3307jum_online_Layout 1 6/20/14 9:00 AM Page 1171
ORIGINAL RESEARCH
Impaired Fetal Myocardial Deformation in Intrahepatic Cholestasis of Pregnancy Xuemei Fan, MD, Qichang Zhou, PhD, Shi Zeng, MD, Jiawei Zhou, MD, Qinghai Peng, PhD, Ming Zhang, PhD, Yiling Ding, PhD
Objectives—To investigate changes in fetal myocardial deformation in intrahepatic cholestasis of pregnancy. Methods—Patients with intrahepatic cholestasis of pregnancy were divided into 2 groups according to the total maternal serum bile acid concentration: mild cholestasis (10–40 μmol/L) and severe cholestasis (>40 μmol/L). Fetal echocardiography and velocity vector imaging were performed on women with cholestasis and control patients. The left ventricular global longitudinal strain and strain rate were measured. Clinical characteristics, maternal serum bile acid levels, and N-terminal pro–brain natriuretic peptide (NT-proBNP) levels in umbilical vein blood were compared between groups. The relationships among fetal myocardial deformation, maternal total bile acids, and cord NT-proBNP were analyzed.
Received July 9, 2013, from the Departments of Ultrasonography (X.F., Q.Z., S.Z., J.Z., Q.P., M.Z.) and Obstetrics and Gynecology (Y.D.), Second Xiangya Hospital, Central South University, Changsha, China. Revision requested August 5, 2013. Revised manuscript accepted for publication October 22, 2013. This study was supported by the State Natural Sciences Foundation of China (grant 30970838). Address correspondence to Qichang Zhou, PhD, Department of Ultrasonography, Second Xiangya Hospital, Central South University, 139 Middle Renmin Rd, 410011 Changsha, Hunan, China. E-mail: [email protected] Abbreviations
LV, left ventricular; NT-proBNP, N-terminal pro–brain natriuretic peptide doi:10.7863/ultra.33.7.1171
Results—Twenty women with mild cholestasis, 20 with severe cholestasis, and 40 control patients were enrolled. There were no significant differences in maternal and gestational ages between the case and control groups. Maternal bile acids and NT-proBNP were significantly higher in fetuses of mothers with cholestasis than control fetuses. The left ventricular longitudinal strain (–10.56% ± 1.83% versus –18.36% ± 1.11%; P < .01), systolic strain rate (–1.63 ± 0.18 versus –2.04 ± 0.18 seconds–1; P < .01), and diastolic strain rate (1.37 ± 0.18 versus 1.83 ± 0.14 seconds–1; P < .01) were significantly decreased in fetuses with severe cholestasis compared with control fetuses. There were positive correlations between fetal myocardial deformation and maternal total bile acids (r = 0.705, 0.643, and 0.690, respectively; P < .01) and between myocardial deformation and NT-proBNP (r = 0.672, 0.643, and 0.647; P < .01). Conclusions—Fetal myocardial deformation is impaired in severe intrahepatic cholestasis of pregnancy. Further investigation is needed to determine whether fetal echocardiography and velocity vector imaging can help predict which fetuses of mothers with cholestasis are likely to have poor outcomes. Key Words—fetal cardiac function; intrahepatic cholestasis of pregnancy; myocardial deformation; obstetric ultrasound; strain; strain rate
I
ntrahepatic cholestasis of pregnancy is a pregnancy-specific liver disorder characterized by maternal pruritus in combination with elevated serum total bile acid levels, abnormal liver function, or both.1 This condition is likely multifactorial in etiology. Genetic, hormonal, and environmental factors may contribute to the pathogenesis of intrahepatic cholestasis of pregnancy.2 The prevalence of this condition varies widely by geographic location and ethnicity. It is most common in South America (particularly in Chile and Bolivia), with incidence rates of 5% to 15%.2
©2014 by the American Institute of Ultrasound in Medicine | J Ultrasound Med 2014; 33:1171–1177 | 0278-4297 | www.aium.org
3307jum_online_Layout 1 6/20/14 9:00 AM Page 1172
Fan et al—Impaired Fetal Myocardial Deformation in Intrahepatic Cholestasis of Pregnancy
