Response-adaptive Randomization Research Articles

Evaluating response-adaptive randomization procedures for recurrent events and terminal event data using a composite endpoint.

Recurrent event and terminal event data commonly arise in clinical and observational studies. To evaluate the efficacy of a treatment effect for both types of events, a composite endpoint has been used as a possible assessment, particularly when faced with high costs and a longer follow-up study. To model recurrent event processes complicated by the existence of a terminal event, joint frailty modeling has been typically employed. In this study, the objective was to develop some target-driven response adaptive randomization strategies using a composite endpoint based on joint frailty modeling. We first implemented a balanced randomized design and then investigated the response adaptive randomization. The former is intuitively first adopted while the latter is expected to be desirable and ethical in terms of allocating more subjects to the more effective treatment. The results show that the proposed procedures using a composite endpoint are capable of reducing the number of trial participants who receive inferior treatment while simultaneously reaching a desired optimal target as compared to a balanced randomized design. The R shiny application for calculating the sample size and allocation probabilities is also available. Finally, two clinical trials were used as pilot datasets to introduce the proposed procedures.

An adaptive clinical trial design to identify the target dose of tenecteplase for treatment of acute pulmonary embolism.

Fibrinolytic therapy with tenecteplase has been proposed for patients with pulmonary embolism but the optimal dose is unknown. Higher-than-necessary dosing is likely to cause excess bleeding. We designed an adaptive clinical trial to identify the minimum and assumed safest dose of tenecteplase that maintains efficacy. We propose a Bayesian adaptive, placebo-controlled, group-sequential dose-finding trial using response-adaptive randomization to preferentially allocate subjects to the most promising doses, dual analyses strategies (continuous and dichotomized) using a gatekeeping approach to maximize clinical impact, and interim stopping rules to efficiently address competing trial objectives. The operating characteristics of the proposed design were evaluated using Monte Carlo simulation across multiple hypothetical efficacy scenarios. Simulation demonstrated response-adaptive randomization can preferentially allocate subjects to doses which appear to be performing well based on interim data. Interim decision-making, including the interim evaluation of both analysis strategies with gatekeeping, allows the trial to continue enrollment when success with the dichotomized analysis strategy appears sufficiently likely and to stop enrollment and declare superiority based on the continuous analysis strategy when there is little chance of ultimately declaring superiority with the dichotomized analysis. The proposed design allows evaluation of a greater number of dose levels than would be possible with a non-adaptive design and avoids the need to choose either the continuous or the dichotomized analysis strategy for the primary endpoint.

GBM AGILE: A global, phase 2/3 adaptive platform trial to evaluate multiple regimens in newly diagnosed and recurrent glioblastoma.

TPS2078 Background: GBM AGILE (Glioblastoma Adaptive, Global, Innovative Learning Environment) is a biomarker based, multi-arm, international, seamless Phase 2/3 Response Adaptive Randomization platform trial designed to rapidly identify experimental therapies that improve overall survival and confirm efficacious experimental therapies and associated biomarker signatures to support new drug approvals and registration. GBM AGILE is a collaboration between academic investigators, patient organizations and industry to support new drug applications for newly diagnosed and recurrent GBM. With its adaptable structure, GBM AGILE has continued trial activation, inclusion of new investigational therapies, and enrollment globally through the challenges of a global pandemic. Methods: The primary objective of GBM AGILE is to identify therapies that effectively improve the overall survival in patients with ND or recurrent GBM. Bayesian response adaptive randomization is used within subtypes of the disease to assign participants to investigational arms based on their performance. Operating under a Master Protocol, GBM AGILE allows multiple drugs from different pharmaceutical/biotech companies to be evaluated simultaneously and/or over time against a common control. New experimental therapies are added as new information about promising new drugs is identified while other therapies are removed as they complete their evaluation. The master protocol/ trial infrastructure includes efficiencies through an adaptive trial design, shared control arm and operational processes such as risk-based monitoring and enhanced remote activities. GBM AGILE has screened over 1000 patients and enrollment rates are 3 to 4 times greater than traditional GBM trials, with active sites averaging 0.75 to 1 patients/sites/month. While enrollment had an initial dip during the early stages of the pandemic (April-May, 2020), with flexible processes including remote based monitoring, minimizing in person visits, and remote provision of IMP, the enrollment rebounded by June, 2020. Through the use of improved and efficient processes allowed within a master protocol/adaptive platform trial infrastructure, GBM AGILE has been seamlessly operating a global trial during a global pandemic. Clinical trial information: NCT03970447.

