Neuroradiology (2003) 45: 393–399 DOI 10.1007/s00234-003-0955-7
Wajanat Jan Robert A. Zimmerman Zhiyue J. Wang Gerard T. Berry Paige B. Kaplan Edward M. Kaye
Received: 30 December 2002 Accepted: 30 December 2002 Published online: 8 May 2003 Springer-Verlag 2003
Presented at RSNA 2001 W. Jan Æ R.A. Zimmerman (&) Z.J. Wang Department of Radiology, University of Pennsylvania School of Medicine, The Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania, USA E-mail:
[email protected] Tel.: +1-215-5902569 Fax: +1-215-5901345 R.A. Zimmerman Department of Radiology, Children’s Hospital of Philadelphia, 34th Street & Civic Center Boulevard, Philadelphia, PA 19104, USA G.T. Berry Æ P.B. Kaplan Æ E.M. Kaye Department of Pediatrics, University of Pennsylvania School of Medicine, The Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania, USA Present address: W. Jan Guy’s Hospital, London, UK Present address: Z.J. Wang Texas Children’s Hospital, Houston, Texas, USA Present address: G.T. Berry Children’s National Medical Center, Washington, DC, USA Present address: E.M. Kaye Genzyme Corporation, Cambridge, Massachusetts, USA
DIAGNOSTIC NEURORADIOLOGY
MR diffusion imaging and MR spectroscopy of maple syrup urine disease during acute metabolic decompensation
Abstract Maple syrup urine disease (MSUD) is an inborn error of amino acid metabolism, which affects the brain tissue resulting in impairment or death if untreated. Imaging studies have shown reversible brain edema during acute metabolic decompensation. The purpose of this paper is to describe the diffusionweighted imaging (DWI) and spectroscopy findings during metabolic decompensation and to assess the value of these findings in the prediction of patient outcome. Six patients with the diagnosis of MSUD underwent conventional MR imaging with DWI during acute presentation with metabolic decompensation. Spectroscopy with long TE was performed in four of the six patients. Follow-up examinations were performed after clinical and metabolic recovery. DWI demonstrated marked restriction of proton diffusion compatible with cytotoxic or intramyelinic sheath edema in the brainstem, basal ganglia, thalami, cerebellar and periventricular white matter and the cerebral cortex. This was accompanied by the presence of an abnormal branched-chain amino acids (BCAA) and branched-chain alphaketo acids (BCKA) peak at 0.9 ppm as well as elevated lactate on proton spectroscopy in all four patients. The changes in all six patients were reversed with treatment without evidence of volume loss or persistent
tissue damage. The presence of cytotoxic or intramyelinic edema as evidenced by restricted water diffusion on DWI, with the presence of lactate on spectroscopy, could imply imminent cell death. However, in the context of metabolic decompensation in MSUD, it appears that changes in cell osmolarity and metabolism can reverse completely after metabolic correction. Keywords Brain, edema Æ Brain, diseases Æ Magnetic resonance, diffusion study Æ Magnetic resonance, spectroscopy Æ Metabolism
394
Introduction Maple syrup urine disease (MSUD) is a rare autosomal recessive inborn error of branched-chain amino acid (BCAA) metabolism. There is a deficiency of the enzyme necessary for the oxidative decarboxylation of branchedchain alpha-keto acids (BCKA) which are produced after the transamination of BCAA. This enzyme deficiency leads to the accumulation of toxic levels of BCAA and BCKA in the body resulting in severe metabolic acidosis, neurologic deterioration and the characteristic maple syrup urine odor [1]. Treatment requires dietary restriction of BCAA and the avoidance of catabolism to prevent protein breakdown. However, the long-term prognosis remains guarded as metabolic decompensation, coma and death may occur during stressful situations such as infection or surgery. The disease can be diagnosed antenatally by measuring the enzyme activity in the leukocytes and fibroblasts [2]. CT and MR imaging show diffuse generalized brain edema with more localized intense MSUD edema involving the cerebellar deep white matter, the posterior part of the brainstem, the cerebral peduncles, the posterior limbs of the internal capsule and the posterior aspect of the centrum semiovale [3]. After treatment, the edema usually subsides leaving atrophic brain substance with widening of the sulci and the fissures [3]. Ultrasound demonstrates similar findings: symmetric increased echogenicity involving the periventricular white matter, the basal ganglia and the thalami [4]. Acute hydrocephalus secondary to impaired cerebrospinal
fluid absorption can contribute to raised intracranial pressure in some unusual cases [5]. Diffusion weighted imaging (DWI) is an MR imaging technique which provides a means of identifying intramyelinic or cytotoxic edema, through reduction in the diffusibility of protons, caused by acute ischemia or infarction. Such edema elicits bright signal on heavily DWI, and is believed to reflect increased intracellular and decrease extracellular fluid resulting from sodium pump failure. Vasogenic or interstitial edema, on the other hand, is associated with unchanged or decreased signal on DWI with increased diffusibility reflected by the apparent diffusion coefficient (ADC). This study was carried out first to assess the type of edema in MSUD and its value in predicting patient outcome using DWI, and second to correlate changes in metabolite levels with imaging features using MR spectroscopy (MRS).
