Comparison of Electrodiagnostic Abnormalities and Criteria in a Cohort of Patients With Chronic Inflammatory Demyelinating Polyneuropathy FREE

CURRENT electrodiagnostic criteria for chronic inflammatory demyelinating polyneuropathy (CIDP) are research oriented and designed for selection of patients to clinical trials.^1- 5 To our knowledge, their sensitivities have not generally been evaluated. More inclusive electrodiagnostic criteria may be needed to ensure that potentially treatable patients with CIDP are not misdiagnosed in routine clinical practice.^6- 7 In this study, we reviewed the electrophysiologic results of 15 patients with clinically diagnosed relapsing sensorimotor CIDP to assess the sensitivity of current electrodiagnostic criteria and to identify a set of electrodiagnostic abnormalities that are shared by patients with CIDP.

We retrospectively reviewed the medical records of patients with the clinical diagnosis of relapsing sensorimotor CIDP treated at the Peripheral Neuropathy Center, New York, NY. Inclusionary criteria included progressive weakness and sensory loss for at least 2 months, a relapsing course, and response to immunotherapy using intravenous immunoglobulin or corticosteroids. Fifteen patients were identified. All had weakness, reduced or absent deep tendon reflexes, and impaired or absent vibratory perception at the toes.

Strength was graded according either to a modified Medical Research Council system (5 indicates normal; 4+, minimal weakness; 4, mild weakness; 4− moderate weakness; 3, moves against gravity without resistance; 2, horizontal movement only without resistance; 1, trace movement only; and 0, no movement), or to the Neurologic Disability Score⁸ (0 indicates no weakness; 1, 25% weakness; 2, 50% weakness; 3, 75% weakness; and 4, 100% weakness). Medical Research Council grades 4+, 4, and 4− correspond to Neurologic Disability Score 1, 2, and 3, respectively. χ² Analysis was used to compare patients with symmetric proximal and distal weakness with those with asymmetric or only distal weakness.

The compound muscle action potentials (CMAPs) evoked from the following stimulation sites were included for analysis: tibial (ankle and popliteal fossa), peroneal (ankle and below the fibular head), ulnar (wrist and below the elbow), and median nerve (wrist and elbow). Possible partial conduction block (ppCB) was defined as partial conduction block (with or without temporal dispersion) with a greater than 20% decline of CMAP amplitude (onset-peak) between 2 stimulation sites. The following segments were used for analysis for ppCB: peroneal nerve ankle–fibular head, tibial ankle–popliteal fossa, ulnar nerve wrist–below the elbow, and median nerve wrist–elbow. Demyelinating range values of conduction velocities, distal latencies, and F-wave minimal latencies were defined according to the Ad Hoc Subcommittee of the American Academy of Neurology AIDS Task Force's (AAN) electrodiagnostic CIDP criteria.¹ Demyelinating range tibial H-reflex minimal latency was defined as exceeding 120% of the upper limit of normal. Mild to moderate values of conduction velocities, distal latencies, F-wave minimal latency, or tibial H-reflex minimal latency were defined as abnormalities that did not meet the demyelinating range values. Distal and proximal CMAP duration data were available for analysis in 10 patients. Given the utility of measurement of the distal CMAP to the return to baseline after the last negative peak that is above the baseline,⁵ we have chosen to obtain duration measurements of both the distal and the proximal CMAPs in this fashion. Demyelinating range distal CMAP durations were defined as 9 milliseconds or more.⁵ Proximal CMAP temporal dispersion, defined as the percentage change between the proximal and distal CMAP durations, was calculated for each nerve. Abnormal temporal dispersion was defined as a proximal distal CMAP duration increase of more than 30%.⁹ Absence of F waves in median or ulnar nerves with distal CMAP amplitudes of 75% or more of the lower limit of normal was considered consistent with demyelination.¹⁰ These data were analyzed using the electrodiagnostic components of the following 5 published CIDP diagnostic criteria: Ad Hoc Subcommittee for the AAN,¹ Hughes et al for the Inflammatory Neuropathy Cause and Treatment Group (INCAT),² Saperstein et al,³ Nicolas et al,⁴ and Thaisetthawatkul et al (distal CMAP).⁵

The clinical findings in the 15 patients are summarized in Table 1. Three patients had an IgG K monoclonal gammopathy with no evidence of cancer on bone marrow and skeletal survey studies. In 2 patients the distal legs were affected (including ankle dorsiflexion), as may be seen in distal acquired demyelinating symmetric neuropathy.³ Five patients (33%) had motor asymmetry, defined as at least 1-point side-side difference in strength in at least 1 muscle group on individual muscle testing, as may be seen in multifocal acquired demyelinating sensory and motor neuropathy.³ χ² Analysis revealed no statistically significant difference in distribution or asymmetry of weakness between the 9 patients who met 1 or more of 4 established CIDP research criteria (AAN,¹ INCAT,² Saperstein et al,³ or Nicolas et al⁴) and the 6 patients who did not. Following immunomodulant therapy, all patients had at least a 1-point improvement in strength in at least 1 muscle group on either the modified Medical Research Council grade or the Neurologic Disability Score. Improvement was seen in all patients following reinstitution of immunotherapy following clinical relapses.

Electrodiagnostic studies were performed prior to the initiation of therapy (Table 1). The number of patients who had particular electrophysiologic abnormalities in at least 1 nerve is listed in Table 2. Results by number of nerves examined are given in Table 3 and Table 4.

