Changes in the Doppler blood flow patterns of the cerebral veins in fetuses affected with severe intrauterine growth restriction before 32 weeks of gestation 

Background

The aim of this project is to evaluate the brain venous circulation in fetuses with severe intrauterine growth restriction (IUGR) before 32 weeks of gestation. Intrauterine growth restriction (IUGR) due to placental insufficiency is associated with hypoxic brain damage in nearly 15 % of cases. Recent studies suggest that this prevalence can increase up to 60% if a wide range of minor and long-term neurological lesions are also included. At present, the ability to predict brain damage prenatally is still poor, mainly due to a limited knowledge on the fetal adaptive processes prior to brain injury. In IUGR fetuses, maintenance of oxygen delivery to the fetal brain is achieved by a blood flow centralisation process, which in clinical practice is generally assessed in the middle cerebral artery (MCA). A reduced MCA-pulsatility index (PI) is considered as compensatory vasodilatation and constitutes one of the stages of the hypoxic progression. Preliminary studies suggested that not only the MCA is affected, but all major cerebral arteries, and, that the adaptive pattern of each vessel might differ in relation to the progression of hypoxia. In general, the fetal brain blood flow increases, thus, suggesting that the venous blood flow should also show physiological adaptive changes and that its evaluation might provide valuable information in the study of the fetal hypoxic adaptation/deterioration process.

The fetal cerebral venous circulation has mainly been explored due to the risk of developing cardiac failure in cases with a Galen vein aneurism. While in this high blood flow velocity pattern, in normal circumstances, its identification is as yet, difficult. Recent technical improvements in ultrasound (US) equipment have allowed a more reliable recognition of the normally slow blood flow of the fetal cerebral veins. Previously, several researchers explored the value of different brain veins in the study of the chronically hypoxic fetus, reporting changes in the blood flow patterns. They also showed that these vessels can be reliably registered.

Although these few available reports agreed that the venous flow is affected in IUGR fetuses, there is still a necessity for more information concerning the specific changes in venous circulation within the cerebral vascular adaptive processes. Therefore, the aims of this study were, to explore the feasibility and reproducibility of the pulsed Doppler evaluation of the fetal brain venous circulation, to evaluate its differences between normally grown and severely affected IUGR fetuses and finally, to investigate the evolving changes in relation to the progression of hypoxia.

Method

Thirty five cases with severe IUGR (<32 weeks of gestation) defined as an estimated growth <10th centile and an abnormal umbilical artery puslatility index (UA-PI, mean _2 SD) were studied longitudinally. In addition, 61 normally grown fetuses were evaluated as controls. None of the IUGR fetuses had abnormal karyotype or structural abnormalities. Maternal age at the time of diagnosis was; (median) 31 years (range 22 - 39 years) and gestational age; (median) 26 weeks+2days (range, 22+3 - 32.0 weeks). The project was approved by the ethics committee and the informed consent of all participants was obtained in all cases.

As the topographic description of the fetal brain venous system has been previously described elsewhere, we were focused on the methodological aspects of its evaluation. Ultrasound (US) and Doppler examinations were performed with a Siemens/Antares US equipment (Siemens Medical Systems, Malvern, PA, USA) with a 6-2 MHz curved linear array. All Doppler examinations were performed in absence of fetal corporal and respiratory movements and with the mother in voluntary suspended respiration. For the Doppler measurements the angle of insonation was maintained below 30 degrees and corrected manually when necessary. Directional colour, or power Doppler US were used to clearly locate the different vessels, and a minimum of 5 consecutive good quality waveforms measured. The mechanical and thermal indices were kept below 1. The mean examination time was 8 minutes (range 4-15 min).

Pulsed Doppler examination of the UA-PI was performed in a free loop of the umbilical cord. The ductus venosus (DV) was located in a transverse or alternatively, in a sagittal view of the fetal abdomen at its emergence from the portal vein, and in the MCA in a transverse view of the fetal head at the level of the sphenoid bone, immediately after its origin from the circle of Willis.

