Postoperative B-type natriuretic peptide monitoring for the assessment of the magnitude of shunting through Blalock-Taussig anastomoses

Abstract Background The Modified Blalock-Taussig shunt (MBTS) is the most common palliative operation performed in patients with complex cardiac defects. Postoperative morbidity and mortality rates are high, mainly due to shunt thrombosis and over-shunting. Over-shunting may be difficult to identify postoperatively based on conventional criteria. Since plasma B-type natriuretic peptide (BNP) concentrations correlate with the magnitude of shunting in various left-to-right shunt cardiac defects, we investigated its ability to identify postoperative MBTS over-shunting. Methods and results This retrospective, observational study included 42 consecutive patients (median age 9.50days, IQR: 6.00–58.25) undergoing MBTS for obstruction of the pulmonary blood flow at a tertiary referral pediatric cardiac center. The BNP threshold concentrations which accurately predicted outcome and MBTS over-shunting were derived using the ROC methodology. 443 BNP concentrations were analysed. The presence of atrio-ventricular valve regurgitation was the most important component of overall variance (72.75%). In 34 patients without regurgitation, BNP concentrations were predictive of a duration of mechanical ventilation >8days and of intensive care stay >11days, with ROC areas of 0.655 [0.597–0.719], 0.650 [0.589–0.711], a negative predictive value for the >1035pgmL −1 threshold of 0.93 and 0.96 respectively. SaO 2 was less accurate for the prediction of both outcomes. In patients in whom the pulmonary flow was entirely MBTS-supplied, a BNP concentrations >1052pgmL −1 was predictive of a pulmonary-to-systemic ratio>2. Conclusion In MBTS patients without atrio-ventricular valve regurgitation, maintaining BNP below 1000pgmL −1 may represent a therapeutic target to avoid over-shunting.