INVOS Surgical and Critical Applications Guide

Detection and Correction of Brain Oxygen Imbalance Surgical and Critical Care Applications of the INVOS™ Cerebral Oximeter

A Pocket Guide for Clinicians

Harvey L. Edmonds, Jr. Ph.D. Professor Emeritus Department of Anesthesiology & Perioperative Medicine University of Louisville School of Medicine

Learning Objectives

Learning Objectives After reading this guide, clinicians should be able to:

»

Describe the link between regional cerebral oxygen balance and the INVOS parameter, regional oxygen saturation (rSO2)

»

Integrate brain rSO2 information with other physiologic & clinical data before, during and after surgery

»

Identify special situations which can influence cerebral rSO2 monitoring

»

Discuss the clinical management and response to cerebral rSO2 monitoring

This resource is intended for educational purposes only. It is not intended to provide comprehensive or patient-specific clinical practice recommendations for rSO2 monitoring technology. The clinical choices discussed in this text may or may not be consistent with your own patient requirements, your clinical practice approaches, or guidelines for practice that are endorsed by your institution or practice group. It is the responsibility of each clinician to make his/her own determination regarding clinical practice decisions that are in the best interest of patients. Readers are advised to review the current product information including the Indications for Use currently provided by the manufacturer. Neither the publisher, authors, nor Covidien assumes any responsibility for any injury and or damage to persons or property resulting from information provided in this text. Dr. Edmonds received compensation from Covidien for his professional time spent preparing this educational piece.

Table of Contents

Table of Contents Executive Overview ... i rSO2-A Clinically-Validated Measure of Regional Brain Oxygen Balance... 1 Brain rSO2 Monitoring Before, During and After General Anesthesia ... 6 Special Issues Impacting Brain rSO2 Monitoring ... 17 Clinical Management: Responding to Brain rSO2 Changes... 22 Summary ... 24 References... 25

Executive Overview

Executive Overview Regional oxygen saturation (rSO2) monitoring systems permit the continuous noninvasive measurement of cerebral regional oxygen balance within the frontal cerebral cortex. Since cerebral rSO2 represents an adjunct physiologic measure, local microcirculatory oxygen balance, it is important to appreciate the fundamental elements of this infrared light-based technology. This technology offers additional insights into patient clinical status; however, the novelty of the technology also makes it imperative for clinicians to review important situations and limitations that may influence rSO2. The non-invasive INVOSTM monitoring system is intended for use as an adjunct monitor of regional hemoglobin oxygen saturation of blood in the brain or in other tissue beneath the sensor. It is intended for use in individuals greater than 2.5 kg at risk for reduced-flow or no-flow ischemic states. The INVOS system should not be used as the sole basis for diagnostic or therapeutic decisions however; both randomized and non-randomized controlled clinical trials have shown the positive impact of INVOS-guided patient management on patient outcomes. Compared with standard clinical practice, INVOS monitoring with a standardized interventional protocol has been shown to provide improved outcomes after surgery and resource utilization.1-4 For complete information about the INVOS system, refer to the full service manual/instructions for use. Additional clinical information and other educational resources can be accessed at www.covidien.com/PACE. Customer support is also available via telephone (US Toll Free) 1-800-635-5267. International contact numbers are available at http://www.covidien.com/covidien/pages.aspx?page=Contact.

i

rSO2-A Clinically-Validated Measure of Regional Brain Oxygen Balance

rSO2-A Clinically-Validated Measure of Regional Brain Oxygen Balance Brain Oxygen Balance Monitoring with Near-infrared Spectroscopy (NIRS) Intracranial regional hemoglobin oxygen saturation (rSO2) measurement by NIRS is possible because the human skull is translucent to infrared light. As with other forms of clinical oximetry, saturation determination relies on multiple wavelengths of light to discriminate between the unique absorption spectra of oxyhemoglobin and deoxyhemoglobin. (Figure 1) Generally speaking, within the wavelength region of interest (i.e., spectrum), the only other competing infrared absorbers (i.e., chromophores) are water and the skin pigment melanin. (Figure 1) INVOS measurements appear to be relatively unaffected by normal skin pigment variation in adult patients; however, pathologic variations in pigmentation or brain water content may influence photon absorption, scatter and the resulting rSO2 measurement.5,6 Figure 1: Photon Absorption in the Cranium

The cranium contains four substances that absorb photons with wavelengths in the near-infrared range (i.e., 680-900 nm).

