Table 1.
Summary of current published studies utilizing CODEX
| Manuscript | Disease state/tissue | Key Findings | Analytic tools |
|---|---|---|---|
| Ferreira et al., JCI Insight, 2021 | Tissue injury/kidney | Distinct anatomic distribution of immune and epithelial cells after acute kidney injury | CODEX-MAV software |
| Phillips et al., Nature Comm, 2021 | Cutaneous T cell lymphoma/skin | Distance between CD4 + PD1 + T cells, tumor cells, and Tregs, quantified by SpatialScore, correlates with response to checkpoint inhibitor | See SpatialScore in Table 2 |
| Gouin et al., Nature Comm, 2021 | Bladder cancer/bladder tumor | Identified CDH12 expressing epithelial tumor cells that predict response to immune checkpoint therapy | Cell types were identified by k-nearest neighbor and then manual gating. Niches were identified by k = 10 nearest cells |
| Schurch et al., Cell 2020 | Colorectal cancer/colon | The two subsets of colorectal cancer has distinct cellular composition and organization. At the tumor boundary, the CD4+ T cell frequency and the CD4+ to CD8+ T cell ratio are prognostic indicators | See Cell Neighborhoods in Table 2 |
| Mondello et al., Blood Cancer Journal, 2021 | Follicular lymphoma, lymph node | Activated central memory T cells within tumor follicles are associated with improved prognosis | CODEX-MAV software for post-processing. Unet neural network for cell segmentation. Phenograph for cellular communities |
| Jiang et al., Front Immunol, 2021 | Ebola infection/rhesus macaque spleen | Validated 21-marker panel to profile multiple immune cell types and Ebola virus in Rhesus Macaques | See CODEX toolkit in Table 2 |
| Mayer et al., Research Square, 2021 | Ulcerative colitis/colon | Identified inflammatory cell types and cellular neighborhoods that persisted despite treatment with TNFa inhibitors in ulcerative colitis. Developed a model utilizing spatial data to predict resistance to TNFa inhibitor therapy in ulcerative colitis | Utilized CODEX and cell neighborhood toolkits (see Table 2). For cell type identification, the authors performed unsupervised X-shift clustering with a manual gating strategy |