**2. Cerebral NIRS**

The neonatal period is a unique time in life as the infant undergoes dramatic physiologic changes during transition from intra- to extra-uterine life, which involve hemodynamics and affect oxygenation, reflected in rSO2. Due to its vulnerability the neonatal central nervous system is the main area of interest for measurements of oxygenation. The majority of articles written on the clinical use of NIRS in neonates include reports on cerebral measurements (crSO2 or cerebral Tissue Oxygenation Index (TOI)).

#### **2.1 Effect of gestational and postnatal age**

The largest body of research investigates cerebral NIRS values. Reports regarding effects of gestational age (pre-term, term, post-term) and postnatal/chronologic age on NIRS values are conflicting.

In a study by McNeill, which was blinded to caregivers and sampled from birth for a maximum of 21 days, baseline rSO2 for preterm infants (gestational age of 29-34 weeks) differed from established pediatric norms, while values for term neonates in the first days of life did not (McNeill et al., 2010, 2011). The observation by McNeill (McNeill et al., 2010, 2011) that cerebral NIRS decreases over time are supported by Roche-Labarbe's findings following weekly spot samples during the first 6 weeks obtained with a different study protocol and different NIRS equipment. (Roche-Labarbe et al., 2010, 2011) Both observations contradict Lemmers' study in which twice daily 60 minute sampling periods found no observed change. (Lemmers et al., 2006)

Naulears found an increase in cerebral oxygenation in premature infants during the first three days. In this study sampling periods were 30 min. NIRS recordings occurred with a different instrument. (Naulaers et al., 2002) Meek's earlier report from 1998 in ventilated babies used NIRS and found an increase in cerebral blood flow over time. (Meek et al., 1998)

A study measuring rSO2-c in transition after delivery found by minute 3 that rSO2 increased and reached a plateau by minute 7. (Urlesberger et al., 2010)

More recently, Takami followed cerebral TOI in extremely low birth weight infants (ELBWs) at 3-6h followed by samples every 6h up to 72h. He observed a decrease in measurements until 12h, then an increase that correlated with similar changes in SVC flow. (Takami et al., 2010).

When reviewing this literature regarding the contradicting study results, possible explanations present themselves: Patient populations are not identical. Protocols vary from study to study. Different sampling times may play an important role in influencing results, especially when spot samples versus long-term continuous data were collected. If studies were not blinded, care giving and subsequently observations might have been influenced. The use of different monitors and probes and probe placement may further lead to different results. Studies were small and data inconclusive. There was some agreement regarding abnormally low values being linked to poor outcome. (Dullenkopf et al., 2003; Sorensen et al., 2008; van Bel et al., 2008; Wolf & Greisen, 2009, also see cerebral hypoxia)

#### **2.2 Variability**

8 Infrared Spectroscopy – Life and Biomedical Sciences

2004; Madsen et al., 2000; McNeill et al., 2010, 2011; Wassenaar et al., 2005) NIRS

Commercially available sensors for neonates have become well tolerated due to smaller size and being lined with a skin friendly adhesive. To provide further skin protection in extremely premature patients probes can be attached to a light-permeable skin barrier

Organs which can be monitored in neonates are brain, kidney, gut, liver and muscle. This

The neonatal period is a unique time in life as the infant undergoes dramatic physiologic changes during transition from intra- to extra-uterine life, which involve hemodynamics and affect oxygenation, reflected in rSO2. Due to its vulnerability the neonatal central nervous system is the main area of interest for measurements of oxygenation. The majority of articles written on the clinical use of NIRS in neonates include reports on cerebral measurements (c-

The largest body of research investigates cerebral NIRS values. Reports regarding effects of gestational age (pre-term, term, post-term) and postnatal/chronologic age on NIRS values

In a study by McNeill, which was blinded to caregivers and sampled from birth for a maximum of 21 days, baseline rSO2 for preterm infants (gestational age of 29-34 weeks) differed from established pediatric norms, while values for term neonates in the first days of life did not (McNeill et al., 2010, 2011). The observation by McNeill (McNeill et al., 2010, 2011) that cerebral NIRS decreases over time are supported by Roche-Labarbe's findings following weekly spot samples during the first 6 weeks obtained with a different study protocol and different NIRS equipment. (Roche-Labarbe et al., 2010, 2011) Both observations contradict Lemmers' study in which twice daily 60 minute sampling periods found no

