Objective. Preterm infants are often presumed to have a pressure passive cerebral circulation implying that a low mean arterial blood pressure (MABP) results in reduced cerebral perfusion. The aim of this study was to determine whether cerebral blood flow (CBF) was compromised in preterm infants whose MABP fell below 30 mm Hg (4 kPa).
Methods. Thirty preterm infants undergoing intensive care were studied within the first 24 hours of life. CBF was measured using near infrared spectroscopy. The infants were analyzed in two groups on the basis of their MABP at the time of study: group 1 had a MABP below 30 mm Hg and group 2 more than 30 mm Hg. CBF in the two groups was compared.
Results. There was no significant difference in the mean CBF between the two groups. In group 1 the median MABP was 27.2 mm Hg (range, 23.7–29.9 mm Hg) and CBF was 13.9 (standard deviation, ±6.9) mL · 100 g−1 · min−1. In group 2 the median MABP was 35.3 mm Hg (range, 30.1–39.3 mm Hg) and CBF was 12.3 (standard deviation, ±6.4) mL · 100 g−1 · min−1. Mortality and incidence of cranial ultrasound scan abnormalities were also not significantly different.
Conclusion. These results indicate that preterm infants undergoing intensive care are able to maintain adequate cerebral perfusion at a MABP in the range of 23.7 to 39.3 mm Hg.
- CBF =
- cerebral blood flow •
- MABP =
- mean arterial blood pressure •
- NIRS =
- near infrared spectroscopy •
- Hbo2 =
- oxyhemoglobin •
- Sao2 =
- arterial oxygen saturation •
- TcPco2 =
- transcutaneous partial pressure of carbon dioxide •
- SD =
- standard deviation
Several studies of cerebral perfusion in sick preterm infants undergoing intensive care have provided evidence of a failure of cerebral autoregulation leading to a pressure-passive circulation.1 2 In these circumstances cerebral blood flow (CBF) has a direct linear relationship with mean arterial blood pressure (MABP). Failure of cerebral autoregulation combined with hypotension has been implicated in the pathogenesis of both ischemic and hemorrhagic cerebral lesions.3 4 As it is not possible in routine clinical practice to determine CBF directly, it has become standard practice to maintain MABP at more than an arbitrary level to minimize the risk of cerebral hypoperfusion. Several authorities have suggested that MABP should not be allowed to fall below 30 mm Hg (4 kPa) in sick infants undergoing intensive care,4-6although there is little published evidence to support this. By contrast, the policy on the neonatal intensive care unit at University College London Hospital has been to tolerate a MABP in the range of 20 to 30 mm Hg (2.7–4.0 kPa), provided the infant was clinically well with no signs of hypovolemia or evidence of reduced perfusion of vital organs.
The aim of this study was to investigate the relation between MABP and CBF during the first 24 hours of life in preterm infants undergoing intensive care, especially at low values of MABP. To investigate the validity of applying a cutoff value of 30 mm Hg, we divided infants into two groups: those with a MABP less than 30 mm Hg (group 1), and those with a MABP greater than 30 mm Hg (group 2).
The study was approved by the University College London Hospital's committee on the ethics of human research, and informed parental consent was obtained before each study.
Thirty preterm infants admitted to the neonatal intensive care unit at University College London Hospital were studied (clinical details in Table 1). The only exclusion criterion was an abnormal cranial ultrasound scan at the time of the study. The median gestational age of the infants was 27.5 weeks (range, 24–34 weeks) and median birth weight was 862 g (range, 460–2100 g). All of the infants were studied within the first 24 hours after delivery and after initial stabilization in intensive care. They were all receiving intermittent positive pressure ventilation with increased levels of inspired oxygen. They were all clinically stable with normal arterial oxygen tensions and no clinical signs of hypovolemia at the time of study. Four infants were receiving dopamine infusions for the treatment of hypotension; 3 at a rate of ≤5 μg · kg−1 · min−1 and 1 at 10 μg · kg−1 · min−1, and in all cases this resulted in an increase in MABP, although only 1 infant receiving dopamine had a MABP more than 30 mm Hg. Twenty-two infants were normally grown for gestational age and 8 had evidence of antenatal placental dysfunction, on the basis of either a birth weight below the 3rd centile or abnormal antenatal umbilical artery Doppler waveforms.