A biomarker-guided Bayesian response-adaptive phase II trial for metastatic melanoma: The Personalized Immunotherapy Platform (PIP) trial design.

TPS9599 Background: Anti-PD1-based immunotherapies have been approved for many cancer types and are now a standard therapy for advanced melanoma. Despite this, ̃50% of advanced melanoma patients (pts) fail to respond or eventually progress after response. It is therefore critical to identify pts with a low likelihood of response to anti-PD1-based therapy and efficiently assess activity of rationally-selected alternative novel immunotherapies. Methods: We designed this investigator-initiated phase II PIP-Trial to evaluate two consecutive biomarker testing platforms, followed by the activity of rationally selected 5 novel agents in pts with advanced melanoma. Two separate pt populations are included: Part-A) treatment-naïve pts predicted to be resistant to either anti-PD-1 alone or combined with ipilimumab using Biomarker Test-1); and Part-B) pts who had progressed on 1 prior line of PD1-based therapy. Part-A) is a Bayesian adaptive multi-arm multi-stage design using response adaptive randomisation after a burn-in period where pts are randomised to the existing arms with equal probability. From then on, regular interim analyses will be carried out with the objective to either drop poorly performing arms or continue. Part-B) is an open platform without control that combined a selection and an expansion phase to identify best novel agent(s) as second-line therapy. Expansion phase decisions will be based on enrichment for biomarker Test-2. Dropping an arm occurs when the posterior probability of observing a clinically significant effect on the primary outcome (i.e. 6-month RECIST objective response rate (ORR)) is low. The operational characteristics of the design were investigated through simulations considering 4 plausible scenarios with 40% ORR in the control arm (anti-PD1 + Anti-CTLA4). Simulations were based on the upcoming R package BATS. Part-A has at least 85% power to detect a 30% absolute improvement in ORR with respect to the control arm (with a max N = 216 – Table below). Part-B will be able to select two promising treatments in the expansion phase and formally test their efficacy against a minimum ORR of 25% at 80% power (max N = 150). [Table: see text]

Feasibility and conduct of INSIGhT, a platform trial of patients with glioblastoma using Bayesian adaptive randomization.

2012 Background: Individualized Screening Trial of Innovative Glioblastoma Therapy (INSIGhT) trial is a phase II platform trial using response adaptive randomization and deep genomic profiling to more efficiently test experimental agents in MGMT unmethylated glioblastoma and potentially accelerate identification of novel therapies for phase III testing. We report on the feasibility and conduct of this approach. Methods: Tumor genotyping was performed prior to treatment assignment on eligible participants with newly diagnosed MGMT-unmethylated glioblastoma to identify biomarker signatures. Initial randomization was 1:1:1:1 between control (temozolomide) and 3 experimental arms (abemaciclib, CC-115, and neratinib). Subsequent randomization was adapted based on Bayesian estimation of biomarker-naïve and biomarker-specific probabilities of treatment impact on progression-free survival (PFS). Ineffective or toxic arms were discontinued by protocol amendment. The primary endpoint was overall survival (OS). Results: INSIGhT randomized 71 patients to the control arm, 73 patients to the abemaciclib arm, 12 patients to the CC-115 arm, and 81 patients to the neratinib arm between 2/9/2017 and 5/14/2021. Following the initial equal randomization period, early data were repeatedly analyzed during the study to capture early signals of treatment effects across the enrolled population or in specific biomarker subgroups. The results of these interim analyses influenced the randomization probability for future enrolled patients. In total, 77% of the participants were randomized before assessing their biomarker profile and 23% were biomarker randomized. The CC-115 arm opened and closed three times during the safety lead-in. The randomization probability to the CC-115 arm decreased based on poor early PFS results and the arm eventually closed after 12 patients due to toxicity. The randomization probability to the abemacicilb arm increased based on promising early PFS results. After the completion of accrual into the abemaciclib arm, the trial switched to block randomization to finish enrolling into the remaining neratinib and control arms. A total of 28 interim analyses and 32 randomization tables were created throughout the course of the trial with 4 adjustments (3 due to CC-115 closures and 1 due to completion of the abemaciclib arm). Biomarker association trends for neratinib and abemaciclib were similar to those seen in preclinical modeling of the trial. Conclusions: Relative to a standard randomization design, the adaptive platform design facilitated more efficient and economical testing of experimental arms by sharing a control arm, decreasing the probability of enrollment to potentially ineffective arms, and increasing the probability of enrollment to potentially effective arms. Additional future arms are planned on INSIGhT. Clinical trial information: NCT02977780.