Naulears found an increase in cerebral oxygenation in premature infants during the first three days. In this study sampling periods were 30 min. NIRS recordings occurred with a different instrument. (Naulaers et al., 2002) Meek's earlier report from 1998 in ventilated babies used NIRS and found an increase in cerebral blood flow over time. (Meek et al., 1998) A study measuring rSO2-c in transition after delivery found by minute 3 that rSO2 increased

More recently, Takami followed cerebral TOI in extremely low birth weight infants (ELBWs) at 3-6h followed by samples every 6h up to 72h. He observed a decrease in measurements until 12h, then an increase that correlated with similar changes in SVC flow. (Takami et al., 2010).

chapter will comment on the most commonly used sites– the brain, kidney and gut.

measurements may differ between probes. (Sorensen et al., 2008)

without interference with measurements. (McNeill et al., 2010, 2011)

rSO2 or cerebral Tissue Oxygenation Index (TOI)).

**2.1 Effect of gestational and postnatal age** 

observed change. (Lemmers et al., 2006)

and reached a plateau by minute 7. (Urlesberger et al., 2010)

**1.4 Safety and feasibility** 

**1.5 Monitoring** 

**2. Cerebral NIRS** 

are conflicting.

Variability is the change in percent of rSO2 away from a calculated baseline. It can be followed over time to know how much time the rSO2 was above or below baseline. The baseline differs from patient to patient. Variability is an area of interest and needs further investigation: Cerebral daily variability is small. Large changes (>20%) off the baseline would raise concern for acute clinical change. (McNeill et al., 2010, 2011) Change in variability may be an indicator of infection (Yanowitz et al., 2006). The change in baseline over the first weeks of life, which is observed in preterm infants, may represent ongoing developmental maturation independent of feeding status. (McNeill et al., 2010, 2011)

#### **2.3 Peripheral blood pressure and oxygenation, impact on autoregulation**

In the research setting cerebral blood flow and blood volume measurements, oxy- and deoxy hemoglobin and fractional extraction of oxygen (FTOE) as well as blood gas samples from central catheters added to detailed understanding of physiology.

Adequate O2 delivery to the brain tissue is most critical. Assessment of O2 delivery and consumption help understand clinical scenarios and their underlying pathophysiology: At the bed side this evaluation can occur by following changes in cerebral rSO2, changes in BP, oxygenation and peripheral blood gases. The below clinical scenarios for monitoring are amongst the more common:

Cerebral autoregulation is a homeostatic phenomenon controlled by the main capacitance vessels in the cerebral circulation. Through dilatation and constriction of these vessels cerebral blood flow and cerebral rSO2 or TOI are maintained at a steady level over a range of changing mean arterial blood pressures (MABP). This range is narrower in neonates, particularly in preterm infants. Cerebral pressure-passivity or loss of autoregulation is associated with low gestational age, low birth weight and systemic hypotension in a large study of 90 patients. (Soul et al., 2007)

If rSO2 or TOI changes correlate with the wave form of MABP autoregulation is lost. Swings in peripheral perfusion will be mirrored in cerebral blood flow and regional saturation readings. This phenomenon, when profound, carries an increased risk for intra-ventricular hemorrhage (IVH) and peri-ventricular leucomalacia (PVL) in preterm infants and generally a poor prognosis for neurodevelopment outcome. The more swings or changes in mean arterial pressure (MAP) and NIRS coincide and mirror each other, the more the waves are in concordance. Several studies link concordance with a more unfavorable prognosis and a higher likelihood of death. (Caicedo et al., 2011; DeSmet et al., 2010; Greisen & Borch, 2001;

Use of Near-Infrared Spectroscopy in the Management of Patients in

measure of short term outcome. (Gilmore et al., 2011)

influencing oxygenation. (Kurth et al., 1995)

**2.4 Cerebral hypoxia** 

with other diagnoses.

2009).

**3. Renal NIRS** 

See figure 2.

