Delayed Parkinsonism after CO Intoxication: Evaluation of the Substantia Nigra with Inversion-Recovery MR Imaging 1

Transcription

1 Note: This copy is for your personal non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, contact us at Hung-Wen Kao, MD Nai-Yu Cho, MS Chun-Jen Hsueh, MD Ming-Chung Chou, PhD Hsiao-Wen Chung, PhD Michelle Liou, PhD Shih-Wei Chiang, MS Shao-Yuan Chen, MD, PhD Chun-Jung Juan, MD, PhD Guo-Shu Huang, MD Cheng-Yu Chen, MD Delayed Parkinsonism after CO Intoxication: Evaluation of the Substantia Nigra with Inversion-Recovery MR Imaging 1 Purpose: Materials and Methods: To quantitatively investigate signal alterations of the substantia nigra in patients with delayed parkinsonism following CO intoxication, as seen on gray matter (GM)- suppressed inversion-recovery (IR) magnetic resonance (MR) images. This prospective study was approved by the local institutional review board, and written informed consent was obtained from all subjects. Thirteen patients with delayed onset of CO-induced parkinsonism (nine men and four women; mean age, 40.3 years), 13 age-matched CO-intoxicated patients without parkinsonism, and 13 age-matched healthy volunteers were examined with GMsuppressed IR MR imaging. The signal intensity of the substantia nigra was normalized to the adjacent normalappearing in the temporal lobe, followed by semiautomatic segmentation into medial, middle, and lateral parts by using a skeleton-based algorithm. Multivariate and univariate analyses and Spearman rank correlation test were performed to examine the relationships between variables. Clinical severity was assessed with the modified Hoehn and Yahr rating scale. Original Research n Neuroradiology 1 From the Department of Radiology, Tri-Service General Hospital, National Defense Medical Center, 325, Section 2, Cheng-Kung Rd., Neihu, Taipei 114, Taiwan, Republic of China (H.W.K., N.Y.C., C.J.H., H.W.C., S.W.C., C.J.J., G.S.H., C.Y.C.); Institute of Biomedical Engineering, National Yang-Ming University, Taipei, Taiwan (N.Y.C.); Department of Medical Imaging and Radiological Sciences, Kaohsiung Medical University, Kaohsiung, Taiwan (M.C.C.); Department of Electrical Engineering, National Taiwan University, Taipei, Taiwan (H.W.C.); Institute of Statistical Science, Academia Sinica, Taipei, Taiwan (M.L.); Department of Hyperbaric Medicine and Neurology, Cardinal Tien Hospital, Taipei, Taiwan (S.Y.C.); and School of Medicine, Fu-Jen Catholic University, Taipei, Taiwan (S.Y.C.). Received December 21, 2011; revision requested February 7, 2012; revision received March 2; accepted March 20; final version accepted March 30. Supported by the National Science Council (grant NSC B MY2). Address correspondence to C.Y.C. ( sandy0928@seed.net.tw). q RSNA, 2012 Results: Conclusion: The normalized signal ratios in the middle and lateral segments of the substantia nigra were significantly higher in those with CO-induced parkinsonism, compared with those with CO intoxication without parkinsonism or normal volunteers (P =.02). For the medial segments, the ratios showed no significant differences among the groups. The normalized signal ratios of substantia nigra were correlated with the severity of parkinsonism, particularly in the lateral segments (r = 0.927, P,.001). CO toxicity to the substantia nigra plays a role in pathophysiologic mechanisms of CO-induced parkinsonism. GM-suppressed IR MR imaging is a useful tool in depicting substantia nigra injury following CO intoxication. q RSNA, 2012 Radiology: Volume 265: Number 1 October 2012 n radiology.rsna.org 215

2 CO intoxication has emerged as a community security threat since the first case of suicide by charcoal burning in Hong Kong in 1998 (1). In Taiwan, this cheap and simple method accounted for 0.22 per suicides in 1999 to 6.48 per suicides in 2006 (2). Inhaled CO rapidly diffuses across the alveolar-capillary membranes into the bloodstream, where reversible binding with hemoglobin occurs and carboxyhemoglobin is formed (3). Endorgan hypoxia ensues from decrease in both oxygen-carrying capacity and oxygen release (4). In addition, CO exerts direct cellular injuries, such as inhibition of electron transport and lipid peroxidation (5,6). Common clinical manifestations of CO intoxication are nonspecific and may include headache, dizziness, weakness, and nausea. Some patients may develop characteristic delayed neuropsychiatric symptoms, such as dementia, incontinence, psychosis, and parkinsonism, days or weeks following acute CO intoxication (7). Although fewer than 10% of the