Cerebral oxygen metabolism in neonatal hypoxic ... - SAGE Journals

Journal of Cerebral Blood Flow & Metabolism (2014) 34, 87–94 & 2014 ISCBFM All rights reserved 0271-678X/14 $32.00 www.jcbfm.com

ORIGINAL ARTICLE

Cerebral oxygen metabolism in neonatal hypoxic ischemic encephalopathy during and after therapeutic hypothermia Mathieu Dehaes1,2, Alpna Aggarwal1,2,8, Pei-Yi Lin3,4, C Rosa Fortuno5,6, Angela Fenoglio1, Nade`ge Roche-Labarbe3,4,9, Janet S Soul5,6, Maria Angela Franceschini3,4 and P Ellen Grant1,3,4,7 Pathophysiologic mechanisms involved in neonatal hypoxic ischemic encephalopathy (HIE) are associated with complex changes of blood flow and metabolism. Therapeutic hypothermia (TH) is effective in reducing the extent of brain injury, but it remains uncertain how TH affects cerebral blood flow (CBF) and metabolism. Ten neonates undergoing TH for HIE and seventeen healthy controls were recruited from the NICU and the well baby nursery, respectively. A combination of frequency domain near infrared spectroscopy (FDNIRS) and diffuse correlation spectroscopy (DCS) systems was used to non-invasively measure cerebral hemodynamic and metabolic variables at the bedside. Results showed that cerebral oxygen metabolism (CMRO2i) and CBF indices (CBFi) in neonates with HIE during TH were significantly lower than post-TH and age-matched control values. Also, cerebral blood volume (CBV) and hemoglobin oxygen saturation (SO2) were significantly higher in neonates with HIE during TH compared with age-matched control neonates. Post-TH CBV was significantly decreased compared with values during TH whereas SO2 remained unchanged after the therapy. Thus, FDNIRS–DCS can provide information complimentary to SO2 and can assess individual cerebral metabolic responses to TH. Combined FDNIRS–DCS parameters improve the understanding of the underlying physiology and have the potential to serve as bedside biomarkers of treatment response and optimization. Journal of Cerebral Blood Flow & Metabolism (2014) 34, 87–94; doi:10.1038/jcbfm.2013.165; published online 25 September 2013 Keywords: cerebral blood circulation; cerebral hemoglobin oxygen saturation; cerebral metabolic rate of oxygen consumption; near infrared spectroscopy; neonatal hypoxic ischemic encephalopathy; therapeutic hypothermia

INTRODUCTION Neonatal hypoxic ischemic encephalopathy (HIE) occurs in B2 to 3 per 1,000 live births in the USA, with devastating consequences including neurodevelopmental disabilities such as cerebral palsy, life-long cognitive impairment, and epilepsy.1 In HIE, there is a period of reduced blood flow (ischemia) and oxygen delivery (hypoxia) followed by reperfusion with transient energy recovery, and then a secondary energy failure.2 Treatment is aimed at preventing the cascade of events that begin with reperfusion and lead to secondary energy failure and inevitable cell death; this is the ‘window of opportunity.’3 Theories for injury mechanisms in this window suggest a central role for excitotoxicity and oxidative stress with the injury evolving over hours to weeks.2 Therapeutic hypothermia (TH) is the treatment of choice for HIE, as it is the only treatment with proven efficacy and acts through a variety of mechanisms, including those associated with decreased neuronal energy metabolism.4 Although early studies showed that TH decreased rates of both mortality and disability, recently published long-term outcomes have demonstrated significance only for decreased mortality rates.5 In addition, a recent

meta-analysis advocated refinement of patient selection and TH duration whereas others have suggested that currently excluded neonates be considered in future trials.6,7 Thus a need for further optimization of TH remains. Magnetic resonance imaging (MRI) is the gold standard for detecting acute brain injury, but is impractical as an early screening tool or as a tool to optimize therapy of HIE due to its non-portability and high cost. Electroencephalography and amplitude-integrated electroencephalography can provide essential measures of neuronal activity but would benefit by additional direct measures of cerebral hemodynamics and cerebral metabolic rate of oxygen consumption (CMRO2) changes that precede or accompany secondary energy failure. Commercial continuous wave near infrared spectroscopy systems have been used in HIE to monitor relative cerebral hemoglobin oxygen saturation (rSO2) and fractional tissue oxygen extraction (FTOE ¼ (SaO2) – rSO2)/ SaO2) where SaO2 is the arterial oxygen saturation.8 However, neither rSO2 nor fractional tissue oxygen extraction accurately reflect CMRO2, as CMRO2 also depends on oxygen delivery (via CBF).

1 Fetal Neonatal Neuroimaging and Developmental Science Center, Division of Newborn Medicine, Boston Children’s Hospital, Boston, Massachusetts, USA; 2Department of Pediatrics, Harvard Medical School, Boston, Massachusetts, USA; 3Athinoula A Martinos Center for Biomedical Imaging, Department of Radiology, Massachusetts General Hospital, Charlestown, Massachusetts, USA; 4Department of Radiology, Harvard Medical School, Boston, Massachusetts, USA; 5Department of Neurology, Boston Children’s Hospital, Boston, Massachusetts, USA; 6Department of Neurology, Harvard Medical School, Boston, Massachusetts, USA and 7Department of Radiology, Boston Children’s Hospital, Boston, Massachusetts, USA. Correspondence: Dr M Dehaes or Dr PE Grant, Division of Newborn Medicine, Boston Children’s Hospital, 300 Longwood Avenue, Boston, MA 02115, USA. E-mail: [email protected] or [email protected] The authors would like to acknowledge funding from NIH/NICHD R21-HD058725 (PEG), R01-HD042908 (MAF), and NIH/NCRR P41-RR14075 (BRR) as well as funding from the Isabelle and Leonard H Goldenson Biomedical Research Foundation (MD) and the William Randolph Hearst Foundation (MD). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health. 8 Alpna Aggarwal is currently at the Cohen Children’s Medical Center, New Hyde Park, New York 11040, USA. 9 Nade`ge Roche-Labarbe is currently at Laboratory PALM, Universite´ de Caen Basse-Normandie, Caen 14032, France. Received 3 May 2013; revised 15 August 2013; accepted 26 August 2013; published online 25 September 2013

