Article Text

Download PDFPDF

Research paper
Electrophysiological diagnosis of Guillain–Barré syndrome subtype: could a single study suffice?
  1. Yusuf A Rajabally1,
  2. Marie-Christine Durand2,
  3. James Mitchell1,
  4. David Orlikowski3,
  5. Guillaume Nicolas4
  1. 1Regional Neuromuscular Clinic, Queen Elizabeth Neurosciences Centre, University Hospitals of Birmingham, Birmingham, UK
  2. 2Department of Neurophysiology, Hôpital Raymond-Poincaré (AP-HP), Université Versailles-Saint-Quentin-en-Yvelines, Garches, France
  3. 3Medical Intensive Care Unit, Hôpital Raymond-Poincaré (AP-HP), Université Versailles-Saint-Quentin-en-Yvelines, Garches, France
  4. 4Department of Neurology, Hôpital Raymond-Poincaré (AP-HP), Université Versailles-Saint-Quentin-en-Yvelines, Garches, France
  1. Correspondence to Dr Yusuf A Rajabally, Regional Neuromuscular Clinic, Queen Elizabeth Neurosciences Centre, University Hospitals of Birmingham, Birmingham B15 2WB, UK; Yusuf.Rajabally{at}uhb.nhs.uk

Abstract

Background Serial electrophysiology has been suggested as essential for accurate diagnosis in Guillain–Barré syndrome (GBS). However, whether more adapted electrophysiological criteria may allow a single study to be sufficient is unknown.

Methods We retrospectively reviewed records of 365 consecutive patients with GBS from Birmingham, UK, and Garches, France, admitted between 1998 and 2013. Electrophysiology was analysed using existing criteria as well as a set of modified criteria, developed using sensitive and specific cut-off values for demyelination and incorporating new knowledge on electrophysiology of axonal GBS. We compared diagnostic rates and classification changes using modified criteria with published literature relating to serial studies.

Results With existing criteria, we found similar proportions of acute inflammatory demyelinating polyradiculoneuropathy (AIDP) (71.5% vs 72%; p=1), axonal GBS (17.5% vs 14.7%; p=0.62) and equivocal forms (9.9% vs 13.3%; p=0.41) to the previous studies considered. With modified criteria, we identified comparable rates of AIDP (56.2% vs 58.7%; p=0.70), axonal GBS (35.1% vs 36%; p=0.89) and equivocal forms (7.7% vs 5.3%; p=0.63) with a single nerve conduction study as compared with when serial electrophysiology was used in previous analyses. We observed an identical diagnostic shift from AIDP to axonal GBS with modified criteria as that described with serial studies (21.5% vs 18.5%; p=0.72). Classification changes with modified criteria correlated significantly with performing of electrophysiology ≤7 days after symptom onset (p=0.045), indicating their greater usefulness in earlier disease stages.

Conclusions A single electrophysiological study may suffice to establish the ultimate electrodiagnosis of GBS subtype if the proposed modified electrodiagnostic criteria are used.

  • GUILLAIN-BARRE SYNDROME
  • NEUROMUSCULAR
  • NEUROPHYSIOL, CLINICAL
  • NEUROPATHY
View Full Text

Statistics from Altmetric.com

Introduction

Guillain–Barré syndrome (GBS) consists of different subtypes, including acute inflammatory demyelinating polyradiculoneuropathy (AIDP), and axonal forms (acute motor axonal neuropathy (AMAN) and acute axonal sensory and motor neuropathy (AMSAN)). In those axonal forms, nodal and paranodal involvement resulting from antibody-mediated attack result in diverse electrophysiological phenomena that are now well described.1 These include axonal loss with reduced compound muscle action potentials (CMAPs), acute motor conduction block (CB) with or without subsequent reversible conduction failure or isolated F-wave absence.1 ,2 Combinations of those features may be present in different nerves of the same patient at one study.

