Article Text
PDF
Abstract
Background A haemodynamically important patent ductus arteriosus (PDA) is a risk factor for brain damage in preterm infants. The authors previously reported lower regional cerebral oxygen saturation (rScO2) in infants with PDA, which recovered after administration of indomethacin. However, PDA ligation has been reported to pose an even higher risk of neurodevelopmental impairment.
Objective To investigate the impact of surgical closure of PDA on rScO2 and cerebral fractional tissue oxygen extraction (cFTOE), measured by near-infrared spectroscopy, and on amplitude-integrated electro-encephalography (aEEG) measured brain activity.
Design/methods In 20 preterm infants (gestational age 24.7–30.4 weeks; birth weight 630–1540 g), blood pressure, arterial saturation, rScO2, cFTOE and aEEG were monitored before, during and up to 24 h after surgery.
Results Before surgery, median (range) rScO2 was 53% (41–68%), and during surgery, but before ductal clipping, it was 46% (31–89%). Eleven infants showed a drop in blood pressure and 13 infants a drop in rScO2 during surgery (range 2–21%), accompanied by a decrease in aEEG amplitude. Twelve infants had rScO2 values below 50% during surgery, with five being below 40%. Only at 24 h after surgery was rScO2 higher (61% (36–85%), p<0.05) and cFTOE values lower (0.38 (0.09–0.61); p<0.05) compared with preclipping values.
Conclusion Ductal ligation poses a risk for a further decrease in already compromised cerebral oxygenation in preterm infants.
Statistics from Altmetric.com
Introduction
A haemodynamically important patent ductus arteriosus (PDA) is a common problem in preterm infants in the first weeks of life and is associated with increased morbidity and mortality.1,–,4 A left-to-right shunt (ductal steal phenomenon) across the duct adversely affects the perfusion and oxygenation of important organ systems,5,–,7 while a PDA-induced increase in pulmonary flow with prolonged ventilatory dependency is an important risk factor for the development of chronic lung disease.4 8 Therefore, a PDA should be treated quickly to achieve early closure and prevent complications in preterm infants,. Preferably, non-invasive closure with indomethacin or ibuprofen should be used, but if this is unsuccessful, surgical closure will be required. Both therapies are considered to be relatively safe, although medical closure may transiently induce cerebral, intestinal and/or renal perfusion changes,6 9 10 while surgical closure has been associated with transient hypotension, pneumothorax and bleeding.11 Recently, however, several studies have suggested that surgical closure of a PDA causes long-term morbidity and poses a high risk for adverse neurodevelopmental outcome.4 8 12
What is already known on this topic
▶ Recent papers have reported a relationship between surgery for a patent ductus arteriosus (PDA) and neurodevelopmental impairment.
▶ Preterm infants with a PDA have lower regional cerebral oxygen saturation (rScO2; mean 62%) compared with infants without a PDA (mean 70%).
What this study adds
▶ This study shows that in a substantial number of infants rScO2 was already low before surgery (mean 53%) and decreased further during surgery.
▶ The present study demonstrates that surgical closure of a PDA poses a risk for a further decrease in the already compromised cerebral oxygenation in infants with PDA.
There is little information on cerebral oxygenation and electrical brain activity during surgery to close a PDA.13 14 A recent study with near-infrared spectroscopy (NIRS) in preterm infants undergoing surgical ductal closure revealed only short-term changes in cerebral volume15 and brain tissue oxygenation16 during and immediately after surgery. The study, however, did not monitor cerebral oxygenation and electrical brain activity over a longer period after surgery.
We therefore investigated the impact of surgical ductal closure on cerebral regional oxygen saturation and extraction, using NIRS, and the effect on electrical brain activity using amplitude-integrated electro-encephalography (aEEG) before, during and up to 24 h after surgical closure of the duct.
Patients and methods
Patients
Twenty-eight infants with a gestational age of less than 32 completed weeks admitted to the neonatal intensive care unit (NICU) of Wilhelmina Children's Hospital between June 2007 and February 2009 were included in this exploratory study. The patients had undergone surgical ligation of a PDA because indomethacin treatment was contraindicated or because the PDA had failed to close after a (repeated) course of indomethacin.