Materials and methods The study group comprised six patients with MSUD who presented acutely with metabolic decompensation between February 1997 and September 2001. The clinical data of the patients are summarized in Table 1. The clinical symptoms and laboratory data were consistent with acute metabolic decompensation. All patients but two (cases 1 and 5) had evidence of concurrent infection. MR examinations were performed on two identical 1.5 T Vision Magnetom Systems (Siemens Medical Systems, Iselin, N.J.) during the acute phase of the patient’s illness (2–11 days after onset of symptoms). MR imaging included axial and coronal T2-weighted images, sagittal and coronal T1-weighted
Table 1 Clinical features in MSUD (Suppl. supplement, Met. decomp. metabolic decompensation, BCAA branched-chain amino acids) Patient no. 1
2
3
4
5
6
Sex Age at diagnosis Treatment
F 8 days
F 14 days
M 3 months
M At birth
M At birth
M 9 days
None
None
None
Diet + suppl.
Diet + suppl.
Treatment compliance Current presentation Chief complaint Clinical diagnosis BCAA Acute treatment
–
–
–
Diet + suppl. Good
Good
Very poor
8 days
9 days
3 months
6 years 4 months
14 years 7 months
Lethargy
Seizure, lethargy Sepsis, MSUD Elevated Antibiotic and dietary control Good recovery 4 weeks
Jackknife episodes Meningitis, MSUD Elevated Metabolic correction
3 years 8 months Fever, lethargy Otitis media Elevated Antibiotics and dietary control Full recovery 3 weeks
Vomiting, seizure
Vomiting, lethargy
Met. decomp.
Sinusitis, Met. decomp.
Elevated Metabolic correction
Elevated Antibiotic and dietary control
Full recovery
Good recovery
5 days
2 weeks
Course and hearing loss Duration of admission
MSUD Elevated Metabolic correction Good recovery ?
Develop. delay 7 weeks
395
scans, coronal FLAIR, followed by contrast enhanced T1weighted imaging in the three orthogonal planes. Two of the patients had MR arteriographic sequences as well. Spectroscopy was recorded using either single voxel SE (PRESS) or 2D SE (PRESS) CSI in four of the six patients, sampling the T2-weighted signal abnormality at the thalamus or the basal ganglia with a TR=1600 ms and TE=135 or 270 ms. The voxel volume ranged from 1.5 to 2 cm3 for CSI and 3.4 to 18.8 cm3 for single-voxel studies. For CSI studies spectra of several voxels were averaged for the anatomic region of interest to enhance the signal-to-noise ratio. Axial DWI was performed using single-shot echo-planar spinecho technique. The diffusion gradients were applied in each of the x, y and z directions with four b values ranging between 0 and 1000 s/mm2. The data were then used to derive ADC maps.