The prognosis for mothers with intrahepatic cholestasis of pregnancy is good; symptoms resolve rapidly, and serum liver function test results normalize after delivery.3 However, the condition is associated with an increased rate of adverse fetal outcomes,4 such as preterm delivery, meconium staining of the amniotic fluid, fetal bradycardia, fetal distress, and fetal loss, particularly when associated with fasting serum bile acid concentrations greater than 40 μmol/L.5 Most notably, sudden fetal death, which occurs in 0.4% to 4.1% of pregnancies with intrahepatic cholestasis,6,7 can be abrupt and cannot be reliably predicted. The pathogenesis of intrahepatic cholestasis of pregnancy–related fetal death is still poorly understood, although several studies have suggested that it is associated with a heart event rather than chronic placental insufficiency. Experimental and clinical studies have demonstrated that the fetal heart is affected by bile acids. In vitro studies on rat cardiomyocytes have shown that elevated bile acids can decrease the contraction rate, reduce the contraction amplitude, prevent cardiomyocyte synchronization,8–10 increase loss of cell integrity,10 and reduce the duration of the action potentials.11 In human fetuses of mothers with intrahepatic cholestasis of pregnancy, bradycardiatachycardia and a longer PR interval were observed.12–15 To our knowledge, no previous human studies have assessed fetal cardiac function in cases of intrahepatic cholestasis of pregnancy. Velocity vector imaging is a novel speckle-tracking– based technique for analyzing myocardial tissue motion and velocity without the limitations of Doppler echocardiography. The technique allows the quantification of myocardial tissue deformation (strain) and the rates of deformation during contraction (systolic strain rate) and relaxation (diastolic strain rate) in the fetal heart.16–20 The purpose of our study was to investigate differences in ventricular myocardial deformation between fetuses of mothers with intrahepatic cholestasis of pregnancy and healthy fetuses and to explore the relationships of any such differences with serum bile acids and N-terminal pro–brain natriuretic peptide (NT-proBNP) levels.
recruited to form the cholestasis group, and 40 age-matched healthy pregnant women were recruited to form the control group. For the purpose of the study, intrahepatic cholestasis of pregnancy was diagnosed when a patient had a combination of pruritus and a serum total bile acid concentration greater than 10 μmol/L.21 Furthermore, Glantz et al5 showed the intrahepatic cholestasis of pregnancy is associated with an increased rate of adverse fetal outcomes, particularly when associated with fasting serum bile acid concentrations greater than 40 μmol/L; therefore, we used a total bile acid concentration of 40 μmol/L to define the severity of the condition and divided the cholestasis group into mild (10–40 μmol/L; n = 20) and severe (>40 μmol/L; n = 20) groups. For patients with cholestasis and control patients, the exclusion criteria consisted of pregestational liver disease, systemic lupus erythematosus, anti-SSA antibody positivity, allergic disorders, diabetes, hypertension, and fetuses with anomalies. During the study, women with cholestasis who had a diagnosis before 37 weeks’ gestation were treated medically, as described by Lee et al.22 The mothers in the cholestasis group underwent lower-segment cesarean delivery at 37 weeks or before the onset of labor. The control group underwent elective lower-segment delivery at term. Data analysis (except cord blood NT-proBNP levels) was based on the time of diagnosis, which was determined before the initiation of pharmacotherapy for the women with cholestasis. Fetal Echocardiography All patients underwent fetal echocardiography and velocity vector imaging at the time of diagnosis and before any therapy. A commercially available system (Acuson Sequoia 512 color Doppler ultrasound system; Siemens Medical Solutions, Mountain View, CA) with a transducer frequency of 2.5 to 6.0 MHz was used. These tests were performed by a single experienced operator (X.F.), who was blinded to the group status of each patient. The pregnancy duration was estimated from the day of the last menstrual period and confirmed by sonographic measurement during the first trimester.