-1-

rSO2-A Clinically-Validated Measure of Regional Brain Oxygen Balance The INVOS™ System The INVOS system employs disposable sensors with an integrated near infrared light source and photodetector that can be applied to each side of the forehead for monitoring blood in the brain. This placement permits monitoring of the ischemia-susceptible cortical tissue in the border zone between the anterior and middle cerebral arteries but, may preclude detection of posterior border zone oxygen imbalance or perfusion abnormality.7,8 Figure 2 illustrates that the mean path of tissue-reflected photons is parabolic. With an INVOS optode light sourcedetector separation of 4 cm, infrared photons penetrate to a depth of ~2 cm [i.e., (light source - detector separation)-½] below the skin surface.9 Thus, INVOS monitoring measures the outer cerebral cortical layers in adults.9 The cerebral tissue sample volume is ~1.5 cc.9 Since there is sufficient photonabsorbing hemoglobin in larger vessels to trap all incident infrared light, surface-detected infrared reflections arise exclusively from blood vessels <100 μm in diameter. Based on recent positron emission tomography (PET) studies, the arterial:venous ratio within this microcirculation appears to be ~30:70.10 Observations during retrograde cerebral perfusion demonstrated that the INVOS monitor rSO2 determination is relatively insensitive to substantial shifts in this ratio.11 In addition, rSO2 differs from other oximetric technologies such as pulse oximetry (SpO2) and jugular venous oxygen saturation (SjvO2) because it does not require actively flowing blood, pulsatile or non-pulsatile. The INVOS system uses a proprietary process termed spatial resolution to suppress the influences of extracerebrally reflected photons. The optode contains a second infrared detector located closer to the light source. Its shallower parabolic path includes less brain and more extracerebral tissue reflections (Figure 2).12 -2-

rSO2-A Clinically-Validated Measure of Regional Brain Oxygen Balance The shallow, deep signal ratio results in a rSO2 measure that is ~70% intracranial and independent of interpatient variation in photon scatter.9,10 Currently, rSO2 provides the only non-invasive method to continuously monitor changes in local brain oxygen balance.13 Figure 2: INVOS™ Detector Orientation

Each INVOS optode uses the proprietary orientation of two infrared detectors to measure light reflected transcranially from a neighboring source. Detector orientation ensures that the parabolic photon paths to both detectors traverse brain tissue.

Validation of rSO2 as a Measure of Brain Oxygen Balance Compared with other oximetric technologies such as arterial (SaO2) and jugular venous (SjvO2) oxygen saturation, verification of brain rSO2 accuracy remains technically challenging. In the absence of a true reference, manufacturers and the U.S. Food & Drug Administration have adopted a proxy called field saturation (fSO2). This metric, k1(SaO2) + k2(SjvO2), was developed by an early INVOS system clinical investigator to assess cerebral oximeter performance.14 The manufacturer-specific constants k1 & k2 are empirical estimates of the arterial and venous blood contribution to proprietary rSO2 algorithms. -3-

rSO2-A Clinically-Validated Measure of Regional Brain Oxygen Balance A peer-reviewed report describes statistically significant correlations between fSO2 and the INVOS monitor rSO2 in healthy adults breathing room air as well as hypoxic and hypercapnic mixtures (Figure 3).15 Without a true reference standard however, regional cerebral oxygen saturation accuracy is indeterminate and the interpretation of device comparisons is complex and uncertain. Clinicians should appreciate that momentary rSO2 values and trending characteristics are machine-specific and are not interchangeable among different oximeter brands.16-18 As a result, it is unjustified to use clinical data generated from one proprietary rSO2 system to “validate” the utility of a competing device.19 Figure 3: Correlation Between rSO2 and fSO2

Cerebral oximeter performance is assessed by comparing rSO2 values to a proxy for brain saturation termed field saturation (fSO2). This graph derived from the study of Kim et al. (2000) shows statistically significant correlation between rSO2 and fSO2 in a group of adult volunteers exposed to graded levels of hypoxia and hypercapnia.

