We describe examination of fetal cardiac function with contemporary functional fetal echocardiography and fetoplacental Doppler ultrasound using the VisualSonics VEVO 2100 microultrasound in a surgically induced model of intrauterine fetal growth restriction in a rabbit.
Fetal intrauterine growth restriction (IUGR) results in abnormal cardiac function that is apparent antenatally due to advances in fetoplacental Doppler ultrasound and fetal echocardiography. Increasingly, these imaging modalities are being employed clinically to examine cardiac function and assess wellbeing in utero, thereby guiding timing of birth decisions. Here, we used a rabbit model of IUGR that allows analysis of cardiac function in a clinically relevant way. Using isoflurane induced anesthesia, IUGR is surgically created at gestational age day 25 by performing a laparotomy, exposing the bicornuate uterus and then ligating 40-50% of uteroplacental vessels supplying each gestational sac in a single uterine horn. The other horn in the rabbit bicornuate uterus serves as internal control fetuses. Then, after recovery at gestational age day 30 (full term), the same rabbit undergoes examination of fetal cardiac function. Anesthesia is induced with ketamine and xylazine intramuscularly, then maintained by a continuous intravenous infusion of ketamine and xylazine to minimize iatrogenic effects on fetal cardiac function. A repeat laparotomy is performed to expose each gestational sac and a microultrasound examination (VisualSonics VEVO 2100) of fetal cardiac function is performed. Placental insufficiency is evident by a raised pulsatility index or an absent or reversed end diastolic flow of the umbilical artery Doppler waveform. The ductus venosus and middle cerebral artery Doppler is then examined. Fetal echocardiography is performed by recording B mode, M mode and flow velocity waveforms in lateral and apical views. Offline calculations determine standard M-mode cardiac variables, tricuspid and mitral annular plane systolic excursion, speckle tracking and strain analysis, modified myocardial performance index and vascular flow velocity waveforms of interest. This small animal model of IUGR therefore affords examination of in utero cardiac function that is consistent with current clinical practice and is therefore useful in a translational research setting.
The burden of cardiovascular disease that results from fetal intrauterine growth restriction (IUGR) cannot be overstated. It is the leading cause of stillbirth after congenital abnormalities.1 IUGR refers to a fetus that fails to reach its growth potential and is commonly a result of placental insufficiency.2 In survivors, cardiovascular ill health is evident across the life span with myocardial dysfunction apparent in infancy and childhood3,4 and subsequent hypertension5, diabetes6, and obesity developing in adult life – all cumulative cardiac risk factors from birth towards premature death from ischemic heart disease.7
Developing animal models to characterize the maternal-fetal communication that establishes IUGR and the subsequent fetal response to reduced substrate availability is necessary if we are to both better understand the pathophysiology of IUGR-related cardiac impairment and to develop cardio-protective strategies to improve fetal and postnatal health. In this regard, many different models across different species have been described.8 IUGR is commonly induced by maternal under nutrition with a low protein diet in rodents,9 surgical ablation or ligation of uterine blood flow in rodents and guinea pigs10 or occlusion of the umbilical artery in sheep.11 However, it is apparent that no model fully recapitulates the human IUGR.12
In this current methodological study, we used a well validated approach of selective uteroplacental vascular interruption in a rabbit13-16 that not only produces fetal cardiovascular responses observed with ultrasound clinically14, but also allows interrogation of fetal cardiac function with novel echocardiography using microultrasound technology of the VisualSonics VEVO 2100. While Doppler ultrasound of fetoplacental vessels remains the cornerstone of current antenatal surveillance of IUGR fetuses17, functional echocardiography is increasingly being utilized to provide new insights into disease pathophysiology and to assess fetal wellbeing.18 Accordingly, here we take these advances from clinical research and describe an animal model that harbors not only this imaging sophistication but also provides the experimental platform to investigate mechanistic pathways and novel therapeutics.
The following experimental protocol is approved by the Animal Ethics Committee, Katholieke Universiteit Leuven, Leuven, Belgium. We followed previously described surgical procedure13 including some methodological changes, especially in anesthesia procedure.
1. Inducing Intrauterine Growth Restriction (IUGR)
2. Performing Fetal Echocardiography and Pulsed-wave Doppler Ultrasound
Flow velocity waveforms (FVW) of the umbilical artery are obtained by locating the umbilical vessels using color Doppler and then placing the pulsed Doppler sample gate over the umbilical artery on a free loop of umbilical cord.25 The middle cerebral artery (MCA) FVW is located by placing the pulsed Doppler sample gate just beyond the origin of MCA once the Circle of Willis is located by color Doppler.26 The ductus venosus (DV) FVW is obtained by placing the pulsed Doppler sample gate at the proximal portion of the DV seen with color Doppler where it originates from the intrahepatic umbilical vein either in a sagittal or transverse view of the fetus.26 The pulsatility index (PI) is calculated offline using the VisualSonics cardiovascular analysis software.
