Deli Wang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Deli Wang has a total of 33 co-authors. Here are the top co-authors by number of shared publications:
Gary Koch
The University of North Carolina at Chapel Hill, Chapel Hill, USA
2 shared publicationsWeining Robieson
AbbVie Inc. Chicago Illinois USA
2 shared publicationsBeatrice Rendenbach-Mueller
Neuroscience, AbbVie Deutschland GmbH , 67061, Ludwigshafen , Germany
1 shared publicationBo Fu
Data and Statistical Science, AbbVie Inc., 1 North Waukegan Road, North Chicago, IL 60064-6075, USA
1 shared publicationBo Yang
Vertex Pharmaceutical
1 shared publicationCatherine Wiener
Department of Biostatistics, The University of North Carolina, Chapel Hill, North Carolina, USA
1 shared publication
Latest Publications
- Open Access
- 1 October 2025
- Published by Wiley in Alzheimer's & Dementia: Translational Research & Clinical Interventions
- Vol. 11 (4) , e70181
Using AI‐generated digital twins to boost clinical trial efficiency in Alzheimer's disease
INTRODUCTION: Machine learning models leverage baseline data to create artificial intelligence (AI)–generated digital twins (DTs)—individualized predictions of each participant's clinical outcomes if they had received placebo. Incorporating DTs may increase statistical power or reduce required sample sizes in Phase 2 or 3 trials, and therefore improve efficiency in Alzheimer's disease (AD) trials. Here we demonstrate these properties using data from an AD Phase 2 clinical trial (AWARE, NCT02880956). METHODS: A conditional restricted Boltzmann machine (CRBM) model was trained on historical clinical trials and observational data from 6736 unique subjects after data harmonization to generate DTs of participants from the AWARE trial. The AWARE trial enrolled 453 subjects with mild cognitive impairment (MCI) or mild AD. DTs were assessed as prognostic covariates to evaluate gains in variance and sample size reduction. RESULTS: Positive partial correlation coefficients were found between DTs and change score from baseline in key cognitive assessments ranging from 0.30 to 0.39 at Week 96 in the AWARE trial. These correlations were consistent with validation results from three independent trials, which ranged from 0.30 to 0.46. Total residual variance was reduced by ~9% to 15% with DTs. While maintaining statistical power, DTs could reduce total sample size by ~9% to 15%, and control arm sample size by 17% to 26% in future AD trials. DISCUSSION: Efficiency was improved in AD clinical trials using machine learning models to generate prognostic DTs by including them in statistical analysis modeling. This methodology aligns with regulatory guidance and represents an application of machine learning models suitable for the analysis of pivotal trial data. Validated DTs have the potential to improve clinical development efficiency in AD and in other neurological indications. Highlights: Digital twins (DTs) were generated by artificial intelligence (AI) models trained on historical datasets. Use of digital twin (DT) as a covariate in the analysis model can reduce treatment effect variability. By coupling DT with the analysis model, trial sample size can be reduced. DT technology was accepted by the U.S. Food and Drug Administration and European Medicines Agency for applications in clinical trials.
- 1 October 2025
- Published by Wiley in Alzheimer's & Dementia: Translational Research & Clinical Interventions
- Vol. 11 (4) , e70181
Open AccessUsing AI‐generated digital twins to boost clinical trial efficiency in Alzheimer's disease
INTRODUCTION: Machine learning models leverage baseline data to create artificial intelligence (AI)–generated digital twins (DTs)—individualized predictions of each participant's clinical outcomes if they had received placebo. Incorporating DTs may increase statistical power or reduce required sample sizes in Phase 2 or 3 trials, and therefore improve efficiency in Alzheimer's disease (AD) trials. Here we demonstrate these properties using data from an AD Phase 2 clinical trial (AWARE, NCT02880956). METHODS: A conditional restricted Boltzmann machine (CRBM) model was trained on historical clinical trials and observational data from 6736 unique subjects after data harmonization to generate DTs of participants from the AWARE trial. The AWARE trial enrolled 453 subjects with mild cognitive impairment (MCI) or mild AD. DTs were assessed as prognostic covariates to evaluate gains in variance and sample size reduction. RESULTS: Positive partial correlation coefficients were found between DTs and change score from baseline in key cognitive assessments ranging from 0.30 to 0.39 at Week 96 in the AWARE trial. These correlations were consistent with validation results from three independent trials, which ranged from 0.30 to 0.46. Total residual variance was reduced by ~9% to 15% with DTs. While maintaining statistical power, DTs could reduce total sample size by ~9% to 15%, and control arm sample size by 17% to 26% in future AD trials. DISCUSSION: Efficiency was improved in AD clinical trials using machine learning models to generate prognostic DTs by including them in statistical analysis modeling. This methodology aligns with regulatory guidance and represents an application of machine learning models suitable for the analysis of pivotal trial data. Validated DTs have the potential to improve clinical development efficiency in AD and in other neurological indications. Highlights: Digital twins (DTs) were generated by artificial intelligence (AI) models trained on historical datasets. Use of digital twin (DT) as a covariate in the analysis model can reduce treatment effect variability. By coupling DT with the analysis model, trial sample size can be reduced. DT technology was accepted by the U.S. Food and Drug Administration and European Medicines Agency for applications in clinical trials.