-4-

rSO2-A Clinically-Validated Measure of Regional Brain Oxygen Balance Normative Brain rSO2 Values Normative brain rSO2 values are an absolute requirement for the definition of abnormality. For the INVOS systems, the normative rSO2 is 71±6% for conscious healthy adults (n=44).15 Edmonds et al. preoperatively determined an rSO2 of 67±10% in 1,000 adult cardiac surgery patients (age 21-91 years, 32% female) (Figure 4).11 Values <50 were thus statistically subnormal. A left vs. right hemisphere asymmetry of >10% occurred in only 5% of patients. Recently, multiple studies have confirmed both the cardiac patient normative values and asymmetry incidence.20,21 It is noteworthy that SjVO2 asymmetry >10% occurs in a majority of patients.22 Thus, physiologically appropriate comparisons of rSO2 and SjvO2 require that both measurements invariably be made from the same side of the head. Figure 4: Sample rSO2 data

Edmonds et al. (2004) published the first large sample rSO2 normative data for adult cardiac surgery patients. Since the data were normally distributed, statistical abnormality is readily defined as the upper and lower 2.5% extremes of the frequency distribution.

-5-

Brain rSO2 Monitoring Before, During and After General Anesthesia

Brain rSO2 Monitoring Before, During and After General Anesthesia Pre-procedure (baseline) rSO2 INVOS monitoring does not require establishment of a preprocedure baseline reference. As with intraoperative blood pressure monitoring however, obtaining baseline information is sound clinical practice.23 Moreover, pre-procedure bilateral rSO2 values can alert the clinician to technical difficulties in need of immediate correction or valid pre-existing symmetrical or asymmetrical subnormal values. Collection of reliable baseline rSO2 values is influenced by proper recording technique. Prior to optode (i.e., disposable sensor) application, patient forehead skin oil should be removed with an INVOS-provided acetone wipe. Oil removal facilitates optimal adherence of the self-adhesive optode to block interference by environmental light. If the forehead is exposed to intense light (i.e., direct forehead illumination by surgical lights) or heat sources (i.e., fluid or body warmers), the optodes should be covered with an opaque material. If practical, the sensors should be positioned above each eye with the long axis parallel to the intra-aural line and the superior optode edge adjacent to the hairline. Consistent positioning in this manner minimizes inter- and intrasubject baseline rSO2 variation and avoids the potentially confounding effects of the frontal sinus on light scattering.24 Repeated optode use is not recommended because the accumulation of epidermal debris on the adhesive surface may have unpredictable effects on extracranial photon scattering.

-6-

Brain rSO2 Monitoring Before, During and After General Anesthesia Prior to monitoring, INVOS system recording quality should be assessed by inspection of the signal strength index (SSI) for each channel (Figure 5). The 5-unit bar scale is non-linear. Thus, the 1-bar signal strength is only 4% of that represented by 5 bars. Adequate signal strength is represented by the continuous display of >1 bar. Figure 5: INVOS™ Monitor

The green vertical 5-bar Signal Strength Indicators are shown for each INVOS recording channel. Signals with a stable reading of >1 bar are sufficiently strong to permit reliable monitoring. Large font numbers are the momentary rSO2 values for each channel, which are updated every 5 seconds. Mid-size font values are the baseline values. Red numbers reflect an oxygen debt alarm and the small font values display the percentage change from baseline.

Positioning A sudden symmetric or asymmetric rSO2 decrease may occur during anesthetic induction, pulmonary artery or central venous catheter insertion or final positioning (Figure 6). Without accompanying change in blood pressure or respiratory gases, precipitous rSO2 decline can help identify an otherwise unrecognized cerebral blood inflow or outflow obstruction.11,25,26 -7-

Brain rSO2 Monitoring Before, During and After General Anesthesia With cardiac and vascular surgery, the unexpected development of regional brain oxygen debt may be the consequence of a malpositioned heart, perfusion cannula, vascular clamp, ligature or cardiac vent.27-29 Figure 6: INVOS™ and Patient Positioning

INVOS monitoring detected rSO2 decline. Oxygen balance returned to normal with restoration of the supine position, and the surgery proceeded without incident.

