Original Paper Eur Neurol 2014;71:299–304 DOI: 10.1159/000357210
Received: July 29, 2013 Accepted: November 10, 2013 Published online: March 21, 2014
Hemichorea Associated with Nonketotic Hyperglycemia: Clinical and Neuroimaging Features in 12 Patients Yan Guo Yan-wei Miao Xiao-fei Ji Ming Li Xuan Liu Xiao-pei Sun Department of Neurology, The First Affiliated Hospital of Dalian Medical University, Dalian, China
Key Words Chorea · Diabetes mellitus · Magnetic resonance imaging
Abstract Background: Nonketotic hyperglycemia is a rare cause of hemichorea. Patients with hemichorea associated with nonketotic hyperglycemia (HCNH) always have a favorable prognosis when given prompt treatment. Methods: We reviewed the medical records of 12 patients with HCNH in our hospital between January 2005 and January 2013. The clinical data, laboratory findings, and imaging features of the patients were collected. Results: All 12 patients were admitted to the hospital with a complaint of involuntary movements. Ten patients had a history of diabetes, while the other 2 patients had not been diagnosed. The mean level of blood glucose on admission was 330.7 ± 107.8 mg/dl, and the ketones were negative. A cranial computed tomography scan showed hyperdensity in the striatum, which quickly resolved. Magnetic resonance imaging showed hyperintensity on T1-weighted images without change over several months. Nearly all of the patients experienced relief from the hemichorea symptoms after correcting hyperglycemia with a combination of dopamine receptor inhibitors and the sedative lorazepam, if necessary. Conclusion: HCNH is a benign disorder, the patho-
© 2014 S. Karger AG, Basel 0014–3022/14/0716–0299$39.50/0 E-Mail [email protected] www.karger.com/ene
genesis of which remains unclear. Radiologic changes can provide guidance for early treatment and generally give an estimation of the degree of injury. © 2014 S. Karger AG, Basel
Introduction
Hemichorea involves a spectrum of continuous, nonpatterned, and involuntary movements on one side of the body. The etiologies include ischemic cerebrovascular lesions, metabolic dysfunction, immunologic and infectious diseases, neurodegenerative disorders, and systemic diseases (lupus erythematosus and thyrotoxicosis) [1, 2]. Nonketotic hyperglycemia is a rare cause of hemichorea, the pathogenic mechanism of which is unclear. Hemichorea associated with nonketotic hyperglycemia (HCNH) usually occurs in elderly diabetic women. It is characterized by unilateral involuntary movements, contralateral striatal abnormalities, and rapid remission of symptoms after the correction of hyperglycemia. Hyperdensity and hyperintensity in the striatum on computed tomography (CT) and magnetic resonance (MR) images, respectively, contribute to the diagnosis. With early treatment, the neuroimages can resolve completeXiao-pei Sun Department of Neurology The First Affiliated Hospital of Dalian Medical University Dalian 116011 (China) E-Mail rxrs69 @ sina.com
Table 1. Clinical features of 12 patients with HCNH
No.
Age, years/ gender
Diabetes melli- Glucose, mg/dl tus history, years
HbA1c, %
Osmolality, mmol/kg
Chorea
CT lesion
MRI lesion
Treatment
1 2 3 4 5 6 7 8 9 10 11 12
81/F 79/F 74/F 82/F 73/F 82/F 76/F 69/F 65/F 70/M 74/M 74/M
0 10 20 0 8 7 8 15 10 8 2 10
9.3 8.6 10.7 9.5 9.6 15.7 28.9 12 13.7 13.6 12.4 13
303 309 299 313 308 296 347 303 340 328 306 312
a a, l a, l f, a, l a, l f, a, l f, a, l a a, l f, a, l a, l a, l
Nor P, CN Nor P, CN P, CN P, CN, GP P P P P P, CN Nor
Nor – P P, CN – P, CN, GP – – P P, GP P, CN –
G G G, H G, H G G, H, C G, D G, T G G, H G, H G, C
214 241 180 239 283 504 277 468 380 355 460 367
– = Not available; f = face; a = arm; l = leg; Nor = normal; P = putamen; CN = caudate nucleus; GP = globus pallidus; G = blood glucose reduction; H = haloperidol; D = diazepam; C = clonazepam; T = tiapride.
ly or partially, and a favorable prognosis is always achieved. The pathogenesis of HCNH is controversial, although many hypotheses have been advanced. The theory of metabolic disorders, being the major hypothesis, suggests that the tricarboxylic acid cycle becomes inactivated during hyperglycemic crises, resulting in a shift in the brain to anaerobic metabolism. To provide an alternative source of energy, the brain metabolizes GABA to succinic acid; however, such a GABA shunt meets only 10–40% of the energy required by the basal ganglia, leading to metabolic acidosis. As a result, acetylcholine is reduced by acetate depletion. Metabolic acidosis, energy depletion, decreased GABA, and acetylcholine levels may cause dysfunction in the basal ganglia, resulting in hemichorea [3]. In addition, upregulation of dopamine receptors, dopaminergic hypersensitivity associated with estrogen, and blood-brain barrier disruption are also thought to be involved in the pathogenesis. Nevertheless, no single mechanism can account for all the symptom characteristics. HCNH is more common clinically than ever before, but it has always been misdiagnosed due to insufficient recognition. In a prospective study involving hyperglycemia-associated movement disorders, all 10 patients were misdiagnosed on initial evaluation by a general practitioner [4]. Because the number of patients with HCNH reported in the literature is limited, we analyzed the clinical presentation, neuroimaging features, and evolution at follow-up of 12 patients with HCNH. 300
Eur Neurol 2014;71:299–304 DOI: 10.1159/000357210
Materials and Methods The medical records at our hospital (the First Affiliated Hospital of Dalian Medical University, Dalian, China) were reviewed between January 2005 and January 2013. Twelve patients were identified who had presented with hemichorea on admission and shown to be in a nonketotic hyperglycemic state without other causes of hemichorea. Clinical data, including laboratory and radiologic findings, were obtained. Data are expressed as means ± SD or frequency.
