USE OF NEAR-INFRARED SPECTROSCOPY IN THE NEONATAL INTENSIVE CARE UNIT. G. Naulaers

USE OF NEAR-INFRARED SPECTROSCOPY IN THE NEONATAL INTENSIVE CARE UNIT.

G. Naulaers

, A. Caicedo

and S. Van Huffel

.

1

Neonatal Intensive Care Unit. University Hospitals Leuven.

2

ESAT-SCD, Department of Electrical Engineering, Katholieke Universiteit Leuven, Leuven, Belgium

3

IBBT-K.U.Leuven Future Health Department, Leuven, Belgium

1. ABSTRACT.

Near-infrared spectroscopy was first described in 1977 as a non-invasive technique to measure the cerebral oxygenation and cytochrome oxydase. Different techniques have been developed resulting in new instruments that make it possible to measure cerebral oxygenation in a non-invasive way. In this chapter the physiology and pathophysiology in relation to the measurement of cerebral oxygenation are explained and the direct possible clinical use enlightened, with special focus on measurement of ischemic cerebral hypoxia. The measurement of other organs like the liver, the bowel and the peripheral circulation are described. At the end, a short overview of future possible bed-side measurements like functional near-infrared spectroscopy, near-infrared imaging and photoacoustic measurements is given.

2. INTRODUCTION.

In this chapter an overview will be given regarding the use of near-infrared spectroscopy in

the neonatal intensive care unit. The focus will be on the use of direct cerebral monitoring and

the topics regarding functional near-infrared spectroscopy and imaging will not be discussed

in detail.

Near infrared spectroscopy (NIRS) is for the first time described for medical use in 1977(1) by Jobsis. In this paper it was used for the measurement of cytochrome oxydase in myocardial insufficiency. The use in neonates to measure cerebral oxygenation was first reported by Brazy et al (2) and Delpy (3). Since than near-infrared spectroscopy started to become more popular for research in neonates, however it remained a difficult task to acquire stable measurements. This was mainly due to the instrumentation that was very sensible for movement artefacts and also because of the applied technique (Beer-Lambert Law) where absolute values are not measured. Different studies in physiology, pathophysiology and the use of medication where performed during the 80’s en 90’s in research settings. It was the introduction of spatially resolved spectroscopy by Matcher and Suzuki (4;5) that made it possible to measure absolute values with spatially resolved spectroscopy. Because this method is less movement-sensitive, easier to use and reporting absolute values it leaded to commercial monitors that are increasingly used in cardiac surgery and also in neonatal intensive care units. The objectives of this chapter are to describe the technique of near- infrared spectroscopy, the use of NIRS in the neonatal intensive care unit for research and the future possibility to use this monitoring for clinical purposes.

3. TECHNIQUE OF NEAR-INFRARED SPECTROSCOPY

The relative transparency of biological tissue to light in the near infrared part of the spectrum (700-1000 nm) enables photons to be detected following passage through tissue at distances of up to 8 cm. As the light passes through the head it is attenuated due to a combination of absorption and scattering. Brain contains three changing chromophores that are present in variable concentrations: oxyhaemoglobin, deoxyhaemoglobin and cytochrome oxidase.

Attenuation of transmitted light in the brain due to other causes (muscle, bone, bilirubin, ...) will be assumed to be constant over the period of monitoring. Consequently any change in observed attenuation is due to a change in the concentration of these chromophores. Because of safe levels of light input, it is difficult to use source-detectors distances of more than 5 cm (6) and because of scattering and acquiring enough depth of measurement the minimum interoptode distance will probably be around 1,5 to 2 cm (7). The depth measured with this method is around half of the source-detector distance, so between 1 and 2,5 cm (8). Different techniques are used to measure the cerebral oxygenation.

3.1 The continuous wave single distance spectrometers : the use of the Beer Lambert Law.

For accurate quantification, the distance travelled by photons through the tissue (the optical pathlength) must be known and remain constant. The distance between the emitting and receiving optodes is known, but light does not travel in straight lines. Instead, photons are scattered by tissue components and diffused through the brain. The Modified Lambert-Beer Law allows to convert changes in absorption and attenuation in changes in concentration of the chromophore (9).

A = Log I

/I =  . c . d . B + G

A is the attenuation measured in optical density. I

is the light intensity incident on the medium and given by the laser or LED light of the spectroscope. I is the light intensity transmitted through the medium and measured by the spectroscope.  is the specific extinction coefficient of the absorbing compound measured in µmolar

. cm

and this is well known and shown in figure 1. C is the concentration of the absorbing compound (HbO

, HbR or cytO) in the solution measured in µ molar. d is the distance between the points where the light enters and leaves the solution measured in cm. B is the differential pathlength factor.

Because light does not travel through tissue in a straight way, d (the geometrical pathlength) has to be multiplied by B to find the differential pathlength which is the true optical distance.

G is the additive term reflecting scattering loss. Part of the light will be scattered, so that the light measured is not only changed due to absorption but also because of scattering. As we cannot measure the scattering, G remains an unknown factor and no absolute values can be measured.

Figure 1.

The only values that can be measured are differential values comparing one moment with another (time resolved spectroscopy). This will give us the following equations:

A

– A

= (c

– c

) . . d. B

Because the specific extinction coefficient is different for the three chromophores the different instruments will use different laser diodes emitting different wavelengths (e.g. at 775, 810, 850 and 910 nm) to calculate the proportional concentration of the chromophores.

The amount of wavelengths used will have an influence on the precision of the measurement.

The majority of the instruments cannot measure the different pathlength (B) and is therefore

given as a constant parameter to the computer. The values for different ages have been published by using models of post mortem measurements with flight-of-time measurements (10;11).

Consequently the instruments using this technology can only measure relative values and will also be more prone to movement-artifacts and difficult to use for long measurements. This explains why they were only used in research settings.

Two adaptations made it possible to record absolute values with this technique. Some instruments use wavelength resolved spectroscopy combined with calibration in vitro or in vivo (babies during cardiac operation or animals). A second possibility is to measure the pathlength factor by intensity modulation or time-of flight. However in the last case, the problem of scattering is still not solved.

The three chromophores that can be measured with this method are HbO

(oxygenated haemoglobin), HbR (deoxygenated haemoglobin) and cytochrome oxydase. These measurements can lead to different parameters.

HbT (total haemoglobin) is the sum of HbO

and HbR and gives an estimation of the total blood volume in the brain.

Two different ways are described to measure cerebral blood flow by means of near-infrared spectroscopy. Both methods are based on the original work of Kety et al, who described the clinical application of the Fick principle for the first time in 1948 using the nitrous oxide uptake technique (12). The first technique described with NIRS was to use oxygen as a “dye”.

The Fick principle states that the amount of a substance (oxygen in this method) taken up by an organ per unit of time is equal to the arterial concentration of the substance minus the venous concentration times the blood flow:

Q(t) : F x (Ca-Cv) dt

Where Q(t) is the amount of the tracer in tissue, F is flow and Ca and Cv are the

concentrations of the tracer in arterial and venous blood, respectively. When the time, t, is less

than the transit time of the organ, the tracer will not have reached the venous side. The flow

(F) can be measured as the ratio between the amount of tracer accumulated in the brain and

the amount of tracer introduced during the period 0-t. With NIRS it is possible to measure

changes in oxygenated and deoxygenated Hb concentrations as well as changes in total

cerebral Hb concentrations. SaO

is measured by a pulse oximeter placed on the right hand of the infant. After a sudden and short increase in the concentration of inspired oxygen, the increase in cerebral oxyhaemoglobin represents the accumulation of tracer. The product of the integral of change in SaO

with respect to time and the arterial Hb concentration (aHb) is the arterial tracer concentration and therefore,

CBF = OI/aHb x Lto (SaO

) dt

Where OI (HbO

– Hb/2) is expressed in mmol/100 g brain wt, aHb in mmol/ml tetraheme and Lto (SaO

) dt in minutes.

