Cardiac Magnetic Resonance Imaging in the Evaluation and Management of Scimitar Syndrome

Abstract

Background

Scimitar syndrome (ScS) features a partial anomalous pulmonary venous connection to the inferior vena cava, with possible septal defects, right lung hypoplasia, and aortopulmonary collaterals. We hypothesize that cardiac magnetic resonance imaging measurements can elucidate ScS physiology and guide surgical decisions.

Methods

This retrospective cohort study included 26 patients diagnosed with ScS from 2011 to 2022 at Children’s Hospital of Philadelphia. Patients with single-ventricle physiology were excluded. Medical records were reviewed for patient characteristics and cardiac magnetic resonance imaging data.

Results

Indexed right ventricular end-diastolic volume (RVEDV) (P = .007) and indexed total shunt volume (P = .029) were significantly greater in the surgical group vs the nonsurgical group. Total shunt volume had a moderate effect (0.43), and RVEDV had a large effect (0.55) on surgery.

Conclusions

These cardiac magnetic resonance imaging-derived measurements demonstrate that patients with surgically corrected ScS had a higher total pulmonary blood flow, total systemic blood flow, and significantly greater RVEDVs and total shunt volumes. RVEDV had the largest effect on surgery. These data mark a primary step in developing diagnostic criteria for ScS surgery referral using cardiac magnetic resonance imaging-derived measurements. Further research is needed to determine clinical cutoffs.

In Short.

• ▪Cardiac magnetic resonance imaging-derived measurements show that patients with surgically corrected scimitar syndrome had higher total pulmonary blood flow-to- total systemic blood flow ratio and significantly greater right ventricular end-diastolic volumes and larger shunt volumes.
• ▪Right ventricular end-diastolic volume had the largest effect on surgery.
• ▪These are key steps toward establishing cardiac magnetic resonance imaging-based criteria for scimitar syndrome surgical referrals.

Scimitar syndrome (ScS) is a rare congenital heart anomaly characterized by partial anomalous pulmonary venous connection (PAPVC) of the right veins most commonly into the inferior vena cava. ScS affects ∼3 children in every 100,000 live births,¹ and is typically identified early in life accompanying other congenital heart defects. This syndrome’s complex manifestations and complications, including right lung hypoplasia, dextroposition, aortopulmonary collaterals, pulmonary arterial hypertension, pulmonary vein stenosis, heart failure, and diaphragmatic hernia, pose intricate management challenges.¹

Diagnosis is suspected with a chest roentgenogram showing a descending pulmonary vein along the right cardiac border, a hypoplastic right lung, and heart dextroposition. Further evaluation can include echocardiography, computed tomographic angiography (CTA), cardiac magnetic resonance imaging (CMRI), and cardiac catheterization.² Echocardiography has limitations in assessing right ventricular (RV) size and function and delineating PAPVC. CTA requires ionizing radiation and has reduced resolution in children with fast heart rates. Cardiac catheterization has procedural risks and limitations in shunt measurement. CMRI is a noninvasive method to delineate the scimitar vein anatomy using magnetic resonance angiography (Figure) and measure various variables, including PAPVC contributions, coexisting septal defects, RV size, shunt volumes, and aortopulmonary collateral flow.²

Figure.

Scimitar syndrome seen on cardiac magnetic resonance imaging in a 7-year-old patient. All right pulmonary veins connect to the inferior vena cava (IVC) above the diaphragm. The black arrow indicates a small atrial communication. The intrahepatic IVC (∗) and hepatic veins (+) are marked. (Ao, aorta; LA, left atrium; PA, pulmonary artery; RA, right atrium; RMPV, right middle pulmonary vein; RUPV, right upper pulmonary vein; ScV, scimitar vein; SVC, superior vena cava.)

This study was conducted to define significant characteristics of ScS patients with surgical vs nonsurgical interventions using CMRI-derived measurements to aid in establishing management guidelines.

Patients and Methods

This retrospective cohort study included 26 ScS patients from 2011 to 2022 at Children’s Hospital of Philadelphia who underwent CMRI. Committees for the Protection of Human Subjects approved the study. Inclusion criteria included double-ventricle physiology and at least 1 preintervention CMRI. Demographic and clinical characteristics extracted from patient medical records were summarized using frequency (%) for categorical variables and median for continuous variables using the Wilcoxon rank sum test. Body surface area (BSA) was used to comparably analyze infants to adults.³ BSA^1.4 was used to normalize the ventricular volume data.⁴ Wilcoxon test effect size was used to analyze effect size.⁵ Scimitar vein flow was measured directly or indirectly by measuring inferior vena cava flow above and below the scimitar connection. Collateral flow was defined as flow ≥10% of aortic flow. Total shunt volume was defined as total pulmonary blood flow (Q_p) minus total systemic blood flow (Q_s). Right pulmonary artery hypoplasia was determined by echocardiogram Z-score data. R 4.2.2 software (R Foundation for Statistical Computing) was used for all statistical analyses.