Results
Demographic Data The patient characteristics are given in table 1. There were 9 women and 3 men. Their mean age was 74.9 ± 5.4 years (range, 65–82 years). With the exception of 2 patients diagnosed with diabetes mellitus for the first time, 10 patients were known to have had diabetes mellitus for 2–20 years. All of the patients had recently had poor control of blood glucose levels. The mean time interval between the onset of hemichorea and admission was 18.3 ± 15.7 days (range, 4–60 days). Clinical Features All 12 patients were admitted to the hospital with complaints of involuntary movements, increasing during nervousness and decreasing during sleep. The involuntary hemichoreic movements involved both upper and lower Guo/Miao/Ji/Li/Liu/Sun
Case 1
Case 3
Case 4
Case 6
Case 9
Case 10
Case 11
Fig. 1. MR imaging was performed in 7 patients at the time of hospitalization (for clinical and topographic lesions, see table 1). MR imaging revealed hyperintensity in the striatum on T1-weighted images except in case 1. Next to each T1-weighted image, there is a T2-weighted image, showing hypointensity (cases 3, 9, and 10), hyperintensity (cases 6, and 11), and isointensity (case 4).
extremities in 10 patients, while the upper extremities alone were affected in the other 2 patients. In 4 patients, facial involvement was described, including tapir mouth, grimacing, and extension of the tongue. On neurologic examination, except for the choreic movements and decreased muscle tone in the affected extremities, no other signs were evident.
Neuroradiologic Findings All 12 patients underwent a brain CT scan on admission, which showed hyperdensity in the striatum in 10 patients, with the exception of 2 patients with no findings. Eight patients with hemichorea had a lesion in the contralateral basal ganglia, while the other 2 patients had a lesion in the bilateral basal ganglia. The mean time interval between the onset of hemichorea and CT scan was
19.2 ± 16.1 days (range, 4–60 days). One patient had a CT scan performed 45 days before the onset of hemichorea because diabetic ketoacidosis occurred while in the hospital. The CT scan had already shown the above changes, indicating that changes in CT scans may be observed before the onset of clinical symptoms. The 2 patients with normal CT scans underwent MR imaging within 5 days of admission; MR imaging in 1 patient revealed hyperintensity in the putamen, while no abnormal signals were demonstrated in the other patient. Except for the 2 abovementioned patients, 5 patients underwent brain MR studies, revealing hyperintense basal ganglia lesions on T1weighted images. All MR images are shown in figure 1. The putamen was involved in all 6 patients with abnormal signals. The caudate nucleus (3 of 6 patients) and globus pallidus (1 of 6 patients) may also be affected. The internal capsule was not involved in our 12 patients. In contrast to the consistent hyperintensity on T1-weighted images, the T2-weighted images showed hypointensity (3 of 6 patients), hyperintensity (2 of 6 patients), and isointensity (1 of 6 patients) in the basal ganglia region. Susceptibility-weighted imaging was performed in 1 patient, which showed a normal signal in the striatum without hypointensity (fig. 2; case 6).
Hemichorea Associated with Nonketotic Hyperglycemia
Eur Neurol 2014;71:299–304 DOI: 10.1159/000357210
Laboratory Findings The mean serum glucose level on admission was 330.7 ± 107.8 mg/dl (range, 180–504 mg/dl). The serum HbA1c levels were elevated by a mean of 11.4 ± 2.3% (range, 8.6– 15.7%) and the mean serum osmolarities were 313.7 ± 16.2 mmol/kg (range, 296–347 mmol/kg). Urinalysis was negative for ketones.
301
Fig. 2. Case 6: an 82-year-old woman was admitted to the emergency department with a complaint of involuntary movements on the right side of her body for 30 days. MR imaging showed hyperintensity on the T1-weighted images (a) and slight hyperintensity on the T2-weighted images (b). Diffusion-weighted imaging (c) revealed slight hypointensity. MR spectroscopy (d) showed a normal signal.
a
c
b
d
Fig. 3. Case 3: a 74-year-old woman was ad-
mitted to the emergency department with a complaint of involuntary movements on the left side of her body for 60 days. a CT on admission was normal, while MR imaging revealed a hyperintensity in the putamen on the T1-weighted images (b) and hypointensity on the T2-weighted images (c). d The lesion is shown in the sagittal section. The symptoms resolved 2 weeks after antidiabetic and haloperidol treatment. One year later, she underwent follow-up MR imaging, which revealed hyperintensity reduced on the T1-weighted images (f) and hypointensity on the T2-weighted images (g). h Sagittal section corresponding to image f. Fifteen months later, followup CT (e) showed slight hypoattenuation in the lesion; no other abnormalities were noted except calcifications.
a
b
c
d
e
f
g
h
Treatment and Follow-Up Visits Four of the 12 patients receiving antidiabetic monotherapy experienced complete relief from hemichoreic symptoms when euglycemia was achieved. We followed these 4 patients for 3–6 months; 3 patients had recurrences within 2 weeks of hospital discharge, and 1 patient who had a normal CT scan on admission was without recurrence. In addition to glucose control, haloperidol treatment was provided as monotherapy or in combination with other medications, such as tiapride, chlorpromazine, or diazepam, in the remaining 8 patients. Hemichorea also completely resolved in these 8 patients within 5 days to 2 months. Two of the 8 patients had recurrent chorea after reducing or discontinuing haloperidol treatment. After restarting haloperidol monotherapy or combination therapy, chorea resolved again. Among the 12 patients, 9 had a follow-up brain CT scan within an interval of 1–18 months. In 8 of the 9 patients, the hyperdense striatal lesions had disappeared completely or near-completely on the CT scans (fig. 3; case 3), while the hyperdensity decreased within 2 months in the remaining patient. Six patients received follow-up 302
Eur Neurol 2014;71:299–304 DOI: 10.1159/000357210
MR imaging. The hyperintense signals disappeared after 6–18 months in 2 patients (fig. 4; case 4), while the hyperintense signals decreased in 6–24 months in the remaining 4 patients.
Discussion
HCNH was first described by Bedwell in 1960 [5] and is characterized by a typical triad, including unilateral involuntary movements, contralateral striatal abnormalities, and rapid remission of symptoms after correction of hyperglycemia. This transient movement disorder is more frequently seen in elderly women, especially Asians, and has been reported as a rare disease over the past 2 decades. Although the pathogenesis of HCNH is still unclear, the strong association between hyperglycemia and hemichorea has been supported in almost all of the cases. Recent studies involving HCNH suggest that it might be a rare complication or the initial manifestation [6] of diabetes mellitus. In our study, whether or not a diabetic history existed, all 12 patients had nonketotic hyperglyGuo/Miao/Ji/Li/Liu/Sun
a
b
c
d
f
g
h
i
e
Fig. 4. Case 4: an 82-year-old woman was admitted to the emergency department with a complaint of involuntary movements on the right side of her body for 4 days. a CT on admission revealed hyperattenuation in the putamen and caudate nucleus. MR imaging showed hyperintensity on the T1-weighted images (b) and isointensity on the T2-weighted images (c). d, f The lesion is shown
in the sagittal and coronal sections. One month after antidiabetic and haloperidol treatment, the hyperattenuation in CT was reduced (f). She underwent follow-up MR imaging (g–i) after 18 months, which revealed that the hyperintensity had resolved, and hypointensity was noted in the primary lesion.
cemia on admission and were diagnosed with diabetes mellitus at the time of hospital discharge. With prompt treatment, all of the patients achieved relief from choreic symptoms, and follow-up neuroimaging showed reversible disappearance of the symptoms in some of the patients. The neuroradiologic findings of HCNH are specific and involve the putamen in all cases, head of the caudate nucleus in most cases, and the globus pallidus in a minority of cases. Lesions are unilateral in the vast majority of patients, though sometimes bilateral involvement has been reported, even in patients who have not yet manifested symptoms, which is consistent with a previous report [3]. To detect the typical lesion of HCNH, MR imaging is more sensitive and superior to CT, revealing reversible hyperintensity in the striatum without signs of a mass effect, edema, volume loss, or internal capsule involvement. Because the lesions correspond to the territory of the lenticulostriate artery, researchers are of the opinion that the striatal lesions may be the result of a vascular insult. Patients with diabetes mellitus and poorly controlled hyperglycemia are at increased risk for cerebrovascular ischemia. However, vascular insufficiency alone does not explain the mechanisms responsible for some bilateral lesions. In addition, calcifications, petechial hemorrhage, myelinolysis, wallerian degeneration, and deposition of manganese have been suggested as possible causes of the
reversible hyperintensity in T1-weighted images [7–12]; however, biopsy and autopsy studies only show selective neuronal loss, gliosis, and reactive astrocytosis [13–15]. Similarly, MR spectroscopy has shown decreased N-acetylaspartate and creatine peaks with a low N-acetylaspartate/creatine ratio, suggesting neuronal dysfunction, loss, or damage, while a high choline/creatine ratio likely indicates gliosis [16]. It has also been postulated that the typical T1-weighted MR images may be due to regional ischemic injury related to hyperglycemia-induced hyperviscosity. Diffusion-weighted imaging and apparent diffusion coefficient maps may show restricted diffusion, which has been ascribed to hyperviscosity [17]. SPECT studies have demonstrated a significant decrease in blood flow in the basal ganglia contralateral to the chorea [18]. 18 F-fluorodeoxyglucose positron emission tomography has shown a reduction in glucose metabolism in the corresponding lesions in T1-weighted MR images in animals [19]. All of the above image changes indicate that the metabolic derangements associated with hyperglycemia play a significant role in the pathogenesis, rather than vascular disease. In the current study, normal susceptibilityweighted imaging excludes the possibility of hemorrhage, thus we consider that metabolic disorders associated with hyperglycemia resulting in hemichorea and vascular insufficiency may contribute to this regional metabolic failure.
Hemichorea Associated with Nonketotic Hyperglycemia
Eur Neurol 2014;71:299–304 DOI: 10.1159/000357210
303
The prognosis of HCNH has been reported to be favorable [3] if there is prompt glycemic control, which is similar to our results. Patients with a slight injury in their brains resulting from hyperglycemia can get a complete remission with simple antidiabetic treatment, although additional neuroleptics are needed for most patients. However, a minority of patients have persistent chorea for years or longer [20], suggesting severe damage in the basal ganglia. In our study, all of the patients experienced symptom relief after early treatment, most of whom (7 of 12 patients) did not have recurrences at the follow-up visit. Four patients received antidiabetic monotherapy and 3 patients with typical abnormal images on admission had recurrence within 2 weeks after hospital discharge. In contrast, the patient without image changes had complete remission at the 1-year follow-up. We therefore speculate that images may predict the degree of injury. For patients
with hyperdense or hyperintense signals within the basal ganglia on CT or MR images, additional symptomatic treatment may be needed. In conclusion, HCNH is a reversible, benign disorder. Radiologic changes may be a predictor of the degree of injury to the brain. Hemichoreic symptoms can be completely or partly ameliorated and hyperintensity in the striatum on T1-weighted images may resolve after euglycemia is restored. Recognition of the unique clinicoradiologic manifestations is important because early correction of the underlying hyperglycemia will contribute to rapid improvement. Although the pathogenesis of HCNH remains uncertain, current evidence suggests that metabolic changes play a main role. Further autopsy data and clinical research based on a sufficient sample size are needed to verify the pathogenesis and nature of neuroradiologic changes.
References 1 Higa M, Kaneko Y, Inokuchi T: Two cases of hyperglycaemic chorea in diabetic patients. Diabet Med 2004;21:196–198. 2 Cheema H, Federman D, Kam A: Hemichorea-hemiballismus in non-ketotic hyperglycaemia. J Clin Neurosci 2011;18:294–196. 3 Oh SH, Lee KY, Im JH, lee MS: Chorea associated with nonketotic hyperglycemia and hyperintensity basal ganglia lesion on T1weighted brain MRI study: a meta-analysis of 53 cases including four present cases. J Neurol Sci 2002;200:57–62. 4 Cervantes-Arriaga A, Arrambide G, Rodríguez-Violante M: A prospective series of patients with hyperglycaemia-associated movement disorders. J Clin Neurosci 2011; 18: 1329–1332. 5 Bedwell SF: Some observations on hemiballism. Neurology 1960;10:619–622. 6 Qi X, Yan Y-Y, Gao Y, Zheng Z-S, Chang Y: Hemichorea associated with non-ketotic hyperglycaemia: a case report. Diabetes Res Clin Pract 2012;95:e1–e3. 7 Sanfield JA, Finkel J, Lewis S, Rosen SG: Alternating choreoathetosis associated with uncontrolled diabetes mellitus and basal ganglia calcification. Diabetes Care 1986;9:100–101. 8 Chang MH, Chiang HT, Lai PH, Sy CG, Lee SS, Lo YY: Putaminal petechial haemorrhage as the cause of chorea: a neuroimaging study. J Neurol Neurosurg Psychiatry 1997; 63: 300– 303.
304
9 Nagai C, Kato T, Katagiri T, Sasaki H: Hyperintense putamen in T1-weighted MR images in a case of chorea with hyperglycemia. AJNR Am J Neuroradiol 1995;16:1243–1246. 10 Sajth J, Ditchfield A, Katifi HA: Extrapontine myelinolysis presenting as acute parkinsonism. BMC Neurol 2006;6:33. 11 Wintermark M, Fischbein NJ, Mukherjee P, Yuh EL, Dillon WP: Unilateral putaminal CT, MR, and diffusion abnormalities secondary to nonketotic hyperglycemia in the setting of acute neurologic symptoms mimicking stroke. AJNR Am J Neuroradiol 2004;25:975– 976. 12 Fujioka M, Taoka T, Matsuo Y, Mishima K, Oqoshi K, Kondo Y, Tsuda M, Asano T, Sakaki T, Miyasaki A, Park D, Siesjö BK: Magnetic resonance imaging shows delayed ischemic striatal neurodegeneration. Ann Neurol 2003; 54:732–747. 13 Shan DH, Ho DM, Chang C, Teng MM: Hemichorea-hemiballism: an explanation for MR signal changes. AJNR Am J Neuroradiol 1998;19:863–870. 14 Ohara S, Nakagawa S, Tabata K: Hemiballism with hyperglycemia and atriatal T1-MRI hyperintensity: an autopsy report. Mov Disord 2001;16:521–525.
Eur Neurol 2014;71:299–304 DOI: 10.1159/000357210
15 Lai PH, Tien RD, Chang MH, Teng MM, Yang CF, Pan HB, Chen C, Liring C, Ling JF, Kong KW: Chorea-ballismus with nonketotic hyperglycemia in primary diabetes mellitus. AJNR Am J Neuroradiol 1996;17:1057–1064. 16 Lai PH, Chen PC, Chang MH, Pan HB, Yang CF, Wu MT, Li JY, Chen C, Liang HL, Chen WL: In vivo proton MR spectroscopy of chorea-ballismus in diabetes mellitus. Neuroradiology 2001;43:525–531. 17 Chu K, Kang DW, Kin DE, Park SH, Roh JK: Diffusion-weighted and gradient echo magnetic resonance findings of hemichorea hemiballismus associated with diabetic hyperglycemia: a hyperviscosity syndrome? Arch Neurol 2002;59:448–452. 18 Chang MH, Li JY, Lee SR, Men CY: Non-ketotic hyperglycaemic chorea: a SPECT study. J Neurosurg Psychiatry 1996;60:428–430. 19 Hsu JL, Wang HC, Hsu WC: Hyperglycemiainduced unilateral basal ganglia lesions with or without hemichorea: a PET study. J Neurol 2004;251:1486–1490. 20 Ahlskog JE, Nishino H, Evidente VGH, Tulloch JW, Forbes GS, Caviness JN, GwinnHardy KA: Persistent chorea triggered by hyperglycemic crisis in diabetics. Mov Disord 2001;16:890–898.
Guo/Miao/Ji/Li/Liu/Sun
Copyright: S. Karger AG, Basel 2014. Reproduced with the permission of S. Karger AG, Basel. Further reproduction or distribution (electronic or otherwise) is prohibited without permission from the copyright holder.
Received: July 29, 2013 Accepted: November 10, 2013 Published online: March 21, 2014
Hemichorea Associated with Nonketotic Hyperglycemia: Clinical and Neuroimaging Features in 12 Patients Yan Guo Yan-wei Miao Xiao-fei Ji Ming Li Xuan Liu Xiao-pei Sun Department of Neurology, The First Affiliated Hospital of Dalian Medical University, Dalian, China
Key Words Chorea · Diabetes mellitus · Magnetic resonance imaging
Abstract Background: Nonketotic hyperglycemia is a rare cause of hemichorea. Patients with hemichorea associated with nonketotic hyperglycemia (HCNH) always have a favorable prognosis when given prompt treatment. Methods: We reviewed the medical records of 12 patients with HCNH in our hospital between January 2005 and January 2013. The clinical data, laboratory findings, and imaging features of the patients were collected. Results: All 12 patients were admitted to the hospital with a complaint of involuntary movements. Ten patients had a history of diabetes, while the other 2 patients had not been diagnosed. The mean level of blood glucose on admission was 330.7 ± 107.8 mg/dl, and the ketones were negative. A cranial computed tomography scan showed hyperdensity in the striatum, which quickly resolved. Magnetic resonance imaging showed hyperintensity on T1-weighted images without change over several months. Nearly all of the patients experienced relief from the hemichorea symptoms after correcting hyperglycemia with a combination of dopamine receptor inhibitors and the sedative lorazepam, if necessary. Conclusion: HCNH is a benign disorder, the patho-
© 2014 S. Karger AG, Basel 0014–3022/14/0716–0299$39.50/0 E-Mail [email protected] www.karger.com/ene
genesis of which remains unclear. Radiologic changes can provide guidance for early treatment and generally give an estimation of the degree of injury. © 2014 S. Karger AG, Basel
Introduction
Hemichorea involves a spectrum of continuous, nonpatterned, and involuntary movements on one side of the body. The etiologies include ischemic cerebrovascular lesions, metabolic dysfunction, immunologic and infectious diseases, neurodegenerative disorders, and systemic diseases (lupus erythematosus and thyrotoxicosis) [1, 2]. Nonketotic hyperglycemia is a rare cause of hemichorea, the pathogenic mechanism of which is unclear. Hemichorea associated with nonketotic hyperglycemia (HCNH) usually occurs in elderly diabetic women. It is characterized by unilateral involuntary movements, contralateral striatal abnormalities, and rapid remission of symptoms after the correction of hyperglycemia. Hyperdensity and hyperintensity in the striatum on computed tomography (CT) and magnetic resonance (MR) images, respectively, contribute to the diagnosis. With early treatment, the neuroimages can resolve completeXiao-pei Sun Department of Neurology The First Affiliated Hospital of Dalian Medical University Dalian 116011 (China) E-Mail rxrs69 @ sina.com
Table 1. Clinical features of 12 patients with HCNH
No.
Age, years/ gender
Diabetes melli- Glucose, mg/dl tus history, years
HbA1c, %
Osmolality, mmol/kg
Chorea
CT lesion
MRI lesion
Treatment
1 2 3 4 5 6 7 8 9 10 11 12
81/F 79/F 74/F 82/F 73/F 82/F 76/F 69/F 65/F 70/M 74/M 74/M
0 10 20 0 8 7 8 15 10 8 2 10
9.3 8.6 10.7 9.5 9.6 15.7 28.9 12 13.7 13.6 12.4 13
303 309 299 313 308 296 347 303 340 328 306 312
a a, l a, l f, a, l a, l f, a, l f, a, l a a, l f, a, l a, l a, l
Nor P, CN Nor P, CN P, CN P, CN, GP P P P P P, CN Nor
Nor – P P, CN – P, CN, GP – – P P, GP P, CN –
G G G, H G, H G G, H, C G, D G, T G G, H G, H G, C
214 241 180 239 283 504 277 468 380 355 460 367
– = Not available; f = face; a = arm; l = leg; Nor = normal; P = putamen; CN = caudate nucleus; GP = globus pallidus; G = blood glucose reduction; H = haloperidol; D = diazepam; C = clonazepam; T = tiapride.
ly or partially, and a favorable prognosis is always achieved. The pathogenesis of HCNH is controversial, although many hypotheses have been advanced. The theory of metabolic disorders, being the major hypothesis, suggests that the tricarboxylic acid cycle becomes inactivated during hyperglycemic crises, resulting in a shift in the brain to anaerobic metabolism. To provide an alternative source of energy, the brain metabolizes GABA to succinic acid; however, such a GABA shunt meets only 10–40% of the energy required by the basal ganglia, leading to metabolic acidosis. As a result, acetylcholine is reduced by acetate depletion. Metabolic acidosis, energy depletion, decreased GABA, and acetylcholine levels may cause dysfunction in the basal ganglia, resulting in hemichorea [3]. In addition, upregulation of dopamine receptors, dopaminergic hypersensitivity associated with estrogen, and blood-brain barrier disruption are also thought to be involved in the pathogenesis. Nevertheless, no single mechanism can account for all the symptom characteristics. HCNH is more common clinically than ever before, but it has always been misdiagnosed due to insufficient recognition. In a prospective study involving hyperglycemia-associated movement disorders, all 10 patients were misdiagnosed on initial evaluation by a general practitioner [4]. Because the number of patients with HCNH reported in the literature is limited, we analyzed the clinical presentation, neuroimaging features, and evolution at follow-up of 12 patients with HCNH. 300
Eur Neurol 2014;71:299–304 DOI: 10.1159/000357210
Materials and Methods The medical records at our hospital (the First Affiliated Hospital of Dalian Medical University, Dalian, China) were reviewed between January 2005 and January 2013. Twelve patients were identified who had presented with hemichorea on admission and shown to be in a nonketotic hyperglycemic state without other causes of hemichorea. Clinical data, including laboratory and radiologic findings, were obtained. Data are expressed as means ± SD or frequency.
Results
Demographic Data The patient characteristics are given in table 1. There were 9 women and 3 men. Their mean age was 74.9 ± 5.4 years (range, 65–82 years). With the exception of 2 patients diagnosed with diabetes mellitus for the first time, 10 patients were known to have had diabetes mellitus for 2–20 years. All of the patients had recently had poor control of blood glucose levels. The mean time interval between the onset of hemichorea and admission was 18.3 ± 15.7 days (range, 4–60 days). Clinical Features All 12 patients were admitted to the hospital with complaints of involuntary movements, increasing during nervousness and decreasing during sleep. The involuntary hemichoreic movements involved both upper and lower Guo/Miao/Ji/Li/Liu/Sun
Case 1
Case 3
Case 4
Case 6
Case 9
Case 10
Case 11
Fig. 1. MR imaging was performed in 7 patients at the time of hospitalization (for clinical and topographic lesions, see table 1). MR imaging revealed hyperintensity in the striatum on T1-weighted images except in case 1. Next to each T1-weighted image, there is a T2-weighted image, showing hypointensity (cases 3, 9, and 10), hyperintensity (cases 6, and 11), and isointensity (case 4).
extremities in 10 patients, while the upper extremities alone were affected in the other 2 patients. In 4 patients, facial involvement was described, including tapir mouth, grimacing, and extension of the tongue. On neurologic examination, except for the choreic movements and decreased muscle tone in the affected extremities, no other signs were evident.
Neuroradiologic Findings All 12 patients underwent a brain CT scan on admission, which showed hyperdensity in the striatum in 10 patients, with the exception of 2 patients with no findings. Eight patients with hemichorea had a lesion in the contralateral basal ganglia, while the other 2 patients had a lesion in the bilateral basal ganglia. The mean time interval between the onset of hemichorea and CT scan was
19.2 ± 16.1 days (range, 4–60 days). One patient had a CT scan performed 45 days before the onset of hemichorea because diabetic ketoacidosis occurred while in the hospital. The CT scan had already shown the above changes, indicating that changes in CT scans may be observed before the onset of clinical symptoms. The 2 patients with normal CT scans underwent MR imaging within 5 days of admission; MR imaging in 1 patient revealed hyperintensity in the putamen, while no abnormal signals were demonstrated in the other patient. Except for the 2 abovementioned patients, 5 patients underwent brain MR studies, revealing hyperintense basal ganglia lesions on T1weighted images. All MR images are shown in figure 1. The putamen was involved in all 6 patients with abnormal signals. The caudate nucleus (3 of 6 patients) and globus pallidus (1 of 6 patients) may also be affected. The internal capsule was not involved in our 12 patients. In contrast to the consistent hyperintensity on T1-weighted images, the T2-weighted images showed hypointensity (3 of 6 patients), hyperintensity (2 of 6 patients), and isointensity (1 of 6 patients) in the basal ganglia region. Susceptibility-weighted imaging was performed in 1 patient, which showed a normal signal in the striatum without hypointensity (fig. 2; case 6).
Hemichorea Associated with Nonketotic Hyperglycemia
Eur Neurol 2014;71:299–304 DOI: 10.1159/000357210
Laboratory Findings The mean serum glucose level on admission was 330.7 ± 107.8 mg/dl (range, 180–504 mg/dl). The serum HbA1c levels were elevated by a mean of 11.4 ± 2.3% (range, 8.6– 15.7%) and the mean serum osmolarities were 313.7 ± 16.2 mmol/kg (range, 296–347 mmol/kg). Urinalysis was negative for ketones.
301
Fig. 2. Case 6: an 82-year-old woman was admitted to the emergency department with a complaint of involuntary movements on the right side of her body for 30 days. MR imaging showed hyperintensity on the T1-weighted images (a) and slight hyperintensity on the T2-weighted images (b). Diffusion-weighted imaging (c) revealed slight hypointensity. MR spectroscopy (d) showed a normal signal.
a
c
b
d
Fig. 3. Case 3: a 74-year-old woman was ad-
mitted to the emergency department with a complaint of involuntary movements on the left side of her body for 60 days. a CT on admission was normal, while MR imaging revealed a hyperintensity in the putamen on the T1-weighted images (b) and hypointensity on the T2-weighted images (c). d The lesion is shown in the sagittal section. The symptoms resolved 2 weeks after antidiabetic and haloperidol treatment. One year later, she underwent follow-up MR imaging, which revealed hyperintensity reduced on the T1-weighted images (f) and hypointensity on the T2-weighted images (g). h Sagittal section corresponding to image f. Fifteen months later, followup CT (e) showed slight hypoattenuation in the lesion; no other abnormalities were noted except calcifications.
a
b
c
d
e
f
g
h
Treatment and Follow-Up Visits Four of the 12 patients receiving antidiabetic monotherapy experienced complete relief from hemichoreic symptoms when euglycemia was achieved. We followed these 4 patients for 3–6 months; 3 patients had recurrences within 2 weeks of hospital discharge, and 1 patient who had a normal CT scan on admission was without recurrence. In addition to glucose control, haloperidol treatment was provided as monotherapy or in combination with other medications, such as tiapride, chlorpromazine, or diazepam, in the remaining 8 patients. Hemichorea also completely resolved in these 8 patients within 5 days to 2 months. Two of the 8 patients had recurrent chorea after reducing or discontinuing haloperidol treatment. After restarting haloperidol monotherapy or combination therapy, chorea resolved again. Among the 12 patients, 9 had a follow-up brain CT scan within an interval of 1–18 months. In 8 of the 9 patients, the hyperdense striatal lesions had disappeared completely or near-completely on the CT scans (fig. 3; case 3), while the hyperdensity decreased within 2 months in the remaining patient. Six patients received follow-up 302
Eur Neurol 2014;71:299–304 DOI: 10.1159/000357210
MR imaging. The hyperintense signals disappeared after 6–18 months in 2 patients (fig. 4; case 4), while the hyperintense signals decreased in 6–24 months in the remaining 4 patients.
Discussion
HCNH was first described by Bedwell in 1960 [5] and is characterized by a typical triad, including unilateral involuntary movements, contralateral striatal abnormalities, and rapid remission of symptoms after correction of hyperglycemia. This transient movement disorder is more frequently seen in elderly women, especially Asians, and has been reported as a rare disease over the past 2 decades. Although the pathogenesis of HCNH is still unclear, the strong association between hyperglycemia and hemichorea has been supported in almost all of the cases. Recent studies involving HCNH suggest that it might be a rare complication or the initial manifestation [6] of diabetes mellitus. In our study, whether or not a diabetic history existed, all 12 patients had nonketotic hyperglyGuo/Miao/Ji/Li/Liu/Sun
a
b
c
d
f
g
h
i
e
Fig. 4. Case 4: an 82-year-old woman was admitted to the emergency department with a complaint of involuntary movements on the right side of her body for 4 days. a CT on admission revealed hyperattenuation in the putamen and caudate nucleus. MR imaging showed hyperintensity on the T1-weighted images (b) and isointensity on the T2-weighted images (c). d, f The lesion is shown
in the sagittal and coronal sections. One month after antidiabetic and haloperidol treatment, the hyperattenuation in CT was reduced (f). She underwent follow-up MR imaging (g–i) after 18 months, which revealed that the hyperintensity had resolved, and hypointensity was noted in the primary lesion.
cemia on admission and were diagnosed with diabetes mellitus at the time of hospital discharge. With prompt treatment, all of the patients achieved relief from choreic symptoms, and follow-up neuroimaging showed reversible disappearance of the symptoms in some of the patients. The neuroradiologic findings of HCNH are specific and involve the putamen in all cases, head of the caudate nucleus in most cases, and the globus pallidus in a minority of cases. Lesions are unilateral in the vast majority of patients, though sometimes bilateral involvement has been reported, even in patients who have not yet manifested symptoms, which is consistent with a previous report [3]. To detect the typical lesion of HCNH, MR imaging is more sensitive and superior to CT, revealing reversible hyperintensity in the striatum without signs of a mass effect, edema, volume loss, or internal capsule involvement. Because the lesions correspond to the territory of the lenticulostriate artery, researchers are of the opinion that the striatal lesions may be the result of a vascular insult. Patients with diabetes mellitus and poorly controlled hyperglycemia are at increased risk for cerebrovascular ischemia. However, vascular insufficiency alone does not explain the mechanisms responsible for some bilateral lesions. In addition, calcifications, petechial hemorrhage, myelinolysis, wallerian degeneration, and deposition of manganese have been suggested as possible causes of the
reversible hyperintensity in T1-weighted images [7–12]; however, biopsy and autopsy studies only show selective neuronal loss, gliosis, and reactive astrocytosis [13–15]. Similarly, MR spectroscopy has shown decreased N-acetylaspartate and creatine peaks with a low N-acetylaspartate/creatine ratio, suggesting neuronal dysfunction, loss, or damage, while a high choline/creatine ratio likely indicates gliosis [16]. It has also been postulated that the typical T1-weighted MR images may be due to regional ischemic injury related to hyperglycemia-induced hyperviscosity. Diffusion-weighted imaging and apparent diffusion coefficient maps may show restricted diffusion, which has been ascribed to hyperviscosity [17]. SPECT studies have demonstrated a significant decrease in blood flow in the basal ganglia contralateral to the chorea [18]. 18 F-fluorodeoxyglucose positron emission tomography has shown a reduction in glucose metabolism in the corresponding lesions in T1-weighted MR images in animals [19]. All of the above image changes indicate that the metabolic derangements associated with hyperglycemia play a significant role in the pathogenesis, rather than vascular disease. In the current study, normal susceptibilityweighted imaging excludes the possibility of hemorrhage, thus we consider that metabolic disorders associated with hyperglycemia resulting in hemichorea and vascular insufficiency may contribute to this regional metabolic failure.
Hemichorea Associated with Nonketotic Hyperglycemia
Eur Neurol 2014;71:299–304 DOI: 10.1159/000357210
303
The prognosis of HCNH has been reported to be favorable [3] if there is prompt glycemic control, which is similar to our results. Patients with a slight injury in their brains resulting from hyperglycemia can get a complete remission with simple antidiabetic treatment, although additional neuroleptics are needed for most patients. However, a minority of patients have persistent chorea for years or longer [20], suggesting severe damage in the basal ganglia. In our study, all of the patients experienced symptom relief after early treatment, most of whom (7 of 12 patients) did not have recurrences at the follow-up visit. Four patients received antidiabetic monotherapy and 3 patients with typical abnormal images on admission had recurrence within 2 weeks after hospital discharge. In contrast, the patient without image changes had complete remission at the 1-year follow-up. We therefore speculate that images may predict the degree of injury. For patients
with hyperdense or hyperintense signals within the basal ganglia on CT or MR images, additional symptomatic treatment may be needed. In conclusion, HCNH is a reversible, benign disorder. Radiologic changes may be a predictor of the degree of injury to the brain. Hemichoreic symptoms can be completely or partly ameliorated and hyperintensity in the striatum on T1-weighted images may resolve after euglycemia is restored. Recognition of the unique clinicoradiologic manifestations is important because early correction of the underlying hyperglycemia will contribute to rapid improvement. Although the pathogenesis of HCNH remains uncertain, current evidence suggests that metabolic changes play a main role. Further autopsy data and clinical research based on a sufficient sample size are needed to verify the pathogenesis and nature of neuroradiologic changes.
References 1 Higa M, Kaneko Y, Inokuchi T: Two cases of hyperglycaemic chorea in diabetic patients. Diabet Med 2004;21:196–198. 2 Cheema H, Federman D, Kam A: Hemichorea-hemiballismus in non-ketotic hyperglycaemia. J Clin Neurosci 2011;18:294–196. 3 Oh SH, Lee KY, Im JH, lee MS: Chorea associated with nonketotic hyperglycemia and hyperintensity basal ganglia lesion on T1weighted brain MRI study: a meta-analysis of 53 cases including four present cases. J Neurol Sci 2002;200:57–62. 4 Cervantes-Arriaga A, Arrambide G, Rodríguez-Violante M: A prospective series of patients with hyperglycaemia-associated movement disorders. J Clin Neurosci 2011; 18: 1329–1332. 5 Bedwell SF: Some observations on hemiballism. Neurology 1960;10:619–622. 6 Qi X, Yan Y-Y, Gao Y, Zheng Z-S, Chang Y: Hemichorea associated with non-ketotic hyperglycaemia: a case report. Diabetes Res Clin Pract 2012;95:e1–e3. 7 Sanfield JA, Finkel J, Lewis S, Rosen SG: Alternating choreoathetosis associated with uncontrolled diabetes mellitus and basal ganglia calcification. Diabetes Care 1986;9:100–101. 8 Chang MH, Chiang HT, Lai PH, Sy CG, Lee SS, Lo YY: Putaminal petechial haemorrhage as the cause of chorea: a neuroimaging study. J Neurol Neurosurg Psychiatry 1997; 63: 300– 303.
304
9 Nagai C, Kato T, Katagiri T, Sasaki H: Hyperintense putamen in T1-weighted MR images in a case of chorea with hyperglycemia. AJNR Am J Neuroradiol 1995;16:1243–1246. 10 Sajth J, Ditchfield A, Katifi HA: Extrapontine myelinolysis presenting as acute parkinsonism. BMC Neurol 2006;6:33. 11 Wintermark M, Fischbein NJ, Mukherjee P, Yuh EL, Dillon WP: Unilateral putaminal CT, MR, and diffusion abnormalities secondary to nonketotic hyperglycemia in the setting of acute neurologic symptoms mimicking stroke. AJNR Am J Neuroradiol 2004;25:975– 976. 12 Fujioka M, Taoka T, Matsuo Y, Mishima K, Oqoshi K, Kondo Y, Tsuda M, Asano T, Sakaki T, Miyasaki A, Park D, Siesjö BK: Magnetic resonance imaging shows delayed ischemic striatal neurodegeneration. Ann Neurol 2003; 54:732–747. 13 Shan DH, Ho DM, Chang C, Teng MM: Hemichorea-hemiballism: an explanation for MR signal changes. AJNR Am J Neuroradiol 1998;19:863–870. 14 Ohara S, Nakagawa S, Tabata K: Hemiballism with hyperglycemia and atriatal T1-MRI hyperintensity: an autopsy report. Mov Disord 2001;16:521–525.
Eur Neurol 2014;71:299–304 DOI: 10.1159/000357210
15 Lai PH, Tien RD, Chang MH, Teng MM, Yang CF, Pan HB, Chen C, Liring C, Ling JF, Kong KW: Chorea-ballismus with nonketotic hyperglycemia in primary diabetes mellitus. AJNR Am J Neuroradiol 1996;17:1057–1064. 16 Lai PH, Chen PC, Chang MH, Pan HB, Yang CF, Wu MT, Li JY, Chen C, Liang HL, Chen WL: In vivo proton MR spectroscopy of chorea-ballismus in diabetes mellitus. Neuroradiology 2001;43:525–531. 17 Chu K, Kang DW, Kin DE, Park SH, Roh JK: Diffusion-weighted and gradient echo magnetic resonance findings of hemichorea hemiballismus associated with diabetic hyperglycemia: a hyperviscosity syndrome? Arch Neurol 2002;59:448–452. 18 Chang MH, Li JY, Lee SR, Men CY: Non-ketotic hyperglycaemic chorea: a SPECT study. J Neurosurg Psychiatry 1996;60:428–430. 19 Hsu JL, Wang HC, Hsu WC: Hyperglycemiainduced unilateral basal ganglia lesions with or without hemichorea: a PET study. J Neurol 2004;251:1486–1490. 20 Ahlskog JE, Nishino H, Evidente VGH, Tulloch JW, Forbes GS, Caviness JN, GwinnHardy KA: Persistent chorea triggered by hyperglycemic crisis in diabetics. Mov Disord 2001;16:890–898.
Guo/Miao/Ji/Li/Liu/Sun
Copyright: S. Karger AG, Basel 2014. Reproduced with the permission of S. Karger AG, Basel. Further reproduction or distribution (electronic or otherwise) is prohibited without permission from the copyright holder.
