ESPE Abstracts

Introduction: Growth hormone deficiency (GHD) is the most frequent endocrine deficit in childhood survivors of brain tumours. However, there is insufficient evidence to guide the timing of growth hormone replacement (GHR). At Great Ormond Street Hospital timing is based on clinical need rather than in relation to oncological treatment. Therefore, sufficient variability in GHR timing is available to analyze its effect on tumour progression and recurrence.

Aim: To examine the association between tumour progression or recurrence and time of GHR initiation.

Methods: Retrospective analysis of children with brain tumours on GH replacement from 2000-2023. GHR timing was defined as time from the end of oncological treatment for primary disease or previous progression/recurrence until GH initiation. Adjusted Hazard Ratios (HR) for first and further recurrence/progression were estimated through Cox Regression.

Results: Data from 75 survivors of brain tumours have been analyzed so far. Gliomas (38.7%), embryonal tumours (25.3%), and craniopharyngiomas (24%) were the most frequent diagnosis. Mean follow-up time was 8.96±3.35 years. GHD was diagnosed by glucagon test, GHR was started at a median of 1.3 years (IQR 0.5-2.2) after the end of oncological treatment. During follow-up, 44% of patients presented relapse/progression: 26.7% during GHR (20% with history of recurrence/progression before GHR, 6.7% with first event after GHR); 17.3% had relapses/progressions before GHR only. Initial analysis showed that starting GHR later post-end of treatment was borderline associated with better 5-year progression-free survival (p 0.05; HR 0.671; 95% CI 0.451-0.999). However, multivariate COX-regression analysis showed no significant decrease in the HR (p 0.431; HR 0.865; CI 0.604-1.240) after adjusting for tumour status, previously progressive disease (≥ 1 previous progression/recurrence), and cranial radiotherapy (CR). Residual disease (p 0.046; HR 5.178; CI 1.032-25.981) and history of previous relapses (p 0.043; HR 3.094, CI 1.038-9.221) were associated with an increase in the adjusted HR for further recurrence/progression, and CR with decreased HR (p 0.016; HR 0.325; CI 0.129-0.814). Analysis of patients without relapse/progression pre-GHR showed no significant increase in progression or recurrence-free survival in relation to GHR timing (p 0.349; HR 0.681, CI 0.305-1.522).

Conclusions: No independent association was found between tumour progression/relapse-free survival and GHR timing when the pre-existing increased risk for tumour progression in patients with residual or previously progressive disease was accounted for. Analysis of the complete cohort of patients is needed to increase the power of the study and drive definitive conclusions.