Multiplexed tissue immunofluorescence data from a clinical trial of gemcitabine and nab-paclitaxel with or without the VDR agonist paricalcitol for metastatic pancreatic cancer

This dataset contains processed multiplexed tissue immunofluorescence data from a clinical trial of gemcitabine and nab-paclitaxel with or without the VDR agonist paricalcitol for metastatic pancreatic cancer (Perez et al. Nature Cancer). Whole slide digital images of metastatic pancreatic biopsy specimens were acquired at 200x magnification using a PhenoImager HT system (Akoya Biosciences, Hopkinton, MA). A pathologist inspected each image after spectral unmixing to confirm the presence of viable cancer and to exclude tissue processing artifacts. Next, images were segmented into tumor epithelial and stromal areas using supervised machine learning, followed by cell detection and segmentation (inForm 2.6, Akoya Biosciences, Hopkinton, MA). Cell phenotypes were assigned via separately trained machine learning algorithms relying upon single cell cytomorphology and protein expression data. Two different multiplex immunofluoresence panels were used for this analysis. One panel enabled the characterization of VDR protein expression, along with the identification of the following cell phenotypes: CD4+ T cells, CD8+ T cells, CD4-CD8- T cells, macrophages (CD68+ and/or CD163+), other immune cells (CD45+CD3-CD68-CD163-), and stromal cells . The second panel was used to identify fibroblasts based on αSMA and FAP positivity. Both panels included a pan-cytokeratin antibody to identify neoplastic tumor cells and DAPI to identify nuclei. The 1.0.1 version of this dataset contains all cells and tissue regions that passed quality control thresholds.