Charcot-Marie-Tooth disease type 1A, or CMT1A, is the most common hereditary neuropathy in the world, caused by a 1.5-million-base-pair duplication on chromosome 17 that includes the PMP22 gene, which encodes a critical protein of peripheral myelin. The result is a dysmyelinating disorder in which nerve conduction slows uniformly across the peripheral nerves, and the large myelinated fibers that power movement and fine touch bear the brunt of the damage. RFC1 disease, by contrast, stems from biallelic AAGGG pentanucleotide repeat expansions in intron 2 of the RFC1 gene and typically announces itself as a late-onset sensory neuronopathy, often packaged together with cerebellar ataxia and vestibular areflexia in the syndrome known as CANVAS. Its true prevalence is likely underestimated because many carriers remain undiagnosed.

Small fibers — the thinly myelinated Aδ fibers and the unmyelinated C fibers — have long been the poor cousins of neurology. They do not show up on conventional nerve conduction studies, yet they mediate pinprick sensation, heat perception, and the automatic control of blood pressure, sweating, and gut function. When they fail, patients can suffer burning pain, insensitivity to injury, dizziness on standing, and devastating gastrointestinal problems. Capturing their dysfunction has always required a patchwork of imperfect tools, which is precisely why the Italian research team behind the new study assembled such an unusually broad battery of tests.

The study enrolled 21 patients with genetically confirmed CMT1A, 16 patients with confirmed biallelic RFC1 expansions, and 21 healthy controls at a single referral center in Genoa. Participants with diabetes, paraproteinemia, vitamin deficiencies, or other neuropathy risk factors were excluded to keep the picture clean. Each patient underwent neurological examination with standardized severity scales, autonomic questionnaires including the COMPASS-31 and the Compound Autonomic Dysfunction Test, electrochemical skin conductance measurement with a Sudoscan device, somatosensory evoked potentials, skin biopsies at the distal leg and thigh, and — most innovatively — two kinds of evoked potential recordings elicited through a micropatterned interdigitated electrode with an inter-rail gap of just 150 micrometers.

That electrode, developed in Italy and applied to the hairy skin of the hand, delivers electrical pulses fine enough to activate intraepidermal nerve endings directly, bypassing the thicker structures that normally dominate a surface stimulus. When driven in sustained rhythmic bursts, it produces what the researchers call nociceptive evoked potentials, whose early N40 component is thought to reflect the arrival of fast Aδ-mediated pain signals at the primary somatosensory cortex. When driven in sparse, randomized bursts, the same electrode elicits pain-related evoked potentials, whose later N1 and N2-P2 components reflect higher-order cortical processing of the painful input. In effect, the team could probe both the express lane and the processing center of the brain’s pain circuitry using a single stimulus site.

The clinical contrasts between the two patient groups were dramatic. Neuropathic pain, defined by a DN4 questionnaire score of four or higher, affected 33 percent of the CMT1A patients but a staggering 81 percent of those with RFC1 disease. Autonomic dysfunction told a similar story: 94 percent of RFC1 patients had abnormal autonomic test scores, with COMPASS-31 symptom scores ranging from 7 to 46, and 63 percent showed reduced electrochemical skin conductance, compared with only 24 percent of CMT1A patients. RFC1 patients commonly reported orthostatic intolerance, bowel and bladder dysfunction, and in severe cases even intestinal volvulus requiring surgery or urinary retention requiring catheterization. Two RFC1 patients had suffered accidental burns because they could no longer feel heat properly — a sobering reminder that losing small fibers can be as dangerous as gaining pain.

The neurophysiology sharpened the distinction further. In CMT1A, the N40 nociceptive responses were either absent in a third of patients or markedly delayed, with latencies exceeding three standard deviations above the control mean, and the later N2 pain-related responses were prolonged and frequently absent as well. This pattern of slowed but surviving pathways fits neatly with the underlying dysmyelinating process, which is known to thin and distort the myelin sheaths of even the small myelinated afferents. In RFC1 disease, the opposite emerged: responses were simply non-recordable in roughly half of patients, while the latencies of the responses that could be recorded were relatively preserved. The researchers interpret this as evidence that in RFC1 disease the nociceptive afferents are not slowed but lost outright, consistent with the neuron death that defines a sensory neuronopathy.

Skin biopsies delivered the most visually compelling confirmation. In CMT1A, the density of intraepidermal nerve fibers — the tiny C-fiber endings that snake up between skin cells — was reduced below the fifth percentile in 12 of 13 biopsied patients, but the loss was concentrated in the distal leg and spared the thigh, a classic length-dependent gradient in which the longest nerves die first. In RFC1 disease, the denervation was catastrophic and indiscriminate: median fiber density collapsed to just 1.0 fiber per millimeter in the distal leg and 0.4 fibers per millimeter in the distal thigh, with nearly complete loss at both proximal and distal sites. That non-length-dependent pattern mirrors the destruction of sensory neuron cell bodies in the dorsal root ganglia rather than a dying-back process along the nerve length, and it aligns with the diffuse allodynia some RFC1 patients reported across the trunk, head, and limbs.

The authors are careful about the limits of their work: the sample was small, the design was cross-sectional, quantitative sensory testing was not performed, and normative values for the new electrode-based evoked potentials do not yet exist, so the technique cannot yet stand alone as a diagnostic biomarker. One intriguing case, however, hints at what longitudinal studies might find — a patient whose burning pain and skin-biopsy denervation appeared years before any abnormality on conventional nerve conduction studies, only to be genetically diagnosed with RFC1 disease later. If small-fiber damage truly precedes large-fiber failure in RFC1 disease, the multimodal toolkit assembled in this study could one day catch the illness at its earliest, most treatable stage. For now, the message is clear: two diseases once defined by their large-fiber damage are also small-fiber diseases, and each writes its signature in a different hand.

Subject of Research: Small nerve fiber dysfunction in CMT1A and RFC1-related hereditary neuropathy

Article Title: Distinct small-fiber dysfunction profiles in CMT1A and RFC1 disease: a multimodal study

Article References: Massucco, S., Leandri, M., Bruzzone, V., Nobbio, L., Hamedani, M., Stara, S., Bellone, E., Geroldi, A., Venturi, C., Gemelli, C., Schenone, A., Marinelli, L., & Grandis, M. (2026). Distinct small-fiber dysfunction profiles in CMT1A and RFC1 disease: a multimodal study. Journal of Neurology, 273(10), Article 602. https://doi.org/10.1007/s00415-026-14145-w

Image Credits: AI Generated

DOI: 10.1007/s00415-026-14145-w

Keywords: CMT1A, RFC1 disease, CANVAS, small fiber neuropathy, peripheral neuropathy, neuropathic pain, autonomic dysfunction, skin biopsy, nociceptive evoked potentials, electrochemical skin conductance, PMP22, sensory neuronopathy