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.

Describing and characterising variability in ALS disease progression

BACKGROUND, OBJECTIVES

Decrease in the revised ALS Functional Rating Scale (ALSFRS-R) score is currently the most widely used measure of disease progression. However, it does not sufficiently encompass the heterogeneity of ALS. We describe a measure of variability in ALSFRS-R scores and demonstrate its utility in disease characterization.

METHODS

We used 5030 ALS clinical trial patients from the Pooled Resource Open-Access ALS Clinical Trials database to calculate variability in disease progression employing a novel measure and correlated variability with disease span. We characterized the more and less variable populations and designed a machine learning model that used clinical, laboratory and demographic data to predict class of variability. The model was validated with a holdout clinical trial dataset of 84 ALS patients (NCT00818389).

RESULTS

Greater variability in disease progression was indicative of longer disease span on the patient-level. The machine learning model was able to predict class of variability with accuracy of 60.1-72.7% across different time periods and yielded a set of predictors based on clinical, laboratory and demographic data. A reduced set of 16 predictors and the holdout dataset yielded similar accuracy.

DISCUSSION

This measure of variability is a significant determinant of disease span for fast-progressing patients. The predictors identified may shed light on pathophysiology of variability, with greater variability in fast-progressing patients possibly indicative of greater compensatory reinnervation and longer disease span. Increasing variability alongside decreasing rate of disease progression could be a future aim of trials for faster-progressing patients.

Identification of potential pathways and biomarkers linked to progression in ALS

OBJECTIVE

To identify potential diagnostic and prognostic biomarkers for clinical management and clinical trials in amyotrophic lateral sclerosis.

METHODS

We analysed proteomics data of ALS patient-induced pluripotent stem cell-derived motor neurons available through the AnswerALS consortium. After stratifying patients using clinical ALSFRS-R and ALS-CBS scales, we identified differentially expressed proteins indicative of ALS disease severity and progression rate as candidate ALS-related and prognostic biomarkers. Pathway analysis for identified proteins was performed using STITCH. Protein sets were correlated with the effects of drugs using the Connectivity Map tool to identify compounds likely to affect similar pathways. RNAi screening was performed in a Drosophila TDP-43 ALS model to validate pathological relevance. A statistical classification machine learning model was constructed using ridge regression that uses proteomics data to differentiate ALS patients from controls.

RESULTS

We identified 76, 21, 71 and 1 candidate ALS-related biomarkers and 22, 41, 27 and 64 candidate prognostic biomarkers from patients stratified by ALSFRS-R baseline, ALSFRS-R progression slope, ALS-CBS baseline and ALS-CBS progression slope, respectively. Nineteen proteins enhanced or suppressed pathogenic eye phenotypes in the ALS fly model. Nutraceuticals, dopamine pathway modulators, statins, anti-inflammatories and antimicrobials were predicted starting points for drug repurposing using the connectivity map tool. Ten diagnostic biomarker proteins were predicted by machine learning to identify ALS patients with high accuracy and sensitivity.

INTERPRETATION

This study showcases the powerful approach of iPSC-motor neuron proteomics combined with machine learning and biological confirmation in the prediction of novel mechanisms and diagnostic and predictive biomarkers in ALS.

Prospective Cohort Study of Nusinersen Treatment in Adults with Spinal Muscular Atrophy

BACKGROUND

The impact of nusinersen therapy on outcomes in adults with Spinal Muscular Atrophy (SMA) remains uncertain.

OBJECTIVE

To demonstrate whether nusinersen therapy, at currently prescribed doses, can stabilize or improve motor function in adults with SMA using existing outcome measures.

METHODS

A single-center prospective cohort study of 6 adults with SMA type 3, with inclusion/exclusion criteria intended to optimize the ability to demonstrate change using established outcome measures. Primary outcomes were the Hammersmith Functional Motor Scale-Expanded (HFMSE) and the Revised Upper Limb Measure (RULM). Secondary outcomes were the PedsQL Fatigue scale, the SMA Functional Rating Scale (SMAFRS), and the 6-minute and 10-meter walk tests (6 MWT and 10 MWT). Estimates of change in HFMSE and RULM mean scores across visits were calculated using a linear mixed effects model. Change from baseline was used for other outcome measures.

