Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2022 Jul;269(7):3858-3878.
doi: 10.1007/s00415-022-11022-0. Epub 2022 Mar 10.

Predicting functional impairment trajectories in amyotrophic lateral sclerosis: a probabilistic, multifactorial model of disease progression

Erica Tavazzi #  1 , Sebastian Daberdaku #  1 , Alessandro Zandonà #  1 , Rosario Vasta #  2 , Beatrice Nefussy  3 , Christian Lunetta  4 , Gabriele Mora  5 , Jessica Mandrioli  6 , Enrico Grisan  1   7 , Claudia Tarlarini  4 , Andrea Calvo  2 , Cristina Moglia  2 , Vivian Drory #  3 , Marc Gotkine #  8 , Adriano Chiò #  2 , Barbara Di Camillo #  9   10 ; Piemonte, Valle d’Aosta Register for ALS (PARALS), for the Emilia Romagna Registry for ALS (ERRALS)
Collaborators, Affiliations

Predicting functional impairment trajectories in amyotrophic lateral sclerosis: a probabilistic, multifactorial model of disease progression

Erica Tavazzi et al. J Neurol. 2022 Jul.

Abstract

Objective: To employ Artificial Intelligence to model, predict and simulate the amyotrophic lateral sclerosis (ALS) progression over time in terms of variable interactions, functional impairments, and survival.

Methods: We employed demographic and clinical variables, including functional scores and the utilisation of support interventions, of 3940 ALS patients from four Italian and two Israeli registers to develop a new approach based on Dynamic Bayesian Networks (DBNs) that models the ALS evolution over time, in two distinct scenarios of variable availability. The method allows to simulate patients' disease trajectories and predict the probability of functional impairment and survival at different time points.

Results: DBNs explicitly represent the relationships between the variables and the pathways along which they influence the disease progression. Several notable inter-dependencies were identified and validated by comparison with literature. Moreover, the implemented tool allows the assessment of the effect of different markers on the disease course, reproducing the probabilistically expected clinical progressions. The tool shows high concordance in terms of predicted and real prognosis, assessed as time to functional impairments and survival (integral of the AU-ROC in the first 36 months between 0.80-0.93 and 0.84-0.89 for the two scenarios, respectively).

Conclusions: Provided only with measurements commonly collected during the first visit, our models can predict time to the loss of independence in walking, breathing, swallowing, communicating, and survival and it can be used to generate in silico patient cohorts with specific characteristics. Our tool provides a comprehensive framework to support physicians in treatment planning and clinical decision-making.

Keywords: Amyotrophic lateral sclerosis; Artificial intelligence; Clinical trajectories; Dynamic Bayesian Networks; Population model; Prognosis modelling.

PubMed Disclaimer

Conflict of interest statement

Dr. Chiò reports personal fees from Biogen, personal fees from Amylyx, personal fees from Denali, outside the submitted work. Dr. Di Camillo reports grants from The Italian Ministry of Foreign Affairs and International Cooperation, grants from the Ministry of Science, Technology and Space of the State of Israel, during the conduct of the study; grants from Italian Ministry of Health (Ministero della Salute), outside the submitted work. Dr. Drory reports grants from the Israel Ministry of Science, Technology and Space, during the conduct of the study. Dr. Grisan reports grants from The Italian Ministry of Foreign Affairs and International Cooperation, grants from the Ministry of Science, Technology and Space of the State of Israel, during the conduct of the study; personal fees from GE Healthcare, grants from Italian Ministry of Education, University and Research—PRIN, grants from Ministry of Health—Applied Research, outside the submitted work. Dr. Lunetta reports personal fees from Italfarmaco, personal fees from Mitsubishi Tanabe Pharma Europe, personal fees from Cytokinetics, outside the submitted work. Dr. Mandrioli reports grants from Emilia Romagna Regional Health Authority, during the conduct of the study. Dr. Beatrice Nafussy reports grants from the Ministry of Science, Technology and Space of the State of Israel during the conduct of the study. Dr. Zandonà reports grants from The Italian Ministry of Foreign Affairs and International Cooperation, grants from the Ministry of Science, Technology and Space of the State of Israel, during the conduct of the study. The tool presented in this paper was patented on 22 July 2020 as PCT/IT2020/000057: “Method for determining the prognosis of disease progression and survival for patients affected by Amyotrophic Lateral Sclerosis”, Di Camillo B, Zandonà A, Daberdaku S, Tavazzi E, Chiò A, Vasta R, Calvo A, Moglia C, Casale F, D’Ovidio F, Mandrioli J, Lunetta C, Drory V, Mora G, and Gotkine M. No other competing interests declared.

