Unlocking the full potential of mesenchymal stromal cell therapy for osteoarthritis through machine learning-based in silico trials.

Despite the potential of mesenchymal stromal cells (MSCs) in osteoarthritis (OA) treatment, the challenge lies in addressing their therapeutic inconsistency. Clinical trials revealed significantly varied therapeutic outcomes among patients receiving the same allogenic MSCs but different treatment regimens. Therefore, optimizing personalized treatment strategies is crucial to fully unlock MSCs' potential and enhance therapeutic consistency. We employed the XGBoost algorithm to train a self-collected database comprising 37 published clinical reports to create a model capable of predicting the probability of effective pain relief and Western Ontario and McMaster Universities (WOMAC) index improvement in OA patients undergoing MSC therapy. Leveraging this model, extensive in silico simulations were conducted to identify optimal personalized treatment strategies and ideal patient profiles. Our in silico trials predicted that the individually optimized MSC treatment strategies would substantially increase patients' chances of recovery compared to the strategies used in reported clinical trials, thereby potentially benefiting 78.1%, 47.8%, 94.4% and 36.4% of the patients with ineffective short-term pain relief, short-term WOMAC index improvement, long-term pain relief and long-term WOMAC index improvement, respectively. We further recommended guidelines on MSC number, concentration, and the patients' appropriate physical (body mass index, age, etc.) and disease states (Kellgren-Lawrence grade, etc.) for OA treatment. Additionally, we revealed the superior efficacy of MSC in providing short-term pain relief compared to platelet-rich plasma therapy for most OA patients. This study represents the pioneering effort to enhance the efficacy and consistency of MSC therapy through machine learning applied to clinical data. The in silico trial approach holds immense potential for diverse clinical applications.

Lung- and diaphragm-protective strategies in acute respiratory failure: an in silico trial.

BACKGROUND: Lung- and diaphragm-protective (LDP) ventilation may prevent diaphragm atrophy and patient self-inflicted lung injury in acute respiratory failure, but feasibility is uncertain. The objectives of this study were to estimate the proportion of patients achieving LDP targets in different modes of ventilation, and to identify predictors of need for extracorporeal carbon dioxide removal (ECCO2R) to achieve LDP targets. METHODS: An in silico clinical trial was conducted using a previously published mathematical model of patient-ventilator interaction in a simulated patient population (n = 5000) with clinically relevant physiological characteristics. Ventilation and sedation were titrated according to a pre-defined algorithm in pressure support ventilation (PSV) and proportional assist ventilation (PAV+) modes, with or without adjunctive ECCO2R, and using ECCO2R alone (without ventilation or sedation). Random forest modelling was employed to identify patient-level factors associated with achieving targets. RESULTS: After titration, the proportion of patients achieving targets was lower in PAV+ vs. PSV (37% vs. 43%, odds ratio 0.78, 95% CI 0.73-0.85). Adjunctive ECCO2R substantially increased the probability of achieving targets in both PSV and PAV+ (85% vs. 84%). ECCO2R alone without ventilation or sedation achieved LDP targets in 9%. The main determinants of success without ECCO2R were lung compliance, ventilatory ratio, and strong ion difference. In silico trial results corresponded closely with the results obtained in a clinical trial of the LDP titration algorithm (n = 30). CONCLUSIONS: In this in silico trial, many patients required ECCO2R in combination with mechanical ventilation and sedation to achieve LDP targets. ECCO2R increased the probability of achieving LDP targets in patients with intermediate degrees of derangement in elastance and ventilatory ratio.

Constructing bilayer and volumetric atrial models at scale.

