Disease Progression In Models Research Articles

Artificial Intelligence for Quantitative Modeling in Drug Discovery and Development: An Innovation and Quality Consortium Perspective on Use Cases and Best Practices.

Recent breakthroughs in artificial intelligence (AI) and machine learning (ML) have ushered in a new era of possibilities across various scientific domains. One area where these advancements hold significant promise is model-informed drug discovery and development (MID3). To foster a wider adoption and acceptance of these advanced algorithms, the Innovation and Quality (IQ) Consortium initiated the AI/ML working group in 2021 with the aim of promoting their acceptance among the broader scientific community as well as by regulatory agencies. By drawing insights from workshops organized by the working group and attended by key stakeholders across the biopharma industry, academia, and regulatory agencies, this white paper provides a perspective from the IQ Consortium. The range of applications covered in this white paper encompass the following thematic topics: (i) AI/ML-enabled Analytics for Pharmacometrics and Quantitative Systems Pharmacology (QSP) Workflows; (ii) Explainable Artificial Intelligence and its Applications in Disease Progression Modeling; (iii) Natural Language Processing (NLP) in Quantitative Pharmacology Modeling; and (iv) AI/ML Utilization in Drug Discovery. Additionally, the paper offers a set of best practices to ensure an effective and responsible use of AI, including considering the context of use, explainability and generalizability of models, and having human-in-the-loop. We believe that embracing the transformative power of AI in quantitative modeling while adopting a set of good practices can unlock new opportunities for innovation, increase efficiency, and ultimately bring benefits to patients.

Disease progression model anchored around clinical diagnosis in longitudinal cohorts: example of Alzheimer’s disease and related dementia

BackgroundAlzheimer’s disease and related dementia (ADRD) are characterized by multiple and progressive anatomo-clinical changes including accumulation of abnormal proteins in the brain, brain atrophy and severe cognitive impairment. Understanding the sequence and timing of these changes is of primary importance to gain insight into the disease natural history and ultimately allow earlier diagnosis. Yet, modeling changes over disease course from cohort data is challenging as the usual timescales (time since inclusion, chronological age) are inappropriate and time-to-clinical diagnosis is available on small subsamples of participants with short follow-up durations prior to diagnosis. One solution to circumvent this challenge is to define the disease time as a latent variable.MethodsWe developed a multivariate mixed model approach that realigns individual trajectories into the latent disease time to describe disease progression. In contrast with the existing literature, our methodology exploits the clinical diagnosis information as a partially observed and approximate reference to guide the estimation of the latent disease time. The model estimation was carried out in the Bayesian Framework using Stan. We applied the methodology to the MEMENTO study, a French multicentric clinic-based cohort of 2186 participants with 5-year intensive follow-up. Repeated measures of 12 ADRD markers stemmed from cerebrospinal fluid (CSF), brain imaging and cognitive tests were analyzed.ResultsThe estimated latent disease time spanned over twenty years before the clinical diagnosis. Considering the profile of a woman aged 70 with a high level of education and APOE4 carrier (the main genetic risk factor for ADRD), CSF markers of tau proteins accumulation preceded markers of brain atrophy by 5 years and cognitive decline by 10 years. However we observed that individual characteristics could substantially modify the sequence and timing of these changes, in particular for CSF level of Abeta _{42}.ConclusionBy leveraging the available clinical diagnosis timing information, our disease progression model does not only realign trajectories into the most homogeneous way. It accounts for the inherent residual inter-individual variability in dementia progression to describe the long-term anatomo-clinical degradations according to the years preceding clinical diagnosis, and to provide clinically meaningful information on the sequence of events.Trial registrationclinicaltrials.gov, NCT01926249. Registered on 16 August 2013.

Interaction of sex and onset site on the disease trajectory of amyotrophic lateral sclerosis.

