To the Editor:
We read with great interest the article by Soehle et al., recently published in this journal. They used three devices employing near-infrared spectroscopy technology (i.e., the INVOS 5100 C and the INVOS 7100 [Medtronic Inc., USA] and the ForeSight Elite [Edwards Lifescience Corp., USA]) to measure regional cerebral oxygenation (rScO2) in a small group of 15 brain-dead subjects in the course of organ procurement. Like other researchers before, Soehle et al. initially observed near-normal rScO2 values that only markedly dropped after the aorta had been clamped to facilitate organ perfusion. The corresponding nadirs that were reached with each technology were 21 ± 8% (INVOS 5100C and INVOS 7100) and 40 ± 8% (ForeSight Elite), respectively. Assuming that the near-infrared spectroscopy signal in brain-dead subjects is predominantly generated by blood flow within the extracerebral space, the authors conclude that all three devices are “unable to detect severe cerebral hypoxia/anoxia under conditions of normal extracerebral oxygenation.” However, they could still be used “to optimize oxygen supply to the head.”
We commend the authors for their novel approach, using these monitors in the extremes of clinical medicine to evaluate potential shortcomings of their inherent technology. Nevertheless, we feel that it requires some additional explanatory notes to fully appraise the findings of this research. First of all, we challenge the authors’ postulation upon which the whole investigation is actually based. They maintain that subjects whose brain functions are irreversibly lost when pronounced brain dead lack any cerebral blood flow. Consequently, they argue that the close-to-normal rScO2 measures that are determined in these situations must primarily be due to a still unaltered perfusion of the scalp and the skull via branches of the external carotid artery.
The diagnosis of brain death, however, is primarily made by a clinical investigation and in many countries does not require detection of cessation of cerebral blood flow. Moreover, irreversible loss of neuronal activity can still occur under persistently inadequate cerebral blood flow not meeting the metabolic demand of cerebral neurons that inevitably causes necrosis despite continuous oxygen transport into the brain. Although the absence of cerebral blood flow is a reliable marker of brain death, a person can be brain dead in spite of maintained or resumed cerebral blood flow. Some neurologists even caution against the use of ancillary tests investigating cerebral perfusion, because in a substantial number of patients, patent cerebral blood flow or opacification of cerebral arteries at the time of brain death can be found. These inconsistencies may cast doubt on the performed clinical brain death exam and can lead to postponements in the procurement of donor organ. The reasons for a persistent or reconstituted cerebral blood flow can be manyfold. Those two patients without detectable cerebral blood flow the authors reported may therefore not be representative for the whole group. Without any information on the cause (primary or secondary brain injury), its initial location (supra- or infratentorial), on supportive measures in the intensive care unit (vasopressors to raise cerebral perfusion pressure, circulatory support with venoarterial extracorporeal membrane oxygenation, and others), and the time between cerebral damage, the diagnosis of brain death, and the final explantation of organs, any allegation on cerebral blood flow must remain hypothetical.
Furthermore, some potential flaw can be introduced by the different clamping sites of the aorta (ascending vs. descending aorta) because this can induce differing shifts in blood volume and affect compartmentalization of intracerebral blood. In addition, the fact that rScO2 does not decline to 15% (the lowest detectable value of the INVOS devices), which has been observed during cardiac arrest or to even lower values with the Foresight Elite, further indicates that rScO2’s background signal in these organ donors may additionally be governed by other contributing factors (e.g., cytochromes within damaged brain tissue, oxyhemoglobin contained in brain vessels that cannot unload due to nonmetabolizing surrounding tissue, and others).
Comparison of absolute rScO2 values derived from three different near-infrared spectroscopy monitors is a further drawback of the current study, especially because the INVOS devices received U.S. Food and Drug Administration clearance as trend monitors only. Conversely, all the factors that can interfere with the rScO2 signal will also be operative when these measures are used as a guide to improve brain oxygenation, particularly when intracerebral blood flow conditions, the integrity of cerebral autoregulation, and compartmentalization are unknown.
In conclusion, due to the many confounders and the lack of critical information, the study of Soehle et al. neither clarifies the role and the extent of extracranial rScO2 signal contamination during cerebral desaturations nor supports assumptions regarding lack of residual brain perfusion in brain-dead organ donors.