| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, Baskent University Faculty of Medicine, Fevzi Cakmak
Cad. 10. Sok No. 45, 06490 Bahçelievler, Ankara, Turkey.
2 Department of Nephrology, Baskent University Faculty of Medicine, Ankara,
Turkey.
Received September 11, 2002;
accepted after revision July 23, 2003.
Address correspondence to N. C. Tarhan.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Seventeen patients with osmotic demyelination syndrome who had undergone hemodialysis at least once and had brain MRI examinations were retrospectively reviewed. Neurologic and MRI examinations were performed during a clinical episode. Serum levels of sodium, creatinine, blood urea nitrogen, and glucose were assessed, and serum osmolality and the ratio of blood urea nitrogen to creatinine (BUN:Cr) were calculated. Follow-up MRI was performed in nine cases. Laboratory and imaging findings were evaluated.
RESULTS. An altered level of consciousness and convulsions were the most common neurologic symptoms. The pons was involved in 11 patients (65%) and extrapontine sites in 12 (71%). Four patients had dysequilibrium syndrome. Follow-up MRI showed complete resolution in six patients and lesion reduction in three within a short time. The most common biochemical changes at the time of MRI were hyponatremia and low BUN:Cr in the blood. Only one patient showed rapid correction of hyponatremia and a rapid change in osmolality during the acute stage.
CONCLUSION. In patients who develop osmotic demyelination syndrome after hemodialysis, the lesions may involve the pons or the pons and extrapontine sites. Most lesions that were followed up resolved rapidly and almost completely, favoring transient edema rather than demyelination. Blood chemistries suggested underlying changes in osmolality, particularly as a result of urea shift from the extracellular fluid.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The MRI findings of osmotic demyelination syndrome in patients with end-stage renal disease have not been documented in detail. The aims of this study were to present the brain MRI findings of osmotic demyelination syndrome at the time of an episode after hemodialysis and at follow-up, and to identify possible factors that may contribute to the development of these lesions in patients with end-stage renal disease.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
All 17 individuals had undergone hemodialysis before they developed neurologic symptoms. The time between hemodialysis and MRI was 14–24 hr (mean, 19 hr).
Clinical Analysis
All patients underwent a neurologic examination after they developed
symptoms. Creatinine (Cr), blood urea nitrogen (BUN), and glucose electrolyte
concentrations were collected, and serum osmolality, BUN/Cr was calculated
several times during the clinical episode in all patients except one. Serum
osmolality was calculated with the formula
[12]:
 |
In four of the eight patients who had follow-up, serum osmolality could not be calculated because serum glucose was not measured at the time of the follow-up MRI examination. Seven patients underwent cerebrospinal fluid evaluation after the episode. Laboratory findings at follow-up were included only for the eight patients who underwent repeated MRI, as described in the following text.
Imaging Analysis
Brain MRI was performed on the same day that each patient developed the
neurologic complaints. A 1.0-T superconductive magnetic field system (Expert
VB33A, Siemens, Erlangen, Germany) was used in all cases. Transverse
T2-weighted (TR/TE, 4,000/99; slice thickness, 6 mm; acquisitions, 3; matrix,
154 x 256) and sagittal T1-weighted (500/15; flip angle, 70°;
acquisitions, 3; matrix, 144 x 256) images were obtained for all
patients. Transverse proton density–weighted (4,000/15; acquisitions, 3;
matrix, 144 x 256) images were obtained in six patients, and transverse
FLAIR (9,000/105; inversion time, 2,200 msec; matrix, 140 x 256;
acquisitions, 1) images were obtained in the other 11 patients. Coronal
T2-weighted images were obtained in six patients. None of the patients was
given an injection of contrast material for evaluation of osmotic
demyelination syndrome. Follow-up MRI was performed in nine patients. Two of
these nine patients were rescanned twice and seven were rescanned once.