Materials and Methods Study Population This investigation was approved by the Ethics Committee of Second Xiangya Hospital, and informed consent was obtained from all participants. All patients were Chinese women with singleton pregnancies. Over a 31-month period (January 2010–July 2012), 40 pregnant women with intrahepatic cholestasis were
1172
Velocity Vector Acquisition and Offline Analysis An acoustic/cine zoomed loop of the apical 4-chamber view incorporating at least one complete cardiac cycle was recorded when there was no fetal movement. The 4-chamber apical view was used with the heart apex up or down and the ventricular septum as vertical as possible. While saving views, the 2-dimensional gain was optimized, and the 2dimensional scan area was minimized to obtain a high
J Ultrasound Med 2014; 33:1171–1177
3307jum_online_Layout 1 6/20/14 9:00 AM Page 1173
Fan et al—Impaired Fetal Myocardial Deformation in Intrahepatic Cholestasis of Pregnancy
frame rate (30–40 frames per second). The image was then stored in the Digital Imaging and Communications in Medicine format, transferred to an offline Siemens KinetDx workstation, and analyzed with Siemens Syngo velocity vector imaging software. The fetal cardiac cycle was determined by the anatomic M-mode. In the specialized workflow step, the definition of the R-wave location was performed. The R wave was set to match with the maximum left ventricular (LV) cavity. With a point-and-click approach, the endocardium of the LV was manually traced at the end of diastole. Tracing began from the edge of the atrioventricular valve annulus, extended to the apex, excluding the papillary muscle, and returned to the other edge of the atrioventricular valve annulus. Cardiac strain and strain rate data were automatically calculated from the velocity vector information and displayed in a 6-segment model of the fetal LV. In addition, the global strain rate and global strain were calculated from the entire velocity vector data set as the average of all segments of ventricular motion and displayed as a separate curve (Figure 1). Strain was defined as the relative deformation of the myocardium, whereas the strain rate was the rate of deformation. The units for strain and the strain rate were percent and seconds–1, respectively.
Figure 1. A, The LV strain curves were automatically shown in a 6segment model, in addition to a separate curve representing the average of all segments (arrow). The peak global strain value is displayed at the top right (red circle). B, Similarly, the strain rate curves were automatically shown in a 6-segment model, in addition to a separate curve representing the average of all segments. The arrows indicate the systolic strain rate (negative value) and diastolic strain rate (positive value). The peak global systolic strain rate value is displayed at the top right (red circle).
N-Terminal proBNP Test A 2-mL sample of umbilical venous blood from the placental end of the cord was collected immediately after delivery. All blood samples were drawn into tubes containing EDTA for further measurement. N-terminal proBNP was measured by an electrochemiluminescence immunoassay for proBNP (Roche Diagnostics Corporation, Indianapolis, IN) on the Elecsys 1010/2010 system (Roche Diagnostics).23 Interobserver and Intraobserver Variability Interobserver variability between 2 independent investigators (X.F. and S.Z.) was assessed for 20 randomly selected data sets. To determine intraobserver variability, 10 randomly selected patients were blindly analyzed 3 times by a single investigator (X.F.) at 1-week intervals. Data and Statistical Analysis The data are presented as mean ± SD or number (percent).The data distribution was normal as assessed by the Kolmogorov-Smirnov test. Unpaired 2-tailed Student t tests for continuous variables and the Fisher exact test for categorical data were performed to compare women with and without intrahepatic cholestasis of pregnancy. The relationships between fetal myocardial deformation and maternal serum total bile acids and between myocardial deformation and cord NT-proBNP were assessed by the Spearman correlation coefficient. Coefficient of variation analysis was used to test interobserver and intraobserver variability. Two-sided P < .05 was considered statistically significant. All statistical analyses were performed with SPSS software for Windows (IBM Corporation, Armonk, NY).
