Presented at the Neonatal Society 2005 Autumn Meeting.
Foran A, Fitzpatrick J, Schmitz S, Franklin J, Pamboucas C, Hajnal J, Edwards AD
Division of Clinical Sciences, Imperial College London, Hammersmith Campus, London, UK
Background: Evidence is emerging that cardiac function seems to be an important determinant of outcome in preterm infants. Low superior vena cava flow is common in the first hours after birth and has been associated with subsequent periventricular or intraventricular haemorrhage (1). However cardiac function in the preterm population is poorly understood. Commonly used bedside tests such as blood pressure and capillary refill time have a very low specificity and sensitivity at detecting low blood flow in preterm infants (2). Echocardiography is invaluable to investigate cardiac morphology but relatively imprecise as a measure of function in individuals. Cardiac magnetic resonance imaging (CMRI) is the method of choice for assessment of cardiac function in adults, but has not been used previously in the preterm population. CMRI in adults is usually done at 1.5 T, however we hypothesised that in preterm infants it would be possible to take advantage of higher field strengths.
Aims: To assess the feasibility of undertaking CMRI in preterm infants and to develop novel approaches to acquire images at 3.0 T.
Methods: All studies were carried out following written parental consent and with approval of Hammersmith Hospital Research Ethics Committee. Eight preterm babies underwent cardiac MRI. We operated the scanner in its normal mode that complies with all relevant safety guidelines. We carried out monitoring in initial examinations to verify that subjects showed no sign of stress, such as elevated temperature during or after the examinations. Gestational age was median 29 weeks (range 26 –33+5 weeks). Median birthweight was 1370g (808 – 2200g). The median corrected gestational age at time of MRI was 35 weeks (32 – 40 weeks). Two of the infants had known patent ductus confirmed on echocardiogram and were oxygen dependent. Infants were fed and allowed to fall into natural sleep. Each scan took approximately 45 minutes. Babies were monitored throughout the scan by a trained neonatologist.
Image Acquisition: All scans were carried out on a Philips 3T Intera system. T1 weighted anatomical sequence was performed followed by a series of balanced fast field echo sequences to assess function.
Data Analysis: Left ventricular volume was measured by area-length calculations. The following left ventricular parameters were then estimated: ventricular ejection fraction (ml), ventricular end-diastolic and end-systolic volumes (ml) ventricular stroke volume (ml) and cardiac output (l/min).
Intra- and inter- observer variability: Two investigators estimated each parameter ten times for a single patient studied and the coefficient of variation (cv) was calculated for each investigator. Comparing estimates for single measurement assessed inter-observer variability.
Results: Imaging was successful in 7/8 infants, and these images were sufficient to allow detailed assessment of cardiac function. Median ejection fraction was 74.6% [69.3-76]; stroke volume 2.9mls [1.5- 10.3]; cardiac output 0.4L/min [0.2-1.6]; end diastolic volume 3.8mls [1.2-11]; end systolic volume 0.3mls [0.3-3.2]. Two dimensional real time CMRI movies provided precise visualisation of cardiac function. Inter-observer variability ranged from 2-9% and intra-observer variability 1-6% for each parameter measured.
Conclusion: This preliminary study demonstrates that CMRI can provide detailed assessment of cardiac function in preterm infants, and this can be achieved at 3.0 T. Infants with PDA showed strikingly increased left ventricular output, achieved by increase in end diastolic volume with unchanged ejection fraction.
References
1.Hunt RW, Evans N, et al J Pediatr. 2004 Nov;145(5):573-5.
2.Osborn DA, Evans N, Kluckow M. Arch Dis Child Fetal Neonatal Ed. 2004 Mar;89(2):F168-73.