Cerebral oxygen metabolism in HIE M Dehaes et al

88 In contrast, frequency domain NIRS (FDNIRS) allows the absolute quantification of hemoglobin oxygen saturation (SO2) and cerebral blood volume (CBV).9 Frequency domain near infrared spectroscopy has been used to demonstrate that CBV is significantly elevated in acute HIE brain injury, whereas SO2 shows no significant difference from neonates without suspicion of brain injury.10 Diffuse correlation spectroscopy (DCS) was developed for in vivo applications to measure microvascular blood flow non-invasively in deep biologic tissue without the injection of external contrasts. In previous work, we used DCS to measure an index of cerebral blood flow (CBFi) in the preterm brain and combined this measurement with FDNIRS measures of SO2 to provide an index of CMRO2 (CMRO2i).11 More recently, we described the relationship between CMRO2i and gestational age in the developing premature brain by using concurrent FDNIRS and DCS measurements.12 FDNIRS–DCS CMRO2i, unlike SO2, is sensitive to early brain development, increasing with gestational age and throughout the first year of life12 and might facilitate early detection of abnormalities and delays in maturation of specific brain regions.13 Using our FDNIRS–DCS bedside instrumentation, the aim of this current study was to determine cerebral hemodynamics and oxygen metabolism in neonatal HIE during TH and compare the parameters to values measured after TH and measured in agematched control. MATERIALS AND METHODS Patients Twelve term neonates diagnosed with moderate to severe HIE were enrolled in a prospective observational study at Boston Children’s Hospital from July 2009 to January 2012 with institutional review board approval from Partners Healthcare, which includes both Massachusetts General Hospital and Brigham and Women’s Hospital and an inter-institutional agreement granted by Boston Children’s Hospital. Each neonate that underwent TH met inclusion criteria based on published studies:14 (1) gestational ageX36 weeks and birth weightX2 kg, (2) evidence of fetal distress including history of acute perinatal event, and (3) evidence of neonatal distress as evidenced by at least one of the following: apgar scorep5 at 10 minutes, pHp7.0 within 1 hour of birth or base deficitp16 mEq/L or need for ventilation for at least 10 minutes after birth, (4) evidence of neonatal encephalopathy by physical exam from the neurologist, and (5) abnormal cerebral function monitor evidenced for example by the presence of seizures identified by amplitude-integrated electroencephalography. Two of the twelve neonates recruited were excluded from analysis due to a change in diagnosis from HIE to a rare metabolic disorder with encephalopathy and myopathy in one neonate and inability to obtain FDNIRS–DCS frontal measurements in the other. All neonates had light skin. Twenty-two control neonates were recruited from the well baby nursery at Brigham and Women’s Hospital between 2009 and 2011 under the same institutional review board. No neonate had signs or symptoms of fetal distress and only routine resuscitation was required after delivery. To further homogenize the healthy population and make it more comparable to the HIE population, five neonates with darker skin color were excluded from the analysis resulting in a total of 17 control neonates. Demographics are detailed in Supplementary Table 1.

Neurodevelopmental Outcomes Gross motor function was assessed using either the gross motor function classification system for cerebral palsy (GMFCS) score15 or the Bayley psychomotor developmental index score whereas cognitive function was assessed using the Bayley mental development index (MDI) score.16 Motor and cognitive scores were then combined yielding a single score that was classified into three categories: normal (Bayley MDI or PDIX85 and GMFCS ¼ 1), mildly abnormal (70pBayley MDI or PDIo85 or GMFCS ¼ 2) or abnormal (Bayley MDI or PDIo70 or GMFCS42, or death). Neurodevelopmental outcomes were assessed in eight infants at 18 months (median), patient #7 was lost to follow-up after NICU discharge and patient #10 died a few days after birth. Thus the outcome score was normal in five infants (5/9, 55.6%), mildly abnormal in three infants (3/9, Journal of Cerebral Blood Flow & Metabolism (2014), 87 – 94

33.3%), and the newborn who died was scored abnormal (1/9, 11.1%). Details are provided in Supplementary Table 1.

Therapeutic Hypothermia Standard TH procedures17 were initiated within 6 hours after birth by passive cooling at the referral hospital or on transport, and by active whole-body cooling upon admission to Boston Children’s Hospital. Body temperature was targeted to 33.51C for 72 hours of TH. Rewarming procedures were then applied for 12 hours, raising the temperature by 0.51C every 2 hours. For all measurements acquired within 90 hours after birth, whole-body temperature waso361C. Both the cooling and rewarming periods are grouped together as a single ‘TH’ period as the patient is technically considered hypothermic (temperatureo361C). PostTH measurements were all acquired at normothermia. Medications used during TH period are listed in Supplementary Table 2 and include sedatives, analgesics, and neurotransmitters.

Optical Measurements Non-invasive FDNIRS–DCS measurements were performed at the bedside during TH (6–78 hours), the rewarming period (78–90 hours), and post-TH (106–305 hours) in neonates with HIE and after birth in healthy neonates (13–77 hours). The hybrid instrumentation has been described previously in literature.11 The power emitted by FDNIRS–DCS was well below the American National Standards Institute light levels and the combined instrumentation was compact enough to be moved to the neonate’s bedside on a small cart. The probe was designed such that the light would penetrate into the neonatal cerebral cortex and subcortical white matter9 while minimizing the error due to extra-cerebral contamination.18 Measurements were obtained from up to three locations on the head including left, middle, and right frontal regions. The probe was positioned on the scalp for 10–16 seconds up to three times per location. Repeating the measurement by slightly moving the probe helped to ensure reproducibility of the data. This procedure was also used to minimize data contamination from local heterogeneities such as hair and extra-cerebral veins.18 Data quality and rejection criteria were based on a published study11 and used to discard spurious source of noises such as motion artifacts and ambient light contamination, resulting in only high-quality data. In neonates with HIE, arterial oxygen saturation (SaO2) was recorded from routine monitors whereas hemoglobin concentration in the blood (HGB) [g/dl] was determined by complete blood count analysis at the time of the measurement and extracted from medical records. In healthy neonates, SaO2 was assumed to be 98%19 and HGB values were derived from clinical reference tables.20 Cerebral blood volume [ml/100 g] and SO2 [%] were computed from the data acquired with the FDNIRS system using a published approach.9 Diffuse correlation spectroscopy data was modeled as a diffusion process and included a set of temporal auto-correlation functions acquired sequentially every second. The diffusion equation was used to derive CBFi [mm2/s] by fitting the normalized intensity temporal auto-correlation function, to the measured signal of multiply scattered light.21,22 Cerebral oxygen metabolism was computed via the calculation of an index of CMRO2 (CMRO2i, [Mol/dl mm2/second]) using Fick’s principle22 such that HGB ðSaO2 SvO2 Þ MWHb HGB 1 ðSaO2 SO2 Þ ¼ CBFi MWHb b