- Top Cited PapersOpen Access
- 2 February 2023
- Published by Oxford University Press (OUP) in Brain
- Vol. 146 (6) , 2275-2284
Tilavonemab in early Alzheimer’s disease: results from a phase 2, randomized, double-blind study
Tau accumulation in patients with Alzheimer’s disease tracks closely with cognitive decline and plays a role in the later stages of disease progression. This phase 2 study evaluated the safety and efficacy of tilavonemab, an anti-tau monoclonal antibody, in patients with early Alzheimer’s disease. In this 96-week, randomized, double-blind, placebo-controlled study (NCT02880956), patients aged 55–85 years meeting clinical criteria for early Alzheimer’s disease with a Clinical Dementia Rating-Global Score of 0.5, a Mini-Mental State Examination score of 22 to 30, a Repeatable Battery for the Assessment of Neuropsychological Status-Delayed Memory Index score of ≤85, and a positive amyloid PET scan were randomized 1:1:1:1 to receive one of three doses of tilavonemab (300 mg, 1000 mg, or 2000 mg) or placebo via intravenous infusion every 4 weeks. The primary end point was the change from baseline up to Week 96 in the Clinical Dementia Rating-Sum of Boxes (CDR-SB) score. Safety evaluations included adverse event monitoring and MRI assessments. A total of 453 patients were randomized, of whom 337 were treated with tilavonemab (300 mg, n = 108; 1000 mg, n = 116; 2000 mg, n = 113) and 116 received placebo. Baseline demographics and disease characteristics were comparable across groups. The mean age was 71.3 (SD 7.0) years, 51.7% were female, and 96.5% were White. At baseline, the mean CDR-SB score was 3.0 (1.2), which worsened through Week 96 for all treatment groups. The least squares mean change from baseline at Week 96 in the CDR-SB score with tilavonemab was not significantly different compared with placebo [300 mg (n = 85): −0.07 (95% confidence interval, CI: −0.83 to 0.69); 1000 mg (n = 91): −0.06 (95% CI: −0.81 to 0.68); 2000 mg (n = 81): 0.16 (95% CI: −0.60 to 0.93); all P ≥ 0.05]. The incidence of any adverse event and MRI findings were generally comparable across groups. Tilavonemab was generally well tolerated but did not demonstrate efficacy in treating patients with early Alzheimer’s disease. Further investigations of tilavonemab in early Alzheimer’s disease are not warranted.
- 2 February 2023
- Published by Oxford University Press (OUP) in Brain
- Vol. 146 (6) , 2275-2284
Top Cited PapersOpen AccessTilavonemab in early Alzheimer’s disease: results from a phase 2, randomized, double-blind study
Tau accumulation in patients with Alzheimer’s disease tracks closely with cognitive decline and plays a role in the later stages of disease progression. This phase 2 study evaluated the safety and efficacy of tilavonemab, an anti-tau monoclonal antibody, in patients with early Alzheimer’s disease. In this 96-week, randomized, double-blind, placebo-controlled study (NCT02880956), patients aged 55–85 years meeting clinical criteria for early Alzheimer’s disease with a Clinical Dementia Rating-Global Score of 0.5, a Mini-Mental State Examination score of 22 to 30, a Repeatable Battery for the Assessment of Neuropsychological Status-Delayed Memory Index score of ≤85, and a positive amyloid PET scan were randomized 1:1:1:1 to receive one of three doses of tilavonemab (300 mg, 1000 mg, or 2000 mg) or placebo via intravenous infusion every 4 weeks. The primary end point was the change from baseline up to Week 96 in the Clinical Dementia Rating-Sum of Boxes (CDR-SB) score. Safety evaluations included adverse event monitoring and MRI assessments. A total of 453 patients were randomized, of whom 337 were treated with tilavonemab (300 mg, n = 108; 1000 mg, n = 116; 2000 mg, n = 113) and 116 received placebo. Baseline demographics and disease characteristics were comparable across groups. The mean age was 71.3 (SD 7.0) years, 51.7% were female, and 96.5% were White. At baseline, the mean CDR-SB score was 3.0 (1.2), which worsened through Week 96 for all treatment groups. The least squares mean change from baseline at Week 96 in the CDR-SB score with tilavonemab was not significantly different compared with placebo [300 mg (n = 85): −0.07 (95% confidence interval, CI: −0.83 to 0.69); 1000 mg (n = 91): −0.06 (95% CI: −0.81 to 0.68); 2000 mg (n = 81): 0.16 (95% CI: −0.60 to 0.93); all P ≥ 0.05]. The incidence of any adverse event and MRI findings were generally comparable across groups. Tilavonemab was generally well tolerated but did not demonstrate efficacy in treating patients with early Alzheimer’s disease. Further investigations of tilavonemab in early Alzheimer’s disease are not warranted.