The most important problem with this method is that is demands a rise in SaO

by means of increasing FiO

. This makes it very difficult to use it in infants with a normal lung function, infants with congenital heart disease and infants with pulmonary hypertension, because a rise in SaO

is very difficult to provoke in these patients. A second problem is the question whether a sudden increase in FiO

does not alter the cerebral blood flow within a few seconds.

The third problem is an ethical issue, whether it is right to increase the FiO

suddenly in very prematurely born infant, known to be at augmented risk for retinopathy of prematurity.

Therefore a second method was developed, using indocyanine green as a tracer. Indocyanine green is a dye that strongly absorbs near-infrared light and is rapidly cleared from the body. It has been used for many years in children and adults without toxic side effects. Injecting ICG and measuring with near infrared spectroscopy the changes in absorption of ICG can measure cerebral blood flow. Validation studies for the FiO2 method (13-15) and the indocyanine green method (16-18) were published.

CBV (cerebral blood volume) can also be measured with this method using a change in oxygen. The change in total haemoglobin can be easily calculated as HbT = HbO

+ Hb and this is often used in studies as one of the parameters to measure the effect of CO

or medication. Wyatt et al described a technique to measure the absolute value of CBV (19). In this measurement, a small but slow change in SaO

is monitored. If the assumption is taken that oxygen consumption and CBV do not change with this small change of SaO

, than the oxygen saturation in each compartment will change by the same amount of SaO

. The

HbO

will than be the equivalent of the change to the product of the total haemoglobin

concentration (tcHb) and the change in fractional SaO

. As described by Elwell et al the

following equations can be calculated (12) :

HbO

= (tcHb) x SaO

The assumption that CBV is constant during the measurement period means that the changes in HbO

and Hb concentration must also be equal and opposite. Hence:

Hb = -(tcHb) x SaO

Combining these two equations will give the following formula:

HbO

- Hb (tcHb) (µmolar) = ---

2x SaO

This can be converted from µmol of haemoglobin per litre of brain tissue into blood volume in ml of blood per 100 g of brain tissue.

Different validation studies found a good correlation (19;20), although other studies did not confirm this (21;22). These measurements are thus usable, although most new instruments do not give the values of HbO

and HbT anymore.

We will not go into the use of HbO

and HbD (HbO

-HbR) as parameters of oxygenation, as they are increasingly been replaced by absolute values for cerebral oxygenation (TOI, rSO

).

We will also not go into an extended discussion about the use of cytochrome aa

(also called

cytochrome oxidase) as a parameter for cerebral oxygenation. Cytaa

indicate changes in the

oxidation-reduction level of the intracerebral enzyme cytochrome oxidase, the terminal

member of the mitochondrial respiratory chain. It can therefore be used as a relative measure

of cellular oxygenation. Mitochondrial cytochrome oxidase is responsible for > 95% of the

oxygen consumption in the body and is essential for the efficient generation of cellular ATP

(23). The enzyme contains four redox-active metal centres and one of these, the Cu

centre,

has a strong absorbance in the near-infrared, which enables it to be detectable by near-infrared

spectroscopy (24). Cytochrome oxidase is known to have a very low level of reduction in

vivo, which means that infants need to undergo clinically very significant saturation changes

to document measurable changes in cytochrome oxidase (25;26). Cytochrome oxidase is

difficult to measure and the values are not available anymore on the new monitors at this

moment. However, future research might enlighten new possibilities for cytochrome oxydase

measurements in intensive care (27).

By using different wavelengths in combination with in vitro or in vivo calibration, it is possible to measure an absolute value : HbO2/HbT reflecting the cerebral oxygenation.

Another way to measure absolute values is spatially resolved spectroscopy.

3.2 Spatially resolved spectroscopy.

Spatially resolved spectroscopy measures the absorption of light at two or more detectors. By using the diffusion equation (4;5) absolute values can be calculated assuming that the scattering for the different distances is constant. Another method is to distract the measurement of the closest detector from the measurement of the further detector to omit the influence of superficial tissue but in these instruments further calibration in vitro or in vivo is still necessary.

Figure 2.

There are different commercial instruments for measurement of cerebral oxygenation on the market. The most used monitors are the NIRO series (Hamamatsu, Phototonics KK,

Hamamatsu city, Japan) measuring TOI (Tissue Oxygenation Index), the INVOS series (somanetics Corporation, Troy, MI, USA) measuring rSO2, FORE-SIGHT (CAS Medical Systems, Branford, CT, USA), Nonin 7600 (Nonin Medical, Plymouth, MN, USA) and frequency domain oxymeters (ISS Inc., Champaign, IL, USA). However they all use different principles of measurement, different wavelengths, different optode distances and different light emitters (laser of LED). It is thus of major importance to know how the instrument works and whether validation studies comparing with jugular venous saturation measurement and with other commercial instruments have been performed.

4. THE MEASUREMENT OF CEREBRAL OXYGENATION IN NEONATOLOGY.

4.1 Validation studies for TOI and rSO

and precision of the instruments

For the validation mostly correlation with the jugular venous saturation is published (28-32),

however this is a correlation, indicating that changes in the cerebral oxygenation reflect

changes in the jugular venous saturation. Studies comparing different instruments, also show

a good correlation between the instruments, although this was also only a correlation.

The precision of the measurements is also an important issue. Different authors showed an important change when the optodes where replaced (33-35).

4.2 Normal values : what is too low and what is too high ?

Different authors described normal values in neonates and preterm infants. Weiss et al described a median TOI of 61 ± 12 % in full term neonates and infants (28). We described an increase of TOI over three days. The median TOI on the first day was 57% (95% CI: 54- 65.7), 66.1% on the second day (95%CI: 61.9-82.3%) and 76.1% on the third day (95%CI 67.8-80.1% (36). Sorensen et al described a mean TOI of 75% with a standard deviation of 10,2% in 253 patients less than 33 weeks (33). Petrova et al described a mean rSO

of 66%

with a standard deviation of 8,8% and Lemmens et al found a mean rSO

of 70% with a mean of 7,7% (37;38). In conclusion we can say that the normal values reported are between 60 and 72 % with the mean standard deviation between 7-12%. Lemmers et al (39) studied the differences between left and right measurements. The Bland and Altman-determined limits of agreement found overall limits of agreement of -8.5 to +9.5% between left and right for rScO2, however during stable systemic oxygenation (i.e., arterial oxygen saturation SaO2 between 85 and 97%) the limits of agreement between left and right improved from -7.8 to +8.2%. It is obvious that these measurements should rather be used for trend measurements instead of precise oxygenation values.

Another approach for measurement is to alarm for the lowest values. We do not know exactly the lowest values but different studies indicate that a rSO2 or TOI less than 50% for more than 30 minutes might be damaging. Hou et al (40) described mitochondrial damage of CA1 neurons in piglets kept on a cerebral saturation of less than 40% for more than 30 minutes.

When cerebral saturation was less than 30% for more than 30 minutes, morphological damage

of CA1 neurons was described together with a decrease in EEG amplitude and an increased

lactate. Kurth et al (41) cerebral hypoxia-ischemia near-infrared spectroscopy thresholds for

functional impairment are SCO2 33% to 44% in hypoxic piglets. Dent et al (42) described

that a prolonged low postoperative cerebral oximetry (<45% for >180 minutes) was

associated with the development of new or worsened ischemia on postoperative magnetic

resonance imaging. In adults those who employed a cut point equivalent to a 20% decrease of

the baseline value or a value < 50% before intervention to reverse desaturation obtained

promising results. Researchers have reported decreases in neurological complications,

incidence of renal failure, other vital organ complications, length of intensive care unit (ICU)

stay, total hospital stay and surgical costs (43). It has to be taken in mind that very low cerebral oxygen saturations are found in the first six minutes after birth as described by Fauchere et al (44).