CMRO2i ¼ CBFi ðCaO2 CvO2 Þ ¼ CBFi

ð1Þ

where CaO2 and CvO2 [Mol/dl] are the arteriolar and venous O2 concentrations, MWHb [g/Mol] is the molecular weight and, SaO2 and SvO2 are the arteriolar and venous O2 saturations. As in our previous studies, we assumed SO2 ¼ aSaO2 þ bSvO2 with a þ b ¼ 1 where the arterial:venous (a:b) contribution ratio is 1/4:3/4.23

Statistical Analysis Non-parametric Wilcoxon signed rank tests were performed to compare hemodynamic and metabolic variables in neonates with HIE during TH (HIE–TH) and after TH (post-TH), and in control neonates (Control) at a significance level of 0.05. Age-matched comparisons were also performed by defining two control age sub-groups with respect to hours of age at the time of measurement: 0oager48 hours and 48oagep96 hours. A paired Wilcoxon signed rank test was performed at a significance level of 0.05 to & 2014 ISCBFM


89 compare each neonate with HIE with its corresponding age-matched control sub-group. Number of measurements for each variable in subsequent comparisons are not identical for all variables because of their independent data quality criteria.11

RESULTS Figure 1 depicts temporal distribution of frontal (A) CMRO2i, (B) CBFi, (C) CBV, and (D) SO2 values in neonates with HIE treated with TH (colored bullets) and control neonates (black squares) with respect to age at the time of measurement. Standard errors of the mean are shown when more than one frontal location was averaged. Therapeutic hypothermia period (light gray background) corresponds to the period between 6 and 90 hours and includes both cooling and rewarming phases whereas normothermic post-TH period (light blue background) starts at 91 hours and ends at the last measurement (at 305 hours). Note that the x-axis is broken twice in post-TH phase, as no measurements were taken during those particular periods. Cerebral oxygen metabolism in HIE–TH was almost always lower than in healthy controls (Figure 1A). Most of the infants had CBFi ranging between 1.5 and 4.5 10 6 mm2/second during TH (Figure 1B). Infant #9 had the highest value (7.1 10 6 mm2/second at 65 hours of age), which coincided with periods of desaturation corresponding to abnormally low partial pressure of oxygen in the arterial blood, per the medical record. The CBV value was low for infant #8 at 20 hours of age (Figure 1C). Infants #6, #7, and #9 had the highest CBV values during TH all around 72 hours of age. Infant #6 had

right temporal lobe and intraventricular hemorrhages. Subpial and intraventricular hemorrhages were reported in infant #7 as well as a subdural hemorrhage within the posterior regions in infant #9. However, none of the infants had intracranial blood products in the frontal cerebral regions where FDNIRS–DCS measurements were acquired. Infant #8 also exhibited the lowest SO2 value during TH (54.5% at 20 hours, Figure 1D) when compared with all other treated or control newborns. Corresponding temporal distribution of the mean frontal oxy(HbO), deoxy- (HbR), and total hemoglobin (HbT) used to derive CMRO2i, CBV and SO2 are shown in Supplementary Figure 1. In addition, individual raw values of regional (middle, left, and right frontal) HbO, HbR, HbT, and CBFi measured in neonates with HIE and healthy neonates are shown in Supplementary Figure 2 and Supplementary Figure 3, respectively. Clinically indicated MRI was performed within 6 days of life in all 10 patients who underwent TH in this study. Two small posterior punctate foci with decreased apparent diffusion coefficient around 3 cm deep from the scalp were reported in infant #3, as well as few scattered foci of decreased apparent diffusion coefficient deep in the cortex and white matter in infant #7. All other neonates had apparent diffusion coefficient maps that were clinically interpreted as normal. Recall that CBFi is calculated using the DCS system whereas CBV and SO2 are calculated using the FDNIRS system. Thus, there are neonates whose data met inclusion criteria for one instrument and not the other and therefore not all subjects have all the same number of CBFi measurements as CBV and SO2 measurements at

CMRO2i [× 10-10 Mol/dl · mm2/s]

CBFi [× 10-6 mm2/s]

8

8

6

6

4

4

2

2

0

0 0

25 50 75 100 125 150 225 300 Hours of age at time of measurement

0

25 50 75 100 125 150 225 300 Hours of age at time of measurement

CBV [ml/100g] 5

SO2 [%]

90

4

80

3 70 2 60

1 0

0

25 50 75 100 125 150 225 300 Hours of age at time of measurement 1

2

9

10

3

4

5

Healthy Control

50

0

6

25 50 75 100 125 150 225 300 Hours of age at time of measurement 7

8

Standard Error

TH Post-TH

Figure 1. Frontal averages of (A) cerebral oxygen metabolism index (CMRO2i), (B) cerebral blood flow index (CBFi), (C) cerebral blood volume (CBV), and (D) hemoglobin oxygen saturation (SO2) in neonates with hypoxic ischemic encephalopathy (HIE) during therapeutic hypothermia (TH; light gray background) and after TH (post-TH; light blue background), and healthy control neonates with respect to hours of age at the time of measurement. Colored bullets represent neonates with HIE and black squares depict measurements for healthy control neonates. Standard errors of the means are shown when more than one measurement was averaged. Note that the x-axis is broken during two periods after TH, as none of the measurement was acquired in these periods. & 2014 ISCBFM