Decisions about the treatment of each infant were made by the attending physician on clinical grounds. The study protocol did not require any modification of existing treatment.
Near Infrared Spectroscopy (NIRS)
CBF was measured by NIRS using the oxyhemoglobin (Hbo 2) method that has been described in detail elsewhere.7 8 In brief, a rapid rise in arterial oxygen saturation (Sao 2) is induced by an increase of the inspired oxygen concentration. The resulting rise in cerebral Hbo 2 concentration is measured by NIRS and CBF is obtained from a modification of the Fick principle. To account for the light scattering properties of the tissue a differential pathlength factor is incorporated into the calculations. A value of 5.13 was used in this study.9
Changes in cerebral chromophore concentrations were monitored using a NIRO500 spectrophotometer (Hamamatsu Photonics KK, Hamamatsu City, Japan). Transmitting and receiving optodes were placed on the infant's head in the temporoparietal or frontal region ∼4 cm apart. The interoptode distance was measured using calipers. Near infrared light at four different frequencies was transmitted via fiber optic bundles to the emitter optode and thereby through the infant's head. Emergent light was collected through the detector optode and transmitted to the photomultiplier tube of the spectrophotometer. Changes in optical densities were measured continuously and recorded for later analysis. Simultaneous measurements were made of MABP through an indwelling arterial catheter connected to a pressure transducer (HP 78834A, Hewlett-Packard, Palo Alto, CA) and transcutaneous carbon dioxide tension (TcPco 2) that was calibrated against an arterial blood sample taken at the time of study. Sao 2 was measured with a pulse oximeter (Nellcor N200, Nellcor Inc, Hayward, CA) modified to provide beat-to-beat measurements. TcPco 2, Sao 2, MABP, and heart rate measurements were obtained simultaneously with the NIRS measurements once every second, and stored by computer.
CBF (mL · 100 g−1 · min−1) was calculated from the equation: Equation 1where K is a constant (0.614) incorporating the molecular weight of hemoglobin and the tissue density and [Hb] is the hemoglobin concentration in g · 100 mL−1. Two to 6 measurements of CBF were made from each infant during a period of 2 to 3 hours and the mean value was calculated. The mean MABP was calculated from recordings collected at the time of each CBF measurement, and the mean value for all measurements was obtained.
Cranial ultrasound scans were performed daily for the first week of life and then weekly until discharge or death. The cranial ultrasound appearance at a corrected gestational age of 40 weeks or at the time of death was recorded.
The infants were divided into two groups on the basis of their MABP at the time of study. For those infants whose MABP fluctuated around 30 mm Hg, the group was assigned according to whether the MABP was more than or less than 30 mm Hg for the majority of the study. The data were tested for normal distribution and for equality of variance and the two groups were compared using an unpaired t test. A power calculation showed that a sample size of 14 in one group and 16 in the other group was sufficient to detect a difference in CBF of 7.5 mL · 100 g−1 · min−1 between the two groups with a power of 80% and a confidence level of P< .05. Multiple linear regression analysis and Spearman rank order correlation were used to determine the dependence of CBF on MABP, TcPco 2, time of study, gestational age, birth weight, and the presence of placental compromise (Jandel Scientific Sigmastat 2.0, Jandel Scientific, San Rafael, CA).
Values for CBF for each infant are displayed in Table 1. Fourteen of the 30 infants had a MABP less than 30 mm Hg (group 1), their median MABP was 27.2 mm Hg (range, 23.7–29.9 mm Hg). The remaining 16 infants (group 2) had a median MABP of 35.3 mm Hg (range, 30.1–39.3 mm Hg). There was no significant difference in gestational age, birth weight, or TcPco 2 between the two groups, but there was a significant difference in the age at the time of study, with a mean of 13.5 hours [±standard deviation (SD), ±6.6 hours] in group 1, and 20.8 (SD, ±7.9) hours in group 2 (P < .01, unpaired t test). Results are summarized in Table 2.
The mean CBF in group 1 was 13.9 (SD, ±6.9) mL · 100 g−1 · min−1 and in group 2 was 12.3 (SD, ±6.4) mL · 100 g−1 · min−1. The SD of CBF measurements from individual infants ranged from 0.6 to 7.1 mL · 100 g−1 · min−1. There was no significant difference in CBF between the two groups (unpairedt test, 29 degrees of freedom). No correlation was found between CBF and MABP (see Fig 1). There was no correlation between CBF and time of study, gestational age, birth weight, or the presence of placental dysfunction, but there was a positive correlation of CBF with TcPco 2(P < .02). Figure 1 shows the relationship of CBF with MABP and Fig 2 with TcPco 2.
Five infants in each group died before discharge. There was no significant difference between the two groups when compared for the incidence of subsequent cranial ultrasound scan abnormalities.
This study found no significant difference in CBF between infants with MABP ≥30 mm Hg and those with MABP in the range 23.7 to 29.9 mm Hg.
The Hbo 2 technique for measuring CBF in the newborn has been validated against the intravenous133Xenon clearance method.10 11 The assumptions underlying the method have been discussed elsewhere.12 All methods for CBF measurement in newborns have limitations in accuracy, which is reflected in the intraindividual variability. However, the range of CBF values we obtained was similar to those measured in very preterm infants using both 133Xenon clearance13and positron emission tomography.14
Our results indicate that infants with MABP in the lower range are able to maintain a satisfactory CBF, suggesting that autoregulation was intact. We did, however, demonstrate a positive correlation with TcPco 2. These findings are consistent with previously published studies of CBF measured shortly after birth in spontaneously-breathing and mechanically-ventilated preterm infants using 133Xenon clearance.13 15 This was a cross-sectional rather than a longitudinal study, which was not designed to detect whether a small subgroup of the sickest infants had a pressure passive circulation. Although the power of the study was insufficient to detect a difference in CBF between the two groups of less than 7.5 mL · 100 g−1 · min−1(P < .05), there was no trend for a lower CBF in the low MABP group. By contrast the study was able to demonstrate the expected correlation of CBF with Paco 2.
Animal models used to study the effects of hypotension on CBF have shown that although autoregulation may be intact, preterm animals have a reduced range during which autoregulation occurs. The lower limit of autoregulation is the same in the term and the preterm fetal lamb;16 17 however, the resting MABP in the preterm lamb lies closer to the lower limit of autoregulation, making it more susceptible to cerebral ischemia. Similar studies in human infants are not possible and the limits of autoregulation in the preterm neonate have not been determined. CBF velocity measurements using Doppler ultrasound have suggested loss of autoregulation in the preterm infant,2 18 although such studies are difficult to interpret because of changes in the diameter of insonated vessels.19
It is generally assumed that there must be a critical level of CBF required to maintain cellular integrity and homeostasis, but this level has not yet been established in preterm infants. Initial reports suggested that a CBF of less than 10 mL · 100 g−1 · min−1 was associated with neurologic impairment.20 However more recent studies have shown that a CBF as low as 5 mL · 100 g−1 · min−1 can be compatible with a normal neurologic and cognitive outcome.14 21 Visual evoked potentials have been shown to be preserved in infants with CBF as low as 4.3 mL · 100 g−1 · min−1.15 In our study 11 infants had a CBF measurement below 10 mL · 100 g−1 · min−1. Of these, only 4 were in the group with lower blood pressure; 5 died and 3 developed ultrasound evidence of cerebral injury.