To enable large in silico trials and personalized model predictions on clinical timescales, it is imperative that models can be constructed quickly and reproducibly. First, we aimed to overcome the challenges of constructing cardiac models at scale through developing a robust, open-source pipeline for bilayer and volumetric atrial models. Second, we aimed to investigate the effects of fibres, fibrosis and model representation on fibrillatory dynamics. To construct bilayer and volumetric models, we extended our previously developed coordinate system to incorporate transmurality, atrial regions and fibres (rule-based or data driven diffusion tensor magnetic resonance imaging (MRI)). We created a cohort of 1000 biatrial bilayer and volumetric models derived from computed tomography (CT) data, as well as models from MRI, and electroanatomical mapping. Fibrillatory dynamics diverged between bilayer and volumetric simulations across the CT cohort (correlation coefficient for phase singularity maps: left atrial (LA) 0.27 ± 0.19, right atrial (RA) 0.41 ± 0.14). Adding fibrotic remodelling stabilized re-entries and reduced the impact of model type (LA: 0.52 ± 0.20, RA: 0.36 ± 0.18). The choice of fibre field has a small effect on paced activation data (less than 12 ms), but a larger effect on fibrillatory dynamics. Overall, we developed an open-source user-friendly pipeline for generating atrial models from imaging or electroanatomical mapping data enabling in silico clinical trials at scale (https://github.com/pcmlab/atrialmtk).

Credibility assessment of computational models according to ASME V&V40: Application to the Bologna Biomechanical Computed Tomography solution.

BACKGROUND AND OBJECTIVE: When a computational model aims to be adopted beyond research purposes, e.g. to inform a clinical or regulatory decision, trust must be placed in its predictive accuracy. This practically translates into the need to demonstrate its credibility. In fact, prior to its adoption for regulatory purposes, an in silico methodology should be proven credible enough for the scope. This has become especially relevant as, although evidence of the safety and efficacy of new medical products or interventions has been traditionally provided to the regulator experimentally, i.e., in vivo or ex vivo, recently the idea to inform a regulatory decision in silico has made its way in the regulatory scenario. While a harmonised technical standard is currently missing in the EU regulatory system, in 2018 the ASME issued V&V40-2018, where a risk-based framework to assess the credibility of a computational model through the performance of predefined credibility activities is provided. The credibility framework is here applied to Bologna Biomechanical Computed Tomography (BBCT) solution, which predicts the absolute risk of fracture at the femur for a subject. BBCT has recently been the object of a qualification advice request to the European Medicine Agency. METHODS: The full implementation of ASME V&V40-2018 framework on BBCT is shown. Starting from BBCT proposed context of use the whole credibility plan is presented and the credibility activities (Verification, Validation, Applicability) described together with the achieved credibility levels. RESULTS: BBCT risk is judged medium, and the credibility levels achieved considered acceptable. The uncertainties intrinsically present in the material properties assignment affected BBCT predictions to the highest extent. CONCLUSIONS: This work provides the practical application of the ASME V&V40-2018 risk-based credibility assessment framework, which could be applied to demonstrate model credibility in any field and support future regulatory submissions and foster the adoption of In Silico Trials.

Beyond the state of the art of reverse vaccinology: predicting vaccine efficacy with the universal immune system simulator for influenza.

When it was first introduced in 2000, reverse vaccinology was defined as an in silico approach that begins with the pathogen's genomic sequence. It concludes with a list of potential proteins with a possible, but not necessarily, list of peptide candidates that need to be experimentally confirmed for vaccine production. During the subsequent years, reverse vaccinology has dramatically changed: now it consists of a large number of bioinformatics tools and processes, namely subtractive proteomics, computational vaccinology, immunoinformatics, and in silico related procedures. However, the state of the art of reverse vaccinology still misses the ability to predict the efficacy of the proposed vaccine formulation. Here, we describe how to fill the gap by introducing an advanced immune system simulator that tests the efficacy of a vaccine formulation against the disease for which it has been designed. As a working example, we entirely apply this advanced reverse vaccinology approach to design and predict the efficacy of a potential vaccine formulation against influenza H5N1. Climate change and melting glaciers are critical due to reactivating frozen viruses and emerging new pandemics. H5N1 is one of the potential strains present in icy lakes that can raise a pandemic. Investigating structural antigen protein is the most profitable therapeutic pipeline to generate an effective vaccine against H5N1. In particular, we designed a multi-epitope vaccine based on predicted epitopes of hemagglutinin and neuraminidase proteins that potentially trigger B-cells, CD4, and CD8 T-cell immune responses. Antigenicity and toxicity of all predicted CTL, Helper T-lymphocytes, and B-cells epitopes were evaluated, and both antigenic and non-allergenic epitopes were selected. From the perspective of advanced reverse vaccinology, the Universal Immune System Simulator, an in silico trial computational framework, was applied to estimate vaccine efficacy using a cohort of 100 digital patients.