- 23 March 2022
- Published by Springer Nature in Therapeutic Innovation & Regulatory Science
- Vol. 56 (3) , 501-516
Application of Meta-analysis to Evaluate Relationships Among ARIA-E Rate, Amyloid Reduction Rate, and Clinical Cognitive Response in Amyloid Therapeutic Clinical Trials for Early Alzheimer’s Disease
Background Removal of the extracellular Aβ plaques in the brain is one of the mechanisms to treat Alzheimer’s disease (AD). Separate clinical trials for several therapeutic compounds that target amyloid plaque reduction have shown noteworthy correlations among plaque removal, the Amyloid-Related Imaging Abnormalities (ARIA) rate, and clinical efficacy of the treatment. The relationships among therapeutic dose levels, the rate of amyloid removal, and the clinical efficacy deserve further investigation across therapeutic agents, particularly for clinical trials to provide insights for strategies to develop amyloid therapies in Alzheimer’s disease. Methods Published data summaries from clinical trials with amyloid therapies of aducanumab, donanemab, lecanemab, and gantenerumab are evaluated with meta-analyses. Linear mixed models for repeated measurements for visits and random study effects are applied to analyze amyloid centiloid value reduction from baseline and clinical cognition change from baseline for treatment groups according to doses and compounds for the clinical trials. Logistic regression analysis is applied to evaluate the relationship between the amyloid removal rate and the ARIA-Edema (ARIA-E) rate across different dose groups and clinical trials. Results The extent of amyloid removal varies among therapeutic agents and their dose levels. Across treatment groups and clinical trials, amyloid centiloid value reductions at Weeks 26 and 52 are strongly correlated with both ARIA-E rate over 78 weeks and the clinical efficacy response in the Clinical Dementia-Rating Scale Sum of Boxes (CDR-SB) score change from baseline at Week 78; and the Spearman rank correlations for amyloid reduction at Week 52 are stronger as − 0.79 with the ARIA-E rate over 78 weeks and 0.73 with the Week 78 CDR-SB score change from baseline. Conclusion Aβ plaques removal in the brain due to amyloid therapy is strongly correlated with a better clinical response in patients with early Alzheimer’s disease and a higher ARIA-E rate for the treatment groups and clinical trials in this meta-analysis. These relationships suggest that the balance between the clinical efficacy response and safety in ARIA-E rate is relevant for the choice of doses for amyloid therapies in confirmatory clinical trials.
