1346354 - Rapid CMR-based simulation successfully guided and predicted the outcome of a complex Fontan intervention

Description of Clinical Presentation: We present herein a 14-year-old Fontan patient with heterotaxy, left atrial isomerism, bilateral superior caval veins (SCVs), hepatic drainage to the common atrium and a  hemiazygos vein to the left SVC. The patient was palliated with a Kawashima-type TCPC at the age of 2 and an 18mm conduit from the liver veins to a position opposing the right SVC (Figure 1A). The patient presented with symptoms of desaturation at rest (81-85%) and dyspnea during mild exercise, which further worsened during the recent year.

Diagnostic Techniques and Their Most Important Findings: CT raised suspicion of widespread pulmonary arterio-venous malformations (PAVM) predominantly in the left lower lobe which were demonstrated on cardiac catheterization, presumed due to significantly diminished hepatic flow (containing hepatic factors) to the left lung. CMR showed very little flow in the hypoplastic central pulmonary artery, indicating a very small fraction of total hepatic blood (approximately 8%) reaching the left lung with PAVM’s.

Computational fluid dynamics (CFD) based on CMR data was used in planning of a 2-staged endovascular intervention, starting first with dilating the central pulmonary artery (Figure 1 B, top). Once the CFD model was constructed, simulation run-times were approximately 8 minutes. CFD predicted negligible improvement of hepatic flow, which was confirmed in post-interventional CMR.

In the planned second stage intervention 6 months later, an open cell stent was placed between the conduit and the right side of the previously inserted stent. Thereafter, a covered stent was placed in the conduit in order to create a 7 mm fenestration at the superior end for hepatic blood flow to RPA (Figure 1B, bottom). The fenestration size was pre-interventionally tuned in CFD based on inflows from CMR to achieve a hepatic flow to LPA of approximately 20% (Figure 1C). Fluoroscopic cine-images of contrast injection showed significant flow through the central PA to the left PA (Figure 1D, right). There were no intraprocedural complications in either of the interventions.

The post-interventional course was uneventful, and the patient was discharged two days after the procedure, with increase in oxygen saturation from 81-85% to 93% at 6 month follow up, and significant improvement in exercise capacity and quality of life.

Post-interventional CMR after 12 months showed that the hepatic flow distribution increased from 8% to 20%, as predicted by pre-interventional CFD. An additional post-interventional CMR 18 months post-intervention showed a further increase to 35% hepatic blood to LPA.

Learning Points from this Case: Rapid CMR-based simulation was extremely useful to provide the optimal, minimally invasive endovascular intervention. Notably, complementary simulations were also performed during the on-going final intervention (Figure 1 C-D) to evaluate results based on hemodynamic findings as they occurred, confirming the utility of lean and clinically integrated CFD in support of invasive interventions.