Cardiothoracic ratio on chest radiograph in pediatric heart disease: How does it correlate with heart volumes at magnetic resonance imaging?

BACKGROUND

The cardiothoracic ratio by chest radiograph is widely used as a marker of cardiac size.

OBJECTIVE

The purpose of this study is to correlate cardiothoracic ratio and cardiac volumes as measured by cardiovascular magnetic resonance (MR) in common structural and myopathic heart disease with increased cardiac size due to volume overload or hypertrophy.

MATERIAL AND METHODS

A retrospective single center study was performed in all patients between 2007 and 2013 with repaired tetralogy of Fallot (TOF), aortic regurgitation, isolated left-to-right shunt and hypertrophic cardiomyopathy (HCM) who underwent cardiovascular MR and chest radiograph within 6 months of each other. Cardiothoracic ratios by chest radiograph (frontal and lateral) were compared to cardiac volumes (indexed for body surface area) by cardiovascular MR.

RESULTS

One hundred twenty-seven patients (mean age: 11.2 ± 5.5 years) were included in this study (76 with TOF, 23 with isolated left-to-right shunt, 16 with aortic regurgitation and 12 with HCM). Frontal cardiothoracic ratio of all groups correlated with indexed right ventricular (RV) end-diastolic volume (EDVI) (r = 0.40, P < 0.01) and indexed total heart volume (THVI) (r = 0.27, P < 0.01). In TOF patients, frontal cardiothoracic ratio correlated with RVEDVI (r = 0.34, P < 0.01; coefficient of variation = 27.6%), indexed RV end-systolic volume (ESVI) (r = 0.44, P < 0.01; coefficient of variation = 33.3%) and THVI (r = 0.35, P < 0.01; coefficient of variation = 19.6%), although RV volumes and THVI showed widespread variation given the high coefficients of variation. In patients with aortic regurgitation, frontal cardiothoracic ratio correlated with left ventricular (LV) EDVI (r = 0.50, P = 0.047), but not with THVI and aortic regurgitant fraction, and widespread variation for LV EDVI (coefficient of variation = 19.2%), LV ESVI (coefficient of variation = 32.5%) and THVI (coefficient of variation = 13.6%) was also observed. Frontal cardiothoracic ratio was not correlated with cardiac volumes or mass in patients with a left-to-right shunt or HCM. Lateral cardiothoracic ratio showed no correlation with any cardiac volume in all four groups.

CONCLUSION

Although increased cardiothoracic ratio on frontal chest radiograph is associated with increased biventricular volumes in patients with pulmonary and aortic regurgitation, significant variation in ventricular volumes and total heart volume for any given frontal cardiothoracic ratio limits the use of cardiothoracic ratio in monitoring the individual patient's heart size. Frontal cardiothoracic ratio did not correlate with cardiac chamber volumes in patients with a left-to-right shunt or HCM and lateral cardiothoracic ratio offered no additional value for cardiac size assessment.