Cardiovascular Magnetic Resonance catheterization derived pulmonary vascular resistance and medium-term outcomes in congenital heart disease

Patients

Median age was 3.6 years (6 days-67 years) and weight 13.8 kg (2.3 -122 kg). There
was a wide range of underlying diagnoses described in Table 1. Seventeen patients had an associated diagnosis of trisomy 21. Ninety-seven patients
had previous surgical or catheter interventions. Fifteen had a single repeat CMR/XMR
study, and 3 patients had 2 repeat CMR/XMR studies. One hundred and fifty four were
combined XMR catheterizations and 13 were solely CMR guided cardiac catheterizations.
Median radiation dose for CMR/XMR catheterization was 3.6 Gycm² (0.72 mSv) (range 0–57.5 Gycm²), with a median screening time of 11.2 minutes (range 0–58.2 minutes). There were
2 procedural complications; pulmonary hemorrhage and pulmonary embolism with good
recovery and no procedural deaths.

Table 1. Primary structural cardiac lesions of the study population

Four patients were lost to long term follow up having moved abroad. Follow up until
death or last medical review was available for the remaining patients over a median
of 4.2 years (4 days to 11 years). The 4-day duration was related to a death from
extracardiac complaints explained later in the manuscript. Eighty-four studies led
to a cardiac intervention (catheter or surgical) at a median interval of 94 days (0–414
days). The remaining patients either had medical or conservative treatment.

Pulmonary vascular resistance and clinical outcomes

Pulmonary vascular resistance results are reported separately for biventricular and
functionally single ventricle circulations.

Biventricular circulations

One hundred and eight PVR studies were performed in biventricular circulations. Forty-seven
led to a cardiac intervention and 61 followed a conservative path or non-cardiac intervention.
Median PVR was 2.8 WU.m² (range 0.4-66.0). Sixty-seven studies were in a group with lesions that allowed shunting
between the systemic and pulmonary circulations or potentially vice-versa. Of these,
41 went on to have a cardiac catheter or surgical intervention to eliminate or substantially
reduce the shunt. The PVR values of this interventional group of patients are displayed
in Figure 1. ROC analysis showed that a baseline Qp:Qs ?2.75 in biventricular circulations with
left-to-right shunts predicted a PVR ?6 WU.m² with 100% sensitivity and 48% specificity (Figure 2 and Figure 3). This translates to a negative predictive value of 100% and a positive predictive
value of 37%.

Figure 1. PVR in patients with biventricular circulations and left to right shunts undergoing
interventions post CMR/XMR catheterization. Patient death is marked in black.

Figure 2. Relationship of PVR to Qp:Qs in biventricular circulations with aorta-to-pulmonary
shunts.

Figure 3. ROC curve for the relationship of PVR to Qp:Qs in biventricular circulations with
aorta-to-pulmonary shunts. The area under the curve is 0.829.

Reversibility with 20 ppm iNO and 100% oxygen were performed in 52 studies with systemic
to pulmonary shunts, with a fall in mean PVR from 3.3 to 2.9 WU.m²(p?=?0.03) and a rise in mean Qp:Qs ratio from 2.2 to 2.4 (p?=?0.02). Further interrogation
showed a mean reduction of PVR by 12.7% (range 0-38%) in those undergoing interventions
and 10.8%(range 0-34%) in those who were managed medically (p?=?0.4).

A baseline PVR of around 6 WU.m² and lower was considered suitable for biventricular surgical repair, accepting the
need for fenestrated closure of left-to-right (L-R) shunt lesions in the borderline
cases. Those with a PVR??6 WU.m² remain alive post intervention with only one death, unrelated to PVR (PVR of 2 WU.m²) in a patient with pulmonary atresia, ventricular septal defect and major aortopulmonary
collaterals who required multiple surgical and catheter interventions to address residual
proximal branch PA narrowing. This low PVR group included four adult patients with
chronic L-R shunts (AVSDs or VSD) who went on to have surgery at a mean age of 46 years
(range 35.6-52.8 yrs) and remain alive with no evidence of pulmonary hypertension.

Six patients with biventricular circulations had a fenestrated, rather than complete
closure of their intracardiac shunt lesions due to the marginally elevated PVR, three
of whom had a PVR just above 6 WU.m² (6.2, 6.3 and 6.7 WU.m²). Of these, one patient died following surgical intervention despite fenestrated
closure, from persistent pulmonary hypertension. This patient had an elevated pre-surgical
PVR of 6.2 WU.m² falling to 4.8 WU.m² in response to maximum pulmonary vasodilation during the PVR study and a Qp:Qs that
remained static at 1.1.