It has been shown that a single electrophysiological study is probably inadequate for correct ultimate diagnosis of disease subtype.3 ,4 However, use of current electrophysiological criteria for GBS may represent the main reason for this due to their lack of specificity for demyelination relating to cut-off values used, particularly for distal motor latency (DML) and motor conduction velocity (MCV). These criteria furthermore do not allow appropriate consideration of the various forms of axonal GBS that are now recognised. Whether the use of more adapted electrophysiological criteria could make earlier accurate diagnosis possible with a single nerve conduction study has not, to our knowledge, been studied.

Methods

We retrospectively reviewed our institutional databases of patients admitted with a diagnosis of GBS between 2007 and 2012 (Birmingham, UK) and 1998 and 2013 (Garches, France).

The diagnosis was made in each case in accordance with established clinical criteria.5 Patients included had undergone electrophysiological testing of at least four motor and two sensory nerves within 21 days of symptom onset. Electrophysiology was performed according to standard methods by a qualified senior physician trained and experienced in electromyography using routine procedures and different neurophysiological equipment over the years of the study. The CMAPs were evoked from the median nerve (stimulating at wrist and elbow, and recording at the Abductor Pollicis Brevis muscle), ulnar nerve (stimulating at wrist and below elbow, and recording at the Abductor Digiti Minimi muscle), common peroneal nerve (stimulating at ankle and fibular neck and recording at the Extensor Digitorum Brevis muscle) and tibial nerve (stimulating at ankle only or ankle and popliteal fossa and recording at the Abductor Hallucis muscle). Measured parameters were MCV, DML, minimum F-wave latency, distal CMAP amplitude and presence of CB as defined within the criteria considered. Results were analysed with our laboratories’ normal values. Fulfilment of each set of electrodiagnostic criteria was ascertained in each case.

We analysed records using current criteria, described by Hadden et al (box 1).6 We, in addition, devised a modified set of criteria with the aim of correctly identifying demyelination as well as axonal variants (box 2). To define demyelination, we used published cut-offs7 excluding temporal dispersion for which we did not have required data in either of our two centres. We chose to use these cut-offs first as the originally described diagnostic sensitivity of these criteria in GBS was of 64%,7 which interestingly approached the rate of AIDP found after serial studies,3 ,4 suggesting what we considered was their potential suitability in correctly differentiating AIDP from axonal GBS with minor demyelinating features. Furthermore, these criteria, which have served as basis for recent electrodiagnostic criteria for chronic inflammatory demyelinating polyradiculoneuropathy (CIDP),8 are of established high specificity for demyelinating versus axonal neuropathy in several populations.7 ,9 ,10 We otherwise attempted to effectively integrate within the modified criteria, characteristics of axonal GBS as recently described.1 This is detailed in box 2. For that purpose, we considered exclusive presence of CBs as indicative of an axonal form. Similarly, if only one CB was present, but one CMAP of one other nerve was low, we categorised this as indicative of axonal GBS. We considered F-wave absence as also suggestive of axonopathy if two or more nerves were involved. We distinguished pure motor forms from motor and sensory axonal forms on the basis of sensory action potentials, which had to be reduced in two nerves at ≤50% of the lower limit of normal.11 Finally, we grouped for the purposes of this analysis, inexcitable forms within the axonal subtype.

Box 1

Hadden et al's6 electrodiagnostic criteria for Guillain–Barré syndrome

  1. Normal

    • (All the following in all nerves tested)

    • DML ≤100% ULN

    • F-wave present with latency ≤100% ULN

    • MCV ≥100% LLN

    • Distal CMAP ≥100% LLN

    • Proximal CMAP ≥100% LLN

    • Proximal CMAP/distal CMAP ratio >0.5

  2. Primary demyelinating

    • (At least one of the following in each of at least two nerves, or at least two of the following in one nerve if all others inexcitable and distal CMAP ≥10% LLN)

    • MCV <90% LLN (85% if Distal CMAP <50% LLN)

    • DML >110% ULN (120% if Distal CMAP <100% LLN)

