See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/7806352

Mannitol in intracerebral hemorrhage: A randomized controlled study

Article  in  Journal of the Neurological Sciences · August 2005

DOI: 10.1016/j.jns.2005.03.038 · Source: PubMed

CITATIONS READS

91 1,178

4 authors, including:

Usha Kant Misra Jayantee Kalita

Sanjay Gandhi Post Graduate Institute of Medical Sciences Sanjay Gandhi Post Graduate Institute of Medical Sciences
522 PUBLICATIONS   9,905 CITATIONS    501 PUBLICATIONS   8,991 CITATIONS   

SEE PROFILE SEE PROFILE

Sujit Kumar Mandal

25 PUBLICATIONS   581 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Hyponatremia following esclicarbazepine therapy View project

mind CSIR View project

All content following this page was uploaded by Jayantee Kalita on 19 June 2019.

The user has requested enhancement of the downloaded file.

 Journal of the Neurological Sciences 234 (2005) 41 – 45
www.elsevier.com/locate/jns

Mannitol in intracerebral hemorrhage: A randomized controlled study

U.K. Misraa,*, J. Kalitaa, P. Ranjana, S.K. Mandalb
a
Department of Neurology, Sanjay Gandhi Post Graduate Institute of Medical Sciences, Raebareily Road, Lucknow-226014, India
b
Department of Biostatistics, Central Drug Research Institute, Lucknow, India

Received 5 May 2004; received in revised form 15 December 2004; accepted 8 March 2005
Available online 3 June 2005

Abstract

Objective: To study the usefulness of mannitol in spontaneous intracerebral hemorrhage (ICH) patients.
Method: 128 CT proven supratentorial ICH patients within 6 days of ictus were randomized into study and control groups. The study group
received mannitol 20%, 100 ml every 4 h for 5 days, tapered in the next 2 days. The control group received sham infusion. Primary endpoint
was 1-month mortality and secondary endpoint functional disability at 3 months assessed by Barthel index score.
Results: There were 65 patients in study and 63 in control groups. The study and control groups were evenly matched regarding age,
Glasgow coma scale (GCS) score, Canadian Neurological Scale (CNS) score, pupillary asymmetry, pyramidal signs on non-hemiplegic side,
and location, midline shift and ventricular extension of hematoma. At 1 month, 16 patients died in each group. The primary and secondary
endpoints were not significantly different between the two groups.
Conclusion: Low dose mannitol does not seem to be beneficial in patients with ICH.
D 2005 Elsevier B.V. All rights reserved.

Keywords: Stroke; Intracerebral hemorrhage; Mannitol; Outcome; Prognosis; Randomized study

1. Introduction The wide popularity of mannitol in stroke is not based on

Mannitol is a hexahydric alcohol related to mannose and
was first isolated by Froust in 1806. Mannitol was first used
in medical practice in 1962 by Wise and Chater [1]. After
intravenous administration, mannitol is filtered by glomeruli
but not appreciably absorbed by tubules. Its high molecular
weight of 182 Da allows it to be rapidly cleared from brain
and CSF thus reducing rebound rise in intracranial tension.
Mannitol is commonly used to reduce intracranial tension in
patients with head injury, brain tumor and stroke. The
beneficial effects of mannitol are attributed not only to
lowering of intracranial pressure but also to its ability to
increase intracranial compliance and ability to reduce red
blood cell diameter by 15% thereby improving tissue
perfusion [2,3].
* Corresponding author. Tel.: +91 522 2668004 8; fax: +91 522 2668017.
E-mail addresses: ukmisra@sgpgi.ac.in, drukmisra@rediffmail.com
(U.K. Misra).
0022-510X/$ - see front matter D 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.jns.2005.03.038
published scientific evidence. In a study on seven patients
with cerebral infarction having midline shift, following
intravenous mannitol, there was clinical improvement as
evidenced by stroke scale in two, GCS score in three and
pupillary asymmetry in two patients. Midline shift did not
improve in any patient [4]. In another study on 20 patients
with ICH admitted in an intensive care unit with raised
intracranial tension, mannitol was not found to be beneficial
[5]. A Chinese study reported that mannitol was not superior
to FCMCK therapy in ICH patients [6]. A randomized
double blind placebo controlled trial in patients with ICH
using intravenous glycerol also did not prove beneficial [7].
In a SPECT study on ICH patients following a mannitol
bolus, blood flow did not change significantly compared to
placebo [8]. In spite of these reports mannitol is commonly
used in ICH. In a survey of 110 neurologists and physicians
from India all used mannitol in ICH [9]. The American
Heart Association and Stroke Council has recommended
mannitol in the management of ICH with raised intracranial
pressure [10]. In the available literature we did not find a
42 U.K. Misra et al. / Journal of the Neurological Sciences 234 (2005) 41 – 45