Virtual patients, digital twins and causal disease models: Paving the ground for in silico clinical trials.

Computational models are being explored to simulate in silico the efficacy and safety of drug candidates and medical devices. Disease models that are based on patients' profiling data are being produced to represent interactomes of genes or proteins and to infer causality in the pathophysiology, which makes it possible to mimic the impact of drugs on relevant targets. Virtual patients designed from medical records as well as digital twins are generated to simulate specific organs and to predict treatment efficacy at the individual patient level. As the acceptance of digital evidence by regulators grows, predictive artificial intelligence (AI)-based models will support the design of confirmatory trials in humans and will accelerate the development of efficient drugs and medical devices.

Calibration of Cohorts of Virtual Patient Heart Models Using Bayesian History Matching.

Previous patient-specific model calibration techniques have treated each patient independently, making the methods expensive for large-scale clinical adoption. In this work, we show how we can reuse simulations to accelerate the patient-specific model calibration pipeline. To represent anatomy, we used a Statistical Shape Model and to represent function, we ran electrophysiological simulations. We study the use of 14 biomarkers to calibrate the model, training one Gaussian Process Emulator (GPE) per biomarker. To fit the models, we followed a Bayesian History Matching (BHM) strategy, wherein each iteration a region of the parameter space is ruled out if the emulation with that set of parameter values produces is "implausible". We found that without running any extra simulations we can find 87.41% of the non-implausible parameter combinations. Moreover, we showed how reducing the uncertainty of the measurements from 10 to 5% can reduce the final parameter space by 6 orders of magnitude. This innovation allows for a model fitting technique, therefore reducing the computational load of future biomedical studies.

Virtual Populations for Quantitative Systems Pharmacology Models.

Quantitative systems pharmacology (QSP) places an emphasis on dynamic systems modeling, incorporating considerations from systems biology modeling and pharmacodynamics. The goal of QSP is often to quantitatively predict the effects of clinical therapeutics, their combinations, and their doses on clinical biomarkers and endpoints. In order to achieve this goal, strategies for incorporating clinical data into model calibration are critical. Virtual population (VPop) approaches facilitate model calibration while faced with challenges encountered in QSP model application, including modeling a breadth of clinical therapies, biomarkers, endpoints, utilizing data of varying structure and source, capturing observed clinical variability, and simulating with models that may require more substantial computational time and resources than often found in pharmacometrics applications. VPops are frequently developed in a process that may involve parameterization of isolated pathway models, integration into a larger QSP model, incorporation of clinical data, calibration, and quantitative validation that the model with the accompanying, calibrated VPop is suitable to address the intended question or help with the intended decision. Here, we introduce previous strategies for developing VPops in the context of a variety of therapeutic and safety areas: metabolic disorders, drug-induced liver injury, autoimmune diseases, and cancer. We introduce methodological considerations, prior work for sensitivity analysis and VPop algorithm design, and potential areas for future advancement. Finally, we give a more detailed application example of a VPop calibration algorithm that illustrates recent progress and many of the methodological considerations. In conclusion, although methodologies have varied, VPop strategies have been successfully applied to give valid clinical insights and predictions with the assistance of carefully defined and designed calibration and validation strategies. While a uniform VPop approach for all potential QSP applications may be challenging given the heterogeneity in use considerations, we anticipate continued innovation will help to drive VPop application for more challenging cases of greater scale while developing new rigorous methodologies and metrics.