Informed parental consent was obtained in all cases. The study was approved by the Medical Ethics Committee of the University Medical Centre, Utrecht.
Clinical data
Obstetric and intrapartum data were collected from hospital records. Neonatal data were collected prospectively. Treatment decisions were made by the attending neonatologist. Arterial oxygen saturation (SaO2) was monitored using pulse oxymetry on a limb, and arterial blood pressure by an indwelling arterial catheter (umbilical, tibial or radial artery) in all infants. Blood pressure support was started according to the decision of the attending neonatologist as indicated by the guidelines used in our NICU.
A blood pressure support scoring system was used to assess the level of blood pressure support.17 In brief, score 0 indicated no support; score 1, volume expansion and/or dopamine ≤5 μg/kg/min; score 2, dopamine >5≤10 μg/kg/min; score 3, dopamine >10 μg/kg/min or dopamine and dobutamine >5 μg/kg/min; score 4, dopamine and dobutamine >10 μg/kg/min; and score 5, additional epinephrine and/or corticosteroids.
The diagnosis of a haemodynamically important PDA was based on clinical indices and confirmed by echocardiography (left atrial and/or left ventricular dilatation: 1:4; internal ductal diameter >1.4 mm/kg; left pulmonary artery end diastolic flow >0.2 m/s).
Monitoring of cerebral tissue oxygenation and oxygen extraction
We used NIRS-determined regional cerebral oxygen saturation (rScO2) when assessing changes in regional cerebral oxygenation. Although rScO2 cannot be used as a robust quantitative estimator of cerebral oxygenation, it can be reliably used in the individual patient to detect substantial changes in cerebral oxygenation.18 19 We used an INVOS 4100/5100 near-infrared spectrometer (Somanetics, Troy, Michigan, USA). A transducer containing a light emitting diode and two distant sensors was attached to the fronto-parietal side of the neonatal head. rScO2 was calculated from the differential signals obtained from these two sensors, expressed as the venous-weighted per cent oxygenated haemoglobin (oxygenated haemoglobin/total haemoglobin (oxygenated haemoglobin+deoxygenated haemoglobin)).20
To investigate the balance between oxygen delivery and oxygen consumption, the relative cerebral fractional tissue oxygen extraction (cFTOE) can be formulated as a ratio: (SaO2−rScO2)/SaO2. Since it is a ratio, an increase might indicate either reduced oxygen delivery to the brain with constant oxygen consumption by the brain, or higher oxygen consumption than oxygen delivery. The opposite is true in the case of a decrease in cFTOE, either reflecting a decrease in oxygen extraction by the brain due to less use of oxygen, or constant oxygen consumption by the brain with increased oxygen delivery to the brain.19
Monitoring electrical brain activity
To monitor cerebral function and electrical brain activity, a single channel EEG cerebral function monitor (CFM 4640; Lectromed Devices, Oxford, UK) was used for determination of the aEEG. aEEG electrodes were applied by a qualified nurse to the neonatal skull. The aEEG was recorded from two parietal electrodes as described by Toet et al.21
Quantitative analysis of the aEEG data was performed with an algorithm as used in modern digital aEEG devices (Brainz monitor): the minimum, mean and maximum voltages (µV) in 15 four-second periods of the aEEG signal were determined. The medians of these 15 values for minimum, mean and maximum voltages were calculated. These data were averaged within selected 10 min periods (ie, average amplitudes).
Study design
Twenty-eight infants with a diagnosis of PDA and subsequent surgical ligation were enrolled in the present study which was part of a larger prospective clinical study, in which consecutively admitted preterm infants (gestational age <32 weeks) were simultaneously monitored for arterial blood pressure, heart rate, SaO2, rScO2, cFTOE and aEEG starting as soon as possible after birth depending on the availability of a NIRS device and parental consent. The surgical procedure always took place on the NICU. Anaesthetic drugs were administered by a paediatric cardio-anaesthesiologist who was unaware of the study results. The end of the study was defined as 24 h after surgical ductal closure. The simultaneously collected data were stored on a personal computer for off-line analysis (sample frequency 10 Hz) using Poly 5 software (Inspektor Research Systems, Amsterdam, The Netherlands). Arterial haemoglobin concentration was measured daily or more frequently if indicated. Arterial blood gasses were measured just before, during and immediately after the surgical procedure and then every 4 h up to 24 h after surgery or more frequently if necessary. Cranial ultrasound studies were performed before surgery and repeated within 24 h after surgery (or more frequently where appropriate).