Studies showed the impact of sex and onset site (spinal or bulbar) on disease onset and survival in ALS. However, they mainly result from cross-sectional or survival analysis, and the interaction of sex and onset site on the different proxies of disease trajectory has not been fully investigated. We selected all patients with repeated observations in the PRO-ACT database. We divided them into four groups depending on their sex and onset site. We estimated a multivariate disease progression model, named ALS Course Map, to investigate the combined temporal changes of the four sub-scores of the revised ALS functional rating scale (ALSFRSr), the forced vital capacity (FVC), and the body mass index (BMI). We then compared the progression rate, the estimated age at onset, and the relative progression of the outcomes across each group. We included 1438 patients from the PRO-ACT database. They were 51% men with spinal onset, 12% men with bulbar onset, 26% women with spinal onset, and 11% women with bulbar onset. We showed a significant influence of both sex and onset site on the ALSFRSr progression. The BMI decreased 8.9 months earlier (95% CI [3.9, 13.8]) in women than men, after correction for the onset site. Among patients with bulbar onset, FVC was impaired 2.6 months earlier (95% CI [0.6, 4.6]) in women. Using a multivariable disease modelling approach, we showed that sex and onset site are important drivers of the progression of motor function, BMI, and FVC decline.

From serological surveys to disease burden: a modelling pipeline for Chagas disease.

In 2012, the World Health Organization (WHO) set the elimination of Chagas disease intradomiciliary vectorial transmission as a goal by 2020. After a decade, some progress has been made, but the new 2021-2030 WHO roadmap has set even more ambitious targets. Innovative and robust modelling methods are required to monitor progress towards these goals. We present a modelling pipeline using local seroprevalence data to obtain national disease burden estimates by disease stage. Firstly, local seroprevalence information is used to estimate spatio-temporal trends in the Force-of-Infection (FoI). FoI estimates are then used to predict such trends across larger and fine-scale geographical areas. Finally, predicted FoI values are used to estimate disease burden based on a disease progression model. Using Colombia as a case study, we estimated that the number of infected people would reach 506 000 (95% credible interval (CrI) = 395 000-648 000) in 2020 with a 1.0% (95%CrI = 0.8-1.3%) prevalence in the general population and 2400 (95%CrI = 1900-3400) deaths (approx. 0.5% of those infected). The interplay between a decrease in infection exposure (FoI and relative proportion of acute cases) was overcompensated by a large increase in population size and gradual population ageing, leading to an increase in the absolute number of Chagas disease cases over time. This article is part of the theme issue 'Challenges and opportunities in the fight against neglected tropical diseases: a decade from the London Declaration on NTDs'.

The temporal event-based model: Learning event timelines in progressive diseases.

Timelines of events, such as symptom appearance or a change in biomarker value, provide powerful signatures that characterise progressive diseases. Understanding and predicting the timing of events is important for clinical trials targeting individuals early in the disease course when putative treatments are likely to have the strongest effect. However, previous models of disease progression cannot estimate the time between events and provide only an ordering in which they change. Here, we introduce the temporal event-based model (TEBM), a new probabilistic model for inferring timelines of biomarker events from sparse and irregularly sampled datasets. We demonstrate the power of the TEBM in two neurodegenerative conditions: Alzheimer's disease (AD) and Huntington's disease (HD). In both diseases, the TEBM not only recapitulates current understanding of event orderings but also provides unique new ranges of timescales between consecutive events. We reproduce and validate these findings using external datasets in both diseases. We also demonstrate that the TEBM improves over current models; provides unique stratification capabilities; and enriches simulated clinical trials to achieve a power of with less than half the cohort size compared with random selection. The application of the TEBM naturally extends to a wide range of progressive conditions.

Resolving the spatial architecture of myeloma and its microenvironment at the single-cell level

In multiple myeloma spatial differences in the subclonal architecture, molecular signatures and composition of the microenvironment remain poorly characterized. To address this shortcoming, we perform multi-region sequencing on paired random bone marrow and focal lesion samples from 17 newly diagnosed patients. Using single-cell RNA- and ATAC-seq we find a median of 6 tumor subclones per patient and unique subclones in focal lesions. Genetically identical subclones display different levels of spatial transcriptional plasticity, including nearly identical profiles and pronounced heterogeneity at different sites, which can include differential expression of immunotherapy targets, such as CD20 and CD38. Macrophages are significantly depleted in the microenvironment of focal lesions. We observe proportional changes in the T-cell repertoire but no site-specific expansion of T-cell clones in intramedullary lesions. In conclusion, our results demonstrate the relevance of considering spatial heterogeneity in multiple myeloma with potential implications for models of cell-cell interactions and disease progression.

Risk Assessment of Kidney Disease Progression and Efficacy of SGLT2 Inhibition in Patients With Type 2 Diabetes.