Results General Condition There were no significant differences in maternal and gestational ages between the case and control groups (P > .05). Serum total bile acids and cord NT-proBNP were significantly increased in the cholestasis group compared with the control group (P < .05). In the severe cholestasis group, the frequency of spontaneous preterm labor, meconium staining of the amniotic fluid, and Apgar scores less than 7 at 1 or 5 minutes were higher than in the control group (P < .05). The clinical characteristics of the women with cholestasis and the control women are compared in Table 1.
J Ultrasound Med 2014; 33:1171–1177
1173
3307jum_online_Layout 1 6/20/14 9:00 AM Page 1174
Fan et al—Impaired Fetal Myocardial Deformation in Intrahepatic Cholestasis of Pregnancy
Fetal Myocardial Deformation The results of tests for fetal LV myocardial deformation are presented in Table 1. The LV global longitudinal strain, systolic strain rate, and diastolic strain rate were decreased significantly in fetuses of mothers with severe cholestasis compared with control fetuses (P < .01). There were no significant differences in the global strain and strain rate between the mild cholestasis and control groups (P > .05). Additionally, Spearman correlation showed significant associations between myocardial deformation parameters (LV global longitudinal strain, systolic strain rate, and diastolic strain rate) and maternal serum total bile acids (r = 0.705, 0.643, and 0.690, respectively; P < .01; Figure 2). Additionally, there were significant correlations between these parameters and cord NT-proBNP (r = 0.672, 0.643, and 0.647; P < .01; Figure 3). Reproducibility The interobserver and intraobserver variability values were 13% and 10% for strain, 10% and 6% for the systolic strain rate, and 9% and 5% for the diastolic strain rate, respectively.
Discussion Among the fetuses with severe intrahepatic cholestasis of pregnancy, we found decreased strain and strain rates in the LV compared with the control population. This finding suggests that both cardiac systolic and diastolic function were impaired, especially when associated with serum total bile acids greater than 40 μmol/L. In mothers with intrahepatic cholestasis of pregnancy, increased maternal cholestasis results in elevated bile acid
levels in the fetal circulation and a reversal in the normal fetalmaternal transplacental bile acid gradient.24,25 The high level of bile acids may be the cause of fetal heart impairment in intrahepatic cholestasis of pregnancy. Several observations in animals have confirmed the effect of bile components on the heart. Deoxycholyltaurine from bile duct–ligated rats reduced the spontaneous contraction of cultured rat cardiac myocytes in a concentration- and time-dependent manner. Higher deoxycholyltaurine concentrations, such as 40 and 60 μmol/L, abolished cardiac myocyte contraction.26 Binah et al11 demonstrated that bile acids reduce the duration of the action potentials in rat ventricular myocytes. In the rabbit sinoatrial node, cholyltaurine (30–300 μmol/L) slowed the sinus rate by reducing diastolic depolarization.27 Similarly, a study in neonatal rat cardiomyocytes10,28 showed that cholyltaurine reduces cardiomyocyte contraction rates by 46% and prevents cardiomyocyte synchronization. Observations of human fetuses have indicated the presence of cardiac electrophysiologic abnormalities, such as bradycardia (≤100 beats per minute),12 atrial flutter,13 supraventricular tachycardia,14 and a longer PR interval.15 These data demonstrate that bile acids alter the “pacemaker” function of cardiac myocytes.29 Regarding myocardial contractile and relaxed function in intrahepatic cholestasis of pregnancy, only an vitro study has been reported. Gorelik et al10 found a reduced contraction amplitude due to taurocholate using scanning ion conductance microscopy. This change was reversible after the removal of taurocholate in adults but not in neonatal cardiomyocytes, which were exposed to higher concentrations (>0.3 mmol/L). Additionally, murine and human embryonic stem cell–derived cardiomyocytes with a more immature (ie, fetal-like) pheno-
Table 1. Clinical Characteristics and Velocity Vector Imaging Findings for Women With and Without Intrahepatic Cholestasis of Pregnancy Characteristic Maternal age, y Nulliparity, n (%) Gestational age at examination, wk Gestational age at delivery, wk Preterm delivery, n (%) Cesarean delivery, n (%) Meconium-stained amniotic fluid, n (%) Apgar
ORIGINAL RESEARCH
Impaired Fetal Myocardial Deformation in Intrahepatic Cholestasis of Pregnancy Xuemei Fan, MD, Qichang Zhou, PhD, Shi Zeng, MD, Jiawei Zhou, MD, Qinghai Peng, PhD, Ming Zhang, PhD, Yiling Ding, PhD
Objectives—To investigate changes in fetal myocardial deformation in intrahepatic cholestasis of pregnancy. Methods—Patients with intrahepatic cholestasis of pregnancy were divided into 2 groups according to the total maternal serum bile acid concentration: mild cholestasis (10–40 μmol/L) and severe cholestasis (>40 μmol/L). Fetal echocardiography and velocity vector imaging were performed on women with cholestasis and control patients. The left ventricular global longitudinal strain and strain rate were measured. Clinical characteristics, maternal serum bile acid levels, and N-terminal pro–brain natriuretic peptide (NT-proBNP) levels in umbilical vein blood were compared between groups. The relationships among fetal myocardial deformation, maternal total bile acids, and cord NT-proBNP were analyzed.
