| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Full Paper |
1 Neuroradiology and 2 Department of Medical Imaging, Toronto Western Hospital, University Health Network, 3 Fell Pavilion, Room 210, 399 Bathurst Street, Toronto, Ontario M5T 2S8, Canada
 |
Abstract |
|---|
 Top Abstract IntroductionMethods and materials/patientsResultsDiscussionConclusionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMethods and materials/patientsResultsDiscussionConclusionReferences |
|---|
Spongiform leukoencephalopathy is the formation of intracytoplasmic vacuoles within the myelin sheath associated with neuronal loss and gliosis [1]. A differential diagnosis of spongiform leukoencephalopathy includes prion disease, metabolic encephalopathies, hypoxic-ischaemic disease and multiple neurodegenerative diseases [1].
A rare complication from inhalation of heated heroin vapour is spongiform leukoencephalopathy. This entity is rare, however, the number of cases that are unrecognized due to non-specific symptoms and often confusing clinical histories is unknown. Spongiform leukoencephalopathy associated with inhalation of heroin vapour has an estimated mortality of 23% [2]. This mortality may be reduced with prompt recognition of this entity and initiation of therapy.
The drug-delivery method of inhaling heated heroin vapour has been around for nearly a century [3]. Chasing is the practice of heating heroin powder over a flame to temperatures between 200°C and 300°C on aluminium foil and inhaling the heated vapour which rises like the tail of a dragon. "Chasing the dragon" is a highly effective mode of drug delivery with rapid, dose-related subjective and objective effects and reproducible bioavailability [4]. Clinical studies of heroin administration have shown chasing to be the most effective drug-delivery method in comparison with other non-intravenous routes, providing approximately 35–45% bioavailability of the heroin dose [4]. The relative potency of chasing to intravenous heroin is not known.
The potentially fatal and disabling complication of spongiform leukoencephalopathy was not recognized until an outbreak in The Netherlands in 1981 [5]. Since the 1980s, much has been written about the clinical, radiographic and pathological nature of this disease. Reports in the literature that include MRI diffusion imaging, ADC mapping and MRS, provide clues to the pathogenesis of heroin-induced spongiform leukoencephalopathy. Nonetheless, the aetiology of the spongiform leukoencephalopathy continues to be elusive.
|
Methods and materials/patients |
|---|
|
TopAbstractIntroductionMethods and materials/patientsResultsDiscussionConclusionReferences |
|---|
ADC images were calculated from the DWI data. Quantitative ADC values were obtained in the same regions of the left occipital lobe and the right frontal lobes in all patients. These values were compared with those from a set of normal volunteers at our institution.
Single-voxel proton spectroscopy (20 mm3) of the left occipital lobe white matter was performed at echo times of 30 m s–1 and 144 m s–1 (MRS; TR=1500 ms, TE=30 ms). All quantitative MRS values were generated at the GE MRI workstation at the time of imaging. Reference quantitative MRS values were obtained from a set of normal volunteers, imaged on the same GE MRI unit.
Contrast enhanced images were acquired at the end of the examination and after administration of 10 cm3 gadolinium DTPA using T1 weighted sequencing in the axial and coronal planes (T1; TR=450 ms, TE=9 ms).
Rapid dynamic susceptibility perfusion MRI was performed with echoplanar gradient echo readout (TR=1750, TE=30, flip angle=90°, voxel size 2 x 2 x 5 mm, 18 slices, 25 temporal frames). An intravenous antecubital injection of 15 cm3 of gadolinium-DTPA was administered at 7 cm3 s–1 at the start of the perfusion scan. Data were processed using "Functool 2" on a GE Advantage Windows workstation (version 4.1). Analysis included evaluation of the "negative enhancement integral" (NEI; a measure of relative cerebral blood volume), the "mean time to enhance" (MTE; a measure of the bolus transit time), and the "time to minimum" (TTM; a measure of the mid-bolus arrival time).
|
Results |
|---|
|
TopAbstractIntroductionMethods and materials/patientsResultsDiscussionConclusionReferences |
|---|
Toxicology screening performed at admission was negative for all substances, including heroin metabolites. His human immunodeficiency virus (HIV) status was negative and all admission laboratory tests were within normal limits, including cerebrospinal fluid (CSF) evaluation from lumbar puncture. He was normotensive and had no documented periods of hypoxia.
His admission head CT was performed without intravenous contrast and showed diffuse symmetric areas of decreased density. The involved areas included the cerebellum, upper medulla, pons, midbrain, thalami, occipital lobes, the posterior aspect of the corpus callosum, the posterior limbs of the internal capsules, and the white matter in the occipital and frontoparietal regions.