Results The DWI findings in the six patients are summarized in Table 2. The percentage decreases in ADC values are given in parentheses for structures relative to age-matched controls in the four patients in whom the ADC information was obtained. Hyperintense signal abnormalities were depicted on the T2-weighted and FLAIR images in the brainstem, middle cerebellar peduncles and basal ganglia. The periventricular white matter was involved in four cases (Figs. 1, 2) and there was marked involvement of the cerebral cortex in two cases. There was no evidence of enhancement and the mass effect caused by the edema was absent or minimal. DWI showed marked restriction of water diffusion of the vital brain structures with low ADC indicating a shift of fluid into the intracellular compartment such as occurs in cytotoxic or intramyelinic edema (Figs. 1, 2). MR spectroscopy was performed in cases 1, 2, 3 and 6. The N-acetyl-aspartate to choline and creatine to phosphocreatine ratios were decreased in all four cases. There was significant elevation of the lactate at 1.3 ppm and the
methyl groups of BCAA and BCKA at 0.9–1 ppm in all four patients. These changes were reversed after treatment (Fig. 3). The follow-up examination showed significant improvement of the T2 signal abnormalities and there was no evidence of cerebral or brainstem atrophy; however, some signal abnormality persisted in the tegmental tracts despite the lack of volume loss and clinical signs (Fig. 2). The DWI and spectroscopy abnormalities resolved completely after clinical recovery. All the patients had a good outcome except patient 3 (Table 1). This most likely related to the late diagnosis of this patient at the age of 3 months. The severity of the neurologic sequelae is strongly correlated with the duration of the acute toxic phase in the neonatal period [9]. All the patients were followed for a minimum of 1 year.
Discussion MSUD is expressed in different forms as classic, intermittent, intermediate, thiamine responsive and dihydrolipoyl dehydrogenase (E3)-deficient [1]. The classic form presents in the first week of life with poor feeding and vomiting, leading to convulsions, lethargy and coma. Clinically there is hypertonicity and muscular rigidity with opisthotonus. Neurologic findings are often mistaken for sepsis or meningitis, and death occurs if untreated. The intermittent form, which develops at times of stress, usually affects older children, but has an acute clinical course similar to the classic form. In the intermediate form of MSUD, the onset of cerebral symptoms is delayed and symptoms are less severe. Rarely, both the intermediate and intermittent forms may respond to high doses of thiamine [2]. The acute neurologic manifestation of MSUD can be reversible [6].
Table 2 Anatomic areas of restricted diffusion. Percentage decreases in ADC values from normal are given in parenthesesa (PR perirolandic, PV periventricular) Patient no.
Medulla Pons Midbrain Cerebellar dentate Thalamus Globus pallidus Putamen Internal capsule Hippocampus Central white matter Cerebral cortex Other a
1
2
3
4
+ (47%) + (60%) + (49%) + (60%) + (53%) + ) + (53%) + + (PR) ) Optic tracts
+ + + + + + ) + ) + (PR) ) Optic chiasm
+ + (dorsal) (48%) + (52%) + (48%) + (22%) + ) + (69%) + (48%) + (PR) (48%) ) Optic radiation
+ + (dorsal) (30%) + + (49%) ) + ) ) ) ) ) Optic radiation
ADC information no longer available for patients 2 and 5
5 + + + + + + + ) + + (PV) + )
6 + (dorsal) + (59%) + (lateral sparing) (63%) ) + (63%) + (67%) + (67%) ) ) ) + (45%) Mamillary bodies and caudate nuclei
396
Fig. 1A–E Case 3. Axial T2-weighted images during acute metabolic decompensation showing high signal abnormality at A the thalami, globus pallidus and B the perirolandic white matter. C, D Corresponding DWI to A and B, demonstrating marked proton diffusion restriction. E Restricted proton diffusion in midbrain and hippocampi. This was also seen on T2-weighted images, but is not shown
