The link between symptoms and long-term outcomes is critical for regulators and payers, but difficult to substantiate and even more so for irare, slowly progressing and chronic illnesses like Sporadic Inclusion Body Myositis (sIBM).
In this project, an exploratory efficacy-to-effectiveness bridging model was developed to predict premature mortality from patients characteristics and symptoms dynamics, using sporadic inclusion body myositis (sIBM) as an illustrative case study. sIBM is a slowly progressive, disabling myopathy which phenotypically exclusively affects the voluntary (skeletal) muscles.
Objectives
To quantify the predictive potential of shorter-term clinical markers on long-term outcomes (excess mortality) in sIBM patients, so to guide study designs and drugs evaluations
Establishing the critical link between symptoms and for a number of rare, slowly progressing and chronic illnesses like sIBM.
Assessing the predictive potential of disease-related characteristics on long-term outcomes, for regulators, payers and prescribers.
Methods & solution
Building a disease model to predict long-term survival from functional short-term outcomes
Candidate variables that may be potentially associated with premature mortality were identified by disease experts and from the sIBM literature.
↓
Based on clinical and economic RWD from 7 countries, a disease model was built to predict long-term survival from functional outcomes via Bayesian survival modelling.
↓
For external validation, model predictions were compared to published mortality data in sIBM patient cohorts.
↓
The final model was used to simulate the increased risk of premature death in sIBM patients.
Results
sIBM patients have higher risk of premature death than non-sIBM subjects with similar characteristics
Presence of dysphagia, aspiration pneumonia, falls, being wheelchair-bound and 6-min walking distance were identified as candidate variables to be used as predictors for premature mortality.
There was limited correlation between these functional performance measures, which were therefore treated as independent variables in the model.
Based on a Bayesian survival model, presence of dysphagia and decrease in 6MWD (6mn walking distance) were associated with poorer survival.
sIBM patients were found to have higher risk of premature death than age/gender matched general population.
Characterization of the burden of sIBM, to design and defend a clinical program design with related endpoints and ascertain medical practice regarding monitoring of sIBM patients.
Impact
Was used to ascertain choices of trial endpoints as surrogates for long term outcomes to be used by payers
- In the absence of long-term data, bridging modelling generated survival predictions by combining relevant information.
- The methodological principle is applicable to any progressive chronic and rare diseases.
- Studies with lifetime follow-up would be needed to further validate the modelling results.
Reference: Capkun et al 2017 “Evidence-based Bayesian modeling to predict survival in patients with sIBM”, Value in Health – https://hal-pasteur.archives-ouvertes.fr/pasteur-02612693/document