RESULTS

HFMSE and RULM scores over 12 months were stable or improved in all participants, with a mean increase of 2 points in each. Other measures showed high intra-individual variability. Adverse events related to the primary diagnosis, including injury and infection, significantly impacted the ability to reliably perform walk tests in the four ambulatory participants.

CONCLUSIONS

HFMSE and RULM show potential as responsive outcome measures of motor function in ambulatory and non-ambulatory adults with SMA type 3. A time-dependent accrual of benefit of nusinersen on motor function was apparent in this cohort. More sensitive alternative measures of quality of life, fatigue, exercise tolerance, stability and ADLs are clearly needed for adults with SMA.

A Patient with Limbic Encephalitis, Ear Perichondritis, and Episcleritis - An Unusual Presentation of Relapsing Polychondritis

Relapsing polychondritis (RPC) is a rare autoimmune disease that is characterized by recurrent inflammation and destruction of cartilaginous tissues. Limbic encephalitis is a rare central nervous system manifestation of RPC that has been mentioned in case reports. Recognition of this association, and reliable methods of diagnosis, including the utility of neuroimaging modalities such as positron emission tomography/computed tomography (PET/CT) can be useful in the evaluation of this diagnostic challenge. We report a patient with limbic encephalitis associated with RPC, where PET/CT was effectively used in the diagnosis, and monitoring of response to treatment. We also demonstrate that it can be a useful modality in certain situations when brain magnetic resonance imaging cannot be done.

Overturning the Paradigm of Spinal Muscular Atrophy as Just a Motor Neuron Disease

Spinal muscular atrophy is typically characterized as a motor neuron disease. Untreated patients with the most severe form, spinal muscular atrophy type 1, die early with infantile-onset progressive skeletal, bulbar, and respiratory muscle weakness. Such patients are now living longer due to new disease-modifying treatments such as gene replacement therapy (onasemnogene abeparvovec), recently approved by the US Food and Drug Administration, and nusinersen, a central nervous system-directed treatment which was approved by the US Food and Drug Administration three years ago. This has created an area of pressing clinical need: if spinal muscular atrophy is a multisystem disease, dysfunction of peripheral tissues and organs may become significant comorbidities as these patients survive into childhood and adulthood. In this review, we have compiled autopsy data, case reports, and cohort studies of peripheral tissue involvement in patients and animal models with spinal muscular atrophy. We have also evaluated preclinical studies addressing the question of whether peripheral expression of survival motor neuron is necessary and/or sufficient for motor neuron function and survival. Indeed, spinal muscular atrophy patient data suggest that spinal muscular atrophy is a multisystem disease with dysfunction in skeletal muscle, heart, kidney, liver, pancreas, spleen, bone, connective tissues, and immune systems. The peripheral requirement of SMN in each organ and how these contribute to motor neuron function and survival remains to be answered. A systemic (peripheral and central nervous system) approach to therapy during early development is most likely to effectively maximize positive clinical outcome.

Trainee Responses to Hurricane Harvey: Correlating Volunteerism With Burnout

Background: Natural disasters take a heavy toll not only on their victims, but also on physicians who suffer vicarious trauma and burnout. New trainees in Houston, from entering PGY1 residents to entering fellows, underwent even more upheaval and stress during Hurricane Harvey. Many responded to calls for volunteer help. Objective: To investigate the impact of Hurricane Harvey on new trainees at our institution, and correlate volunteerism with measures of burnout and resilience. Methodology: Thirty three new trainees out of 90 (43% of population) from all specialties in our institution voluntarily responded to an online survey on the impact of Hurricane Harvey on their lives, whether or not they volunteered and in what form, and answered questions drawing from the abbreviated Maslach burnout survey and Resiliency Quiz. Statistical analyses were conducted using GraphPad Prism and Excel data analysis. Results: The top areas impacted were emotional health (32%), eating habits (29%), family (25%) and finances (25%). The main voluntary activities were covering for colleagues who could not make it to hospital (50%), donating money and supplies (36%), and cleaning and rebuilding (36%). Volunteering was associated with feelings of appreciation (76%), happiness (62%), thankfulness (57%), purposefulness (43%) and pride (33%). Fewer volunteers scored lowly in personal achievement as compared to non-volunteers (10 vs. 38%, p = 0.05). Significance: Hurricane Harvey affected health, finances and family of new trainees, more than half of whom volunteered to help. Volunteers had a greater sense of personal achievement as compared to non-volunteers. This may be due to having more volunteers among less burnt-out trainees or because volunteering reduced burnout and stress responses/trauma. These results suggest that volunteer opportunities should be made available in programs targeting resident burnout.

