Examination of PHOX2B in adult neuroendocrine neoplasms reveals relatively frequent expression in phaeochromocytomas and paragangliomas

John P Lee,Yin P Hung,Thomas M O'Dorisio,James R Howe,Jason L Hornick,Andrew M Bellizzi
DOI: https://doi.org/10.1111/his.13243
Histopathology
Abstract:Aims: Paired-like homeobox 2b (PHOX2B) is a transcription factor with expression outside of the central nervous system restricted to neurons and chromaffin cells of the autonomic nervous system. Germline mutations cause congenital central hypoventilation syndrome and predispose to neuroblastoma and Hirschsprung disease. Among paediatric small round cell tumours, PHOX2B is neuroblastoma-specific. Two studies of adult autonomic nervous system tumours (n = 62) produced conflicting results (all tumours stained in one; expression restricted to 40% of paragangliomas in the other). We examined PHOX2B expression in a large cohort of phaeochromocytomas and paragangliomas, as well as well-differentiated neuroendocrine tumours (WDNETs) and poorly differentiated neuroendocrine carcinomas (PDNECs). Methods and results: Tissue microarrays (TMAs) were constructed from 609 tumours: 111 phaeochromocytomas, 146 paragangliomas, 250 WDNETs, and 102 PDNECs. PHOX2B immunohistochemistry was scored for extent (%) and intensity (0-3+), and an H-score (extent × intensity) was calculated. PHOX2B expression was seen in 32% of phaeochromocytomas and in 47% of paragangliomas. Mean/median H-scores for these tumours were in the 30-55 range (i.e. weak to moderate staining). No WDNETs and only 7% of PDNECs stained, the latter often strongly. In a representative cohort of corresponding whole sections (n = 55), the results in WDNETs and PDNECs were unchanged, whereas half of the phaeochromocytomas/paragangliomas that were negative on TMAs became focally, weakly positive. Conclusions: We found frequent, weak to moderate PHOX2B expression in phaeochromocytomas/paragangliomas and no expression in WDNETs, which could be diagnostically useful in the distinction of these tumours. Expression in a minority of PDNECs probably reflects the transcription factor lineage infidelity that is characteristic of this tumour class.
What problem does this paper attempt to address?