patients had CO-induced parkinsonism, the symptoms can be devastating, manifesting as akinetic mutism, bradykinesia, mask face, and gait ataxia (8). The pathophysiology of CO-induced parkinsonism remains unclear despite the fact that the combined tissue anoxia and direct CO toxicity to cells lead to necrosis preferentially in iron-rich regions, such as the substantia nigra and globus pallidus (7,9). Previous magnetic resonance (MR) imaging studies have demonstrated a Advances in Knowledge CO-induced parkinsonism could be caused by injuries to the substantia nigra, which can be evaluated with gray matter suppressed inversion-recovery (IR) MR imaging. CO-induced substantia nigra injuries occur mainly in the middle and lateral segments, which is in parallel with the findings seen in idiopathic parkinsonism. The degree of the signal change in the substantia nigra correlates with the severity of parkinsonism. spectrum of brain abnormalities in acute stage following CO intoxication, findings that included acute necrosis of globus pallidus, deep edema, and less frequently, edema of other deep-seated nuclei such as putamen, caudate nucleus, thalamus, and hippocampus (10 12). Acute necrosis of substantia nigra was reported in two patients who demonstrated a pallidoreticular pattern of brain injuries and later onset of parkinsonism (9). The imaging findings of delayed encephalopathy often encompassed deep edema, while substantia nigra signal alteration has not been reported, even in cases with delayed parkinsonism (13). The subtle change of the substantia nigra can be assessed by using inversion-recovery (IR) MR technique in the diagnosis and follow-up of idiopathic Parkinson disease (14). In this study, we hypothesized that delayed parkinsonism caused by CO intoxication could be due to damage of the iron-rich substantia nigra, and the damage could be evaluated with gray matter (GM)-suppressed IR MR imaging. Therefore, the purpose of our study was to quantitatively investigate signal alterations of the substantia nigra in patients with parkinsonism following CO intoxication, as seen on GM-suppressed IR MR images. Materials and Methods Subjects This prospective study was approved by a local institutional review board, and written informed consent was Implications for Patient Care Gray matter suppressed IR MR imaging of the substantia nigra may help to assess brain abnormality in delayed parkinsonism caused by CO intoxication and justifies the use of dopamine agonists as a therapeutic option. The knowledge that signal alterations of the substantia nigra correlate with the severity of parkinsonism implies its practical use as an imaging marker for CO-induced parkinsonism. obtained from all subjects. During the period of December 2008 to November 2010, 13 patients with delayed onset of CO-induced parkinsonism (nine men and four women; age range, years; mean age, 40.3 years [standard deviation]), 13 age-matched CO-intoxicated patients without delayed parkinsonism (eight men and five women; age range, years; mean age, 38.4 years ), and 13 age-matched healthy volunteers (nine men and four women; age range, years; mean age, 39.9 years ) were enrolled in this study. All CO-intoxicated patients underwent follow-up MR imaging examinations within 1 3 months after CO exposure (mean, 27 days ). The clinical severity of parkinsonism was assessed by an experienced neurologist (S.Y.C.) within 1 week before MR imaging by using the modified Hoehn and Yahr (H&Y) rating scale. CO-induced parkinsonism was staged, from a score of 1 to a score of 4 (average score, 1.9). Serum levels of carboxyhemoglobin measured in the emergency room were obtained whenever possible. Patients with previous use of neuroleptics, family history of parkinsonism, or history of stroke, head Published online before print /radiol Content code: Radiology 2012; 265: Abbreviations: FLAIR = fluid-attenuated inversion recovery GM = gray matter H&Y = Hoehn and Yahr IR = inversion recovery ROI = region of interest Author contributions: Guarantors of integrity of entire study, H.W.K., C.J.H., C.Y.C.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, H.W.K., N.Y.C., C.Y.C.; clinical studies, H.W.K., N.Y.C., C.J.H., M.C.C., S.Y.C., C.J.J., G.S.H., C.Y.C.; experimental studies, N.Y.C., M.C.C., H.W.C., S.W.C.; statistical analysis, H.W.K., N.Y.C., M.C.C., M.L.; and manuscript editing, H.W.K., N.Y.C., H.W.C., C.Y.C. Potential conflicts of interest are listed at the end of this article. 216 radiology.rsna.org n Radiology: Volume 265: Number 1 October 2012