To develop a risk assessment tool to identify patients with type 2 diabetes (T2D) at higher risk for kidney disease progression and who might benefit more from sodium-glucose cotransporter 2 (SGLT2) inhibition. A total of 41,204 patients with T2D from four Thrombolysis In Myocardial Infarction (TIMI) clinical trials were divided into derivation (70%) and validation cohorts (30%). Candidate predictors of kidney disease progression (composite of sustained ≥40% decline in estimated glomerular filtration rate [eGFR], end-stage kidney disease, or kidney death) were selected with multivariable Cox regression. Efficacy of dapagliflozin was assessed by risk categories (low: <0.5%; intermediate: 0.5 to <2%; high: ≥2%) in Dapagliflozin Effect on Cardiovascular Events (DECLARE)-TIMI 58. There were 695 events over a median follow-up of 2.4 years. The final model comprised eight independent predictors of kidney disease progression: atherosclerotic cardiovascular disease, heart failure, systolic blood pressure, T2D duration, glycated hemoglobin, eGFR, urine albumin-to-creatinine ratio, and hemoglobin. The c-indices were 0.798 (95% CI, 0.774-0.821) and 0.798 (95% CI, 0.765-0.831) in the derivation and validation cohort, respectively. The calibration plot slope (deciles of predicted vs. observed risk) was 0.98 (95% CI, 0.93-1.04) in the validation cohort. Whereas relative risk reductions with dapagliflozin did not differ across risk categories, there was greater absolute risk reduction in patients with higher baseline risk, with a 3.5% absolute risk reduction in kidney disease progression at 4 years in the highest risk group (≥1%/year). Results were similar with the 2022 Chronic Kidney Disease Prognosis Consortium risk prediction model. Risk models for kidney disease progression can be applied in patients with T2D to stratify risk and identify those who experience a greater magnitude of benefit from SGLT2 inhibition.

BSBM-17 NOVEL VALIDATED PATIENT-DERIVED MODELS OF A RECTAL CANCER BRAIN METASTASIS

Abstract Colorectal adenocarcinomas (CRC) rarely metastasize to the brain, however, survival rates in patients with colorectal brain metastases (CBM) are short, with a median survival of only 4 months from diagnosis. Currently, no evidence-based treatment options against these rare brain metastases are available (PMC). Therefore, characterization of these tumors and development of targeted therapies are imperative. Optimal models for neoplasms heighten the notoriously low success rate of bench-to-bedside translation. Historically, brain cancer metastases models are established cell lines grown on plastic and transplanted as xenografts (Valiente; PMC7572582). Genomic drift, lack of genetic accuracy and heterogeneity, caused by serial passaging on plastic and in high-serum media, are frequent limitations of these models. Growth in defined media without serum and as 3-dimensional (3-D) spheroids or organoids is thought to improve the fidelity of cancer cell lines (Lee; PMID:16697959; Lo; PMC820864). Despite significant efforts of large collaborative projects (https://hcmi-searchable-catalog.nci.nih.gov/), 3-D rectal brain cancer metastasis models have not been reported. To address this unmet medical need, we developed 3-D (suspension, organoid) cultures of a rectal adenocarcinoma brain metastasis with moderate to focally poor differentiation. Next-generation sequencing, expression profiling, and immunohistochemistry were performed to validate our models. Our data confirmed that the PDO model contains tumor cells; pathogenic alterations and expression profiles in the models largely match that of the original tumor. Overall, we detected a significant phenotypic overlap with the original tumor sample. Spatial profiling and PDO-derived xenografts are in progress and will facilitate experimental therapeutics. Taken together, our models retained tumor heterogeneity and fidelity from the primary tumor, making this one of the first studies to establish 3-D patient-derived models for disease progression in rectal cancer brain metastases. Collectively, these models will be valuable translational tools for future exploration of targeted therapy, which remains absent in this disease process.