CO2 Influence on rSO2 Cerebral arteries in the healthy brain are exquisitely sensitive to hydrogen ion shifts and consequently CO2 change. CO2 accumulation results in arteriolar vasodilation and attendant rSO2 increase.30,31 Of note, the CO2-mediated rSO2 rise accompanying endotracheal intubation provides a simple method to verify bihemispheric normal vascular responsiveness (Figure 7).11

-8-

Brain rSO2 Monitoring Before, During and After General Anesthesia Figure 7: INVOS™ and CO2 Accumulation

The inset graph at upper left shows the normal response to pre-oxygenation and anesthetic induction. The large graph also shows large rSO2 increase with endotracheal intubation, suggesting normal cerebral arteriole CO2 reactivity. Multiple hypocapnic episodes consistently resulted in brain oxygen debt. Each was promptly corrected by appropriate adjustments of the respiratory rate (RR) and tidal volume (Vt).

Since cerebral CO2 reactivity is a pre-condition for autoregulation, its absence signifies increased risk of potentially injurious oxygen imbalance and hypoperfusion (Figure 8).32 Armed with this knowledge, rSO2-guided blood pressure management may then be used to help avoid hypoperfusion injury. The individualized CO2:rSO2 relationships are also important during cardiopulmonary bypass to optimize acid-base management.33 With CO2unreactive cerebral arterioles, the risk of brain hypoperfusion is increased and the perfusionist has a diminished opportunity to improve brain oxygen balance via adjustments in acidbase management.

-9-

Brain rSO2 Monitoring Before, During and After General Anesthesia Figure 8: INVOS™ and Hypoperfusion

rSO2 trends were notable initially for abnormally low and asymmetric baseline values. The trends then precisely quantified the extent of brain oxygen debt associated with three failed intubation attempts and documented the ultimately successful fourth attempt.

Systemic & Regional Hypoxemia Influences on rSO2 The physiologic properties of brain rSO2 make it uniquely suited for the early detection of developing hypoxemia. Inspection of the familiar oxygen dissociation curve emphasizes that venous SvO2 or venous-weighted rSO2 will change more than arterial SaO2 or SpO2 to a fixed decline in blood oxygen partial pressure (Figure 9). This fact combined with the extraordinarily high brain oxygen demand results in observation that developing hypoxemia often appears first in brain rSO2 (Figure 8).34,35 Even with the extensive physiologic monitoring used during cardiac surgery, evidence of inadequate oxygen delivery may be first observed by a declining cerebral rSO2.36

-10-

Brain rSO2 Monitoring Before, During and After General Anesthesia Figure 9: INVOS™ and Hypoxemia

The oxygen dissociation curve illustrates the differential sensitivity of arterial and venous-dominant O2 saturation measures to small changes in oxygen partial pressure. This differential sensitivity helps explain the observation that cerebral rSO2 routinely detects developing oxygen imbalance before pulse oximetry.

Blood Product and Fluid Management Influences on rSO2 In its low-to-normal range (i.e., hematocrit <30%), hemoglobin and rSO2 are linearly related, while at a higher hematocrit, their relationship vanishes or may become inverse.37 This hemoglobin-dependency explains the oft-observed transient rSO2 decline at the onset of cardiopulmonary bypass. Initial passage of a crystalloid prime solution through the cerebral circulation momentarily lowers brain hemoglobin. It should also be appreciated that blood product administration will not invariably result in a rSO2 increase. Naidech et al. (2008) noted a wide variation in brain rSO2 responses to administration of packed red cells (Figure 10).8 -11-

Brain rSO2 Monitoring Before, During and After General Anesthesia Occasional declines in rSO2 should be expected, since stored red cells may have their oxygen-carrying capacity diminished by up to 90%.38 Figure 10: rSO2 and Blood Product Administration

The results of this small study illustrate the marked variation in rSO2 response to administration of two units of packed red cells (PRC). (The graph is based on data from Naidech et al. 2008.)

