Electrophysiological sensory demyelination in typical chronic inflammatory demyelinating polyneuropathy

Background: The presence of electrophysiological demyelination of sensory nerves is not routinely assessed in the evaluation of suspected chronic inflammatory demyelinating polyneuropathy (CIDP). Whether this can be useful is unknown. Methods: We compared, using surface recording techniques, in 19 patients with typical CIDP and 26 controls with distal large fibre sensory axonal neuropathy, the forearm median sensory conductions, sensory nerve action potential (SNAP) amplitudes and durations and sensory nerve conduction velocities (SNCVs) of median, radial and sural nerves. Results: Median nerve sensory conduction block (SCB) across the forearm was greater in CIDP patients than in controls (P = 0.005). SNAP durations were longer in CIDP patients for median (P = 0.001) and sural nerves (P = 0.004). Receiver operating characteristic (ROC) curves provided sensitive (>40%) and specific (>95%) cut-offs for median nerve SCB as well as median and sural SNAP durations. SNCVs were significantly slower for median and sural nerves in CIDP patients, but ROC curves did not demonstrate cut-offs with useful sensitivities/specificities. Median SCB or prolonged median SNAP duration or prolonged sural SNAP duration offered a sensitivity of 73.7% for CIDP and specificity of 96.2%. Used as additional parameters, they improved diagnostic sensitivity of the American Academy of Neurology (AAN) criteria for CIDP of 1991, from 42.1% to 78.9% in this population, with preserved specificity of 100%. Discussion: Sensory electrophysiological demyelination is present and may be diagnostically useful in typical CIDP. SCB detection and SNAP duration prolongation appear to represent more useful markers of demyelination than SNCV reduction.