Prediction of postoperative hypoglycemia in patients with pheochromocytoma resection: a retrospective study
摘要
Hypoglycemia after pheochromocytoma resection is one of the most common complications. The factors predicting postoperative hypoglycemia are not well established. The objective of this study was to develop a model to predict hypoglycemia in patients after pheochromocytoma resection.
MethodsThis single-center retrospective study enrolled adult patients who underwent pheochromocytoma resection between September 2019 and August 2025. Logistic regression identified risk factors for postoperative hypoglycemia, based on which a nomogram model was developed. Discrimination and calibration analyses were conducted to assess the performance of the model.
ResultsA total of 114 patients were enrolled in the final analysis. The median age was 56.5 years, and 53.5% were females. The overall incidence of hypoglycemia after pheochromocytoma resection was 27.2%. The length of hospital stay after surgery was longer in patients with hypoglycemia than those without (P = 0.020). In the multivariable logistic regression analysis, patients who developed hypoglycemia were younger (odds ratio [OR] = 0.95, 95% confidence interval [CI] 0.92 to 0.99, P = 0.015), and were more likely to have elevated preoperative plasma metanephrine level (OR = 7.73, 95% CI 2.00 to 29.85, P = 0.003) and to receive beta blocker intraoperatively (OR = 3.67, 95% CI 1.40 to 9.57, P = 0.008). The nomogram model showed good discrimination (area under receiver operating characteristic curve: 0.798, 95% CI: 0.714 to 0.882) and calibration (Hosmer-Lemeshow test P = 0.720) for predicting postoperative hypoglycemia.
ConclusionsHypoglycemia was common after pheochromocytoma resection and was associated with poor prognosis. The developed nomogram model incorporating age, elevated preoperative plasma metanephrine level, and intraoperative beta-blocker use can effectively predict postoperative hypoglycemia in patients undergoing pheochromocytoma surgery.
Clinical trial numberNot applicable.