Impact of Carbohydrate Counting Error on Glycemic Control in Open-Loop Management of Type 1 Diabetes: Quantitative Assessment Through an In Silico Trial.

BACKGROUND: In the management of type 1 diabetes (T1D), systematic and random errors in carb-counting can have an adverse effect on glycemic control. In this study, we performed an in silico trial aiming at quantifying the impact of different levels of carb-counting error on glycemic control. METHODS: The T1D patient decision simulator was used to simulate 7-day glycemic profiles of 100 adults using open-loop therapy. The simulation was repeated for different values of systematic and random carb-counting errors, generated with Gaussian distribution varying the error mean from -10% to +10% and standard deviation (SD) from 0% to 50%. The effect of the error was evaluated by computing the difference of time inside (∆TIR), above (∆TAR) and below (∆TBR) the target glycemic range (70-180mg/dl) compared to the reference case, that is, absence of error. Finally, 3 linear regression models were developed to mathematically describe how error mean and SD variations result in ∆TIR, ∆TAR, and ∆TBR changes. RESULTS: Random errors globally deteriorate the glycemic control; systematic underestimations lead to, on average, up to 5.2% more TAR than the reference case, while systematic overestimation results in up to 0.8% more TBR. The different time in range metrics were linearly related with error mean and SD (R2>0.95), with slopes of ßMEAN=0.21, ßSD=-0.07 for ∆TIR, ßMEAN=-0.25, ßSD=+0.06 for ∆TAR, and ßMEAN=0.05, ßSD=+0.01 for ∆TBR. CONCLUSIONS: The quantification of carb-counting error impact performed in this work may be useful understanding causes of glycemic variability and the impact of possible therapy adjustments or behavior changes in different glucose metrics.

Efficient Closed Loop Simulation of Do-It-Yourself Artificial Pancreas Systems.

BACKGROUND: Numerical simulations, also referred to as in silico trials, are nowadays the first step toward approval of new artificial pancreas (AP) systems. One suitable tool to run such simulations is the UVA/Padova Type 1 Diabetes Metabolic Simulator (T1DMS). It was used by Toffanin et al. to provide data about safety and efficacy of AndroidAPS, one of the most wide-spread do-it-yourself AP systems. However, the setup suffered from slow simulation speed. The objective of this work is to speed up simulation by implementing the algorithm directly in MATLAB®/Simulink®. METHOD: Firstly, AndroidAPS is re-implemented in MATLAB® and verified. Then, the function is incorporated into T1DMS. To evaluate the new setup, a scenario covering 2 days in real time is run for 30 virtual patients. The results are compared to those presented in the literature. RESULTS: Unit tests and integration tests proved the equivalence of the new implementation and the original AndroidAPS code. Simulation of the scenario required approximately 15 minutes, corresponding to a speed-up factor of roughly 1000 with respect to real time. The results closely resemble those presented by Toffanin et al. Discrepancies were to be expected because a different virtual population was considered. Also, some parameters could not be extracted from and harmonized with the original setup. CONCLUSIONS: The new implementation facilitates extensive in silico trials of AndroidAPS due to the significant reduction of runtime. This provides a cheap and fast means to test new versions of the algorithm before they are shared with the community.

A multi-step and multi-scale bioinformatic protocol to investigate potential SARS-CoV-2 vaccine targets.

The COVID-19 pandemic has highlighted the need to come out with quick interventional solutions that can now be obtained through the application of different bioinformatics software to actively improve the success rate. Technological advances in fields such as computer modeling and simulation are enriching the discovery, development, assessment and monitoring for better prevention, diagnosis, treatment and scientific evidence generation of specific therapeutic strategies. The combined use of both molecular prediction tools and computer simulation in the development or regulatory evaluation of a medical intervention, are making the difference to better predict the efficacy and safety of new vaccines. An integrated bioinformatics pipeline that merges the prediction power of different software that act at different scales for evaluating the elicited response of human immune system against every pathogen is proposed. As a working example, we applied this problem solving protocol to predict the cross-reactivity of pre-existing vaccination interventions against SARS-CoV-2.