Follow-up MRI examination time from the initial MRI ranged from 2 weeks to 4
months (mean, 5 weeks) in eight patients. In one patient, 2.5 years later new
neurologic symptoms developed, so the second MRI was performed 2.5 years after
the first.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
All 17 individuals showed clinical improvement during follow-up. Neurologic findings in the nine patients who had follow-up imaging were as follows: Patient 1 had headache, nausea and vomiting, obtundation, and convulsions immediately after hemodialysis. The neurologic findings resolved within a week. Patient 2 initially had nausea and vomiting, obtundation, ataxia, paraparesis, and convulsion during hemodialysis. At follow-up in 3 weeks, all neurologic findings had subsided. Patient 3 had obtundation and stupor, quadriparesis, third cranial nerve palsy, a positive Babinski's sign, and convulsions. Three days later, she developed left hemiplegia. At follow-up after 1 month, her left hemiparesis and partial third nerve palsy remained, but all other neurologic findings had resolved. Patient 4 had obtundation, a positive Babinski's sign, and convulsions with abnormal EEG findings at the time of presentation. At follow-up 3 weeks later, her EEG findings returned to normal and she had only mild paraparesis and peripheral neuropathy. Patient 5 had nausea, bilateral facial hypoesthesia, and progressive facial paralysis, a positive Babinski's sign on the left side, ataxia, dysarthria, difficulty swallowing, and quadriparesis, and all the findings occurred suddenly. All clinical findings resolved in 2–3 weeks. Patient 6 developed obtundation, agitation, and left-sided monoparesis during dialysis. All clinical findings returned to normal in 3 weeks. Patient 7 had obtundation, quadriparesis, and convulsions at the time of presentation. Quadriparesis decreased in severity and she had only peripheral neuropathy at follow-up 4 months after treatment. Patient 8 had dizziness and hypoesthesia at presentation that had resolved at follow-up in 40 days. Patient 9 had clinical improvement during follow-up, and MRI was not performed during the course of the disease. However, she developed new neurologic findings of dizziness and nausea, so MRI was performed 2.5 years after the initial study.
MRI Findings at Presentation
MRI signal changes were observed in all patients, all of whom were
diagnosed as having osmotic demyelination syndrome primarily on the basis of
imaging findings. Patients had hypo- to isointense lesions on the T1-weighted
images and hyperintense lesions on the T2-weighted, proton
density–weighted, and FLAIR images. Most signal changes were centrally
located in the pons (Fig. 1A,
1B), medulla oblongata, and
mesencephalon. In one patient, the pontine lesion was located posteriorly
(Fig. 2A,
2B,
2C). In cerebellar and
supratentorial lesions, bilateral symmetric involvement was mostly noted. In
three patients, the extrapontine lesions were asymmetric but were detected on
both sides (Fig. 3A,
3B). The pontine lesions
exhibited the classic trident shape on axial images. The pontine tegmentum and
ventrolateral pons were preserved, which is characteristic of osmotic
demyelination syndrome (Fig.
1A,
1B). Eleven patients showed
pons involvement (65%), and in five the pons was the only site affected (Fig.
1A,
1B). In six patients (35%),
only extrapontine sites were involved (Fig.
4A,
4B).
|
|
|
|
|
|
|
|
|
MRI Findings at Follow-Up
As stated, repeated MRI was performed in nine cases. Two patients underwent
two repeated MRI examinations. Patient 1 had both pontine and extrapontine
lesions. These lesions were significantly reduced at 1 week after treatment,
and rescanning at 1 month showed complete recovery. Patient 2, who had
extrapontine lesions, showed complete recovery at 3 weeks. In patient 3,
follow-up examination at 2 weeks showed a new area of edema in the right
putamen and resolution of lesions that had been detected in the medulla
oblongata, pons, mesencephalon, posterior limb of internal capsule, and
periventricular white matter (Fig.
5A,
5B,
5C). At 1 month, the patient
showed clinical improvement but the putaminal lesion remained unchanged.
Because this patient developed left hemiparesis in the interval between
treatment and the first follow-up MRI examination, the lesion in the right
putamen was interpreted as an infarct rather than as myelinolysis. Patient 4
exhibited both pontine and extrapontine changes. By 3 weeks, the size of the
pontine lesion was reduced and the other lesions had resolved completely.
Patient 5 had only a pontine lesion (Figs.
2A and
2B), and it had disappeared by
the time MRI was repeated at 3 weeks (Fig.
2C). In patient 6, 17 days after initial MRI the lesions in the
pons, basal ganglia, and internal and external capsules had resolved, and
those in the periventricular white matter and centrum semiovale were smaller.
In patient 7, the lesions in the thalami had completely disappeared 4 months
after treatment. Patient 8 had bilateral asymmetric lesions in the basal
ganglia, and repeated MRI at 40 days showed reduction of lesions (Figs.
3A and
3B). Patient 9 had edema in the
central pons at the time of diagnosis (Fig.