Of the 26 patients with systemic to pulmonary shunt lesions who were not operated,
11 were because of an elevated PVR of 6 WU.m² (median 15 WU.m² range 6.6-66 WU.m²) and remain alive on medical treatment. The remaining 15 patients with PVR 6 WU.m² (median 2.4 WU.m², range 0.6-5.3 WU.m²) were managed conservatively as the defects were judged to be small. There was one
death in this low PVR group in a patient with a ventricular septal defect, VACTERL
association (Vertebral anomalies, Anal atresia, Cardiac defects, Tracheo-oesophageal fistula and or Eosophageal atresia, Renal and radial anomalies and Limb defects) and chronic lung disease (PVR 1.6 WU.m²) with a laryngeal cleft and airway difficulties.

Forty one patients had a biventricular circulation without a lesion that allowed shunting
between the systemic and pulmonary circulations or potentially vice-versa (median
PVR 3.1 WU.m² (0.6-31.4 WU.m²). Four patients with low PVR went on to have catheter or surgical intervention with
3 stenting of the right ventricular outflow tract or branch pulmonary arteries, and
one balloon of the pulmonary valve. Two patients with a PVR greater than 6 WU.m² (19 and 6.2 WU.m²) underwent catheter-guided occlusion of aortopulmonary collaterals. The patient with
a PVR of 19 WU.m² had a previously repaired hypoplastic aortic arch and died from pulmonary hypertension
despite embolization of a right MAPCA and initiation of sildenafil. The second patient
with Scimitar syndrome survived.

Of the remaining patients without shunt lesions, eight had no cardiac interventions
but underwent a liver transplant with good outcomes. The remaining 27 patients have
had no interventions and remain alive at last follow up. Eleven had normal PVR and
were managed conservatively (median PVR 1.8 WU.m², range 1.0-3.4) and 16 were treated with oral pulmonary vasodilators for varying
periods of time (median PVR 8.0, range 2.0-42.3 WU.m²).

Functionally univentricular circulations

Standard care for the preoperative assessment of patients undergoing single ventricle
surgical palliation at our center has not involved cardiac catheterization since 2002.
Patients progressing along the path of surgical single ventricle palliation undergo
a CMR with simultaneous measurement of central venous pressure by means of jugular
venous cannulation. Fifty-nine catheter studies were performed in 50 patients with
functionally univentricular circulations where there were concerns regarding an elevated
PVR. Thirty-seven studies led to a cardiac intervention, and 22 followed a conservative
course. Median PVR was 2.3 WU.m² (range 0.9-6.6 WU.m²).

Forty-two studies were performed in 35 patients prior to completion of Hemifontan
or Fontan. Twenty patients went on to have completion of the Fontan circulation (Figure 4) despite a raised PVR in 4 patients (PVR 4.6, 3.9, 3.3 and 3.1 WU.m²). The median pulmonary artery pressure pre-Fontan was 10.4 mmHg (range 6-18 mmHg).
It is our institutional practice to routinely fenestrate all Fontans. There was one
long-term death in a patient with a PVR of 4.6 WU.m², reducing to 3.6 WU.m² with iNO and 100% oxygen undergoing a high risk Fontan on sildenafil therapy. There
was one death (PVR 1.8 WU.m²) due to plastic bronchitis 85 days post completion of Fontan.

Figure 4. PVR in patients with univentricular circulations undergoing staged surgical palliation
to completion of Fontan post CMR/XMR catheterization. Patient deaths are marked in black.

Of the 15 patients who did not have Fontan completion, 3 had a PVR??4 WU.m² and were medically treated for pulmonary hypertension, with one death. Of patients
managed conservatively with a PVR??4 WU.m², one is awaiting surgery for completion of Fontan upon reduction of body mass index,
and two were deemed unsuitable for Fontan completion despite a PVR 4 WU.m² due to primary ciliary dyskinesia/cystic fibrosis in one, and branch pulmonary arteries
that were considered too small in the other. Six patients remain palliated with cavopulmonary
shunts and have not been referred for pre-Fontan assessment as yet. Three patients
died pre Fontan from complications unrelated to PVR. The first patient died following
surgery for unrelated endocarditis of the mitral valve. The second patient had a sudden
arrhythmic death. The third patient was a neonatal death following withdrawal of care
40 days after a redo of pulmonary artery banding and atrial septectomy due to multiple
congenital anomalies and associated complications from necrotising enterocolitis.

Fifteen patients had 17 studies post-Fontan completion (median 2.1 WU.m², range 0.9-4.3 WU.m²) of whom 12 were treated conservatively or with medical therapy. The remaining 3
led to catheter interventions, which consisted of coil occlusion of aortopulmonary
collaterals (PVR 2.2 WU.m²), lateral tunnel fenestration creation and stenting (PVR 2.9 WU.m²) and fenestration closure (PVR 1.4 WU.m²). All remain alive.