A Case of Persistence of Normal Tissue Oxygenation Monitored by Near-Infrared Spectroscopy (NIRS) Values Despite Prolonged Perioperative Cardiac Arrest

Background

Cerebral oximetry is a non-invasive monitoring method that can include the use of near-infrared spectroscopy (NIRS) technology to provide an estimation of the regional cerebral tissue oxygen saturation (rSO2). NIRS is simple to use and is not operator-dependent when compared with other brain monitoring devices, although an understanding of the technology behind the measurements is essential for clinicians to correctly interpret NIRS values. Extensive reviews of NIRS technology and the limitations of these devices in estimating cerebral oxygenation have previously been published [1–3].

Infrared radiation (IR) is electromagnetic radiation (EMR) and has longer wavelengths than visible light (700–950 nm) and passes through tissue due to the relative transparency to light in this range of wavelengths. Once emitted, the light is absorbed by surrounding tissues according to their molecular properties, each having distinct absorption spectra in the NIRS. These absorbing molecules, called chromophores, include oxyhemoglobin (HbO2) and deoxyhemoglobin, with a maximal absorption spectrum between 700–850 nm [4]. All NIRS-based monitoring techniques rely on a measure of optical attenuation, caused by light absorption and scattering. Using complex algorithms, NIRS monitors convert the measured changes of light attenuation into a numerical and trend display signal of rSO2.

One of the main challenges in using NIRS technology is the elimination of artifacts originating from the extra-cranial tissue. Two photodetectors are positioned at two different distances from the light source to differentiate the contribution of superficial tissue from brain tissue to the rSO2 value. Assuming that photons transmitted through a sphere (the skull) will follow an elliptical path, the near detector measures the photons from the superficial tissues and the far detector measures the photons that come from the brain. Ideally, the subtraction between the two detectors allows an accurate measurement of brain tissue oxygenation [5].

Also, as between 70–80% of the blood in the brain is in the venous compartment, NIRS values mainly reflect venous blood oxygenation [6]. Also, there is an important intra-individual and inter-individual baseline variability. Normal values for NIRS values are between 60–75% but physiological values of 55–60% have been reported in some cardiac patients [7]. Therefore, trends in values are considered more important in clinical practice than the absolute values [8].

The use of NIRS, especially in vascular and cardiac surgery, has been extensively described [9]. While the current evidence does not show that the use of NIRS improves neurological outcomes after cardiac surgery [10], it has been shown that NIRS allows for the intraoperative assessment of the limits of cerebral autoregulation [11], and might predict postoperative cognitive dysfunction [12] and neurological complications [13,14].

Most recently, attention has been given to the use of NIRS during cardiopulmonary resuscitation (CPR) and the post-cardiac arrest period, as an indirect indicator of impaired cerebral perfusion or low arterial oxygen content. There is evidence in the literature that rSO2 has potential as a neuro-monitoring tool during cardiopulmonary resuscitation (CPR) and can also be used to guide neuroprotective therapeutic strategies during cardiac arrest [15–21]. Several studies have documented the feasibility of cerebral saturation measurements using NIRS during CPR [16]. These previous studies have supported the use of NIRS values as a guide to the quality of CPR maneuvres [15,17] as a predictor of the return of spontaneous cardiopulmonary circulation [16,18,21], for monitoring brain oxygenation during patient transport [21] and as an indicator to post-arrest neurologic outcome [18–20]. The reliability of continuous cerebral oximetry to reflect the real-world clinical situations is important for informed clinical decision-making. The influence of certain clinical situations that may distort values must be taken into account by the practitioner [22].

A case report is presented that demonstrates the persistence of normal tissue oxygenation monitored by NIRS values despite a prolonged perioperative cardiac arrest with normal continuous cerebral oximetry values based on NIRS monitoring despite the prolonged absence of systemic blood flow.

Case Report

A 65-year-old man, with peripheral artery disease, presented with sudden onset of jaw and neck pain followed by facial seizures and a 5-minute loss of consciousness, which was witnessed by his wife. On the arrival of the emergency services to his home, the patient was noted to have dysarthria, left facial ptosis, left hemiplegia, and hypotension of 75/50 mmHg (measured in both arms). After fluid administration and treatment with an adrenaline bolus of 30 µg, the patient was rapidly transferred to the Emergency Department at the University Hospital.

