Research Article: Quantitative FLAIR MRI in Amyotrophic Lateral Sclerosis

Date Published: October 1, 2017

Publisher: Association Of University Radiologists

Author(s): Jeremy Fabes, Lucy Matthews, Nicola Filippini, Kevin Talbot, Mark Jenkinson, Martin R. Turner.

http://doi.org/10.1016/j.acra.2017.04.008

Abstract

T2-weighted magnetic resonance imaging (MRI) hyperintensity assessed visually in the corticospinal tract (CST) lacks sensitivity for a diagnosis of amyotrophic lateral sclerosis (ALS). We sought to explore a quantitative approach to fluid-attenuated inversion recovery (FLAIR) MRI intensity across a range of ALS phenotypes.

Thirty-three classical ALS patients, 10 with a flail arm presentation, and six with primary lateral sclerosis underwent MRI at 3 Tesla. Comparisons of quantitative FLAIR intensity in the CST and corpus callosum were made between 21 healthy controls and within patient phenotypic subgroups, some of whom were studied longitudinally.

Mean FLAIR intensity was greater in patient groups. The cerebral peduncle intensity provided the strongest subgroup classification. FLAIR intensity increased longitudinally. The rate of change of FLAIR within CST correlated with rate of decline in executive function and ALS functional rating score.

FLAIR MRI encodes quantifiable information of potential diagnostic, stratification, and monitoring value.

Partial Text

The neurodegenerative disorder amyotrophic lateral sclerosis (ALS) remains predominantly a clinical diagnosis with significant heterogeneity in rate of disability accumulation. Therapeutic trials rely on survival or change in disability accumulation rate as the primary endpoints. Biomarkers are therefore a research priority (1). Research-based magnetic resonance imaging (MRI) techniques, in particular diffusion tensor imaging (DTI), have demonstrated a consistent involvement of the corticospinal tracts (CSTs) and corpus callosum (CC) in ALS across a range of phenotypes (reviewed in Reference (2)).

A total of 49 patients and 21 controls were available for analysis. Twenty-one patients had at least one follow-up scan. Participant details including patient subgroups are shown in Table 1. The control, cross-sectional, and longitudinal demographics were comparable. The patient cohort was older (60.9 ± 11.1 vs 51.1 ± 12.7 years, P = 0.002), and a correction for this was made in the analysis.Table 1(a) Participant Characteristics (Cross-Sectional Group). (b) Participant Characteristics (Longitudinal Group)Table 1(a) Participant Characteristics (Cross-Sectional Group)Controls (n = 21)Cross-Sectional Patients (n = 49)Mean ± SDRangeMean ± SDRangeP†Age (years)*51.1 ± 12.728–7260.9 ± 11.131–830.002Classical ALS——59.3 ± 11.031–770.021“Flail arm” ALS——61.7 ± 12.541–830.042PLS——68.8 ± 6.262–760.003Disease duration (months)*——59 ± 66.45–366—UMN score*——9.5 ± 4.70–25—ALSFRS-R*——33.3 ± 5.818–44—TMT B − A*——36.8 ± 220–86—n%n%PMale1152.4%3163.3%0.433‡Classical ALS——3367.4%—“Flail arm” ALS——1020.4%—PLS——612.2%—Disease Duration (Months ± SD)Classical ALS35.1±29.2“Flail arm” ALS64.7±34.6PLS180.5±114.9(b) Participant Characteristics (Longitudinal Group)Controls (n = 21)Single-Scan Patients (n = 28)Longitudinal Patients (n = 21)Mean ± SDRangeMean ± SDRangeP†Mean ± SDRangeP§Age (years)*51.1 ± 12.728–7260.9 ± 10.631–770.14361.6 ± 12.239–830.954Disease duration (months)*——37.3 ± 39.15–190—91.5 ± 8421–3660.007UMN score*——10.0 ± 5.02–25—8.7 ± 4.70–150.494ALSFRS-R*——33.2 ± 7.218–44—33.3 ± 3.426–370.906TMT B − A*——39.5 ± 22.58–81—36.4 ± 19.910–860.472Number of scans—————3.6 ± 1.12–5—n%n%P‡n%P¶Male1152.4%1967.7%0.3761152.4%0.553Classical ALS——2485.7%—942.9%0.002“Flail arm” ALS——310.7%—733.3%0.076PLS——13.6%—523.8%0.072Disease Duration (Months ± SD)Disease DurationP‖Classical ALS29.8± 25.049.4±36.00.222“Flail arm” ALS46.8± 30.572.3±35.40.665PLS189.9—178.6128.40.980Longitudinal Patient Clinical ProgressionnMean ± SDRangeRate of change of ALSFRS**20−0.379 ±9 0.25−1.07 to −0.11Rate of change of TMT B − A**140.31 ± 1.7−4.09 to 1.95Rate of change of CST FLAIR††211.47 ± 2.78—Rate of change of CC FLAIR††210.58 ± 3.08—ALS, amyotrophic lateral sclerosis; ALSFRS-R, revised Amyotrophic Lateral Sclerosis Functional Rating Scale; CC, corpus callosum; CST, corticospinal tract; FLAIR, fluid-attenuated inversion recovery; PLS, primary lateral sclerosis; SD, standard deviation; TMT, Trail Making Test; UMN, upper motor neuron.ALS vs PLS: P = 0.049.ALS vs Flail arm: P = 0.589.Flail arm vs PLS: P = 0.149.*Values at first scan.†Comparison to control population, independent samples t test.‡Comparison to control population, Fisher’s exact test.§Comparison to cross-sectional patient population that underwent a single scan (n = 28), independent samples t test.¶Comparison to cross-sectional patient population that underwent a single scan (n = 28), Fisher’s exact test.‖Comparison between same disease subgroup to cross-sectional patient population that underwent a single scan (n = 28), independent samples t test.**Individual patient rate of change of clinical score per month of disease.††Population mean rate of change of magnetic resonance imaging intensity per month of disease × 10−3.

This novel quantitative analysis demonstrated that FLAIR intensity is increased in the CST and the CC in ALS patients compared to healthy controls. FLAIR intensity within the cerebral peduncles and genu of the CC most improved the accuracy of the diagnostic discriminant analysis. The small subgroup of PLS patients demonstrated some of the highest FLAIR intensities, but there was no simple relationship with burden of UMN signs. The rate of change of CST intensity in the patients correlated with the rate of change of both physical disability and executive dysfunction.

DTI remains the leading white matter analysis for the exploratory deep phenotyping of established ALS cases, but FLAIR offers a routine clinical sequence frequently used during the diagnostic work-up of those suspected to have ALS that might offer some added value. An important initiative will be to study such routinely acquired clinical images to see if quantitative analysis can be applied meaningfully (22). Particular challenges will be the development of standardized external templates for coregistration, and automation of intensity standardization across the whole brain. Validation will necessitate applying the parameters derived from this classification analysis to an independent cohort of patients to assess diagnostic accuracy. Comparison with subjects with other neurological diseases, rather than healthy controls, is another key aspect of the aspiration of clinical translation of these findings.

 

Source:

http://doi.org/10.1016/j.acra.2017.04.008

 

Leave a Reply

Your email address will not be published.