Journal of Cerebral Blood Flow & Metabolism (2014), 87 – 94


90 CMRO2i [× 10-10 Mol/dl · mm2/s] 6

CBFi [× 10-6 mm2/s] 8

p < 10-5

p < 0.01 4

6

p < 0.01

p < 0.01

4 2 0

2 HIE-TH n = 19

Post-TH n = 11

HIE-TH n = 19

Control n = 15

0

HIE-TH n = 20

CBV [ml/100g] 5 4 3 2 1 0

p < 0.01

Post-TH n = 11

HIE-TH n = 20

Control n = 16

SO2 [%] p < 10-3

90 p = 0.014

80 70 60

HIE-TH n = 20

Post-TH n = 12

HIE-TH n = 20

Control n = 18

50

HIE-TH n = 20

Post-TH n = 12

HIE-TH n = 20

Control n = 18

Figure 2. Boxplots of (A) cerebral oxygen metabolism index (CMRO2i), (B) cerebral blood flow index (CBFi), (C) cerebral blood volume (CBV), and (D) hemoglobin oxygen saturation (SO2) in neonates with hypoxic ischemic encephalopathy (HIE) during therapeutic hypothermia (TH; HIE–TH) and after TH (post-TH), and in healthy control neonates (Control). On each box, the central mark is the median, the black square is the mean, the edges of the box are the 25th and 75th percentiles, and the whiskers show the 95% confidence interval. Empty circles denote outliers and significant statistical comparisons are indicated with its corresponding P value as well as the number of measurements for each group/condition.

every time point. Figure 2 shows boxplots of (A) CMRO2i, (B) CBFi, (C) CBV, and (D) SO2 values between HIE–TH and post-TH conditions (left box), and between HIE–TH and Control groups (right box) with corresponding number of measurements. On each box, the central mark is the median, the black square is the mean, the edges of the box are the 25th and 75th percentiles, and the whiskers show the 95% confidence interval. Outliers are depicted by empty circles whereas P value is provided in the box when the comparison is significant. Neonates with HIE had significantly lower CMRO2i during TH than both post-TH (Po0.01) and Control conditions (Po10 5; Figure 2A). The low CMRO2i in HIE–TH compared with Control remained significant when age-matched (Po10 4). Consistent with the low CMRO2i, CBFi in HIE–TH was also significantly lower than post-TH values (Po0.01) and Control neonates (Po0.01; Figure 2B). The difference between HIE–TH and Control conditions remained significant when age-matched (Po0.01). Surprisingly, CBV was significantly elevated despite the low CBFi in the HIE–TH condition compared with both post-TH (Po0.01) and Control conditions (Po10 3; Figure 2C). Agematched comparison also demonstrated significantly higher CBV in HIE–TH than Control groups (Po10 3). Although SO2 is the parameter monitored most frequently with commercial systems, differences in SO2 were the least discriminating and were only significantly elevated in the comparison between HIE– TH and Control (P ¼ 0.014; Figure 2D), with this significance remaining when age-matched (P ¼ 0.019). A similar analysis was also performed with HbO, HbR, and HbT measurements. Oxy hemoglobin and HbT in HIE–TH were significantly higher than in post-TH condition (Po0.01 and P ¼ 0.01, respectively) whereas HbR was not significantly different. None of the comparisons between HbO, HbR, and HbT in HIE–TH versus Control was significantly different.

DISCUSSION Indices of Cerebral Oxygen Metabolism and Cerebral Blood Flow In this prospective pilot study of 10 term neonates undergoing TH for HIE and 17 healthy control neonates, bedside FDNIRS–DCS measurements of CMRO2i were significantly lower in neonates with HIE during hypothermia (HIE–TH) compared with postJournal of Cerebral Blood Flow & Metabolism (2014), 87 – 94

hypothermia (post-TH) and age-matched controls (Control). In fact, of all hemodynamic parameters studied, CMRO2i was the parameters that best discriminated among the different conditions and groups. To our knowledge, this is the first report of bedside human neonatal CMRO2i measures that demonstrates the expected lower CMRO2i during TH compared with post-TH measures and age-matched healthy controls. These data suggest that FDNIRS–DCS instrumentation has the potential to be a good bedside monitor of newborns with HIE undergoing TH or other therapies. This low CMRO2i we observed in neonates with HIE during TH contrasts sharply with our prior study of untreated neonates with HIE. In particular, we found a higher relative CMRO2 (rCMRO2) in acute neonatal HIE without TH compared with neonates without suspicion of brain injury.10 Although data were acquired between 1 and 12 days of birth all neonates were within 5 days of symptom onset and all had acute injury documented by diffusion-weighted imaging near the time of the FDNIRS measures. In addition, rCMRO2 was estimated using the Grubb’s relationship,24 as the DCS system was not available at the time of the study. However, in a previous MRI study, we have shown increased cerebral perfusion in untreated neonatal HIE.25 Thus with prior studies showing higher rCMRO2 and increased cerebral perfusion in neonatal HIE, it was important to discover that both CBFi and CMRO2i were decreased during TH. This observation was also consistent with a previous MRI study that showed decreased cerebral perfusion in acute neonatal HIE with hypothermia.26 Although MRI methods for estimating CMRO2 have been published in the adult literature,27–31 there is currently little data on the MRI measurement of CMRO2 in neonates. One recent abstract reported results for 6 normothermic preterm infants without acute brain injury.32 These MRI methods are based on a relationship between venous T2 imaging and venous oxygenation that is described through a calibration process determined in vitro. As NIRS, this calibration method depends on hemoglobin (hematocrit).33,34 These methods provide whole brain estimates of CMRO2. Although a regional variant of this method is being developed it has not been applied to neonates.35 In addition, it is currently not practical to screen for abnormal CMRO2 with MRI before the necessary rapid initiation of TH, nor is MRI a practical tool to monitor response to TH or other future therapies because & 2014 ISCBFM


91 of high cost, difficulty in scheduling rapid studies and need to transport unstable infants outside the NICU. Few reports of neonatal CMRO2 using gold standard techniques such as positron emission tomography36 or Xenon clearance37 exist and as a result of increasing concerns for radiation exposure in neonates, no new data has been published since 1993. Thus although other, arguably more accurate, approaches exist, these approaches are currently not acceptable for research purposes in neonates. The calculation of CMRO2 is strongly dependent on the technique utilized to measure CBF. Here the subscript ‘i’ in CMRO2i is used to acknowledge that CMRO2 is derived using a CBF index (CBFi), which although a flux (in mm2/second), correlates well with other measures of cerebral perfusion (in ml/100 g per minute). In fact, DCS CBFi correlates well with MR arterial spin labeling38 and with phase-encoded velocity mapping (VENC) MRI measures of relative blood flow in the jugular veins and superior vena cava in neonates/children with congenital heart defect.39 Further support for the use of DCS CBFi measures was provided by the high correlation (r2 ¼ 0.73, Po0.01) between relative CBFi and Xenon CT cerebral blood perfusion measures in critically brain-injured adults.40 These comparisons show that CBFi measured by DCS offers reasonable agreement with other modalities, in particular those providing perfusion measurements, and therefore provides a reliable index of CMRO2.