randomized controlled trial evaluating the role of mannitol was 1-month mortality. The cause of death, i.e. brain
in ICH. The present study therefore has been undertaken to herniation, pneumonia, septicemia, airway obstruction and
evaluate the role of mannitol in spontaneous supratentorial others, were ascertained as far as possible. Secondary
ICH patients. outcome measure was functional disability at the end of 3
months assessed by Barthel Index (BI) score. The outcome
was classified as poor if BI score was below 12, partial
2. Patients and methods BI = 12 –19 and complete BI = 20 [14]. Sample size was
calculated considering 20% reduction in mortality in the
The patients with CT proven primary supratentorial ICH study group compared to control. Forty patients in each
within 6 days of ictus were enrolled in a tertiary care group in each analysis were planned. The first interim
teaching hospital. The patients with renal and hepatic analysis was planned after recruiting 80 and final analysis
failure, hyperglycemia >250 mg%, past history of ICH, after 160 patients or 2 years whichever is earlier. If there
infratentorial hematoma, vascular malformation, tumor or were superiority of treatment or its absence, the study
anticoagulant bleed and those in whom consent could not be would be terminated. Statistical analysis was done using Z
obtained were excluded. test of proportion with an alternative that the treatment is
All the patients were subjected to detailed neurological effective. The power of test in the first stage was 0.86 and
evaluation. The consciousness was assessed by Glasgow in the final stages 0.96. The level of significance in the
coma scale (GCS) and severity of stroke by Canadian interim analysis was 0.01 and the final analysis 0.04. To
Neurological Scale (CNS) score [11]. Pulse, blood pressure, evaluate the significance of treatment one-tail test was
history of gastric hemorrhage and presence of surrogate used.
markers of raised intracranial pressure such as spontaneous
hyperventilation, papillary asymmetry and pyramidal signs
on the nonhemiplegic side were also noted. Cranial CT scan 3. Results
was carried out using a third generation spiral CT scanner
and 10 mm axial sections were obtained parallel to One hundred and sixty-four patients with ICH were
orbitomeatal line. The location of hematoma on CT scan, admitted during 2000 –2001; 36 of whom were excluded
its size, ventricular extension and midline shift were noted. because of renal failure in 5, diabetes in 1, recurrent ICH in
The size of hematoma was calculated as 4/3p A  B  C, 2, infratentorial location in 20, late admission 3 and lack of
where A is the largest diameter of hematoma in cm rounded consent in 5. In the study group there were 65 patients and
to nearest half cm, B is the diameter at 90- to A in cm in the control group 63. The age of the patients ranged
rounded to nearest half cm, and C is the number of between 35 and 80 (mean 56) years, 32 were females. The
parenchymal hematoma seen in 1 cm slice [12]. The
hematoma were classified into small (<20 ml), medium Table 1
(20 –40 ml) and large (>40 ml) [13]. The patients were Various clinical and radiological features in patients with intracerebral
managed conservatively and none underwent surgery. Anti- hemorrhage, receiving mannitol and placebo
hypertensive medication was prescribed if BP exceeded Patients details Mannitol (n = 65) Control (n = 63)
180/110 mm of Hg using atenolol or amlodepine. Ranitidine Age mean (range) years 55.0 (35 – 77) 57 (35 – 80)
or sucralfate were prescribed to all the patients. Patients’ Sex (female) 18 14
hydration was ensured by administering 2.0– 2.5 L isotonic Time of admission (days) 0.5 – 6 0.5 – 6
glucose saline daily. Caloric intake was maintained between Hypertension 51 49
Diabetes mellitus 9 7
1800 and 2000 Kcals. Blood sugar was monitored and if
Ischemic heart disease 5 6
exceeded 160 mg% insulin was administered. Seizure 2 1
The patients were randomized using random table Mean GCS (range) 9.6 (3 – 15) 9.4 (3 – 15)
numbers after getting informed consent from the first- Mean CNS score (range) 3.0 (2 – 10) 3.0 (2 – 9.5)
degree relatives if the patient was unable to give consent Pupillary asymmetry 7.0 9.0
Contralateral signs 28 28
because of aphasia or altered sensorium. The Institute’s
Septicemia (FDP+) 9 8
ethics committee approved the trial. The patients were Hematoma
randomized into study and control group. The random- Small 19 22
ization was done by one investigator (JK) and evaluation by Medium 23 21
another (PR). The hematoma was classified into small, Large 23 20
Hematoma location
medium and large and the randomization for each group was
Putaminal 42 42
done separately. The study group received mannitol 20% Thalamic 19 15
100 ml IV every 4 h for 5 days which was tapered in the Lobar 2 2
next 2 days as 100 ml every 8 h on sixth and every 12 h on Caudate 2 4
seventh day. The control group received a sham infusion of Midline shift 39 36
Ventricular extension 34 29
the same amount of normal saline. The primary endpoint
 U.K. Misra et al. / Journal of the Neurological Sciences 234 (2005) 41 – 45 43