    • Proximal CMAP/distal CMAP ratio <0.5 and distal CMAP ≥20% LLN

    • F-response latency >120% ULN

  3. Primary axonal

    • None of the above features of demyelination in any nerve (except one demyelinating feature allowed in one nerve if distal CMAP <10% LLN) and

    • Distal CMAP <80% LLN in at least two nerves

  4. Inexcitable

    • Distal CMAP absent in all nerves (or present in only one nerve with distal CMAP <10% LLN)

  5. Equivocal

    • Does not exactly fit criteria for any other group

CMAP, compound muscle action potentials; DML, distal motor latency; LLN, lower limit of normal; MCV, motor conduction velocity; ULN, upper limit of normal.

Box 2

Proposed modified set of electrodiagnostic criteria for Guillain–Barré syndrome (based on Van den Bergh and Piéret,7 for demyelinating cut-offs and incorporating use of new knowledge on axonal GBS to define primary axonal forms1,2)

  1. Normal

    • (All the following in all nerves tested)

    • DML ≤100% ULN

    • F-wave present with latency ≤100% ULN

    • MCV ≥100% LLN

    • Distal CMAP ≥100% LLN

    • Proximal CMAP/distal CMAP ratio >0.7 (excluding the tibial nerve)

  2. Acute inflammatory demyelinating polyradiculoneuropathy (AIDP)

    • At least one of the following in at least two nerves:

      • MCV <70% LLN

      • DML >150% ULN

      • F-response latency >120% ULN, or >150% ULN (if distal CMAP <50% of LLN)

    • OR

      • F-wave absence in two nerves with distal CMAP ≥20% LLN, with an additional parameter, in one other nerve

    • OR

      • Proximal CMAP/distal CMAP ratio <0.7 (excluding the tibial nerve), in two nerves with an additional parameter, in one other nerve

  3. Axonal GBS including inexcitable forms

    • Axonal GBS:

    • None of the above features of demyelination in any nerve (except one demyelinating feature allowed in one nerve if distal CMAP <10% LLN), and at least one of the following:

      • Distal CMAP <80% LLN in two nerves

      • F-wave absence in two nerves with distal CMAP ≥20% LLN, in absence of any demyelinating feature in any nerve

      • Proximal CMAP/distal CMAP ratio <0.7, in two nerves (excluding the tibial nerve)

      • F-wave absence in one nerve with distal CMAP ≥20% LLN OR proximal CMAP/distal CMAP ratio <0.7 (excluding the tibial nerve), in one nerve; with IN ADDITION, distal CMAP <80% LLN in one other nerve

    • Inexcitable:

    • If distal CMAP absent in all nerves (or present in only one nerve with distal CMAP <10% LLN)

  4. Equivocal

    • Abnormal range findings however not fitting criteria for any other group

CMAP, compound muscle action potentials; DML, distal motor latency; GBS, Guillain–Barré syndrome; LLN, lower limit of normal; MCV, motor conduction velocity; ULN, upper limit of normal.

We classified patients with AIDP, axonal GBS or with ‘equivocal’ electrophysiology with current criteria and established diagnostic rates and classification changes with use of modified criteria. Findings were compared with recent published literature on the resulting classification modifications from the use of serial electrophysiological studies.3 ,4 Comparison of proportions was performed using Fisher's exact tests. Changes in diagnostic sensitivity were assessed by McNemar's test. Correlations of electrophysiological findings were performed with timing of nerve conductions using Spearman's rank correlation analysis. Significance level was set at p values <0.05.

This study, which was part of a larger retrospective analysis of clinical and electrophysiological features of GBS at our institutions, was registered and approved by our respective relevant institutional boards. Ethics committee approval was not required.

Results

We included 365 patients (48 from Birmingham from 2007 to 2012 and 317 from Garches from 1998 to 2013). One hundred and nine patients were excluded in total on the basis of incomplete clinical details, delayed electrophysiology performed >21 days after disease onset, insufficiently exhaustive electrophysiology (<4 motor nerves and <2 sensory nerves tested), a diagnosis of Miller–Fisher syndrome or a diagnosis of acute-onset CIDP. There were 210 males and 155 females, corresponding to a gender ratio of 4:3. Mean age was 52.1 years (SD 17.8). The median interval between symptom-onset and performing of the electrophysiological studies was 9 days (IQR 6).

The main findings are shown in figure 1. In the initial analysis, using Hadden et al's criteria, 71.5% of patients (261/365) were diagnosed with AIDP, 17.5% (64/365) had axonal GBS and 9.9% (36/365) had equivocal electrophysiology. Four patients (1.1%) had normal electrophysiology. Using modified criteria, the number of patients with AIDP decreased from 261 to 205 (70.9% to 56.2%). The number of patients with axonal GBS increased from 64 to 128 (17.5% to 35.1%). Among those, 97 (75.8%) had a pure motor form and 31 (24.2%) had a sensory and motor form (AMSAN). The number of equivocal cases was reduced from 36 to 28 (9.9% to 7.7%). The number of cases with normal electrophysiology remained unchanged at 4 (1.1%). The main classification shift occurred from an initial diagnosis of AIDP to a subsequent one of axonal GBS (56 patients; ie, 21.5% of initially diagnosed AIDP cases). There was also a proportion of patients diagnosed with equivocal electrophysiology by Hadden et al's criteria, who were rediagnosed with axonal GBS (eight patients; 22.2% of the equivocal group). There were no shifts from an initial diagnosis of axonal GBS or equivocal electrophysiology to AIDP. All abovementioned results obtained were comparable in our two centres considered individually.

Figure 1

Proportions of Guillain–Barré syndrome subtypes and changes in the electrophysiological, classification in 365 patients with Guillain–Barré syndrome from Birmingham, UK, and Garches, France, by, use of proposed modified criteria after initial application of Hadden et al's6 criteria.

An initial diagnosis of equivocal electrophysiology with Hadden et al's criteria otherwise correlated significantly with performing of early electrophysiology, defined as ≤7 days after symptom-onset (Spearman's r=0.134; p=0.017). Similarly, there was a significant correlation between the occurrence of classification changes with use of the modified criteria and early electrophysiology (Spearman's r=0.113; p=0.045), with 43/64 (67.2%) of all diagnostic changes occurring in patients having undergone electrophysiology ≤7 days. There was a significant increase in diagnostic sensitivity for non-equivocal GBS (89% to 91.3%), resulting from the reduction in number of equivocal cases (McNemar's test: p=0.013).

A comparative analysis with previous serial studies3 ,4 is shown in table 1. There were no significant differences for proportions of AIDP (71.5% vs 72%; p=1.0), axonal GBS (17.5% vs 14.7%; p=0.62) or equivocal forms (9.9% vs 13.3%; p=0.41) using Hadden et al's criteria. Use of the modified criteria with a single set of nerve conductions demonstrated similar rates of AIDP (56.2% vs 58.7%; p=0.70), axonal GBS (35.1% vs 36%; p=0.89) and equivocal electrophysiology (7.7% vs 5.3%; p=0.63) with those obtained with these serial electrophysiological studies using Hadden et al's criteria. The rate of classification change was comparable (17.5% vs 24%; p=0.20), and the rate of reclassification from AIDP to axonal GBS was also similar (21.5% vs 18.5%; p=0.72).

Table 1

Changes in the electrophysiological classification in 365 patients with Guillain–Barré syndrome from Birmingham, UK, and Garches, France, by use of a proposed modified set of electrophysiological criteria compared with that resulting from the use of serial studies with use of Hadden et al's criteria,6 in a total of 75 patients from Italy and Malaysia3 ,4

