Redefining hemodynamic instability in pheochromocytoma surgery through latent class analysis

OBJECTIVE: Hemodynamic instability (HDI) during pheochromocytoma surgical resection remains a major anesthesiologic challenge. The definition of HDI is often arbitrary, limiting comparability and risk stratification. We aimed to develop an objective, data-driven classification of intraoperative HDI. METHODS: Hemodynamic variables, including systolic blood pressure (SBP), diastolic blood pressure (DBP), heart rate (HR), and mean arterial pressure (MAP), were collected at 5 intraoperative time points. Latent class analysis (LCA) was used to classify patients into 2 groups: low-risk (l-HDI) and high-risk (h-HDI) HDI classes. We compared 4 LCA models based on different combinations of variables (SBP + DBP, SBP + DBP + HR, MAP, MAP + HR) to identify the optimal classification of HDI, using AIC and BIC for model selection. RESULTS: Among 92 patients, the LCA model based solely on intraoperative MAP yielded the best classification of HDI (AIC: 3783; BIC: 3824). Two latent classes were identified: high-risk HDI (n = 32, 34.8%) and low-risk HDI (n = 60, 65.2%). MAP values were significantly different between the 2 groups during anesthesia induction (89 vs 102 mmHg, P < .001), pneumoperitoneum induction (83 vs 94 mmHg, P = .002), gland manipulation (95 vs 131 mmHg, P < .001), and immediately post-adrenal vein ligation (79 vs 83 mmHg, P = .039). Multivariable logistic regression showed that patients' age (OR 1.04 per year; 95% CI 1.01-1.02; P = .047) and symptomatic presentation (OR 15.35; 95% CI 4.19-56.2; P < .001) were strongly associated with HDI. CONCLUSION: This study proposes a reproducible and objective classification of intraoperative HDI in pheochromocytoma surgery based on MAP-derived LCA.