patients were admitted within 6 days of ictus (mean-3, range
12 h– 6 days). 62 patients were admitted within 24 h, 30
between 24 and 72 h and 26 between 72 h and 6 days.
History of hypertension was present in 61 but on examina-
tion hypertension was detected in 100 patients. Twenty-nine
patients consumed alcohol. The GCS score ranged between
3 and 15 (mean 9.5) and three patients had generalized tonic
clonic seizure. Mean CNS score was 3.0 (range 2 – 10). The
majority of patients had medium or large hematoma with
dense hemiplegia and had severe stroke as assessed by CNS
score. Pupillary asymmetry was present in 16, spontaneous
hyperventilation in 38 and motor signs on the nonhemiple-
gic side in 56 patients. These features were present mainly
in large or medium size hematoma. On CT scan 41 patients
had small, 44 medium and 43 large hematoma. The location
of hematoma was putaminal in 84, thalamic in 34, lobar in
four and caudate in six patients. Midline shift was present in
75 patients, which was noted in patients with large and
medium size hematoma. Ventricular extension of hematoma
was present in 63 patients. The study and control groups
were evenly matched with respect to their clinical and
radiological parameters. The clinical and radiological
variables other than size of hematoma were not taken into
account before randomization; however these variables were
not different between the two groups (Table 1).
The profile of this trial is presented in Fig 1. Interim
analysis after recruiting 80 patients did not reveal any
significant difference in the primary endpoint between study
and control group. There were 10 deaths in study and 9 in
control group. By the end of second year although 164
patients were recruited, but 128 patients were eligible on
whom the final analysis was based. There was no difference
in primary outcome in mannitol and control group (Z = 0.12,
OR 1.042, CI 1.004 – 1.080, P = 0.80, power of test 80.36%).
16 deaths occurred in placebo and control groups respec-
tively. Three-month outcome between study and control
group was also not significantly different (X 2 = 2.83, df = 2,
P = 0.25). In study group, 23 patients had poor, 18 partial
and 8 complete recovery; and in control group 18 poor, 20
partial and 9 had poor recovery. The comparison of
mortality and 3-month outcome in study and control group
is presented in Table 2.