**2.5 Cerebral hyperoxia** 

Neonatal Intensive Care Units – An Example of Implementation of a New Technology 11

Hahn et al., 2010; Lemmers et al., 2006; Morren et al., 2003; Munro et al., 2004, 2005; O'Leary et al., 2009; Seri, 2006; Tsuji et al., 2000; Wong et al., 2008) In a recent study 23 infants with a mean gestational age of 26.7 +/-1.4 weeks were observed with NIRS. They were found to have periods of loss of cerebral autoregulation which were more profound with lower, longer lasting MABPs. There was no correlation with head ultrasound (HUS) findings as

A study followed changes in cerebral NIRS in ventilated preterm infants and found frequent periods of loss of autoregulation. (Lemmers et al., 2006). Vanderhaegen stresses the important contribution of pCO2 to cerebral blood flow, which may possibly override autoregulation. (Vanderhaegen et al., 2010) Hoffmann manipulated pCO2 in neonates undergoing cardiac surgery to improve cerebral blood flow. (Hoffman et al., 2005) According to another study by Vanderhaegen in 11 ELBWS blood glucose may play a role in

Cerebral hypoxia is a feared event as it translates to long-term morbidity and mortality*.* There is not enough data available linking a specific duration of hypoxia and levels of rSO2 or TOI while in the NICU with outcomes. There are no absolute numbers as reference in the human neonate. A piglet study from 2007 demonstrated changes seen on brain autopsy 72h after the animal spent 30 min. with rSO2-c of <40%. (Hou et al., 2007) It is not certain whether observations of concerning low levels of r-SO2/TOI in cardiac patients (Dullenkopf et al., 2003; Sorensen et al., 2008; van Bel et al., 2008; Wolf & Greisen, 2009) apply to infants

Cerebral hyperoxia in the critically ill neonate may occur by 2 mechanisms: either as hyperoxygenation during the reperfusion phase of severe hypoxic ischemic encephalopathy most commonly occurring in neonates after perinatal birth depression or from decreased brain metabolism as seen in critical patients when blood flow is uncoupled from O2 (Toet, 2006; Wolf & Greisen, 2009). Either scenario is concerning for a poor long-term prognosis. The overall clinical situation needs to be taken into consideration as cerebral rSO2 in well preterm neonates has also been reported to be high in the first days of life. (Sorensen et al.,

Renal rSO2 is higher than cerebral rSO2. McNeill reported that trends in cerebral and renal NIRS during the first 21 days of life mirror each other. Short-term and long-term variability of r-SO2 is small. Saturation changes exceeding >20% from baseline would be reason for concern and may indicate compromised perfusion. Several investigators report use in patients with shock or during surgery. Measurements of the renal rSO2 give insight into peripheral perfusion in general and into renal end-organ function. Using renal rSO2 in conjunction with cerebral rSO2 has been reported to give more and sometimes earlier insights into evolving pathology such as shock. (Cohn et al., 2003; Hoffman et al., 2003, 2004)

Fig. 1a. Example 1: Patient with loss of autoregulation and concordance of MAP and NIRS measurement of intravascular oxygenation (HbD). This patient had an unfavorable outcome.

Fig. 1b. Example 2: Maintenance of autoregulation (Tsuji, 2000)

Hahn et al., 2010; Lemmers et al., 2006; Morren et al., 2003; Munro et al., 2004, 2005; O'Leary et al., 2009; Seri, 2006; Tsuji et al., 2000; Wong et al., 2008) In a recent study 23 infants with a mean gestational age of 26.7 +/-1.4 weeks were observed with NIRS. They were found to have periods of loss of cerebral autoregulation which were more profound with lower, longer lasting MABPs. There was no correlation with head ultrasound (HUS) findings as measure of short term outcome. (Gilmore et al., 2011)

A study followed changes in cerebral NIRS in ventilated preterm infants and found frequent periods of loss of autoregulation. (Lemmers et al., 2006). Vanderhaegen stresses the important contribution of pCO2 to cerebral blood flow, which may possibly override autoregulation. (Vanderhaegen et al., 2010) Hoffmann manipulated pCO2 in neonates undergoing cardiac surgery to improve cerebral blood flow. (Hoffman et al., 2005) According to another study by Vanderhaegen in 11 ELBWS blood glucose may play a role in influencing oxygenation. (Kurth et al., 1995)
