Electrophysiological profile of polyneuropathy, organomegaly, endocrinopathy, monoclonal protein, and skin changes syndrome: demyelinating features and lower-limb-predominant response loss

Background and aims Peripheral neuropathy is a major manifestation of polyneuropathy, organomegaly, endocrinopathy, monoclonal protein, and skin changes (POEMS) syndrome. We characterized its electrophysiological phenotype and examined whether baseline serum vascular endothelial growth factor (VEGF) reflected concurrent neurological or electrophysiological severity. Methods We retrospectively reviewed 32 patients with POEMS syndrome. Clinical findings, baseline serum VEGF, nerve conduction studies, needle electromyography, and median nerve terminal latency index (TLI) were analyzed. Motor conduction findings were additionally assessed according to the 2021 EAN/PNS electrodiagnostic criteria. Results Among patients with evaluable motor conduction data, 75.0% fulfilled criteria strongly supportive of demyelination. Peroneal and tibial CMAPs were unrecordable on at least one side in 13/24 patients (54.2%) for each nerve, whereas median and ulnar CMAPs remained recordable in all evaluable patients. Sural SNAPs were unrecordable in 14/24 patients (58.3%). Active denervation was identified in 23/24 patients (95.8%) with available needle electromyography records, with lower-limb involvement in 22/23 (95.7%) of those showing active denervation. Median nerve TLI was >0.38 in 13/22 patients (59.1%). Ankle dorsiflexion strength showed a moderate positive correlation with peroneal CMAP amplitude ( ρ = 0.537, p = 0.015), and disease duration was inversely correlated with sural SNAP amplitude ( ρ = −0.461, p = 0.023). Other clinical–electrophysiological associations were not significant. Baseline serum VEGF was not consistently associated with cross-sectional neurological or electrophysiological severity. Interpretation In this POEMS cohort, demyelinating abnormalities coexisted with lower-limb-predominant motor and sensory response loss. Frequent active denervation supports concomitant motor axonal involvement. Median nerve TLI provided an additional descriptive feature, whereas a single baseline VEGF measurement may not closely reflect concurrent peripheral nerve dysfunction.