An asymmetrical growth restricted fetus and placenta from uteroplacental vascular ligation is compared to a normal control fetus and placenta in Figure 1F. Asymmetrical growth is confirmed by reduced neonatal birth weight and increased head circumference:abdominal circumference ratio to controls. Representative results from fetoplacental Doppler studies are shown in Figure 2. A normal low resistance pattern of positive end-diastolic flow in a control fetus is shown. With progressive increases in placental resistance seen in IUGR fetuses, absence and then reversed end diastolic flow is apparent. Figure 3 demonstrates a normal high resistance middle cerebral artery Doppler signal in a healthy fetus and a positive a wave ductus venosus in the same fetus. In IUGR fetuses, an increased pulsatility index of the ductus venosus and reversal of the a wave can be seen.14 Representative results from M-mode fetal echocardiography are then shown in Figure 4. This lateral view allows calculation of internal ventricular diameters and volumes. The apical view allows calculation of TAPSE and MAPSE. Figure 5 demonstrates speckle tracking of velocity vectors and the resultant regional strain curves to calculate strain, strain rate, displacement and velocity.
Figure 1. Surgical technique of creating IUGR in a rabbit model. A: Midline laparotomy exposing rectus sheath, arrow = mammory glands; B: Safe entry into the abdominal cavity; C: arrow = uteroplacental vessels supplying gestational sac; D: suture method; E: arrow = completed ligation of uteroplacental vessel; F: representative control and IUGR fetus and placenta.
Figure 2. Doppler ultrasound of the umbilical artery. A: positive end-diastolic flow (EDF); B: absent end-diastolic flow (AEDF); C: Reversed end-diastolic flow (REDF).
Figure 3. A: Doppler ultrasound of the ductus venosus, s = s wave (ventricular systolic contraction), d = d wave (early ventricular diastole), a = a wave (atrial contraction); B: Doppler ultrasound of the middle cerebral artery.
Figure 4. M mode echocardiography. A: lateral four-chamber view, LVID = left ventricular internal diameter, IVSD = intraventricular septal diameter, RVID = right ventricular septal diameter, ESD = end systolic diameter, EDD = end diastolic diameter; B: apical view demonstrating tricuspid annular plane systolic excursion (TAPSE); C: apical view demonstrating mitral annular plane systolic excursion (MAPSE).
Figure 5. Fetal cardiac strain analysis. A: Region of interest defined by endo- and epi-cardium of the left ventricle; B: Strain rate curves of six myocardial segments, SR = peak systolic strain rate; C: Strain curves of six myocardial segments, Str = peak systolic strain. Cardiac motion depicted by M mode demonstrating end diastole (ED) and end systole (ES). Click here to view larger figure.
We have used a previously validated approach of surgically reducing uteroplacental blood flow in a rabbit to produce IUGR13-16 and later examining fetal cardiac function14 to describe microultrasound technology and cardiac function analysis available on the VisualSonics VEVO 2100. The ability to reproduce fetoplacental Doppler changes of human IUGR fetuses in a small animal model and to then allow examination of cardiac function with recently described echocardiography is likely to progress fetal cardiac research in a clinically relevant way.
Small animal models commonly rely on maternal caloric restriction or low protein consumption9, however these are limited by an inability to demonstrate reduced placental blood flow, the primary mechanism of IUGR in developed countries.27 Further, surgical bilateral uterine artery ligation in rats, while commonly reported, does not reproducibly result in growth restriction.28 In this current methodology in rabbit fetuses, we show placental insufficiency to be evident by an absent or reversed end-diastolic flow of the umbilical artery Doppler (UA AREDF), consistent with sonographic findings in human IUGR. It has been shown experimentally that increased resistance in this Doppler signal reflects down stream impedance to blood flow in the placental vascular bed and is indicative of placental insufficiency.29 The presence of UA AREDF is apparent when 50 – 70% of the villous vasculature is dysfunctional.30,31 Clinically, UA AREDF is a powerful predictor of hypoxia and poor perinatal outcome, and there is level 1-evidence to support its use in the management of high-risk pregnancy.17
In this model, if UA AREDF is observed the researcher can be confident that the primary surgery was successful in producing severe placental insufficiency and that further echocardiographic assessment is likely to be fruitful. Fetal echocardiography has recently branched out from a predominantly diagnostic domain of congenital abnormalities to now providing detailed functional assessments of cardiac function.18 Both Doppler and M-mode can be used to assess fetal ventricular function and derive measures of cardiac output.32 More recently, novel indices of cardiac performance in the fetus have been described such as speckle tracking and strain measurement33, the myocardial performance index24,34, tissue Doppler35 and three-dimensional (3D) techniques.32 An important feature of this current study is that these recent advances can also be performed on this small-scale rabbit model using the Visualsonics VEVO 2100 microultrasound and cardiac function analysis software. Furthermore, as previously described14 this model also allows assessment of the fetal hemodynamic response in other vascular territories, in particular the middle cerebral artery, ductus venosus and aortic isthmus, which are used widely in clinical practice when examining the growth-restricted fetus.36 Similarly, the effects of administering glucocorticoids in preparation for preterm birth may also be examined.37,38 This rabbit model offers further advantages in terms of internal controls in the opposite horn of the uterus, a similar villous and hemochorial placentation to human pregnancy13, low cost, availability and relatively easy handling.
