Pre-existing diabetes mellitus (DM) is associated with increased PCa-specific and all-cause mortality in men with prostate cancer (PMID: 27652121). However, the underlying reasons are unclear. We hypothesized that the transcriptomic profile of metastatic PCa patients (pts) diagnosed with DM before the diagnosis of metastatic disease and the start of systemic therapy would be different from those without DM.


This IRB-approved retrospective study included advanced prostate cancer pts with available RNA profiling of treatment naïve tumor tissue through a CLIA-certified laboratory. Based on pre-existing DM before the onset of metastatic disease, pts were grouped into DM and non-DM. Differential gene expression analysis between the two groups was performed using DeSeq2. Results were subjected to Gene Set Enrichment software analysis (GSEA) to identify pathways enriched in each cohort. Gene ontology analysis using TopGO software was done to determine the biological process occurring at the molecular level of these differentially expressed genes. All bioinformatic analysis was conducted in R studio, version 4.1.1.


75 pts were included: 20 DM vs 55 non-DM. Baseline characteristics (DM vs. non-DM) were as follows: median age 63.5 vs. 64 years; median PSA at diagnosis 20 vs. 18.85ng/mL; de novo disease: 55% vs. 43.6%; Gleason score ≥8: 60% vs. 74.5%. DM pts had upregulation of the following pathways: TNF alpha signaling, inflammatory response, IL-6 JAK STAT3 signaling, heme metabolism, and the p53 pathway vs. non-DM pts (table 1). Gene ontology analysis and individual differential gene expression profiles will be reported at the meeting.

Table 1.Tumor normalized GSEA score between DM vs non-DM prostate cancer pts.
Pathway Normalized enrichment score P-value Q-value
TNFA Signaling via NFKB 2.24 <0.001 <0.001
Inflammatory response 2.07 <0.001 <0.001
IL-6 JAK-STAT 3 Signaling 1.99 <0.001 <0.001
Heme metabolism 1.67 <0.001 <0.001
p53 Pathway 1.44 0.002 0.003


Our study found that pre-existing DM is associated with the upregulation of inflammatory pathways in pts with PCa. These hypothesis-generating results need external validation. Identifying transcriptomic biomarkers in these subsets of pts may help with future drug development.