Results

Population Characteristics

The demographics and clinical characteristics of the cohort are summarized in Supplemental Table 1. The average age at diagnosis was 6.84 years. Approximately 70% of patients had RV dilation, defined as >100 cm³/BSA^1.4.⁶

Subpopulation Characteristics

Anatomical and flow characteristics of the surgical and nonsurgical groups are reported in the Table. The indexed total shunt volume was significantly greater in the surgical group (3.02 L/min/m²) vs the nonsurgical group (1.91 L/min/m²; P = .029). The indexed RVEDV, as well as the indexed RV/left ventricular measurements, were significantly greater in the surgical group (P = .007) vs the nonsurgical group (P = .034). One patient did not have RVEDV or shunt volume data to report on. Total shunt volume had a moderate effect (0.43) and RVEDV had a large effect⁵ (0.55) on surgery vs no surgery (Supplemental Figure). Average collateral flow is reported in Supplemental Table 2.

Table.

Characteristic Comparison of Surgery vs No Surgery Groups

Characteristic	No.	No Surgery	Surgery	P Valuea
		(n = 17)	(n = 9)
Age at most recent CMRI, y	26	15.4 (2.6-36.0)	3.8 (1.4-10.0)	.200
BSA, m2	26	1.47 (0.59-1.90)	0.70 (0.39-0.87)	.300
Scimitar anatomy
All RPVs to scimitar	20	14 (82.3)	6 (66.6)
1 RPV connecting normally to left atrium	4	1 (5.9)	3 (33.0)
Additional anomalous RUPV to azygos/SVC	2	2 (12)	0 (0)
Qp indexed to BSA, L/min/m2	26	4.75 (4.26-6.47)	6.78 (6.15-8.22)	.130
Qs indexed to BSA, L/min/m2	26	3.10 (2.52-3.33)	3.06 (2.60-3.33)	>.900
Qp/Qs	26	1.72 (1.46-1.93)	2.00 (1.85-3.16)	.063
Average collateral flow, L/min	9	0.62 (0.33-0.63)	0.40 (0.28-0.65)	>.900
Average collateral flow indexed to aortic flow, %	9	17 (12-17)	16 (13-24)	>.900
RVEDV indexed to BSA1.4 (mL/m2.8)	25	88 (79-106)	130 (102-180)	.007
RVEDV (mL) Z score	25	1.3 (−0.1 to 3.7)	5.3 (2.8-7.0)
LVEDV indexed to BSA1.4 (mL/m2.8)	25	54 (48-65)	56 (54-67)	.500
RVEDV/LVEDV	25	1.65 (1.38-1.83)	1.98 (1.82-2.47)	.034
Right ejection fraction average	25	0.629 (0.589-0.679)	0.628 (0.577-0.642)	.800
Left ejection fraction average	25	0.693 (0.660-0.731)	0.683 (0.668-0.706)	.500
Total shunt indexed to BSA (L/min/m2.8)	26	1.91 (1.46-2.85)	3.02 (2.20-7.89)	.029
PFO/ASD contribution, %	25	−5 (−7 to 19)	10 (−9 to 22)	.800
Collateral contribution, %	9	28 (20-38)	19 (14-31)	.700
Scimitar vein, contribution, %	26	83 (37-100)	86 (70-104)	.900

Data are presented as n (%) or median (interquartile range).

ASD, atrial septal defect; BSA, body surface area; CMRI, cardiac magnetic resonance imaging; LVEDV, left ventricular end-diastolic volume; PFO, patent foramen ovale; Q_p, total pulmonic blood flow; Q_s, total systemic blood flow; RPV, right pulmonary vein; RUPV, right upper pulmonary vein; RVEDV, right ventricular end-diastolic volume; SVC, superior vena cava.

^aWilcoxon rank sum test.

RVEDV Characteristics

Presence of an atrial septal defect was significantly associated with RVEDV (β = 37; 95% CI, 3.3-7.1; P = .033) (Supplemental Table 3). Presence of a ventricular septal defect was also significantly associated with RVEDV (β = 90; 95% CI, 35-144; P = .002). Q_p/Q_s was significantly associated with RVEDV (β = 23; 95% CI, 7.5-38, P = .005) as well as the indexed total shunt volume (β = 14; 95% CI, 10-18, P < .001).

