Longitudinal imaging in C9orf72 mutation carriers: Relationship to phenotype

Abstract

Expansion mutations in the C9orf72 gene may cause amyotrophic lateral sclerosis (ALS), frontotemporal dementia (FTD), or mixtures of the two clinical phenotypes. Different imaging findings have been described for C9orf72-associated diseases in comparison with sporadic patients with the same phenotypes, but it is uncertain whether different phenotypes have a common genotype-associated imaging signature. To address this question, 27 unrelated C9orf72 expansion mutation carriers (C9 +) with varied phenotypes, 28 age-matched healthy controls and 22 patients with sporadic ALS (sALS) underwent 3T MRI scanning and clinical phenotyping. Measures of brain volumes and cortical thickness were extracted from T1 images. Compared to healthy controls and sALS patients, symptomatic C9 + subjects had greater ventricular volume loss and thalamic atrophy for age, with diffuse, patchy cortical thinning. Asymptomatic carriers did not differ from controls. C9 + ALS and ALS-FTD patients had less thinning of the motor cortex than sALS patients, but more thinning in extramotor regions, particularly in frontal and temporal lobes. C9 + ALS patients differed from sporadic ALS patients in the thickness of the superior frontal gyrus and lateral orbitofrontal cortex. Thickness of the precentral gyrus was weakly correlated with the revised ALS functional rating scale. Thickness of many cortical regions, including several frontal and temporal regions, was moderately correlated with letter fluency scores. Letter fluency scores were weakly correlated with ventricular and thalamic volume. To better understand how imaging findings are related to disease progression, nineteen C9 + subjects and 23 healthy controls were scanned approximately 6 months later. Ventricular volume increased in C9 + patients with FTD and ALS-FTD phenotypes and remained stable in asymptomatic C9 + subjects. We conclude that diffuse atrophy is a common underlying feature of disease associated with C9orf72 mutations across its clinical phenotypes. Ventricular enlargement can be measured over a 6-month time frame, and appears to be faster in patients with cognitive impairment.

Keywords: ALS, amyotrophic lateral sclerosis; ALSFRS-R, ALS functional rating scale — revised; ANCOVA, analysis of covariance; ANOVA, analysis of variance; Amyotrophic lateral sclerosis; C9orf72; C9 +,  subjects with C9orf72 expansion mutations; CSF, cerebrospinal fluid; Cortical thickness; DRS-2, Mattis dementia rating scale; DTI, diffusion tensor imaging; Diffusion tensor imaging; FBI, frontobehavioral inventory; FDR, false discovery rate correction; FTD, frontotemporal dementia; Frontotemporal dementia; MRI, magnetic resonance imaging; SD, standard deviation; TIV, total intracranial volume; Ventricular volume; bvFTD, behavioral variant frontotemporal dementia; sALS, sporadic ALS.

Fig. 1

Ventricular volumes expressed as the percent of Total Intracranial Volume (TIV) at baseline (A–C) and longitudinally (D–F). A) At baseline, C9 + subject group had larger ventricles than healthy controls (HC; *p < 0.05). Sporadic ALS (sALS) did not differ from controls. B) Within the C9 + subgroups, C9 + ALS-FTD patients had larger ventricles than asymptomatic C9 + carriers (Asymp) and HC (*p < 0.05). C) Ventricular volume increased with disease duration in C9 + subjects (filled circles, solid lines) and sALS (open squares, dashed line) with a similar slope, but C9 + had larger ventricles, adjusted for age, compared to sALS and HC. D) Change in ventricular volume over 6 months in C9 + subjects. E) Rate of ventricular enlargement per month, calculated from baseline to 6 month-follow-up of 19 C9 + subjects and 23 HC (individual values, line indicates group mean). F) Ventricular volumes all scans C9 + subjects, including longitudinal scans for asymptomatic C9 + (filled circles), C9 + ALS (open circles), C9 + bvFTD (open triangles) and C9 + ALS-FTD (filled squares) plotted by symptom duration.

Fig. 2

Regions of cortical thinning in whole-brain analyses of C9 + subjects compared to other groups are indicated in blue. A) Symptomatic C9 + subjects (n = 20) have diffuse patchy thinning, with broad swaths of thinning in frontal and temporal regions compared to healthy controls (n = 28). B) C9 + patients with ALS or ALS-FTD (n = 17) have patchy regions of thinning with a frontal predominance compared to sporadic ALS patients (n = 22). In contrast, sALS patients have thinner motor cortex than C9 + patients, best seen in this inflated view (yellow-orange areas). Vertex-wise comparison using FreeSurfer. (Fischl and Dale, 2000) The color scale indicates the negative log of the p values (threshold p = 0.05, FDR corrected).