Response-adaptive treatment randomization for multiple comparisons of treatments with recurrent event responses.

Recurrent event responses are frequently encountered during clinical trials of treatments for certain diseases, such as asthma. The recurrence rates of different treatments are often compared by applying the negative binomial model. In addition, a balanced treatment-allocation procedure that assigns the same number of patients to each treatment is often applied. Recently, a response-adaptive treatment-allocation procedure has been developed for trials with recurrent event data, and has been shown to be superior to balanced treatment allocation. However, this response-adaptive treatment allocation procedure is only applicable for the comparison of two treatments. In this paper, we derive response-adaptive treatment-allocation procedures for trials which comprise several treatments. As pairwise comparisons and multiple comparisons with a control are two common multiple-testing scenarios in trials with more than two treatments, corresponding treatment-allocation procedures for these scenarios are also investigated. The redesign of two clinical studies illustrates the clinical benefits that would be obtained from our proposed response-adaptive treatment-allocation procedures.

An alternative metric for evaluating the potential patient benefit of response-adaptive randomization procedures.

When planning a two-arm group sequential clinical trial with a binary primary outcome that has severe implications for quality of life (e.g., mortality), investigators may strive to find the design that maximizes in-trial patient benefit. In such cases, Bayesian response-adaptive randomization (BRAR) is often considered because it can alter the allocation ratio throughout the trial in favor of the treatment that is currently performing better. Although previous studies have recommended using fixed randomization over BRAR based on patient benefit metrics calculated from the realized trial sample size, these previous comparisons have been limited by failures to hold type I and II error rates constant across designs or consider the impacts on all individuals directly affected by the design choice. In this paper, we propose a metric for comparing designs with the same type I and II error rates that reflects expected outcomes among individuals who would participate in the trial if enrollment is open when they become eligible. We demonstrate how to use the proposed metric to guide the choice of design in the context of two recent trials in persons suffering out of hospital cardiac arrest. Using computer simulation, we demonstrate that various implementations of group sequential BRAR offer modest improvements with respect to the proposed metric relative to conventional group sequential monitoringalone.

Personalized Risk-Based Screening Design for Comparative Two-Arm Group Sequential Clinical Trials.

Personalized medicine has been emerging to take into account individual variability in genes and environment. In the era of personalized medicine, it is critical to incorporate the patients’ characteristics and improve the clinical benefit for patients. The patients’ characteristics are incorporated in adaptive randomization to identify patients who are expected to get more benefit from the treatment and optimize the treatment allocation. However, it is challenging to control potential selection bias from using observed efficacy data and the effect of prognostic covariates in adaptive randomization. This paper proposes a personalized risk-based screening design using Bayesian covariate-adjusted response-adaptive randomization that compares the experimental screening method to a standard screening method based on indicators of having a disease. Personalized risk-based allocation probability is built for adaptive randomization, and Bayesian adaptive decision rules are calibrated to preserve error rates. A simulation study shows that the proposed design controls error rates and yields a much smaller number of failures and a larger number of patients allocated to a better intervention compared to existing randomized controlled trial designs. Therefore, the proposed design performs well for randomized controlled clinical trials under personalized medicine.