Cerebral Blood Volume Our current observation of elevated CBV in HIE during TH has also been observed in one term newborn with moderate HIE between 2 and 8 hours of age during TH compared with pre-treatment value using commercial continuous wave near infrared spectroscopy.41 This increase was followed by a decrease in CBV between 68 and 74 hours of age during TH, and another increase during the rewarming phase. However, the study did not specify if patient or table values for HGB were used to derive CBV. In a previous study, van Bel et al,42 found CBV lowered during cooling and increased during rewarming in 12 neonates and infants (age 0 to 11 months) undergoing open-heart surgery using commercial continuous wave near infrared spectroscopy.42 However, none of these neonates had documented brain injury, the degree of hypothermia was inconsistent among the patient population and there was no detail on the duration of cooling. An increase in CBV during rewarming was also noted using commercial continuous wave near infrared spectroscopy in seven out of eight neonates. However these neonates were preterm infants (o31-week gestational age) with very low birth weight presenting with hypothermia at admission and none had documented brain injury.43 Elevated CBV with TH is interesting as CBFi is decreased rather than increased. Possible reasons for increased CBV include mediations such as dopamine, which has a similar effect as cocaine,44 increased carbon dioxide partial pressure and increased venous pressure. In addition, there is a variable relationship between CBV and CBF depending on cerebral perfusion pressure and the effectiveness of autoregulation. If CBF decreases sufficiently to begin impacting oxygen supply, CBV may increase to compensate. A similar phenomenon has been seen in other mechanisms of brain injury.45,46 Thus there are many factors that must be considered and explored in future work. Cerebral blood volume is determined using HbT measured with our FDNIRS system and HGB. To determine if HGB changes may be driving the increase in CBV, we first provided values of HbO, HbR, and HbT in Supplementary Figure 1. We confirmed HbT to be significantly higher in neonates with HIE during TH (P ¼ 0.01) compared with post-TH values. However, HbT in HIE–TH was not significantly higher than in Control group. We further discuss the & 2014 ISCBFM

accuracy of HGB measurements in the next section presenting the modeling assumptions. The clinical significance and etiology of increased CBV remain unclear. The role of medications and the potential of inadequate cerebral perfusion leading to vasodilatation needs to be explored in neonates with HIE undergoing TH. This observation raises questions about the optimal CBF and CBV during therapy.47 Cerebral Hemoglobin Oxygen Saturation Near infrared spectroscopy-measured rSO2 have been used to monitor neonates with HIE before the clinical standardization of TH. Results among the literature are variable. In a previous study, we measured higher SO2 values in neonates with HIE within 5 days of symptom onset compared with neonates without suspicion of brain injury; however, the difference was not significant.10 In a time window between 24 and 48 hours of age in neonates with HIE, rSO2 was significantly higher in neonates with adverse outcomes as compared with neonates with good outcomes.8 However, this difference was not significant between 6 and 18 hours of age. In a recent study of neonates with HIE who underwent TH, rSO2 was reported as slightly higher in HIE with adverse outcomes at 6, 12, and 24 hours of life compared with neonates with good outcomes; however the difference was only significant at 12 hours.48 In the current study, SO2 in neonates with HIE was higher than post-TH, but the difference was not significant. With an elevation of SO2 occurring both with and without TH, SO2 alone appears to be less sensitive to the therapy and metabolic disturbances occurring in the first few days after birth. Therefore, FDNIRS–DCS parameters presented in this study have potential to compliment SO2 measures to provide a more complete picture of the hemodynamic and metabolic changes in neonates with HIE treated with TH. Medication Effects The primary role of analgesics and sedatives is to prevent secondary brain injury during the treatment by maintaining an adequate mean arterial pressure, which in turn will influence CBF and volume to lower cerebral metabolic demand.49 For example, barbiturates (Phenobarbital) decrease CMRO2i and benzodiazepines (midazolam or lorazepam) are commonly used as a sedative in ventilated pediatric patients. They have minimal adverse cardiovascular effects. Opioids (fentanyl, morphine, or methadone) produce analgesia and comfort.50 In rhesus monkeys, propofol anesthesia concurrent with TH did not impair the coupling between cerebral metabolism and blood flow.51 Neurotransmitters are also commonly used in neonatal population for their action on the cardiovascular system. For example, Dopamine is known to increase systemic and regional CBF as well as blood pressure.52 Cardiovascular effects of Dopamine in piglets were maintained during mild hypothermia and were similar to responses in normothermia.53 Given the wide range and inconsistent use of medications in the neonates presented here (see Supplementary Table 2), we believe that medication is less likely the dominant factor causing the low oxygen metabolism we observed. However, future studies are necessary to further assess the effects of medications on FDNIRS–DCS measures and to better understand flow–volume–metabolism coupling. Modeling Assumptions and Statistical Analysis Factors that could contribute to CMRO2i errors include shifts in arterial:venous volume ratios during the therapy. For example, as CBFi decreases and CBV increases during TH, there could be a relative increase in the venous compartment from 75% to 85%. Recalculating CMRO2i during TH with 85% venous content further reduces CMRO2i by 12% and therefore CMRO2i remains significantly lower than controls (Po10 8). However, the 1/4:3/4 Journal of Cerebral Blood Flow & Metabolism (2014), 87 – 94


