Amyotrophic lateral sclerosis (ALS) engenders adult-onset, progressive degeneration of both upper and lower motor neurons, leading to muscle weakness, paralysis, and death usually within 3–5 years. No effective treatment for ALS is currently available, despite the limited efficacy of riluzole [1] and edaravone [2]. Increasing numbers of genes responsible for ALS have been identified in recent years; among those, the gene coding the Cu/Zn superoxide dismutase (SOD1) is the first ALS-causative gene found in 1993, and the second most frequent genetic cause for ALS after C9orf72 in patients with European ancestry, while it is the most frequently mutated gene in Asian populations [3]. To date, multiple therapeutic approaches have targeted SOD1-related ALS. The antisense oligonucleotide tofersen showed promising results in a recent phase I/II trial [4], significantly reducing the SOD1 concentration in the cerebrospinal fluid. Given the clinical heterogeneity among different SOD1 mutations, it is important to clinically characterize distinct mutations in order to evaluate the efficacy of SOD1-targeting therapies. Moreover, the increasing evidence of ethnicity-based epidemiologic and genetic differences motivated us to describe and compare disease-causing mutations and related phenotypes between Asians and Caucasians.
In this study, we enrolled ALS patients from two prospectively-established cohorts from China[5] and Germany[6]. Only genetic-confirmed ALS patients with SOD1 mutations were included in further analysis and comparison of natural history. Demographic and clinical data were collected at enrollment and updated at following-ups every 3–6 months. Detailed methods are provided in Supplementary Methods.
We identified a total of 69 distinct SOD1 known mutations (61), variants of uncertain significance (5), and likely pathogenic variants (3). The most frequent mutation present in both populations was p.His47Arg (8 Chinese and 2 German). Other common mutations seen in both populations included p.His44Arg, p.Leu85Phe, p.Glu41Gly, p.Asn87Ser, p.Ile114Thr and p.Gly148Asp. All common mutations featured consistent phenotypes, including an aggressive form of ALS in p.Gly148Asp and slow progression forms in p.Glu41Gly, p.His47Arg and p.Asn87Ser (Supplementary Table S1). The majority of mutations in Chinese patients were located in exon 2 while those in German patients in exon 4. There was a significant difference of average age of onset between Chinese and German patients who carried mutations in exon 4 (37.4 vs 49.9 years, p < 0.001). However, diagnostic delay and survival were not significantly different among the exons (Supplementary Figure S1).
Among 66 Chinese and 84 German patients with SOD1 mutations, the mean (95% CI) age of onset was 47.2 (45.2–49.1) years (Chinese: 43.9 [41.6–46.2] vs German: 49.9 [47.0-52.8], p = 0.002), and the proportion of young-onset ALS, defined as onset between 25 and 45 years, was 47.5% (62.5% Chinese and 30.7% German, p < 0.001). The median (IQR) diagnostic delay was 12.0 (6.0–35.0) months (Chinese: 14.5 [6.0-36.5] vs German: 11.0 [6.0–32.0], p = 0.59), and the median (IQR) survival was 141.0 (21.0-364.0) months (Chinese: cannot be calculated because of more than 60% censored cases at censoring date March 31, 2019; German: 198.0 [22.0-364.0]; p = 0.90). Despite lower riluzole prescription rates in the Chinese population (28.3%) compared to the German population (81.3%, p < 0.001), no survival differences determined by log-rank test were observed between both populations (p = 0.90, Table 1).
The majority of cases (94.3%) had spinal onset, while there was no proportional difference of site of onset between Chinese and German patients or between male and female patients. 20% of patients were characterized with a pure LMN phenotype (Chinese: 17.3%, German: 25.0%, p = 0.41; Table 1). Some of these patients progressed slowly (p.His47Arg), while others showed an aggressive pattern (p.Ala5Val and p.His44Arg). Another notable difference was that male patients featured a significantly shorter diagnostic delay (p = 0.01) and survival (p = 0.005) compared to females (Supplementary Figure S2).
Results of uni- and multivariate Cox regression analysis are summarized in Supplementary Table S2 and Supplementary Figure S3. Because of the significant relationship between diagnostic delay, early progression rate, and late progression rate, only sex, age of onset (HR 1.02, p = 0.054), site of onset, and late progression rate were included in the multivariate Cox regression analysis, which revealed that patients with bulbar onset (HR 10.31, p = 0.01) and higher late progression rate (HR 2.42, p = 0.003) had a much shorter survival time.
The present study has three major genetic and clinical implications. First, to our best knowledge, this study constitutes the first direct comparison of mutational and clinical characteristics of Asian and European ALS patients with SOD1 mutations based on two large hospital-based registry cohorts. Distinct mutational distribution between Chinese (mainly in exon 2) and German patients (mainly in exon 4) displayed the relative consistent phenotype regarding the identical common mutations and no survival difference among the five exons. The dominate mutation p.Ala5Val in North America [7] was rare in China and absent in Germany.
Second, our findings demonstrate a significant lower age of onset and higher proportion of young-onset cases in the Chinese SOD1-mutant patients compared with their German counterparts, possibly reflecting a higher burden of genetic and environmental risk factors[8]. A younger age of onset featured in SOD1-mutant patients in each cohort compared to the overall ALS patients in each ethnic population, while at the same time age of onset also displayed ethnicity-specific features known from sporadic ALS, i.e., younger age of onset in Chinese compared to German patients.
Third, we have confirmed several previous known prognostic factors in this SOD1-ALS cohort, i.e., site of onset and disease progression rate (especially late progression rate). Of note, some known prognostic factors in sporadic ALS population, including BMI and age of onset, were not associated with survival, indicating that the spectrum of disease-modifying factors may be different from sporadic ALS. SOD1-G93A murine models of different genetic backgrounds have also been found to feature significantly different disease progression rates and survival, highlighting a set of potential key genes and molecular pathways acting as disease modifiers and possibly contributing to the heterogeneity of ALS phenotypes[9].
We acknowledge that the cognitive status and potential modified genes were not included and analyzed in this study, despite that few cases carrying SOD1 mutations have been reported to feature cognitive dysfunction[10] and that it is not common in SOD1-mutant patients to carry another causative gene mutation. On the other hand, we regard the existence of large synchronized databases and that the patients were derived from specialized ALS centers in both countries as main strengths of our study. In the future, it will be necessary to establish the detailed genotype-phenotype database of SOD1 mutation in different populations in order to clarify the underlying mechanisms and develop precise strategy for SOD1-targeted clinical trials for ALS.