The cerebral venous circulation was studied in the superior sagittal sinus (SS) localised in a mid-sagittal view with an anterior projection of the fetal head, in the GV located in a middle transverse plane with a posterior projection of the fetal head, just below the turbulence produced after its confluence with the inferior sagittal sinus, in the straight sinus (STR) located in the same anatomical projection as the continuation of the GV, and before the junction with the transverse sinus, and in the transverse sinus (TS) located below the occipital bone in a transverse view of the posterior fossa, below the cerebellum. The pulse repetition frequency of the Doppler US was lowered enough to detect the slow blood movement in these vessels. The presence or absence of pulsatility was recorded. A pulsatile venous blood flow pattern was defined as a minimal difference of 15% between systolic and diastolic flows, as suggested by Dubiel et al. In all cases, peak systolic velocity (PSV) and time-averaged maximum velocity (TAMV) were estimated, and in cases where the blood flow was defined as pulsatile, the pulsatility index for veins (PIV) was also calculated. Caution was taken in avoiding to exert excessive pressure over the fetal head, as the Doppler waveforms can be altered.

In the IUGR group, a sub-division in relation to the progression of hypoxia was done as:

  • stage 1 (n=32) - umbilical artery pulsatility index (UA-PI) >2 SD or umbilical artery absent end diastolic flow (UA-AEDF), and middle cerebral artery PI (MCA-PI) mean ± 2SD;
  • stage 2 (n=62) - UA-PI >2 SD or UA-AEDF and MCA-PI <2SD;
  • stage 3 (n=38 ) - reversed end diastolic UA blood flow and MCA-PI <2SD with present DV-atrial flow; and
  • stage 4 (n=14) - reversed or absent atrial flow in the ductus venous (DV).

At diagnosis, each case was assigned to one of the hypoxic stages and modified according to the clinical evolution. The median number of examinations per case was 2 (range 1-12).

For reproducibility analysis, in 15 normal cases, 95% limits of agreement (LA) between observers with 95% confidence intervals (CI) as suggested by Bland and Altman, were estimated for the TS-PI measurement. Kappa analysis was used to estimate the agreement between observations in defining a pulsatile blood flow pattern in the GV.

The results of the MCA, DV and UA Doppler examination were available to the clinicians, but not the remaining measurements. The decision to deliver the fetus was taken at the discretion of managing physicians on the basis of current clinical protocols, either when the pregnancy reached 28 weeks of gestation and there was cerebral vasodilatation (MCA-PI mean <2SD) and/or reversed UA end diastolic flow, or with absence or reversed atrial flow in the DV at any stage of pregnancy.

Data were stored in databases and analysed with the SPSS 12.0 Statistical Package (SPSS Inc., Chicago, Illinois, USA). Comparisons between IUGR and NG fetuses were tested with paired T test. Differences within the IUGR hypoxic stages were estimated with analysis of variance and Bonferroni Post-Hoc test. A p value <0.05 was considered significant.

Results

The survival rate in the IUGR group was 85% (30/35), there were 2 fetal and 3 neonatal deaths. The two stillbirths showed reversed atrial flow in the ductus venosus before 28 weeks, and despite detailed counselling, the parents decided not to deliver the fetus.

Doppler signals from the 4 cerebral veins could be recorded in all IUGR cases. In controls, the TS and GV could be examined in all cases, whereas the straight sinus, in 52 / 61 (81%), and the sagittal sinus in 48 / 61 (75%) of cases. Interobserver 95% limits of agreement for TS-PI were –0.157 (95% CI, –0.154 to –0.160) to 0.170 (95% CI, 0.167 to 0.173) (mean difference, 0.006 (SD, 0.09)) and Kappa agreement for a pulsatile blood flow pattern in the GV was 0.82 (95% CI, 0.67-0.94).

