Endocrine Abstracts

Background: It has been reported that some Magnetic Resonance Imaging (MRI) pituitary findings may reflect specific endocrinological alterations, as in the case of adult-onset growth hormone deficiency (GHD), which has already been associated with lower MRI-derived pituitary height and volume in pediatric patients.

Purpose: The objective of this study was to investigate potential associations between pituitary morphology and signal intensity on MRI with GH secretory function in a cohort of patients with varying degrees of adiposity.

Methods: We conducted a retrospective observational study on 297 patients (235 females, median age 46 years, IQR: 20 years), admitted to our institution between January 2015 and December 2023, who had signs and symptoms suggestive of GHD. Our cohort included 243 patients with obesity and 54 age- and sex- matched controls. We assessed GH-IGF1 axis and pituitary morphology with MRI. To quantify the mean and standard deviation (SD) of pituitary signal intensity, we used Horos software, employing T2-weighted sequences and gray matter intensity as a normalizer. In addition, we measured pituitary height and calculated pituitary area in coronal section as a surrogate for pituitary volume.

Results: In the entire study population, we found an inverse correlation between BMI and pituitary area (r=-0.357, P=.000), height (r=-0.406, P=.000) and normalised pituitary signal intensity (r=-0.197, P=0.05). Patients with obesity showed significantly lower pituitary height and area (P=.000) along with a higher prevalence of empty sella (X2= 13.996, P=.000) compared to controls. Regarding GH secretory capacity, we found a direct correlation between the area under the curve of the GHRH+arginine test and pituitary area (r = 0.479, P=.000) and height (r=0.513, P=.0.00) and a negative correlation with BMI (r= -0.481, P=.000). Finally, by fitting ROC curves, we identified cut-offs for pituitary area (lower than 27.8 mm^2, AUC= 0.696, P=.000) and pituitary height in coronal scan (lower than 3.1 mm, AUC= 0.729, P=.000) as predictors of GHD with a sensitivity of 69% and 72% and specificity of 61% and 62%, respectively. After stratifying the cohort by the degree of adiposity, we showed a progressive reduction in GH secretory capacity and pituitary size from the control group to gradually higher degrees of obesity (P=.000 for all parameters).

Conclusion: Consistent with current evidence, our study demonstrates that patients with obesity have impaired GH-IGF1 axis and reduced pituitary size compared to controls, suggesting a close relationship between pituitary morphology and functional capacity.