Abstract PD8-07: Evaluation of Tucatinib + (Paclitaxel + Pertuzumab + Trastuzumab) followed by AC in high-risk HER2 positive (HER2+) stage II/III breast cancer: Results from the I-SPY 2 TRIAL

Abstract Background: I-SPY 2 is a multicenter, phase 2 trial using response-adaptive randomization within molecular subtypes defined by receptor status and MammaPrint (MP) risk to evaluate novel agents as neoadjuvant therapy for women with high-risk breast cancer. Tucatinib is a potent HER2 (ErbB2) tyrosine kinase inhibitor, selective for HER2 vs. epidermal growth factor receptor (EGFR) and is active vs. brain metastases. Safety and efficacy of tucatinib combined with paclitaxel, pertuzumab, and trastuzumab are unknown and were tested in a planned 10 patient (pt) safety run-in of the I-SPY 2 trial. Methods: Women with tumors ≥ 2.5cm were eligible for screening. Only pts with tumors that were HER2+ by FISH were eligible for this treatment. Treatment included tucatinib (max dose 300 mg) BID for 12 weeks with weekly paclitaxel 80 mg/m2 and trastuzumab (2 mg/kg weekly following loading), and pertuzumab (420 mg every 3 weeks following loading), followed by doxorubicin/cyclophosphamide (AC) every 2 weeks x 4. The control arm was weekly paclitaxel and trastuzumab with pertuzumab for 12 weeks followed by AC every 2 weeks x 4. All pts undergo serial MR imaging and response at 3 & 12 weeks is combined with real time pCR data to estimate, and continuously update, the predicted pCR rate for each trial arm. The goal of the trial is to identify/graduate regimens with ≥85.% Bayesian predictive probability of success (i.e. demonstrating superiority to control) in a future 300-patient phase 3 neoadjuvant trial with a pCR endpoint. This run-in arm was conducted to determine safety of combining tucatinib with paclitaxel/trastuzumab/pertuzumab, monitoring special adverse events of interest including LFT elevations and gastrointestinal toxicities.Methods: The I-SPY 2 methods have been previously published. Results: 20 pts were evaluable in tucatinib treatment arm. The control arm included 329 historical controls enrolled since April 2010. The initial tucatinib dose was 300 mg BID. After enrollment of the first 8 pts, there were 3 pts with grade 3 LFT elevations, 2 pts with grade 2/3 diarrhea, 1 pt with grade 2 neutropenia, and 1 pt with grade 3 nausea. After this safety review, the tucatinib dose was lowered to 250 mg BID. Among 5 additional pts enrolled, 3 developed grade 2/3 LFT abnormalities. The protocol was then modified to tucatinib 150 mg BID days 1-28 and then 250 mg BID days 29-84; 7 pts were treated. Safety data were reviewed after 20 pts were enrolled; the arm was then suspended due to similar LFT elevations regardless of tucatinib dose reduction or schedule. 7 of 20 pts (35%) had reversible Grade 3 or higher ALT/AST elevation (Table). No pt met criteria for Hy’s Law. In terms of efficacy, 12 of 14 evaluable pts had > 80% reduction of tumor volume by 12 weeks, measured by MRI assessment of functional tumor volume (FTV). Conclusion: The goal of the run-in arm was to determine the safety of adding tucatinib to the combination of paclitaxel/trastuzumab/pertuzumab. The addition of tucatinib resulted in unacceptable but reversible LFT elevations despite tucatinib dose reduction. Tucatinib containing therapy resulted in >80% decline in tumor volume at 12 weeks in 86% of pts. Tucatinib showed a high level of activity when combined with paclitaxel/trastuzumab/pertuzumab, but the combination is not feasible. Table: Number of pts with grade 2, 3, and 4 LFT elevations by treatment schedule (highest grade per patient per event, ALT or Treatment scheduleGrade 2 LFT elevationGrade 3 LFT elevationGrade 4 LFT elevationTucatinib 300 mg BID030Tucatinib 250 mg BID210Tucatinib 150 mg BID days 1-28 followed by 250 mg BID days 29 to 84112 Citation Format: David A Potter, Erin Roesch, Christina Yau, Ruixiao Lu, Denise Wolf, Susan Samson, Debra Stafford, Kathy S Albain, Claudine Isaacs, Meghana Trivedi, Douglas Yee, Judy Boughey, Alexandra Thomas, A. Jo Chien, Nola Hylton, Wen Li, Angela DeMichele, Jane Perlmutter, W. Fraser Symmans, Dawn L Hershman, Michelle Melisko, Laura J van 't Veer, Amy Wilson, Smita M Asare, Donald A Berry, Richard Schwab, Hope S Rugo, Laura J Esserman. Evaluation of Tucatinib + (Paclitaxel + Pertuzumab + Trastuzumab) followed by AC in high-risk HER2 positive (HER2+) stage II/III breast cancer: Results from the I-SPY 2 TRIAL [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr PD8-07.