Figures

Fig. 1
Fig. 1
Graph representations of the A ITIS and B IT DBNs, representing the conditional dependencies among the variables over time. The loops on the four MiToS domain variables represent the dependency on the values of the same variable from the previous time-step. The red edges represent the dependencies defined as mandatory in the network learning stage
Fig. 2
Fig. 2
Area Under the time-dependent ROC curve (AU-ROC) for the MiToS impairments and survival on the subjects of the A ITIS and B IT test sets, computed on a 3-month time step up to 96 months since the disease onset. For each clinical outcome, the integral of the AU-ROC (iAU-ROC) computed up to 24, 36, and 96 months is also reported
Fig. 3
Fig. 3
Cumulative probability of impairment in the four MiToS domains and of tracheostomy/death overtime in the A ITIS and B IT test sets (orange line) and in the simulated population (green line: mean values over population; shaded region: standard deviation), based on probabilities modelled by the DBN
Fig. 4
Fig. 4
Density and cumulative probability plots of the times A to MiToS swallowing impairment for the patients with bulbar and spinal onset from the ITIS test set, B to MiToS walking/self-care impairment for the patients from the IT test set with FVC at diagnosis lower than 84%, between 84 and 101%, and higher than 101%, C to MiToS breathing impairment for the patients from the IT test set with FVC at diagnosis lower than 84%, between 84 and 101%, and higher than 101%, and D to MiToS communication impairment for the patients from the ITIS test set with and without walking/self-care impairment at the first visit. Most patients experience the impairment in correspondence with the maximum of the probability density curve (mode). For each patient, we ran 100 different simulations of the disease progression. While the density curves focus for convenience on the first months of the time span (where the distributions were more significant) the cumulative curves are shown until they reach the maximum values of 1
Fig. 5
Fig. 5
Example of single-patient ALS prognosis prediction using the web application we developed on the DBN built on the IT dataset. The figure shows the impairment probability evolution in time (months) in each of the four MiToS domains for two hypothetical patients with very similar characteristics, differing only in the onset site of the disease. Different tabs are available and allow visualisation of the probabilistic predictions of the 4 MiToS impairments and the survival over all the repetitions in terms of cumulative probability, histogram of frequencies, and density plot. The dashboard was implemented using the Shiny framework for R
See this image and copyright information in PMC

References

    1. Talbot K. Motor neuron disease. Pract Neurol. 2009;9(5):303–309. doi: 10.1136/jnnp.2009.188151. - DOI - PubMed
    1. Chio A, Calvo A, Moglia C, Mazzini L, Mora G, PARALS Study Group Phenotypic heterogeneity of amyotrophic lateral sclerosis: a population based study. J Neurol Neurosurg Psychiatry. 2011;82(7):740–746. doi: 10.1136/jnnp.2010.235952. - DOI - PubMed
    1. Al-Chalabi A, Jones A, Troakes C, King A, Al-Sarraj S, van den Berg LH. The genetics and neuropathology of amyotrophic lateral sclerosis. Acta Neuropathol. 2012;124(3):339–352. doi: 10.1007/s00401-012-1022-4. - DOI - PubMed
    1. Al-Chalabi A, Hardiman O, Kiernan MC, Chiò A, Rix-Brooks B, van den Berg LH. Amyotrophic lateral sclerosis: moving towards a new classification system. Lancet Neurol. 2016;15(11):1182–1194. doi: 10.1016/S1474-4422(16)30199-5. - DOI - PubMed
    1. Küffner R, et al. Crowdsourced analysis of clinical trial data to predict amyotrophic lateral sclerosis progression. Nat Biotechnol. 2015;33(1):51–57. doi: 10.1038/nbt.3051. - DOI - PubMed