3 trauma, or encephalitis were excluded from the study. MR Image Acquisition All subjects were imaged with a 1.5-T MR unit (Vision Plus; Siemens, Erlangen, Germany). To increase contrast between the substantia nigra and the surrounding structures, a GM-suppressed IR sequence (repetition time, 2140 msec; echo time, 30 msec; inversion time, 420 msec; matrix size, ; section thickness, 3 mm; intersection gap, 0.21 mm) was used with four axial sections perpendicular to the longitudinal axis of the midbrain, with reference to a midsagittal scout image of the brain stem. In addition, routine images of the whole brain, including T1-weighted, T2-weighted, diffusionweighted, and fluid-attenuated inversion recovery (FLAIR) images, were also obtained. Semiautomatic Skeleton-based Segmentation and Image Analysis To assess the signal alteration of the substantia nigra, a semiautomatic segmentation and quantification method was used. Out of four consecutive axial midbrain sections on the GM-suppressed images, three board-certificated diagnostic neuroradiologists (C.Y.C., C.J.H., and H.W.K., with 22, 11, and 6 years, respectively, of clinical experience in neuroimaging interpretation) blinded to the clinical information of the subjects, independently chose a section that represented the best contrast between the substantia nigra and the surrounding structures. They manually defined the bilateral substantia nigra as the region of interest (ROI), which contained approximately 40 pixels. To increase the reliability of the ROIs, each rater outlined the ROIs three times in three different sessions at least 1 week apart for each subject. A skeleton of each ROI was computer-generated with two lines perpendicular to the skeleton, dividing the ROI into medial, middle, and lateral segments of equal length. The mean pixel intensity value of each segment was automatically computed and displayed. The three raters also drew an ROI of normal-appearing Figure 1 Figure 1: (a) GM-suppressed IR MR image shows an ROI selected at the right normal-appearing temporal lobe for internal normalization (oval). (b) The midbrain (rectangle in a) is magnified in demonstrating the semiautomatic segmentation of the substantia nigra according to the length of the generated skeleton. The mean pixel value within each ROI was calculated and the normalized signal ratio of the ROIs (substantia nigra/temporal lobe ) was computed. in the temporal lobe on the chosen GM-suppressed IR image, as an internal reference. Then, the ratio of the mean pixel intensity of each segment to the normal-appearing temporal lobe was calculated for normalization. Figure 1 illustrates the semiautomatic skeleton-based segmentation, which was performed by using software (Matlab 7.0; Mathworks, Natick, Mass). On FLAIR images, hyperintensities in the cerebral cortex, cerebellum, globus pallidus, and cerebral were documented by the three raters in consensus. Statistical Analysis To evaluate the difference of the normalized ratios of the three substantia nigra segments among the groups, the multivariate analysis of covariance was used, with age and sex as covariates. It was followed by post hoc comparisons between the groups, with Bonferroni method to control the familywise error rate at a of.05. Although the test assumes that the multivariate data were in normal distribution, the equal sample size in our study groups made the test more robust to the normality assumption. Ma-Whitney U tests were used to compare carboxyhemoglobin levels and intervals (in days) between CO exposure and MR imaging between the two groups with CO intoxication and to test the difference of the normalized ratios of the substantia nigra between patients with hyperintensities of the globus pallidus and those without. Fisher exact test of independence was used to compare the incidence of FLAIR hyperintensities between the groups in different locations, including the cerebral cortex, cerebellum, globus pallidus, and cerebral. The correlations between the normalized ratios and the modified H&Y scores in those with COinduced parkinsonism were measured by using Spearman rank correlation tests. The intraclass correlation coefficient was used to evaluate interoperator agreement of the selected ROIs. All statistical analyses were performed by using SPSS statistical software (SPSS for Windows, release 18.0; SPSS, Chicago, Ill). Results The clinical variables and MR findings of CO-intoxicated patients are summarized in the Table. None of our patients showed signal alterations in bilateral temporal lobe. The carboxyhemoglobin levels, intervals of CO exposure to MR imaging, Radiology: Volume 265: Number 1 October 2012 n radiology.rsna.org 217

