Multi-omics integration unravels four molecular subgroups of corticotroph pituitary neuroendocrine tumours with distinct clinicopathological features

Abstract

Abstract Corticotroph pituitary neuroendocrine tumours (PitNETs)/adenomas are heterogeneous sellar neoplasms. Currently established histopathological classification approaches are often considered limited in fully capturing the clinical and biological complexity of these tumours. Thus far, a molecular-based classification has not been established in corticotroph PitNETs. We compile molecular data of 270 corticotroph PitNETs (111 internal, 159 external), encompassing epigenome, transcriptome, and proteome profiles. Comprehensive integrative analyses are performed to identify, validate and characterise definitive molecular subgroups. Corticotroph PitNETs separate into four robust and clinicopathologically distinct molecular subgroups, which are broadly distinguishable by microscopy using SSTR1, GATA3 and SSTR5 immunohistochemistry. An integrated stratification model incorporating these molecular subgroups demonstrates significant prognostic utility. Our findings support the establishment of a refined molecular-based corticotroph PitNET classification, the full clinical value of which will require validation in prospective studies. To facilitate future research, we provide an easy-to-use epigenomic classifier for corticotroph PitNETs.