An emergent MRI was performed with DWI/ADC imaging, single-voxel proton spectroscopy of the occipital white matter, and contrast enhancement. Again noted were the symmetrical white matter changes. There was increased T2 weighted (T2W) signal corresponding to the areas of low density on the head CT, with preservation of the sub-cortical white matter (Figure 1a
). There was no abnormal gadolinium enhancement. The DWI and ADC mapping showed normal water diffusion in the affected areas (Left occipital lobe (LO): ADC=814.9, SD=29.3, normal=788, SD=52; Right frontal lobe (RF): ADC=754.3, SD=24.3, normal=774, SD=46) (Figure 2a). MR spectroscopy showed significant lactate accumulation within the occipital white matter, with an elevated Lac/Cr ratio of 0.91 (normal=0.5). The NAA/Cr and the Cho/Cr ratios were both decreased from normal (1.22, normal=2.41; and 1.0, normal=1.35, respectively). There was a large peak in myo-inositol with a mI/Cr ratio of 0.8 (normal=0.65). There was no excessive accumulation of lipids (Figures 3a and 4).
|
|
|
|
A repeat MRI was performed nearly 6 months after the original examination. The T2W images showed continued abnormal increased signal in the same regions as on the initial examination (Figure 1a
). However, there was an unexpected expanded region of abnormally increased signal intensity in the frontoparietal white matter. The DWI/ADC images showed increased water diffusion in contrast to the normal water motion on the prior exam (LO: ADC=1003.5, SD 62.5; RF: ADC=1041.9, SD=35.3) (Figure 2b). Despite this expanded region of white matter damage, his subjective brain volume was unchanged and within normal limits. MRS showed persistently decreased, but improved, ratios of NAA/Cr (1.72, normal=2.41) and myo-inositol (0.73, normal=0.65). The Cho/Cr ratio had worsened, decreasing to 0.76 (normal=1.35).
He showed significant clinical improvement at a follow-up clinic visit nearly 4 months after admission. He was able to converse fluently, walk 20 steps, had increased upper and lower extremity power, and improved tone. He still had some spasticity in the upper extremities, persistent difficulty with rapid alternating movements, finger-to-nose and heel-to-shin tests. His gait remained ataxic with abnormal posturing and a wide-spread gait with arching of his back.
Case 2
Case 2 is a 24-year-old man who presented with a 1 month history of progressive dizziness, nausea, vomiting, ataxia, mild dysarthria and social withdrawal. His electrolytes, haematology, HIV and LP tests were all negative. He was normotensive. He had no documented periods of hypoxia. Toxicology screen was positive for morphine and diphenhydramine. Electroencephalography (EEG) during wakefulness was within normal limits, showing no focus of seizure activity. He did not receive any preceding vaccinations.
Admission head CT, without intravenous contrast, showed symmetrical low densities involving the white matter within the cerebral hemispheres, the thalami, and the bilateral internal capsules. The initial differential diagnosis included: adrenoleukodystrophy, lymphoma, progressive multifocal leukoencephalopathy, vasculitis, PRES and toxic exposure.
MRI was performed in what was clearly the sub-acute phase of his illness. The MRI showed symmetrical non-enhancing T2W increased-signal abnormalities involving the white matter tracts of the posterior fossa, brain stem, occipital lobes, and internal capsules (Figure 1b). This process extended into the posterior centrum semiovale, with preservation of the sub-cortical white matter. There was abnormal increased T2W signal within the grey matter of the thalami. The DWI/ADC imaging (Figure 5) showed corresponding areas of abnormal signal without restricted water movement (LO: ADC=788.1, SD=27.4; RF: ADC=881.6, SD=46.8). The MR perfusion data were within normal limits.
|
). He later admitted to a 2-year history of heroin abuse of approximately 0.5 g per day via inhalation of heroin vapour. He was treated with coenzyme Q, had a gastrostomy tube placed and was discharged to a long-term care facility nearly 2 months after his initial presentation.
Case 3
Case 3 is a 43-year-old male who was admitted to the intensive care unit with the non-specific diagnosis of encephalopathy. He was found unresponsive on the floor. He was last seen functioning normally the day before his collapse. His Glasgow Coma Scale upon admission was 4. His pupils were dilated to 5 mm and had a sluggish response to light stimulation. His toes were up-going. He responded to painful stimuli. His airway was patent and his initial transcutaneous blood oxygen saturation was 91%. He was initially hypotensive. It was not known how long his level of consciousness had been depressed or if he suffered from any periods of hypoxia. His airway was protected by endotracheal intubation and he was given intravenous thiamine (100 mg) and naloxone (2 mg).