Abstract GS4-02: Analysis of clinical outcomes and expression-based immune signatures by race in the I-SPY 2 trial

Abstract Background Transcriptomic immune-related gene signatures have been associated with achievement of pathologic complete response (pCR) and prognosis in the neoadjuvant setting. I-SPY 2 is a multicenter, phase 2 platform trial using response-adaptive randomization within subtypes defined by receptor status (HR/HER2) and MammaPrint (MP) risk to evaluate novel agents as neoadjuvant therapy for women with high-risk breast cancer. Given racial disparities in mortality from breast cancer and the paucity of racial demographic data from clinical trials, we aimed to evaluate the association between racial groups and baseline characteristics, including expression-based subtypes and immune signatures, treatment response, and prognosis of patients enrolled in the I-SPY 2 TRIAL. Methods Our study population included 990 I-SPY 2 patients. 15 patients identified as part of a racial group with <10 patients enrolled in the trial and were excluded from analysis. Pre-treatment expression data was available for 971 patients. Follow-up data was available for 907 patients; median follow-up time of 4.4 yrs. Chi-square test was used to assess associations between racial groups and pre-treatment SBR grade, HR/HER2 defined subtypes, intrinsic subtype (defined by BluePrint 80-gene molecular subtyping) and residual cancer burden (RCB) class. Logistic regression was used to evaluate race association with pCR. Cox proportional hazard modeling was used to assess the association between racial groups and event free survival (EFS) in a univariate setting, adjusting for pCR status. Association between racial groups and 28 expression signatures related to immune, proliferation, ER and HER2 pathway was analyzed using ANOVA with post-hoc Tukey test in the overall population and in each receptor subtype. Results Of 975 patients included in our analysis, 787 (81%) were White, 68 (7%) were Asian, and 120 (12%) were Black or African American. No significant associations between race and pre-treatment SBR grade (p=0.49), HR/HER2 defined subtypes (p=0.09), or expression-based subtypes (p=0.25) were observed. pCR rates do not significantly differ by racial groups (Odds ratio of pCR relative to White: 1.00 for Asian and 0.89 for Black or African American); and no significant differences in RCB class distribution by race was observed (p=0.88). Event free survival was not associated with patient racial group in a univariate Cox model (Hazard ratio relative to White: 1.10, p=0.73 for Asian and 1.37, p=0.13 for Black or African American). Among the 28 expression signatures evaluated, four were differentially expressed among racial groups within the overall population (F-test p<0.05): IFN module, B cell signature, Dendritic cell signature, and Mitotic score. Pairwise comparisons between racial groups with post-hoc Tukey test identified significant differences in IFN module expression between Black or African American vs. White (p=0.019) and Dendritic cell signature expression between Asian vs White (p=0.047). Among patients in the TNBC subtype, three signatures (dendritic cell signature, macrophage signature and ERBB2 module) were differentially expressed between Black or African American and White patients (p=0.002, 0.016 and 0.007). Conclusion Our analysis demonstrates that among women with high risk breast cancer, race does not affect subtype specific response rates nor event free survival. Distribution of subtypes previously shown to be associated with pCR in the I-SPY2 trial did not significantly differ among racial groups indicating race is less likely than tumor biology to predict response. The decreased expression of immune signatures observed in Black or African American women with TNBC suggests possible differential sensitivity to immunotherapy plus combination chemotherapy. Tumor immune multiplex studies are underway to further investigate. Citation Format: Beverly Kyalwazi, Christina Yau, Olufunmilayo Olopade, A. Jo Chien, Anne Wallace, Andres Forero-Torres, Lajos Pusztai, Erin Ellis, Kathy Albain, Anne Blaes, Barbara Haley, Judy Boughey, Anthony Elias, Amy Clark, Claudine Isaacs, Rita Nanda, Hyo Han, Rachel Yung, Debu Tripathy, Kristen Edmiston, Rebecca Viscusi, Donald Northfelt, Qamar Khan, Ashish Sanil, Scott Berry, Smita Asare, Amy Wilson, Gillian Hirst, Nola Hylton, Michelle Melisko, Jane Perlmutter, Hope Rugo, Fraser Symmans, Laura van ‘t Veer, Donald Berry, Laura Esserman. Analysis of clinical outcomes and expression-based immune signatures by race in the I-SPY 2 trial [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr GS4-02.