4 Summary of Clinical Variables and MR Imaging Findings of CO-intoxicated Patients Patient No./Sex/ Age (y) COHb (%)* Major Clinical Symptoms 1/F/ Cognitive impairment, left limb weakness Lesion Location on MR Images CO-induced Parkinsonism Cerebral cortex, globus pallidus, 2/M/47 16 Cognitive impairment Globus pallidus, 3 3/F/ Bizarre behavior, Globus pallidus 1 4/M/ Bizarre behavior, urinary Cerebellum, white 1.5 and fecal incontinence, matter 5/M/39 Not available Bizarre behavior Globus pallidus, 4 6/M/49 Not available Unremarkable Globus pallidus, 2.5 7/F/ Cognitive impairment Globus pallidus, 3 8/F/ Bizarre behavior, mutism, right limb weakness Cerebral cortex, globus pallidus, Globus pallidus, Modified H&Y Score 9/M/31 38 Bizarre behavior, 1 10/M/28 28 Unremarkable Globus pallidus 1 11/M/43 16 Bizarre behavior Gglobus pallidus, 3 12/M/37 Not available Bizarre behavior, White matter 1 13/M/ Cognitive impairment Globus pallidus 1 CO Intoxication without Parkinsonism 1/F/ Disorientation, cognitive Globus pallidus, Not applicable impairment 2/ M/ Unremarkable Globus pallidus Not applicable 3/F/ Bizarre behavior, Unremarkable Not applicable 4/M/45 Not available Bizarre behavior Globus pallidus Not applicable 5/M/ Cognitive impairment Unremarkable Not applicable 6/M/ Cognitive impairment Cerebellum, globus Not applicable pallidus 7/F/ Unremarkable Cerebral cortex Not applicable 8/F/ Cognitive impairment White matter Not applicable 9/M/ Bizarre behavior White matter Not applicable 10/M/ Unremarkable Globus pallidus, Not applicable 11/F/43 24 Cognitive impairment, Globus pallidus, Not applicable urinary incontinence 12/M/ Unremarkable Globus pallidus, Not applicable 13/M/59 48 Cognitive impairment White matter Not applicable * COHb = carboxyhemoglobin. 2 1 and incidences of FLAIR hyperintensities in the globus pallidus did not show significant differences between the two groups (P =.81,.72, and.20, respectively). The interrater reliability indexes of the three substantia nigra segments for all subjects were high (intraclass correlation coefficient = 0.997, 0.997, and for medial, middle, and lateral segments, respectively). Since there was no significant difference in the normalized ratios of the substantia nigra between the paired left and right segments in each group, the averaged ratios of the paired segments were used for analysis. The averaged ratios of the paired lateral segments of the substantia nigra were in the group with COinduced parkinsonism, in the group with CO intoxication without parkinsonism, and in the group of normal volunteers; those of the paired middle segments were , , and , respectively, and those of the paired medial segments were , , and , respectively. Figure 2 demonstrates the distribution of the normalized ratios of the averaged paired segments in the three groups. The results of multivariate analysis of covariance showed a significant difference in the normalized ratios between the groups (Wilks lambda = 0.639, F 6,64 = 2.676, P =.02). The significance of Wilks lambda was mainly contributed by the middle and lateral segments (P =.01 and.001, respectively). Although we included age and sex as covariates, their correlations with the normalized ratios were not significant (P =.131 and.659, respectively). The post hoc comparisons with Bonferroni method to control familywise error rate suggested that the group with CO-induced parkinsonism significantly differed from the group with CO intoxication without parkinsonism and the group of normal volunteers in the lateral and middle segments, while the three groups did not show significant differences in the medial segments (Fig 2). The normalized ratios of the substantia nigra were not significantly different between patients with FLAIR hyperintensities of 218 radiology.rsna.org n Radiology: Volume 265: Number 1 October 2012

5 Figure 2 Figure 2: Box plot shows distribution of the normalized ratios of the paired segments in the three groups. Multivariate analysis of covariance indicated presence of significant differences among groups, with P,.005 in the lateral and middle segments of the substantia nigra ( ). the globus pallidus and those without (P =.657,.824, and.267 for the medial, middle, and lateral segments of substantial nigra, respectively). On GM-suppressed IR images, the borders of the substantia nigra appeared relatively indistinct in those with CO-induced parkinsonism, particularly in the lateral segments. Figure 3 shows signal alteration of the substantia nigra in a 29-year-old woman with CO-induced parkinsonism, compared with a 31-year-year old woman with CO intoxication without parkinsonism. The normalized ratios of the substantia nigra had a strong positive correlation with the modified H&Y scores in the lateral segments (r = 0.927, P,.001, Fig 4) and weak correlations in the middle and medial segments (r = and 0.576; P =.007 and.04, respectively). Discussion Our study results demonstrate several important clinical implications with regard to delayed neurologic com plications. First, we confirm that substantia nigra is vulnerable to CO intoxication, and the lateral and middle segments of the substantia nigra were more susceptible to CO exposure than the medial segment. Second, although globus pallidus, the most common site of abnormality in the brain after CO intoxication, is often considered to account for the parkinsonian symptoms, the anatomic basis of CO-induced parkinsonism is not well established (10,15). In our series, the abnormality in globus pallidus is not a requisite for CO-induced parkinsonism and can be found in CO-intoxicated patients without parkinsonism. In addition, a significant correlation between hyperintensities of globus pallidus on FLAIR images and the signal abnormalities of the substantia nigra on GM-suppressed IR images does not seem to exist. Third, in contrast to the ambiguous origin of CO-induced parkinsonism, idiopathic Parkinson disease is pathologically characterized by loss of dopaminergic neurons and presence of Lewy bodies in the substantia nigra in a lateral-tomedial gradient (16,17). Our study, to our knowledge, is the first to use a semiquantitative GM-suppressed IR imaging method and shows comparable findings of substantia nigra changes to idiopathic Parkinson disease, which may suggest a vital role of the substantia nigra in the pathogenesis of CO-induced parkinsonism. Various MR imaging techniques have been developed to assess changes of the substantia nigra relative to Parkinson disease. Some studies measured the width or volume of the substantia nigra, with conflicting results (18 20), probably due to issues of partial volume averaging and a nonlinear correlation between the degree of neuronal loss and clinical symptoms (16,21). A number of studies investigated the signal changes of the substantia nigra based on increased iron deposition in Parkinson disease (22,23). However, the range of signal changes may overlap between normal aging and Parkinson disease (24). In the studies of Hutchinson et al, structure changes of pars compacta in Parkinson disease can be revealed by signal intensity ratios of GM-suppressed-to-- suppressed images (14,25,26). In our study, the low signal bands of substantia nigra were considered as a whole, because we could not confidently separate the pars compacta from pars reticulata in such a small area. Instead of using suppressed images as references, we normalized the signal intensity of the substantia nigra by referencing to normal-appearing temporal lobe. This analytic method was deemed sufficient and was expected to eliminate potential confounding factors when the images were compared quantitatively among different subjects. The IR MR images are sensitive to changes of spin-lattice relaxation time T1, which is postulated to mainly reflect water proton relaxation in the presence of protein and lipid-rich intracellular space (27). In normal subjects, bands of hypointensity can be seen in the substantia nigra regions on GM-suppressed IR MR images, supposedly to be due to Radiology: Volume 265: Number 1 October 2012 n radiology.rsna.org 219

6 the abundant dopaminergic neurons in the regions of pars compacta. Hence, the disappearance of hypointensity in the substantia nigra in patients with CO-induced parkinsonism may be explained by the depletion of dopaminergic neurons, tissue loss, and deposition of hemoglobin degradation products due to previous petechial hemorrhage from reperfusion injury; degradation of protein; or presence of free radicals or cellular components of necrotic tissue, as described in postischemic brain injuries (28,29). The pattern of the signal change of the substantia nigra in the group with CO-induced parkinsonism was characterized by a lateral-to-medial regional vulnerability. Although the cause remains unclear, this preferential lateral nigral change may correlate with increased gene expression of mitochondrial electron transport in the lateral segment, which can be directly inhibited by CO with increased oxidative