Registered/ Eligible pts (N=164)

Not randomized (n=36)

Reasons -(renal failure 5,diabetes1, recuurent ICH 2,infratentorial location 20,

late admission 5 and lack of consent 3

Randomization (n=128)

Group I (Mannitol) Group II (Normal saline)

Received standard intervention as Received standard intervention as allocated

allocated (n=65) (n=63)

Did not receive standard intervention as Did not receive standard intervention as

allocated (n=0) allocated (n=0)

Followed up (n=65) Followed up (n= 63)

Timing of primary and secondary outcome: Timing of primary and secondary outcome:

1 month mortality, 3 month outcome 1 month mortality, 3 month outcome

Withdrawn (n=0) Withdrawn (n=0)

Lost to followup (n=0) Lost to followup (n=0)

Other (n=0) Other (n=0)

Completed trial (n=65) Completed trial (n=63)

Fig. 1. Profile of randomized controlled trial to evaluate the role of mannitol in ICH.
44 U.K. Misra et al. / Journal of the Neurological Sciences 234 (2005) 41 – 45

Table 2
Outcome of patients with ICH receiving mannitol and placebo
Outcome Study (65) Control (63)
Small (n = 19) Medium (n = 23) Large (n = 23) Small (n = 22) Medium (n = 21) Large (n = 20)
Death 0 2 14 0 2 14
Outcome
Poor 2 12 9 1 11 6
Partial 10 8 0 14 6 0
Complete recovery 7 1 0 7 2 0

Comparison of death in different sizes of hematoma was 4. Discussion

also not significance. In the small hematoma there was no
death (Z = 0, P = 0.99), in median size hematoma there were
2 deaths each in mannitol and control group (Z = 0.09,
P > 0.8, OR 0.905, CI 0.862 – 0.948, power of test of
79.8%); and in the large hematoma there were 14 deaths
in each group (Z = 0.269, P > 0.50, OR 0.666, CI 0.593 –
0.739). Three-month outcome in different subgroups was
also not significant in mannitol and control group. In the
large hematoma all the surviving patients had poor outcome
in both the groups. Evaluating the primary outcome of
patients as per the time of recruitment (within 24 h, 2 – 3
days and 4 – 6 days) was not significantly different
(X 2 = 3.73, df = 2, P = 0.16). The secondary outcome that is
poor (X 2 = 0.8, df = 2, P = 0.96), partial (X 2 = 0.17, df = 2,
P = 0.92) and complete recovery (X 2 = 0.12, df = 2, P = 0.94)
was not significantly different in control and study group.
On analyzing the timing of death, 10 patients in each group
died within 7 days of admission and the cause of death was
brain herniation in all except one who had pneumonia and
septicemia. After 7 days of admission 6 patients in each
group died and the death was due to infection in 11 and
cardiac arrhythmia in 1 (Fig 2). The analysis of death in first
7 days and later did not show any difference. Two patients
died after 2 months and have been included in poor outcome
group.
In our study, no serious adverse effect necessitating
withholding the treatment was noted. All the patients who
survived received full treatment. Hyponatraemia was noted
in two patients in the study group and six patients in the
control group (X 2 = 1.302, df = 1, NS).
12
Number of patients
Our study reveals that in ICH patients, there was no
significant difference in mortality or 3-month outcome in
mannitol and placebo group. In earlier studies on large
hemispheric infarction, mannitol aggravated the midline
shift and resulted in neurological deterioration [15,16]. In
another study on six patients with middle cerebral artery
territory infarction with midline shift, administration of 1.5
g/kg mannitol resulted in significant reduction of brain
volume at 50 – 55 min. The non-infarcted hemisphere shrank
more than the infarcted hemisphere. The clinical implica-
tions of this study however could not be ascertained [16].
Single dose mannitol in another study also failed to
demonstrate significant alteration in midline shift or neuro-
logical status in seven patients with large cerebral infarction
[4]. Mannitol failed to improve the outcome of 20 patients
with intracerebral hemorrhage admitted in an intensive care
unit [5]. The lack of benefit of mannitol in patients with ICH
may be due to associated alteration of blood – brain barrier
around the hematoma. Progressive entry of administered
solute from plasma into the damaged brain may increase the
osmolality which may draw water from plasma into brain
thereby defeating the intended purpose for which mannitol
was administered.
In our study, we have defined primary endpoint as 1-
month mortality. In the early stage of ICH, death may be due
to the effects of brain herniation however later it may be
influenced by aspiration, infection such as urinary and
pulmonary, septicemia and pulmonary thromboembolism
[17]. In the present study 10 patients each in the mannitol
and control groups died in the first 7 days, whereas 6
patients each died in both study and control groups after 7
days. The earlier death was most likely due to herniation