Another important issue is whether high values also may be damaging for the brain. Different studies show that the premature infant is especially vulnerable for oxidative stress (45). In very premature infants (46) higher cerebral oxygen saturations are measured than in term infants. Special consideration should be given for patients treated for apnea as recently described by Baerts et al (47) who described very high values after increase of oxygen because of desaturation in case of apneas. The cerebral oxygen monitor can measure much more accurately the changes than the peripheral oxygen saturation monitor.

4.3 Fractional oxygen extraction.

The cerebral tissue oxygen delivery (OD) is the total amount of O2 delivered to the cerebral vessels. It depends on both the arterial oxygen content (CaO2) and the CBF by

OD= CaO2 x CBF

The OD to the brain is thus affected by the CBF and CaO2. In physiologically stable conditions, the OD is in balance with the cerebral metabolic requirements and consumption of the newborn, defined as the cerebral metabolic rate of oxygen (CMRO2), i.e. the total amount of oxygen used by the brain. A CMRO2 less than 1mL/100g/min has been reported in term and preterm newborns with no evidence of brain damage (48). This is rather ‘low’ and below the threshold of brain viability in adults (49). The magnitude of abnormality of cerebral oxidative metabolism has been related to the severity of the adverse outcome at one year of age in postaphyxiated newborns(50). Yoxall et al have described an increase in cerebral oxygen consumption (CvO2) with gestational age and in addition they suggested that the measurement of CvO2 gives more information than the measurement of CBF for the prevention of cerebral hypoxia and ischemia (51).

By the Fick principle, CMRO2 is defined as the product of CBF and the cerebral arteriovenous

difference (CaO2-CvO2):

CMRO2 or VO2= CBF x (CaO2-CvO2)

where CaO2 is the arterial oxygen content and CvO2 the venous oxygen content (mmol O2/ml blood).

CaO2= [(Hb) x (SaO2/100) x 1.39] + (PaO2 x 0.00003) CvO2= [(Hb) x (SvO2/100) x 1.39] + (PvO2 x 0.00003)

Hb is the hemoglobin concentration (g/ml), SaO2 and SvO the percentage arterial and venous hemoglobin saturation, PaO2 and PvO2 the arterial and venous partial pressure of oxygen (mmHg), 1.39 the stoichometric value of oxygen for Hb and 0.00003 the coefficient of oxygen solubility (ml of O2/ ml of blood).

To ensure stable CMRO2, the brain can extract more oxygen from the perfusing blood, or CBF could increase. The brain relies largely on the latter(52). The increase in CBF has the effect of maintaining tissue oxygen delivery (OD); unless it is sufficient, CMRO2 will inevitably fall.

The balance between OD and CMRO2 is reflected by the fractional oxygen extraction, ie the fraction of oxygen removed from the arterial blood, as a fraction of that delivered

Using the Fick principle FOE = VO2/OD = (CaO2-CvO2)/CaO2

This method was first used by Yoxall et al (53), where the cerebral venous saturation was measured using the jugular venous occlusion technique. The technique is well validated and resulted in different studies regarding the cerebral fractional oxygen extraction (51;54-58).

However, the measurement is not continuous and it demands a jugular venous occlusion.

Therefore we searched whether the newer cerebral oxygenation parameters could be used to

measure the fractional oxygen extraction. As peripheral arterial oxygen saturation (SaO

)

reflects CaO

and the cerebral oxygen saturation (TOI or RSO

) reflects CVO2, we can

measure the fractional oxygen saturation using continuous monitoring as followed

FOE = VO2/OD = (CaO2-CvO2)/CaO2 ≈ (SaO

– TOI/ SaO

) = FTOE

A good correlation between FOE and FTOE was described in a validation study in piglets (32)

Since the brains need for oxygen is large and relative constant, the brain is dependent upon a good balance between OD and CMRO2. Changes in SaO2, Hb or CBF however, may disturb this balance and cause cerebral injury. Cerebral hypoxia can thus be caused by low CaO2 (hypoxic hypoxia), low CBF (ischemic hypoxia) and by low Hb (anemic hypoxia).

5. THE USE OF NEAR-INFRARED SPECTROSCOPY IN THE NEONATAL INTENSIVE CARE UNIT.

5.1 Physiological and pathophysiological influences on the cerebral oxygenation.

Figure 3

5.1.1 Hypoxic hypoxia.

Hypoxic hypoxia is hypoxia caused by a decrease in oxygen content (CaO2).

As an adaptive response redistribution of the blood flow occurs to ensure adequate oxygen and substrate supply to vital organs such as the brain. In hypoxic conditions in neonatal lambs, an increase in CBF(59) with constant CMRO2 (60;61) has been reported, to ensure a stable OD and FOE. At the point when CBF reaches its maximum , OD decreases and FOE has to increase to maintain CMRO2. When this mechanism is also exhausted, CMRO2 eventually decreases. The fact that OD is not maintained during hypoxic hypoxia in the immature model suggests that important regulatory mechanisms are not fully developed in the immature brain, making it more vulnerable to hypoxic injury (61).

Hypoxic hypoxia is been well described using near-infrared spectroscopy. Normally a

decrease in cerebral oxygen saturation with a concomitant decrease in arterial oxygen

saturation, resulting in a stable FTOE is found. This is described by Petrova et al (38;62) and

direct postnatally by Fauchere et al (44).

As described above, special consideration should also be given for hyperoxia. No immediate upper limit is given, but values above 85 are probably too high. Special attention should be given to hyperoxygenation after apnea and reanimation.

5.1.2 Anemic hypoxia.

Anemic hypoxia is hypoxia caused by a decrease in haemoglobin (Hb).

At hemoglobin levels <6-7 g/dl in the preterm infant, oxygen delivery may be compromised and blood transfusion is recommended. In case of anemic hypoxia, oxygen delivery is initially maintained by an increase in CBF – even up to a hematocrite range of 20-30% (60;63) – as shown in the lamb (60) as well as in the human newborn (57;64). Assuming constant CMRO2, FOE does not change with moderate anemia. However, below a critical threshold of Hb concentration, the increase in CBF is no longer able to meet the tissue oxygen requirements and FOE will increase. With severe anemia, even a decrease in CMRO2 has been reported because of inadequate tissue oxygenating (65). In neonates no change in fractional oxygen extraction was found above hemoglobine of 8 g/dl. (57), however other reported a risk when hemoglobine fell under 9,7 g/dl (66). Van Hoften et al found an important increase in rSO

and decrease in FTOE after transfusion in patients with haemoglobin below 6.0 mmol/l (9,7 g/dl) (66).

5.1.3 Ischemic hypoxia

Ischemic hypoxia is hypoxia caused by a decrease in CBF (ie ischemia). This will be seen by a decrease in cerebral oxygenation without a decrease in arterial oxygen saturation, resulting in an increase in FTOE.

