Optimal timing of blood samplings to detect GH inhibition during oral glucose tolerance test.

Abstract

In patients with suspected acromegaly, evaluation of IGF-I is recommended as first-line test, while the assessment of GH-nadir during oral glucose tolerance test (OGTT) is advised as confirmatory test. The procedure of this test generally involves GH measurement every 30min (30') from baseline to +120' or +180'. However, the optimal timing of samplings for the distinction between patients with or without active acromegaly is still a matter of debate. Sixty-seven healthy subjects and 46 acromegalic patients who achieved documented and persistent long-term cure were enrolled. A greedy algorithm was used to identify the minimal subset of time-points that sufficed to correctly detect GH suppression. The sampling at 90' was the one in which a GH level < 1μg/L was most frequently achieved (i.e., in 91.3% of cured acromegalic patients and in 91.0% of healthy subjects). Considering the whole cohort, the best combination of 2 time-points was +90' and +150' and achieved 95.6% accuracy; the best combination of 3 time-points was +60', +90' and +150' and achieved 99.1% accuracy. The minimal subset of GH determinations that demonstrated perfect accuracy (100%) needed the inclusion of 4 time-points, namely +60', +90', +120' and +150'. A subset of 4 time-points (60' - 90' - 120' - 150') was identified as the most relevant to detect GH suppression at OGTT, with a perfect classification of 100% of subjects. This supports the possibility to restrict the blood samplings to these time-points when assessing disease cure, with possible advantages in terms of saving time and lowering costs.