Advanced neuroimaging in Balo's concentric sclerosis: MRI, MRS, DTI, and ASL perfusion imaging over 1 year

Balo concentric sclerosis (BCS) is a rare, atypical demyelinating disease, which may rapidly progress to become severe and fatal. Advanced neuroimaging has proven helpful for early diagnosis, classification, prognostication, and monitoring of progression in multiple sclerosis, but has not been fully explored in BCS. We present the case of a 27-year-old woman with BCS in whom advanced neuroimaging was used to correlate the evolution of disease with clinical findings over the course of 1 year. Magnetic resonance imaging, magnetic resonance spectroscopy (MRS), diffusion tensor imaging (DTI), and arterial spin labeling cerebral perfusion were obtained at presentation (Day 0), and at Day 67 and Day 252. Imaging features include multilayered concentric ring lesion, reduced diffusion along the rim, hypoperfusion with possible mild central hyperperfusion, and MRS findings of increased choline, decreased N-acetylaspartate (NAA), and possible presence of lactate and/or lipid peak. DTI tractography and relative apparent diffusion coefficient analyses correlated with clinical symptoms and may help to determine extent of white matter tract injury and prognosis.

Discussing edaravone with the ALS patient: an ethical framework from a U.S. perspective

The recent approval of edaravone by the United States Food and Drug Administration has generated a mix of hope tempered by reality. The costs of the drug, both monetarily and with regard to intensity of treatment, are high. The benefits, while modest, will be viewed through a very different lens by individuals depending on their goals of care. By virtue of our training and experience, physicians are ideally suited to understand and explain new treatments to our patients. As healthcare providers with a fiduciary responsibility to our patients, we must make sure they are fully informed about both the costs and benefits of non-curative therapies such as edaravone, and be prepared to discuss these in the context of their goals of care and potential impact on quality of life. Respect for our patients' autonomy is critical when discussing these issues, but we should always be guided by the ethical principles of beneficence and non-maleficence.

Recurrent Cerebral Hemorrhage in Normal Pregnancy Secondary to Mycotic Pseudoaneurysms Related to Choriocarcinoma

BACKGROUND

Choriocarcinoma coexisting with or after normal pregnancy is extremely rare. To our knowledge, our case report is the first time cerebral mycotic pseudoaneurysms from choriocarcinoma have been proven angiographically.

CASE DESCRIPTION

A 38-week pregnant 26-year-old woman presented with an acute left frontal hemorrhage. She underwent emergency cesarean section, followed by hematoma evacuation and resection of what grossly appeared to be a medium-sized arteriovenous malformation at the time of surgery. Angiogram before and after resection showed no obvious vascular pathology. One month later, she returned with status epilepticus, and an acute parenchymal hematoma posterior to the surgical resection cavity was identified. Angiography showed a multilobulated pseudoaneurysm along the left middle cerebral artery. This was resected and found on histopathology to have choriocarcinoma within and around the blood vessels. Serum human chorionic gonadotrophin levels increased daily. Pan computed tomography showed a left lung lobular mass. The diagnosis was stage 4 World Health Organization score 9 high-risk metastatic choriocarcinoma requiring radiation followed by multiagent chemotherapy. Two weeks later, she had another seizure. An angiogram showed an unruptured pseudoaneurysm along the right posterior cerebral artery, which was embolized.

CONCLUSIONS

Metastatic choriocarcinoma is rarely considered during a viable pregnancy but is almost always fatal if unrecognized. Early recognition enhances the chances of cure with chemotherapy. Arteriovenous malformations are typically considered in young women with intracerebral hemorrhages and have higher risk of rupture in pregnant women, but physicians should also be aware of metastatic choriocarcinoma and the development of mycotic aneurysms in peripartum women with intracerebral hemorrhages.