On initial arrival at the Emergency Department, the patient’s blood pressure was 60/40 mmHg. Computed tomography (CT) angiography showed a Stanford type A aortic dissection, beginning at the coronary artery level and extending to the brachiocephalic trunk and the right common carotid artery with a 16 mm hemopericardium (Figure 1). The patient was immediately transferred to the operating room.

During preparation for anesthesia the patient had a second seizure followed by cardiac arrest with pulseless electrical activity (PEA). Cardiopulmonary resuscitation (CPR) was immediately begun and the trachea was intubated. The patient was transferred to the operating room. The femoral artery and vein were cannulated to initiate an extracorporeal circulation, with a delay of 20 mins. A median sternotomy was performed to allow the pericardium to be opened. The surgeons observed a large ascending aortic tear of 5 cm associated with approximately 3 liters of blood pooled in the left pleural space. Continuous intra-thoracic hemorrhage with the absence of sufficient venous return to the extracorporeal pump, despite aggressive transfusion of fluids, led to a decreased mean arterial pressure (MAP) (<30 mmHg), anemia (Hb of 63 g/l), reduced blood oxygen saturation (SpO2 of 65.9%), and hypocapnia (PaCO2 of 2.82 kPa). After 21 minutes of inefficient extracorporeal circulation, and a total of 44 minutes of extremely low systemic blood perfusion values assessed clinically and with concomitant monitoring devices, the attending physicians made a decision to discontinue therapeutic interventions. The patient died in the operating room soon afterward.

In this case, the near-infrared spectroscopy (NIRS) sensor, the FORE-SIGHT ELITE Absolute Tissue Oximeter (Fore-Sight, London, UK) was applied to the forehead shortly after the initiation of CPR. The initial rSO2 value was 60% and rose to 70% at the initiation of the extracorporeal circulation. After the termination of resuscitation and extracorporeal circulation, with no systemic blood flow, and a clearly cyanotic patient, rSO2 values remained above 68%. Patient oxygen saturation (SpO2) during the period of extracorporeal circulation and at the time of the decision to withhold any further therapeutic action was approximately 75%. The paradoxical NIRS values, although confusing, did not influence the final decision to end treatment.

Discussion

Near-infrared spectroscopy (NIRS) has been used to acquire real-time information on the cerebral circulation and metabolism in various clinical conditions of cerebral ischemia-reperfusion [23–26]. Adequate cerebral oxygenation depends upon sufficient cerebral blood flow and oxygen content. Any factors disturbing either of these will result in a reduction in cerebral oxygenation and a reduction in cerebral oximetry values. We report a case of normal continuous cerebral oximetry values based on NIRS monitoring despite the prolonged perioperative cardiac arrest.

Any factors disturbing either of these will result in a reduction in cerebral oxygenation and a reduction in cerebral oximetry values. Low cerebral regional oxygen saturation has been reported in clinical conditions associated with compromised cerebral perfusion such as unstable hemodynamics, inadvertent carotid compression, displacement of the extracorporeal membrane oxygenation or extracorporeal circulation (ECMO/ECC) cannula, surgical procedures such as deep hypothermic circulatory arrest or aortic/carotid cross-clamping [27–32]. Also, low rSO2 values were also found in a patient with global hypoxia or increased cerebral oxygen consumption [22,33].

In the case described, the impaired return flow to the venous cannula of extracorporeal circulation was associated with decreased arterial pressure, major hemorrhage, anemia, reduced blood oxygen saturation and hypocapnia. These different events should have been followed by a significant decrease in cerebral oximetry since cerebral saturation values are a representation of oxygenation in the cerebral microcirculation. Factors impairing oxygen delivery to the brain associated with extra-corporeal circulation have been described during cardiac surgical procedures. Among others, impaired flow through the arterial or venous cannula of extracorporeal circulation, as well as micro-embolism and macro-embolism, hemodilution and severe anemia, can all effect NIRS values [34,35].

Hypotheses that might explain the situation of normal NIRS values despite prolonged perioperative cardiac arrest include the absence of cerebral metabolism during the event, preservation of the extracranial circulation, or a maintained residual cerebral perfusion. The non-metabolizing tissue hypothesis is based on the lack of oxygen required by ischemic or necrotic cerebral tissue. It would be expected that if there is reduced systemic and cerebral blood flow, a rapid decrease in arterial and central venous blood oxygenation would automatically occur. Blood oxygenation would be compromised and oxygen extraction would rise. Therefore, NIRS monitoring would then measure a concomitant reduction in cerebral blood oxygenation. In the case presented in this report, the seizures experienced by the patient might have been symptoms of impaired cerebral circulation as these were among the first symptoms that occurred when at home. The potential rapid development of altered brain tissue failing to extract oxygen from the cerebral blood could be a contributing factor, as the NIRS correctly measures fully oxygenated blood pooled in a non-metabolizing and non-extracting altered brain tissue areas, despite the absence of cerebral blood flow.

Forensic research groups have worked on HbO2 in cadavers showing that a non-metabolizing brain can have a high oxygen content because of sequestered cerebral venous blood in capillaries and venous capacitance vessels [36]. This was confirmed by oximetry on blood samples of 214 forensic autopsy cases, including blood from the sagittal venous sinus of the dura (cerebral venous blood), where HbO2 values up to 95.1% were reported. HbO2 values in these cases depended on total hemoglobin content, cause of death, and cadaver storage conditions [37]. A case report on a living patient with a right internal carotid artery occlusion and an infarct in the middle cerebral artery territory with confirmed absence of perfusion to the infarct showed rSO2 values between 60–65%. This nonmetabolizing tissue hypothesis may offer a plausible explanation for the normal rSO2 values in this case [38].

An alternative explanation in this patient’s case was the persistence of an extracranial circulation. This phenomenon was first described on a brain-dead patient without blood flow to the proximal or middle cerebral arteries using transcranial Doppler (TCD), but with normal rSO2 values [39]. As the SpO2 in this patient was 98–99% throughout the study, it was concluded that the discrepancies occurred because the NIRS device was significantly influenced by the oxygenation of the scalp and skull [39]. This explanation was confirmed in a similar study, which also showed a concomitant decrease in rSO2 with the SpO2 after discontinuation of mechanical ventilation [40]. Ultrasound-tagged near-infrared spectroscopy (UT-NIRS) was also unable to differentiate signals from extracranial structures [41]. This explanation may be helpful in our case, as the patient had a SpO2 of approximately 75% during extracorporeal circulation and after its discontinuation, when the NIRS value was still 68%, suggesting that some residual perfusion was maintained in cerebral circulation or that there was no cerebral circulation, and the NIRS values originated solely from extracerebral tissue.

The presence of residual global circulation is a possible explanation in the present case. However, extracorporeal circulation was unable to achieve adequate systemic blood flow, and cardiopulmonary resuscitation (CPR) was not feasible once the chest was opened for surgery. It might be hypothesized that the low degree of metabolism by the brain enabled minimal oxygenated blood circulation or possible pooling of blood in the brain even though, with inadequate perfusion, there were normal values from cerebral oximetry. It should be noted that in the context of resuscitation, massive hemorrhage, and emergency surgery, it would be extremely complex to obtain a direct measurement of cerebral blood flow to exclude this possibility.

Despite the limitations of the use of NIRS in situations where there may be brain death [41], Weigl et al. recently reported a promising technique for the confirmation of cerebral circulation cessation using bolus tracking of an optical contrast agent, indocyanine green, and time-resolved NIRS [42]. As this case report has shown, NIRS is a promising method that opens a new era of research in the assessment of the cerebral circulation in critical situations.

Conclusions

Near-infrared spectroscopy (NIRS) is a powerful non-invasive technology that has an established role in perioperative care. However, an understanding of the underlying technology is essential, as specific clinical circumstances could lead to confusing values that do not match the clinical context. Further research using optical contrast agents could enhance the ability of NIRS devices to assess cerebral circulation in complex clinical situations.