Serum and CSF GQ1b antibodies in

isolated ophthalmologic syndromes

Marianna Spatola, MD
Renaud Du Pasquier, MD
Myriam Schluep, MD
Axel Regeniter, MD

Correspondence to
Dr. Spatola:
spatola@clinic.ub.es

ABSTRACT

Objective: To establish the sensitivity and specificity of serum and CSF antibodies targeting the
gangliosides GQ1b (GQ1bAb) in isolated ophthalmologic syndromes, such as acute ophthalmoplegia (AO) and optic neuritis (ON), caused by disorders other than Miller-Fisher syndrome (MFS).

Methods: We measured serum and CSF GQ1bAb in patients with MFS and with AO or ON caused
by other disorders than MFS.
Results: Twenty-one patients with AO (21 serum, 9 CSF), 13 with ON (13 serum, 13 CSF), and 12
with MFS (12 serum, 10 CSF) were included in the study. There were no significant differences in
age, sex, and CSF findings between the AO and MFS groups. Elevated serum GQ1b titers occurred
in 11 of 12 patients with MFS but in only 1 of the 34 patients without MFS. Sensitivity was 92%
(95% confidence interval [CI] 62%100%) and specificity 97% (95% CI 85%100%). In CSF,
GQ1bAb were identified in 2 of 10 patients with MFS but in none with other disorders. Sensitivity
was 20% (95% CI 2%56%) and specificity 100% (95% CI 85%100%).

Conclusions: Increased serum GQ1bAb are highly specific for MFS. Measurement of GQ1bAb in
CSF does not improve diagnosis.

Classification of evidence: This study provides Class III evidence that serum GQ1bAb accurately
distinguish MFS from other disorders (sensitivity 92%, 95% CI 62%100%; specificity 97%,
95% CI 85%100%). Neurology 2016;86:17801784
GLOSSARY
AO 5 acute ophthalmoplegia; CI 5 confidence interval; GQ1bAb 5 GQ1b antibodies; Ig 5 immunoglobulin; LH 5 Lausanne
University Hospital; MFS 5 Miller-Fisher syndrome; ON 5 optic neuritis.

Anti-ganglioside GQ1b immunoglobulin G (IgG) antibodies (GQ1bAb) are sensitive markers

of Miller-Fisher syndrome (MFS).1 Ocular manifestations of MFS include ophthalmoplegia,
pupillary dysfunction, and, infrequently, optic neuritis (ON).2,3 Although GQ1bAb seem to
correlate with the severity of ocular motor dysfunction,4 the role of these antibodies in isolated
acute ophthalmoplegia (AO) or other pathologies affecting the eyes, such as ON, is still unclear.
MFS is caused by an autoimmune attack to the peripheral nerve, through a mechanism that is not
completely understood. Either GQ1bAb are directly pathogenic or they act as biomarkers of an
immune-mediated disruption of the nerve sheath integrity, which results in the exposition of antigens
normally hidden to the immune surveillance and consequently in the production of autoantibodies.
GQ1b gangliosides are abundant in ocular motor and optic nerves.5,6 We thus studied the
titers of GQ1bAb in disorders other than MFS associated with AO and ON to further investigate the specificity of serum and CSF GQ1bAb.
METHODS Patients. We identified 3 groups of adult patients: (1) those with AO, (2) those with ON, and (3) those with MFS. Patients
with AO were identified prospectively, from June 2011 to June 2012, at the Lausanne University Hospital (LH). AO was defined as palsy of
the ocular motor muscles developing within 7 days. Patients with ON (visual acuity #0.5) were retrospectively identified from the LH
Inflammatory Bio Bank, which comprises samples from patients who required blood and CSF studies for any neurologic condition. Patients
with MFS were identified retrospectively from the LH Inflammatory Bio Bank. Definite diagnosis was based on these criteria7: (1) self-limited
From the University of Lausanne (UNIL) (M. Spatola), Lausanne, Switzerland; Institut dInvestigacions Biomdiques August Pi i Sunyer
(IDIBAPS) and Hospital Clinic (M. Spatola), Neuroimmunology Program, University of Barcelona, Barcelona, Spain; University Hospital of
Lausanne (CHUV) (R.Du Pasquier, M. Schluep), Department of Clinical Neuroscience, Service of Neurology, Lausanne, Switzerland; and
University Hospital of Basel (A.Regeniter), Department of Laboratory Medicine, Basel, Switzerland.
Go to Neurology.org for full disclosures. Funding information and disclosures deemed relevant by the authors, if any, are provided at the end of the article.
1780

2016 American Academy of Neurology

2016 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

symmetric ophthalmoplegia and ataxia developing within 4 weeks,

(2) decreased or absent deep tendon reflexes, (3) absence of limb
weakness, and (4) exclusion of other etiologies. If ataxia was absent,
a probable diagnosis was considered if at least one other cranial nerve
was involved.
The 3 groups were defined solely by clinical criteria.
GQ1bAb results were never used to classify patients as MFS or
non-MFS (AO and ON).

Table 1

Comparison between ELISAs for serum

GQ1bAb measurements in 21 patients
with acute ophthalmoplegiaa
ELISA

Patient

IgG

IgM

Mixed IgG/IgM

4.24

4.45

7.75

4.62

6.62

9.01

4.91

2.66

6.63

9.53

3.28

9.76

4.62

2.39

6.65

4.47

3.28

6.51

4.97

4.25

9.60

review board. All patients gave written informed consent for sample collection and publication of clinical information.

4.81

3.58

6.72

8.52

2.85

7.65

Sample collection and antibody analysis. Serum and CSF

samples were obtained within 4 weeks from symptom onset
and before immunotherapy. GQ1bAb were measured in
the laboratory of Basel University Hospital with the GanglioCombi ELISA (BHLMANN Laboratories, Schnenbuch,
Switzerland), which uses an IgG/IgM mix conjugate. Antibody
titers were expressed as % ratio, as previously described8; 50%
cutoff value defined a positive test.
In the AO group, GQ1bAb were additionally measured with
the BHLMANN GanglioCombi-GM, which uses antibodies
specific for IgG and IgM. Since GQ1bAb values were higher with
the mixed than with the specific IgG ELISA (except for patient 9,
see table 1), we used only the mixed assay measurements for ON
and MFS groups.

10

4.69

3.47

7.90

11

5.14

4.71

17.11

12

4.64

3.08

8.71

13

4.55

3.76

6.75

14

4.47

3.25

7.01

15

4.72

7.57

6.44

16

4.81

3.4

6.63

17

162.62

11.39

NA

18

7.07

4.66

10.46

19

8.35

21.77

28.52

20

4.47

2.55

6.63

Statistical analysis. We calculated sensitivity, specificity, 95%

21

4.24

2.71

5.65

Primary research question. Can serum or CSF GQ1bAb measurements distinguish MFS from other disorders in patients with
isolated ophthalmologic syndrome? This study provides Class III
evidence that serum GQ1bAb accurately distinguish MFS from
other disorders (sensitivity 92%, 95% confidence interval [CI]
62%100%; specificity 97%, 95% CI 85%100%).

Standard protocol approvals, registrations, and patient

consents. The study was approved by the LH institutional

CIs, analyses of variance, and post hoc Tukey testing with R,

3.1 (R Development Core Team, 2015), and also test efficiency,
as defined in the following equation:
Efficiency in % 5

TP 1 TN
3 100
TP 1 TN 1 FP 1 FN

where TP/TN 5 true positive/negative, FP/FN 5 false positive/

negative.
Differences between the AO and MFS groups were assessed
with the t test and 2-tailed Fisher exact test. Results ,0.05 were
regarded as statistically significant.

GQ1bAb were measured in 21 patients

with AO (21 serum, 9 CSF), 13 with ON (13 serum,
13 CSF), and 12 with MFS (12 serum, 10 CSF).
GQ1bAb titers were much higher in MFS
(median 5 330.6%) than in ON (median 5 5.7%)
or AO (median 5 7.6%), with substantial differences
between the non-MFS (AO and ON) and MFS
groups (analysis of variance and Tukey test, p ,
0.001; after log transformation).
Only 1 of the 34 patients without MFS had high
serum GQ1bAb (patient 17, who presented with acute
multicranial neuropathy without ataxia or areflexia)
compared to 11 of 12 patients with MFS who had
strongly positive results (tables 2 and 3). Only 2 of
10 patients with MFS had detectable CSF GQ1bAb

RESULTS

Abbreviations: GQ1bAb 5 GQ1b antibodies; Ig 5 immunoglobulin; NA 5 not assessed.

a
Separated IgG and IgM were measured by the GanglioCombi-GM and the mixed IgG/IgM by the GanglioCombi
assay (BHLMANN Laboratories AG, Schnenbuch, Switzerland). Serum GQ1bAb concentrations in the mixed IgG/
IgM ELISA are higher than those of the specific IgG ELISA
(except for patient 9).

compared to none of the 22 patients without MFS. In

serum, GQ1bAb sensitivity was 92% (95% CI 62%
100%) and specificity 97% (95% CI 85%100%, test
efficiency 95.6%). In CSF, sensitivity was 20% (95%
CI 2%56%) and specificity 100% (95% CI 85%
100%, test efficiency 75%).
There was no difference between the MFS and AO
groups (tables 2 and 3) in age (mean 47.6 vs 55.8 years,
p 5 0.25), sex (female 25% vs 52%, p 5 0.16), or CSF
features, such as albuminocytologic dissociation (30%
vs 56%, p 5 0.36), oligoclonal bands (10% vs 22%,
p 5 0.58), and pleocytosis (20% vs 0%, p 5 0.47).
DISCUSSION The principal finding of this study is
that increased serum GQ1bAb are specific for MFS.
The specificity increases in CSF but at the expense
of a much lower sensitivity. Thus, CSF measurements
provide no further benefit.
Neurology 86

May 10, 2016

1781

2016 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

Table 2

Characteristics of 34 patients without MFS: Patients 121 presented with acute ophthalmoplegia and patients 2234 with ON

Patient

Sex

Age, y

OMN, VA

Additional clinical
findings

Etiology

Serum GQ1bAb

CSF GQ1bAb

CSF

69

III

Bil INO

Primary CNS angiitis

7.75

6.00

OCB

55

VI

Microangiopathic (diabetic)

9.01

NA

Normal

74

III

Microangiopathic

6.63

NA

NA

47

III

Up-beat Ny

Myasthenia gravis

9.76

6.13

EP (529)

59

III

Microangiopathic

6.65

5.91

EP (527)

58

III, VI

Myasthenia gravis

6.51

6.64

Normal

49

IV

Myasthenia gravis

9.60

NA

Normal

44

IV

Undetermined

6.72

NA

NA

75

III Bil

Dysp and dysph

Myasthenia gravis

7.65

NA

NA

10

42

III, IV

VII, multidir. Ny

Myasthenia gravis

7.90

NA

OCB

11

76

III, IV

Microangiopathic

17.11

5.66

EP (499)

12

75

IV

Mesencephalic stroke

8.71

8.7

Normal

13

32

III

Undetermined

6.75

6.50

Normal

14

86

III

Multidir. Ny, cs

Mesencephalic and MCA stroke

7.01

NA

NA

15

33

VI

Undetermined

6.44

5.76

Normal

16

28

VI

Undetermined

6.63

NA

NA

17

60

III

II, V, VII

Undetermined

162.62

9.22

Normal

18

25

VI Bil

V, VII, cs

Postinfectious RE

10.46

NA

Normal

19

44

IV, VI

Myasthenia gravis

28.52

NA

NA

20

75

VI

Microangiopathic (diabetic)

6.63

NA

EP (503)

21

65

VI Bil

II

Microangiopathic (diabetic)

5.65

NA

EP (516)

22

33

,0.1

PO, RAPD

MS

5.8

5.1

OCB

23

40

,0.1

MS

5.3

5.1

OCB

24

38

0.4

MS

5.4

5.5

EWBC (12), EP (572), OCB

25

28

0.5

MS

7.2

5.2

EWBC (13), EP (750), OCB

26

17

,0.1

MS

6.6

5.6

OCB

27

26

,0.1

MS

5.4

Normal

28

56

0.3

CC scotoma

Isolated ON

7.7

5.2

Normal

29

20

,0.1

MS

5.7

5.6

Normal

30

34

,0.1

Postinfectious ON

7.9

5.5

Normal

31

21

,0.1

NMO

5.6

EP (485)

32

39

0.5

MS

5.7

5.6

EWBC (7), EP (470), OCB

33

18

,0.1

MS

4.9

EWBC (60), EP, OCB

34

34

,0.1

MS

5.6

5.4

OCB

Abbreviations: Bil 5 Bilateral; CC 5 cecocentral; cs 5 corticospinal signs; dysp 5 dyspnea; dysph 5 dysphagia; EP 5 elevated CSF proteins (.450 mg/L);
EWBC 5 elevated CSF white blood cells (.5 cells/mm3); GQ1bAb 5 GQ1b antibodies; INO 5 internuclear ophthalmoplegia; MCA 5 middle cerebral artery;
MFS 5 Miller-Fisher syndrome; MS 5 multiple sclerosis; multidir. 5 multidirectional; NA 5 not assessed; NMO 5 neuromyelitis optica; Ny 5 nystagmus;
OCB 5 oligoclonal bands in CSF; OMN 5 ocular motor nerve palsy; ON 5 optic neuritis; PO 5 papilledema; RAPD 5 relative afferent pupillary deficit; RE 5
rhombencephalitis; VA 5 visual acuity.
Mixed IgG/IgM GQ1bAb results in serum and CSF are displayed (except for patient 17, for whom IgG is shown).
a
Positive GQ1bAb value (patient 17).

Serum GQ1bAb are highly sensitive markers of

MFS.1,9 Some studies suggested that they are also
specific to MFS, because they were not found in
healthy individuals4 or patients with other nonneurologic or neurologic conditions without ocular
1782

Neurology 86

dysfunction. Elevated GQ1bAb are associated with

typical MFS, clinically defined by ataxia, areflexia,
and ophthalmoplegia, but their role in isolated syndromes such as AO or ON is less clear. For instance,
it is debatable whether elevated GQ1bAb in patients

May 10, 2016

2016 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

Table 3

Characteristics of 12 patients (3546) with Miller-Fisher syndrome

Patient

Sex

35

Age, y
44

OMN
VI Bil

Additional clinical findings

Ar, dys, acroparesthesias

Serum GQ1bAb
a

247.4
a

CSF GQ1bAb

CSF findings

NA

Normal

36

26

III Bil

Ar, VII

200

NA

Normal

37

40

III, IV, VI Bil

At, Ar, IX, X

286.9a

5.5

EWBC (6), OCB, EP (785)

38
39

F
M

64
33

III Bil, IV, VI

III Bil, IV, VI

At, Ar, hemibody sensory loss

At, Ar, VII

286

11.8

Normal

5.4

Normal

224.5

40

16

VI

At, Ar, convergence paresis,

Ny retractorius, Bil IX, X

374.3

6.8

Normal

41

82

III Bil, IV, VI

At, Ar

428.2a

74a

Normal

42

34

III, VI Bil

Ar, At

433.5

13.6

Normal

43

81

VI Bil

Ar, At, VII Bil

482.6a

54.6a

EP (491)

