Quantitative magnetic resonance evaluation of the trigeminal nerve in familial dysautonomia

Cranial Nerve Thinning Distinguishes RFC1-Related Disorder from Other Late-Onset Ataxias

BACKGROUND

RFC1-related disorder (RFC1/CANVAS) shares clinical features with other late-onset ataxias, such as spinocerebellar ataxias (SCA) and multiple system atrophy cerebellar type (MSA-C). Thinning of cranial nerves V (CNV) and VIII (CNVIII) has been reported in magnetic resonance imaging (MRI) scans of RFC1/CANVAS, but its specificity remains unclear.

OBJECTIVES

To assess the usefulness of CNV and CNVIII thinning to differentiate RFC1/CANVAS from SCA and MSA-C.

METHODS

Seventeen individuals with RFC1/CANVAS, 57 with SCA (types 2, 3 and 6), 11 with MSA-C and 15 healthy controls were enrolled. The Balanced Fast Field Echo sequence was used for assessment of cranial nerves. Images were reviewed by a neuroradiologist, who classified these nerves as atrophic or normal, and subsequently the CNV was segmented manually by an experienced neurologist. Both assessments were blinded to patient and clinical data. Non-parametric tests were used to assess between-group comparisons.

RESULTS

Atrophy of CNV and CNVIII, both alone and in combination, was significantly more frequent in the RFC1/CANVAS group than in healthy controls and all other ataxia groups. Atrophy of CNV had the highest sensitivity (82%) and combined CNV and CNVIII atrophy had the best specificity (92%) for diagnosing RFC1/CANVAS. In the quantitative analyses, CNV was significantly thinner in the RFC1/CANVAS group relative to all other groups. The cutoff CNV diameter that best identified RFC1/CANVAS was ≤2.2 mm (AUC = 0.91; sensitivity 88.2%, specificity 95.6%).

CONCLUSION

MRI evaluation of CNV and CNVIII using a dedicated sequence is an easy-to-use tool that helps to distinguish RFC1/CANVAS from SCA and MSA-C.

Loss of Elp1 disrupts trigeminal ganglion neurodevelopment in a model of familial dysautonomia

Familial dysautonomia (FD) is a sensory and autonomic neuropathy caused by mutations in elongator complex protein 1 (ELP1). FD patients have small trigeminal nerves and impaired facial pain and temperature perception. These signals are relayed by nociceptive neurons in the trigeminal ganglion, a structure that is composed of both neural crest- and placode-derived cells. Mice lacking Elp1 in neural crest derivatives ('Elp1 CKO') are born with small trigeminal ganglia, suggesting Elp1 is important for trigeminal ganglion development, yet the function of Elp1 in this context is unknown. We demonstrate that Elp1, expressed in both neural crest- and placode-derived neurons, is not required for initial trigeminal ganglion formation. However, Elp1 CKO trigeminal neurons exhibit abnormal axon outgrowth and deficient target innervation. Developing nociceptors expressing the receptor TrkA undergo early apoptosis in Elp1 CKO, while TrkB- and TrkC-expressing neurons are spared, indicating Elp1 supports the target innervation and survival of trigeminal nociceptors. Furthermore, we demonstrate that specific TrkA deficits in the Elp1 CKO trigeminal ganglion reflect the neural crest lineage of most TrkA neurons versus the placodal lineage of most TrkB and TrkC neurons. Altogether, these findings explain defects in cranial gangliogenesis that may lead to loss of facial pain and temperature sensation in FD.