10
8
while latter was contributed by septicemia though no
6
Mannitol autopsy studies were carried out. The analysis of death in
Saline first 7 days of admission and after did not show any
4
2
difference between the two groups.
0
In our study, higher frequency of hyponatremia was
noted in the control group compared to those receiving
s

ys

ia
n
ay

io

io

tio

hm
da

mannitol. Disturbance of water balance and hyponatremia

ct

at
7d

ec
fe

>7

yt
ni
0-

rh
er
In

In

occurs in one-third of patients with massive stroke

H

Ar

Cause of death in different time scale particularly subarachnoid haemorrhage [18]. These compli-
Fig. 2. The cause of death in intracerebral hemorrhage patients was mainly cations are most likely to develop in critically ill patients
brain herniation within 7 days of admission and infection after 7 days in with large hemorrhages. Hyponatraemia in ICH is attributed
both mannitol and control groups. to syndrome of inappropriate secretion of antidiuretic
 U.K. Misra et al. / Journal of the Neurological Sciences 234 (2005) 41 – 45
hormone (SIADH) and cerebral salt wasting due to atrial
natriuretic factor [19]. The lower frequency of hyponatremia
in patients on mannitol therapy in the present study could be
due to osmotic diuretic effect of mannitol and resulted in
loss of both water and sodium, which compensated each
other [20,21]. Higher frequency of hyponatremia in the
control group could be due to lack of diuretic effect of
normal saline and additional effects of antidiuretic hormone
and atrial natriuretic factor resulting in natriuresis.
In the present study, the dose was 100 ml of 20%
mannitol every 4 h for 5 days tapered in the next 2 days. We
did not measure plasma osmolality in our patients, but the
dose response effects have been reported. Mannitol in a dose
of 1 g/kg administered over 10 –15 min results in raised
serum osmolality of 20 – 30 mosM/l, which returns to
normal in about 3 h [22]. We have chosen the small
frequent dosage of mannitol (0.25 – 0.5 g/kg) as per
published recommendations to avoid rebound phenomena
[10]. The effect of higher dose of mannitol and admin-
istration at early stage of stroke needs further studies.
The bias in the present study was eliminated by
randomization by one and evaluation by another investi-
gator. The outcome was evaluated by robust endpoint of 1-
month mortality and 3-month ADL score which is quite
objective. The final analysis was based on 128 patients
instead of 160, because of time constraint. This has reduced
the power of test from 0.96 to 0.90 although the conclusion
remained the same. Intracranial pressure monitoring and
plasma osmolality measurement have not been done in our
study, which could have provided additional objective
documentation of the mechanism of action of mannitol.
It can be concluded that low dose mannitol within 6 days
in patient with supratentorial ICH does not seem to be
beneficial, though it does not produce any serious side
effects. Effect of higher dose of mannitol in ICH needs
further study.
Acknowledgement
We acknowledge Rakesh Kumar Nigam for the prepara-
tion of the manuscript.
References
[1] Wise BL, Chater M. The nature of hypertonic mannitol solution in
decreasing brain mass and lowering of cerebrospinal fluid pressure.
J Neuro Surg 1962;10:1038 – 43.
[2] Mendehow AD, Teasdale GM, Russel T, Floof J, Paterson J, Murhay
GD. Effect of mannitol on cerebral perfusion pressure in human head
injury. J Neuro Surg 1985;63:43.
View publication stats
45
[3] Muizelaar JP, Wei EP, Kontos HA, Becker DP. Mannitol causes
