Twenty-one patients were included. Five of them were controlled after surgery, 10 received LA-SSA monotherapy, and six were on combination therapy of LA-SSA + pegvisomant (LA-SSA + PEG). Baseline characteristics are summarized in Table 1. In our population, three patients had central hypothyroidism, and three patients had primary hypothyroidism; all were treated with levothyroxine. Six patients had central hypogonadism, four of them were postmenopausal women without hormonal replacement therapy, one of the two men was on testosterone substitution, and the other one was not because of a urologic contra-indication. Five patients received hydrocortisone substitution for central hypocortisolism. All patients were on stable substitution doses for at least 6 months. None of the participants had diabetes insipidus. There was no significant difference between the different treatment groups for central hypogonadism, central hypothyroidism or hypocortisolism.
Although BMI and WHR were numerically lower in the medically treated groups, there were no significant differences in baseline anthropometric parameters. There was a borderline significant difference in serum IGF1 levels (p = .05) with post hoc analysis showing a lower IGF1 in SUR versus LA-SSA (p = .018). These differences persisted after correction for BMI (overall p = .019; post hoc p = .006). Differences in parameters of glucose metabolism are presented in Table 2. No significant differences were observed between groups for HbA1c, fasting serum glucose, fasting serum insulin and lipid levels or fasting serum GIP and GLP-1 levels nor for glucose disposal rate during the hyperinsulinaemic-euglycaemic clamp.
The results of the mixed meal test are shown in Figure 1. Despite the graphical impression for a trend towards a higher glucose and insulin in the LA-SSA group and a higher GIP in the surgery group, the postprandial responses in glucose, insulin, triglycerides (TG), GIP or GLP-1 levels after the mixed meal test did not differ significantly in the whole group. The post hoc analysis for AUC GIP did show a higher AUC GIP in the surgery group versus the LA-SSA + PEG group (p = .044). Figure 2 shows associations between IGF1 levels and parameters reflecting glucose metabolism, insulin sensitivity and postprandial substrate metabolism. In the whole cohort, significant positive correlations between serum IGF1 levels and HbA1c, fasting serum glucose and insulin and postprandial AUC of glucose and insulin were found. In addition, there was an inverse association between IGF1 levels and M-value. There were neither significant correlations between serum IGF1 levels and lipid levels or postprandial incretin responses, nor were there significant correlations between IGF1 levels and WHR or BMI (all p > .05 with the lowest p = .147 for AUC TG).
Results of the mixed meal test. Al results are represented as median ± interquartile range. AUC, area under the curve; GIP, glucose-dependent insulinotropic polypeptide; GLP-1, glucagon-like peptide 1; LA-SSA, long-acting somatostatin analogues; PEG, pegvisomant; SUR, surgery; TG, fasting triglyceride
Correlation between IGF1 levels and parameters reflecting glucose metabolism and insulin sensitivity (A, correlation between IGF1 and fasting serum glucose; B, correlation between IGF1 and fasting serum insulin; C, correlation between IGF1 and HbA1c; D, correlation between IGF1 and M-value, E, correlation between IGF1 and AUC glucose; F, correlation between IGF1 and AUC Insulin). AUC, area under the curve; HbA1c, haemoglobin A1c; IGF1, insulin growth factor 1
Clin Endocrinol. 2021;95(1):65-73. © 2021 Blackwell Publishing