Cerebral blood flow is determined by different factors. In all species and at all ages there is a

coupling between cerebral blood flow and metabolism. An increase in metabolism will

immediately result in an increase in cerebral blood flow resulting in a stable balance between

oxygen delivery and oxygen consumption. This explains the increase in cerebral blood flow in

the first three days (67) which is correlated with the increase in cerebral metabolism as seen

as the increase in the EEG-voltage and a stable cerebral oxygen extraction as described by

Victor et al (56). It is important to keep in mind that by measuring cerebral blood flow, it has

to be related to the cerebral activity (sleep, awake, sedation). Munger et al found a close

relationship between the sleep state changes and changes in total cerebral haemoglobin and cerebral oxyhaemoglobin, reflecting the changes in blood flow. There was a close relationship between sleep state changes and changes in total cerebral haemoglobin concentration, which increased from active to quiet sleep and decreased from quiet to active sleep. Changes in total cerebral haemoglobin were due, in the most part, to changes in the cerebral oxyhaemoglobin concentration. In conclusion, sleep states influence the cerebral haemoglobin concentration (68).

CO

is an important regulator of the cerebral blood flow as showed in 1948 by Kety et al. (12) Hypocapnia causes a decrease in cerebral blood flow, resulting in a decrease in oxygen delivery. This is measured as a decrease in cerebral oxygen saturation (rSO

or TOI) and an increase in oxygen extraction, which is shown in different NIRS studies (57;69-76). Unless transcutaneous CO

or end-tidal CO

measurements are performed, monitoring the cerebral oxygenation might be one of the most sensitive non-invasive monitoring to warn for hypocapnia.

A second important parameter in premature infants is a patent ductus arteriosus. When there is an hemodynamic important ductus an important left to right shunt can cause a decrease in cerebral blood flow. This results in an important decrease of cerebral oxygenation (77;78). In this way, the cerebral oxymeter can become a screening tool for detecting an important patent ductus (78) and a possible important extra parameter in the decision process for treatment of patent ductus arteriosus.

Hypotension is another parameter causing low cerebral blood flow, however the low range of

blood pressure is dependent on the autoregulation curve (57;79-84). There is still discussion

whether autoregulation is seen in most of the premature infants and how to measure it. (85-

90). We do not know at which blood pressure there is an important decrease in cerebral blood

flow with cerebral ischemia as complication. Tyszczuk et al showed that there were normal

cerebral blood flows with sometimes very low blood pressures (84) and fractional oxygen

extraction remains stable with different mean arterial blood pressures (91). The use of fluids

or inotropics to increase the blood pressure will increase the mean arterial blood pressure and

the left ventricular output, but not necessarily the cerebral blood flow (92). On the other hand

increase in cerebral blood flow and cerebral oxygenation is been described after the infusion

of dopamine or epinephrine (82). Another important aspect of dopamine is that there are

different studies suggesting a restoration of the autoregulation in infants with low blood

pressure after infusion of dopamine (80;93). Further studies are necessary to show the importance of dopamine and to investigate whether NIRS might play a role in the decision to start inotropics in hypotensive preterm infants.

5.1.4 The effect of medication.

NIRS also plays a role in pharmacodynamic research. Many medications used in neonatology are not been tested for the brain. The effect of surfactant (94-100), caffeine (101;102), ibuprofen (103-105), indomethacin (105-109), morphine (110), midazolam (110) and propofol (111) are described. However, as shown in the effects of propofol, this can be dependent of postmenstrual age, postnatal age and the other hemodynamic parameters like blood pressure, PCO

, hematocrit and patent ductus arteriosus. Therefore the daily use of these monitors at the moment that medication is given can give us individual information regarding the individual reaction on the drug in that specific situation.

In contrast to studies in adults, mainly in cardiac surgery and neurosurgery, where a positive effect of the monitoring on the outcome could be proved, we do not have today any study that proves that outcome in neonates improves when cerebral oxygenation monitoring is used.

However, the same critic can be given to other monitoring devices daily used in the unit. We

therefore advice to use the cerebral oxygenation monitoring in combination with the other

parameters like peripheral oxygen saturation, blood pressure, heart rate and carbon dioxide

measurements to get a better understanding of the patient and the treatment given. As hypoxic

hypoxia is mainly quickly detected by a monitoring the peripheral oxygen saturation,

ischemic hypoxia might be missed with conventional monitoring. Therefore the following

scheme could be used when combining the different parameters :

In this way the cerebral oxymeter will be one of the monitors guiding the clinician in its daily work.

5.2. Relation between measurement of cerebral oxygenation and brain lesions.

The main reason to measure the cerebral oxygenation is to decrease the frequency of brain lesions. Kissack et al described a high fractional oxygen extraction (FOE), measured with NIRS, in two infants who developed a periventricular haemorrhagic infarction, while changing FOE’s, decreasing from day 1 tot day 2 and increasing from day 2 to day 3, were found in 9 infants with intraventricular haemorrhage. They did not find a significant change in children with periventricular leucomalacia (112). Verhagen et al described that r(c)SO was lower and FTOE higher in infants with GMH-IVH on Days 1, 2, 3, 4, 5, 8, and 15. The largest difference occurred on Day 5 with r(c)SO median 64% in infants with GMH-IVH versus 77%

in control subjects and FTOE median 0.30 versus 0.17. R(c)SO and FTOE were not affected

by the grade of GMH-IVH (113)

This confirms the findings of Sorensen et al who also

describe a lower TOI in premature infants that later developed an intraventricular

haemorrhage (114).

6. FUTURE DEVELOPMENTS IN THE MEASUREMENT OF THE CEREBRAL OXYGENATION AND CIRCULATION.

6.1 Measurement of the cerebral autoregulation.

Cerebral autoregulation refers to the property of the brain to regulate the CBF in presence of changes in intracranial perfussion pressure (CPP). As changes in CPP are related to changes in MABP, cerebral autoregulation can then be defined as the capacity of the brain to regulate the CBF in presence of changes in MABP. It was first described by Lassen (115) in 1959. In common terms it means that the brain arteries will dilate in presence of a decrease in MABP and will constrict in presence of an increase in MABP. However, the autoregulative properties are limited to some values in MABP. Those limits are imposed by the capacity of the brain arteries to dilate and contract. In adults, the lower and upper limit of mean arterial blood pressures where cerebral autoregulation is active are 60mmHg and 160 mmHg respectively.

In infants those limits are unknown. Beyond those limits the brain lacks autoregulative protection (116).

Although some studies indicated that Cerebral Autoregulation is lost in premature sick infants(117), other studies have found no evidence of a decrease in CBF in presence of low values of MABP (84). Moreover, when comparing lamb studies with postnatal results in human it is suggested that cerebral autoregulation can be developed earlier in the human than in the fetal lamb (118). This indicate that cerebral autoregulation is more complex in premature infants than originally thought (119).

Tsuji et al (120) were the first to report on the use of NIRS to assess cerebral autoregulation, they found a high correlation between the cerebral intravascular oxygenation (dHbD) and mean arterial blood pressure (MABP) changes, indicating loss in autoregulation. Loss in autoregulation was then correlated with clinical outcome, i.e. the frequency of severe intraventricular bleeding. Moreover, Tusji et al (81) validated changes in dHbD as a measure of CBF in an experimental model. However, due to the sensitivity of dHbD to movement artifacts TOI and rScO2 represent better option to assess CBF using NIRS (121;122).

There are several approaches to assess cerebral autoregulation. Linear techniques address the

problem using methods to quantify the linear relation between MABP and CBF, these

techniques are simple and the results are easy to interpret, therefore, they are suited for

clinical use. However, in presence of nonlinearities the results can be misleading. Nonlinear techniques, on the other hand, take into account the possible nonlinearities present in the common dynamics of MABP and CBF, they use physiological models and more complex mathematical methods. Although the results are more difficult to interpret than in the linear case, they can provide a better understanding of the underling processes involved in cerebral autoregulation. Furthermore, there has been a growing interest in the use of multivariable techniques for cerebral autoregulation, where it is seen as the coupled response of several mechanisms involving changes in CBF due to changes in MABP as well as changes in partial pressure of CO

(PCO

), and partial pressure of O

(PO

). However, it is difficult to assess the properties of those mechanism due to their highly coupled dynamics.