1A); at follow-up 2.5 years later, pontine edema had disappeared
completely (Fig. 1B).
|
|
|
Laboratory Data
Eight patients were hyponatremic (110–136 mEq/L) at the time of the
episode. During and immediately after the episode in one patient, sodium
levels rose rapidly (28 mEq/L in 48 hr)
[2]. The serum glucose level
was elevated (135–281 mg/dL) in eight patients at the time of the
episode. These patients had diabetic nephropathy.
Serum BUN levels were minimally increased (30–60 mg/dL) in 12 patients and were very high (> 60 mg/dL) in five at the time of the episode. In two of these latter five patients with follow-up MRI, the BUN level decreased.
In four cases, serum osmolality was within normal limits (275–295 mOsm/L) at the time of the episode. In one individual, serum osmolality was very low initially (245 mOsm/L) but rose rapidly (to 301 mOsm/L) soon after the episode. In 10 patients, serum osmolality was slightly elevated during the episode (295–320 mOsm/L). In one case, serum osmolality was very high initially (> 320 mOsm/L) but decreased during follow-up (296 mOsm/L).
At the time of the episode, seven patients had a ratio of BUN to creatinine (BUN:Cr) in the 9–13.5 range, four patients (all hyponatremic) had ratios of 15–40, and six patients had ratios of 4–8. Four of the latter six patients were normonatremic and had normal serum glucose levels. Three of these four underwent repeated MRI. Of these three individuals, one showed a marked reduction and the other, resolution of lesions; the BUN:Cr in these cases rose to the 9–13.5 range at follow-up. The third patient showed minimal lesion changes on repeated MRI, and the BUN:Cr remained low at follow-up. Two of the six patients who were in the 4–8 BUN:Cr range were hyponatremic. In both of these patients, follow-up MRI showed that the lesions resolved but the BUN:Cr remained low. One individual had normal serum sodium and glucose levels, and a BUN:Cr ratio of 9–13.5, but very high serum osmolality at the time of the episode. At follow-up, MRI showed decreased edema, and laboratory tests revealed lower serum osmolality.
In five of the seven patients who underwent cerebrospinal fluid examination, the findings were normal. The sample from one patient showed slightly elevated protein (62 mg/dL), which was considered a nonspecific finding. Another patient exhibited elevated cerebrospinal fluid pressure (300 cm H2O) and had a clinical diagnosis of pseudotumor cerebri.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
When hyponatremia causes intracellular volume alterations in the brain, exchangeable intracellular solutes are redistributed by osmotic adaptation to prevent cerebral edema, leading to gradual loss of intracellular organic osmotically active particles. More than 5 days are required to rebuild the stores of these osmotically active particles, so when serum osmolality increases rapidly the brain is at risk for osmotic dehydration [15]. In one study in rats, Rojiani et al. [16] showed transient opening of the tight junctions between endothelial cells in the cerebral microvasculature that interfere with the blood–brain barrier by injecting hypertonic saline after hyponatremia. In classic osmotic demyelination syndrome, it was suggested that the hypertonic fluid that causes the edema remains in the extracellular space until endothelial integrity is restored, and that this fluid may have toxic effects on myelin and oligodendrocytes. Naing Oo et al. [17] suggested that uremic patients after rapid correction of chronic hyponatremia might be less prone to develop demyelination.
In an autopsy series of patients with end-stage renal disease who were undergoing hemodialysis that did not include radiologic studies, Endo et al. [1] reported an incidence of osmotic demyelination syndrome of 14%. Most patients with end-stage renal disease are subject to rapid osmotic fluctuations after hemodialysis. In our study, we documented rapid increases in serum sodium (the major known cause of osmotic demyelination syndrome) in just one patient. But as mentioned before, all 17 patients had had at least one hemodialysis session, and all the patients developed symptoms during, or within 24 hr after, the hemodialysis. In four patients who lost consciousness and developed convulsions during or immediately after hemodialysis, the clinical findings were consistent with dysequilibrium syndrome. This syndrome is a complication that is usually seen after rapid dialysis, especially when a patient has recently started undergoing hemodialysis [18]. The syndrome is considered to result from an osmotic gradient of unidentified osmotically active agents resulting in a hypotonic extracellular fluid [19]. These patients may show the classic MRI findings of osmotic demyelination syndrome [10] that were seen in four of our patients.