92 (arterial:venous) ratio has not been validated for in vivo circulation in neonates and TH could also lead to a decrease in the venous content. Assuming a fixed ratio of 1/3:2/3 (arterial:venous) leads to an increase of 12% in CMRO2i but CMRO2i remains significantly lower compared with controls (Po10 5). Thus our major findings remain unchanged. Errors in CMRO2i propagate with respect to all physiologic variables present in its modeling (equation (1)). Assuming a coefficient of variation of 13%, 2.3%, 5.5%, and 4% for CBFi,11 HGB,54 SaO2,55 and SO2,11,18 respectively, variation in CMRO2i was estimated to be 19.9%. Error of 12% in rCMRO2 has been calculated in rats during focal ischemia using a similar instrumentation and was particularly sensitive to the error in relative oxygen extraction fraction.22 The major source of error in our method is likely due to the variability of CBFi. Furthermore, the fitting process used to derive CBFi requires the knowledge of the average reduced scattering coefficient of the cerebral tissue, which is quantified with the FDNIRS system. We verified that the reduced scattering coefficient was not statistically different between neonates with HIE treated with TH and healthy controls, thus differences in CBFi between the groups are due to differences in CBF and not confounded by differences in scattering. In this study, we used the average scattering coefficient for each population as the input for the fitting process. As a result, future research is aimed at improving the accuracy of CBFi measurements. The calculation of both CBV and CMRO2i necessitates knowledge of HGB values. We found consistently lower HGB in neonates with HIE than normal clinical table values and believe this may be due to the high number of blood draws and intravenous fluids these neonates receive, as well as pre-existing anemia in some HIE newborns. To determine if elevated CBV and decreased CMRO2i in HIE–TH compared with healthy controls were solely due to HGB values in Control we performed an additional analysis using HGB lower (13.5 g/dl) and upper (21.5 g/dl) limits from the reference table values.20 We confirmed the persistence of a statistically significant difference between HIE–TH and Control groups for CMRO2i (see Supplementary Figure 4(a)). It is only when HGB in Control is dropped below the 5th percentile that the increase in CBV in HIE–TH compared with Control increase is not significant (see Supplementary Figure 4(b)). Our results may lose significance only if our normal neonates have very low HGB (below the 5th percentile). As blood is not drawn on healthy neonates, in future a direct measurement of HGB using non-invasive techniques would help improve the accuracy of CBV and CMRO2i measures. In the statistical analysis, the selection of the two sub-groups about the age at measurement is justified by prior studies showing significant decreases in cerebral rSO2 in 26 healthy control neonates during the first 5 days of age.56 This result is consistent with our data: SO2 in control neonates from birth to 2 days old is significantly higher than SO2 from 48 to 96 hours of age (Po0.05). We identified a similar relationship with CBV in controls (Po10 3). However, control CBFi and CMRO2i data did not show a significant difference between the two age sub-groups (P40.1 and P ¼ 0.06, respectively). Future Studies HIE is associated with a complex cascade of events including vascular and metabolic changes.2 This evolution depends on the severity and timing of the hypoxic ischemic insult. In this pilot study, the aim was to compare hemodynamic and metabolic variables in neonatal HIE during and following the therapy. The addition of time-points during the first week after birth would better characterize individual injuries and their evolution. For example, continuous monitoring of FDNIRS–DCS parameters during the therapy would provide more details of the time course of changes.8 Furthermore, with the primary goal of TH to reduce Journal of Cerebral Blood Flow & Metabolism (2014), 87 – 94

the extent of brain injury,17 bedside measurement of hemodynamic and metabolic variables might be useful concurrent measures with electroencephalography or amplitude-integrated electroencephalography prior the therapy to aid in the decision to initiate TH in neonates with HIE. However, this clinical translation would be facilitated with a simplified device and standardization of the measured parameters. The role of animal studies in providing pre-, intra- and postmeasurements with and without TH might also help to refine and optimize inclusion criteria and TH parameters.57 The rate of normal outcome (56.6%) from the current study was similar to the normal rate reported in literature.17 However, the rate of mildly abnormal outcome (33.3%) was higher compared with literature and the rate of abnormal outcome (1 death, 11.1%) was lower than reported trials.5,17 This discrepancy is most likely due to the small sample size included in this particular study, as the overall outcome of neonatal HIE treated with TH at our institution is similar to published data. Due to the low number of patients, scores can only be used to describe individuals but predictive modeling will be implemented for a planned larger cohort. CONCLUSION In summary, we have demonstrated that human term neonates with HIE who meet inclusion criteria for TH have low CMRO2i and CBFi while undergoing TH when compared with post-TH and healthy age-matched controls. In contrast, CBV in neonates with HIE treated with TH was significantly higher than post-TH and Control values whereas SO2 was only significantly higher in HIE–TH compared with Control group. We believe that these hemodynamic and metabolic changes observed during TH are of multi-factorial nature including factors related to mechanisms associated with TH, hypoxia-ischemia, timing of measurements and potentially, medications administered. In conclusion, FDNIRS– DCS parameters have the potential to serve as bedside biomarkers for monitoring individual response and optimization of TH but further studies are needed to explore the effects of medications on individual responses as well as long-term outcome. DISCLOSURE/CONFLICT OF INTEREST Maria Angela Franceschini hold patents on the technology developed in this article. The other authors declare no conflict of interest.

ACKNOWLEDGMENTS The authors thank all the families and the nurses, physicians, and staff in the NICU, the Special Care Nursery, Pediatric Neurology, and the maternity units at Brigham and Women’s Hospital and Boston Children’s Hospital. In particular, the authors thank Breda Hayes, Deirdre O’Reilly, Marcia Kocienski-Filip, Maddy Artunduaga, Linda J Van Marter, Robert M Insoft, Jonathan H Cronin, Julianne Mazzawi, and Steven A Ringer. The authors would also like to thank Barry D Kussman, David A Boas, Stefan A Carp, Erin M Buckley and Bruce R Rosen for their discussion and support, and Marie Drottar for reading through the manuscript.

REFERENCES 1 Wyatt JS, Gluckman PD, Liu PY, Azzopardi D, Ballard R, Edwards AD et al. Determinants of outcomes after head cooling for neonatal encephalopathy. Pediatrics 2007; 119: 912–921. 2 Ferriero DM. Neonatal brain injury. N Engl J Med 2004; 351: 1985–1995. 3 Lampe JW, Becker LB. State of the art in therapeutic hypothermia. Annu Rev Med 2011; 62: 79–93. 4 Erecinska M, Thoresen M, Silver IA. Effects of hypothermia on energy metabolism in mammalian central nervous system. J Cereb Blood Flow Metab 2003; 23: 513–530. 5 Shankaran S, Pappas A, McDonald SA, Vohr BR, Hintz SR, Yolton K et al. Childhood outcomes after hypothermia for neonatal encephalopathy. N Engl J Med 2012; 366: 2085–2092.