The presence of pulsatile blood flow was variable. The transverse and straight sinuses showed a pulsatile blood flow pattern in 96% and 98% of controls, respectively, and in all IUGR fetuses. In the Galen vein and in the sagittal sinus, the frequency of pulsatile blood was higher in IUGR fetuses, as compared with controls (74% vs. 35%, p=0.0001; 83% vs. 60%, p= 0.001, respectively). In these 2 veins, the rate of a pulsatile blood flow pattern increased significantly in relation to the progression of hypoxia.

The sagittal and straight sinuses, and the Galen vein from IUGR fetuses showed similar values, as compared with controls and only the transverse sinus PI in IUGR fetuses was significantly reduced (p=0.0008). In relation to the progression of hypoxia, PI values in all cerebral veins from IUGR fetuses tended to decrease towards stage 2, showing an opposite pattern as that of ductus venosus. However, from stage 3 onwards, they increased again. This tendency was statistically significantly only in the TSPI. DV-PI in IUGR fetuses showing a gradual increase in relation to the progression of hypoxia that became more evident from stage 3.

In IUGR fetuses analysed as a whole, PSV and TAMV values in all veins were significantly increased, as compared with controls. In relation to the progression of hypoxia, PSV and TAMV increased and became significant at stage 2, decreasing later in stages 3 and 4, thus, mirroring the PI values.

Discussion

The results of this study showed that in early stages of hypoxia, the fetal brain veins reflect the increment in cerebral blood flow as an increase in velocity and a reduction in the pulsatility index. In more advanced hypoxic stages, there is an increase in pulsatility, similarly to other veins such as the ductus venosus, and a trend for reduced velocities, suggesting a subclinical state of hypoxic cardiac dysfunction.

The transverse sinus was the first fetal brain vein to be systematically studied. Laurichesse-Delmas and cols. constructed normal reference velocity and PI values, and Senat et al. found a significant association between TS-PIV with DVPI and low Apgar scores. Conversely, Cheema and cols. in a group of 109 high risk pregnancies, found that only 9 of them showed an abnormal TS-PI with no increment in the prevalence of adverse perinatal outcome. In our study, we found a significantly reduced PI value in the TS-PI from IUGR fetuses with a biphasic trend in relation to the progression of hypoxia. The reduced PI values in stage 2 reflected the enhanced cerebral flow observed in the arterial system, and their further increment might correspond to the reduction of right heart compliance in the context of fetal hypoxic deterioration. Other potentially contributing factors could be a progressive lack of compliance of the vein wall.

Dubiel et al. studied the blood flow pattern in the Galen vein and found an increase in the prevalence of pulsations in high risk pregnancies associated with a higher risk of perinatal complications. These results were later corroborated by Cheema et al. who also pointed at the Galen vein as the more robust parameter associated with an adverse perinatal outcome.

According to our study, a pulsatile blood flow pattern in the Galen vein was statistically more frequent in IUGR fetuses and its prevalence increased as the hypoxic insult progressed.

However, the relatively high frequency of a pulsatile blood flow pattern of the GV in normal pregnancies precludes its clinical application. We observed a significantly increased PSV and TAMV in the GV from IUGR fetuses. While the presence of a pulsatile pattern is not a clear indicator of the risk, the combined estimation of increased velocity and presence/ absence of a pulsatile blood flow pattern might offer better clinical information in cases at a higher risk of perinatal complications.

The superior sagittal sinus and the straight sinus were the most difficult veins to locate in normal pregnancies, while in the IUGR they were easier to find, due to increased velocities, together with fewer fetal movements. One of the drawbacks in locating the SS, is that for a clear recording, a clear complete mid-sagittal view of the fetal head is needed. Similar to the other brain veins, the blood flow pattern was not always pulsatile, in normal pregnancies was about 60% whereas in IUGR fetuses, increased up to 80%. In relation to the progression of hypoxia, SS showed the same trend as the Galen vein.