Application of machine learning methods in clinical trials for precision medicine.

ObjectiveA key component for precision medicine is a good prediction algorithm for patients’ response to treatments. We aim to implement machine learning (ML) algorithms into the response-adaptive randomization (RAR) design and improve the treatment outcomes.Materials and MethodsWe incorporated 9 ML algorithms to model the relationship of patient responses and biomarkers in clinical trial design. Such a model predicted the response rate of each treatment for each new patient and provide guidance for treatment assignment. Realizing that no single method may fit all trials well, we also built an ensemble of these 9 methods. We evaluated their performance through quantifying the benefits for trial participants, such as the overall response rate and the percentage of patients who receive their optimal treatments.ResultsSimulation studies showed that the adoption of ML methods resulted in more personalized optimal treatment assignments and higher overall response rates among trial participants. Compared with each individual ML method, the ensemble approach achieved the highest response rate and assigned the largest percentage of patients to their optimal treatments. For the real-world study, we successfully showed the potential improvements if the proposed design had been implemented in the study.ConclusionIn summary, the ML-based RAR design is a promising approach for assigning more patients to their personalized effective treatments, which makes the clinical trial more ethical and appealing. These features are especially desirable for late-stage cancer patients who have failed all the Food and Drug Administration (FDA)-approved treatment options and only can get new treatments through clinical trials.

Time trends with response-adaptive randomization: The inevitability of inefficiency.

Response-adaptive randomization, which changes the randomization ratio as a randomized clinical trial progresses, is inefficient as compared to a fixed 1:1 randomization ratio in terms of increased required sample size. It is also known that response-adaptive randomization leads to biased treatment effects if there are time trends in the accruing outcome data, for example, due to changes in the patient population being accrued, evaluation methods, or concomitant treatments. Response-adaptive-randomization analysis methods that account for potential time trends, such as time-block stratification or re-randomization, can eliminate this bias. However, as shown in this Commentary, these analysis methods cause a large additional inefficiency of response-adaptive randomization, regardless of whether a time trend actually exists.

The Contribution of Artificial Intelligence in Achieving the Sustainable Development Goals (SDGs): What Can Eye Health Can Learn From Commercial Industry and Early Lessons From the Application of Machine Learning in Eye Health Programmes.

Achieving The United Nations sustainable developments goals by 2030 will be a challenge. Researchers around the world are working toward this aim across the breadth of healthcare. Technology, and more especially artificial intelligence, has the ability to propel us forwards and support these goals but requires careful application. Artificial intelligence shows promise within healthcare and there has been fast development in ophthalmology, cardiology, diabetes, and oncology. Healthcare is starting to learn from commercial industry leaders who utilize fast and continuous testing algorithms to gain efficiency and find the optimum solutions. This article provides examples of how commercial industry is benefitting from utilizing AI and improving service delivery. The article then provides a specific example in eye health on how machine learning algorithms can be purposed to drive service delivery in a resource-limited setting by utilizing the novel study designs in response adaptive randomization. We then aim to provide six key considerations for researchers who wish to begin working with AI technology which include collaboration, adopting a fast-fail culture and developing a capacity in ethics and data science.