In Silico Model Estimates the Clinical Trial Outcome of Cancer Vaccines.

Over 30 years after the first cancer vaccine clinical trial (CT), scientists still search the missing link between immunogenicity and clinical responses. A predictor able to estimate the outcome of cancer vaccine CTs would greatly benefit vaccine development. Published results of 94 CTs with 64 therapeutic vaccines were collected. We found that preselection of CT subjects based on a single matching HLA allele does not increase immune response rates (IRR) compared with non-preselected CTs (median 60% vs. 57%, p = 0.4490). A representative in silico model population (MP) comprising HLA-genotyped subjects was used to retrospectively calculate in silico IRRs of CTs based on the percentage of MP-subjects having epitope(s) predicted to bind ≥ 1-4 autologous HLA allele(s). We found that in vitro measured IRRs correlated with the frequency of predicted multiple autologous allele-binding epitopes (AUC 0.63-0.79). Subgroup analysis of multi-antigen targeting vaccine CTs revealed correlation between clinical response rates (CRRs) and predicted multi-epitope IRRs when HLA threshold was ≥ 3 (r = 0.7463, p = 0.0004) but not for single HLA allele-binding epitopes (r = 0.2865, p = 0.2491). Our results suggest that CRR depends on the induction of broad T-cell responses and both IRR and CRR can be predicted when epitopes binding to multiple autologous HLAs are considered.

Modeling-Based Decision Support System for Radical Prostatectomy Versus External Beam Radiotherapy for Prostate Cancer Incorporating an In Silico Clinical Trial and a Cost-Utility Study.

The aim of this study is to build a decision support system (DSS) to select radical prostatectomy (RP) or external beam radiotherapy (EBRT) for low- to intermediate-risk prostate cancer patients. We used an individual state-transition model based on predictive models for estimating tumor control and toxicity probabilities. We performed analyses on a synthetically generated dataset of 1000 patients with realistic clinical parameters, externally validated by comparison to randomized clinical trials, and set up an in silico clinical trial for elderly patients. We assessed the cost-effectiveness (CE) of the DSS for treatment selection by comparing it to randomized treatment allotment. Using the DSS, 47.8% of synthetic patients were selected for RP and 52.2% for EBRT. During validation, differences with the simulations of late toxicity and biochemical failure never exceeded 2%. The in silico trial showed that for elderly patients, toxicity has more influence on the decision than TCP, and the predicted QoL depends on the initial erectile function. The DSS is estimated to result in cost savings (EUR 323 (95% CI: EUR 213-433)) and more quality-adjusted life years (QALYs; 0.11 years, 95% CI: 0.00-0.22) than randomized treatment selection.

On the Sensitivity Analysis of Porous Finite Element Models for Cerebral Perfusion Estimation.

Computational physiological models are promising tools to enhance the design of clinical trials and to assist in decision making. Organ-scale haemodynamic models are gaining popularity to evaluate perfusion in a virtual environment both in healthy and diseased patients. Recently, the principles of verification, validation, and uncertainty quantification of such physiological models have been laid down to ensure safe applications of engineering software in the medical device industry. The present study sets out to establish guidelines for the usage of a three-dimensional steady state porous cerebral perfusion model of the human brain following principles detailed in the verification and validation (V&V 40) standard of the American Society of Mechanical Engineers. The model relies on the finite element method and has been developed specifically to estimate how brain perfusion is altered in ischaemic stroke patients before, during, and after treatments. Simulations are compared with exact analytical solutions and a thorough sensitivity analysis is presented covering every numerical and physiological model parameter. The results suggest that such porous models can approximate blood pressure and perfusion distributions reliably even on a coarse grid with first order elements. On the other hand, higher order elements are essential to mitigate errors in volumetric blood flow rate estimation through cortical surface regions. Matching the volumetric flow rate corresponding to major cerebral arteries is identified as a validation milestone. It is found that inlet velocity boundary conditions are hard to obtain and that constant pressure inlet boundary conditions are feasible alternatives. A one-dimensional model is presented which can serve as a computationally inexpensive replacement of the three-dimensional brain model to ease parameter optimisation, sensitivity analyses and uncertainty quantification. The findings of the present study can be generalised to organ-scale porous perfusion models. The results increase the applicability of computational tools regarding treatment development for stroke and other cerebrovascular conditions.