The severity of the neurologic sequelae is strongly correlated with the duration of the acute toxic phase in the neonatal period [7]. In untreated cases, there is macroscopic evidence of brain edema with severe status spongiosus of the white matter which is most pronounced at the gray-white matter junction [8]. Myelination is delayed, particularly in the corticospinal tracts, but there is no evidence of myelin breakdown [3]. In treated cases, the status spongiosus is markedly reduced or absent and the myelination is compatible with the patient’s age [9]. Similar changes are seen in cattle with MSUD, suggesting that the severe myelin edema seen early in the course of the disease indicates that the toxic
effect of BCAA and BCKA is accompanied by or mediated via a disturbance of the fluid retention mechanisms of the myelin sheath [10]. Apoptosis has been detected in BCKA-treated fibroblasts. Fibroblast viability is reduced when exposed to a mixture of MSUD metabolites, while, in contrast, fibroblasts from healthy human donors are unaffected. This suggests that BCKA triggers apoptosis [11]. Apoptosis is thought to be important in the pathogenesis of cerebral ischemia. The mechanism of apoptosis induction can be preferentially triggered by mild/ moderate microcirculatory disturbances; however, selective cell populations may be more vulnerable to subtle reductions of cerebral microcirculation that otherwise do not disturb metabolic or functional integrity [12]. Cellular dysfunction is caused by inhibition of cellular energy metabolism. This mechanism is known to induce cellular swelling (cytotoxic edema) which causes a low ADC. The mechanism of temporal progression of ischemia to irreversible brain damage is expressed as a
397
Fig. 2A–C Case 4. A Axial DWI showing restrictive diffusion at the pons and dentate nuclei. B Corresponding T2-weighted image showing signal changes at the same regions. C The follow-up scan a year later shows persistence of the signal abnormality at the tegmental tracts
sequence of pathophysiologic stages, beginning with perfusion deficit, followed by imbalance of the cerebral energy metabolism and cellular dysfunction, and finally leading to cell death [13, 14]. DWI is a technique where diffusion of water molecules can be detected. It is most commonly used in acute stroke where shortly after an episode of ischemia, diffusion of water is highly restricted due to shift of fluid into the intracellular compartment leading to cytotoxic edema. The ADC of brain tissue is significantly reduced in acute cytotoxic edema. Over several days, the rapid initial drop in ADC is followed by a return to ‘‘pseudonormal’’ values at approximately 1 week and subsequently, elevated ADC values are seen at chronic time points [15]. Regions of vasogenic edema demonstrate increased ADC because the diffusion of water in the extracellular compartment is less than that in free water but nonetheless is significantly higher than in the intracellular compartment [16]. The T2 shine-through effect must be considered when one evaluates hyperintensity on DWI because lesions that are hyperintense on T2weighted images also appear hyperintense on DWI. However, when ADC maps show restricted diffusion, the T2 shine-through effect is unlikely to be the cause of the hyperintensity, as is the case in MSUD edema. Cytotoxic edema induced by tissue hypoperfusion is seen in cerebral venous occlusion, but as the clinical outcome from venous infarction has been quite variable, the question of the predictive value of cytotoxic edema in the outcome prognosis has also been raised [17]. Encephalitic lesions [18, 19] show restricted diffusion of
the involved gray matter, which implies direct viral invasion of the cells and axons of the white matter. In stroke-like lesions such as in some cases of MELAS, the DWI may demonstrate vasogenic rather than cytotoxic edema with higher proton mobility, which corresponds to the high signals on ADC maps [20]. The high mobility of protons in this circumstance suggests that the impaired metabolic activity of mitochondria in the endothelial and smooth muscle cells of blood vessels may disturb the autoregulatory mechanism, resulting in interstitial edema, seen as high signal on T2-weighted images and vasogenic edema on DWI [21]. Similar