Serial magnetic resonance imaging in hypoplastic left heart syndrome gives valuable insight into ventricular and vascular adaptation

Abstract

Objectives: This study sought to investigate changes in magnetic resonance imaging (MRI) ventricular volumes and vascular dimensions before hemi-Fontan (HF) and before total cavopulmonary connection (TCPC) in children with hypoplastic left heart syndrome (HLHS).

Background: The systemic right ventricle (RV) in HLHS is subject to significant changes in volume loading throughout the surgical stages of palliation, particularly after the HF.

Methods: Fifty-eight patients had paired pre-HF and pre-TCPC MRI for assessment of changes of RV volumes, neoaortic flow, and vascular dimensions.

Results: Comparison of pre-HF and pre-TCPC MRI results showed a decrease of indexed RV end-diastolic volume and end-systolic volume (98 ml/m(2) to 87 ml/m(2) and 50 ml/m(2) to 36 ml/m(2), respectively) with stroke volume remaining constant (49 ml/m(2) vs. 51 ml/m(2)), leading to an increased RV ejection fraction (51% vs. 59%). These findings persisted after excluding the 3 patients who underwent tricuspid valve repair as part of their HF procedure. Indexed RV end-diastolic volume plotted against neoaortic stroke volume demonstrated a Frank-Starling-like curve that shifted upward after HF. The indexed distal left and right cross-sectional pulmonary artery areas were reduced after HF.

Conclusions: In HLHS, serial MRI shows the adaptation of the systemic RV after HF with volume reduction in the context of a preserved stroke volume and an increased ejection fraction. The staged palliation in HLHS may be a risk factor particularly for reduced left pulmonary artery growth in itself as no factors investigated in this study were found to significantly impact on this.

Figure 1

Measurement of Branch Pulmonary Arteries After Norwood and Hemi-Fontan, and Assessment of Proximal and Distal Aortic Dimensions (A) Measurement of branch pulmonary arteries after the Norwood procedure showing the 2 planes for measuring (i) distal right pulmonary artery; (ii) proximal left pulmonary artery; and (iii) narrowest left pulmonary artery. (B) Measurement of branch pulmonary arteries after the hemi-Fontan showing the 2 planes for measuring (i) distal right pulmonary artery; (ii) proximal left pulmonary artery; and (iii) narrowest left pulmonary artery. (C) Assessment of proximal aortic dimensions showing the 2 planes for measuring (i) native aorta; (ii) neoaortic root; (iii) ascending aorta; and (iv) transverse arch. (D) Assessment of distal aortic dimensions showing the 2 planes for measuring (i) upper descending aorta; and (ii) descending aorta at the diaphragm.

Figure 2

Patients Undergoing TCPC Between February 2003 and July 2010 at Evelina Children's Hospital The 71 patients indicated by (*) had magnetic resonance imaging (MRI) before total cavopulmonary connection (TCPC). The 58 patients indicated by (§) were used to assess remodeling before and after hemi-Fontan (HF). Five patients had assessment of pulmonary vascular resistance in a combined MRI-cardiac catheterization procedure.

Figure 3

Indexed RV EDV Plotted Against Indexed Neoaortic Stroke Volume Indexed right ventricular (RV) end-diastolic volume (EDV) plotted against indexed neoaortic stroke volume showing an upward and leftward shift in the curve (which resembles the Frank-Starling curve) after hemi-Fontan (HF), suggesting improved efficiency. Curves are calculated using fractional polynomial regression based on the 58 paired magnetic resonance imaging (MRI) scans; solid lines = predicted fit, dashed lines = 95% confidence intervals. MRI1 is pre-HF; MRI2 is pre-total cavopulmonary connection.

Figure 4

Within-Patient Change in Loading Conditions, RV iEDV, and iSV Between MRI1 and MRI2 Change in loading conditions, right ventricle (RV) indexed end-diastolic volume (iEDV), and indexed stroke volume (iSV) within individual patients (n = 58) between magnetic resonance imaging (MRI) scan 1 (MRI1), before hemi-Fontan (HF), and MRI2, before total cavopulmonary connection (TCPC). Each patient acts as his or her own control with the baseline volumes for each patient before HF zeroed to the reference point (0,0). Thus, each vector relates to the change in volume loading for an individual patient between MRI1 and MRI2, with an improvement in cardiac efficiency highlighted by vectors traveling to the left of the line of identity, which occurs in the majority of patients (67%).