All the methods mentioned before can be used to assess static cerebral autoregulation as dynamic cerebral autoregulation. Static cerebral autoregulation refers to the relation between CBF and MABP in a long time interval, while dynamic cerebral autoregulation refers to how the autoregulative mechanism behaves during the first 10 seconds after a change in MABP is presented (123). Different linear methods like correlation, coherence and partial correlation have been used to look at autoregulation (86-88;124-126).

Transfer function analysis derivates from system theory. In this framework the autoregulative properties of the brain are treated as a linear system, where the input is the MABP and the output is the CBF. As measurements of the system input-output are available, the properties of this system can be accessed via identification techniques. In order to interpret the complex values given by the transfer function, its gain and phase are studied. The gain of the transfer function represents the strength of the relationship between the system input and output.

However, in contrast with the correlation and the coherence that only measure the presence of a linear relationship, the gain is proportional to the slope of their linear relation. Hence, a high gain value implies a high variation in the output driven by a small variation in the input, while a small gain implies a low variation in the output driven by a big variation in the input. This property of the gain is of great interest for the analysis of cerebral autoregulation as the autoregulative plateau in the cerebral autoregulation curve is not completely flat but presents a small inclination. This inclination can lead to high values of correlation and coherence but will produce small values in the gain factor.

On the other hand, the phase represents the delay between the input and the output per each

frequency component. This value can be an indication of causality in the system.

The transfer function method allows to include more than one input and one output in the analysis of the system. This is an important property that permits to study the interaction of other variables in the underlying mechanism of cerebral autoregulation.

6.2 Measurement of the neurovascular coupling.

Looking at cerebral oxygenation we take into account that there is normal neurovascular coupling in neonates. However there is not much good evidence of one of both. Roche Labarbe et al looked at the EEG during quiet sleep and found a good correlation between the well known bursts or SATs and the spontaneous waves of HbO2 that are seen in different NIRS measurements. (127). Wong et al described a positive role for dopamine for promoting neurovascular coupling in neonates; (93)

Another model to study neurovascular coupling might be convulsions. The seizure was studied in a neonate and was characterized by a first increase in [HHb] followed by an increase in [HbO(2)] and [HbT]. [HHb] returned to baseline at the end of the seizure and decreased thereafter (128)

To study neurovascular coupling, new instruments measuring EEG and NIRS together have to be developed for use in the neonatal intensive care unit (129)

6.3 Functional near-infrared spectroscopy

Functional near-infrared spectroscopy is possible in neonates although still in research settings. (130;131). Different functions like smell(132;133), taste, hearing(134;135), speech(136), vision(137) and motor cortex (138) can be tested with near-infrared spectroscopy. This gives a lot of future possibilities in research, trying to understand the development of the different functions and for functional testing in the future.

6.4 Near-infrared imaging.

There is also an increasing interest in near-infrared imaging, which would make it possible to image the brain bedside, overting difficult transports to MRI or CT facilities. (139;140).

6.5 Photo-acoustic spectroscopy.

Photo-acoustic methods are starting to be developed for a more accurate imaging. Hereby the Bell-principle is used that light will also cause a energy input producing a oscillatory wave that can be detected with echography. As oscillatory waves are more homogeneous, omitting the scattering of light, this will give more accurate images. (141)

7. THE MEASUREMENT OF THE OXYGENATION OF THE LIVER AND THE GASTRO-INTESTINAL SYSTEM;

Near-infrared spectroscopy measurements of the liver are feasible because it is a homogenous organ that lies directly beneath the skin. The tissue oxygenation index can be measured with spatially resolved spectroscopy in a non-invasive way. When measuring the oxygen saturation of the liver, the arterial as well as the venous oxygen saturation will be measured.

The liver is irrigated by three different sources of blood supply. The portal vein drains the venous blood of the splanchnic system. It receives blood of the splenic, the superior mesenteric, the right and left gastric, the prepyloric, the cystic and the paraumbilical veins.

The superior mesenteric vein drains the jejunal, ileal, ileocolic, right colic and middle colic veins. It is joined by the right gastroepiploic and the pancreaticoduodenal veins to form the portal vein. The hepatic artery and its branches join the branches of the portal vein at the level of the sinusoids and are in this way distributed to the same territory. The hepatic veins convey blood from the liver to the inferior vena cava (142). The global hepatic blood supply is derived for 75% from the portal vein and hence for 25% from the hepatic artery. Low cardiac output or shock reduce dramatically the intestinal perfusion and hence decrease portal vein oxygen saturation and increase oxygen extraction rate in the liver resulting in a fall of liver tissue oxygenation.(143).

The same group described the distribution of the different flows to the liver and its consequences (144). If the hepatic artery was clamped, a decrease in haemoglobin oxygen saturation was seen with a recovery later. This was explained by the “reciprocal relationship”

between the hepatic artery and the portal vein, such that an increase in blood flow through one

circuit leads to an increased inflow resistance in the other, thus helping to maintain a constant

blood flow in the liver. However, when the portal vein was clamped, a much greater decrease

in haemoglobin oxygen saturation was seen and there was no recuperation, suggesting that the

hepatic artery cannot immediately compensate for a loss of portal venous inflow. This shows the great importance of the inflow of the portal vein for the hepatic oxygenation. In these studies the NIRS probes were placed upon the surface of the liver. Seifalian and El-Desoky et al described a decrease in HbO

and an increase in HbR, measured with NIRO-500 (Hamamatsu, Hamamatsu City, Japan) during graded hypoxia (145). The same group found a good correlation between HbO

and the arterial oxygen pressure (146) and also between HbO

and the hepatic vein oxygen partial pressure in rabbits during graded hypoxia (147).

All these experiments where performed with the NIRS optodes placed on the surface of the liver. Beilman et al described a good correlation between the regional oxygen saturation of the liver and the arterial oxygen delivery (OD) during haemorrhagic shock in piglets (148).

Schultz and Weiss measured the oxygen saturation of the liver by means of the tissue oxygenation index, measured with NIRO 300 transcutaneously, during cardiac catheterisations. They described a good correlation between the TOI of the liver and the central venous oxygen saturation, measured in the right atrium, in children during cardiac catheterisation (149;150). They did not find a good correlation between TOI of the liver and the hepatic vein oxygenation (149). However, they performed single point measurements and no trend evaluations were done. We confirmed that the tissue oxygenation index of the liver correlates well with the mixed venous oxygen saturation, measured in the pulmonary artery in this study. A good relation was also found with CvO

, the mixed venous oxygen content(151).

No relation was found with the arterial oxygen saturation, which is in contrast to earlier descriptions (146;148). However in these studies HbO

was measured during graded hypoxia and thus mainly reflecting hypoxic hypoxia. By increasing the temperature and the induction of hypocapnia, we tested changes in liver oxygen saturation in a more physiological way.

Arterial oxygenation remained stable so that changes in TOI were mainly caused by changes in splanchnic blood flow or changes in oxygen consumption(151).