His admission toxicology screening was positive for cocaine, benzodiazepines and lidocaine, but not for heroin. The patient's family reported chronic cocaine abuse with the patient using cocaine 2–3 times per week. Lumbar puncture, serum antinuclear antibody and HIV tests were negative.
An emergent admission head CT without contrast showed small rounded areas of decreased density within the globus pallidi (Figure 6). An emergent MRI exam was performed with DWI/ADC mapping showing restricted water movement within the globus pallidi (Figure 7a), corresponding to the areas of low density on the preceding head CT. The differential diagnosis for such lesions includes hypoxia, mitochondrial susceptibility, carbon monoxide poisoning and methanol exposure. Additionally, there was diffuse abnormal T2W signal with restricted diffusion involving the periventricular white matter, right-sided putamen, splenium of the corpus callosum, and small punctate areas within the bilateral anterior thalami (LO: ADC=469.8, SD=43.4; RF: ADC=321.5, SD=39.2).
|
|
On follow-up head CT performed 5 days later, the bilateral areas of low density within the globus pallidi were less well defined (Figure 6). However, the periventricular white matter showed new low density corresponding to the T2W abnormalities on admission MRI.
A repeat MRI was performed approximately 11 days after the admission MRI. The T2W images were even more dramatic with increasing signal involving the entire supratentorial white matter (sparing only the sub-cortical zones), the splenium of the corpus callosum and the bilateral globus pallidi (Figure 1c). The DWI mirrored the abnormally increased signal as seen on the T2W sequence (Figure 7b). There were no areas of restricted water movement. In fact, there was a striking reversal of the restricted water diffusion that was seen 11 days prior during the acute phase of his encephalopathy (Figure 7a,b) (LO: ADC=829.2, SD=65.4; RF: ADC=885.8, SD=27.7). MRS showed a markedly increased lactate peak with a Lac/Cr ratio of 3.76 (normal=0.5). Additionally, there was a large lipid peak and a large choline peak (Cho/Cr=2.2, normal=1.35), consistent with severe acute demyelination (Figures 3c and 4).
His overall neurological prognosis was guarded. Tracheostomy and gastrostomy tubes were placed and he was discharged to a nursing facility nearly 6 months after admission.
|
Discussion |
|---|
|
TopAbstractIntroductionMethods and materials/patientsResultsDiscussionConclusionReferences |
|---|
Cases 1 and 2 had documented heroin use via the "chasing the dragon" technique. Their initial MRI examinations showed T2W features nearly identical to cases of chasing toxicity in the literature [2, 5–7, 10]. Although the T2W features in case 3 were also dramatic, the brainstem and cerebellum were not affected. The cerebellar involvement is a distinguishing feature and must be present for the diagnosis of chasing toxicity [7]. PRES can be excluded if there is no history of hypertensive crisis.
Au-Yeung and Lai have described three clinical stages of heroin vapour-related leukoencephalopathy [8]. Stage 1 is described as mostly cerebellar symptoms. Stage 2 includes both cerebellar and extrapyramidal symptoms. Stage 3 progresses to stretching spasms and akinetic mutism or hypotonic mutism [8]. These clinically recognized stages are associated with the extent of white matter involvement on MRI.
DWI/ADC images
Wolters et al evaluated the neuropathological findings in nine patients from Amsterdam who presented with leukoencephalopathy after chasing heroin [5]. They reported spongiform degeneration of all the CNS white matter in all patients. The extent and intensity of the spongiform change was variable. Countless vacuoles were observed in the damaged white matter, with many areas coalescing to form larger vacuolar cavities. There was a reduction in the number of axons and of oligodendroglia surrounding the empty vacuoles, however, there were no products of myelin breakdown. Electron microscopy showed swollen mitochondria and distended endoplasmic reticulum within the remaining cytoplasm of the degenerated multivacuolar oligodendroglia and axons.
The neuropathology findings are essential in understanding the DWI/ADC data, which in turn may also be helpful in defining the aetiology and natural history of spongiform leukoencephalopathy. Vella et al [9] evaluated DWI data without ADC correlation, and postulated that the DWI changes were a non-specific marker, not necessarily representing a specific assay for ischaemia. Chen et al performed acute-phase DWI and ADC mapping showing restricted water diffusion [10]. They postulated that this was due to "fluid entrapment within the myelin lamellae without demyelination" [10]. Building on the postulate by Chen et al, it is conceivable that the acute-phase restricted water movement is only seen in the acute phase of spongiform leukoencephalopathy during the initial development of small vacuoles within the myelin lamellae. As the pathology progresses into the sub-acute and chronic stages, these vacuoles become larger and more numerous, ultimately coalescing to form larger cavities [5]. At this stage, a swollen mitochondria and endoplasmic reticulum would result in increased water diffusion within the affected areas, as seen in both of our patients with known heroin exposure.
