Brain imaging determinants of functional prognosis after severe endocarditis: a multicenter observational study

Area of Science:

• Neurology
• Radiology
• Infectious Diseases

Background:

• Infective endocarditis (IE) can lead to severe neurological complications in critically ill patients.
• Cerebral imaging is crucial for assessing these complications, but its prognostic value is not fully understood.

Purpose of the Study:

• To develop a detailed imaging phenotype of cerebral complications in severe IE patients.
• To determine if specific neuroimaging patterns can predict patient prognosis.

Main Methods:

• 192 critically ill patients with definite left IE and neurological complications underwent cerebral imaging.
• Voxel-based lesion-symptom mapping (VLSM) was used to define lesion types, volumes, and locations.
• Logistic regression analyses explored associations between imaging features, prognostic variables, and 6-month outcomes (modified Rankin Scale - mRS).

Main Results:

• Ischemic lesions were most common (75%), followed by microbleeds (50%) and hemorrhages (20%).
• Univariate analysis showed associations between hemorrhagic transformation volume, leukopathy severity, and motor pathway lesions with poor mRS.
• Multivariate analysis revealed age and cardiac surgery status, not imaging patterns, as independent predictors of mRS.

Conclusions:

• No specific neuroimaging pattern independently predicted prognosis in severely ill IE patients with neurological complications.
• Prognosis was primarily influenced by patient age and cardiac surgery status.