The straight sinus was also studied by Cheema et al. They could not find an increased prevalence in the adverse perinatal outcome in cases with increased maximum velocity in this vein. In their study, as in ours, the STR was recorded less successfully than the TS and the Galen vein. In relation to the progression of hypoxia, we observed a continuous increment in the STR blood velocities, starting from hypoxic stage 2 onwards, probably reflecting the vascular response to the overall enhanced brain blood flow. The blood flow pattern in this vein was almost always pulsatile in normally grown and IUGR fetuses. However, cases with no-pulsatile blood flow were more often seen after 31 weeks of gestation. Variations in the presence of a pulsatile blood flow pattern in normal fetuses can be mainly related to gestational age, as previously suggested by Laurichesse-Delmas and cols. A continuous blood flow is more frequently observed towards the end of pregnancy in all fetal cerebral veins. Nevertheless, it appears that the Galen vein shows preferentially a continuous blood flow pattern, regardless of the stage of pregnancy whereas the transverse and straight sinuses showed preferentially a pulsatile pattern. The sagittal sinus showed a wide variation in its blood flow pattern.

Feasibility to record the Doppler waveforms from the brain vessels has previously been addressed by different authors. Laurichesse-Delmas reported a 98% success rate to obtain the Doppler waveform from the TS and Cheema and cols reported a 100% success in getting signals form the Galen vein and 82% success rate for the straight sinus. We were able to record all veins from all IUGR fetuses, and in nearly 85% of the normally grown fetuses. The most challenging vessel to record was the SS, followed by the STR, as previously mentioned. We were always able to obtain recordings from the GV, and TS in both groups. Reproducibility analyses showed a small disagreement when two operators measured the TSPI, and a good agreement in defining a pulsatile blood flow pattern, meaning that in the hands of trained operators the venous cerebral circulation can be reliably evaluated. While in clinical practice it seems that the GV and TS might offer potentially useful information, the evaluation of the straight and sagittal sinuses might contribute important physiological information concerning the cerebral vascular redistribution process.

In this study, we measured all main cerebral vessels, which allowed us to get a full picture of brain venous changes under hypoxia. Interpretation of changes in venous vessels is difficult and several factors must be considered. Firstly, the territories drained are ill-defined and therefore, interpretation of regional changes is even more limited than with arteries. Secondly, while changes observed in velocity are probably mainly determined by blood flow, the changes in pulsatility have a mixed component. In early stages of hypoxic deterioration, the PI tended to decrease, as opposite to vessels from the inferior venous circulation, such as DV suggesting that PI changes are probably influenced by the increase in blood flow. However, in more advanced stages of hypoxic deterioration, the influence of reduced cardiac compliance seems to prevail over increased flow, and brain veins follow the pattern of the ductus venosus, though probably with a lower magnitude.

Our definition of the progression of hypoxia was based on experimental and clinical studies where, in the presence of a continuous hypoxic insult, the cardio-vascular adaptation is first manifested as changes in the umbilical artery (UA) blood flow, followed by a vascular centralisation process and finally, as changes in the fetal venous return. While by noninvasive methods, it is extremely difficult to properly classify a fetus in relation to its hypoxic status, we considered that using this stratification allowed us to reliable include cases in similar stages of the hypoxic progression. Nevertheless, we must accept some degree of overlapping in the stages described here. In summary, from our results we can conclude that the fetal cerebral vascular redistribution process in IUGR is also expressed as an increment in the venous blood flow velocities in all cerebral veins, as an increased pulsatility index in the transverse sinus and as an increment in the prevalence of a pulsatile blood flow pattern in the Galen vein. These results may suggest that in venous vessels velocity estimations are probably a more reliable indicator of increased blood flow in the chronically hypoxic fetus. This information should be analysed in conjunction with arterial blood flow changes and with other techniques addressing regional blood flow perfusion, to better understand the cerebral vascular adaptive process in fetuses complicated with severe growth restriction.

This information is not meant to replace the advice of any physician or qualified health professional. The information provided by Cerebra is for information purposes only and is not a substitute for medical advice or treatment for any medical condition. You should promptly seek professional medical assistance if you have concerns regarding any health issue.

Page last updated: 16/12/2011 13:30 
 
© 2012 Cerebra | Registered charity No: 1089812 | Company Limited by Guarantee, registered in England and Wales 4336208