Neuropathy due to bi-allelic SH3TC2 variants: genotype-phenotype correlation and natural history

Abstract

Recessive SH3TC2 variants cause Charcot-Marie-Tooth disease type 4C (CMT4C). CMT4C is typically a sensorimotor demyelinating polyneuropathy, marked by early onset spinal deformities, but its clinical characteristics and severity are quite variable. Clear relationships between pathogenic variants and the spectrum of disease manifestations are to date lacking. Gene replacement therapy has been shown to ameliorate the phenotype in a mouse model of CMT4C, emphasizing the need for natural history studies to inform clinical trial readiness. Data, including both genetic information and clinical characteristics, were compiled from the longitudinal, prospective dataset of the Inherited Neuropathy Consortium, a member of the Rare Diseases Clinical Research Network (INC-RDCRN). The Charcot Marie Tooth Neuropathy Score (CMTNS), Examination Score (CMTES) and the Rasch-weighted CMTES (CMTES-R) were used to describe symptoms, neurological examinations and neurophysiological characteristics. Standardized response means were calculated at yearly intervals and a mixed model for repeated measures was used to estimate the change in CMTES and CMTES-R over time. Fifty-six individuals (59% female), median age 27 years (range 2-67 years) with homozygous or compound heterozygous variants in SH3TC2 were identified, including 34 unique variants, 14 of which have not previously been published. Twenty-eight participants had longitudinal data available. While there was no significant difference in the CMTES in those with protein truncating versus non-protein truncating variants, there were significant differences in the mean ulnar nerve compound muscle action potential amplitude, the mean radial sensory nerve action potential amplitude, and in the prevalence of scoliosis, suggesting the possibility of a milder phenotype in individuals with one or two non-protein-truncating variants. Overall, the mean value of the CMTES was 13, reflecting moderate clinical severity. There was a high rate of scoliosis (81%), scoliosis surgery (36%), and walking difficulty (94%) among study participants. The CMTES and CMTES-R appeared moderately responsive to change over extended follow-up, demonstrating a standardized response mean of 0.81 standard deviation units or 0.71 standard deviation units, respectively, over 3 years. Our analysis represents the largest cross-sectional and only longitudinal study to date, of the clinical phenotype of both adults and children with CMT4C. With the promise of upcoming genetic treatments, these data will further define the natural history of the disease and inform study design in preparation for clinical trials.

Keywords: CMT4C; Charcot-Marie-Tooth disease; genotype-phenotype; longitudinal; natural history.

Figure 1

The relationship between the CMTES and CMTES-R at the time of study entry and years since reported onset of walking difficulty as an estimate of neurological symptom onset. The CMTES and CMTES-R at the baseline examination correlated with the number of years since the onset of walking difficulty (CMTES: R = 0.41, P < 0.01; CMTES-R: R = 0.44, P < 0.01).

Figure 2

All variants included in this study arranged by cDNA position. The variant 286°C>T (R954*) was seen far more than any other variant, 61 times in this study. The figure has been truncated for legibility purposes. No other variant was seen more than seven times. Gene sequence and functional annotations data obtained from NCBI refseq, MANE isoform. Black = nonsense or out-of-frame variations; orange = splice site mutation; brown = in-frame deletion; green = missense.

Figure 3

Mixed models for repeated measures analyses with categorical time effects, showing change in mean CMTES and CMTES-R at each year after the baseline examination. A significant difference in the mean CMTES and CMTES-R was seen beginning at Year 3 (P = 0.01 and P < 0.01 for CMTES and CMTES-R, respectively). See also Table 5 for all values at each year after the baseline examination.