Multi-Task Deep Evidential Sequence Learning for Trustworthy Alzheimer’s Disease Progression Prediction

Alzheimer’s disease (AD) is an irreversible neurodegenerative disease. Providing trustworthy AD progression predictions for at-risk individuals contributes to early identification of AD patients and holds significant value in discovering effective treatments and empowering the patient in taking proactive care. Recently, although numerous disease progression models based on machine learning have emerged, they often focus solely on enhancing predictive accuracy and ignore the measurement of result reliability. Consequently, this oversight adversely affects the recognition and acceptance of these models in clinical applications. To address these problems, we propose a multi-task evidential sequence learning model for the trustworthy prediction of disease progression. Specifically, we incorporate evidential deep learning into the multi-task learning framework based on recurrent neural networks. We simultaneously perform AD clinical diagnosis and cognitive score predictions while quantifying the uncertainty of each prediction without incurring additional computational costs by leveraging the Dirichlet and Normal-Inverse-Gamma distributions. Moreover, an adaptive weighting scheme is introduced to automatically balance between tasks for more effective training. Finally, experimental results on the TADPOLE dataset validate that our model not only has a comparable predictive performance to similar models but also offers reliable quantification of prediction uncertainties, providing a crucial supplementary factor for risk-sensitive AD progression prediction applications.

Cost-Effectiveness of Single-Inhaler Triple Therapy (FF/UMEC/VI) versus Tiotropium Monotherapy in Patients with Symptomatic Moderate-to-Very Severe COPD in the UK.

For patients with chronic obstructive pulmonary disease (COPD) who remain symptomatic despite maintenance treatment, clinical management guidelines recommend a stepwise escalation from monotherapy to dual therapy, and from dual therapy to triple therapy. However, in clinical practice, patients are often escalated directly from monotherapy to triple therapy based on disease severity. This study evaluated the cost-effectiveness of once-daily, single-inhaler fluticasone furoate, umeclidinium, and vilanterol (FF/UMEC/VI) triple therapy compared with long-acting muscarinic antagonist monotherapy with once-daily tiotropium (TIO) in patients with symptomatic moderate-to-very severe COPD, from a UK National Health Service perspective. The validated GALAXY-COPD disease progression model was populated with patient baseline characteristics and treatment effect data from the 12-week GSK Study 207626 comparing FF/UMEC/VI with TIO in patients with moderate-to-very severe COPD. UK unit costs and drug costs (British Pound, 2021) were applied to healthcare resource utilization and treatments. The base case analysis was conducted over a lifetime horizon, and costs and health outcomes (except for life years [LYs]) were discounted at 3.5% per year. Model outputs included exacerbation rates, healthcare costs, LYs, quality-adjusted LYs (QALYs), and incremental cost-effectiveness ratios. Overall, treatment with FF/UMEC/VI resulted in increased clinical benefit (reduction in total exacerbations and increased overall survival and QALYs), coupled with cost savings (derived from lower maintenance and exacerbation healthcare costs) compared with TIO monotherapy. In the base case analysis, FF/UMEC/VI provided an additional 0.393 LYs (95% range: 0.176, 0.655) and 0.443 QALYs (0.246, 0.648), at a cost saving of £880 (£54, £1608) versus TIO. FF/UMEC/VI remained the cost-effective (dominant) treatment option across sensitivity and scenario analyses. FF/UMEC/VI offers greater clinical benefits and is a cost-effective treatment option compared with TIO for the treatment of adult patients with COPD with persistent symptoms and/or who are at risk of exacerbation in the UK.

Clinical Pharmacology Applications of Real-World Data and Real-World Evidence in Drug Development and Approval-An Industry Perspective.

Since the 21st Century Cures Act was signed into law in 2016, real-world data (RWD) and real-world evidence (RWE) have attracted great interest from the healthcare ecosystem globally. The potential and capability of RWD/RWE to inform regulatory decisions and clinical drug development have been extensively reviewed and discussed in the literature. However, a comprehensive review of current applications of RWD/RWE in clinical pharmacology, particularly from an industry perspective, is needed to inspire new insights and identify potential future opportunities for clinical pharmacologists to utilize RWD/RWE to address key drug development questions. In this paper, we review the RWD/RWE applications relevant to clinical pharmacology based on recent publications from member companies in the International Consortium for Innovation and Quality in Pharmaceutical Development (IQ) RWD Working Group, and discuss the future direction of RWE utilization from a clinical pharmacology perspective. A comprehensive review of RWD/RWE use cases is provided and discussed in the following categories of application: drug-drug interaction assessments, dose recommendation for patients with organ impairment, pediatric plan development and study design, model-informed drug development (e.g., disease progression modeling), prognostic and predictive biomarkers/factors identification, regulatory decisions support (e.g., label expansion), and synthetic/external control generation for rare diseases. Additionally, we describe and discuss common sources of RWD to help guide appropriate data selection to address questions pertaining to clinical pharmacology in drug development and regulatory decision making.

Cost-effectiveness of integrated treatment for hepatitis C virus (HCV) among people who inject drugs in Norway: An economic evaluation of the INTRO-HCV trial.

The INTRO-HCV randomized controlled trial conducted in Norway over 2017-2019 found that integrated treatment, compared with standard-of-care hospital treatment, for hepatitis C virus (HCV) with direct-acting antivirals (DAAs) improved treatment outcomes among people who inject drugs (PWID). We evaluated cost-effectiveness of the INTRO-HCV intervention. A Markov health state transition model of HCV disease progression and treatment with cost-effectiveness analysis from the health-provider perspective. Primary cost, utility, and health outcome data were derived from the trial. Costs and health benefits (quality-adjusted life-years, QALYs) were tracked over 50 years. Probabilistic and univariate sensitivity analyses investigated DAA price reductions and variations in HCV treatment and disease care cost assumptions, using costs from different countries (Norway, United Kingdom, United States, France, Australia). PWID attending community-based drug treatment centers for people with opioid dependence in Norway. Incremental cost-effectiveness ratio (ICER) in terms of cost per QALY gained, compared against a conventional (€70 000/QALY) willingness-to-pay threshold for Norway and lower (€20 000/QALY) threshold common among high-income countries. Integrated treatment resulted in an ICER of €13 300/QALY gained, with 99% and 71% probability of being cost-effective against conventional and lower willingness-to-pay thresholds, respectively. A 30% lower DAA price reduced the ICER to €6 900/QALY gained, with 91% probability of being cost-effective at the lower willingness-to-pay threshold. A 60% and 90% lower DAA price had 36% and >99% probability of being cost-saving, respectively. Sensitivity analyses suggest integrated treatment was cost-effective at the lower willingness-to-pay threshold (>60% probability) across different assumptions on HCV treatment and disease care costs with 30% DAA price reduction, and became cost-saving with 60%-90% price reductions. Integrated hepatitis C virus treatment for people who inject drugs in community settings is likely cost-effective compared with standard-of-care referral pathways in Norway and may be cost-saving in settings with particular characteristics.

Disease progression modelling of Alzheimer’s disease using probabilistic principal components analysis

The recent biological redefinition of Alzheimer’s Disease (AD) has spurred the development of statistical models that relate changes in biomarkers with neurodegeneration and worsening condition linked to AD. The ability to measure such changes may facilitate earlier diagnoses for affected individuals and help in monitoring the evolution of their condition. Amongst such statistical tools, disease progression models (DPMs) are quantitative, data-driven methods that specifically attempt to describe the temporal dynamics of biomarkers relevant to AD. Due to the heterogeneous nature of this disease, with patients of similar age experiencing different AD-related changes, a challenge facing longitudinal mixed-effects-based DPMs is the estimation of patient-realigning time-shifts. These time-shifts are indispensable for meaningful biomarker modelling, but may impact fitting time or vary with missing data in jointly estimated models. In this work, we estimate an individual’s progression through Alzheimer’s disease by combining multiple biomarkers into a single value using a probabilistic formulation of principal components analysis. Our results show that this variable, which summarises AD through observable biomarkers, is remarkably similar to jointly estimated time-shifts when we compute our scores for the baseline visit, on cross-sectional data from the Alzheimer’s Disease Neuroimaging Initiative (ADNI). Reproducing the expected properties of clinical datasets, we confirm that estimated scores are robust to missing data or unavailable biomarkers. In addition to cross-sectional insights, we can model the latent variable as an individual progression score by repeating estimations at follow-up examinations and refining long-term estimates as more data is gathered, which would be ideal in a clinical setting. Finally, we verify that our score can be used as a pseudo-temporal scale instead of age to ignore some patient heterogeneity in cohort data and highlight the general trend in expected biomarker evolution in affected individuals.

