Describing and characterising variability in ALS disease progression

Predictive Analysis of Amyotrophic Lateral Sclerosis Progression and Mortality in a Clinic Cohort From Singapore

INTRODUCTION

There is currently no comprehensive Amyotrophic Lateral Sclerosis (ALS) patient database in Singapore comparable to those available in Europe and the United States. We established the Singapore ALS registry (SingALS) to draw meaningful inferences about the ALS population in Singapore through developing statistical and machine learning-based predictive models.

METHODS

The SingALS registry was established through the retrospective collection of demographic, clinical, and laboratory data from 72 ALS patients at Tan Tock Seng Hospital (TTSH) and combining it with demographic and clinical data from 71 patients at Singapore General Hospital (SGH). The SingALS was compared against international ALS registries. Using comparative studies including survival and temporal feature analysis, we identified key factors influencing ALS survival and developed a machine learning model to predict survival outcomes.

RESULTS

Compared to Caucasian-dominant registries, such as the German Swabia registry, SingALS patients had longer average survival (50.51 vs. 31.0 months), younger age of onset (56.18 vs. 66.6 years), and lower bulbar onset prevalence (20.98% vs. 34.10%). Singaporean males had poorer outcomes compared to females, with a hazard ratio (HR) of 3.12 (p = 0.008). Patients who died within 24 months had an earlier need for being bedbound (p < 0.004), percutaneous endoscopic gastrostomy (PEG) insertion (p = 0.004) and non-invasive ventilation (NIV) (p < 0.001). Machine learning and statistical analysis indicated that a steeper ALSFRS-R slope, higher alkaline phosphatase (ALP), white blood cell (WBC), absolute neutrophil counts, and creatinine levels are associated with worse mortality.

DISCUSSION

We developed a comprehensive Singaporean ALS registry and identified key factors influencing survival.

A neuromuscular clinician's primer on machine learning

Artificial intelligence is the future of clinical practice and is increasingly utilized in medical management and clinical research. The release of ChatGPT3 in 2022 brought generative AI to the headlines and rekindled public interest in software agents that would complete repetitive tasks and save time. Artificial intelligence/machine learning underlies applications and devices which are assisting clinicians in the diagnosis, monitoring, formulation of prognosis, and treatment of patients with a spectrum of neuromuscular diseases. However, these applications have remained in the research sphere, and neurologists as a specialty are running the risk of falling behind other clinical specialties which are quicker to embrace these new technologies. While there are many comprehensive reviews on the use of artificial intelligence/machine learning in medicine, our aim is to provide a simple and practical primer to educate clinicians on the basics of machine learning. This will help clinicians specializing in neuromuscular and electrodiagnostic medicine to understand machine learning applications in nerve and muscle ultrasound, MRI imaging, electrical impendence myography, nerve conductions and electromyography and clinical cohort studies, and the limitations, pitfalls, regulatory and ethical concerns, and future directions. The question is not whether artificial intelligence/machine learning will change clinical practice, but when and how. How future neurologists will look back upon this period of transition will be determined not by how much changed or by how fast clinicians embraced this change but by how much patient outcomes were improved.

Self-assessment of amyotrophic lateral sclerosis functional rating scale on the patient's smartphone proves to be non-inferior to clinic data capture

OBJECTIVE

To investigate self-assessment of the amyotrophic lateral sclerosis functional rating scale-revised (ALSFRS-R) using the patient's smartphone and to analyze non-inferiority to clinic assessment.

METHODS

In an observational study, ALSFRS-R data being remotely collected on a mobile application (App-ALSFRS-R) were compared to ALSFRS-R captured during clinic visits (clinic-ALSFRS-R). ALS progression rate (ALSPR)-as calculated by the monthly decline of ALSFRS-R-and its intrasubject variability (ALSPR-ISV) between ratings were used to compare both cohorts. To investigate non-inferiority of App-ALSFRS-R data, a non-inferiority margin was determined.

RESULTS

A total of 691 ALS patients using the ALS-App and 1895 patients with clinic assessments were included. Clinical characteristics for the App-ALSFRS-R and clinic-ALSFRS-R cohorts were as follows: Mean age 60.45 (SD 10.43) and 63.69 (SD 11.30) years (p < 0.001), disease duration 38.7 (SD 37.68) and 56.75 (SD 54.34) months (p < 0.001) and ALSPR 0.72 and 0.59 (p < 0.001), respectively. A paired sample analysis of ALSPR-ISV was applicable for 398 patients with clinic as well as app assessments and did not show a significant difference (IQR 0.12 [CI 0.11, 0.14] vs 0.12 [CI 0.11, 0.14], p = 0.24; Cohen's d = 0.06). CI of IQR for App-ALSFRS-R was below the predefined non-inferiority margin of 0.15 IQR, demonstrating non-inferiority.

CONCLUSIONS

Patients using a mobile application for remote digital self-assessment of the ALSFRS-R revealed younger age, earlier disease course, and faster ALS progression. The finding of non-inferiority of App-ALSFRS-R assessments underscores, that data collection using the ALS-App on the patient's smartphone can serve as additional source of ALSFRS-R in ALS research and clinical practice.

Predicting amyotrophic lateral sclerosis (ALS) progression with machine learning

OBJECTIVE

To predict ALS progression with varying observation and prediction window lengths, using machine learning (ML).

METHODS

We used demographic, clinical, and laboratory parameters from 5030 patients in the Pooled Resource Open-Access ALS Clinical Trials (PRO-ACT) database to model ALS disease progression as fast (at least 1.5 points decline in ALS Functional Rating Scale-Revised (ALSFRS-R) per month) or non-fast, using Extreme Gradient Boosting (XGBoost) and Bayesian Long Short Term Memory (BLSTM). XGBoost identified predictors of progression while BLSTM provided a confidence level for each prediction.

RESULTS

ML models achieved area under receiver-operating-characteristics curve (AUROC) of 0.570-0.748 and were non-inferior to clinician assessments. Performance was similar with observation lengths of a single visit, 3, 6, or 12 months and on a holdout validation dataset, but was better for longer prediction lengths. 21 important predictors were identified, with the top 3 being days since disease onset, past ALSFRS-R and forced vital capacity. Nonstandard predictors included phosphorus, chloride and albumin. BLSTM demonstrated higher performance for the samples about which it was most confident. Patient screening by models may reduce hypothetical Phase II/III clinical trial sizes by 18.3%.

CONCLUSION

Similar accuracies across ML models using different observation lengths suggest that a clinical trial observation period could be shortened to a single visit and clinical trial sizes reduced. Confidence levels provided by BLSTM gave additional information on the trustworthiness of predictions, which could aid decision-making. The identified predictors of ALS progression are potential biomarkers and therapeutic targets for further research.