A Nomogram for Predicting Patent Foramen Ovale-Related Stroke Recurrence

Area of Science:

• Cardiology
• Neurology
• Medical Informatics

Background:

• Patent foramen ovale (PFO) is prevalent in cryptogenic stroke, suggesting a causal link.
• Lack of recognized recurrence risk factors hinders identification and management of high-risk patients.
• This study addresses the need for a predictive model for PFO-related stroke recurrence.

Purpose of the Study:

• To establish a reliable prediction model for PFO-related stroke recurrence.
• To identify key clinical and biochemical factors associated with stroke recurrence in PFO patients.
• To facilitate personalized risk assessment and clinical management strategies.

Main Methods:

• A training set of 392 PFO-related stroke patients and a validation set of 164 patients were used.
• Two nomograms were developed using Cox regression and LASSO regression.
• Model performance was evaluated using concordance index (C-index), AUC, calibration curves, and decision curve analysis (DCA).

Main Results:

• Nomogram 1, based on homocysteine (Hcy), hsCRP, and albumin (ALB), showed strong predictive performance (C-index 0.861 in training, 0.862 in validation).
• Nomogram 1 demonstrated good discrimination with 2-year AUC of 0.863 and 5-year AUC of 0.777 in the training set.
• Calibration and DCA confirmed Nomogram 1's clinical utility and ability to stratify patients into low and high-risk groups.

Conclusions:

• Nomogram 1, utilizing Hcy, hsCRP, and ALB, offers a clinically realistic prognostic prediction for PFO-related stroke.
• This model aids in personalized prognostic evaluations for patients with PFO-related stroke.
• The model facilitates improved patient stratification and management for PFO-related stroke recurrence.