Regarding the relation with intestinal blood flow, a good correlation was found between the

blood flow in the distal ileum and the mid-gut of the small bowel and the tissue oxygenation

of the liver in our study(151). No correlation was found with the blood flow in the stomach

and the proximal part of the jejunum. These correlations reflect the important role of the

portal vein in the oxygenation of the liver as was described by Tokuka et al (144). In a study

in rabbits, where the mesenteric artery was clamped for 30 minutes, a significant decrease in

liver TOI was seen after 90 minutes of occlusion of the SMA and is likely to be the

consequence of bowel ischemia (152) . The further decrease after reperfusion might reflect

reperfusion injury. The most important confounding factor is the arterial hepatic buffer reaction (153;154). When there is a decrease in portal blood flow, the hepatic arterial blood flow will increase. This leads also to an increase in oxygenation in the liver and a delay in decrease in liver oxygenation. However measuring the liver oxygenation might still tell us more about the splanchnic circulation and oxygenation in a non-invasive way. Teller et al (155;156) were the first to use the measurement of TOI of the liver as a possible parameter of intestinal flow. They found a decrease of TOI of the liver after feeding a bolus of breast milk.

It must be stressed that changes in tissue oxygenation of the liver reflect changes in the whole splanchnic bed and also, most importantly, changes in oxygen consumption of the liver itself.

If there is ischaemia in other parts of the splanchnic system, the change in TOI will reflect mainly this ischaemia.

Another way to look at the splanchnic oxygenation and circulation is to measure the oxygenation under the umbilicus. Petros and Fortune were the first to use this method and they used specifically the difference between the cerebral and splanchnic oxygenation, called the cerebrosplanchnic oxygen ratio (CSOR = (157;158). They described that CSOR had a 90% (56-100%) sensitivity to detect splanchnic ischemia, indicating that this might be a non- invasive way to detect necrotising enterocolitis. Dave et al looked at the CSOR during feeding and found that CSOR and splanchnic tissue oxygenation, but not cerebral oxygenation increased during feeding in stable infants (159).

Further clinical research is needed whether liver oxygenation or CSOR might be a good indicator for splanchnic ischaemia.

8. MEASUREMENT OF THE PERIPHERAL OXYGENATION.

NIRS can also be used to measure the peripheral oxygenation, however the mechanisms for

the adequate provision of oxygen to the peripheral tissues remain (160). The best validated

way to measure the peripheral venous saturation is with the partial venous occlusion

technique as shown by Yoxall et al (53). Most measurements are made on the forearm. The

method is well validated against co-oximetry (161). The main problem is that this will give

only intermittent values and can only be performed by trained clinicians. Therefore the use of

spatially resolved spectroscopy might be used and FTOE, as explained higher, can be derived

from this measurements. This might also be used for measurement of peripheral perfusion and

oxygenation. Interestingly, recommendations for use of NIRS in peripheral measurements were published, which is still not been done for cerebral measurements (162).

Peripheral FOE increases with hypocapnia(163), , symptomatic anaemia (164) and increased viscosity. There is no relation with cardiac output(54), hypotension(163) or low temperature(165). Whether this method will be used in the future for clinical management of shock and hypotension cannot be foreseen at this moment.

9. CONCLUSION.

Near-infrared spectroscopy has evolved from a purely research instrument to a monitor that is increasingly introduced into the clinical environment. It is mainly been used for the measurement of the cerebral oxygenation and circulation and it becomes more and more a monitor that helps the clinician in understanding the effect of different pathologies and treatments on the cerebral oxygenation. Especially in the situation of cerebral ischemic hypoxia NIRS will be a monitor that gives extra information.

In the future combination with other parameters like blood pressure and EEG will help to measure cerebral autoregulation and neurovascular coupling in the clinical setting.

The last important future prospect is that NIRS can also measure other organs like the liver, the gastro-intestinal system, the muscle, the kidney and the overall peripheral venous oxygenation in a non-invasive way.

10. REFERENCE LIST.

Reference List

(1) Jobsis FF. Noninvasive, infrared monitoring of cerebral and myocardial oxygen sufficiency and circulatory parameters. Science 1977; 198(4323):1264-1267.

(2) Brazy JE, Lewis DV. Changes in cerebral blood volume and cytochrome aa3 during hypertensive peaks in preterm infants. J Pediatr 1986; 108(6):983-987.

(3) Delpy DT, Cope MC, Cady EB, Wyatt JS, Hamilton PA, Hope PL et al. Cerebral monitoring in newborn infants by magnetic resonance and near infrared spectroscopy.

Scand J Clin Lab Invest Suppl 1987; 188:9-17.

(4) Matcher J, Kirkpatrick PJ, Nahid M, Cope M, Delpy DT. Absolute quantification

methods in tissue near infrared spectroscopy. Proc SPIE 1995; 2389:486-495.

(5) Suzuki S, Takasaki S, Ozaki T, Kobayashi Y. A tissue oxygenation monitor using NIR spatially resolved spectroscopy. Proc SPIE 1999; 3597:582-592.

(6) Wolf M, Greisen G. Advances in near-infrared spectroscopy to study the brain of the preterm and term neonate. Clin Perinatol 2009; 36(4):807-34, vi.

(7) Faris F, Thorniley M, Wickramasinghe Y, Houston R, Rolfe P, Livera N et al. Non- invasive in vivo near-infrared optical measurement of the penetration depth in the neonatal head. Clin Phys Physiol Meas 1991; 12(4):353-358.

(8) Choi J, Wolf M, Toronov V, Wolf U, Polzonetti C, Hueber D et al. Noninvasive determination of the optical properties of adult brain: near-infrared spectroscopy approach. J Biomed Opt 2004; 9(1):221-229.

(9) Elwell C. A practical users guide to near infrared spectroscopy. Hamamatsu Phototonics KK. 1995. Hamamatsu Phototonics KK.

(10) Benaron DA, Kurth CD, Steven JM, Delivoria-Papadopoulos M, Chance B.

Transcranial optical path length in infants by near-infrared phase-shift spectroscopy. J Clin Monit 1995; 11(2):109-117.

(11) Wyatt JS, Cope M, Delpy DT, van der ZP, Arridge S, Edwards AD et al. Measurement of optical path length for cerebral near-infrared spectroscopy in newborn infants. Dev Neurosci 1990; 12(2):140-144.

(12) Kety SS, Schmidt CF. The effects of altered arterial tensions of carbon dioxide and oxygen on cerebral blood flow and cerebral oxygen consumption of normal young men. J Clin Invest 1948; 27(4):484-492.

(13) Bucher HU, Edwards AD, Lipp AE, Duc G. Comparison between near infrared spectroscopy and 133Xenon clearance for estimation of cerebral blood flow in critically ill preterm infants. Pediatr Res 1993; 33(1):56-60.

(14) Skov L, Pryds O, Greisen G. Estimating cerebral blood flow in newborn infants:

comparison of near infrared spectroscopy and 133Xe clearance. Pediatr Res 1991;

30(6):570-573.

(15) Goddard-Finegold J, Louis PT, Rodriguez DL, David Y, Contant CF, Rolfe P.

Correlation of near infrared spectroscopy cerebral blood flow estimations and microsphere quantitations in newborn piglets. Biol Neonate 1998; 74(5):376-384.

(16) Brown DW, Picot PA, Naeini JG, Springett R, Delpy DT, Lee TY. Quantitative near infrared spectroscopy measurement of cerebral hemodynamics in newborn piglets.

Pediatr Res 2002; 51(5):564-570.

(17) Kuebler WM, Sckell A, Habler O, Kleen M, Kuhnle GE, Welte M et al. Noninvasive

measurement of regional cerebral blood flow by near-infrared spectroscopy and

indocyanine green. J Cereb Blood Flow Metab 1998; 18(4):445-456.

(18) Roberts I, Fallon P, Kirkham FJ, Lloyd-Thomas A, Cooper C, Maynard R et al.

Estimation of cerebral blood flow with near infrared spectroscopy and indocyanine green. Lancet 1993; 342(8884):1425.

(19) Wyatt JS, Cope M, Delpy DT, Richardson CE, Edwards AD, Wray S et al.

