Balancing pulmonary blood flow: Theory, in vitro measurements, and clinical correlation of systemic-to-pulmonary shunt banding.

Background: Size mismatch between body and a systemic-to-pulmonary shunt can result in excessive pulmonary blood flow, compromising systemic oxygen delivery. Previously reported techniques to mechanically restrict shunt flow lack precision and reproducibility. We developed a formula for shunt banding and assessed its efficacy and reproducibility by in vitro and clinical measurements. Methods: Formulas to determine diameter reduction, length of banding, and effect on the ratio of pulmonary blood flow (Q(p)) to systemic blood flow (Q(s)) were established. In vitro measurements of different shunt grafts were performed. Results were compared with calculations and clinical data. Clinical outcome was retrospectively assessed in all patients (n = 8) who underwent a shunt banding procedure at our institution between 2008 and 2012. Results: Our formulas can adequately predict the length of the band based on the desired diameter and shunt type or on the Q(p):Q(s) mismatch. In vitro measurements correlated with the manufacturer's specifications in small shunts (<= 5 mm diameter; 0.45 mm mean wall thickness). The calculated diameters of these shunts were closely correlated with in vitro measurements (r = 0.953; P = .001). Arterial saturation, pH, and calculated Q(p):Q(s) decreased significantly with banding (P = .026, .002, and .004, respectively). Clinical effects varied among patients, with hemodynamically stable patients achieving the most benefit. Adjustment of the band was required in 1 patient. No shunt thrombosis or shunt banding-related complications were noted. Conclusions: Our formulas and surgical strategy offer a new approach to controlling excessive pulmonary blood flow in shunt-dependent circulations in an effective and predictable way. The best reproducibility was achieved in small, thin-walled shunts. This strategy was most effective in patients with pulmonary overcirculation without hemodynamic decompensation.