Cerebral edema and intracranial hypertension continue to account for unacceptably high morbidity and mortality rates among patients with fulminant hepatic failure (FHF). Unfortunately, only a few small, randomized controlled trials dealing with these complications have been performed for human FHF, and results of most of the studies have not been confirmed. Instead, many clinical interventions continue to be instituted on an intuitive basis or simply because they are adopted from other critical care settings.1 The lack of evidence-based guidelines not only is a problem in the management of individual patients with FHF, but also constitutes a major problem when data from liver transplantation centers across the world is compared and discussed.
For many decades, it has been known that the brain in patients with liver failure is highly susceptible to hypoxia because of arterial hypotension, respiratory failure, or critically reduced cerebral blood flow (CBF) attributable to intracranial hypertension. Indeed, the primary goal in resuscitating patients with FHF includes optimal oxygenation and maintenance of cerebral perfusion pressure of more than 40 mmHg and of intracranial pressure (ICP) at levels lower than 20 mmHg.2 To achieve these goals, fluid therapy and inotropic support remain cornerstones in critical care management, but the type of fluid and inotrope, as well as their rate of delivery, remains essentially unknown. Yet, such basic knowledge is of potential clinical value because aggressive volume expansion and an increase in hydrostatic pressure in brain capillaries are assumed to aggravate edema. Two important clinical studies in this issue of HEPATOLOGY deal with these topics.3, 4
Terlipressin is a vasopressin analog that is increasingly used to reverse hepatorenal syndrome in patients with chronic liver disease. Renal failure also is common in FHF, and the value of terlipressin to prevent or treat renal failure is of considerable interest. Before launching a randomized controlled trial, Shawcross et al.3 carried out an important safety study, reported in this issue of HEPATOLOGY at the Scottish Liver Transplant Unit in Edinburgh that included six patients with FHF. Both CBF and ICP were recorded before and after intravenous injection of a minimal dose of terlipressin that did not affect systemic hemodynamics. In this elegant fashion, the potential confounding factor of increased blood pressure could be disregarded. They found that CBF increased approximately 17%, probably by stimulation of cerebral V2 receptors that dilate cerebral arterioles, findings that support observations made in an experimental study.5 The ICP increased in five of the patients from a median of 15 mmHg (range, 13–18 mmHg) to 20 mmHg (range, 16–23 mmHg). The authors recommend that extreme caution should be exercised if terlipressin is to be used in patients with FHF and low cerebral compliance as a result of brain swelling. This is another important observation that highlights the danger of the uncritical adoption of therapies from other critical illnesses in patients with FHF. However, the question arises of whether terlipressin should be contraindicated in patients with FHF. The answer probably depends on the situation, on the indication, and on the alternatives. Although it may be rational to refrain from using terlipressin in hepatorenal syndrome in patients with FHF, it is questionable whether terlipressin affects the brain more severely than norepinephrine, the use of which may also increase ICP.6 In fact, the use of terlipressin may be of some value in hypotensive FHF patients with severe lactic acidosis that is resistant to volume expansion and in FHF patients with catecholamine-resistant shock resulting from sepsis in order to initiate extracorporeal filtration. Although this study may not necessarily be the final verdict for the use of terlipressin in FHF, it does indicate that terlipressin should be used only during monitoring of ICP and under close observation for signs of cardiac, gastrointestinal, and extremity ischemia.
It is encouraging that the neurobiologic molecular changes7 and the exact sequence of pathophysiologic events that result in brain swelling8 are about to be unraveled in experimental models of FHF. It is still unclear how this knowledge will affect the future management of patients with FHF. Indeed, a wide variety of different strategies currently are used world wide to ameliorate cerebral edema in FHF, including head elevation, sedation, hyperventilation, diuretics, hypothermia, indomethacin, and ventriculostomy with cerebrospinal fluid drainage. However, each initiative has significant limitations or potential serious adverse effects. Mannitol currently is the most commonly used osmotic agent, although its exact mechanism of action is unknown. Mannitol has been associated with acute renal failure, hyperkalemia, hypotension, and rebound rises in ICP, and the increase in adverse effects beyond a serum osmolarity of 320 mOsm/L also limits its usefulness unless renal replacement therapy is instituted. Could hypertonic saline (HTS) be an alternative agent to control ICP in FHF? The group headed by Dr. Jules Wendon has extensive experience managing patients with FHF, and, in this issue of HEPATOLOGY,4 her group studied the efficacy of hyperosmolar therapy on ICP using a continuous infusion of 30% saline. This prophylactic intervention was hypothesized to be of value for both restitution and maintenance of an osmotic pressure gradient across the blood–brain barrier to avoid the development of brain edema. This approach seems rational because the blood–brain barrier is better able to exclude HTS due to a higher polarity and the presence of tight gap junctions, resulting in a reflection coefficient of 1.0 for sodium chloride, as compared with 0.9 for mannitol. Although only 30 patients with FHF were randomized to receive standard intensive care or HTS, one important point emerged: Inducing moderate hypernatremia with 30% saline lowered ICP from baseline and reduced the incidence of clinically significant intracranial hypertension in FHF.
The astrocyte is at the center of the pathophysiology of brain edema in FHF.7, 8 Why could swelling of astrocyte be affected by changes in the tonicity of plasma induced by HTS? To answer this question, it may be helpful to recall some basic pathophysiologic changes of the brain in FHF. Among other important functions, the astrocytes secure a constant composition of the extracellular fluid and are involved in the reuptake of neurotransmitters and ions from the synaptic cleft after depolarization. Ammonia entering the brain by diffusion and through ion channels, as seen in FHF,9 is neurotoxic and can induce seizures, probably by release of glutamate from astrocytes.10 Astrocytes are the primary site of ammonia detoxification in brain, where ammonia is eliminated via amidation of glutamate to form glutamine, or by transamination of pyruvate to alanine.11 Astrocytes—as other eukaryotic cells—cannot resist the membrane tension generated by disturbances in the osmotic pressure gradient across its cell membrane, and because glutamine and alanine are organic osmolytes, ammonia detoxification takes place at the expense of astrocyte swelling. However, dynamic processes are initiated rapidly to counteract such cell volume changes by restoration of an optimal osmotic balance between the intracellular and extracellular milieu.12 The intracellular osmotic load is decreased by the combined activation of chloride and potassium channels13 and later by release of organic osmolytes.14–16 These processes are highly energy dependent and require optimal metabolic conditions to operate. In FHF, such a favorable milieu is not present. In fact, the normal aerobic glycolytic pathway is compromised.11, 17, 18 The consequences not only are imminent “power failure” but also a failure of regulatory volume decrease of astrocytes. During HTS infusion, the extracellular osmolality increases as HTS extracts water from the extracellular space to the vascular bed. The resulting osmotic gradient between the astrocyte and extracellular space in this turn likewise extracts water from the astrocytes until a new equilibrium is obtained. This dehydrating effect of HTS is not energy dependent but simply modulates osmotic gradients within brain compartments.
Before HTS can be recommended as part of routine management of patients with FHF, some potential adverse effects of HTS should be considered. It is well known that multiple small hemorrhages, vein thromboses, and obliteration of sulci and fissures resulting in brain areas of encephalomalacia may develop after administration of HTS in the normal brain.19, 20 Such complications resulting from acute shrinkage of brain tissue and tearing of the falx and venous sinus were not noted in the present study. Another important concern is demyelination in the pons, that is, central pontine myelinolysis, because patients with FHF often suffer from hyponatremia.21 The pathophysiology of central pontine myelinolysis involves destruction of myelinated structures after a rapid rise in serum sodium. Human trials using HTS for ICP control in conditions other than FHF generally have been performed avoiding rapid rises in serum sodium; central pontine myelinolysis has not been noted, but should be considered when instituting hyperosmotic therapy in FHF. Other concerns are the potential induction of hyperchloremic acidosis, hematologic abnormalities, aggravation of coagulopathy, and red cell lysis20 in patients with FHF. So although the conclusions made by Murphy et al.4 seem valid and rational and undoubtedly will affect future handling of patients with FHF, it also raises many questions that need to be addressed in future randomized controlled trials targeting the brain.
As highlighted by the two papers in this issue,3, 4 both CBF and the colloid osmotic gradient across the blood–brain barrier are of central pathophysiologic importance to the development of brain edema and high ICP in patients with FHF.1, 8 The consequences of changing plasma tonicity or increasing CBF by administration of vasoactive drugs may be important to emphasize at a time when the number of proposed experimental treatments, using complex and costly artificial and bioartificial liver support systems to secure brain viability, has considerably increased. Before patients benefit from such advanced therapy and before randomized, controlled studies can be compared and validated among centers, it would be highly desirable that optimal critical care management is fully established. In this context, it is of considerable importance that critical answers are still being obtained by carrying out clinical research, even under the quite difficult circumstances of human FHF.