44

38

III, IV, VI Bil

Ar, I (hyposmia), VII (hypogeusia),

Bil IX, X

482.1a

33

EP (692)

45

65

III, IV, VI Bil

Ar, At, VII, dysph

374.8a

7.8

EP (647)

46

48

VI

At, Ar, VII, IX, X

6.2

7.6

EWBC (11), EP (856)

Abbreviations: At 5 ataxia; Ar 5 hypo-/areflexia; Bil 5 Bilateral; dys 5 dysarthria; dysph 5 dysphagia; EP 5 elevated CSF proteins (.450 mg/L); EWBC 5
elevated CSF white blood cells (.5 cells/mm3); GQ1bAb 5 GQ1b antibodies; NA 5 not assessed; Ny 5 nystagmus; OCB 5 oligoclonal bands in CSF; OMN 5
ocular motor nerve palsy.
Mixed IgG/IgM GQ1bAb results in serum and CSF are displayed.
a
Positive GQ1bAb values.

with AO define an MFS spectrum or are a nonspecific

finding. One study found anti-GQ1b seroreactivity
in 2 of 16 patients with ocular motor nerves palsy of
undetermined origin, but not in 34 patients with
known origin, which included vascular etiology,
myasthenia gravis, or multiple sclerosis.9 However,
few studies reported increased serum GQ1bAb in
patients with ophthalmoplegia of miscellaneous origins, such as diabetes, Tolosa-Hunt syndrome, or
opsoclonus myoclonus.10 These apparently discordant results can be explained by considering the levels
of antibodies. It is likely that low serum concentrations of GQ1bAb, which occur in several disorders,
are nonspecific markers of damage to ocular motor
nerve sheaths, while high serum concentrations are
specific for MFS.
Unlike AO, ON seems to be a rare manifestation
of MFS. It has been described as part of a broader
clinical constellation,2,3 as supported by immunolocalization studies.5 Our findings suggest that
GQ1bAb are not associated with isolated ON.
Even if methodologies for GQ1bAb detection
are not standardized, ELISA assays are used in clinical routine. In our study, the GanglioCombi
mixed IgG/IgM correlated well with the total
IgG, which is the most relevant antibody isotype
for MFS diagnosis, making it a viable alternative
to separate IgG and IgM assays, but at half the cost
and effort.
Retrospective identification of patients with MFS
and ON represents a limitation of our study.
Nevertheless, patients with AO were included

prospectively. Since comparison between ELISA

techniques was performed in this group, this consistently reduced potential sources of bias.
In conclusion, elevated serum GQ1bAb are specific to MFS. Measurement in CSF does not offer
any additional information. Unlike neuromyelitis optica, in which identification of anti-aquaporin-4 antibodies greatly contributed to expand the range of
neuromyelitis opticalike syndromes, GQ1bAb do
not seem to substantially widen the MFS spectrum,
and our results do not support that isolated ON is
part of the supposed incomplete MFS spectrum.
However, this is different for AO. We found
GQ1bAb in only 1 of 21 patients with AO but they
occurred in 1 of 5 patients with AO of unknown
etiology. These findings, substantiated by previous
reports,9 would call for including GQ1bAb in the
workup of patients with AO of undermined origin.
AUTHOR CONTRIBUTIONS
All authors gave substantial intellectual contribution to the submitted
manuscript. Marianna Spatola contributed to the design and conceptualization of the study, performed the analysis and interpretation of the data,
and drafted the manuscript. Renaud Du Pasquier designed and conceptualized the study, contributed to the interpretation of the data and to
the manuscript draft, and revised the manuscript for intellectual content.
Myriam Schluep contributed to the design and conceptualization of the
study and to data collection. Axel Regeniter performed the tests, contributed to the interpretation of the data, and revised the statistics and the
manuscript for intellectual content.

ACKNOWLEDGMENT
The authors warmly thank Professor Josep Dalmau, at IDIBAPS Barcelona, for his critical and constructive input and his help in revising the
manuscript.
Neurology 86

May 10, 2016

1783

2016 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

STUDY FUNDING
No targeted funding reported.

DISCLOSURE

5.

M. Spatola has received financial support from Societ e Academ

ique
Vaudoise and Fondation Pierre Mercier pour la Science. R. Du Pasquier,
M. Schluep, and A. Regeniter report no disclosures relevant to the manuscript. Go to Neurology.org for full disclosures.

6.

Received October 8, 2015. Accepted in final form January 28, 2016.

7.
REFERENCES
1. Chiba A, Kusunoki S, Shimizu T, Kanazawa I. Serum IgG
antibody to ganglioside GQ1b is a possible marker of
Miller Fisher syndrome. Ann Neurol 1992;31:677679.
2. Sasaki Y, Takubo H, Arai T, et al. Atypical Fisher syndrome with optic nerve involvement [in Japanese]. No To
Shinkei 2001;53:571573.
3. Toshniwal P. Demyelinating optic neuropathy with
Miller-Fisher syndrome: the case for overlap syndromes
with central and peripheral demyelination. J Neurol
1987;234:353358.
4. Chiba A, Kusunoki S, Obata H, et al. Serum anti-GQ1b
IgG antibody is associated with ophthalmoplegia in Miller

8.

9.

10.

Fisher syndrome and Guillain-Barr syndrome: clinical

and immunohistochemical studies. Neurology 1993;43:
19111917.
Chiba A, Kusunoki S, Obata H, et al. Ganglioside composition of the human cranial nerves, with special reference
to pathophysiology of Miller Fisher syndrome. Brain Res
1997;745:3236.
Yoshino H, Maeda Y, King M, et al. Sulfated glucuronyl
glycolipids and gangliosides in the optic nerve of humans.
Neurology 1993;43:408411.
Ito M, Kuwabara S, Odaka M, et al. Bickerstaffs brainstem encephalitis and Fisher syndrome form a continuous
spectrum: clinical analysis of 581 cases. J Neurol 2008;
255:674682.
Kollewe K, Wurster U, Sinzenich T, et al. Anti-ganglioside
antibodies in amyotrophic lateral sclerosis revisited. PLoS
One 2015;10:e0125339.
Suzuki T, Chiba A, Kusunoki S, Chikuda M, Fujita T,
Misu K. Anti-GQ1b ganglioside antibody and ophthalmoplegia of undetermined cause. Br J Ophthalmol 1998;82:
916918.
Morales-Briceo H, Rodrguez-Violante M, CervantesArriaga A, et al. Opsoclonus myoclonus syndrome associated
with GQ1b antibodies. Mov Disord 2012;27:16151616.

Complimentary CME and Self-assessment Resources:

An Exclusive AAN Membership Benefit!
Looking for ways to earn important CME credits in 2016? Look no further, because AAN
membership provides FREE* access to the AANs suite of online learning programs: NeuroSAE,
NeuroLearnSM, and NeuroPISM. Access from virtually anywherehome or officeto meet your
CME needs, as well as take the necessary steps toward fulfilling your maintenance of certification
(MOC) requirements, as mandated by the ABPN. Visit AAN.com/view/MOC today!
*Free access is limited to one course per program at a time.

Get Your Front Row Seat to the Premier Event on

Sports Concussion
The worlds leading experts on sports concussion are once again hosting the years premier event on
sports concussion. The American Academy of Neurologys 2016 Sports Concussion Conference,
set for July 8 through 10 in Chicago, features a concussion boot camp and is poised to be the go-to
meeting for all disciplines involved in the prevention, diagnosis, and treatment of professional,
collegiate, and high school sports concussion, including neurologists, athletic trainers, and other
medical professionals. Early registration discounts end June 14, so visit AAN.com/view/
ConcussionConference to secure your spotand savingstoday!

1784

Neurology 86

May 10, 2016

2016 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.