A KNN Algorithm Based Predictive Model for Heart Disease Progression

Background: Disease diagnosis was a vital responsibility in healthcare. Machine learning classification methods would considerably improve the healthcare industry by providing a quick diagnosis of disease. Thus, time could be saved for doctors. Nearly 17.9 million people expire due to heart disease every year. Objectives: World Health Organization predicted that rate of death might increase by 24.5 million in 2030. Since heart illness was the major cause of death in comparison with other diseases today, it was the most challenging disease to diagnose. Methods: One of the reasons for death due to heart disease was due to the fact that risks were not identified in the earlier stage. Earlier diagnosis of disease was very much important. Machine Learning algorithms were used for predicting the prognosis of disease. Results: Here K-NN algorithm was used to predict the presence of heart disease in an individual. Thus, patients were classified as either positive or negative for heart disease and this model enhanced medical care and reduced the cost. This gave us significant knowledge that helps us to predict the patients with heart disease. Conclusion: The Python sci-kit library was used to implement this in Anaconda Navigator's Spyder Integrated Development Environment. Experiments revealed that technique worked well and was more accurate than before.

Predictori ai evoluţiei clinice a schizofreniei

Introduction. Schizophrenia is a serious disorder that influences all aspects of the patients’ life. The most important goals in schizophrenia are remission, recovery, and improving psychosocial functioning and quality of life. Unfortunately, there are important barriers to achieving optimal long-term outcomes in this disease due to patient-related factors and treatment-related factors. Lack of insight is considered the main reason for partial/non-adherence and, therefore, for the long-term outcome. Aim. The aim of this study was to identify the predictive factors involved in long-term schizophrenia, and especially the role of the awareness of illness. Materials and method. Eighty patients (44 males and 36 females) recruited from the First and Second Psychiatric Clinic from Cluj-Napoca, diagnosed according to ICD-10 criteria with schizophrenia, participated in this study. A semi-structured interview collected the demographical data. Psychotic symptoms were evaluated using Positive and Negative Syndrome Scale (PANSS), severity of the disease using Clinical Global Impression (CGI) and insight using the Scale for the Assessment of Unawareness of Mental Disorder (SUMD), the Schedule for Assessment of Insight-Expanded version (SAI-E) and Beck Cognitive Insight Scale (BCIS). Results. Our results showed that, among potential predictors of psychotic symptoms change measured in this research, those significantly correlated with changes are the level of insight measured with SUMD scale (r=-0.41; p&lt;0.01), presence of family history (r=0.24; p&lt;0.05), and belonging to urban areas (r=0.23; p&lt;0.05). The level of insight, as one predictor, with PANSS initially controlled, explained 16% of variance of improving psychotic symptoms during hospitalization. Also, patients from urban areas seem to have a slight tendency towards higher magnitudes of the changes (low correlation, but statistically significant). Again, the square of the correlation obtained (coefficient of determination) indicates that urban membership explaines 5% of variance when symptoms improve. Also, it seems that the family history is positively associated with the magnitude of symptoms’ change, but the level of this relationship is also reduced. Conclusions. The awareness of illness is one of the predictive factors for long-term schizophrenia, and the best predictive model of disease progression is composed of variables SUMD total and PANSS total on admission.

Semi-supervised calibration of noisy event risk (SCANER) with electronic health records

ObjectiveElectronic health records (EHR), containing detailed longitudinal clinical information on a large number of patients and covering broad patient populations, open opportunities for comprehensive predictive modeling of disease progression and treatment response. However, since EHRs were originally constructed for administrative purposes not for research, in the EHR-linked studies, it is often not feasible to capture reliable information for analytical variables, especially in the survival setting, when both accurate event status and event times are needed for model building. For example, progression-free survival (PFS), a commonly used survival outcome for cancer patients, often involves complex information embedded in free-text clinical notes and cannot be extracted reliably. Proxies of PFS time such as time to the first mention of progression in the notes are at best good approximations to the true event time. This leads to difficulty in efficiently estimating event rates for an EHR patient cohort. Estimating survival rates based on error-prone outcome definitions can lead to biased results and hamper the power in the downstream analysis. On the other hand, extracting accurate event time information via manual annotation is time and resource intensive. The objective of this study is to develop a calibrated survival rate estimator using noisy outcomes from EHR data. Materials and MethodsIn this paper, we propose a two-stage semi-supervised calibration of noisy event rate (SCANER) estimator that can effectively overcome censoring induced dependency and attains more robust performance (i.e., not sensitive to misspecification of the imputation model) by fully utilizing both a small-labeled set of gold-standard survival outcomes annotated via manual chart review and a set of proxy features automatically captured via EHR in the unlabeled set. We validate the SCANER estimator by estimating the PFS rates for a virtual cohort of lung cancer patients from one large tertiary care center and the ICU-free survival rates for COVID patients from two large tertiary care centers. ResultsIn terms of survival rate estimates, the SCANER had very similar point estimates compared to the complete-case Kaplan Meier estimator. On the other hand, other benchmark methods for comparison, which fail to account for the induced dependency between event time and the censoring time conditioning on surrogate outcomes, produced biased results across all three case studies. In terms of standard errors, the SCANER estimator was more efficient than the KM estimator, with up to 50% efficiency gain. ConclusionThe SCANER estimator achieves more efficient, robust, and accurate survival rate estimates compared to existing approaches. This promising new approach can also improve the resolution (i.e., granularity of event time) by using labels conditioning on multiple surrogates, particularly among less common or poorly coded conditions.