& 2014 ISCBFM


93 6 Gancia P, Pomero G. Therapeutic hypothermia in the prevention of hypoxicischaemic encephalopathy: new categories to be enrolled. J Matern Fetal Neonatal Med 2012; 25(Suppl 4): 94–96. 7 Jacobs SE, Berg M, Hunt R, Tarnow-Mordi WO, Inder TE, Davis PG. Cooling for newborns with hypoxic ischaemic encephalopathy. Cochrane Database Syst Rev 2013; 1: CD003311. 8 Toet MC, Lemmers PMA, van Schelven LJ, van Bel F. Cerebral oxygenation and electrical activity after birth asphyxia: their relation to outcome. Pediatrics 2006; 117: 333–339. 9 Franceschini MA, Thaker S, Themelis G, Krishnamoorthy KK, Bortfeld H, Diamond SG et al. Assessment of infant brain development with frequency-domain nearinfrared spectroscopy. Pediatr Res 2007; 61(5 Pt 1): 546–551. 10 Grant PE, Roche-Labarbe N, Surova A, Themelis G, Selb J, Warren EK et al. Increased cerebral blood volume and oxygen consumption in neonatal brain injury. J Cereb Blood Flow Metab 2009; 29: 1704–1713. 11 Roche-Labarbe N, Carp SA, Surova A, Patel M, Boas DA, Grant PE et al. Noninvasive optical measures of CBV, StO(2), CBF index, and rCMRO(2) in human premature neonates’ brains in the first six weeks of life. Hum Brain Mapp 2010; 31: 341–352. 12 Roche-Labarbe N, Fenoglio A, Aggarwal A, Dehaes M, Carp SA, Franceschini MA et al. Near-infrared spectroscopy assessment of cerebral oxygen metabolism in the developing premature brain. J Cereb Blood Flow Metab 2012; 32: 481–488. 13 Lin PY, Roche-Labarbe N, Dehaes M, Fenoglio A, Grant PE, Franceschini MA. Regional and hemispheric asymmetries of cerebral hemodynamic and oxygen metabolism in newborns. Cereb Cortex 2013; 23: 339–348. 14 Higgins RD, Raju TNK, Perlman J, Azzopardi DV, Blackmon LR, Clark RH et al. Hypothermia and perinatal asphyxia: executive summary of the National Institute of Child Health and Human Development workshop. J Pediatr 2006; 148: 170–175. 15 Palisano R, Rosenbaum P, Walter S, Russell D, Wood E, Galuppi B. Development and reliability of a system to classify gross motor function in children with cerebral palsy. Dev Med Child Neurol 1997; 39: 214–223. 16 Bayley N. Bayley Scales of Infant Development. 2nd edn. Psychological Corp: San Antonio, TX, 1993. 17 Shankaran S, Laptook AR, Ehrenkranz RA, Tyson JE, McDonald SA, Donovan EF et al. Whole-body hypotherma for neonates with hypoxic-ischemic encephalopathy. N Engl J Med 2005; 353: 1574–1584. 18 Dehaes M, Grant PE, Sliva DD, Roche-Labarbe N, Pienaar R, Boas DA et al. Assessment of the frequency-domain multi-distance method to evaluate the brain optical properties: Monte Carlo simulations from neonate to adult. Biomed Opt Express 2011; 2: 552–567. 19 Poets CF, Stebbens VA, Lang JA, O’Brien LM, Boon AW, Southall DP. Arterial oxygen saturation in healthy term neonates. Eur J Pediatr 1996; 155: 219–223. 20 Jopling J, Henry E, Wiedmeier SE, Christensen RD. Reference ranges for hematocrit and blood hemoglobin concentration during the neonatal period: data from a multihospital health care system. Pediatrics 2009; 123: e333–e337. 21 Cheung C, Culver JP, Takahashi K, Greenberg JH, Yodh AG. In vivo cerebrovascular measurement combining diffuse near-infrared absorption and correlation spectroscopies. Phys Med Biol 2001; 46: 2053–2065. 22 Culver JP, Durduran T, Furuya D, Cheung C, Greenberg JH, Yodh AG. Diffuse optical tomography of cerebral blood flow, oxygenation, and metabolism in rat during focal ischemia. J Cereb Blood Flow Metab 2003; 23: 911–924. 23 Watzman HM, Kurth CD, Montenegro LM, Rome J, Steven JM, Nicolson SC. Arterial and venous contributions to near-infrared cerebral oximetry. Anesthesiology 2000; 93: 947–953. 24 Grubb RL, Raichle ME, Eichling JO, Ter-Pogossian MM. The effects of changes in PaCO2 on cerebral blood volume, blood flow, and vascular mean transit time. Stroke 1974; 5: 630–639. 25 Pienaar R, Paldino MJ, Madan N, Krishnamoorthy KS, Alsop DC, Dehaes M et al. A quantitative method for correlating observations of decreased apparent diffusion coefficient with elevated cerebral blood perfusion in newborns presenting cerebral ischemic insults. NeuroImage 2012; 63: 1510–1518. 26 Wintermark P, Hansen A, Gregas MC, Soul J, Labrecque M, Robertson RL et al. Brain perfusion in asphyxiated newborns treated with therapeutic hypothermia. AM J Neuroradiol 2011; 32: 2023–2029. 27 Bolar D, Rosen B, Sorensen A, Adalsteinsson E. Quantitative Imaging of Extraction of Oxygen and Tissue Consumption (QUIXOTIC) using Velocity Selective Spin Labeling. Proc Intl Soc Mag Reson Med: Hawaii, 2009, p 6050. 28 Davis TL, Kwong KK, Weisskoff RM, Rosen BR. Calibrated functional MRI: mapping the dynamics of oxidative metabolism. Proc Natl Acad Sci USA 1998; 95: 1834–1839. 29 Fan AP, Benner T, Bolar DS, Rosen BR, Adalsteinsson E. Phase-based regional oxygen metabolism (PROM) using MRI. Magn Reson Med 2012; 67: 669–678.