Pulmonic and Systemic Flow Characteristics

Indexed RVEDV was significantly associated with Q_p (β = 0.05, 95% CI, 0.03-0.06; P < .001) (Supplemental Table 4). Presence of an atrial septal defect was significantly associated with Q_p (β = 2.7, 95% CI, 1.0-4.4, P = .003). Presence of a ventricular septal defect was also significantly associated with Q_p (β = 3.3, 95% CI, 0.52-6.1, P = .022). An atrial communication was significantly associated with Q_p/Q_s (β = 0.24, 95% CI, 0.05-0.43, P = .014) (Supplemental Table 5).

Comment

CMRI analysis revealed significant physiological differences between surgical and nonsurgical ScS management, aiding to develop surgical referral criteria. Indexed total shunt volume was significantly greater in the surgery group (P = .029), indicating that cumulative larger shunts, regardless of etiology, may warrant surgical correction. An indexed total shunt volume of >3 L/min/m² seems reasonable for considering surgery. The Q_p/Q_s between groups approached significance (P = .063), suggesting that it may be less sensitive than indexed total shunt volume in determining surgical referral.

ScS patients are typically referred for surgery because of hemodynamic factors such as RV volumes and ejection fraction, Q_p/Q_s, and cardiac index, based on clinical symptoms such as respiratory distress, heart failure, and exercise intolerance or from a combination of these factors. Our study focuses on whether CMRI as a quantitative tool can provide hemodynamic data that could directly influence patient referral to surgery or correlate with factors underlying the referral. This research is important because it represents a cohort of patients undergoing comprehensive CMRI flow quantification for ScS management.

Historically, CTA and cardiac catheterization were primarily used for anatomical assessment, exposing patients to ionizing radiation and its risks. Cardiac catheterization also cannot reliably measure Q_p/Q_s in cases involving right lower-lobe pulmonary supply by an aortic collateral due to dual pulmonary arterial supply or coexisting PAPVC of the right upper pulmonary vein to the azygous vein due to pulmonary venous blood that can contaminate systemic venous saturation sampled within the superior vena cava.

CMRI can measure cumulative shunt volume and its components in L/min, whereas cardiac catheterization may inaccurately measure absolute flows due to errors in estimating total oxygen consumption. CMRI can accurately measure Q_p/Q_s in all instances and is the gold standard for quantifying RV volumes and ejection fraction, providing crucial insights into factors affecting ScS management.

Our study showed that surgically managed ScS patients had significantly higher indexed RVEDVs (88 vs 130 mL/m^2.8, P = .007) and an increase in RVEDV assessed by Z score. RVEDV effect size was greater than total shunt volume in relationship to surgery (0.55 vs 0.43), indicating that increased RVEDV may drive surgical referrals. An indexed RVEDV of >130 mL/m^2.8 may also be a surgical indication.

Interestingly, some of our values align with guidelines for pulmonary valve replacement. Specifically, a right ventricular-to-left ventricular EDV ratio >2 is an indication for pulmonary valve replacement, seen in our cohort. An indexed RVEDV >150 to 160 mL/m² is also a surgical indication for pulmonary valve replacement.^7,8 This is understandably slightly higher than the median indexed RVEDV that we saw in our surgical cohort, which indexed the RVEDV to a higher value, BSA^1.4. These correlated metrics stress that severe RV dilation plays an outsized role in the decision to refer for surgical treatment.

The indexed Q_p of the nonprocedural group was 4.74 L/min/m² and was 6.79 L/min/m² for the procedural group, indicating that those with interventions had higher total pulmonary blood flow, possibly contributing to larger RVEDV. In our study, indexed RVEDVs and total shunt volumes emerged as the most critical variables influencing surgical referral.

Our results demonstrate that CMRI flow quantification can also measure the individual contributions of coexisting shunts from aortic collaterals and septal defects. This ability is particularly useful for patients who may not be suitable for surgical repair but could benefit from catheter-based interventions. These findings are the first step toward developing management guidelines for ScS patients.

Our CMRI flow data diverge from the traditional approach and suggest that indexed total shunt volumes may be a better marker of disease severity than Q_p/Q_s, because they better discriminate between the surgical and nonsurgical groups and show stronger correlation with indexed RV volumes. This provides valuable clinical guidance and the basis for which to power future multicenter clinical trials in surgical outcomes.

The study’s limitations include a small sample size, not large enough to provide the statistical power to determine whether ScS patients will benefit from surgery and whether preoperative CMRI data will correlate with postoperative outcomes. Not all patients undergo CMRI, and referral reasons remain unclear. Larger, multicenter studies are needed to determine surgical efficacy and precise surgical indications. This study highlights the central role CMRI measurements have in the management of ScS patients.

Supplementary Data

Supplementary Figure.