Brain death or brain dying?

For the past 50 years, the medical profession has understood ‘‘brain death’’ to represent the endpoint of a neuropathologic vicious cycle. An initial major brain injury sets off a mutually exacerbating cascade of cerebral edema, increased intracranial pressure, and decreased cerebral blood flow, which advances beyond some point-of-no-return to a state of no cerebral blood flow and total brain infarction (death of the brain, or ‘‘brain death’’). This pathophysiology was considered so well established that the Swedish Committee on Defining Death chose that very term—‘‘total brain infarction’’—as its official name for the condition, avoiding the semantic ambiguities inherent in the term ‘‘brain death.’’ Regardless of the name, this endpoint has been traditionally understood to correspond to brain-based statutory definitions of death everywhere in the world. In the United States, virtually all states have adopted some variation on the Uniform Determination of Death Act proposed by the President’s Commission in 1981, namely: ‘‘irreversible cessation of all functions of the entire brain, including the brain stem.’’ In clinical practice, both the irreversibility and totality of nonfunction are considered established by inference from sure knowledge that the neuropathologic vicious cycle has already reached its endpoint of no intracranial blood flow and total brain infarction. The past several decades have witnessed a search for practical clinical criteria to guarantee this inference in individual cases. Most developed countries have standardized protocols, which continue to evolve, although important controversies remain. But reality is rarely as straightforward as theory. In this issue of the Journal of Child Neurology, Suzuki and colleagues present evidence that ‘‘total brain necrosis might not be present in children at the time of clinical diagnosis of brain death.’’ Their hospital routinely follows serial serum levels of neuron-specific enolase, a standard marker of neuronal cell death, in children admitted with acute brain injuries. In a retrospective review spanning 14 years, the authors found 3 children who met clinical criteria for brain death following cardiopulmonary arrest and survived more than 2 months. For comparison, they selected 3 cases with cardiopulmonary arrest and poor outcomes short of brain death. If the vicious cycle ending in total brain infarction universally plays itself out over the course of hours or a few days at most, and if the clinical diagnostic criteria reliably determine that the endpoint has already been reached, then one would expect serum neuron-specific enolase levels to reflect an initial massive outpouring of neuron-specific enolase, followed by rapid exponential decay due to cessation of both intracranial blood flow and cerebrospinal fluid production and flow. The investigators report three findings: two interesting but not surprising, and one very surprising. (1) The children with brain death had higher peak neuron-specific enolase levels than the controls. (2) The time to peak neuron-specific enolase level was not significantly different between the groups (4-10 days vs 5-8 days). (3) Remarkably, all 3 children with brain death had persistent elevation of neuron-specific enolase at 4 weeks (>400 ng/mL) and 8 weeks (>50 ng/mL), in contrast to the two control survivors in the vegetative state, whose neuron-specific enolase decreased to <50 ng/mL within 4 weeks. The authors acknowledge the study’s obvious limitations resulting from its retrospective methodology and the small number of patients that was insufficient for statistical power. They also point out that in the context of brain death, the time course of serum neuron-specific enolase might not exactly reflect that of neuronal cell death. An additional caveat not mentioned in the article is that the patients with chronic brain death they followed may not be (and probably are not) representative of all patients with brain death, particularly those that systemically deteriorate to asystole within a few days despite all therapeutic measures. The study needs to be replicated and extended, although that is much easier said than done. Clearly, this is a wide-open new area for clinical research, for which Japan is uniquely suited, as it is probably the only country where extended intensive care unit support of patients with brain death still occurs, out of respect for traditional societal values. The neuron-specific enolase findings reinforce and contribute a novel temporal dimension to the growing neuropathologic evidence that at the time of fulfillment of clinical brain death criteria, the brain infarction may be far from ‘‘total.’’