Objective: Neurosurgery is the first-line treatment for acromegaly. Whether metabolic disorders are reversible after neurosurgery is still debated. The meta-analysis aimed to address the following questions: 1) Does neurosurgery affect glycolipid metabolism? 2) Are these effects related to disease control or follow-up length?
Design: a meta-analysis and systematic review of the literature.
Methods: Three reviewers searched up databases until August 2019 for prospective trials reporting glycometabolic outcomes after neurosurgery. Three other extracted outcomes, all assessed the risk of bias.
Results: Twenty studies were included. Neurosurgery significantly reduced fasting plasma glucose (FPG) [effect size (ES) -0.57 mmol/L, 95% CI -0.82 to -0.31; P<0.001], glucose load [ES -1.10 mmol/L, 95% CI -1.66 to -0.53; p<0.001], glycosylated haemoglobin (HbA1c) [ES -0.28%, 95% CI -0.42 to -0.14; P<0.001], fasting plasma insulin (FPI) [ES -10.53 mU/L, 95% CI -14.54 to -6.51; P<0.001], homeostatic model assessment of insulin resistance (HOMA-IR) [ES -1.98, 95% CI -3.24 to -0.72; P=0.002], triglycerides (TGDs) [ES -0.28 mmol/L, 95% CI -0.36 to -0.20; P<0.001] and LDL-cholesterol (LDL-C) [ES -0.23 mmol/L, 95% CI -0.45 to -0.02 mmol/L); P=0.030] and increased HDL-cholesterol (HDL-C) [ES 0.21 mmol/L, 95% CI 0.14 to 0.28; P<0.001]. Meta-regression analysis showed that follow-up length - not disease control - had a significant effect on FPG, with the greatest reduction in the shortest follow-up (beta=0.012, SE=0.003; P=0.001).
Conclusions: Neurosurgery improves metabolism with a significant decrease in FPG, glucose load, HbA1c, FPI, HOMA-IR, TGDs, and LDL-C and increase in HDL-C. The effect on FPG seems to be more related to follow-up length than to disease control.