Fig. 2.

Pan-genomic classification of PPGLs. By in-depth transcriptome profiling, PPGLs generally adhere to one out of four principal expressional clusters, the Wnt pathway cluster, the kinase-associated cluster, the pseudohypoxia-associated cluster, and the cortical admixture cluster. The Wnt cluster is enriched for pheochromocytomas with somatic MAML3 gene fusions as well as CSDE1 gene mutations, and these tumors have an increased risk of metastasizing compared to kinase associated and cortical admixture cluster lesions. The kinase associated cluster is built-up primarily of pheochromocytomas with low metastatic potential. Tumors in this cluster exhibit mutations in kinase-associated pathways, and genes include NF1, RET, HRAS, MAX, TMEM127, and KIF1B. The pseudohypoxia cluster includes pheochromocytomas and paragangliomas with tricarboxylic acid (TCA) cycle mutations (SDHx gene family, multiple other TCA cycle related enzymes) or mutations in hypoxia-inducible factor (HIF)-associated signaling networks (VHL, EPAS1/HIF2-α, PHD1/EGLN2, and PHD2/EGLN1). This cluster contains a subgroup of cases (mostly TCA cycle aberrant) with global hypermethylation, which is a feature generally associated to metastatic disease and worse clinical outcomes. The fourth group (cortical admixture) is represented by pheochromocytomas with kinase-associated mutations and low risk of metastatic disease. In all, 24.3% of patients with PPGLs adhering to the pseudo-hypoxia cluster exhibit metastatic disease, compared to 11.4% of patients with Wnt cluster PPGLs and 4.1% of patients with PPGLs associated to the kinase cluster [5]. In contrast, patients with PPGLs associated to the cortical admixture cluster almost always demonstrate a benign clinical course [29]