One group of researchers documented a cerebral volume increase without any symptoms or apparent MRI changes in patients who had undergone regular hemodialysis [20]. Those researchers performed a volumetric MRI study with T1-weighted images and found an approximately 3% increase in brain volume after hemodialysis. This change was thought to be the result of a rapid decrease in plasma urea concentration compared with the concentration in the brain cells. During hemodialysis, the plasma becomes hypotonic in relation to brain cells, and the sudden shift of water from the extracellular space to the brain cells causes edema [18]. The most likely pathogenetic mechanism behind osmotic demyelination syndrome in patients with end-stage renal disease may be the rapid change in plasma solute levels, resulting in the change of serum osmolality that occurs during or after hemodialysis. Only four patients had serum osmolality in the normal range at the time of their episodes. More than the known effects of a rapid rise in serum sodium, a patient's BUN and glucose levels before and after hemodialysis may be important because these three parameters are used to measure serum osmolality (see the formula in Materials and Methods). In particular, BUN level has a greater effect when calculating the osmolality, so a low BUN:Cr after hemodialysis that would show a relatively high urea shift may indicate a major osmotic gradient. As our results show, even when serum glucose and sodium levels are normal, low BUN:Cr may be linked to changes on MRI.
Although evidence supports these factors being involved in the pathogenesis of osmotic demyelination syndrome, it is still not clear why some hemodialysis patients develop the condition and others do not. The incidence of osmotic demyelination syndrome is thought to be higher in patients with chronic debilitating diseases [1, 6, 8–10]. End-stage renal disease and chronic hemodialysis are in this category, but not all of our patients had been on long-term hemodialysis. Seven of them had been undergoing dialysis for less than 3 months.
In our study group, the pons was the most commonly involved brain area (65% of cases). Other commonly involved sites were the cerebral periventricular and subcortical white matter, thalami, basal ganglia, internal capsule, and the cerebellar peduncles and white matter. All of these sites contain interdigitated gray and white matter. This interdigitation may interfere with the diffusion of hypertonic edema fluid into adjacent white matter; thus, these areas are especially vulnerable to rapid fluid and electrolyte changes [16, 21, 22].
There are a few reports on diffusion-weighted imaging in osmotic demyelination syndrome [23–25]. In studies of Cramer et al. [23] with two cases and Chu et al. [24], decreased apparent diffusion coefficient values, especially early in the disease, suggested cytotoxic edema. Those authors stated that conditions with increased apparent diffusion coefficient values, such as tumors, acute disseminated encephalomyelitis, and multiple sclerosis, which might be mistaken clinically for osmotic demyelination syndrome, can be differentiated by diffusion-weighted imaging. In the study of Chua et al. [25] of six patients with decreased serum sodium levels, apparent diffusion coefficient values were increased, and the authors stated that those increased values had similarities to multiple sclerosis. We could not perform diffusion-weighted imaging in our patients at the time of MRI because the technique was unavailable. With the major BUN shift and osmolality changes found in our patient group after hemodialysis, a major intracellular fluid increase is consistent with cytotoxic edema, so we thought it would be more likely to obtain decreased apparent diffusion coefficient values in osmotic demyelination syndrome patients with end-stage renal disease after hemodialysis. Apparent diffusion coefficient maps may quantify these changes in osmotic demyelination syndrome [26].
The classic symptoms of osmotic demyelination syndrome are spastic quadriparesis and pseudobulbar palsy, which occur as a result of corticospinal and corticobulbar tract involvement in the pons and internal capsule [2, 7]. Most patients show early impairment of consciousness, which sometimes leads to coma. Bilateral cerebral or thalamic damage, or lesions in the ascending reticular activating system in the brainstem or pons, may result in altered consciousness [2]. This type of impairment was the most frequent neurologic finding (10/17 patients) in our series.
We found that the radiologic findings in osmotic demyelination syndrome may not always match the neurologic signs [13]. In five patients, the diagnosis was based on specific radiologic findings of osmotic demyelination syndrome rather than on clinical findings. In three of these cases, patients' brain lesions of osmotic demyelination syndrome probably developed without any noticeable symptoms. One of these individuals had bilateral papilledema and increased cerebrospinal fluid pressure but extrapontine lesions. In this patient, the clinical diagnosis was pseudotumor cerebri, and the patient was treated accordingly. In another two individuals, we saw that the clinical findings were only partially linked to MRI findings. Contrary to its classic description, osmotic demyelination syndrome may develop without any presenting signs and symptoms in patients with end-stage renal disease who have undergone recent hemodialysis. MRI may detect these asymptomatic transient changes during hemodialysis in patients' brains because these five patients did not have apparent signs or symptoms to explain the edema. In only one patient did we detect an abscess that was likely responsible for the neurologic deficit.
