Endocrine Abstracts

JOINT3036

Background: Growth hormone (GH) is secreted in a pulsatile manner from the anterior pituitary in healthy individuals. In acromegaly, most commonly caused by a GH-secreting pituitary adenoma, the ensuing GH patterns are traditionally described as erratic and irregular due to a substantial shift toward non-pulsatile release. Given the clinical reliance on serum GH and insulin-like growth factor I (IGF-I) as diagnostic and therapeutic biomarkers, it is crucial to establish evidence-based data on GH secretion dynamics in acromegaly. Further investigation into the temporal patterns of GH release may refine diagnostic criteria and improve treatment monitoring in affected individuals.

Aim: The aim of the study is to investigate the day-to-day reproducibility and variability of GH secretion.

Methods: We analyzed medical records of acromegaly patients who received standard treatment and care between 2012 – 2021, including those with a minimum of 3 three GH profiles in conjunction with an oral glucose tolerance test (OGTT) composed of 11 measurements over 3 hours following an overnight fast. To quantify variability in GH secretion patterns, cross correlation analysis and dynamic time warping (DTW), a non-linear alignment algorithm, were applied. DTW distances were computed to assess intra-individual stability across repeated measurements and inter-individual variability among each patient.

Results: We obtained and analyzed 650 GH profiles from 106 patients, of whom the diagnostic GH profile was available in 21 patients. 33 patients were controlled by surgery alone and 72 patients were treated medically with a somatostatin analogue (SA) alone (n = 54) or in combination with a GH antagonist (n = 18). The inter-individual DTW distance was 0.136 with 95%Cl (0.135; 0.138) whereas the intra-individual DTW distance was 0.050 (95%Cl 0.046; 0.055), yielding a ratio of the intra-/inter-individual DTW distance of 0.367 (95%Cl 0.335; 0.403), p <0.0001. The intra-/inter-individual DTW distance was independent of treatment modality. Also, the intra-individual cross-correlation coefficients were significantly higher (0.166 (95%Cl 0.137; 0.195)) than inter-individual correlations (0.003 (95%Cl -0.001; 0.007)), P < 0.0001, indicating greater consistency in GH secretion patterns within individuals over time.

Conclusions: 1. The unique and highly conserved GH pattern in serum from individual patients challenges the dogma of disorderly GH secretion in acromegaly.

2. This periodicity is independent of disease control and treatment modality suggesting an intrinsic clock within the pituitary somatotrope that is preserved after tumorous transformation.

3. It remains to be studied if our findings depend on the oral glucose load.