(M1330-11-62) ILDsym^®, a Quantitative Systems Pharmacology (QSP) Platform, Successfully Simulates the Pathophysiology of Systemic Sclerosis-Interstitial Lung Disease (SSc-ILD) and Inter-patient Variability

Purpose: Systemic sclerosis (SSc) is a rare connective tissue and autoimmune disease associated with inflammation of the skin and internal organs. Interstitial lung disease (ILD), a frequent complication of SSc, is associated with increased risk of morbidity and mortality¹. The course of SSc-ILD is highly variable; some patients remain stable while others progress rapidly. The two current FDA-approved treatments for SSc-ILD patients, nintedanib and tocilizumab (TCZ), demonstrate efficacy but do not reverse disease². Mechanistic, mathematical modeling approaches can support the development of new drug treatments by improving understanding of disease pathophysiology, identifying mechanistic drivers of SSc-ILD, interpreting clinical treatment results, and assessing efficacy for novel treatments as monotherapies or in combination with existing treatments.

Methods: ILDsym, a QSP model of SSc-ILD, was developed based on publicly available data. Applications combine predicted drug exposure in the lung with drug pharmacodynamic (PD) properties to modulate SSc-ILD pathophysiology and predict efficacy of novel therapeutics. To facilitate evaluating the effects of inter‑individual patient variability on predicted treatment response, a simulated population (SimPops™) of SSc-ILD patients was generated reflecting variability in disease pathophysiology. Patients were screened by assessing simultaneous consistency with multiple clinical data sets, including lung function, disease extent, cell populations, biomarkers, and standards of care

Results: ILDsym dynamically captures changes to lung architecture, susceptibility to inflammation and/or fibrosis, and several key clinical outcome measures such as forced vital capacity (FVC) and diffusing capacity of the lung for carbon monoxide (DLCO). Over 700 simulated patients were generated to represent SSc-ILD patients at various stages of disease (baseline %FVC and %DLCO values of 74.3% ± 12.2% and 54.8% ± 19.1%, respectively (Figure 1)), and the simulated patients span appropriate ranges of alveolar epithelial cells, endothelial cells, macrophages, and myofibroblasts, mediators and biomarkers in accordance with published data for SSc-ILD patients; a selection of these validation results are shown in Figure 2. Over a one-year simulation, lung function remains stable in some patients and declines in others, with average FVC losses of -0.142 ± 0.112 L/year. Patients were sub-selected (i.e., SimCohorts generation) to match the baseline lung function characteristics of patients in phase 3 trials for nintedanib³. Simulations of compound exposure and PD effects in SimCohorts resulted in lesser lung function decline and stabilized lung function, respectively, consistent with nintedanib clinical data (Figure 3).

Conclusion: ILDsym is well-positioned to support SSc-ILD drug development by providing a dynamic representation of SSc-ILD disease progression. The flexible framework of ILDsym enables its use in exploring novel targets and therapies as needed. ILD SimPops capture expected clinical inter-patient variability and can be utilized to optimize clinical trial protocols, explore mechanistic hypotheses, and evaluate combination therapies in addition to monotherapies.

References: 1. Asano. J Clin Med. 2020; 9(9): 2687-2714.
2. Khanna. Arthritis Rheumatol. 2022; 74(1): 13-27.
3. Distler. NEJM. 2019; 380(26): 2518-28.
4. Mathai. Lab Invest. 2010; 90 (6): 812-23.
5. Dantas. Disease Markers. 2016.
6. Dzaidzio. Rheum. 2005; 44 (12) 1518-24.
7. Kuznik-Trocha. BioMed Res Int. 2020.
8. Hoffman-Vold. Respir Res. 2020; 21 (1): 197.
9. Hoffman-Vold. Respir Crit Care. 2019; 200 (10): 1258-66.
10. Young-Min. Ann Rheum Dis. 2001; 60 (9): 846-51.
11. Montagana. Clin Biochem. 2007; 40 (12): 837-42.
12. Gourh. Arth Res Therapy. 2009; 11 (5): R147.
13. De Lauretis. J. Rheum. 2013; 40 (4): 435-46.
14. Sato. J Derm Science. 2001. 27 (2): 140-46.
15. Clark. Arth Rheum. 2019; 71.
16. Khan. Ann Rheum Disease. 2012: 71 (7): 1235-42.
17. Matsushita. J Rheum. 2006;33 (2): 275-84.

Acknowledgements:All authors are employees of Simulations Plus.


Figure 1: Comparison of SimPops virtual patient clinical output individual values with corresponding clinical data for FVC and DLCO.


Figure 2: Comparison of selected SimPops virtual patient mediator individual values with clinical data


Figure 3: Comparison of simulated untreated and Nintedanib-treated cohorts (n=444) with respective clinical trial groups (n=288)