changes have been seen in hypertensive encephalopathy and eclampsia [22, 23]. A possible explanation of this restricted diffusion pattern of edema is that there is an effect on the myelin leading to intramyelinic edema. This corresponds in young patients (cases 1, 2 and 3) to areas that are fully myelinated at the time of injury. A second explanation is that the neurotoxic effect of BCAA and BCKA affects the neurons of the gray matter in the acute phase resulting in energetic failure. The lesion distribution suggests the contribution of a toxic or vasoactive substance that affects specific nuclei selectively and metabolically, rather than a direct osmotic effect of the BCAA and BCKA. This may play a role in young patients, but convincing evidence of this is seen in the older patients with cortical involvement (cases 5 and 6). The changes in myelin osmolarity in the oldest patient (case 6), where the white matter is not involved, are unexplained, but may be due to changes with time in the susceptibility of white matter to injury. In 1993, Felber et al. [24] first reported the presence of a reversible methyl resonance peak at 0.9–1.0 ppm in a patient with MSUD in acute metabolic decompensation. They found associated elevation of lactate, which they
398
explained by the presence of anerobic glycolysis and a breakdown of the blood-brain barrier. They also noticed a low NAA/Cr level, which is an indicator of neuronal death. The persistence of T2 signal abnormality even after the reversal of the clinical, biochemical and spectral abnormalities argues against the theory that cerebral edema is caused solely by the accumulation and toxicity of metabolites [24]. Furthermore, Heindel et al. [25] managed to isolate the methyl peak at 0.9 ppm in the absence of MR signal abnormality using a TE= 136 ms, where the signal was inverted. The use of a long TE is important because long chain amino acid signals are obscured in the short TE spectra as a result of lipids and macromolecules present in the normal brain. A TE of 270 ms provides more specific detection of BCAA and BCKA because lipid signal could still be detected with TE=136 ms. The ability to detect residual BCAA and BCKA is limited by the noise level in the spectra. The lactate elevation and the 0.9 ppm methyl peak are significant as they correlate with the patient’s clinical condition [26]. Other reports have shown, however, that slightly elevated BCAA and BCKA may remain in patients after recovery, if extensive signal averaging is used [27]. The reversible changes of the NAA and lactate appear to be related to the compromise of mitochondrial function during metabolic decompensation. In our cases the spectroscopy changes correlated with the reversible changes in the DWI; however, the T2-weighted signal abnormalities lagged behind.
Conclusion Diffusion-weighted imaging depicts the T2 abnormal signal lesions as areas of restricted diffusion. These changes were reversible with no evidence of subsequent volume loss. The elevated lactate and low NAA occurred at the same time as the elevated brain and blood levels of BCAA, indicating mitochondrial dysfunction during metabolic decompensation. Unfortunately, the pathogenesis of the reversible diffusion restriction could not be fully clarified in this study, but it has an implication for the received wisdom that diffusion abnormalities with low ADC values are due to cytotoxic edema and are a definitive precursor of cell death. The so-called MSUD edema needs to be investigated further. Fig. 3A, B Case 6. Spectroscopy was performed with a 1.5·1.5·1.5 cm single voxel placed over the region of the lateral ventricles and bilateral anterior thalami using a TR=1600 ms and a TE=270 ms. A At day 2, there is a peak at 0.9 representing the methyl group of the BCAA/BCKA with elevation of the lactate (1.3 ppm). B At day 12, despite the persistence of the thalamic signal abnormality, the spectrum has returned to its normal baseline
399
