Osmotherapy: A Call to Arms
To the Editor:
We read with interest the report by Bereczki and colleagues,1 who failed to find evidence supporting the use of mannitol in acute stroke after a thorough search of the available medical literature. The authors found only one “unconfounded” randomized clinical trial evaluating mannitol in stroke. Not surprisingly, mannitol holds a preeminent position in the limited clinical armamentarium of neurological critical care, despite the virtual absence of well-designed studies addressing its role in stroke. The rationale of osmotherapy in brain ischemia is mainly inferred from two circumstances: (1) brain edema is intricately involved in cellular damage during ischemia and (2) the blood-brain barrier (BBB) remains basically intact during the first few hours after the onset of ischemia. Perivascular glial edema occurs within minutes after the induction of ischemia and is followed by neuronal edema.2 Due to the small pore size of cerebral endothelium, the BBB has a very low hydraulic conductivity, which limits the passage of small molecules and even restricts the permeability of water.3 A hydrostatic pressure gradient across the BBB develops early after ischemia, leading to the accumulation of interstitial fluid.4 In addition, an osmotic gradient develops 3 to 6 hours after ischemia.4 Nevertheless, experiments using Evans blue and Na-fluorescein fail to demonstrate leakage of dye into brain parenchyma, indicating that BBB remains intact during the initial phase of ischemia.5
The choice of an osmotic agent should not remain a matter of tradition, since some rational points can be made. The ability of solutes to cross a membrane can be expressed by the osmotic reflection coefficient (ORC). The ORC value ranges from 1 for an “ideal” osmotic agent to 0 for substances that provoke no osmotic drive.6 The ORC of mannitol is estimated to be close to 0.5, while that of NaCl is almost 1. However, these parameters were mostly obtained in experiments using circulatory circuits different than the brain, like kidney and muscle. It is not known whether these values can be applied to the BBB. The ideal osmotic agent should not accumulate in the edematous brain parenchyma, since it may create more edema and “rebound phenomenon.”6
Although no controlled trials have investigated hypertonic saline in acute stroke, numerous controlled trials have compared glycerol with placebo and found no net effect on outcome (Table⇓). All of these trials, however, share the same essential flaws: none has demonstrated an actual change in blood osmolarity, blood viscosity, or intracranial pressure (including midline shift measurements on imaging studies).7 Another pertinent effect of glycerol in acute stroke, hyperglycemia, has not been systematically addressed. Finally, the number of enrolled patients has been small, compromising the statistical power of the studies.7
Despite its early failures, osmotherapy holds a potential that should be rigorously investigated. Future trials should employ a rational, physiological approach to therapy as opposed to arbitrary, nonphysiological regimens. They should use enough doses of agents to exert an efficient osmotic drive, guided by physiological parameters such as intracranial pressure monitoring, evidence of decreased lateral shift of the brain parenchyma on neuroimaging, and serial determinations of serum osmolarity, viscosity, glucose, sodium, BUN, and creatinine.
- Copyright © 2001 by American Heart Association
Bereczki D, Liu M, Fernandes do Prado G, Fekete I. Cochrane Report: A systematic review of mannitol therapy for acute ischemic stroke and cerebral parenchymal hemorrhage. Stroke. 2000;31:2719–2722.
Hatashita S, Hoff JT. Role of blood-brain barrier permeability in focal ischemic brain edema. In: Long DM, ed. Brain Edema: Pathogenesis, Imaging and Therapy. Advances in Neurology. New York, NY: Raven Press; 1990:327–333.
Bourke RS, Kimelberg HK, Nelson LR, Barron KD, Auen EL, Popp AJ. Biology of glial cells in experimental brain edema. In: Cervos-Navarro J, Ferszt R, eds. Brain Edema. Advances in Neurology. New York, NY: Raven Press; 1980:99–109.
Hatashita S, Hoff JT, Salamat SM. An osmotic gradient in ischemic brain edema. In: Long DM, ed. Brain Edema: Pathogenesis, Imaging and Therapy. Advances in Neurology. New York, NY: Raven Press; 1990:85–92.
Reulen HJ, Tsuyumu M, Prioleau G. Further results concerning the resolution of vasogenic brain edema. In: Long DM, ed. Brain Edema: Pathogenesis, Imaging and Therapy. Advances in Neurology. New York, NY: Raven Press; 1990:375-381.
Restrepo L, Balish M. Role of glycerol in cerebrovascular disease. Neurologist. 2000;6:338–347.
We thank Drs Restrepo and Silverman for drawing attention to several pathophysiological aspects of osmotherapy and to some clinical trials on glycerol in acute stroke. We fully agree with their statements that osmotherapy might have a potential in stroke therapy, that this therapeutic option should be rigorously investigated, and that the clinical application of an osmotic agent should not remain a matter of tradition. In fact, these were among our reasons for performing a systematic review of unconfounded randomized trials of mannitol therapy in acute stroke.R1
As far as glycerol is concerned, in addition to the trials Drs Restrepo and Silverman referred to in their letter, several other randomized trials have been performed. Eleven randomized trials are included in the systematic review of Righetti et al,R2 and this review comes to a slightly different conclusion than that of Drs Restrepo and Silverman. Righetti et al found that among patients with definite or probable ischemic stroke, glycerol was associated with a significant reduction in the odds of death during the scheduled treatment period, although this favorable outcome could not be detected at the end of the scheduled follow-up period. They recommended that because of the small number of patients, the results should be interpreted cautiously, and emphasized that the lack of evidence of benefit in long-term survival does not support the routine or selective use of glycerol treatment in patients with acute stroke.
The fact that we could not identify a sufficient amount of evidence to prove that mannitol influences short- or long-term outcome after stroke should not be considered as a proof against the effect (beneficial or harmful) of mannitol in acute stroke. It simply means that there is not enough evidence from randomized controlled trials to come to a final conclusion, and therefore more studies are needed. We agree with the recommendations of Drs Restrepo and Silverman that when designing such trials the applied optimal dose should be based on previous pathophysiological observations and neuroimaging should be a part of the study protocol. Because midline shift was not altered by mannitol in patients with cerebral edema after a large hemispheric infarction,R3 measuring changes in lateral shift does not seem to be a useful method for monitoring osmotherapy. A noninvasive transcranial Doppler monitoring of osmotherapy might be promising.R4 Although it might be useful to monitor physiological and serum parameters during osmotherapy, in clinical trials the primary outcome measures that are important for the patients and their families—such as survival, dependency, and quality of life—should be evaluated.
Bereczki D, Liu M, do Prado GF, Fekete I. Cochrane report. A systematic review of mannitol therapy for acute ischemic stroke and cerebral parenchymal hemorrhage. Stroke. 2000;31:2719–2722.
Righetti E, Celani MG, Cantisani T, Sterzi R, Boysen G, Ricci S. Glycerol for acute stroke (Cochrane review). In: The Cochrane Library, issue 4, 2000. Oxford, UK: Update Software.
Manno EM, Adams RE, Derdeyn CP, Powers WJ, Diringer MN. The effects of mannitol on cerebral edema after large hemispheric cerebral infarct. Neurology. 1999;52:583–587.