- 23 March 2022
- Published by Springer Nature in Therapeutic Innovation & Regulatory Science
- Vol. 56 (3) , 501-516
Application of Meta-analysis to Evaluate Relationships Among ARIA-E Rate, Amyloid Reduction Rate, and Clinical Cognitive Response in Amyloid Therapeutic Clinical Trials for Early Alzheimer’s Disease
Background Removal of the extracellular Aβ plaques in the brain is one of the mechanisms to treat Alzheimer’s disease (AD). Separate clinical trials for several therapeutic compounds that target amyloid plaque reduction have shown noteworthy correlations among plaque removal, the Amyloid-Related Imaging Abnormalities (ARIA) rate, and clinical efficacy of the treatment. The relationships among therapeutic dose levels, the rate of amyloid removal, and the clinical efficacy deserve further investigation across therapeutic agents, particularly for clinical trials to provide insights for strategies to develop amyloid therapies in Alzheimer’s disease. Methods Published data summaries from clinical trials with amyloid therapies of aducanumab, donanemab, lecanemab, and gantenerumab are evaluated with meta-analyses. Linear mixed models for repeated measurements for visits and random study effects are applied to analyze amyloid centiloid value reduction from baseline and clinical cognition change from baseline for treatment groups according to doses and compounds for the clinical trials. Logistic regression analysis is applied to evaluate the relationship between the amyloid removal rate and the ARIA-Edema (ARIA-E) rate across different dose groups and clinical trials. Results The extent of amyloid removal varies among therapeutic agents and their dose levels. Across treatment groups and clinical trials, amyloid centiloid value reductions at Weeks 26 and 52 are strongly correlated with both ARIA-E rate over 78 weeks and the clinical efficacy response in the Clinical Dementia-Rating Scale Sum of Boxes (CDR-SB) score change from baseline at Week 78; and the Spearman rank correlations for amyloid reduction at Week 52 are stronger as − 0.79 with the ARIA-E rate over 78 weeks and 0.73 with the Week 78 CDR-SB score change from baseline. Conclusion Aβ plaques removal in the brain due to amyloid therapy is strongly correlated with a better clinical response in patients with early Alzheimer’s disease and a higher ARIA-E rate for the treatment groups and clinical trials in this meta-analysis. These relationships suggest that the balance between the clinical efficacy response and safety in ARIA-E rate is relevant for the choice of doses for amyloid therapies in confirmatory clinical trials.
- 29 January 2019
- Published by Wiley in Pharmaceutical Statistics
- Vol. 18 (4) , 407-419
Statistical considerations in a delayed‐start design to demonstrate disease modification effect in neurodegenerative disorders
There has been a paradigm shift in diagnostic conceptualization of Alzheimer's disease (AD) based on the current evidence suggesting that structure and biology changes start to occur before clinical symptoms emerge. Consequently, therapeutic drug development is also shifting to treat early AD patients using biomarkers for enrichment in clinical trials. A similar paradigm shift is occurring for Parkinson disease. In the absence of acceptable biomarkers that could be combined with a clinical endpoint to demonstrate a disease modification (DM) effect in neurodegenerative disorders, a delayed‐start design can be applied to demonstrate a lasting effect on the disease course. The delayed‐start design includes two treatment periods, where in period 1, patients are randomized to receive an active treatment or placebo, and in period 2, placebo patients are switched to the active treatment while patients in the active treatment arm will continue the same treatment. The hypothesis is that patients who start the active treatment later will fail to catch up to the treatment benefit achieved by patients who receive the active treatment in both periods. A usual analytical approach has sought to demonstrate the divergence of slope during period 1 and the parallelism of slopes during period 2 as the DM effect. However, due to heterogeneity in timing and the magnitude of maximal effect among patients, nonlinear response over time could be observed within the two treatment arms in both periods. We propose an approach to evaluate the DM effect with the linearity assumption for treatment differences, but not for each arm separately.
- 29 January 2019
- Published by Wiley in Pharmaceutical Statistics
- Vol. 18 (4) , 407-419
Statistical considerations in a delayed‐start design to demonstrate disease modification effect in neurodegenerative disorders
There has been a paradigm shift in diagnostic conceptualization of Alzheimer's disease (AD) based on the current evidence suggesting that structure and biology changes start to occur before clinical symptoms emerge. Consequently, therapeutic drug development is also shifting to treat early AD patients using biomarkers for enrichment in clinical trials. A similar paradigm shift is occurring for Parkinson disease. In the absence of acceptable biomarkers that could be combined with a clinical endpoint to demonstrate a disease modification (DM) effect in neurodegenerative disorders, a delayed‐start design can be applied to demonstrate a lasting effect on the disease course. The delayed‐start design includes two treatment periods, where in period 1, patients are randomized to receive an active treatment or placebo, and in period 2, placebo patients are switched to the active treatment while patients in the active treatment arm will continue the same treatment. The hypothesis is that patients who start the active treatment later will fail to catch up to the treatment benefit achieved by patients who receive the active treatment in both periods. A usual analytical approach has sought to demonstrate the divergence of slope during period 1 and the parallelism of slopes during period 2 as the DM effect. However, due to heterogeneity in timing and the magnitude of maximal effect among patients, nonlinear response over time could be observed within the two treatment arms in both periods. We propose an approach to evaluate the DM effect with the linearity assumption for treatment differences, but not for each arm separately.
- 1 November 2015
- Published by Elsevier in Contemporary Clinical Trials
- Vol. 45, 13-20
Overview of multiple testing methodology and recent development in clinical trials
- 1 November 2015
- Published by Elsevier in Contemporary Clinical Trials
- Vol. 45, 13-20
Overview of multiple testing methodology and recent development in clinical trials