References 1. Chuang DT, Shih VE (2001) Maple syrup urine disease (branched-chain ketoaciduria). In: Scriver, Beaudet, Valle, et al (eds) The metabolic and molecular bases of inherited disease, 8th edn. McGraw-Hill, New York, pp 1971–2005 2. Rezvani I, Auerbach VH (1987) Defects in metabolism of amino acids. In: Behrman RE, Vaughan VC (eds) Nelson’s textbook of pediatrics, 13th edn. WB Saunders, Philadelphia, pp 289–291 3. Brismar J, Aqeel A, Brismar G, Coates R, Gascon G, Ozand P (1990) Maple syrup urine disease: findings on CT and MR scans of the brain in 10 infants. AJNR Am J Neuroradiol 11: 1219–1228 4. Fariello G, Dionisi-Vici C, Orazi C, et al (1996) Cranial ultrasonography in maple syrup urine disease. AJNR Am J Neuroradiol 17: 311–315 5. Mikati MA, Dudin GE, Kaloustian VMD, Benson PF, Fensom AH (1982) Maple syrup urine disease with increased intracranial pressure. Am J Dis Child 136: 642–643 6. Backhouse O, Leitch RJ, Thompson D, et al (1999) A case of reversible blindness in maple syrup urine disease. Br J Ophthalmol 83: 250–251 7. Muller K, Kahn T, Wendel U (1993) Is demyelination a feature of maple syrup urine disease? Pediatr Neurol 9: 375– 382 8. Martin JJ, Schlote W (1972) Central nervous system lesions in disorders of amino-acid metabolism. A neuropathological study. J Neurol Sci 15: 49–76 9. Kandall BE (1992) Disorders of lysosomes, peroxisomes, and mitochondria. AJNR Am J Neuroradiol 13: 621–653 10. Harper PAW, Helay PJ, Dennis JA (1990) Animal model of human disease. Maple syrup urine disease (branched chain ketoaciduria). Am J Pathol 136: 1445–1447
11. Jouvet P, Rustin P, Felderhoff U, et al (1998) Maple syrup urine disease metabolites induce apoptosis in neural cells without cytochrome c release or changes in mitochondrial membrane potential. Biochem Soc Trans 26: S341 12. Vexler ZS, Roberts TP, Bollen AW, Derugin N, Arieff AI (1997) Transients cerebral ischemia. Association of apoptosis induction hypoperfusion. J Clin Invest 99: 1453–1459 13. Rumpel H, Ferrini B, Martin E (1998) Lasting cytotoxic edema as an indicator of irreversible brain damage: a case of neonatal stroke. AJNR Am J Neuroradiol 19: 1636–1638 14. Phillips MD, Zimmerman RA (1999) Diffusion imaging in pediatric hypoxicischemic injury. Neuroimaging Clin North Am 9: 41–52 15. Albers GW (1998) Diffusion-weighted MRI for evaluation of acute stroke. Neurology 51: S47–49 16. Zimmerman RA (2000) Pediatric cerebrovascular disease. JBR-BTR 83: 245– 252 17. Forbes KPN, Pipe JG, Heiserman JE (2001) Evidence for cytotoxic edema in the pathogenesis of cerebral venous infarction. AJNR Am J Neuroradiol 22: 450–455 18. Tsuchiya K, Katase S, Yoshino A, Hachiya J (1999) Diffusion-weighted MR imaging of encephalitis. AJR Am J Roentgenol 173: 1097–1099 19. Tsuchiya K, Katase S, Yoshino A, Hachiya J (2000) MRI of influenza encephalopathy in children: value of diffusion-weighted imaging. J Comput Assist Tomogr 24: 303–307 20. Yoneda M, Maeda M, Kimura H, Fujii A, Katayama K, Kuriyama M (1999) Vasogenic edema on MELAS: a serial study with diffusion-weighted MR imaging. Neurology 53: 2182–2184
21. Ohshita T, Oka M, Iman Y, et al (2000) Serial diffusion-weighted imaging in MELAS. Neuroradiology 42: 651–656 22. Schaefer PW, Buonanno F, Gonzales RG, Schwamm LH (1997) Diffusionweighted imaging discriminates between cytotoxic and vasogenic edema in a patient with eclampsia. Stroke 28: 1082– 1085 23. Schwartz RB, Mulkern RV, Gudbjartsson H, Jolesz F (1998) Diffusionweighted MR imaging in hypertensive encephalopathy: clues to pathogenesis. AJNR Am J Neuroradiol 19: 859–862 24. Felber SR, Sperl W, Chemelli A, Murr C, Wendel U (1993) Maple syrup urine disease: metabolic decompensation monitored by proton magnetic resonance imaging and spectroscopy. Ann Neurol 33: 396–401 25. Heindel W, Kugel H, Wendel U, Roth B, Benz-Bohm G (1995) Proton magnetic resonance spectroscopy reflects metabolic decompensation in maple syrup urine disease. Pediatr Radiol 25: 296–299 26. Wang ZJ, Zimmerman RA (1998) Proton MR spectroscopy of pediatric brain metabolic disorders. Neuroimaging Clin North Am 8: 781–807 27. Wang ZJ, Zimmerman RA, Sauter R (1996) Proton MR spectroscopy of the brain: clinically useful information obtained in assessing CNS diseases in children. AJR Am J Roentgenol 167: 191–199