The percentage of patients fulfilling each electrodiagnostic criteria set was 40% for the AAN,¹ 47% for Saperstein et al,³ 53% for Nicolas et al,⁴ and 60% for the INCAT.² Six patients (40%) did not satisfy any of the aforementioned criteria. Seven (70%) of 10 patients met the Thaisetthawatkul et al⁵ distal CMAP duration criteria. Combining the aforementioned 4 criteria with the distal CMAP criteria resulted in the following increased combined sensitivities: 70% for the AAN, 70% for Saperstein et al, 70% for Nicolas et al, and 80% for INCAT. With regard to the criteria of INCAT and Nicolas et al, the full recommended number of 8 motor nerves or nerve segments was not uniformly studied, but our approach is more typical of how these criteria would be applied in clinical practice rather than for inclusion in a clinical research study.

All 15 patients had abnormalities of CMAP amplitude, distal latency, conduction velocity, or F-wave or H-reflex minimal latencies in at least 3 nerves. Additionally, they had ppCB (11 patients) or demyelinating range (11 patients) abnormalities in at least 1 nerve (Table 5). In the 4 patients who had ppCB as the only demyelinating feature, the proximal–distal CMAP amplitude decline was more than 30%, although enrollment criteria required only a greater than 20% amplitude decline.

Historically, CIDP has been a clinical diagnosis based on the presence of progressive weakness and sensory loss, and the patient's response to prednisone therapy.¹¹ The typical phenotype is that of symmetric proximal and distal weakness, but as many as 50% of the CIDP-affected patients have atypical features such as asymmetry, distal weakness, or only sensory loss, although all respond to immunotherapy.² The presence of demyelination in a biopsy specimen of a nerve or an increased cerebrospinal fluid protein concentration can help make the diagnosis, but are not always present and are not mandatory.^2,11 For this study, we included only patients who had a relapsing course, with weakness and sensory loss that responded to immunotherapy on more than 1 occasion, making misdiagnosis unlikely.

In comparison with the findings of 2 previous studies of CIDP,^12- 13 our patients had fewer demyelinating abnormalities, possibly owing to differences in inclusionary criteria, disease severity, or the number of nerves or nerve segments evaluated. Demyelination, for example, may be more commonly missed in proximal nerve segments that are not commonly evaluated or accessible to routine electrodiagnostic studies.^14- 15 Patient 2, for example (Table 1), had 3 definite sites of conduction block in the cervical root to the Erb point segments on electrical cervical root stimulation, and only 1 ppCB in more distal segments. This patient met none of the 5 CIDP electrodiagnostic criteria, even though 4 extremities and proximal (median and ulnar) nerve segments (Erb point and axilla) were evaluated.

A cutoff of 20% for ppCB was chosen in this study based on AAN criteria. However, analysis of these data indicates that a greater than 30% decline cutoff would have been sufficient. This cutoff is likely to adequately preserve specificity as suggested by data from Cappellari et al,¹⁶ with maximal distal–proximal CMAP declines of less than 20% (median and ulnar nerves) and less than 30% (peroneal nerve) in amyotrophic lateral sclerosis, and less than 10% (median and ulnar nerves) and 23% (peroneal nerve) in healthy subjects. In computer modeling¹⁷ physiological temporal dispersion may result in an amplitude decline exceeding 50%, but application of more stringent criteria would result in loss of sensitivity. In our patients, defining CB as a proximal CMAP amplitude decline of greater than 40% or more than 50%, would have reduced the sensitivity to 93% or 87%, respectively. Tibial distal–proximal CMAP decrement was excluded as evidence of ppCB from our minimal criteria, as popliteal fossa stimulation may be submaximal for purely technical reasons. However, one may consider using tibial CB with a proximal CMAP decrement exceeding 50%⁹ if care is taken to provide a supramaximal popliteal fossa stimulus.

The findings of this study indicate that published electrodiagnostic criteria may fail to diagnose a large proportion of patients having CIDP seen in clinical practice. The sensitivity of established electrodiagnostic criteria in this cohort was only 40% to 70%, similar to that recently reported by Thaisetthawatkul et al.⁵ Thirty percent to 60% of the patients would not have been diagnosed as having CIDP or treated for CIDP if any of the current criteria had been applied, as is required for approval of therapy by some payors.¹⁸ In the clinical setting such diagnostic sensitivity may be unacceptable given that CIDP may cause secondary axonal involvement and lead to disability.¹⁹

Based on the findings in this cohort, the diagnosis of CIDP could be considered in patients with chronic sensorimotor neuropathy of otherwise unknown origin, with electrodiagnostic abnormalities in at least 3 nerves and ppCB or demyelinating range abnormalities in at least 1 nerve (Table 5). These abnormalities, however, may not be exhibited in patients with certain subtypes of CIDP such as sensory CIDP,^20- 21 or their presence might not help distinguish between CIDP and other causes for neuropathy, such as diabetic neuropathy where more stringent diagnostic criteria may be required.^22- 23 Proposed diagnostic criteria would have to be tested in prospective clinical trials.

Accepted for publication July 23, 2003.

Author contributions: Study concept and design (Drs Magda, Latov, Brannagan, and Sander); acquisition of data (Drs Magda, Weimer, and Sander); analysis and interpretation of data (Drs Magda, Latov, Brannagan, Chin, and Sander); drafting of the manuscript (Drs Magda, Weimer, and Sander); critical revision of the manuscript for important intellectual content (Drs Magda, Latov, Brannagan, Chin, and Sander); study supervision (Drs Magda, Latov, Brannagan, Weimer, Chin, and Sander).

Coresponding author: Paul Magda, DO, Peripheral Neuropathy Center, Department of Neurology, Weill Medical College of Cornell University, 635 Madison Ave, Suite 400, New York, NY 10022 (e-mail: pam2003@med.cornell.edu).