To reduce the data to manageable proportions and allow comparisons over time, 10 min periods of monitored data were selected and averaged for mean arterial blood pressure (MABP), SaO2, rScO2, cFTOE and the aEEG signal before surgery (presurg), just before clipping (preclip), just after clipping (clip) and at 1, 2, 6, 12 and 24 h after clipping. Data were also available from 14 infants before the development of PDA (pre-PDA) and are shown for descriptive purposes only (figure 1).
(A) Patterns (Box–Whisker plots) of arterial oxygen saturation (SaO2), (B) mean arterial blood pressure (MABP), (C) regional cerebral oxygen saturation (rScO2) and (D) cerebral fractional tissue oxygen extraction (cFTOE) before surgery (presurg), during surgery before clipping (preclip), just after clip placement (clip) and at 1, 2, 6 and 24 h after clipping of the duct (*p<0.05 vs presurg; #p<0.05 vs preclip) in 20 infants. The white Box–Whisker plots show the values of 14 infants in the period without a haemodynamically significant patent ductus arteriosus (median (range) postnatal age: 8.5 h (2–17 h)).
Statistical analysis
Data are summarised as mean values±SD or as median values (ranges) where appropriate. Differences between the MABP, rScO2, cFTOE and aEEG values over time were evaluated by analysis of variance for repeated measurements. Adjustments for multiple comparisons were made using a Bonferroni test. A p<0.05 was considered statistically significant. SPSS v 15.0 was used for statistical analysis.
Results
Clinical characteristics
The complete data set was not available in eight of the 28 infants due to technical and registration problems. These infants were therefore excluded from further analysis. Of the 20 remaining infants, 14 were initially treated with indomethacin and one with ibuprofen. Five infants had contraindications for medical treatment (intestinal perforation/necrotising enterocolitis), unstable grade 2 intraventricular haemorrhage, pulmonary haemorrhage or echocardiographic suspicion of a slight stenosis at the level of the aortic isthmus (not established during surgery). All 20 infants included in this study had a haemodynamically important PDA with a significant left-to-right shunt diagnosed by the criteria described in the Patients and methods section. Data before the development of PDA were available for 14 infants in the analysed group. The clinical data of the 20 infants are shown in table 1. Median (range) gestational age and birth weight were 26.7 (24.7–30.4) weeks and 907 (630–1540) g, respectively. Median (range) postnatal age during surgical closure was 7 (4–39) days. All infants were on mechanical ventilation during the study period and anaesthetised with at least one dose of sufentanyl (1 µg/kg) and atracurium (0.5 mg/kg). They were subsequently sedated with low dose morphine (10 µg/kg/h) after the surgical procedure. Haemoglobin concentrations were always stable and within the normal range before and after surgery (data not shown). Mean (±SD) pCO2 values before and immediately after surgery were 49 mm Hg (±7) and 53 mm Hg (±12), respectively.
Clinical characteristics
Pattern of arterial blood pressure
Systolic, mean and diastolic blood pressure immediately but transiently increased after clipping (p<0.05) but remained stable thereafter at the presurgical level during the entire study period as shown in figure 1B (MABP). In 25% (n=5) of the infants, blood pressure support was necessary before surgery, increasing to 35% (n=7) up to 2 h after clipping (blood pressure support score ranged from 1 to 2). After 24 h, only two of the infants (10%) needed blood pressure support. Data are shown in table 2. During the surgical procedure, MABP decreased in 50% of the patients, but only in three patients did it decrease below 30 mm Hg, considered the lower limit of acceptable MABP. Except for a short increase just after clipping (p<0.05), heart rate was always stable at the various time points (data not shown).
Blood pressure support scores
Patterns of arterial saturation, cerebral oxygen saturation and oxygen extraction
Figure 1A,C,D depicts SaO2, rScO2 and cFTOE at the various time points. SaO2 was quite stable during the study period (figure 1A). The rScO2 values of 14 infants in the period before having a significant PDA (pre-PDA) (median (range) postnatal age: 8.5 h (2–17)) were higher compared with the presurgical values as shown in figure 1C). Median (range) rScO2 before surgery was 53% (41–68%) and tended to decrease during anaesthesia before clipping. In the hours after clipping, median rScO2 showed a sustained increase and at 24 h after clipping median (range) rScO2 was 61% (36–85%) (p<0.05 vs rScO2 before clipping). The results are summarised in figure 1C. The cFTOE pattern mirrored that of rScO2 and is summarised in figure 1D: the median (range) cFTOE values 24 h after clipping were lower compared with the values before clipping (preclip: 0.47 (0.08–0.68); at 24 h after clipping: 0.38 (0.09–0.61); p<0.05, preclip vs values at 24 h, respectively). However, when individual patterns were analysed, during surgery four infants had rScO2 values below 40%, which put them at risk of hypoxic brain damage (also see Discussion section).
Pattern of aEEG activity
Figure 2 shows the maximum, mean and minimum aEEG amplitudes at the various time points. All infants had a burst suppression-like pattern, with minimum activity values of below 5 µV before surgery, probably due to the low dose morphine administered (10 µg/kg/h). During surgery, aEEG amplitudes tended to decrease, but were only significantly lower at 2 h after clipping as compared to presurgical values (p<0.05). Only after 24 h did the aEEG amplitudes recover (p<0.05 vs maximum amplitude at 2 h). Analysis of individual aEEG amplitudes during surgery revealed that the decrease in electrical brain activity was most pronounced in those infants with the lowest rScO2 values, but was not significantly different when comparing values lower and higher than 50%. Figure 3 illustrates this, showing two distinct MABP, rScO2 and aEEG patterns in two individual infants. In patient A (figure 3A) with low cerebral saturations (<40%) before and during surgery and MABP <35 mm Hg, a further decrease in aEEG amplitude occurred during surgery: the aEEG pattern changed from a burst suppression-like pattern to a continuous low voltage pattern. Upon clipping, rScO2, MABP and aEEG amplitude quickly recovered. Patient B (figure 3B), who had values of rScO2 well above 40% and stable MABPs, showed stable aEEG amplitude patterns during the entire surgical procedure.
The patterns of maximal (orange filled circle), mean (green filled circle) and minimum (blue filled circle) aEEG amplitude (mean±SD) before, during and up to 24 h after clipping of the duct in 20 patients (*p<0.05 vs presurg, #p<0.05 vs 2 h). aEEG, amplitude-integrated electro-encephalography.
Individual patterns of mean arterial blood pressure (MABP), regional cerebral oxygen saturation (rScO2) and aEEG patterns of (A) a 27.3-week, 1050 g male infant and (B) a 28-week, 1000 g male infant during the studied period. Note the low rScO2 (49%) before surgery and the subsequent drop to extremely low values of rScO2 and MABP in patient A, accompanied by a substantial drop in aEEG amplitude during the surgical period, and the quite stable patterns of rScO2, MABP and aEEG amplitude in patient B. aEEG, amplitude-integrated electro-encephalography.
Discussion
The results of the present study in infants undergoing surgical ligation of a PDA suggest there were no substantial changes in cerebral oxygen supply and extraction (measured by NIRS) during the surgical procedure. After clipping of the duct, there was sustained recovery of cerebral oxygenation. From 24 h after clipping, rScO2 was significantly higher as compared to preclipping values and comparable to the values in preterm infants without PDA.18 aEEG amplitude showed a significant drop 2 h after surgery as compared to presurgery values, but recovered afterwards. It is important, however, to emphasise that these conclusions are derived from the median values as shown in figures 1 and 2. A subgroup of four infants exhibited substantially lower cerebral saturation values during the surgical procedure before ductal clipping, with a consequent drop in the aEEG amplitude (see also individual tracings in figure 3A).