An optimal design in a two-stage ethical allocation based on U-statistics

The present article is related to a comparison between two competing treatments, say A and B in the context of a Phase III clinical trial. This comparison is made through a U-statistic estimator of the functional , where X and Y represent the continuous responses on treatments A and B respectively. Suppose the subjects are coming sequentially to a clinic and they are to be treated either by A or B. The recruitment of the subjects is done by following a two-stage ethical allocation design with an optimal sequential allocation at the first stage. Here a sequential sampling scheme with equal randomization is adopted to optimize the stage 1 sample size for a given the total sample size of the trial. Then, a response-adaptive randomization rule is employed at the second-stage with a view to achieve ethical gain measured in terms of the proportion of allocations on the better treatment. The motivation of the proposed two stage design is to assign a larger number of subjects to the more promising treatment. Here we study various asymptotic results associated with the proposed design and compare its performance with a single stage equal allocation design. Also, the proposed procedure is illustrated through a real-life data set on carcinoma in the oropharynx.

Generalisations of a Bayesian decision-theoretic randomisation procedure and the impact of delayed responses

The design of sequential experiments and, in particular, randomised controlled trials involves a trade-off between operational characteristics such as statistical power, estimation bias and patient benefit. The family of randomisation procedures referred to as Constrained Randomised Dynamic Programming (CRDP), which is set in the Bayesian decision-theoretic framework, can be used to balance these competing objectives. A generalisation and novel interpretation of CRDP is proposed to highlight its inherent flexibility to adapt to a variety of practicalities and align with individual trial objectives. CRDP, as with most response-adaptive randomisation procedures, hinges on the limiting assumption of patient responses being available before allocation of the next patient. This forms one of the greatest barriers to their implementation in practice which, despite being an important research question, has not received a thorough treatment. Therefore, motivated by the existing gap between the theory of response-adaptive randomisation (which is abundant with proposed methods in the immediate response setting) and clinical practice (in which responses are typically delayed), the performance of CRDP in the presence of fixed and random delays is evaluated. Simulation results show that CRDP continues to offer patient benefit gains over alternative procedures and is relatively robust to delayed responses. To compensate for a fixed delay, a method which adjusts the time horizon used in the optimisation objective is proposed and its performance illustrated.

Baseline characteristics for CLARITY AD: A phase 3 placebo‐controlled, double‐blind, parallel‐group, 18‐month study evaluating lecanemab (ban2401) in early Alzheimer's disease

AbstractBackgroundLecanemab (BAN2401) is a humanized IgG1 monoclonal antibody that preferentially targets soluble aggregated Aβ. In a recent multinational, phase 2, double‐blind, placebo‐controlled, study utilizing a Bayesian design with response adaptive randomization, 10‐mg/kg bi‐weekly lecanemab reduced brain Ab levels and slowed decline in cognition and function on Alzheimer’s Disease Composite Score (ADCOMS), Alzheimer’s Disease Assessment Scale‐Cognitive Subscale (ADAS‐cog), and Clinical Dementia Rating‐Sum of Boxes (CDR‐SB) at 18 months in participants with early Alzheimer’s disease (AD). Following these results, the CLARITY AD phase 3 study was designed to confirm the efficacy and safety of lecanemab in participants with early Alzheimer’s disease. Herein, we describe the baseline characteristics for participants in the ongoing CLARITY AD study.MethodsCLARITY AD is an 18‐month treatment, multicenter, double‐blind, placebo‐controlled, parallel‐group study with open‐label extension in participants with early AD. Baseline clinical and demographic data were summarized descriptively and compared to the lecanemab phase 2 study population.ResultsAs of a data cutoff of January 6, 2021, 1540 participants were randomized in CLARITY AD. The median age of participants was 72 years (range: 50‐90 years), with 81% of participants 65 years of age or older. Overall, 52% of participants were female and 75% were Caucasian. Mean (SD) baseline values for clinical endpoints were 3.2 (1.3) for CDR‐SB, 0.4 (0.1) for ADCOMS, 25.4 (7.3) for ADAS‐Cog, 25.6 (2.2) for MMSE, and 0.6 (0.2) for Global CDR. Aggregate baseline characteristics are similar to the lecanemab phase 2 study (median age 72 years [range: 50‐90 years]; 80% 65 years of age or older; 50% female; 90% Caucasian; clinical endpoints: 3.0 [1.4] for CDR‐SB, 0.4 [0.2] for ADCOMS, 22.2 [7.4] for ADAS‐Cog, 25.6 [2.4] for MMSE, and 0.6 [0.2] for Global CDR). Comparisons of the study populations will be presented.ConclusionBuilding on the encouraging findings from the lecanemab phase 2 study, the phase 3 CLARITY AD study is designed to confirm clinical efficacy and safety of lecanemab versus placebo in participants with early AD. Baseline characteristics after enrollment of 1540 participants are consistent with the phase 2 study and representative of an early AD population. Enrollment is ongoing.

