Biotin-Responsive Basal Ganglia Disease: Neuroimaging Features before and after Treatment

Discussion

BBGD is an autosomal recessive neurometabolic disorder.⁷ It was first described in 1998 in 10 patients from the Middle East.¹ Recently, the disease was recognized as a panethnic condition that usually occurs in preschool and school-aged children.^2,4

The typical clinical presentation of patients with BBGD is recurrent subacute episodes of encephalopathy, often triggered by febrile illness or mild trauma and characterized by confusion, seizures, dystonia, external ophthalmoplegia, and dysphagia, eventually leading to coma and even death. Less frequently, patients with BBGD present with a chronic or slowly progressive condition characterized by dystonia, seizures, and psychomotor delay.¹³

The neurologic history of most our patients (86.7%) was in the form of subacute encephalopathy similar to the first 10 patients described by Ozand et al.¹ However, 2 patients (13.3%) presented with acute crisis with severe neurologic deficit, generalized dystonia, quadriplegia, and coma. A pre-existing trigger factor (febrile illness or trauma) was found in 46.7% of our cases.

Early diagnosis of BBGD is crucial because early administration of biotin and thiamine results in partial or complete improvement within days. Delay or lack of treatment may result in death or severe neurologic sequelae, including dystonia and quadriparesis. Moreover, patients require counseling regarding life-long treatment.¹³

In the original report by Ozand et al,¹ the MR imaging findings consisted of bilateral necrosis in the basal ganglia, particularly at the central part of the caudate heads and part or all of the putamen with severe edema during the acute/subacute crisis in addition to white matter involvement at the gray–white matter junction. Subsequent reports confirmed these findings and added the involvement of the thalami, cerebellum, and brain stem.^2,5 Our results supported the findings of these previous reports regarding the MR imaging distribution of lesions in all cases. In addition, the globus pallidi were spared in all patients in agreement with most of previous reports.^1,2,13

Alfadhel et al⁷emphasized diffuse involvement of the cortical and subcortical white matter in the acute phase and atrophy and necrosis of the basal ganglia in the chronic phase in all cases. The thalami and brain stem were involved in one-third of their patients. In the present study, we demonstrated discrete abnormal signal changes in the mesencephalon, cerebral cortical-subcortical regions, and in the medial dorsal nuclei of the thalami in 80% of cases. In 53%, when the disease was advanced, patchy deep white matter affection was found. The cerebellum along the cerebellar cortex and vermis was involved in only 13.3% of our cases compared with one-third of the patients of Alfadhel et al.⁷

Follow-up of our patients under treatment confirmed the evolution of the basal ganglia from swelling into atrophy and necrosis with resolution of the abnormal high signals of the cerebral cortex and subcortical areas together with regression of the cerebellar abnormal signal in patients who showed cerebellar involvement. However, the abnormal signal lesions of the caudate and putamen persisted. These findings are in absolute agreement with earlier reports.^{1⇓⇓–4,7,13}

Despite the extensive abnormal findings observed on the brain MR imaging, 80% of our patients with biotin-responsive encephalopathy remained asymptomatic under treatment. This observation suggests that enough tissue may be preserved in the target areas of the disease, to permit neurologic function.¹

Interestingly, in the report by Alfadhel et al,⁷1 patient demonstrated spinal cord involvement with increased T2 signal intensity particularly at the cervical region, which was not reported on earlier studies. They thought that the spinal cord involvement could be part of the disease and advised performing spinal MR imaging for all patients with BBGD. However, 7 of our patients were investigated by spinal MR imaging and none showed spinal cord involvement, which strongly argues against spinal cord involvement as a part of the disease. The absence of spinal cord lesions in our patients is consistent with all previously described cases, except the 1 reported by Alfadhel et al.⁷

Spectroscopy (¹H-MR) was performed in 6 of our cases at time of onset of symptoms and before initiation of therapy. It depicted lactate peaks and decreased NAA in all cases. The lactate doublet could be interpreted as sign of ongoing apoptosis. It has disappeared in the single posttherapy MR spectroscopy. Decreased NAA could be explained by early brain injury and neuronal loss.¹³ We found no appreciable differences between the MR spectrum in basal ganglia and in parieto-occipital cortical-subcortical lesions. As increased CNS lactate can be found in other pathologic processes including mitochondrial diseases,¹⁴ these findings are considered nonspecific.¹³

In our study, because assessment of the response to therapy was limited clinically to the reversibility of symptoms and radiologically to the regression of the swelling and vasogenic edema of the affected regions, none of our patients had follow-up MR spectroscopy except 1. Interestingly, in this patient (Fig 4), the lactate peak has almost disappeared reflecting that MR spectroscopy could provide additional information as a monitor of response to therapy. The short half-time of lactate could be valuable as an early indicator of response to treatment in the future.

Bilateral and symmetrical basal ganglia lesions are caused by different systemic or metabolic conditions that share some findings with BBGD. However, the dramatic response to biotin therapy and persistence of caudate and putaminal signal abnormality without neurologic sequelae are characteristic features of BBGD. Wernicke encephalopathy (WE) is an example of these diseases. It can involve the medial dorsal nucleus in thalami, periventricular regions of the third ventricle, brain stem, central gray matter, basal ganglia, and cerebellum similar to BBGD. However, the main MR imaging differences between the two are that in BBGD, the mammillary bodies are spared when usually involved in WE and the supra- and infratentorial cortical involvement in BBGD is more extensive compared with WE.¹⁵

Mitochondrial disorders can also mimic BBGD because of similarity in deep gray matter involvement. They include Leigh disease, MELAS (mitochondrial myopathy, encephalopathy, lactic acidosis, and strokelike episodes), and MERRF (myoclonic epilepsy associated with ragged red fibers). Like BBGD, Leigh disease typically causes symmetric putaminal involvement, which may be associated with abnormalities of caudate nuclei, thalami, and brain stem; however, it less frequently involves the cerebral cortex and rarely involves the white matter.¹⁶ In MELAS, the multiple cortical and subcortical infarctlike lesions that cross vascular boundaries are a distinctive feature.¹⁶ In MERRF, the imaging findings are nonspecific. The cerebral and cerebellar white matter is reported to show patchy T2 prolongation on MR imaging. Cerebral and cerebellar atrophy are almost always present. Involvement of the deep gray matter nuclei with degeneration and calcification of dentate and globus pallidus are the most common manifestations.¹⁷ On the contrary, in BBGD the globus pallidus is usually not affected.^1,2,13,18 This also differentiates the disease from toxic causes of encephalopathy such as carbon monoxide poisoning, which has propensity to affect the globus pallidus.¹⁹

Although the disease described responds to biotin, biotinidase deficiency is absent. The brain imaging findings are also different from those seen in biotinidase deficiency, in which the neuroradiologic changes indicate diffuse low attenuation of the white matter followed by progressive marked cerebral atrophy.²⁰

In contrast to the swelling of the caudate heads and putamen in patients with BBGD, the main radiologic feature in juvenile Huntington disease is caudate atrophy and increased proton attenuation and T2 signal in the atrophic caudate nuclei and putamina.²¹ In patients with acute destructive lesions of the basal ganglia, ie, the so-called benign form of infantile bilateral striatal necrosis, the basal ganglia lesions might return to normal²² while they persist after treatment in BBGD.⁴

Generally, BBGD should be thought of in any pediatric patient with neurologic symptoms and bilateral and symmetrical affection of caudate nuclei and putamina in MR imaging. Biotin and thiamine treatment is then suggested until either BBGD is excluded or 1 of the other differential diseases has been clearly identified.²³

In our study, 12 cases (80%) had a good (favorable) outcome with no clinical sequelae, whereas 3 patients had an unfavorable outcome, 2 (13.3%) had dystonia and dysarthria, and 1 patient (6.7%) died from severe encephalopathy because of delayed diagnosis and treatment. These data suggested better outcome results when compared with Alfadhel et al⁷ who had 4 of 18 (22%) patients who died, 6 (33.3%) patients who showed mild to moderate neurologic deficit, and 2 (11%) cases in which the patients had severe neurologic deficits.

The gap between the age of disease onset and the date of starting treatment by biotin and thiamine correlates directly with the neurologic outcome. Two of our 15 patients who had delayed diagnosis displayed neurologic deficits and another patient died for the same reason; the remaining patients who were diagnosed early and received immediate treatment achieved good outcomes and became asymptomatic.

It is important to check for the presence of BBGD in children with acute onset extrapyramidal symptoms as it can be managed without further neurologic deterioration.³ Because there is no known biochemical marker for the disease, the only way to confirm the diagnosis is the detection of a mutation in the SLC19A3 gene.⁶