Clinical parameters-based machine learning models for predicting intraoperative hemodynamic instability in hypertensive pheochromocytomas and paragangliomas patients - Scorecard - MDSpire

Clinical parameters-based machine learning models for predicting intraoperative hemodynamic instability in hypertensive pheochromocytomas and paragangliomas patients

  • By

  • Houming Zhao

  • Lu Tang

  • Zhuoran Li

  • Xintao Li

  • Tongyu Jia

  • Jin Luo

  • Yujie Dong

  • Shangwei Li

  • Xin Ma

  • Peng Zhang

  • September 15, 2025

  • 0 min

Share

Clinical Scorecard: Machine Learning Models Utilizing Clinical Parameters to Forecast Intraoperative Hemodynamic Instability in Patients with Hypertensive Pheochromocytomas and Paragangliomas

At a Glance

CategoryDetail
ConditionHypertensive pheochromocytomas and paragangliomas (PPGLs)
Key MechanismsExcessive catecholamine secretion causing systemic effects and intraoperative hemodynamic instability
Target PopulationPatients with sustained hypertension and unilateral PPGLs undergoing laparoscopic or robotic surgery
Care SettingPreoperative and intraoperative surgical care in tertiary hospital setting

Key Highlights

  • Sustained hypertensive PPGL patients have a threefold higher risk of intraoperative hypertensive emergencies and hypotensive episodes compared to normotensive PPGL patients.
  • Catecholamine secretion affects inflammatory responses, coagulation, and erythrocyte morphology, which may predict intraoperative hemodynamic instability.
  • Machine learning models can integrate clinical, laboratory, and tumor parameters to predict intraoperative hemodynamic instability risk in hypertensive PPGL patients.

Guideline-Based Recommendations

Diagnosis

  • Confirm PPGL diagnosis postoperatively via pathological examination.
  • Define sustained hypertension as office SBP ≥140 mmHg and/or DBP ≥90 mmHg on ≥2 visits or 24-h ambulatory SBP/DBP ≥130/80 mmHg.
  • Evaluate urinary catecholamines and blood catecholamine metabolites preoperatively using high-performance liquid chromatography methods.

Management

  • Preoperative adrenergic alpha-receptor blockade (phenoxybenzamine or terazosin) for at least two weeks before surgery.
  • Use laparoscopic or robotic surgical approaches for tumor excision.
  • Exclude patients with severe inflammation, infection, or hematologic disorders to reduce confounding risks.

Monitoring & Follow-up

  • Monitor intraoperative blood pressure closely to detect hemodynamic instability.
  • Assess inflammatory markers (e.g., SII, WLR, NPR, MLR, PLR) and coagulation parameters (APTT, PT, INR, TT) preoperatively.
  • Use machine learning predictive models trained on clinical and laboratory data to stratify intraoperative risk.

Risks

  • Intraoperative hemodynamic instability may lead to life-threatening complications such as stroke, myocardial infarction, and multiple organ failure.
  • Sustained hypertensive PPGL patients have increased risk of hypertensive emergencies and hypotensive episodes during surgery.
  • Uncontrolled catecholamine release during anesthesia and surgery can precipitate hemodynamic instability.

Patient & Prescribing Data

Patients with sustained hypertension and unilateral PPGLs undergoing laparoscopic or robotic surgery

Preoperative alpha-adrenergic blockade for at least two weeks is standard to reduce intraoperative hemodynamic instability risk.

Clinical Best Practices

  • Exclude patients with active infection, severe inflammation, or hematologic disorders when assessing inflammatory and coagulation markers.
  • Collect comprehensive preoperative clinical, laboratory, and tumor data within one week before surgery for accurate risk prediction.
  • Utilize machine learning models to integrate multidimensional data for individualized intraoperative hemodynamic instability risk assessment.
  • Ensure anesthesia management by an experienced team to maintain blood pressure stability during surgery.

References

Original Source(s)

Related Content