Exploring the Potential of Generative Adversarial Networks for Synthesizing Radiological Images of the Spine to be Used in In Silico Trials.

In silico trials recently emerged as a disruptive technology, which may reduce the costs related to the development and marketing approval of novel medical technologies, as well as shortening their time-to-market. In these trials, virtual patients are recruited from a large database and their response to the therapy, such as the implantation of a medical device, is simulated by means of numerical models. In this work, we propose the use of generative adversarial networks to produce synthetic radiological images to be used in in silico trials. The generative models produced credible synthetic sagittal X-rays of the lumbar spine based on a simple sketch, and were able to generate sagittal radiological images of the trunk using coronal projections as inputs, and vice versa. Although numerous inaccuracies in the anatomical details may still allow distinguishing synthetic and real images in the majority of cases, the present work showed that generative models are a feasible solution for creating synthetic imaging data to be used in in silico trials of novel medical devices.

Towards a Clinical Decision Support System for External Beam Radiation Oncology Prostate Cancer Patients: Proton vs. Photon Radiotherapy? A Radiobiological Study of Robustness and Stability.

We present a methodology which can be utilized to select proton or photon radiotherapy in prostate cancer patients. Four state-of-the-art competing treatment modalities were compared (by way of an in silico trial) for a cohort of 25 prostate cancer patients, with and without correction strategies for prostate displacements. Metrics measured from clinical image guidance systems were used. Three correction strategies were investigated; no-correction, extended-no-action-limit, and online-correction. Clinical efficacy was estimated via radiobiological models incorporating robustness (how probable a given treatment plan was delivered) and stability (the consistency between the probable best and worst delivered treatments at the 95% confidence limit). The results obtained at the cohort level enabled the determination of a threshold for likely clinical benefit at the individual level. Depending on the imaging system and correction strategy; 24%, 32% and 44% of patients were identified as suitable candidates for proton therapy. For the constraints of this study: Intensity-modulated proton therapy with online-correction was on average the most effective modality. Irrespective of the imaging system, each treatment modality is similar in terms of robustness, with and without the correction strategies. Conversely, there is substantial variation in stability between the treatment modalities, which is greatly reduced by correction strategies. This study provides a 'proof-of-concept' methodology to enable the prospective identification of individual patients that will most likely (above a certain threshold) benefit from proton therapy.

Clinical testing of propofol geriatic dose for sedation designed via in silico trial.

The geriatric population shows significant physiological changes due to aging and the multiple co-morbidities that they often present. Conventionally the propofol sedation dose for patients older than 65 years is 80% of the adult dose. We performed an in silico trial for elderly population and the results showed that the necessary simulated dose of propofol was lower than the conventional dose; therefore, a clinical trial was implemented to test three different propofol doses, two of them lower than the conventional dose, during a pacemaker implantation. The clinical trial showed that there was no clinical difference between the effects of the doses. A BIS monitor was used to measure the level of sedation, which proved to be adequate and well maintained by all patients. All the patients maintained an acceptable level of sedation, measured by a BIS monitor. Since propofol has some dose-dependent secondary effects, the use of lower doses, especially the ones designed for this age group, helps to avoid them.