Both of our known heroin abusers presented in the sub-acute phase of their encephalopathy. Both patients were discharged to long-term care facilities after a prolonged hospital stay. Our first case returned for repeat MRI nearly 6 months after his original examination. The follow-up MR unexpectedly showed progression of the T2W and DWI/ADC findings anteriorly into the frontoparietal white matter (Figures 1a and 2b). The apparent progression of his disease was paradoxical, given that he was experiencing steady clinical improvement. Also somewhat paradoxical was his lack of apparent brain atrophy in the presence of diffuse white matter disease. The subjective maintenance of his brain volume was thought to be due to the expansion of the vacuoles coinciding with the loss of axons and oligodendroglia.
MRS
MRS is another important tool in evaluating these patients and may assist in differentiating between cases which mimic heroin vapour induced spongiform leukoencephalopathy. Kriegstein et al described the clinical course, MR imaging/spectroscopy, and pathology in the first American patients to be diagnosed with this syndrome [7]. Kriegstein evaluated levels of N-acetyl aspartate (NAA, a neuron-specific marker with a decrease indicating axonal injury or mitochondrial toxicity), choline (Cho, a measure of membrane turn-over and lipid metabolism), lactate (Lac, a measure of anaerobic/aerobic metabolism and mitochondrial function) and creatine (Cr, serves as a reserve for high-energy phosphates and thus, plays a role in maintaining energy-dependent systems) [11].
Within the affected white matter, Kriegstein described increased Lac and decreased NAA. In the clinically most affected patients, Kriegstein reported significantly decreased NAA/Cr. Cortical Cho/Cr was low in all patients but was normal in the cerebellum. Cerebellar Lac/Cr was significantly increased in the clinically worst patient. Cortical Lac/Cr was only increased in the clinically worst patient and dramatically decreased at a 24-month follow-up with corresponding clinical improvement [7].
In general, decreased NAA connotes axonal injury and mitochondrial toxicity. Increased Lac suggests conversion of aerobic to anaerobic metabolism such as in a hypoxic/ischaemic condition or in cases of mitochondrial dysfunction. Kriegstein postulated that this evidence supported a metabolic effect from a heroin-related toxin causing mitochondrial dysfunction [7]. Further anecdotal evidence supporting this theory comes from the clinical improvement in some patients undergoing antioxidant treatment (oral coenzyme Q, 30 mg qid), an effect that has been observed in mitochondrial disorders with increased CSF lactate [7]. Demyelination was not considered an important feature since clinical, electrophysiological and MRS did not support demyelination.
MRS changes have been described in other patients with substance abuse [12, 13]. However, the symmetric pattern observed in cases of chasing heroin toxicity (and in all of our cases) is not present in patients without complicated substance abuse. Haelhorst et al describe decreases in NAA in the frontal lobe grey matter of intravenous heroin-dependent patients [13]. Grey matter NAA was decreased by approximately 7% in comparison with normal controls, while NAA levels in the white matter were not different than those in the control population [13]. They also report reduced NAA levels in frontal grey matter of prior cocaine users and in frontal white matter of prior methamphetamine users [13]. MRS in symptomatic patients who inhaled heated heroin vapour shows reduced NAA levels primarily within the white matter [13]. This suggests a very different mechanism for patients chasing heroin versus other heroin techniques or the use of other illicit substances.
The MRS data from our two known cases of heroin vapour exposure were identical (Figure 4). In both, the MRS data were consistent with a diffuse white matter process, with high lactate peaks indicating metabolic stress. The NAA peak was decreased, indicating neuronal loss, axonal injury and/or mitochondrial toxicity at the time of imaging. Myo-inositol is a glial metabolite and its elevation probably reflects gliosis. Curiously, the combination of decreased NAA and elevated myo-inositol is present in patients with Alzheimer's disease [11]. The large lactate peaks clearly show a breakdown in energy metabolism at the cellular level within the heroin patients. There was no evidence for demyelination given the essentially normal choline peaks and the absence of lipid breakdown products.
The persistent elevation of lactate and increased T2W signal abnormalities in our first case is somewhat paradoxical, especially in the setting of clinical improvement. There are several possibilities that would support these findings. The most likely is that the initial MRI did not reveal the maximum extent of damage incurred by the heroin toxicity. Additionally, there may be a prolonged metabolic imbalance from delayed toxic effects, causing persistent elevation of the lactate levels within the damaged white matter. Another possibility includes that of persistent heroin use via chasing despite living in a protective environment. His clinical improvement is judged from his initial discharge condition and, although we documented improvement, his condition may have regressed or reached a plateau that is not known to us due to clinical visits that are months apart.
The MRS data in our third patient do not conform to the pattern of chasing toxicity. Our data from the confirmed cases of heroin chasing do not show demyelination, despite damage and vacuolization of the intraperiod lines. MRS findings from case 3 clearly demonstrated acute florid demyelination. Thus, MRS evidence of demyelination may be helpful in excluding spongiform leukoencephalopathy due to chasing heroin in similar-appearing cases.
Although our last case did not have a confirmed exposure to heroin, the T2W findings had a pattern similar to that of spongiform leukoencephalopathy from chasing toxicity. The DWI/ADC pattern was also similar to that in chasing toxicity published by Chen et al, showing a reversal of restricted water movement on acute imaging to increased movement upon follow-up [10]. The case reported by Chen et al [10], as with our third patient, may have been confounded by a hypoxic event. Hypoxic complications are a possibility in any case of illicit drug abuse causing respiratory depression.
The most important difference between the heroin patients and the patient with cocaine abuse is the lack of cerebellar T2W signal abnormality in the cocaine abuser. Even though DWI/ADC and MRS images are helpful in distinguishing cases of drug-induced leukoencephalopathy, it is ultimately the increased T2W signal in the cerebellum that is the distinguishing feature, along with the supratentorial white matter findings, that makes the MRI diagnosis of spongiform leukoencephalopathy associated with toxicity from chasing the dragon.
|
Conclusion |
|---|
|
TopAbstractIntroductionMethods and materials/patientsResultsDiscussionConclusionReferences |
|---|
MRI with MRS and DWI/ADC imaging is helpful to evaluate for possible drug-induced leukoencephalopathy, especially in the case of heroin chasing toxicity. The most distinguishing MRI feature of chasing toxicity is the pattern of diffuse increased T2W signal within the white matter tracts of the cerebellum, brainstem and supratentorial brain including the deep grey matter structures. The key to differentiating chasing from other mimickers is the involvement of the cerebellum in a clinically non-hypertensive patient. MRS can be helpful in differentiating cases as well, since chasing toxicity is not associated with demyelination. Instead, MRS is consistent with a breakdown in energy metabolism at the cellular level (increased Lac) with a loss of neurons/mitochondrial injury (decreased NAA) and gliosis (increased mI). DWI/ADC values are less helpful in differentiating chasing toxicity from other cases, since restricted diffusion can be present in acute chasing toxicity and hypoxia. The DWI/ADC data in our chasing cases are consistent with a subacute process, showing increased water diffusion, thought to be due to progressive vacuolization of the affected white matter.
An early diagnosis of chasing toxicity may help in limiting morbidity and mortality by initiation of appropriate supportive therapy, such as coenzyme Q [7]. This is especially important given the possibility of a delayed effect from heroin toxicity, as evidenced by the sub-acute progression of white matter changes in one of our cases months after apparent abstinence of heroin and, paradoxically, during clinical improvement.
Received for publication December 31, 2004. Revision received March 3, 2005. Accepted for publication May 13, 2005.
|
References |
|---|
|
TopAbstractIntroductionMethods and materials/patientsResultsDiscussionConclusionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  D Kondziella, E R Danielsen, and P Arlien-Soeborg Fatal encephalopathy after an isolated overdose of cocaine BMJ Case Reports, July 1, 2009; 2009(jul01_1): bcr0620092003 - bcr0620092003. [Abstract] [Full Text]
|
|
|
|
 |
|
 D. R. Shprecher, K. M. Flanigan, A. G. Smith, S. M. Smith, T. Schenkenberg, and J. Steffens Clinical and Diagnostic Features of Delayed Hypoxic Leukoencephalopathy J Neuropsychiatry Clin Neurosci, November 1, 2008; 20(4): 473 - 477. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D Kondziella, E R Danielsen, and P Arlien-Soeborg Fatal encephalopathy after an isolated overdose of cocaine J. Neurol. Neurosurg. Psychiatry, April 1, 2007; 78(4): 437 - 438. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| BJR | DMFR | IMAGING | ALL BIR JOURNALS |