Fine‐grained ordering of white matter degeneration in the Alzheimer’s disease pathophysiological cascade

AbstractBackgroundThe role of white matter (WM) neurodegeneration in Alzheimer’s disease (AD) is poorly understood. Fine‐grained sequencing of WM neurodegeneration could provide new avenues for early diagnosis and novel disease‐modifying therapies to slow or halt AD progression. Previous approaches are limited by their coarse granularity and incomplete disease coverage. In this study, we apply event‐based disease progression modelling (DPM) (Fonteijn et al, Neuroimage, 2012), which can provide fine‐grained ordering, to estimate progression of regional WM abnormality in AD.MethodDiffusion tensor imaging (DTI) markers of WM abnormality and Freesurfer volumetric markers of grey matter (GM) atrophy were obtained from ADNI. DTI markers were hemisphere‐averaged and GM volumes were normalised by head size. Eleven WM ROIs with reported AD abnormality and 4 reference GM ROIs were selected for sequencing (abbreviations in Table 1), yielding complete data for 83 cognitively normal subjects, 40 AD patients and 102 MCI subjects. Event‐based DPM (github.com/ucl‐pond/kde_ebm) was used to estimate a data‐driven sequence of neurodegeneration and its uncertainty (positional variance), and individuals were assigned to their most likely stage (Young et al., Brain, 2014). Resampling analysis assessed the robustness of the estimated sequence.ResultFig. 1 shows the first data‐driven sequence of WM abnormalities in AD, including positional variance, with GM abnormality events for reference. The earliest WM events are increases in mean diffusivity (MD) in CGH and axial diffusivity (AxD) in FX and BCC. Subsequently, a cascade of AxD events in corpus callosum regions precedes abnormality in other medial temporal lobe WM structures.Fig. 2 shows the sequence is relatively robust under resampling, with low positional variance and similar median positions for most ROIs. Subject staging (Fig. 3, 4) shows that the fine‐grained approach splits MCI subjects with hippocampal atrophy into those with hippocampal atrophy and those with additional increased MD in CGH.ConclusionThis study finds WM microstructural abnormality among the earliest events in AD. Early hippocampal WM abnormality and corpus callosum ordering are consistent with early memory deficits in AD and inferences from cruder group‐based comparisons. The order suggests Wallerian degeneration rather than retrogenesis is the primary driver of WM neurodegeneration in AD.

Modeling Approach with Therapeutic Drug Monitoring Data to Describe Time Course of Clozapine Exposure and Positive and Negative Syndrome Scale Scores

The Positive and Negative Syndrome Scale (PANSS) is commonly used to assess the severity of the clinical symptoms of schizophrenia (SCZ). This study aimed to develop a pharmacokinetic (PK)/pharmacodynamic (PD) model based on therapeutic drug monitoring (TDM) data to characterize the relationship between clozapine exposure and the PANSS scores in patients with SCZ. TDM data for clozapine and PANSS scores from 45 patients with SCZ were included in this modeling analysis using NONMEM. Based on published data, intensive PK sampling data collected up to 12 hours postdose from 23 patients was incorporated into the PK data set to improve the fitting of absorption and disposition. For PD model development, the PANSS score was assessed at baseline, followed by 8 and 18 weeks after the initiation of clozapine dosing. Visual predictive check plots, the precision of parameter estimates, and decreases in the minimum objective function values were used for the model evaluation. A 2-compartment model with an absorption lag and a combined error model adequately described the PK of clozapine. The implementation of disease progression with placebo and drug effects improved the model's ability to describe the time course of the PANSS scores. In the final PK/PD model, Weibull and maximum effect (E max ) models were selected as disease progression models for the placebo and drug effect models, respectively. The model evaluation results supported the adequacy of the final model. A clozapine PK/PD model based on clinical settings adequately described the PANSS time course in patients with SCZ. These findings may aid the development of treatment strategies for patients with SCZ.

