Endocrine Abstracts

Introduction: In 2017 the term ‘Atypical adenoma’ was removed from the WHO classification, and ‘High-risk adenomas’ was introduced as a term to classify those with rapid growth, radiological invasion, and a high Ki-67 proliferation index. In this retrospective cohort study we have sought to identify the clinical and histological characteristics which are associated with worse outcomes.

Methods: Data: We created a dataset of 1793 patients who were discussed at the Pituitary MDT at King’s College Hospital, a tertiary centre in central London from the electronic medical record. The nature and date of treatment was recorded as well as clinical characteristics which were hypothesised to be significant.

Survival analyses were performed using cox regression, events were recorded as time of re-operation or radiotherapy, patients were censored at time of last follow-up within the electronic patient record. Incidences when radiotherapy or re-operation which occurred less than one year after the initial treatment were not included within the analysis as these were treated as primary treatment failure rather than recurrence.

Results: Ki67 was strongly associated with recurrence in a multivariate model (P < 0.05). Hazard ratio 1.08 (95% CE 1.04–1.14). We created risk groups of low (<3%), intermediate (3–9%) and high (<9%) which were also statistically significant (P < 0.05, (when compared with <3%). Hazard ratios were1.7240 ( 95% CE 1.1597–2.563) and 2.8109 (1.6804–4.702) respectively. Male sex was also significantly associated with reccurance (P = 0.003, HR = 1.5873 (1.1648–2.163). The above findings remained the case when only NFPAs were analysed. For non-functioning adenomas, initial presentation with a visual field deficit was a significant hazard for recurrence (P < 0.05, HR = 1.6300 (1.0740–2.474)). There was no significant increase in hazard between silent corticotroph and other types of non-functioning adenoma.

Discussion: These results demonstrate that the previous ‘atypical adenoma’ category may be of clinical value. In this cohort Ki67 could be used to stratify patients by risk.