Anesthetic Influences on rSO2 Nearly two-thirds of cerebral oxygen is utilized to support interneuronal signal transmission. Thus, anesthetic influences on rSO2 depend on the neuropharmacologic properties of each agent and its dose. Volatile halogenated anesthetics, barbiturate hypnotics and propofol have profound cortical suppressant activity, while opioid analgesics and benzodiazepine amnestics generally do not. Rising doses of the powerful cortical suppressant anesthetics may increase rSO2 as oxygen consumption is decreased.39 Conversely, a sudden rSO2 decrease may signify decline in anesthetic effect (Figure 11).

-12-

Brain rSO2 Monitoring Before, During and After General Anesthesia Figure 11: INVOS™ and Anesthesia

At the onset of total cardiopulmonary bypass, brain responses to an initially unrecognized empty anesthetic vaporizer are shown. The increased cerebrocortical neuronal activity bilaterally increased EEG bispectral index (BIS) and decreased brain oxygen balance (i.e., rSO2). All values normalized with vaporizer refilling.

Brain Temperature Management Influences on rSO2 The neuroprotection afforded by hypothermia, in part, is due to reduced brain oxygen demand. However, individual patient responses to cooling vary widely. Thus, decreasing cranial temperature does not automatically ensure an adequately neuroprotective cerebral hyperoxic state.40 Wide variation in cooling response is due to patient-specific cerebral hemodynamics as well as mechanical perfusion strategy/ tactics.41 For example, the enhanced cerebral blood flow and cooling efficacy afforded by temperature-corrected (i.e., pH-stat) acid-base management improve neurologic outcome in both pediatric and adult patient cohorts undergoing cardiovascular surgery with deep-hypothermic circulatory arrest.29,42 Yet, the magnitude of hypothermic neuroprotection in individual patients depends on the bi-hemispheric responsiveness of cerebral arterioles to change in hydrogen ions and CO2.43

-13-

Brain rSO2 Monitoring Before, During and After General Anesthesia Cerebral oximetry gives anesthesia providers and perfusionists this key information at the start of surgery to guide patient care plans and optimize hypothermia management (Figure 12).41 Regional brain hypoperfusion associated with suboptimal cooling may lead to transient cerebral vasoparesis (i.e., vasoneural uncoupling).44,45 As a result, during re-warming an inverse relationship between brain temperature and rSO2 has been described in both adult and pediatric cardiac surgery patients. Prompt detection and treatment of this flow-metabolism mismatch may help avoid ischemic brain injury.46,47 Figure 12: INVOS™ and Brain Temperature

The left graph shows the expected inverse relationship between cranial temperature and brain rSO2. Note that the rSO2 rise reaches an asymptote at ~23° C and that further cooling doesn’t create more regional hyperoxia. At lower right, the graph illustrates an optimal cooling response conducted with pH-stat acid-base management in a patient with CO2-reactive cerebral arterioles. Marked hyperoxia prevented oxygen debt development during later total circulatory arrest. In contrast, the upper right graph depicts suboptimal cooling with alpha-stat acid-base management. Minimal hyperoxia resulted in a large oxygen debt during total circulatory arrest.

-14-

Brain rSO2 Monitoring Before, During and After General Anesthesia Supplemental Cerebral Perfusion Influences on rSO2 During deep-hypothermic circulatory arrest, rSO2 declines of >30% below baseline are highly associated with new neurologic deficit.48 Numerous studies have demonstrated that the “safe time” for systemic circulatory arrest may be extended with the use of bilateral rSO2 monitoring to ensure adequate retrograde or selective antegrade cerebral perfusion (Figure 13).49,50 Figure 13: Supplemental Cerebral Perfusion

Because of the hemodynamic characteristics of this acute Type I aortic dissection, upper body cerebral perfusion was mechanically supported through a right axillary perfusion cannula. However, cerebral perfusion was not symmetrically supported. Cooling and selective antegrade cerebral perfusion were more efficacious on the right side. rSO2 verified adequate right hemisphere perfusion throughout the emergent procedure.

-15-