Multivariate disease progression modeling with longitudinal ordinal data.

Disease modeling is an essential tool to describe disease progression and its heterogeneity across patients. Usual approaches use continuous data such as biomarkers to assess progression. Nevertheless, categorical or ordinal data such as item responses in questionnaires also provide insightful information about disease progression. In this work, we propose a disease progression model for ordinal and categorical data. We built it on the principles of disease course mapping, a technique that uniquely describes the variability in both the dynamics of progression and disease heterogeneity from multivariate longitudinal data. This extension can also be seen as an attempt to bridge the gap between longitudinal multivariate models and the field of item response theory. Application to the Parkinson's progression markers initiative cohort illustrates the benefits of our approach: a fine-grained description of disease progression at the item level, as compared to the aggregated total score, together with improved predictions of the patient's future visits. The analysis of the heterogeneity across individual trajectories highlights known disease trends such as tremor dominant or postural instability and gait difficulties subtypes of Parkinson's disease.

Risk factors for intensive care admission in patients with COVID-19 pneumonia: A retrospective study

BackgroundCoronavirus disease 2019 (COVID-19)-associated mortality is predominantly due to respiratory failure. However, risk factors and predictive models for disease progression in patients with COVID-19 are not consistent across the globe. In this study, we aimed to assess the risk factors associated with intensive care (ICU) admission and mortality in patients with COVID-19 pneumonia. MethodsInformation was retrieved from the database of all patients admitted with COVID-19 pneumonia between March 2020 and July 2020 at a tertiary care center in Saudi Arabia. The patients’ demographic, clinical, laboratory and radiological characteristics were analyzed. ResultsOf 1054 patients admitted with PCR proven COVID-19, 254 patients (24%) with radiological evidence of pneumonia were enrolled. The median age was 55, with 25.6% above 65 years and 55.1% males. The comorbidities included hypertension (45%), diabetes (43%), dyslipidemia (24%), solid organ and bone marrow transplantation (14.5%), malignancy (13.4%), ischemic heart disease (10.6%) and chronic kidney disease (9.4%). The mortality rate was 4.7%, and 22.8% were admitted to the ICU.The risk factors for ICU admission were> 65 years of age (RR: 1.74, CI 95%, 1.10–2.74, p = 0.017), diabetes melitus (RR: 1.66, CI 95% 1.06–2.62, p = 0.028), heart failure (RR: 2.51, CI 95%, 1.28–4.93, p = 0.007), respiratory rate> 25 (RR: 2.75, CI 95%, 1.66–4.55, p < 0.001), upper lobe involvement (RR: 1.68, CI 95%, 1.02–2.77, p = 0.043), and C-reactive protein (CRP)> 140 (RR: 1.89, CI 95%, 1.14–3.13, p = 0.013).The risk factors for mortality were> 65 years of age (RR: 5.82, CI 95%, 1.81–18.68, p = 0.003), upper lobe involvement on chest radiography (RR:4.40, CI 95%, 1.22–15.86, p = 0.016), diffuse chest computed tomography changes (RR: 7.36, CI 95%, 2.31–23.46, p < 0.011), ischemic heart disease (RR: 4.20, CI 95%, 1.36–13.04, p = 0.028), chronic kidney disease (RR: 6.85, CI 95%, 2.35–19.90, p < 0.003), cerebrovascular disease (RR:13.61, CI 95%, 5.01–36.96 p < 0.001), respiratory rate> 25 (RR: 3.94, CI 95%, 1.32–11.78 p = 0.023), oxygen saturation< 90% on admission (RR: 12.19, CI 95%, 3.71–40.01, p < 0.001), thrombocytopenia (RR:4.16, CI 95%, 1.37–12.64, p = 0.013), and elevated troponin (RR: 6.20, CI 95%, 1.73–22.24, p = 0.003). ConclusionsIn this study, nearly a quarter of the patients with COVID-19 pneumonia required intensive care. We identified several risk factors associated with ICU admission and mortality that may be useful for predicting, triaging, and managing COVID-19 pneumonia patients. However, these findings need to be validated prospectively.