Facebook Twitter YouTube


Left and Right Ventricular Longitudinal Cardiac Strain In Fetuses of Diabetic Mothers

Friday, October 19, 2012
Room 275-277 (Morial Convention Center)
Anjali Chelliah, MD, Jodi I. Pike, MD, Lowell H. Frank, MD, Anita Krishnan, MD, Caroline Stallings, MSW, Craig A. Sable, MD and Mary T. Donofrio, MD, Pediatric Cardiology, Children's National Medical Center, Washington, DC

Purpose: Prior studies suggest that fetuses of women with pregestational diabetes mellitus demonstrate evidence of impaired myocardial performance and diastolic cardiac function. However, cardiac strain has not previously been quantified in this population. The aim of this study was to determine whether left and right ventricular (LV and RV) cardiac strain measurements in fetuses of women with pregestational diabetes differ from those of normal controls.

Methods: Syngo Velocity Vector Imaging software was used by two blinded observers to measure total longitudinal ventricular strain on singleton fetuses with maternal diabetes and no evidence of cardiac disease as well as gestational age-matched normal controls referred for a family history of congenital heart disease. Two-tailed paired t-tests were used to compare global LV and RV longitudinal strain among the two populations. LV and RV strain were also compared between the subpopulation of fetuses of insulin-dependent diabetic mothers and normal controls. Intra- and inter-observer reliability were calculated using intraclass correlation coefficients (ICC). 

Results: Data from 80 fetuses (40 fetuses of diabetic mothers and 40 age-matched controls) with a median gestational age of 22 weeks (range 20-37 weeks) was analyzed. The total longitudinal LV and RV strain in fetuses of diabetic mothers, respectively, was -11.4 +/- 5.6% and -10.2 +/- 5.9%. Total longitudinal LV and RV strain in normal controls was -11.5 +/- 5% and -11.8 +/- 5.4%. There was no significant difference in global LV (p=.91) or RV (p=0.12) strain between the two groups. LV and RV strain in the subpopulation of fetuses with insulin-dependent diabetic mothers also did not differ from that of normal controls. Intra-observer agreement was strong for both LV (ICC= 0.79, 95% confidence interval 0.69-0.86) and RV (ICC=0.8, 95% CI 0.71-0.87) strain measurements. Inter-observer agreement measured on a 20% sample of patients was also strong for both LV (ICC = 0.91, 95% CI 0.76-0.97) and RV (ICC= 0.82, 95% CI 0.5-0.94). 

Conclusion: Fetuses of diabetic mothers have overall similar total longitudinal strain to normal fetuses. However, further analysis of higher-risk subgroups, such as fetuses that are large for gestational age or have mothers with poorly controlled diabetes, should be pursued to better characterize this population.