& 2014 ISCBFM

30 Jain V, Langham MC, Wehrli FW. MRI estimation of global brain oxygen consumption rate. J Cereb Blood Flow Metab 2010; 30: 1598–1607. 31 Xu F, Ge Y, Lu H. Noninvasive quantification of wholebrain cerebral metabolic rate of oxygen (CMRO2) by MRI. Magn Reson Med 2009; 62: 141–148. 32 Liu P, Huang H, Rollins N, Jeon T, Chalak L, Lu H. Quantitative Assessment of Cerebral Metabolism Rate of Oxygen in Neonates. Proc Intl Soc Mag Reson Med: Salt Lake City, 2013, p 0175. 33 Krishnamurthy LC, Liu P, Xu F, Uh J, Dimitrov I, Lu H. Dependence of blood T2 on oxygenation at 7 T: in vitro calibration and in vivo application. Magn Reson Med 2013; doi:10.1002/mrm.24868. 34 Lu H, Xu F, Grgac K, Liu P, Qin Q, van Zijl P. Calibration and validation of TRUST MRI for the estimation of cerebral blood oxygenation. Magn Reson Med 2011; 67: 42–49. 35 Krishnamurthy LC, Liu P, Ge Y, Lu H. Vessel-specific quantification of blood oxygenation with T 2-relaxation-under-phase-contrast MRI. Magn Reson Med 2013; doi:10.1002/mrm.24750. 36 Altman DI, Perlman JM, Volpe JJ, Powers WJ. Cerebral oxygen metabolism in newborns. Pediatrics 1993; 92: 99–104. 37 Greeley W, Kern F, Ungerleider R, Boyd III J, Quill T, Smith L et al. The effect of hypothermic cardiopulmonary bypass and total circulatory arrest on cerebral metabolism in neonates, infants, and children. J Thorac Cardiovasc Surg 1991; 101: 783–794. 38 Durduran T, Zhou C, Buckley EM, Kim MN, Yu G, Choe R et al. Optical measurement of cerebral hemodynamics and oxygen metabolism in neonates with congenital heart defects. J Biomed Opt 2010; 15: 037004. 39 Buckley EM, Hance D, Pawlowski T, Lynch J, Wilson FB, Mesquita RC et al. Validation of diffuse correlation spectroscopic measurement of cerebral blood flow using phase-encoded velocity mapping magnetic resonance imaging. J Biomed Opt 2012; 17: 037007. 40 Kim MN, Durduran T, Frangos S, Edlow BL, Buckley EM, Moss HE et al. Noninvasive measurement of cerebral blood flow and blood oxygenation using near-infrared and diffuse correlation spectroscopies in critically brain-injured adults. Neurocrit Care 2010; 12: 173–180. 41 Ancora G, Maranella E, Locatelli C, Pierantoni L, Faldella G. Changes in cerebral hemodynamics and amplitude integrated EEG in an asphyxiated newborn during and after cool cap treatment. Brain Dev 2009; 31: 442–444. 42 van Bel F, Zeeuwe PE, Dorrepaal CA, Benders MJ, van de Bor M, Hardjowijono R. Changes in cerebral hemodynamics and oxygenation during hypothermic cardiopulmonary bypass in neonates and infants. Biol Neonate 1996; 70: 141–154. 43 Naulaers G, Cossey V, Morren G, Van Huffel S, Casaer P, Devlieger H. Continuous measurement of cerebral blood volume and oxygenation during rewarming of neonates. Acta Paediatr 2004; 93: 1540–1542. 44 Luo F, Schmidt KF, Fox GB, Ferris CF. Differential responses in CBF and CBV to cocaine as measured by fMRI: implications for pharmacological MRI signals derived oxygen metabolism assessment. J Psychiatr Res 2009; 43: 1018–1024. 45 Bouma GJ, Muizelaar JP. Cerebral blood flow, cerebral blood volume, and cerebrovascular reactivity after severe head injury. J Neurotrauma 1992; 9(Suppl 1): S333–S348. 46 Cahill WJ, Calvert JH, Zhang JH. Mechanisms of early brain injury after subarachnoid hemorrhage. J Cereb Blood Flow Metab 2006; 26: 1341–1353. 47 Liem KD, Greisen G. Monitoring of cerebral haemodynamics in newborn infants. Early Hum Dev 2010; 86: 155–158. 48 Ancora G, Maranella E, Grandi S, Sbravati F, Coccolini E, Savini S et al. Early predictors of short term neurodevelopmental outcome in asphyxiated cooled infants. A combined brain amplitude integrated electroencephalography and near infrared spectroscopy study. Brain Dev 2013; 35: 26–31. 49 Rhoney DH, Parker D. Use of sedative and analgesic agents in neurotrauma patients: effects on cerebral physiology. Neurol Res 2001; 23: 237–259. 50 Brinker D. Sedation and comfort issues in the ventilated infant and child. Crit Care Nurs Clin N Am 2004; 16: 365–377. 51 Ouchi T, Ochiai R, Takeda J, Tsukada H, Kakiuchi T. Combined effects of propofol and mild hypothermia on cerebral metabolism and blood flow in rhesus monkey: a positron emission tomography study. J Anesth 2006; 20: 208–214. 52 Nachar RA, Booth EA, Friedlich P, Borzage M, Soleymani S, Wider MD et al. Dose-dependent hemodynamic and metabolic effects of vasoactive medications in normotensive, anesthetized neonatal piglets. Pediatr Res 2011; 70: 473–479. 53 Filseth OM, How O-J, Kondratiev T, Gamst TM, Sager G, Tveita T. Changes in cardiovascular effects of dopamine in response to graded hypothermia in vivo. Critical Care Medicine 2012; 40: 178–186.



94 54 Bourner G, Dhaliwal J, Sumner J. Performance evaluation of the latest fully automated hematology analyzers in a large, commercial laboratory setting: a 4-way, side-by-side study. Lab Hematol 2005; 11: 285–297. 55 Schmitt HJ, Schuetz WH, Proeschel PA, Jaklin C. Accuracy of pulse oximetry in children with cyanotic congenital heart disease. J Cardiothor Vasc An 1993; 7: 61–65.

56 Bernal NP, Hoffman GM, Ghanayem NS, Arca MJ. Cerebral and somatic nearinfrared spectroscopy in normal newborns. J Pediatr Surg 2010; 45: 1306–1310. 57 Cheng G, Sun J, Wang L, Shao X, Zhou W. Effects of selective head cooling on cerebral blood flow and metabolism in newborn piglets after hypoxia-ischemia. Early Hum Dev 2011; 87: 109–114.

Supplementary Information accompanies the paper on the Journal of Cerebral Blood Flow & Metabolism website (http:// www.nature.com/jcbfm)


& 2014 ISCBFM