Quantitation of cerebral blood volume in human infants by near-infrared spectroscopy.

J Appl Physiol 1990; 68(3):1086-1091.

(20) Barfield CP, Yu VY, Noma O, Kukita J, Cussen LJ, Oates A et al. Cerebral blood volume measured using near-infrared spectroscopy and radiolabels in the immature lamb brain. Pediatr Res 1999; 46(1):50-56.

(21) Brun NC, Moen A, Borch K, Saugstad OD, Greisen G. Near-infrared monitoring of cerebral tissue oxygen saturation and blood volume in newborn piglets. Am J Physiol 1997; 273(2 Pt 2):H682-H686.

(22) Wolf M, Weber O, Keel M, Golay X, Scheidegger M, Bucher HU et al. Comparison of cerebral blood volume measured by near infrared spectroscopy and contrast enhanced magnetic resonance imaging. Adv Exp Med Biol 1999; 471:767-773.

(23) Babcock GT, Wikstrom M. Oxygen activation and the conservation of energy in cell respiration. Nature 1992; 356(6367):301-309.

(24) Cooper CE, Delpy DT, Nemoto EM. The relationship of oxygen delivery to absolute haemoglobin oxygenation and mitochondrial cytochrome oxidase redox state in the adult brain: a near-infrared spectroscopy study. Biochem J 1998; 332 ( Pt 3):627-632.

(25) Adcock LM, Wafelman LS, Hegemier S, Moise AA, Speer ME, Contant CF et al.

Neonatal intensive care applications of near-infrared spectroscopy. Clin Perinatol 1999; 26(4):893-903, ix.

(26) Edwards AD, Brown GC, Cope M, Wyatt JS, McCormick DC, Roth SC et al.

Quantification of concentration changes in neonatal human cerebral oxidized cytochrome oxidase. J Appl Physiol 1991; 71(5):1907-1913.

(27) Tachtsidis I, Tisdall MM, Pritchard C, Leung TS, Ghosh A, Elwell CE et al. Analysis of the Changes in the Oxidation of Brain Tissue Cytochrome-c-Oxidase in Traumatic Brain Injury Patients during Hypercapnoea : A Broadband NIRS Study. Adv Exp Med Biol 2011; 915:9-14.

(28) Weiss M, Dullenkopf A, Kolarova A, Schulz G, Frey B, Baenziger O. Near-infrared spectroscopic cerebral oxygenation reading in neonates and infants is associated with central venous oxygen saturation. Paediatr Anaesth 2005; 15(2):102-109.

(29) Nagdyman N, Fleck T, Barth S, Abdul-Khaliq H, Stiller B, Ewert P et al. Relation of cerebral tissue oxygenation index to central venous oxygen saturation in children.

Intensive Care Med 2004; 30(3):468-471.

(30) Moran M, Miletin J, Pichova K, Dempsey EM. Cerebral tissue oxygenation index and superior vena cava blood flow in the very low birth weight infant. Acta Paediatr 2009;

98(1):43-46.

(31) Shimizu N, Gilder F, Bissonnette B, Coles J, Bohn D, Miyasaka K. Brain tissue oxygenation index measured by near infrared spatially resolved spectroscopy agreed with jugular bulb oxygen saturation in normal pediatric brain: a pilot study. Childs Nerv Syst 2005; 21(3):181-184.

(32) Naulaers G, Meyns B, Miserez M, Leunens V, Van HS, Casaer P et al. Use of tissue oxygenation index and fractional tissue oxygen extraction as non-invasive parameters for cerebral oxygenation. A validation study in piglets. Neonatology 2007; 92(2):120- 126.

(33) Sorensen LC, Greisen G. Precision of measurement of cerebral tissue oxygenation index using near-infrared spectroscopy in preterm neonates. J Biomed Opt 2006;

11(5):054005.

(34) Sorensen LC, Leung TS, Greisen G. Comparison of cerebral oxygen saturation in premature infants by near-infrared spatially resolved spectroscopy: observations on probe-dependent bias. J Biomed Opt 2008; 13(6):064013.

(35) Dullenkopf A, Kolarova A, Schulz G, Frey B, Baenziger O, Weiss M. Reproducibility of cerebral oxygenation measurement in neonates and infants in the clinical setting using the NIRO 300 oximeter. Pediatr Crit Care Med 2005; 6(3):344-347.

(36) Naulaers G, Morren G, Van Huffel S, Casaer P, Devlieger H. Measurement of tissue oxygenation index during the first three days in premature born infants. Adv Exp Med Biol 2003; 510:379-383.

(37) Lemmers PM, Toet M, van Schelven LJ, van Bel F. Cerebral oxygenation and cerebral oxygen extraction in the preterm infant: the impact of respiratory distress syndrome.

Exp Brain Res 2006; 173(3):458-467.

(38) Petrova A, Mehta R. Near-infrared spectroscopy in the detection of regional tissue oxygenation during hypoxic events in preterm infants undergoing critical care. Pediatr Crit Care Med 2006; 7(5):449-454.

(39) Lemmers PM, van Bel F. Left-to-right differences of regional cerebral oxygen

saturation and oxygen extraction in preterm infants during the first days of life. Pediatr Res 2009; 65(2):226-230.

(40) Hou X, Ding H, Teng Y, Zhou C, Tang X, Li S et al. Research on the relationship between brain anoxia at different regional oxygen saturations and brain damage using near-infrared spectroscopy. Physiol Meas 2007; 28(10):1251-1265.

(41) Kurth CD, Levy WJ, McCann J. Near-infrared spectroscopy cerebral oxygen saturation thresholds for hypoxia-ischemia in piglets. J Cereb Blood Flow Metab 2002; 22(3):335-341.

(42) Dent CL, Spaeth JP, Jones BV, Schwartz SM, Glauser TA, Hallinan B et al. Brain magnetic resonance imaging abnormalities after the Norwood procedure using regional cerebral perfusion. J Thorac Cardiovasc Surg 2006; 131(1):190-197.

(43) Taillefer MC, Denault AY. Cerebral near-infrared spectroscopy in adult heart surgery:

systematic review of its clinical efficacy. Can J Anaesth 2005; 52(1):79-87.

(44) Fauchere JC, Schulz G, Haensse D, Keller E, Ersch J, Bucher HU et al. Near-infrared spectroscopy measurements of cerebral oxygenation in newborns during immediate postnatal adaptation. J Pediatr 2010; 156(3):372-376.

(45) Saugstad OD. Oxidative stress in the newborn--a 30-year perspective. Biol Neonate 2005; 88(3):228-236.

(46) Sorensen LC, Greisen G. The brains of very preterm newborns in clinically stable condition may be hyperoxygenated. Pediatrics 2009; 124(5):e958-e963.

(47) Baerts W, Lemmers PM, van Bel F. Cerebral Oxygenation and Oxygen Extraction in the Preterm Infant during Desaturation: Effects of Increasing FiO(2) to Assist

Recovery. Neonatology 2010; 99(1):65-72.

(48) Greisen G. Ischaemia of the preterm brain. Biol Neonate 1992; 62(4):243-247.

(49) Altman DI, Perlman JM, Volpe JJ, Powers WJ. Cerebral oxygen metabolism in newborns. Pediatrics 1993; 92(1):99-104.

(50) Roth SC, Baudin J, Cady E, Johal K, Townsend JP, Wyatt JS et al. Relation of deranged neonatal cerebral oxidative metabolism with neurodevelopmental outcome and head circumference at 4 years. Dev Med Child Neurol 1997; 39(11):718-725.

(51) Yoxall CW, Weindling AM. Measurement of cerebral oxygen consumption in the human neonate using near infrared spectroscopy: cerebral oxygen consumption increases with advancing gestational age. Pediatr Res 1998; 44(3):283-290.