compensatory vasoconstriction and vasodilatation in response to blood
viscosity changes. J Neuro Surg 1983;59:822.
[4] Manno EM, Adams RE, Derdeyn CP, Powers WJ, Diringer MN. The
effect of mannitol on cerebral edema after large hemispheric cerebral
infarction. Neurology 1999;52:583 – 7.
[5] Gigliuto CM, Stone KE, Algus M. the use of mannitol in intracerebral
bleed in the medical ICU. N Engl J Med 1991;88(1):48 – 51.
[6] Wang J. Medical treatment of acute haemorrhagic stroke—observation
of 44 cases with FCMCK therapy. Zhongguoyi Xue Ke Xue Yuan Xue
Bao 1990;12(1):61 – 7.
[7] Yu YI, Kumana CR, Lauder IJ, Cheung YK, Chan FL, Kou MHV.
Treatment of acute cerebral hemorrhage with intravenous glycerol: a
double blind placebo controlled randomized trial. Stroke 1992;
23:967 – 71.
[8] Kalita J, Misra UK, Ranjan P, Pradhan PK, Das BK. Effect of mannitol
on regional cerebral blood flow in patients with intracerebral
haemorrhage. J Neurol Sci 2004;224:19 – 22.
[9] Kalita J, Misra UK, Ranjan P. Prescribing pattern of antioedema
therapy in stroke by neurologists and general physicians. Neurol India
2004;52:186 – 8.
[10] Broderick JP, Adams Jr HP, Barson U, Feinberg W, Relfman E, Grotta
J, et al. Guidelines for the management of spontaneous intracerebral
haemorrhage a statement for health care professional from a special
writing group of the stroke council. American Heart Association.
Stroke 1999;30:905 – 11.
[11] Cote R, Battista RN, Wolfson C, Boucher J, Adam J, Hachinski V.
The Canadian neurological scale—validation and reliability assess-
ment. Neurology 1989;39:638 – 43.
[12] Broderick JP, Brott TG, Duldner JE, Tomsick T, Huster G. Volume of
intracerebral haemorrhage. A powerful and easy to use predictor of 30
days mortality. Stroke 1993;24:987 – 93.
[13] Caplan LR. Intracerebral haemorrhage. Lancet 1992;339:656 – 8.
[14] Misra UK, Kalita J. Evoked potential studies in thalamic haemorrhage.
Clin Neurol Neurosurg 1996;98:291 – 8.
[15] Frank JL. Large hemispheric infarctions, deterioration and intracranial
pressure. Neurology 1995;45:1286 – 90.
[16] Videen Tu, Zazulia AR, Manno EM, Derdeyn CP, Adam RE, Diringer
MN, et al. Mannitol bolus preferentially shrinks non infarcted brain in
patients with ischemic stroke. Neurology 2001;57:2120 – 2.
[17] Bamford J, Dennis M, Sandercock P, Burn J, Warlow C. The
frequency, cause and timing of death within 30 days of a first stroke:
the Oxfordshire community project. JNNP 1990;53:824 – 9.
[18] Adam Jr H, Love B. Medical management of aneurysmal SAH. In:
Barnett H, Mohr J, Stein B, Yatsu F, editors. Stroke—pathophysiol-
ogy, diagnosis and management. 3rd edition. Philadelphia’ Churchill
Livingstone; 1998. p. 1243 – 62.
[19] Hanson S, Degraba T, Villar-Cordova C, Yatsu F. Medical complica-
tion of stroke. In: Barnett H, Mohr J, Stein B, Yatsu F, editors.
Stroke—pathophysiology, diagnosis and management. 3rd edition.
Philadelphia’ Churchill Livingstone; 1998. p. 1121 – 32.
[20] Wijman C, Kase C. Intracerebral haemorrhage—medical consider-
ation. In: Barnett H, Mohr J, Stein B, Yatsu F, editors. Stroke—
pathophysiology, diagnosis and management. 3rd edition. Phila-
delphia’ Churchill Livingstone; 1998. p. 1359 – 72.
[21] Jackson E. Diuretics. In: Hardman J, Limbard L, Molinoff P, et al,
editors. Goodman and Gilman’s—the pharmacological basis of
therapeutics. 9th edition. Mc Graw Hills; 1996. p. 685 – 713.
[22] McGraw CP, Alexander EP, Howard C. Effect of dose and dose
schedule on the response of intracranial pressure to mannitol. Surg
Neurol 1978;10:127 – 30.