocrine markers INS, GCG, SST, and WSCD2 were all reduced in the diabetic cohort (** p < 0.05, * p < 0.1) (B) Correlation analysis between duration of diabetes (days) and INS (r = 0.36); GCG (r = −0.06). (C) Correlation analysis between GCG and INS (r = 0.79; p < 0.001), and SST and INS (r = 0.84; p < 0.001) for the diabetic cohort. Gene expression levels are represented as RSEM-normalized counts.

Figure 5

Gene expression analysis of exocrine markers across cohorts. (A) Genes associated with the exocrine tissue had similar expression across cohorts (p > 0.05) except KRT19, which was significantly reduced in the diabetic cohort (** p < 0.05). (B) Comparison of clinical outcomes for PDAC with diabetes cohort versus no diabetes cohort. Clinical data were downloaded from the Firehose server and plotted as a Kaplan–Meier curve based on presence or absence of diabetes. Mean survival for the diabetes cohort = 498 days, no diabetes cohort = 614 days. Time = days, p value = 0.3.

Figure 6

Gene expression analysis for genetic markers of immunogenic subtype. (A) PDCD1 expression had a binary pattern in the diabetic group (p > 0.05). Correlation analysis between PDCD1 and CTLA4 expression in the diabetic cohort was significant (r = 0.9, p < 0.001) (B). No correlation was observed between PDCD1 and duration of diabetes (r = 0.4), similarly between PDCD1 and INS in the diabetic cohort (r = 0.09).

Figure 7

Comparison of relative leukocyte populations in the tumour microenvironment of PDAC (with and without diabetes) and pancreatitis as determined by CIBERSORT. No statistical significance was observed but the PDAC with diabetes group had higher levels of CD4 T cell markers. Round dots indicate no diabetes, and squares indicate with diabetes.