Introduction

Staged palliation for patients with functionally univentricular hearts concludes with the Fontan procedure, with the goal of completely separating the systemic and pulmonary circulations. Survival metrics following the Fontan procedure have improved in the current era with advancement in surgical techniques and post-operative care [1, 2]. However, there remains significant post-operative morbidity including arrhythmias, pleural effusions, ventricular dysfunction, liver dysfunction, thrombotic events, protein-losing enteropathy, and renal failure [2, 3]. This highlights the importance of careful patient evaluation to determine which patients are suitable for Fontan completion. Traditional pre-Fontan assessment has included echocardiography coupled with cardiac catheterization to define anatomy and assess hemodynamic parameters [4]. There is significant practice variation in children with single ventricle circulation prior to their Fontan palliation [5]. Cardiac MRI has proven to be a useful, non-invasive means for assessing single ventricle patients including anatomy, ventricular function, and hemodynamics [6]. Specifically, cardiac MRI is the only available tool by which a quantitative measurement of aortopulmonary collateral flow can be obtained [7]. Several studies have argued that cardiac catheterization may no longer be required prior to Fontan [8,9,10]. We hypothesize that pre-Fontan assessment combining both cardiac MRI and catheterization provides a complete anatomic and hemodynamic assessment while reducing ionizing radiation exposure, iodine-based contrast load, and affording the opportunity to perform targeted interventions prior to the Fontan without affecting short-term surgical outcomes.

Materials and Methods

Patients

This is a single-center, retrospective cohort study. The catheterization database was queried for single ventricle patients undergoing pre-Fontan cardiac catheterization and cardiac MRI between October 2018 and April 2022. Institutional review board approval was obtained for the study through Baylor College of Medicine. Early in our experience, referral for pre-Fontan evaluation with a combined assessment versus a catheterization only was solely determined by the preference of the referring cardiologist and the interventional cardiologist. Patients who were known in advance to require angioplasty, stent placement or angioplasty of an existing stent were more likely to be referred for catheterization only. More recently, we have adopted the approach of performing combined cardiac MRI-catheterization procedures in all children presenting for pre-Fontan assessment unless there is a contraindication to performing cardiac MRI such as pacemaker or predetermined need for prolonged intervention at the time of catheterization. Query of the catheterization database revealed 41 unique combined procedures during the study period. These procedures were matched randomly by month of procedure to patients undergoing cardiac catheterization only. There were three patients during the study period who had undergone multiple pre-Fontan catheterizations. Therefore, only the most recent catheterization being used to evaluate Fontan candidacy was included in the study. In the combined group, the MRI and catheterization were performed on the same day. Patient demographic, procedural, and imaging data were collected. In addition, patients were reviewed to determine if they have undergone Fontan completion through May 2022. For these patients, surgical procedural and post-operative data were collected.

MRI Protocol

All patients in the combined group were placed under general anesthesia and intubated prior to their cardiac MRI. Cardiac MRI was performed on a Siemens 1.5-T magnet. The MRI protocol is listed in Fig. 1.

Fig. 1
figure 1

MRI Protocol describes the 4 components of our group’s MRI protocol. MRI magnetic resonance imaging

Full size image

With four distinct sections of the cardiac MRI, a full assessment of single ventricle anatomy and function was performed. Gadolinium contrast was used to enhance imaging of the branch pulmonary arteries, systemic and pulmonary veins, collateral vessels, and the aortic arch. Special attention was given to aortic arch anatomy in patients who have previously undergone aortic arch reconstruction as a part of their single ventricle palliation. Following image acquisition, image processing was performed in real time by the MRI cardiologist or radiologist using Vitrea software. Flow data calculations including the ratio of pulmonary to systemic blood flow, differential pulmonary blood flow, collateral burden estimates, and magnetic resonance angiograms were reviewed prior to initiating cardiac catheterization.

Catheterization Protocol

Cardiac catheterization procedures for patients in both groups were routinely performed with patients under general anesthesia. Pre-Fontan catheterization was performed routinely with internal jugular, femoral venous, and femoral arterial access when necessary if prograde left heart catheterization could not be performed. Following venous and arterial access, patients were administered heparin to achieve activated clotting times of > 250 s.

Hemodynamic measurements were performed with patients mechanically ventilated and without supplemental oxygen. Pressure and oxygen saturations were measured in the superior vena cava, pulmonary arteries, inferior vena cava, atria, single ventricle, pulmonary vein(s), and aorta. The Lafarge standard oxygen consumption table was used to estimate patients’ oxygen consumption [11]. A complete blood count was performed on every patient to determine hemoglobin concentration at the time of the catheterization procedure. Standard hemodynamic data were calculated using the Fick principle including cardiac index, ratio of pulmonary to systemic blood flow, transpulmonary gradient, and pulmonary vascular resistance.

Standard angiograms for patients who did not have cardiac MRI prior to cardiac catheterization included superior vena cava, inferior vena cava, ascending aorta, descending aorta, and pulmonary vein(s). In patients in the combined group, an angiogram was routinely performed in the superior vena cava. An aortic root angiogram was performed at the discretion of the interventional cardiologist.