Received July 9, 2013, from the Departments of Ultrasonography (X.F., Q.Z., S.Z., J.Z., Q.P., M.Z.) and Obstetrics and Gynecology (Y.D.), Second Xiangya Hospital, Central South University, Changsha, China. Revision requested August 5, 2013. Revised manuscript accepted for publication October 22, 2013. This study was supported by the State Natural Sciences Foundation of China (grant 30970838). Address correspondence to Qichang Zhou, PhD, Department of Ultrasonography, Second Xiangya Hospital, Central South University, 139 Middle Renmin Rd, 410011 Changsha, Hunan, China. E-mail: [email protected] Abbreviations
LV, left ventricular; NT-proBNP, N-terminal pro–brain natriuretic peptide doi:10.7863/ultra.33.7.1171
Results—Twenty women with mild cholestasis, 20 with severe cholestasis, and 40 control patients were enrolled. There were no significant differences in maternal and gestational ages between the case and control groups. Maternal bile acids and NT-proBNP were significantly higher in fetuses of mothers with cholestasis than control fetuses. The left ventricular longitudinal strain (–10.56% ± 1.83% versus –18.36% ± 1.11%; P < .01), systolic strain rate (–1.63 ± 0.18 versus –2.04 ± 0.18 seconds–1; P < .01), and diastolic strain rate (1.37 ± 0.18 versus 1.83 ± 0.14 seconds–1; P < .01) were significantly decreased in fetuses with severe cholestasis compared with control fetuses. There were positive correlations between fetal myocardial deformation and maternal total bile acids (r = 0.705, 0.643, and 0.690, respectively; P < .01) and between myocardial deformation and NT-proBNP (r = 0.672, 0.643, and 0.647; P < .01). Conclusions—Fetal myocardial deformation is impaired in severe intrahepatic cholestasis of pregnancy. Further investigation is needed to determine whether fetal echocardiography and velocity vector imaging can help predict which fetuses of mothers with cholestasis are likely to have poor outcomes. Key Words—fetal cardiac function; intrahepatic cholestasis of pregnancy; myocardial deformation; obstetric ultrasound; strain; strain rate
I
ntrahepatic cholestasis of pregnancy is a pregnancy-specific liver disorder characterized by maternal pruritus in combination with elevated serum total bile acid levels, abnormal liver function, or both.1 This condition is likely multifactorial in etiology. Genetic, hormonal, and environmental factors may contribute to the pathogenesis of intrahepatic cholestasis of pregnancy.2 The prevalence of this condition varies widely by geographic location and ethnicity. It is most common in South America (particularly in Chile and Bolivia), with incidence rates of 5% to 15%.2
©2014 by the American Institute of Ultrasound in Medicine | J Ultrasound Med 2014; 33:1171–1177 | 0278-4297 | www.aium.org
3307jum_online_Layout 1 6/20/14 9:00 AM Page 1172
Fan et al—Impaired Fetal Myocardial Deformation in Intrahepatic Cholestasis of Pregnancy
The prognosis for mothers with intrahepatic cholestasis of pregnancy is good; symptoms resolve rapidly, and serum liver function test results normalize after delivery.3 However, the condition is associated with an increased rate of adverse fetal outcomes,4 such as preterm delivery, meconium staining of the amniotic fluid, fetal bradycardia, fetal distress, and fetal loss, particularly when associated with fasting serum bile acid concentrations greater than 40 μmol/L.5 Most notably, sudden fetal death, which occurs in 0.4% to 4.1% of pregnancies with intrahepatic cholestasis,6,7 can be abrupt and cannot be reliably predicted. The pathogenesis of intrahepatic cholestasis of pregnancy–related fetal death is still poorly understood, although several studies have suggested that it is associated with a heart event rather than chronic placental insufficiency. Experimental and clinical studies have demonstrated that the fetal heart is affected by bile acids. In vitro studies on rat cardiomyocytes have shown that elevated bile acids can decrease the contraction rate, reduce the contraction amplitude, prevent cardiomyocyte synchronization,8–10 increase loss of cell integrity,10 and reduce the duration of the action potentials.11 In human fetuses of mothers with intrahepatic cholestasis of pregnancy, bradycardiatachycardia and a longer PR interval were observed.12–15 To our knowledge, no previous human studies have assessed fetal cardiac function in cases of intrahepatic cholestasis of pregnancy. Velocity vector imaging is a novel speckle-tracking– based technique for analyzing myocardial tissue motion and velocity without the limitations of Doppler echocardiography. The technique allows the quantification of myocardial tissue deformation (strain) and the rates of deformation during contraction (systolic strain rate) and relaxation (diastolic strain rate) in the fetal heart.16–20 The purpose of our study was to investigate differences in ventricular myocardial deformation between fetuses of mothers with intrahepatic cholestasis of pregnancy and healthy fetuses and to explore the relationships of any such differences with serum bile acids and N-terminal pro–brain natriuretic peptide (NT-proBNP) levels.
recruited to form the cholestasis group, and 40 age-matched healthy pregnant women were recruited to form the control group. For the purpose of the study, intrahepatic cholestasis of pregnancy was diagnosed when a patient had a combination of pruritus and a serum total bile acid concentration greater than 10 μmol/L.21 Furthermore, Glantz et al5 showed the intrahepatic cholestasis of pregnancy is associated with an increased rate of adverse fetal outcomes, particularly when associated with fasting serum bile acid concentrations greater than 40 μmol/L; therefore, we used a total bile acid concentration of 40 μmol/L to define the severity of the condition and divided the cholestasis group into mild (10–40 μmol/L; n = 20) and severe (>40 μmol/L; n = 20) groups. For patients with cholestasis and control patients, the exclusion criteria consisted of pregestational liver disease, systemic lupus erythematosus, anti-SSA antibody positivity, allergic disorders, diabetes, hypertension, and fetuses with anomalies. During the study, women with cholestasis who had a diagnosis before 37 weeks’ gestation were treated medically, as described by Lee et al.22 The mothers in the cholestasis group underwent lower-segment cesarean delivery at 37 weeks or before the onset of labor. The control group underwent elective lower-segment delivery at term. Data analysis (except cord blood NT-proBNP levels) was based on the time of diagnosis, which was determined before the initiation of pharmacotherapy for the women with cholestasis. Fetal Echocardiography All patients underwent fetal echocardiography and velocity vector imaging at the time of diagnosis and before any therapy. A commercially available system (Acuson Sequoia 512 color Doppler ultrasound system; Siemens Medical Solutions, Mountain View, CA) with a transducer frequency of 2.5 to 6.0 MHz was used. These tests were performed by a single experienced operator (X.F.), who was blinded to the group status of each patient. The pregnancy duration was estimated from the day of the last menstrual period and confirmed by sonographic measurement during the first trimester.
Materials and Methods Study Population This investigation was approved by the Ethics Committee of Second Xiangya Hospital, and informed consent was obtained from all participants. All patients were Chinese women with singleton pregnancies. Over a 31-month period (January 2010–July 2012), 40 pregnant women with intrahepatic cholestasis were
1172
Velocity Vector Acquisition and Offline Analysis An acoustic/cine zoomed loop of the apical 4-chamber view incorporating at least one complete cardiac cycle was recorded when there was no fetal movement. The 4-chamber apical view was used with the heart apex up or down and the ventricular septum as vertical as possible. While saving views, the 2-dimensional gain was optimized, and the 2dimensional scan area was minimized to obtain a high
J Ultrasound Med 2014; 33:1171–1177
3307jum_online_Layout 1 6/20/14 9:00 AM Page 1173
Fan et al—Impaired Fetal Myocardial Deformation in Intrahepatic Cholestasis of Pregnancy
frame rate (30–40 frames per second). The image was then stored in the Digital Imaging and Communications in Medicine format, transferred to an offline Siemens KinetDx workstation, and analyzed with Siemens Syngo velocity vector imaging software. The fetal cardiac cycle was determined by the anatomic M-mode. In the specialized workflow step, the definition of the R-wave location was performed. The R wave was set to match with the maximum left ventricular (LV) cavity. With a point-and-click approach, the endocardium of the LV was manually traced at the end of diastole. Tracing began from the edge of the atrioventricular valve annulus, extended to the apex, excluding the papillary muscle, and returned to the other edge of the atrioventricular valve annulus. Cardiac strain and strain rate data were automatically calculated from the velocity vector information and displayed in a 6-segment model of the fetal LV. In addition, the global strain rate and global strain were calculated from the entire velocity vector data set as the average of all segments of ventricular motion and displayed as a separate curve (Figure 1). Strain was defined as the relative deformation of the myocardium, whereas the strain rate was the rate of deformation. The units for strain and the strain rate were percent and seconds–1, respectively.
Figure 1. A, The LV strain curves were automatically shown in a 6segment model, in addition to a separate curve representing the average of all segments (arrow). The peak global strain value is displayed at the top right (red circle). B, Similarly, the strain rate curves were automatically shown in a 6-segment model, in addition to a separate curve representing the average of all segments. The arrows indicate the systolic strain rate (negative value) and diastolic strain rate (positive value). The peak global systolic strain rate value is displayed at the top right (red circle).
N-Terminal proBNP Test A 2-mL sample of umbilical venous blood from the placental end of the cord was collected immediately after delivery. All blood samples were drawn into tubes containing EDTA for further measurement. N-terminal proBNP was measured by an electrochemiluminescence immunoassay for proBNP (Roche Diagnostics Corporation, Indianapolis, IN) on the Elecsys 1010/2010 system (Roche Diagnostics).23 Interobserver and Intraobserver Variability Interobserver variability between 2 independent investigators (X.F. and S.Z.) was assessed for 20 randomly selected data sets. To determine intraobserver variability, 10 randomly selected patients were blindly analyzed 3 times by a single investigator (X.F.) at 1-week intervals. Data and Statistical Analysis The data are presented as mean ± SD or number (percent).The data distribution was normal as assessed by the Kolmogorov-Smirnov test. Unpaired 2-tailed Student t tests for continuous variables and the Fisher exact test for categorical data were performed to compare women with and without intrahepatic cholestasis of pregnancy. The relationships between fetal myocardial deformation and maternal serum total bile acids and between myocardial deformation and cord NT-proBNP were assessed by the Spearman correlation coefficient. Coefficient of variation analysis was used to test interobserver and intraobserver variability. Two-sided P < .05 was considered statistically significant. All statistical analyses were performed with SPSS software for Windows (IBM Corporation, Armonk, NY).
