|
|
||||||||
|
by the American Institute of Ultrasound in Medicine J Ultrasound Med 21:739-746 0278-4297 Fetal Flow Redistribution to the Brain in Response to Malaria InfectionDoes Protection of the Fetus Against Malaria Develop Over Time?Department of Nuclear Medicine and Ultrasound, Institut National de la Santé et de la Recherche Médicale, Centre Hospitalier Universitaire Trousseau, Tours, France (P.A., N.T., S.H., M.G., F.P.); and Section of Obstetrics and Gynecology, Hôpital André Bouron, Saint-Laurent du Maroni, French Guiana (G.C., F.B.). Address correspondence and reprint requests to Philippe Arbeille, MD, PhD, Département Médecine Nucléaire & Ultrasons, Institut National de la Santé et de la Recherche Médicale 316, Centre Hospitalier Universitaire Trousseau, 37044 Tours, France.
Objective. Malaria during pregnancy induces deterioration of placental function, resulting in transient fetal hypoxia. Our objective was to evaluate the sensitivity and specificity of fetal Doppler indices for prediction of abnormal fetal heart rate at delivery and to compare the amplitude of the fetal vascular response to malaria in 2 groups of fetuses (1994 and 1996) separated by an interval of 2 years. Methods. Every day during the crisis, the umbilical and cerebral resistance indices, the cerebral-umbilical resistance ratio, and the hypoxic index (mean percent cerebral-umbilical resistance ratio change x crisis duration) were calculated. Results. In group 2 (1996), the duration of the flow redistribution period was about 7 days (mean cerebral-umbilical resistance ratio change ± SD, 7% ± 4%; hypoxic index, 49 ± 26; premature, 35%; and abnormal fetal heart rate, 17.5%). A hypoxic index greater than 150 predicted abnormal fetal heart rate with high sensitivity and specificity (group 1, 80% and 85%; and group 2, 100% and 91%). Moreover, in group 2 (1996), the amplitude of the fetal vascular response and the rate of long-duration crisis were significantly lower than in group 1 (1994; P < .01). Nevertheless, the hypoxic index was much more predictive of fetal heart rate at delivery than the amplitude or duration (i.e., crisis duration) of the flow redistribution. Conclusions. The hypoxic index value during the crisis allowed prediction of abnormal fetal heart rate at delivery. In group 2, the absence of a long-term flow redistribution period and the smaller hemodynamic changes (lower hypoxic index) associated with a lower occurrence of abnormal fetal heart rate could be related to improvement of pregnancy management, acquired protection during the interval between the 2 studies, or both.
Key Words: Doppler sonography fetus hypoxia malaria Abbreviations: aFHR, abnormal fetal heart rate CRI, cerebral resistance index C/U, cerebral/umbilical resistance FHR, fetal heart rate HI, hypoxic index URI, umbilical resistance index
Malarial placental lesions result in a decrease in fetoplacental exchange. In early pregnancy, severe acute hypoxia may be observed, which induces abortion or death of the fetus. Toward the end of gestation, hypoxia may also be observed, either acute (several days) or chronic, and may be associated with premature birth and growth restriction, with a risk of cerebral lesions of hypoxic origin.16 Doppler examination of the fetal blood vessels is widely used to confirm the existence of hemodynamic abnormalities at the placental level accompanying intrauterine growth restriction and to evaluate the cerebrovascular response to hypoxia. Many studies have shown the usefulness of the umbilical resistance index (URI) and cerebral resistance index (CRI) in detection of intrauterine growth restriction and hypoxia.714 An initial published study of 23 pregnancies complicated by malaria was performed in 1994.15 The results showed good sensitivity and specificity of the Doppler hypoxic index (HI), which integrates both the duration and the intensity of the fetal flow redistribution in response to hypoxia, for predicting abnormal fetal heart rate (aFHR) at delivery. Our first objective was to confirm the sensitivity and specificity of the Doppler HI on a second cohort of fetuses investigated 2 years later in the same center but for whom the clinical consequences of malaria were considerably lower than for the first study. The second objective was to compare the amplitude of the fetal vascular response to malaria in the 2 groups of 23 fetuses investigated, separated by an interval of 2 years, given that the recruitment of the pregnancies was within the same type of population and that the management of the pregnancies was similar and performed by the same medical team.
Population The study, designed as prospective and observational, was carried out in the obstetric department of a government hospital in French Guiana. The population consisted of 23 pregnancies investigated in 1996 (group 2), all complicated by a malaria infection (Plasmodium falciparum). The results were compared with those obtained in similar population of 23 pregnancies investigated 2 years earlier.