Cardiac MRI and Cardiac Catheterization Protocol

The combined procedures were performed in a combined MRI and catheterization suite under the same ventilator conditions, on room air. The MRI magnet is separated from the catheterization lab by a specialized “barn” door. In the combined group, cardiac MRI was performed first immediately following intubation. The process map for combined procedures is shown in Fig. 2.

Fig. 2
figure 2

cMRI + CC process map shows the process map for patients undergoing a cMRI + CC beginning with general anesthesia induction and intubation and ending with cardiac catheterization. cMRI + CC cMRI and cardiac catheterization

Full size image

After completion of the cardiac MRI patients were moved to the cardiac catheterization laboratory under the same anesthesia. Patients were then prepped for catheterization which was undertaken as the second component of the procedure. Patients were extubated after completion of the cardiac catheterization.

Statistical Analysis

Data were compared between the combined group and catheterization alone group. Statistical analyses were performed using SPSS version 28 (IBM, Armonk, NY). Normality of continuous variables was assessed using histograms. Descriptive statistics are presented as medians with interquartile range for continuous variables with nonparametric distribution, means with standard deviation for continuous variables with parametric distribution, and aggregates with percentages for categorical variables. Demographic information was compared using Chi-square or Fisher Exact for categorical variables where appropriate, Mann–Whitney- U test for non-parametric continuous variables, and t-test for parametric continuous variables with p < 0.05 demonstrating statistical significance.

Results

During the study period, there were 77 unique patients with 37 in the combined group and 40 in the catheterization alone group.

Patient demographics and procedural data are summarized in Table 1.

Table 1 Demographics and Procedure data
Full size table

There was no significant difference in age and weight between the groups. Duration of intubation and total anesthesia time were significantly longer in the combined group. Patients in-lab time, defined as total time in the catheterization lab, and catheterization procedure time, defined as the start of access to sheath removal, were significantly lower in the combined group. Total contrast dose/kilogram and total fluoroscopy time were significantly lower in the combined group as well. The median radiation exposure measured by dose area product was lower in the combined group but was not statistically significant.

There were significantly more interventions performed during catheterization in the catheterization only group. In the combined group, 7/37 (19%) procedures included at least one intervention while in the catheterization only group, 19/40 (48%) procedures included intervention (p = 0.008). The interventional procedures in the combined group included 2 aortopulmonary collateral occlusions, 4 venovenous collateral occlusions, 1 pulmonary artery angioplasty, and 1 internal jugular vein angioplasty. One patient in this group who underwent aortopulmonary collateral occlusion had greater than 30% collateral burden as calculated by MRI and an elevated single ventricle end diastolic pressure. The second patient who underwent AP collateral occlusion did not have greater than 30% aortopulmonary collateral burden calculated by MRI but the single ventricle end diastolic pressure was severely elevated therefore aortopulmonary collateral occlusion was performed. The interventional procedures in the catheterization only group included 11 aortopulmonary collateral occlusions, 4 venovenous collateral occlusions, 1 pulmonary artery angioplasty, 1 main pulmonary artery test occlusion, 3 coarctation stent angioplasties, and 1 pulmonary vein stent angioplasty. The AP collateral occlusions in this group were performed due to subjective assessment of high collateral burden on angiography. Without objective data from the MRI, the decision to occlude AP collaterals is operator dependent. All patients in the catheterization only group who underwent AP collateral occlusion excluding one had a normal single ventricle end diastolic pressure. The pulmonary artery angioplasty was performed due to anatomic narrowing visualized on angiography. Coarctation angioplasty was performed due to an elevated pressure gradient across the site.

The median time from cardiac catheterization to Fontan operation was 6 months. Table 2 describes outcomes for Fontan completion in patients who have undergone the final stage of single ventricle palliation.

Table 2 Fontan outcomes
Full size table

The number of patients who have undergone Fontan completion to date is not significantly different between the groups. Additionally, bypass time, intensive care unit length of stay, and duration of chest tubes were also not significantly different between the groups.

Discussion

This study shows the utility of obtaining cardiac MRI in patients undergoing pre-Fontan catheterization prior to Fontan palliation. The 2 groups were similar in demographic characteristics. This study shows that using cardiac MRI as a diagnostic tool prior to catheterization significantly decreases the catheterization procedure time, iodinated contrast use, and intervention rate (aorto-pulmonary collateral occlusion) in children with single ventricle physiology undergoing evaluation for Fontan. The median radiation dose was lower in the combined procedure group but was not statistically significant. In patients who have undergone Fontan completion, there was no difference in the ICU length of stay or chest tube duration in children who underwent catheterization only compared to children in the combined group.

This study highlights the use of cardiac MRI as complimentary method to assess Fontan candidacy. In our study, MRI was used to assess the aortopulmonary collateral flow prior to catheterization. Cardiac MRI has been used previously to accurately assess the amount of collateral flow in children with single ventricle circulation. Grosse-Wortmann et al. described the use of MRI to assess aortopulmonary collateral flow in patients prior to Fontan completion [12]. In this study, the aortopulmonary collateral flow was correlated with hospital length of stay and duration of chest tube drainage. Patients whose chest tube drainage lasted less than a week had less aortopulmonary collateral flow before the Fontan operation compared to patients with a longer duration of chest tube drainage [12]. Similarly, Glatz et al. demonstrated an association of systemic to pulmonary collateral flow obtained via cardiac MRI with length of stay and chest tube drainage [13]. Goldstein et al. in the multicenter C3PO registry, demonstrated that aortopulmonary collateral occlusion remained one of the most common interventions performed during pre-Fontan catheterization [5].

While MRI likely did not affect the surgical management or the peri-operative ICU management, the pre-operative medical management was different in patients who underwent both a catheterization and MRI. There were notably less aortopulmonary collateral occlusions performed in the combined group compared to the cardiac catheterization only group. In the combined group, objective MRI collateral flow data were used to guide intervention for AP collateral occlusion. Greater than or equal to 30% collateral flow by MRI calculation was used as an indication for AP collateral occlusion. In the catheterization only group, decision regarding AP collateral occlusion was made by subjective assessment of quantity of collateral flow as well as single ventricular end diastolic pressure (> 10 mmHg). This likely led to increased rate of AP collateral occlusion in the catheterization only group. Cardiac MRI aortopulmonary collateral flow data are more precise, resulting in decreased number of interventions by virtue of the ability to perform targeted interventions based on objective data. This can result in a significant decrease in catheterization procedure time, contrast use, and only performing clearly defined/indicated interventions. In patients who underwent Fontan completion, there was no difference in the post-operative length of stay or pleural drainage when the two study groups were compared.

The actual catheterization procedure time was significantly less in the combined group vs the catheterization only group. This is due to decreased need for angiographic assessment of the aortic arch, aortopulmonary collaterals, pulmonary vein, hepatic veins, and inferior vena cava anatomy. This may also be associated with decreased interventions required in the combined group. However, the total anesthesia time was considerably longer in the combined group as the patient spent additional time under anesthesia in the MRI scanner. This is an important consideration that can be optimized with improved workflow to decrease the transition time from the MRI suite to catheterization laboratory. Further enhancement in the MRI protocol could further reduce the time spent in the MRI suite including obtaining magnetic resonance angiography early in the MRI for ease of acquisition of phase contrast flows earlier in the procedure.

The iodinated contrast dose per kilogram was significantly less in the combined group than the catheterization group, therefore decreasing the adverse effects. This is due to decreased need for angiography as well as decreased intervention rates in the combined group. While patients in the combined group also received Gadolinium, this is a non-iodinated contrast agent and was administered at a low dose of 1–2 mL per patient. The multicenter C3PO data published the median contrast dose, fluoroscopy time and case duration [5]. The C3PO study shows comparable data to our catheterization only group and lower contrast dose, fluoroscopy time as well as case duration in the combined group.

Our study also demonstrates that children who have a combined assessment compared to catheterization only had similar postoperative outcomes. The ICU length of stay and chest tube duration was similar in both groups.

Fogel et al. has previously demonstrated that single ventricle patients who do not require intervention can have successful Fontan palliation when preoperative physiology and anatomy was assessed by echo and MRI only as compared to children who underwent catheterization only [14]. Cardiac MRI provides a good anatomical assessment as well as additional physiological data such as pulmonary to systemic blood flow ratio, relative pulmonary blood flow, collateral burden, and ventricular function. In the catheterization lab, the assessment of these parameters, aside from pulmonary to systemic blood flow ratio, is usually subjective and operator dependent. A cardiac MRI provides more consistent estimate of these parameters. Like our study, Fogel et al. observed similar surgical outcomes in these patients with no difference in length of stay or incidence of pleural effusion [14]. Our study did not assess any group without catheterization as catheterization is considered a standard of care prior to Fontan completion at our institution. Also, this study focused on obtaining cardiac MRI at the time of pre-Fontan catheterization which has not been demonstrated before.

Limitations

The retrospective nature of the study limits the ability to derive causative relationships between the group characteristics and outcomes. Additionally, the patient grouping in our early experience was dependent on the preference of the referring cardiologist and interventionalist, creating a selection bias.

Consideration must also be given to the added cost of performing cardiac MRI in these children who are already burdened with lifelong testing and care. Although we did not specifically assess the financial impact of adding cardiac MRI, it is possible that the cost of MRI would be offset by the decreased need for intervention being performed. In the future, the ability to perform catheterization in the MRI suite could lead to decreased cost as well as total anesthesia time.

Conclusion

In conclusion, pre-Fontan cardiac MRI provides valuable information which leads to the ability to perform targeted interventions as well as a decrease in contrast use, fluoroscopy time, and catheterization procedure time. In our experience, there is no difference in short term surgical outcomes when combined cardiac MRI catheterization procedures are performed as a part of the pre-Fontan assessment. In the future, longitudinal data will be collected to assess long-term outcomes including survival data, surgical re-intervention and catheter based for re-intervention post-Fontan including the need for collateral occlusions.