NEW YORK — With a combination of bulk and single-cell omic analyses, researchers have teased out more details about the pancreatic cancer tumor microenvironment, providing insights that could improve disease therapies.
Researchers from Washington University in St. Louis and elsewhere examined more than six dozen samples of pancreatic ductal adenocarcinoma (PDAC), which has an 11 percent five-year survival rate. Using multiple omic approaches, including spatial transcriptomics, they uncovered heterogeneity within the tumors and identified multiple tumor cell subpopulations, including transitional ADM and PanIN cell populations. As they reported on Monday in Nature Genetics, the researchers found some subpopulations were more common among patients who had undergone treatment and the presence of other subpopulations hinted at a means of overturning an immunosuppressive PDAC microenvironment.
"Our work provides a resource to identify new targets of clinical relevance," senior author Li Ding from WUSTL and colleagues wrote in their paper.
For their analysis, the researchers collected 73 PDAC samples and adjacent normal tissue samples from 21 patients. The patients included seven who were treatment naïve, but the others had undergone standard treatments such as neoadjuvant FOLFIRINOX or neoadjuvant gemcitabine and nab-paclitaxel therapy. All samples underwent single-cell RNA sequencing, and 64 samples underwent whole-exome sequencing and 65 bulk RNA sequencing.
Single-nucleus RNA-seq with spatial transcriptomics was also conducted for a portion of samples, and a subset of about 30 samples also underwent proteomic analysis.
Using the samples' expression profiles, the researchers clustered the samples and used marker gene expression to identify cell subpopulations. Based on pathway enrichment analyses, they further uncovered clusters with specific functions. Some subpopulations expressed, for instance, proliferation signatures, while others were marked by KRAS signaling or epithelial-to-mesenchymal transition signatures.
The researchers in particular homed in on cells expressing acinar markers, as one theory posits that PDAC arises from acinar cells that have undergone acinar-to-ductal metaplasia (ADM). The ADM cell population, they found, expresses both oncogenes and tumor suppressor genes, suggesting it represents a transitional or dynamic state. Additionally, they noted that the development of a driver KRAS mutation appears to move ADM cells forward toward PDAC.
Cell populations representing pancreatic intraepithelial neoplasia, a precursor to PDAC, similarly occupied a transitional state.
Meanwhile, the researchers examined cancer-associated fibroblasts (CAFs) found within the tumor samples, as the role of CAFs is unclear since they have both cancer driver and suppressor activities. Within the samples, they identified iCAFs, myCAFs, and apCAFs and noted different markers and pathways that are differentially expressed in the subtypes that are being studied in clinical trials.
For instance, the researchers found that the inflammatory-related iCAFs were present at higher levels in samples from patients who had undergone treatment, and that patients treated with gemcitabine and nab-paclitaxel had upregulation of metallothionein genes in iCAFs, which are associated with chemoresistance.
At the same time, the researchers examined the lack of immune checkpoint blockade activity in PDAC, a cancer type for which immunotherapies are not currently effective. By examining the immunosuppressive PDAC tumor microenvironments, the researchers uncovered strong expression of exhaustion markers in NK, CD4+ and CD8+ T, and regulatory T cells. They also found high expression of NECTIN receptors in all tumor cells and TIGIT in T regulatory and exhausted T cells, suggesting that the NECTIN-TIGIT axis could be targeted to bolster anti-tumor T cell activity.
"Our study reveals a deeper understanding of the intricate substructure of pancreatic ductal adenocarcinoma tumors that could help improve therapy for patients with this disease," Ding and colleagues wrote.
They added that future work that includes uniform treatment regiments and clinical-response data could additionally help identify treatment-linked resistance signatures.