Our MRI studies showed mostly symmetric lesions with low to intermediate signal on T1-weighted images and high signal on T2-weighted and proton density–weighted or FLAIR images. No mass effect was seen in any of the 17 patients. Most of the pontine lesions exhibited the characteristic trident shape on axial images. Some reports have described peripheral rim enhancement after contrast administration [5, 27], but the diagnosis was made without contrast material in all our patients.
The intensity changes observed on MRI of patients with osmotic demyelination syndrome after hemodialysis are thought to represent a combination of edema and demyelination. After the acute phase of osmotic demyelination syndrome, the lesions become smaller because the edema subsides and some degree of remyelination occurs [4]. In six of our nine patients who underwent follow-up imaging, the lesions completely resolved. Two others showed significant lesion reduction, and one showed minimal reduction. The literature contains only two reported cases of complete disappearance of initial MRI findings, but neither of these patients had end-stage renal disease [27, 28]. In six of our nine patients who showed resolution of the initial MRI findings, reduction or complete resolution of the lesions occurred within 1 month after the episode. Reduction of lesions within a month is faster than has been reported in previous cases of osmotic demyelination syndrome in any patient group [4, 8, 9]. The rapid reduction of lesions in our patients seems to indicate that most of the lesions represented edema rather than myelinolysis. Myelinotoxic agents are released in osmotic demyelination syndrome as a result of osmotic endothelial damage [29]. Endothelial integrity may be restored faster than with other diseases after transient blood–brain barrier damage in the hemodialysis setting, thus reversing the process before myelinolysis starts.
Our study had limitations. First, we could not perform diffusion-weighted imaging or MR spectroscopy in any of our patients because at the time our patients underwent MRI, these imaging techniques were not available at our MRI unit. Diffusion-weighted imaging may help separate cytotoxic from vasogenic edema and may potentially improve the sensitivity and specificity of MRI for the diagnosis of osmotic demyelination syndrome. Second, all patients included in our study were diagnosed as having osmotic demyelination syndrome on the basis of imaging findings. Therefore, we did not have a gold standard diagnostic test for comparison, so it was not possible to calculate sensitivity or specificity in this study.
In conclusion, the results of our study show that patients with end-stage renal disease may have foci of edema in various areas of the brain after recent hemodialysis. The typical MRI findings along with neurologic signs and symptoms, and the pattern and areas of brain involvement, were diagnostic for osmotic demyelination syndrome. Osmotic demyelination syndrome in hemodialysis patients may be asymptomatic. The lesions in patients with end-stage renal disease after hemodialysis differed from other causes of osmotic demyelination syndrome in that most lesions were rapidly reversible (mean, 5 weeks). These rapidly reversible MRI findings indicate that the lesions represented edema rather than myelinolysis in this patient group. As has been documented previously, with any cause of osmotic demyelination syndrome the pons is the most commonly involved site. Main laboratory findings in our patient group were hyponatremia and a low BUN:Cr, so we believe that the mechanism of osmotic demyelination syndrome in patients with end-stage renal disease involves rapid osmotically active particle changes—mainly, BUN shifts from plasma during hemodialysis. The results of serum osmolality of the patients were variable. It is important to perform MRI in patients who experience rapid serum osmolarity changes after hemodialysis.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  C.J. da Silva, A.J. da Rocha, S. Jeronymo, M.F. Mendes, F.T. Milani, A.C.M. Maia Jr, F.T. Braga, Y.A.S. Sens, and L.A. Miorin A Preliminary Study Revealing a New Association in Patients Undergoing Maintenance Hemodialysis: Manganism Symptoms and T1 Hyperintense Changes in the Basal Ganglia AJNR Am. J. Neuroradiol., September 1, 2007; 28(8): 1474 - 1479. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C.L. Chen, P.H. Lai, K.J. Chou, P.T. Lee, H.M. Chung, and H.C. Fang A Preliminary Report of Brain Edema in Patients with Uremia at First Hemodialysis: Evaluation by Diffusion-Weighted MR Imaging AJNR Am. J. Neuroradiol., January 1, 2007; 28(1): 68 - 71. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. M. Murray, D. E. Tupper, D. S. Knopman, D. T. Gilbertson, S. L. Pederson, S. Li, G. E. Smith, A. K. Hochhalter, A. J. Collins, and R. L. Kane Cognitive impairment in hemodialysis patients is common. Neurology, July 25, 2006; 67(2): 216 - 223. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
