This was a single-center, retrospective analysis of medical information obtained in Japan aiming to investigate the correlation between the baseline magnitude of muscle impairment and the change in gross motor function after initiation of nusinersen treatment.
Subjects
The medical records of the patients with SMA who visited the department of Child Neurology, National Center of Neurology and Psychiatry (NCNP), Tokyo, Japan between 2000 to November 2021 were reviewed. This identified a total of 56 patients, of which two had deceased. Of the 54 patients, 16 individuals received nusinersen treatment in the department of Child Neurology, NCNP and all of them were included in the study. The clinical records were retrospectively obtained from the medical records. Following the approved nusinersen regimen for non-infantile SMA in Japan, patients were intrathecally injected with the drug at weeks 0, 4, 12 for loading, and every 6 months thereafter. All the patients received 12mg/time of nusinersen as indicated in the manufacturer’s instruction (https://pins.japic.or.jp/pdf/newPINS/00066991.pdf). The individuals underwent pelvic and thigh MR imaging studies, motor function, and X-ray exams for scoliosis at baseline. X-rays to examine scoliosis were taken either in a supine or sitting position or in a position where it was safe for the patient. Cobb angles were manually measured as shown elsewhere. These tests were followed up every time or occasionally as they received nusinersen.
Muscle MRI scoring
MRI exams were done in 3.0 T MRI (either in Achieva; Philips Healthcare, Best, the Netherlands; or MAGNETOM Verio; Siemens, Erlangen, Germany). For the analysis, we used turbo-spin echo T1-weighted sequences on the axial plane of the bilateral pelvis and thighs in contiguous slices (5 mm thickness, gap 3–11 mm, TR 550–728 ms, and TE 8.4–8.4 ms). Most of the patients except for those who could not remain still during the MRI procedure did not receive any sedation.
Using the Mercuri classification, the examined parts through their whole extension were as follows: muscles of the pelvis (gluteus maximus, gluteus medius, gluteus minimus) and the thighs (vastus lateralis, vastus intermedius, vastus medialis, rectus femoris, gracilis, sartorius, adductor longus, adductor brevis, adductor magnus, semitendinosus, biceps femoris, semimembranosus) [14, 28]. That was to semi-quantitatively evaluate the areas of fatty infiltration of the muscle. Specifically, fatty infiltration of muscles was staged as follows [14]:
Stage 0: normal appearance
Stage 1: scattered small areas of or increased density by MRI
Stage 2a (2.0): numerous discrete areas of increased density < 30% of the volume of the muscle
Stage 2b (2.5): numerous discrete areas of increased density with beginning confluence, 30–60% of the volume of the muscle
Stage 3: washed-out appearance due to confluent areas of increased density with muscle still present at the periphery
Stage 4: end-stage appearance, muscle replaced by increased density connective tissue and fat
To assess the atrophy in a method that has been previously reported, thigh muscles were grouped to anterior (rectus femoris, vastus lateralis, vastus intermedius, vastus medialis), medial (adductor magnus, adductor brevis, adductor longus, sartorius, gracilis), and posterior (semimembranosus, semitendinosus, biceps femoris) compartments [14]. Specifically, muscle atrophy was graded as follows [14]:
Grade 0: normal muscle
Grade 1: peripheral muscle volume loss
Grade 2: <50% of muscle volume loss
Grade 3: >50% of muscle volume loss
Muscle MR images were assessed independently by two evaluators (one pediatric neurologist and one radiologist), who had experiences of examining muscle images of neuromuscular disorders. The summed number of fatty infiltration stage or atrophy grades was considered as the score for the muscle damage (0 being normal and 69 being the most severe), and the average of the scores from two evaluators was adopted for further assessment.
Motor function assessment
The Hammersmith Functional Motor Scale-Expanded (HFMSE), a widely used scale for patients with SMA, was used to assess gross motor function [36]. The HFMSE includes 33 items, with a maximum motor function score equal to 66, whereas the lowest motor function score is 0. The HFMSE was evaluated by physical therapists that were trained to assess patients with neuromuscular disorders. As previous studies have shown that the changes beyond ± 2 points on HFMSE are a clinically meaningful change in SMA types 2 and 3 [17–19], individuals who had HFMSE score change of three or more were considered to have obtained MCID by nusinersen in this study.
Statistical analysis
Analyses were conducted using SPSS Statistics 23 (Armonk, NY: IBM Corp). Testing for normality was done with the Shapiro–Wilk test. The comparison of variables between the individuals with apparent improvement and those who did not after nusinersen treatment was assessed with two tailed Student’s t-test for continuous variables and Fisher’s exact test for categorical variables. The correlation between age-Cobb angle, age-MRI score, MRI score-Cobb angle was evaluated either by Spearman’s rank or Pearson correlation coefficient depending on the normality of the values.
The MRI score (fatty infiltration and atrophy scores) was non-parametric data; therefore, interrater correlation of the scores from two evaluators was assessed by Spearman’s rank correlation coefficient. The comparison of MRI scores of each muscle between the individuals with apparent improvement with nusinersen and those with unclear response was assessed by Mann-Whitney U test. The relationship between baseline muscle MRI score, Cobb angle, age, and HFMSE score change (baseline-15 months of treatment) was modeled using either simple or multiple linear regression analysis.