Protocol After leaving the hospital, the gravida was seen as regularly as possible and at each visit underwent a complete sonographic and Doppler examination. After delivery, the perinatal data were collected.
Fetal Hemodynamic Data
At each daily Doppler examination, we measured the umbilical and cerebral resistances, and then the C/U ratio and HI were calculated. Uterine vascular resistance was only measured at admission of the gravida. Doppler examinations were performed by the 2 obstetricians in charge of the pregnancies using an EUB 44 system (Hitachi, Ltd, Tokyo, Japan).
Perinatal Data
Data Processing
The 23 pregnancies (group 2, 1996) were investigated according to the same protocol as described above and in the same center as for the previous study (1994).15 The duration of the study was 1 year. Forty pregnancies complicated by malaria were investigated, but the complete Doppler and clinical data needed for the study were collected in only 23 of them. The other patients assessed by Doppler imaging during their stay at the hospital returned to their villages, and their deliveries could not be monitored.
Perinatal data are presented in Table 1
The fetal umbilical and cerebral Doppler traces were successfully recorded at each daily Doppler session. Doppler data at admission and percent changes of these parameters during the crisis are presented in Table 2
Sensitivity, specificity, and positive and negative predictive values of the lowest C/U absolute value (abnormal when <1.1), crisis duration (>7 days), and HI (>150) for predicting aFHR at delivery are presented in Table 3
Figure 1
The perinatal and Doppler data of group 2 (1996) did not differ from those of the short crisis in group 1 (1994), except for the rate of oligoamnios and prematurity. Conversely, there was a significant difference between groups 1 and 2 for most of the perinatal and Doppler parameters except for parity, age of delivery, oligoamnios, parasitemia grade, Apgar score, intrauterine growth restriction, uterine and cerebral resistance at admission, and percent C/U decrease during the crisis. In both groups the HI (>150) predicted abnormal FHR at delivery with good sensitivity and specificity. Conversely, the lowest C/U value (<1.1) and crisis duration (>7 days) did not predict aFHR with good sensitivity or specificity, despite the finding that the very low C/U value and long-duration crisis were more frequently associated with aFHR (Table 3
The fetal Doppler imaging performed during the crisis showed that the malarial infection induced a range of transient hemodynamic changes at the placental and cerebral levels. In all cases, there was an increase of the umbilical resistances due to the degradation of the vascular bed reported in previous studies.714 Relatively high uterine resistance indices were also found, but only in 34% of the pregnancies in group 2 and 48% in group 1, which is in agreement with previous studies, which reported the existence of placental lesions in 50% of cases.16 The uterine resistance at admission was not correlated (in any group) with the parasitemia, fetal cerebral and umbilical hemodynamic changes during the crisis, or crisis duration.
The umbilical resistance at admission in group 2 was significantly lower than in group 1, which suggests that the increase in placental vascular resistance induced by malaria was lower in group 2. During all crises, umbilical resistance increased, but this increase was significantly higher in a long crisis (short crisis, 5%8%; and long crisis, 5%15%; Fig 1A As a response to the deterioration of the fetomaternal exchanges, a reduction of the cerebral vascular resistance by vasodilation was observed, which resulted in an increase in the cerebral perfusion and thus oxygen supply. The cerebral resistance at admission was not different within the groups (1 and 2), whereas the amplitudes of its mean decrease during the crisis were similar for short crises but significantly higher in group 1 long crises.
Last, the C/U ratio, which varies proportionally to the fetal PO2, decreased.17 The C/U ratio at admission was significantly lower in group 2 than in group 1, which confirms that the amplitude of the mean fetal flow redistribution at the beginning of the crisis was lower after 2 years of endemia. The C/U ratio has been tested on hypertensive pregnancies, on pregnancies with idiopathic growth restriction, and on twin pregnancies,1921 and in all cases the sensitivity of this parameter (abnormal when <1.1) for the detection of intrauterine growth restriction was comparable (sensitivity, 85%; and specificity, 98%). This parameter has subsequently been tested as a predictor of the occurrence of acute fetal distress at birth and of neonatal complications. In this application, the sensitivity was 90% for the C/U ratio, 78% for the cerebral index, and 80% for the umbilical index.21 This study confirms the superiority of the C/U ratio for detecting fetal vascular abnormalities associated with hypoxia.22,23 In group 2, umbilical and cerebral resistance and C/U ratio were abnormal in 17%, 39%, and 48%, respectively, whereas in group 1, these parameters were abnormal in 31% of the cases (umbilical resistance), 56% (cerebral resistance), and 83% (C/U). In group 2 (1996), only 47% of the pregnant patients had a C/U ratio less than the normal limit (1.1), whereas in group 1 (1994), there were 83%. The minimum C/U ratio was used to predict the occurrence of FHR abnormalities at the end of gestation, but the sensitivity (group 2, 10%; and group 1, 47%) and specificity (group 2, 77%; and group 1, 75%) of this parameter for predicting the occurrence of aFHR were not very high (Table 3 Another factor for evaluating hypoxia is the mean C/U ratio variation during the crisis. This variation, expressed as a percentage of the normal value at the end of the crisis, measures the flow redistribution amplitude and thus the drop in PO2.17,18 In group 2, the mean C/U variation ± SD during the crisis was similar to that observed in the group 1 short crisis (7% ± 4% versus 8% ± 8%). Nevertheless, this parameter predicted the occurrence of aFHR with rather modest sensitivity and specificity: 10% and 77%, respectively, in group 2 and 47% and 75% in group 1. Moreover, there was no direct relationship between the amplitude of the C/U ratio variation and grade of parasitemia, parity, gestational age at crisis, date and mode of delivery, and fetal weight.
The duration of the crisis was partially correlated with aFHR at delivery, but the sensitivity and specificity of the crisis duration (>7 days) for predicting aFHR were rather modest (group 1, 54% and 70%; and group 2, 33% and 82%; Table 3
The HI, combining both the amplitude of the C/U ratio variation (fall in PO2) and the crisis duration (flow redistribution duration, i.e., C/U <1.1), has thus been calculated and tested as a factor for predicting aFHR at delivery. The mean HI (percent change of C/U ratio x crisis duration in days) was approximately 40 to 70 in short crises (e.g., HI = 56, equivalent to 8 days of 7% PO2 reduction [8 x 7] or 14 days of 4% PO2 reduction [14 x 4]). In long crises, the HI was approximately 140 to 220 (e.g., HI = 182, equivalent to 13 days of 14% PO2 reduction [13 x 14] or 7 days of 26% PO2 reduction [7 x 26]). Only crises of long duration and with a substantial C/U decrease (flow redistribution) were associated with a high rate of aFHR, oligoamnios, cesarean delivery, prematurity, and poor Apgar scores (Tables 1 and 2 There was no severe fetal distress at delivery and no sign of neurologic disability after delivery, which supports the hypothesis that consequences of the transient hypoxia induced by malaria are limited to functional disturbances such as aFHR. Conversely, the transient hypoxia had a limited effect on fetal growth (30% of the cases and between the 5th and 10th percentiles). Moreover, the duration of the crisis (group 1) was not correlated with the degree of growth restriction. This observation is in agreement with the fact that the crisis took place several weeks before delivery, a period during which the fetus could have normal growth, even if portions of the fetal physiologic systems (autonomic nervous system) did not recover in the interim. The gestational age at which the crisis occurred was not correlated with crisis duration, the rate of aFHR, or the amplitude of the fetal flow redistribution. This observation suggests that the capacity of the fetus to respond to acute hypoxia and the vulnerability of the fetal physiologic functions were similar between 27 and 35 weeks.
In the population studied, malaria affected multiparas and primiparas equally, which indicates that successive pregnancies do not induce any protection against malaria. Moreover, the rate of malaria crisis and the maternal symptoms were not different in the 2 groups. However, in the second study (2 years after the first), the pregnancies complicated with malaria were investigated and treated earlier at the beginning of the crisis than in group 1 ( In conclusion, close monitoring of the fetal circulation by Doppler sonography during a malaria crisis allows measurement of the duration and amplitude of the fetal vascular disturbance induced by the malaria, which can predict the consequences of this disease on fetal development and outcome.
Received October 9, 2001, from the Department of Nuclear Medicine and Ultrasound, Institut National de la Santé et de la Recherche Médicale, Centre Hospitalier Universitaire Trousseau, Tours, France (P.A., N.T., S.H., M.G., F.P.); and Section of Obstetrics and Gynecology, Hôpital André Bouron, Saint-Laurent du Maroni, French Guiana (G.C., F.B.). This work was supported by grants from Fondation Cordet and Institut National de la Santé et de la Recherche Médicale.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |