Reviewer: Thashi T Chang
Major Compulsory Revisions
HE is a diagnosis of exclusion. The authors do not indicate whether the other well
recognised autoimmune encephalopathies have been excluded by testing for NMDAR,
VGKC, AMPAR, GABABR, GAD, Gly-R antibodies in their patients (see review by
Angela Vincent and others in Lancet Neurol 2011;10:759-72), whether PCR for viral
antigens including HSV was done on CSF to exclude viral encephalitides and whether
demyelinating diseases such as ADEM were excluded. The misclassification of
patients as HE and their subsequent clinical characterization will further distort the
nosology of HE. The authors claim that not all HE patients present with
encephalopathy but this too could be due to the misclassification of disease given that
anti-thyroid antibodies do occur in the normal population. Furthermore, in some
patients it appears that an elevated anti-thyroid antibody titre has been the only
criterion for diagnosis of HE (eg. patient 2).
A: Thanks for your valuable comments! Indeed, HE is a diagnosis of exclusion.
Therefore, all 15 patients have been through careful differential diagnosis from
alternative infectious, toxic, metabolic, vascular or neoplastic etiology related to
the neurological symptoms. For viral encephalitis, there was no fever or
pleocytosis in all patients. IgG and IgM antibodies against toxoplasma, HSV,
CMV, RV were tested in both CSF and serum. For each patient, viral
encephalitis and ADEM were among the most important disorders in differential
diagnosis, and clinical manifestations and investigations didn’t support the
diagnosis of viral encephalitis and ADEM. More details of patients No.1, 2, 4, 14,
15 were described as following.
However, as the reviewer stated, this study suffers from the limitations of a
retrospective analysis. In recent years, the field of autoimmune encephalopathy
has expanded rapidly. Studies discovered that some neurological disorders
including limbic encephalitis, Faciobrachial dystonic seizures, Morvan’s
syndrome, neuromyotonia etc. are associated with autoantibodies directed
against neuronal proteins, including VGKC-antibodies, AMPAR-antibodies,
GABABR-antibodies, GAD-antibodies etc.[1-3] Presence of these antibodies
highly prompted the associated neurological disorders. However, since most of
the association between these autoantibodies and corresponding neurological
disorders was established in recent years, the tests of these antibodies were not
available here in our hospital when these patients were admitted between 2005
and 2009. It is a defect of this retrospective study that these patients did not
undertake the tests of these antibodies. However, all patients were carefully
differentiated from other autoimmune encephalopathies. It is now recognized
that the autoimmune encephalopathies associated with these antibodies have
their own characteristic manifestations which were not present in our patients.
We discussed them in detail as following:
(1) NMDAR antibodies: The NMDAR antibody-associated encephalitis is a
severe encephalopathy. The clinical presentation can be very characteristic and
divided into three stages, including diverse infections in a prodromal phase;
psychosis, confusion, amnesia, and psychiatrists in an early stage; and then
within 1-2 weeks, a later stage characterized by movement disorder, autonomic
instability, hypoventilation and often reduced consciousness[3]. Recovery of this
kind of encephalopathy is often slow, even with immunotherapy[3]. The majority
of our patients had suffered various encephalopathic symptoms more than one
month on admission, only patient NO.6 and patient NO.7 had a short duration,
who did not have an infection before onset and had complete remission after
immunotherapy.
(2) VGKC antibodies: For limbic encephalitis associated with VGKC antibodies,
almost all the cases were found to be over 50 years of age, more commonly male
than female (2:1). A serum hyponatremia and MRI T2/FLAIR high signal in the
medial temporal lobe were characteristic features, present in around 60% of
patients[1]. Before the manifestations of limbic encephalitis, some patients have a
specific seizure, called faciobrachial dystonic seizure (FBDS), especially for
patients with cognitive impairments [1, 4]. It was reported recently that 77% of
these patients developed FBDS prior to the onset of amnesia or other cognitive
changes [4]. The seizures caused by VGKC antibodies often show a poor
response to antiepileptic drugs[3]. In our study, no patient had hyponatremia or
FBDS, though 73.3% of our patients had cognitive impairments. Only three
patients had MRI abnormal signals in temporal lobe (Patient NO. 4, 9 and 10),
and they were all relatively young for VGKC antibodies associated
encephalopathy. A total of four patients in our study had seizures; two of them
had no further seizures through antiepileptic drugs, one had no response to
immunotherapy, and only one patient, a 43-year-old female, with normal MRI,
had complete remission after immunotherapy.
(3) AMPAR antibodies and GABABR antibodies: Most of the patients with
AMPAR or GABABR antibodies had tumors [1, 3]. In this study, all patients
underwent multiple investigations for tumor, including serum tumor markers,
serum and CSF anti-Hu, anti-Ri, anti-Yo, anti-Ma2, anti-amphiphysin, anti-CV2
antibodies, and CSF and urine cytologic examination. No positive findings were
reported. Chest CT and abdominal ultrasound also revealed no positive findings.
(4) GAD antibodies: Patients with limbic encephalitis and epilepsy associated
with GAD antibodies are often young adult women[3]. All patients with limbic
encephalitis associated with GAD antibodies had temporal lobe seizures, and
even presented with seizures only[3, 5]. The seizures associated with GAD
antibodies could not be controlled with anticonvulsive treatment [5]. These
disorders are associated with the high signal on T2 MRI in the medial temporal
lobes[3]. In our 4 patients with seizures, only one patient had abnormal signal in
the temporal lobe, who had no further seizure attacks just with single
antiepileptic drug.
(5) Gly-R antibodies: The CNS disorder associated with GlyR antibodies, as part
of the stiff-person syndrome spectrum, is progressive encephalomyelitis with
rigidity and myoclonus[3]. The clinical features include stiffness and rigidity,
excessive startle in response to various stimuli, and brainstem involvement with
oculomotor dysfunction[3]. These clinical manifestations were not present in our
patients.
More details of patients No.1, 2, 4, 14, 15 were described as following.
Patient NO.1 was admitted with the complaints of memory impairment,
psychiatric symptoms and insomnia. She had a history of hypothyroidism and
takes levothyroxine continuously. She was euthyroid on admission, however, her
serum levels of TPO antibody and TG antibody were significantly elevated (TPO
antibody >1300U/ml and TG antibody >500U/ml). After systematic examinations,
we confirmed that she had cognitive deficits (bachelor degree, MMSE=21). MRI
showed nonspecific abnormal signals in periventricular white matter. CSF tests
showed the protein level was elevated (55 mg/dL, normal reference range: 15-45
mg/dL), and the IgG synthesis rate was also elevated (4.4740/24hr, reference
range -9.900-3.300/24hr). Based on these MRI and CSF positive findings, we
diagnosed this patient as having HE.
Patient NO.2 This 32-year-old female was admitted to our department because of
generalized tonic-clonic seizures for 6 months. She had a history of goiter. The
emission computed tomography of thyroid showed bilateral thyroid enlargement,
multiple hot and cool nodules, and abnormal tracer uptake. Her antithyroid
antibodies levels were extraordinarily high (TPO antibody >1300U/ml and TG
antibody >500U/ml). The diagnosis of Hashimoto’s thyroiditis was established by
the endocrinologist. Her CSF protein level was elevated (70 mg/dL, normal
reference range: 15-45 mg/dL). No alternative infectious, toxic, metabolic,
vascular or neoplastic evidences were found. Combining the history of
Hashimoto’s thyroiditis, remarkably elevated TPO antibody, encephalopathic
manifestation (seizure), elevated CSF protein, the seizures of this patient was
considered symptomatic and the diagnosis of HE was made.
Patient No. 4 This 27-year-old male was admitted because of memory loss
impairment for 3 weeks (MMSE 25). No headache, fever or seizures. Brain MRI
showed abnormalities in splenium corporis callosi and bilaterial hippocampus.
VEEG was normal. There was no pleocytosis and CSF protein was elevated
(97mg/dl, normal reference range: 15-45 mg/dL). TPO-Ab was remarkably
elevated (>1300U/ml) and TG-Ab was normal (38.8U/ml). The serum sodium
level was normal. Investigations for alternative infectious, toxic, metabolic,
vascular or neoplastic etiology revealed no positive findings. Considering his
mild symptoms, steroid therapy was not applied. In 3-month follow-up, his
memory impairment recovered (MMSE 30).
For patient No. 14, symptoms began in mid-adult years, with subacute onset of
dysarthria and hemiparesis. There were no prodromal infections, fever or
vaccination. The onset of the symptoms was subacute rather than acute. No
impairment of consciousness. There was no pleocytosis or increased protein in
CSF. No oligoband (OB) and IgG index was normal. Brain MRI showed
abnormal signals in bilateral basal ganglia (Fig. 1 CD), without white matter
lesions. The clinical manifestations, CSF and MRI investigations didn’t support
the diagnosis of ADEM. Careful blood, urine and CSF investigations also
excluded other infectious, toxic, metabolic, vascular or neoplastic etiology related
to the neurological symptoms. The patient has a history of Hashimoto thyroiditis
and high titers of TPO-Ab (>1300 U/ml) and TG-Ab (219.0 U/ml). Her symptoms
improved and the TPO-Ab and TG-Ab titers decreased after steroid therapy.
Based on these manifestations, the diagnosis of HE was made.
Patient No. 15 has been previously reported[6]. In this case, symptoms began in
mid-adult years, with gradual progression of gait and limb ataxia, scanning
dysarthria, focal dystonia, lingual fasciculations and mild weakness, suggesting
the diagnosis of spinocerebellar ataxia. Since his symptoms accelerated and he
became wheelchair-bound in 2 weeks before admission, other causes must be
considered. Infectious, autoimmune, toxic, biochemical, haematological, and
paraneoplastic analyses revealed no findings. With the significantly elevated
TPO-Ab, response to steroid therapy, fluctuating symptoms, the diagnosis of HE
was made.
Indeed, as the reviewer commented, limited by the retrospective study design,
our patients did not undertake the tests of these antibodies associated with
autoimmune encephalopathy. The possibility of coincident two pathologies could
not be excluded. However, with careful differential, our patients did not conform
to the common characteristics of the encephalopathies associated with these
antibodies. According to Occam’s razor, after exclusion of infectious,
autoimmune, toxic, biochemical, haematological, and paraneoplastic etiologies,
Hashimoto’s encephalopathy could be responsible for all the clinical features.
Minor essential revisions
The authors state the ‘upper limits’ for thyroid antibodies in their institution but these
differ from the reference ranges that are subsequently stated. This is confusing and
need clarification.
A: Thanks for your comments! The levels of thyroid antibodies were measured
using chemiluminescence immunoassay method in the department of clinical
laboratory of our hospital. The ‘upper limits’ means the highest detectable levels
of thyroid antibodies by this method. The levels upper than this limit are
reported as >1300U/ml for TPO antibody and >500U/ml for TG antibody. The
‘reference ranges’ means the normal ranges. We have clarified this in the
manuscript.
The EEG and MRI changes must be related to the patients’ clinical profiles in the
table.
A: Thanks for your suggestion! We have added the description of EEG and MRI
changes in Table 1.
The hippocampal signal changes in the MRI Figure 1 (AB) are better demonstated in
coronal cuts through the hippocampus. The authors must explain how this is different
from VGKC antibody-associated limbic encephalitis.
A: Thanks for your suggestions! We agree with the reviewer that the
hippocampal signal changes are better demonstrated in coronal cuts, which was
not available in this patient. We will be careful with this in future practice.
Figure 1 (AB) corresponds to patient No. 4. Details of this patient and the
characteristics of VGKC antibodies-associated limbic encephalitis were
described in the ‘Major Compulsory Revisions’ section. This patient was just 27year-old, relatively young for VGKC antibodies-associated limbic encephalitis.
He had no hyponatremia or FBDS, though memory impairment was his only
symptom. Limited by the retrospective study design, this patient did not
undertake the test of VGKC antibodies. However, the common characteristics of
VGKC antibodies-associated limbic encephalitis were not present in this patient.
The authors must explain why the MRI appearance of figure 1 (CD) is not ADEM.
A: Thanks for your comments! For patient No. 14, symptoms began in mid-adult
years, with subacute onset of dysarthria and hemiparesis. There were no
prodromal infections, fever or vaccination. The onset of the symptoms was
subacute rather than acute. No impairment of consciousness. There was no
pleocytosis or increased protein in CSF. No oligoband (OB) and IgG index was
normal. Brain MRI showed abnormal signals in bilateral basal ganglia (Fig. 1
CD), without white matter lesions. The clinical manifestations, CSF and MRI
investigations didn’t support the diagnosis of ADEM. Careful blood, urine and
CSF investigations also excluded other infectious, toxic, metabolic, vascular or
neoplastic etiology related to the neurological symptoms. The patient has a
history of Hashimoto thyroiditis and extremely high titers of TPO-Ab (>1300
U/ml) and TG-Ab (219.0 U/ml). Her symptoms improved and the TPO-Ab and
TG-Ab titers decreased after steroid therapy. Based on these manifestations, the
diagnosis of HE was made.
The authors state ‘complete’, ‘partial’ recovery and ‘relapses’ during recovery
without stating objective assessments such as MMSE. This needs to be substantiated.
A: Thanks for your suggestion! We have added the detailed assessments of our
patients at follow-up in Table 1, including the scores of cognitive tests.
The authors must explain how they excluded limbic encephalitis in patients with
memory disorders.
A: Thanks! Limbic encephalitis was mainly regarded as paraneoplastic
previously. Thus, all of our patients underwent multiple investigations for tumor,
including serum tumor markers, serum and CSF anti-Hu, anti-Ri, anti-Yo, antiMa2, anti-amphiphysin, anti-CV2 antibodies, and CSF and urine cytologic
examination. No positive findings were reported. Chest CT and abdominal
ultrasound also revealed no positive findings. Until recently, multiple antibodies
were found to be associated with nonparaneoplastic limbic encephalitis. The
differential with these antibodies associated limbic encephalitis was explained in
the ‘Major Compulsory Revisions’ section.
The authors must explain the rationale for treating patient 1 with Levothyroxine given
that the patient was euthyroid, the rationale for nerve growth factor in patient 4 and 11,
and the basis of the dexamethasone dose in patient 12.
A: Thanks for your comments! Patient NO.1 was admitted with the complaints
of memory impairment, psychiatric symptoms and insomnia. She had a history
of hypothyroidism and takes levothyroxine continuously. Although she was
euthyroid on admission, the treatment of Levothyroxine was continued. For
Patient 4 and 11, nerve growth factor is approved for the treatment of nerve
injury and encephalopathy in China. Therefore, it was used in patient 4 and 11
as adjunctive treatment. For patient 12, since there is no acknowledged guideline
for the use of corticosteroids in neurological disorders currently, the selection
and initial dose of corticosteroids are mostly based on personal experience.
According to the chronic onset and slow progression of symptoms, initial dose of
10mg of dexamethasone was used in this patient.
References
1.
2.
3.
4.
5.
6.
Irani SR, Vincent A: Autoimmune encephalitis -- new awareness,
challenging questions. Discov Med 2011, 11(60):449-458.
Granerod J, Ambrose HE, Davies NW, Clewley JP, Walsh AL, Morgan D,
Cunningham R, Zuckerman M, Mutton KJ, Solomon T et al: Causes of
encephalitis and differences in their clinical presentations in
England: a multicentre, population-based prospective study. Lancet
Infect Dis 2010, 10(12):835-844.
Vincent A, Bien CG, Irani SR, Waters P: Autoantibodies associated with
diseases of the CNS: new developments and future challenges.
Lancet Neurol 2011, 10(8):759-772.
Irani SR, Michell AW, Lang B, Pettingill P, Waters P, Johnson MR, Schott JM,
Armstrong RJ, A SZ, Bleasel A et al: Faciobrachial dystonic seizures
precede Lgi1 antibody limbic encephalitis. Ann Neurol, 69(5):892-900.
Malter MP, Helmstaedter C, Urbach H, Vincent A, Bien CG: Antibodies to
glutamic acid decarboxylase define a form of limbic encephalitis.
Annals of neurology 2010, 67(4):470-478.
Tang Y, Chu C, Lin MT, Wei G, Zhang X, Da Y, Huang H, Jia J: Hashimoto's
encephalopathy mimicking spinocerebellar ataxia. J Neurol 2011,
258(9):1705-1707.