stress in cells (5,30). On the basis of clinical features and functional disabilities, the modified H&Y rating scale is widely used to describe the severity of Parkinson disease (31). In our results, the linear association between the signal alterations in the substantia nigra and the modified H&Y scale suggested the potential of GM-suppressed IR MR imaging to be a biomarker in patients with CO-induced parkinsonism. Similar correlations have been reported between the H&Y scale and positron emission tomographic findings (32). However, each increment of H&Y scale does not necessarily represent a higher degree of overall motor dysfunction. For example, patients with stage 1 disability in the dominant hands are more functionally impaired than patients with disability in the nondominant hands. Hence, the degree of image alterations does not definitely indicate overall functional decline. Furthermore, the IR MR imaging used in our study did not show a significant difference between the bilateral substantia nigra in patients with unilateral disability, modified H&Y score of 1, a finding potentially reflecting a more global effect of CO toxicity. There were several limitations in our study. First, the sample size was small, Figure 3 Figure 3: GM-suppressed IR MR images. (a) Image shows symmetric hypointense bands (arrows) of the substantia nigra posterior to cerebral peduncles in a 31-year-old CO-intoxicated woman without parkinsonism. (b) The substantia nigra appears blurred with increased signal intensity (arrows) in a 29-year-old woman with CO-induced parkinsonism. Figure 4 Figure 4: Scatterplot shows significant correlation between the signal alteration in the lateral aspect of the substantia nigra and the modified H&Y scale by using a generalized linear model (y = x). which warrants further investigations with larger sample sizes. Second, the serum level of carboxyhemoglobin was not obtained in every enrolled patient to assess CO exposure. However, the level may decrease substantially due to high dose of oxygen delivery to the patients before arriving at the emergency department and, thus, was found not closely related to the severity of symptoms in previous studies (7). Third, the simple modified H&Y rating scale was utilized instead of the comprehensive Unified Parkinson s Disease Rating Scale. This may limit in some aspects of its clinical use as far as comparisons among other studies are concerned. Last, we chose the normalappearing temporal lobe as an internal reference to GM-suppressed IR imaging. This method is thus limited 220 radiology.rsna.org n Radiology: Volume 265: Number 1 October 2012

7 to patients with normal-appearing white matter in bilateral temporal lobes. In conclusion, substantia nigra injury plays a role in CO-induced parkinsonism with a lateral to medial regional vulnerability, which can be depicted on GM-suppressed IR MR images. The degree of signal alteration correlates with the severity of clinical symptoms. Further longitudinal studies are required to investigate prognostic implications of the imaging findings in relation to hyperbaric oxygen therapy. Disclosures of Potential Conflicts of Interest: H.W.K. No potential conflicts of interest to disclose. N.Y.C. No potential conflicts of interest to disclose. C.J.H. No potential conflicts of interest to disclose. M.C.C. No potential conflicts of interest to disclose. H.W.C. Financial activities related to the present article: none to disclose. Financial activities not related to the present article: author and institution received research grants from the National Science Council, author received payment from Tri-Service General Hospital for educational lectures on MRI physics, author received payment from the National Science Council for traveling expenses to scientific conferences; part of research grants. Other relationships: none to disclose. M.L. No potential conflicts of interest to disclose. S.W.C. No potential conflicts of interest to disclose. S.Y.C. No potential conflicts of interest to disclose. C.J.J. No potential conflicts of interest to disclose. G.S.H. No potential conflicts of interest to disclose. C.Y.C. No potential conflicts of interest to disclose. References 1. Liu KY, Beautrais A, Caine E, et al. Charcoal burning suicides in Hong Kong and urban Taiwan: an illustration of the impact of a novel suicide method on overall regional rates. J Epidemiol Community Health 2007;61(3): Pan YJ, Liao SC, Lee MB. Suicide by charcoal burning in Taiwan, J Affect Disord 2010;120(1-3): Thom SR, Keim LW. Carbon monoxide poisoning: a review epidemiology, pathophysiology, clinical findings, and treatment options including hyperbaric oxygen therapy. J Toxicol Clin Toxicol 1989;27(3): Peey DG. Hemodynamic response to carbon monoxide. Environ Health Perspect 1988;77: Zhang J, Piantadosi CA. Mitochondrial oxidative stress after carbon monoxide hypoxia in the rat brain. J Clin Invest 1992;90(4): Thom SR. Carbon monoxide-mediated brain lipid peroxidation in the rat. J Appl Physiol 1990;68(3): Ernst A, Zibrak JD. Carbon monoxide poisoning. N Engl J Med 1998;339(22): Choi IS. Parkinsonism after carbon monoxide poisoning. Eur Neurol 2002;48(1): Kawanami T, Kato T, Kurita K, Sasaki H. The pallidoreticular pattern of brain damage on MRI in a patient with carbon monoxide poisoning. J Neurol Neurosurg Psychiatry 1998;64(2): O Doell P, Buxton PJ, Pitkin A, Jarvis LJ. The magnetic resonance imaging appearances of the brain in acute carbon monoxide poisoning. Clin Radiol 2000;55(4): Piccini P, Pavese N, Canapicchi R, et al. White matter hyperintensities in Parkinson s disease: clinical correlations. Arch Neurol 1995;52(2): Lo CP, Chen SY, Chou MC, et al. Diffusion-tensor MR imaging for evaluation of the efficacy of hyperbaric oxygen therapy in patients with delayed neuropsychiatric syndrome caused by carbon monoxide inhalation. Eur J Neurol 2007;14(7): Chang KH, Han MH, Kim HS, Wie BA, Han MC. Delayed encephalopathy after acute carbon monoxide intoxication: MR imaging features and distribution of cerebral white matter lesions. Radiology 1992;184(1): Hutchinson M, Raff U. Structural changes of the substantia nigra in Parkinson s disease as revealed by MR imaging. AJNR Am J Neuroradiol 2000;21(4): Prockop LD. Carbon monoxide brain toxicity: clinical, magnetic resonance imaging, magnetic resonance spectroscopy, and neuropsychological effects in 9 people. J Neuroimaging 2005;15(2): Ma SY, Rie JO, Collan Y, Röyttä M, Rie UK. A quantitative morphometrical study of neuron degeneration in the substantia nigra in Parkinson s disease. J Neurol Sci 1996;140(1-2): Hsiao CL, Kuo HC, Huang CC. Delayed encephalopathy after carbon monoxide intoxication: long-term prognosis and correlation of clinical manifestations and neuroimages. Acta Neurol Taiwan 2004;13(2): Duguid JR, De La Paz R, DeGroot J. Magnetic resonance imaging of the midbrain in Parkinson s disease. A Neurol 1986;20(6): Menke RA, Scholz J, Miller KL, et al. MRI characteristics of the substantia nigra in Parkinson s disease: a combined quantitative T1 and DTI study. Neuroimage 2009;47(2): Adachi M, Hosoya T, Haku T, Yamaguchi K, Kawanami T. Evaluation of the substantia nigra in patients with Parkinsonian syndrome accomplished using multishot diffusion-weighted MR imaging. AJNR Am J Neuroradiol 1999;20(8): Marsden CD. Parkinson s disease. Lancet 1990;335(8695): Baudrexel S, Nürnberger L, Rüb U, et al. Quantitative mapping of T1 and T2* discloses nigral and brainstem pathology in early Parkinson s disease. Neuroimage 2010;51(2): Gorell JM, Ordidge RJ, Brown GG, Deniau JC, Buderer NM, Helpern JA. Increased iron-related MRI contrast in the substantia nigra in Parkinson s disease. Neurology 1995;45(6): Antonini A, Leenders KL, Meier D, Oertel WH, Boesiger P, Anliker M. T2 relaxation time in patients with Parkinson s disease. Neurology 1993;43(4): Hutchinson M, Raff U. Parkinson s disease: a novel MRI method for determining structural changes in the substantia nigra. J Neurol Neurosurg Psychiatry 1999;67(6): Hutchinson M, Raff U. Detection of Parkinson s disease by MRI: spin-lattice distribution imaging. Mov Disord 2008;23(14): Hutchinson M, Raff U, Lebedev S. MRI correlates of pathology in parkinsonism: segmented inversion recovery ratio imaging (SIRRIM). Neuroimage 2003;20(3): Fujioka M, Okuchi K, Hiramatsu KI, Sakaki T, Sakaguchi S, Ishii Y. Specific changes in human brain after hypoglycemic injury. Stroke 1997;28(3): Komiyama M, Nakajima H, Nishikawa M, Yasui T. Chronological changes in nonhaemorrhagic brain infarcts with short T1 in the cerebellum and basal ganglia. Neuroradiology 2000;42(7): Duke DC, Moran LB, Pearce RK, Graeber MB. The medial and lateral substantia nigra in Parkinson s disease: mrna profiles associated with higher brain tissue vulnerability. Neurogenetics 2007;8(2): Hoehn MM, Yahr MD. Parkinsonism: onset, progression and mortality. Neurology 1967;17(5): Eidelberg D, Moeller JR, Ishikawa T, et al. Assessment of disease severity in parkinsonism with fluorine-18-fluorodeoxyglucose and PET. J Nucl Med 1995;36(3): Radiology: Volume 265: Number 1 October 2012 n radiology.rsna.org 221