Results General Condition There were no significant differences in maternal and gestational ages between the case and control groups (P > .05). Serum total bile acids and cord NT-proBNP were significantly increased in the cholestasis group compared with the control group (P < .05). In the severe cholestasis group, the frequency of spontaneous preterm labor, meconium staining of the amniotic fluid, and Apgar scores less than 7 at 1 or 5 minutes were higher than in the control group (P < .05). The clinical characteristics of the women with cholestasis and the control women are compared in Table 1.
J Ultrasound Med 2014; 33:1171–1177
1173
3307jum_online_Layout 1 6/20/14 9:00 AM Page 1174
Fan et al—Impaired Fetal Myocardial Deformation in Intrahepatic Cholestasis of Pregnancy
Fetal Myocardial Deformation The results of tests for fetal LV myocardial deformation are presented in Table 1. The LV global longitudinal strain, systolic strain rate, and diastolic strain rate were decreased significantly in fetuses of mothers with severe cholestasis compared with control fetuses (P < .01). There were no significant differences in the global strain and strain rate between the mild cholestasis and control groups (P > .05). Additionally, Spearman correlation showed significant associations between myocardial deformation parameters (LV global longitudinal strain, systolic strain rate, and diastolic strain rate) and maternal serum total bile acids (r = 0.705, 0.643, and 0.690, respectively; P < .01; Figure 2). Additionally, there were significant correlations between these parameters and cord NT-proBNP (r = 0.672, 0.643, and 0.647; P < .01; Figure 3). Reproducibility The interobserver and intraobserver variability values were 13% and 10% for strain, 10% and 6% for the systolic strain rate, and 9% and 5% for the diastolic strain rate, respectively.
Discussion Among the fetuses with severe intrahepatic cholestasis of pregnancy, we found decreased strain and strain rates in the LV compared with the control population. This finding suggests that both cardiac systolic and diastolic function were impaired, especially when associated with serum total bile acids greater than 40 μmol/L. In mothers with intrahepatic cholestasis of pregnancy, increased maternal cholestasis results in elevated bile acid
levels in the fetal circulation and a reversal in the normal fetalmaternal transplacental bile acid gradient.24,25 The high level of bile acids may be the cause of fetal heart impairment in intrahepatic cholestasis of pregnancy. Several observations in animals have confirmed the effect of bile components on the heart. Deoxycholyltaurine from bile duct–ligated rats reduced the spontaneous contraction of cultured rat cardiac myocytes in a concentration- and time-dependent manner. Higher deoxycholyltaurine concentrations, such as 40 and 60 μmol/L, abolished cardiac myocyte contraction.26 Binah et al11 demonstrated that bile acids reduce the duration of the action potentials in rat ventricular myocytes. In the rabbit sinoatrial node, cholyltaurine (30–300 μmol/L) slowed the sinus rate by reducing diastolic depolarization.27 Similarly, a study in neonatal rat cardiomyocytes10,28 showed that cholyltaurine reduces cardiomyocyte contraction rates by 46% and prevents cardiomyocyte synchronization. Observations of human fetuses have indicated the presence of cardiac electrophysiologic abnormalities, such as bradycardia (≤100 beats per minute),12 atrial flutter,13 supraventricular tachycardia,14 and a longer PR interval.15 These data demonstrate that bile acids alter the “pacemaker” function of cardiac myocytes.29 Regarding myocardial contractile and relaxed function in intrahepatic cholestasis of pregnancy, only an vitro study has been reported. Gorelik et al10 found a reduced contraction amplitude due to taurocholate using scanning ion conductance microscopy. This change was reversible after the removal of taurocholate in adults but not in neonatal cardiomyocytes, which were exposed to higher concentrations (>0.3 mmol/L). Additionally, murine and human embryonic stem cell–derived cardiomyocytes with a more immature (ie, fetal-like) pheno-
Table 1. Clinical Characteristics and Velocity Vector Imaging Findings for Women With and Without Intrahepatic Cholestasis of Pregnancy Characteristic Maternal age, y Nulliparity, n (%) Gestational age at examination, wk Gestational age at delivery, wk Preterm delivery, n (%) Cesarean delivery, n (%) Meconium-stained amniotic fluid, n (%) Apgar