There are several points with regard to the results of the present study, especially in light of recent evidence that surgical ligation of a PDA carries increased risks for chronic lung disease and adverse neurodevelopmental outcome.4 8 12 First, it was found that presurgical cerebral oxygenation (rScO2) was substantially lower and extraction higher (cFTOE) in the surgical ligation cohort as compared with pretreatment values in our recently reported cohort of infants with successful pharmacological treatment: median rScO2: 63% versus 51%, p<0.05; median cFTOE: 0.48 versus 0.38, p<0.05 respectively.5 This may be due to the longer existence and/or left-to-right shunt of the duct in the ligation group, especially in those infants where pharmacological treatment had failed (n=15), leading to prolonged perturbation in oxygenation of the immature brain. However, it must be mentioned that the ligation group had somewhat lower gestational ages and lower birth weights (albeit not significant). Nonetheless, it is interesting that our group reported some years ago that presurgery oxygenation of the brain in neonates who underwent an arterial switch operation was predictive for neurodevelopmental outcome.22
Second, we noticed that in a subgroup of infants in the present study, cerebral oxygenation (and to a lesser extent blood pressure) fell to quite low values during surgery in the preclipping period with, in a few cases, an accompanying drop in electrical brain activity to extremely low voltages. The surgical procedure, which can induce haemodynamic perturbation, together with the effects of anaesthesia, may have attenuated the already low cerebral oxygenation. There is also evidence that the pharmacodynamics of medication, particularly opioids, are different during different postmenstrual and postnatal ages,23 24 giving rise to excessive sufentanyl levels in some extremely preterm patients. This could have had a negative impact on the blood pressure of these infants. However, it remains uncertain whether or not this was the case, since re-analysis of our data did not reveal a correlation between gestational or postnatal age and blood pressure. Furthermore, we cannot exclude a causative negative effect of anaesthetics such as opioids on electrical brain activity as shown in figure 2A. However, aEEG measured electrical brain activity recovered simultaneously with improved cerebral oxygenation and blood pressure while the anaesthetics were still effective. Blood pressure-dependent perfusion of the brain may also have played a role as suggested by Tsuji et al.25 There is increasing evidence from studies in newly born piglets and from human studies that rScO2 values below 40% for 30 min are related to functional and morphological brain damage and adverse neurodevelopmental outcome.26
Finally, cerebral oxygenation remained lower and extraction higher during the first hours after clipping of the duct and only normalised to expected values 24 h after completion of surgery.18 It has been reported that cardiac function deteriorates during prolonged existence of a duct27 and during surgical duct closure.28 It may be that a suboptimal haemodynamic condition together with the extended effect of anaesthetics (see also above) in the postsurgical period are responsible for the suboptimal cerebral oxygen supply and increased cerebral oxygen extraction in several of our studied infants. This suggestion is supported by the fact that seven of the 20 infants needed blood pressure support in the first hours after clipping of the duct.
Our study showed several instances of low cerebral saturations and aEEG amplitudes during surgery just before clipping. Although in this study we did not examine whether low cerebral saturations and aEEG amplitudes gave rise to (ischaemic) brain damage, we recommend an active policy towards these apparently low blood pressures, especially when coinciding with low rScO2 values, during the preclip period, with adequate blood pressure support.
In conclusion, surgical closure of a haemodynamically important patent ductus is essential in selected cases and necessary to restore normal haemodynamics. However, it is important to emphasise that the current results also suggest that the surgical procedure itself and possible adverse effects of the anaesthetics used during surgery may have undesirable side effects on cerebral oxygenation and blood pressure. Nonetheless, the results of the present exploratory study suggest that the procedure itself usually has no harmful effects on cerebral oxygenation and electrical brain activity, although in a minority of cases exceptionally and potentially dangerously low cerebral oxygenation was detected with extremely low electrical brain activity. These observations may partly explain the results of follow-up studies that report surgical closure of the duct is a risk factor for neurodevelopmental outcome in extremely preterm infants. A prospective study with long-term follow-up is warranted to elucidate the impact of surgical duct clipping on neurodevelopmental outcome.
References
Footnotes
-
Competing interests None.
-
Ethics approval This study was conducted with the approval of the Medical Ethics Committee of University Medical Center, Utrecht.
-
Provenance and peer review Not commissioned; externally peer reviewed.