(52) Purves MJ. The physiology of the cerebral circulation. Monogr Physiol Soc 1972;

(28):1-414.

(53) Yoxall CW, Weindling AM. The measurement of peripheral venous oxyhemoglobin saturation in newborn infants by near infrared spectroscopy with venous occlusion.

Pediatr Res 1996; 39(6):1103-1106.

(54) Kissack CM, Garr R, Wardle SP, Weindling AM. Cerebral fractional oxygen

extraction is inversely correlated with oxygen delivery in the sick, newborn, preterm infant. J Cereb Blood Flow Metab 2005; 25(5):545-553.

(55) Naulaers G, Meyns B, Miserez M, Leunens V, Van HS, Casaer P et al. Use of tissue oxygenation index and fractional tissue oxygen extraction as non-invasive parameters for cerebral oxygenation. A validation study in piglets. Neonatology 2007; 92(2):120- 126.

(56) Victor S, Marson AG, Appleton RE, Beirne M, Weindling AM. Relationship between blood pressure, cerebral electrical activity, cerebral fractional oxygen extraction, and peripheral blood flow in very low birth weight newborn infants. Pediatr Res 2006;

59(2):314-319.

(57) Wardle SP, Yoxall CW, Weindling AM. Determinants of cerebral fractional oxygen

extraction using near infrared spectroscopy in preterm neonates. J Cereb Blood Flow

Metab 2000; 20(2):272-279.

(58) Wardle SP, Yoxall CW, Weindling AM. Cerebral oxygenation during cardiopulmonary bypass. Arch Dis Child 1998; 78(1):26-32.

(59) Koehler RC, Jones MD, Jr., Traystman RJ. Cerebral circulatory response to carbon monoxide and hypoxic hypoxia in the lamb

142. Am J Physiol 1982; 243(1):H27-H32.

(60) Jones MD, Jr., Traystman RJ, Simmons MA, Molteni RA. Effects of changes in arterial O2 content on cerebral blood flow in the lamb. Am J Physiol 1981;

240(2):H209-H215.

(61) Gleason CA, Hamm C, Jones MD, Jr. Effect of acute hypoxemia on brain blood flow and oxygen metabolism in immature fetal sheep. Am J Physiol 1990; 258(4 Pt 2):H1064-H1069.

(62) Petrova A, Mehta R. Regional tissue oxygenation in association with duration of hypoxaemia and haemodynamic variability in preterm neonates. Arch Dis Child Fetal Neonatal Ed 2010; 95(3):F213-F219.

(63) Fan FC, Chen RY, Schuessler GB, Chien S. Effects of hematocrit variations on regional hemodynamics and oxygen transport in the dog. Am J Physiol 1980;

238(4):H545-22.

(64) Herold S, Brozovic M, Gibbs J, Lammertsma AA, Leenders KL, Carr D et al.

Measurement of regional cerebral blood flow, blood volume and oxygen metabolism in patients with sickle cell disease using positron emission tomography. Stroke 1986;

17(4):692-698.

(65) Van Ameringen MR, Fouron JC, Bard H, Le Guennec JC, Prosmanne J. Oxygenation in anemic newborn lambs with high or low oxygen affinity red cells. Pediatr Res 1981;

15(12):1500-1503.

(66) van Hoften JC, Verhagen EA, Keating P, ter Horst HJ, Bos AF. Cerebral tissue

oxygen saturation and extraction in preterm infants before and after blood transfusion.

Arch Dis Child Fetal Neonatal Ed 2010; 95(5):F352-F358.

(67) Meek JH, Tyszczuk L, Elwell CE, Wyatt JS. Cerebral blood flow increases over the first three days of life in extremely preterm neonates. Arch Dis Child Fetal Neonatal Ed 1998; 78(1):F33-F37.

(68) Munger DM, Bucher HU, Duc G. Sleep state changes associated with cerebral blood volume changes in healthy term newborn infants. Early Hum Dev 1998; 52(1):27-42.

(69) Dietz V, Wolf M, Keel M, Siebenthal K, Baenziger O, Bucher H. CO2 reactivity of the cerebral hemoglobin concentration in healthy term newborns measured by near infrared spectrophotometry. Biol Neonate 1999; 75(2):85-90.

(70) Fallon P, Roberts IG, Kirkham FJ, Edwards AD, Lloyd-Thomas A, Elliott MJ.

Cerebral blood volume response to changes in carbon dioxide tension before and during cardiopulmonary bypass in children, investigated by near infrared

spectroscopy. Eur J Cardiothorac Surg 1994; 8(3):130-134.

(71) Greisen G. Cerebral blood flow and energy metabolism in the newborn. Clin Perinatol 1997; 24(3):531-546.

(72) Li J, Zhang G, Holtby H, Bissonnette B, Wang G, Redington AN et al. Carbon dioxide--a complex gas in a complex circulation: its effects on systemic

hemodynamics and oxygen transport, cerebral, and splanchnic circulation in neonates after the Norwood procedure. J Thorac Cardiovasc Surg 2008; 136(5):1207-1214.

(73) Mosca FA, Colnaghi M, Lattanzio M, Bray M, Pugliese S, Fumagalli M. Closed versus open endotracheal suctioning in preterm infants: effects on cerebral oxygenation and blood volume. Biol Neonate 1997; 72(1):9-14.

(74) Skov L, Ryding J, Pryds O, Greisen G. Changes in cerebral oxygenation and cerebral blood volume during endotracheal suctioning in ventilated neonates. Acta Paediatr 1992; 81(5):389-393.

(75) Vanderhaegen J, Naulaers G, Vanhole C, De Smet D, Van Huffel S, Vanhaesebrouck S et al. The effect of changes in tPCO2 on the fractional tissue oxygen extraction--as measured by near-infrared spectroscopy--in neonates during the first days of life. Eur J Paediatr Neurol 2009; 13(2):128-134.

(76) Wyatt JS, Edwards AD, Cope M, Delpy DT, McCormick DC, Potter A et al. Response of cerebral blood volume to changes in arterial carbon dioxide tension in preterm and term infants. Pediatr Res 1991; 29(6):553-557.

(77) Lemmers PM, Toet MC, van Bel F. Impact of patent ductus arteriosus and subsequent therapy with indomethacin on cerebral oxygenation in preterm infants. Pediatrics 2008; 121(1):142-147.

(78) Underwood MA, Milstein JM, Sherman MP. Near-infrared spectroscopy as a screening tool for patent ductus arteriosus in extremely low birth weight infants.

Neonatology 2007; 91(2):134-139.

(79) Alfonso I, Papazian O, Dunoyer C, Yaylali I. [New techniques used to monitor cerebral function in the newborn]. Rev Neurol 2002; 34(1):27-30.

(80) Munro MJ, Walker AM, Barfield CP. Hypotensive extremely low birth weight infants have reduced cerebral blood flow. Pediatrics 2004; 114(6):1591-1596.

(81) Tsuji M, duPlessis A, Taylor G, Crocker R, Volpe JJ. Near infrared spectroscopy detects cerebral ischemia during hypotension in piglets. Pediatr Res 1998; 44(4):591- 595.

(82) Pellicer A, Valverde E, Elorza MD, Madero R, Gaya F, Quero J et al. Cardiovascular support for low birth weight infants and cerebral hemodynamics: a randomized, blinded, clinical trial. Pediatrics 2005; 115(6):1501-1512.

(83) Weindling AM, Kissack CM. Blood pressure and tissue oxygenation in the newborn

baby at risk of brain damage. Biol Neonate 2001; 79(3-4):241-245.