Blood pressure nomograms as a function of birth weight have been published for term and preterm infants.6 22-25 However there are discrepancies between different studies, probably because of differences in the clinical treatment of preterm infants. Some studies of infants with a birth weight of <1500 g have shown that a MABP of less than 30 mm Hg was more than 2 SD below the mean MABP,6 25 but the concept of a normal range for MABP in preterm infants undergoing intensive care is problematic. A MABP lower than 30 mm Hg for any significant period has been associated with periventricular hemorrhage,4 6 22 cerebral ischemic lesions, and death.4 Although there is evidence of a statistical association between low MABP and adverse outcome, there may not be a causal link. The statistical association may simply reflect the fact that the sickest infants tend to have the lowest blood pressures. It is therefore not surprising that the treatment of blood pressure in preterm infants varies widely. It is common practice in many neonatal intensive care units to maintain the MABP more than 30 mm Hg using plasma expanders or pressor agents to ensure adequate perfusion to vital organs, particularly the brain.
An alternative recommendation from the Joint Working Group of the British Association of Perinatal Medicine was that the MABP in mm Hg should not fall below the gestational age of the infant in weeks.26 This recommendation was based on a consensus decision and there are no studies to confirm these lower limits of MABP. In our study 9 infants had a MABP lower than their gestational age. The mean CBF (±SD) in this group was 15.2 ± 7.7 mL · 100 g−1 · min−1 compared with 14.6 ± 7.0 mL · 100 g−1 · min−1in the remainder. Our study does not support the assumption that a MABP in the range of 23.7 to 29.9 mm Hg is necessarily associated with a reduction in cerebral perfusion. Fluid overload and edema are frequent results of excessive use of plasma expanders. The direct effects of pressor agents such as dopamine on cerebral vasculature in preterm infants have not been fully investigated. Recent evidence has demonstrated that these agents may have a vasoconstrictive action27 and may even reduce cerebral perfusion.28 29 Although dopamine causes a rise in MABP this may be a result of either improved left ventricular output30 or of increased systemic vascular resistance. Hence, therapeutic measures such as inotrope infusions to maintain MABP more than arbitrary levels may be inappropriate and even harmful.
Our data show a strong dependence of CBF on Paco 2 and emphasize the need to monitor Paco 2 continuously and to use ventilatory strategies that minimize the risk of hypocarbia.
We found no relationship between CBF and MABP in very preterm infants with MABP in the range of 23.7 to 39.3 mm Hg. By contrast, there was a positive relationship between CBF and TcPco 2. These data suggest that cerebral hypoperfusion is more likely to be associated with hypocarbia than with hypotension, and that the use of pharmacologic interventions to maintain MABP more than an arbitrary limit has little proven basis.
This work was funded by the United Kingdom Medical Research Council and a personal donation in memory of Filippo Galassi (1972–1992).
The authors acknowledge the contribution to this work of Dr Ann Stewart, Prof David Edwards, and Dr David McCormick, as well as colleagues in the Departments of Paediatrics and Medical Physics, and the staff of the Neonatal Unit at University College London Hospital Trust. We would also like to thank Prof E. O. R. Reynolds, Dr Jane Hawdon, and Dr Nick Evans for valuable discussions.
- Received October 28, 1997.
- Accepted March 13, 1998.
Reprint requests to (J.M.) Department of Paediatrics, University College London School of Medicine, University St, London WC1E 6JJ, UK.
- Miall-Allen VM,
- De Vries LS,
- Whitelaw AGL
- ↵Greenough A, Morley JC, Roberton NRC. Acute respiratory disease in the newborn. In: Roberton NRC, ed. Textbook of Neonatology. 2nd ed. London and Edinburgh, UK: Churchill Livingstone, Longman Group; 1992:408
- ↵Elwell CE. Data selection criteria. In: A Practical Users Guide to Near Infrared Spectroscopy. Hamamatsu City, Japan: Hamamatsu Photonics KK; 1995:85
- Anthony MY,
- Evans DH,
- Levene MI
- Lou HC,
- Skov H,
- Henrikson L
- Altmann DI,
- Perlmann JM,
- Volpe JJ,
- Powers WJ
- Versmold Hans T, Kitterman JA, Phibbs RH, Gregory GA, Tooley WH
- Report of a Joint Working Group of the British Association of Perinatal Medicine and the Research Unit of the Royal College of Physicians
- Roze JC,
- Tohier C,
- Maingueneau C,
- Lefevre M,
- Mouzard A
- Copyright © 1998 American Academy of Pediatrics