We present a case with a male born with univentricular heart (double inlet, double outlet left ventricle). Main pulmonary artery banding during childhood had failed to prevent pulmonary hypertension. Thus, establishment of Fontan circulation (connection of the superior and inferior vena cava to the pulmonary arteries) was contraindicated. Therefore, at 21 years of age, palliation of cyanosis was obtained with a bidirectional Glenn anastomosis (connection of the superior vena cava to right pulmonary artery). Due to shunting from the superior to the inferior vena cava, he developed increasing cyanosis which motivated closure of the azygos vein with a catheter-delivered metal coil. At 40 years of age, he developed signs of heart failure and had visible venous cutaneous collaterals. We examined his suitability for heart transplantation. We wanted to determine the pulmonary flow and vascular resistance, but measurement of the pulmonary blood flow is challenging in patients with complex cyanotic congenital heart disease with pulmonary blood supply from separate sources with different oxygen saturations. In these patients, standard invasive measurements using thermo dilution or Fick’s principle cannot be applied. In some patients, direct flow measurements in the pulmonary arteries can be done by phase-contrast MRI. However, many patients have metal implants that cause severe artifacts on MRI making flow measurements impossible. In this case, MRI flow measurements were attempted but were inconclusive due to artifact. Figure 1 shows an MRI scan of the superior mediastinum with profound artifact from a surgical metal clip. To examine the cardiopulmonary circulation and perfusion, a dynamic 15O-water PET scan was performed using bolus injection of 15O-water. Four hundred megabecquerels in 5 mL saline (1 mL s−1) was injected in a peripheral vein using an automatic injection system, followed by a 35-mL saline flush (2.0 mL s−1) as previously described (Nielsen et al. 2019). Together with the PET scan, a low-dose CT scan was performed for anatomical reference which allowed for the identification of pulmonary arteries, lung lobes, cardiac chambers, and the Glenn anastomosis. Regions of interest were drawn on the CT images and transferred to the dynamic PET.
Figure 2 shows a dynamic CT angiography scan (left image) after a single contrast bolus and a 15O-water PET scan (right images) of the heart and lungs up to 70 s after bolus injection of 15O-water performed on separate dates. CT and PET scans are shown as maximum intensity projections (MIP). While the dynamic CT angiography scan showed that both lungs were indeed perfused, the 15O-water PET scan demonstrated that the blood from the superior vena cava primarily was led to the upper lobe of the left lung (Fig. 2, arrow—LL) which was the only lobe properly perfused. This is further illustrated by the time-activity curves of the lung lobes (Fig. 3). The time-activity curves show that a significantly lesser amount of blood was led to the left lower lobe compared to the left upper lobe and almost nothing reached the right lung. Moreover, the perfusion of the lobes of the right lung was delayed. Eventually, small quantities of blood reached the right lung approximately 30 s later than the left (Fig. 2, arrow—RL). This indicates that the blood from the superior vena cava reached the ventricle before it reached the right lung. Figure 4 shows fused images of the 15O-water PET scan and a low-dose CT scan performed simultaneously of the mediastinum and demonstrates the obtained PET information related to the anatomy of the mediastinum. The myocardial blood flow (MBF) of the ventricle was the same as a normal left ventricle (0.9 ml/ml/min). Time-activity curves are extracted from the major vessels and used for calculation of transit times and stroke volume using the distance between centroids and the area under the curve, respectively (Nielsen et al. 2019). The stroke volume index (SVI) in the aorta was low at 28 ml/m2, whereas 36 ml/m2 would be expected from a healthy person measured with this method. Furthermore, the mean transit time from the superior vena cava to the left ventricle was 28 s, which was approximately four times longer than would be expected in a healthy person. The scan demonstrated a comprehensive shunting of blood from the superior vena cava to the inferior vena cava through the internal thoracic vein and venous cutaneous collaterals (Fig. 2 and Fig. 4, arrow—collateral). Thus, approximately 35% of blood from the superior vena cava bypassed the lungs which was unknown before the PET scan.
15O-water is the gold standard for PET quantification of regional tissue perfusion. As a freely diffusible tracer, 15O-water has been intensively used to study the cardiopulmonary system including MBF and lung blood flow (Bergmann et al. 1984; Driessen et al. 2017; Harms et al. 2011; Heinonen et al. 2013). To our knowledge, this is the first report demonstrating that 15O-water PET can be used for evaluation of cardiopulmonary perfusion in complex congenital cyanotic heart disease.