DescriptionPituitary neuroendocrine tumours (PitNETs) are common intracranial tumours with widespread heterogeneity. Current recommendations on histopathological evaluation of PitNETs includes an assessment of Ki-67 proliferation index (PI) and immunohistochemical detection of anterior pituitary hormones as a minimum. Ki-67 has been utilized in the histopathological assessment of PitNETs since the 1980s and remains the first line proliferative marker in PitNETs. However, despite reliance on Ki-67 as a prognostic biomarker controversy remains about its clinical value and the recommended cut-off of ≥3%. The aim of this study was to perform an anonymized retrospective analysis of histopathological data from resected PitNETs in a tertiary referral centre and subsequently pilot the utility of QuPath digital image analysis software in Ki-67 PI assessment. We present available histopathological data on anterior pituitary hormone expression in 271 PitNET samples resected between January 2011 and August 2019. Using Northern Ireland Biobank infrastructure 196 MIB1 antibody stained whole face sections completed as standard best care were subsequently retrieved, scanned and digital images then examined using QuPath digital image analysis software. Hormone negative PitNETs (n=110) were the most common resected PitNET by immunohistochemical analysis, while thyrotropinoma was least common (n=5). QuPath software was successfully used to identify PitNET cells and calculate Ki-67 PI. An inverse relationship was found between cell count and Ki-67 PI which may impact Ki-67 cut-offs. This highlights the importance of standardized approaches to Ki-67 assessment and we propose that digital image analysis strategies will have a role in complimenting expert neuropathological evaluation of PitNETs in the future.
|Period||27 Nov 2020|
|Event title||Irish Endocrine Society 44th Annual Meeting|