There are, nevertheless, several limitations of this model that must be addressed. The major limitation is iatrogenic fetal bradycardia during echocardiography. Maternally administered inhalational isoflurane can result in fetal bradycardia39 and should be kept to a minimum or as in our case not used at all during echocardiography. Instead, we substituted this for an intravenous infusion of ketamine and xylazine, which has recently been shown in rabbits not to alter mean arterial pressure40, thereby presumably maintaining placental (and thus fetal) perfusion. Despite this approach, exposure of the gestational sac to the external environment, handling and pressure from the ultrasound transducer can all cause fetal bradycardia temporarily. We describe in the method ways to minimize this effect, however for accurate results we believe echocardiographic and fetoplacental Doppler ultrasounds are best limited to around 5 min of total exposure. With increasingly complex echocardiographic techniques and learning curves now described in the clinical literature41, meticulous planning is required beforehand to ensure standardized results. Several of the echocardiographic measurements described in this study, for example valvular annular plane systolic excursion and speckle tracking with strain analysis, are not established in current obstetric clinical practice, despite their use in adult cardiology.42,43 Nevertheless, given the recent research interest in these novel measurements in fetal medicine, we chose to include them in our methodology to inform researchers that they are feasible to obtain when using this rabbit model. The analysis of these specific parameters in IUGR fetuses is beyond the scope of this methodological study. Another limitation relates to the natural tendency in the rabbit for IUGR based on fetal position44, therefore the well-perfused ovarian and vaginal ends of each horn are recommended for fetal case allocation. Furthermore, in this methodology, a gestational age of 25 days is used to surgically induce IUGR. This is based on previously described successful experiments, however the expected mortality rate is 50%.13,14 Lastly, the fetal size precludes chronic instrumentation that is useful in sheep models for the placement of hemodynamic flow probes for later non-euthanized measurements and repeated blood sampling.11 In conclusion, selective ligature of the uteroplacental vessels to produce IUGR in the rabbit with subsequent microultrasound examination of cardiac function represents an animal model that is consistent with contemporary clinical practice and therefore useful to bidirectional translational researchers.
The authors have nothing to disclose.
This work is supported by a Hamilton-Fairley NHMRC Fellowship (RH, AL); the Victorian Government’s Operational Infrastructure Support Program (RH, EW) and the Marie Curie Industria-Academia Partnership and Pathways grant sponsored by the European Commission (ME, PD). The authors would like to thank Dr. Andre Miyague, Dr. Francesca Russo, Ms. Rosita Kinnart and Mr. Ivan Laermans for their technical expertise in producing this video.
Name of Reagent/Material | Company | Catalog Number | Comments |
Ketamine | Ceva Sante Animale | http://www.ceva.com/en | |
Xylazine | Ceva Sante Animale | http://www.ceva.com/en | |
Depot Provera | Pharmacia Upjohn | ||
Penicillin G | Kela Pharma | http://www.kela.be | |
Lidocaine | B Braun Medical | http://www.bbraun.com/ | |
Temgesic | Schering-Plough | http://www.merck-animal-health-usa.com/ | |
Isolurane | Isoba Vet; Abbott Laboratories Ltd | http://www.abbottanimalhealth.com/index.html | |
Ethicon | Johnson and Johnson | http://www.ethiconproducts.co.uk/products/sutures | |
Ethicon | Johnson and Johnson | http://www.ethiconproducts.co.uk/products/sutures | |
Ethicon | Johnson and Johnson | http://www.ethiconproducts.co.uk/products/sutures | |
VEVO 2100 | VisualSonics | SN100-0032 | http://www.visualsonics.com/ |
Aquasonic Gel | Parker Laboratories | 01 02 | http://www.parkerlabs.com/ultrasound_products.html |
Nellcor N-20PA Pulse oximeter | Covidien | http://www.nellcor.com/prod/PRODUCT.ASPX?S1=POX&S2=MON&id=282&V |