Endocrine Abstracts

Introduction: Acromegaly is a chronic disease caused by growth hormone (GH) oversecretion. A pituitary macroadenoma is found in mostly of cases, but the secretion pattern of GH and insulin-like growth factor 1 (IGF-1) and the natural history of somatotropinomas are heterogeneous. The objective of the present study was to evaluate the relationship between tumor size, GH and IGF-1 levels in patients diagnosed with acromegaly.

Methods: A cross-sectional, observational, descriptive study was carried out in acromegaly patients (n=25) attending our Neuroendocrinology Outpatient Clinic at the local university hospital. Patients’ anthropometric data, biochemical investigations and adenoma size by MRI at diagnosis were recorded. Basal GH and IGF-1 blood levels (chemiluminescent immunometric assays) were obtained; IGF-1 relative to the upper limit of the normal range (IGF-1_ULN) and the mean standard deviation scores (IGF-1_SDS) were calculated from each patient. ROC curves analysis were performed to determine the discriminative ability of GH, IGF-1, IGF-1_ULN or IGF-1_SDS in predicting the presence of a GH-secreting Macroadenoma (≥10 mm). The results were expressed as sensitivity (S) and specificity (Sp).

Results: Based on the tumor size MRI classification, 76% of the tumors were macroadenomas (1.85±1.08 cm maximal diameter; mean±S.D.). The biochemical variable IGF-1 exhibited the highest area under the curve (AUC=0.88, P<0.01) with a cut-off point of 703.6 ng/ml (S=88%; SP=83%). IGF-1_mean and IGF-1_ULN presented cut off of 4.2 and 2.8, respectively with identical AUC values (0.80); P=0.03, S=82%, SP=83%. GH level’s cut-off was 4.520 ng/ml (AUC 0.73, P=0.11; S=94.12%; SP=50%). The most specific cut-off value for IGF-1 to discriminate between micro and macroadenomas was 1681 ng/ml (SP=100%).

Conclusion: The basal GH level could not discriminate adenoma sizes, while IGF-1 values were able to indicate the presence of macroadenoma. The absolute IGF-1 blood concentration was more effective than relative values.