It is not possible to establish an etiology in around 30-40% of ischemic Cerebrovascular Accidents (CVA). This happens frequently, even following exhaustive clinical research. It is possible that a fraction of these CVAs, called cryptogenic strokes, occurs after an episode of paroxysmal atrial fibrillation (AF), which cannot be documented. AF-related CVAs are generally serious, with an estimated mortality of 50% in the first year. Recurrence rate is high and leads to morbidity and significant costs. Correct identification of a cardioembolic etiology in this type of patients is important, since it has been demonstrated that oral anticoagulants, compared with antiaggregating therapy, reduces the risk of recurring CVAs in patients with nonvalvular AF, and may be associated with less serious CVAs in case of recurrence.

Current complementary diagnostic means have a low capacity for detecting paroxysmal AF. Despite the Holter test being superior to the electrocardiogram, in a systematic review, Holter’s capacity (24-72 hours) to detect paroxysmal FA was 4.6%. Longer monitoring (4-7 days) has detected AF in a further 8% of patients. Prolonged 7-day long recordings, at 0 and 6 months following a CVA, have detected AF in 14% of patients.

Accordingly, it is necessary to consider alternative ways of detecting an eventual cardioembolic etiology in cryptogenic CVA. The importance of correctly identifying a CVA as cardioembolic has led to the possibility of using biomarkers in its diagnostic.

Previous studies have demonstrated a rise in the serum concentrations of the Brain Natriuretic Peptide (BNP) during the acute phase of ischemic CVA when compared with those found in controls and with the cut-off points established for heart failure. The BNP is a neurohormone that has been studied and disseminated as a natriuretic cardiac hormone. However, recent studies suggest it may play an important role in the physiopathology of acute CVA.
In chemical terms, it is a peptide initially produced as a prohormone called pro-BNP. Pro-BNP is subsequently divided into biologically active BNP and N-terminal-proBNP (NT-proBNP), which has no biological activity.

Most tests measure NT-proBNP, and until now the production of BPN has been attributed to two areas: the brain and, above all, the heart. Here, myocyte stretching is the main stimulus for BNP synthesis and secretion. Under a normal physiological state, the auricula is the main place for cardiac production. Catecholamine, angiotensin II and endothelin may stimulate the secretion of BNP, through paracrine or endocrine mechanisms. It is presumed that NT-proBNP is eliminated from circulation through kidney excretion. The biological effects of BNP include natriuresis, diuresis, vasodilation, inhibition of the do renin-angiotensin-aldosterone axis and the sympathetic nervous system. This peptide is currently used as a marker of left ventricular and prognostic dysfunction in patients with congestive heart failure (CHF) and with acute coronary syndromes. High plasmatic levels of NT-proBNP were noted in patients suffering from kidney failure, essential hypertension and disrhythmias such as AF. Advanced age and low levels of haemoglobins are associated to increased plasmatic levels of BNP.

Despite BNP having been firstly identified in brain tissue, current information about this natriuretic peptide in cerebral-vascular disease is limited. Some works to date have shown a sharp rise in NT-pro-BNP during the acute phase of ischemic CVA.

Four possibilities have been put forward for increasing NT-proBNP during acute ischemic CVA:

– These patients often suffer from heart failure, and BNP increase might reflect ventricular dysfunction.
– It might be related to AF, which is a risk factor for CVA and a cause for BNP increase.

– As the brain is a place for BNP production, it might occur after lesion in the cerebral parenchyma.

– Due to its sympathetic-inhibitive effects, BNP increase may emerge as a response to increased sympathetic activity.

In a previous work carried out as part of an original dissertation in the Master Degree in Neurosciences at the Faculty of Medicine of the University of Lisbon (FMUL), we attempted to determine whether NT-proBNP increase in the acute phase of ischemic CVA might have a cardiac origin. The results suggested that the increase in the serum concentration of NT-proBNP, which occurs in the context of CVA, has a cardiac origin. We noted that patients with ischemic cardioembolic CVA had statistically significant higher average concentrations of NT-proBNP compared with patients with non-cardioembolic ischemic CVA.

No association was found between the serum concentration of NT-proBNP and the extent and location of the CVA, or blood pressure values. NT-proBNP showed a very good capacity to diagnose cardioembolic CVA associated with atrial fibrillation in the first 72 hours after ischemic CVA, with the ROC curve analysis showing two cut-off points with great sensitivity and specificity.

To identify a CVA of cardioembolic etiology, a value of area below the AUC curve of 0.77±0.06 was achieved. The cut-off point with the greatest specificity and sensitivity was established at 265.50 pg/mL (71.4% and 74%, respectively). To identify a CVA of cardioembolic etiology with AF, an AUC of 0.92 ± 0.03 was obtained. The NT-proBNP value established as a cut-off point was 265.50 pg/mL (sensitivity 94.4%, specificity 72.9%) and 912.0 pg/mL (sensitivity 55.5%, specificity 97.9%).

A further study in a different sample including a higher number of patients is now in progress, in order to validate the established cut-off points. The objective is also to identify possible sources of variation in NT-proBNP, and to establish the profile for temporal evolution of NT-proBNP following ischemic CVA. The work is supervised by Professor José Ferro and Professor Dulce Brito, with the collaboration of Dr. Sampaio Matias and of the doctors from the Cerebrovascular Accidents Unit of Santa Maria Hospital Neurology Service.
Ana Catarina Fonseca
Neurology Service
Santa Maria Hospital
catarinagfonseca@gmail.com