Discussion

The main finding of our study is that use of the proposed modified criteria for GBS alters, in this large patient cohort from two tertiary centres, the electrodiagnosis in similar proportions and in an identical manner to that resulting from serial studies. This suggests the modified electrodiagnostic criteria may be more adequate, enabling improved early and accurate electrodiagnosis using a single nerve conduction study performed in the first 3 weeks after symptom onset. The use of the modified criteria appears of maximal diagnostic benefit in the first 7 days of the disease. This latter result is consistent with the modified criteria allowing earlier accurate electrodiagnosis, comparable with that which may be achieved with delayed studies at 3–10 weeks.3 Also, with the modified criteria, the proportion of equivocal cases is reduced, resulting in an improved sensitivity for a non-equivocal diagnosis of GBS, that is, of AIDP or axonal GBS. This is helpful as it offers a greater likelihood, with a single study, of reaching a firm diagnosis, which remains in practice the main purpose of electrophysiology in the clinical setting.12

Our study has a number of limitations starting with its retrospective design. The long study period, number of different electromyographers, use of different electrophysiological equipment and absence of a pre-established nerve conduction study protocol may all have impacted upon the findings. Disease severity and timing of treatment was not considered specifically here. We were in addition unable to consider both distal and proximal temporal dispersion within the modified criteria, which however clearly represent sensitive markers of demyelination.7 ,13 Preferably also, it could be argued that our results should have been internally validated by a subgroup of patients having undergone serial studies to ascertain confirmation of our findings with the modified criteria on a case-by-case basis. We are now considering such an analysis that was unfortunately not possible due to serial electrophysiology not being performed routinely at set intervals at our institutions in the past. Finally, a further limitation of our study was its purely electrophysiological focus, without, in particular, microbiological and immunological components. These may have been of interest particularly in the cases diagnosed with axonal GBS after an initial diagnosis of AIDP, given the significantly higher rate of antiganglioside antibody positivity described in axonal compared with demyelinating forms.3 ,14

A comparative analysis with existing published data relating to delayed diagnostic change with serial electrophysiology, as we did in this study, represents, in our opinion, a necessary starting point in validating the use of the proposed modified criteria. The justification for serial studies in rectifying the diagnostic subtype in a proportion of cases of GBS has convincingly been put forward in recent years,3 ,4 and comparison with such serial analyses therefore appears a logical validating method for any new set of criteria. No definite gold standard is available for determining correctness of the electrodiagnosis of GBS, as AIDP and axonal forms cannot be distinguished on clinical grounds or therapeutic response and do not have absolute biomarkers. In terms of cut-off values we used for each parameter, it appears difficult, despite the limitations of such methods, not to use predefined limits to diagnose demyelination or nodo-paranodopathy, as we have done here, in order to have a meaningful interpretation of the electrophysiological data. For that purpose, use of cut-offs that have shown robustness in CIDP, or that are based on newly available and well-documented data on axonal GBS subtypes, as we used here, although potentially arbitrary, does as a result, appear ultimately appropriate. Our results seem to confirm this in the present study.

We believe that despite the drawbacks, our findings bring new insight into the electrodiagnosis of GBS, indicating for the first time in a large series that a single, early, nerve conduction study may suffice for precise subtype classification with use of the proposed modified criteria. Outside considerations of elucidation of pathophysiology, long-term outcome or of clinico-electrophysiological correlations, this importantly indicates that, from a purely clinical and diagnostic point of view, only a single study may be needed in over 90% of patients with GBS for an accurate diagnosis. Further, prospective studies are now required to test the validity of our new proposed criteria.

References

View Abstract

Footnotes

  • Contributors YAR: conception, organisation and execution; design, statistical analysis; writing of manuscript. M-CD: organisation and execution; design, review and critique of manuscript. JM: organisation and execution; review and critique of manuscript. DO: design, review and critique of manuscript. GN: conception; organisation and execution; design, writing of manuscript.

  • Competing interests YAR has received speaker/consultancy honoraria from LfB France, Griffols and BPL and has received educational sponsorships from LfB France, CSL Behring and Baxter. M-CD, JM and DO have no disclosures. GN has received departmental research support/honoraria from Debiopharm, GSK, LfB France, Ipsen and Novartis.

  • Ethics approval Clinical Audit Registration Office.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles