Paradoxical Growth Hormone Response to an Oral Glucose Load – Not always Acromegaly!

Background: The normal response of growth hormone (GH) to oral glucose (OGTT) is an initial suppression reaching a nadir within 2 hours followed by a stimulatory rebound after 2 hours. Paradoxical elevation of GH has been reported in patients with acromegaly (1-4). However the literature with currently available GH assays is sparse. 

Hypothesis: We studied the paradoxical response of GH to oral glucose loading in patients with and without acromegaly in order to assess its utility in confirming the diagnosis. 

Methodology: Retrospective chart review of GH suppression tests by oral glucose performed from 2000 to 2012 at the Cleveland Clinic. Acromegaly was excluded based on normal IGF-1or GH nadir <0.4 ng/ml or tumor negative for GH stain. Patients with an inconclusive work up or on medical therapy for acromegaly were excluded. Patients were considered to have a paradoxical GH response if their post-glucose GH levels were higher than the corresponding basal GH level within the first 2 hours without a preceding nadir and an absolute post-glucose GH value was > 1ng/ml.

Results: Of the 104 patients studied, 40 (38%) had acromegaly and of the 8 (7.6%) patients who had a paradoxical GH response, 6 patients had acromegaly but 2 did not. In these 2, 1F & 1M, BMI was 20 and 53 kg/m² respectively. In the morbidly obese patient, IGF-1 was elevated at 365 ng/ml (Z score = 3.2) at the time of the OGTT. Basal GH level was 0.15 ng/ml, peak GH level at 60 min was 4.74 and reached a nadir of 0.38 ng/ml at the end of 2 hours. On follow up in 6 months a repeat IGF-1 level was 246 ng/ml (Z score = 1.3), MRI of the pituitary was normal and a repeat OGTT showed a GH nadir of 0.09 ng/ml at 60 minutes followed by a rise to 7.38 at 90 minutes. In the other patient the IGF-1 level was 443 ng/ml (normal <267), basal GH was 0.15 ng/ml and the peak GH level at 60 minutes was 1.71 and reached a nadir to 0.12 ng/ml at the end of 2 hours. Pituitary MRI was normal. In the acromegaly group (3M, 3F) with the paradoxical response, median (range) basal GH and IGF-1 levels was 2.35 (0.9-30) ng/ml and 484 (327-1030) ng/ml respectively. There was a statistical trend for a paradoxical GH response to occur more frequently in acromegaly patients compared to non-acromegaly patients (P 0.052).

Conclusion: A paradoxical GH response to oral glucose load cannot be used by itself to confirm the diagnosis of acromegaly. The paradoxical GH rise was not reproducible in the non-acromegaly patients.