CTNI-05. PRELIMINARY RESULTS OF THE NERATINIB ARM IN THE INDIVIDUALIZED SCREENING TRIAL OF INNOVATIVE GLIOBLASTOMA THERAPY (INSIGHT): A PHASE II PLATFORM TRIAL USING BAYESIAN ADAPTIVE RANDOMIZATION

Abstract BACKGROUND EGFR is amplified in over 50% of glioblastoma and 20-30% have EGFRvIII mutations. Neratinib is a potent inhibitor of EGFR/HER2 approved for metastatic HER2+ breast cancer. To efficiently evaluate the potential impact of neratinib on overall survival (OS) in newly-diagnosed glioblastoma and to simultaneously develop information regarding potential genomic biomarker associations, neratinib was included as an arm on the Individualized Screening Trial of Innovative Glioblastoma Therapy (INSIGhT) trial. INSIGhT is a phase II platform trial using response adaptive randomization and deep genomic profiling to more efficiently test experimental agents in MGMT unmethylated glioblastoma and accelerate identification of novel therapies for phase III testing. Initial randomization was equal between neratinib, control, and two other experimental arms but subsequent randomization was adapted based on efficacy as determined by progression-free survival (PFS). We report preliminary results for the neratinib arm. METHODS Patients with newly diagnosed MGMT-unmethylated glioblastoma were randomized to receive either radiotherapy with concomitant and adjuvant temozolomide or standard radiochemotherapy followed by adjuvant neratinib (240 mg daily). Treatment continued until progression or development of unacceptable toxicities. The primary endpoint was OS. Association between neratinib efficacy and EGFR amplification was also investigated. RESULTS There were 144 patients (70 control; 74 neratinib). Neratinib was reasonably well-tolerated with no new toxicity signals identified. PFS was compared (HR 0.84; p=0.38, logrank test – not significant) between the neratinib (median 6.05 months) and control (median 5.82 months) arms. For patients EGFR pathway activation the PFS HR was 0.53 (p-value=0.03 – significant, median PFS: neratinib, 6.21 months, control, 5.26 months). However, there was no significant improvement in OS in EGFR amplified/mutated patients (HR 1.05; p-value 0.87) between neratinib (median 14.2) compared to the control arm (median 14.6). CONCLUSION Neratinib prolonged PFS in the EGFR positive subpopulation but there was no overall PFS benefit, or any OS improvement.

Using biomarkers to allocate patients in a response-adaptive clinical trial

In this paper, we discuss a response adaptive randomization method, and why it should be used in clinical trials for rare diseases compared to a randomized controlled trial with equal fixed randomization. The developed method uses a patient’s biomarkers to alter the allocation probability to each treatment, in order to emphasize the benefit to the trial population. The method starts with an initial burn-in period of a small number of patients, who with equal probability, are allocated to each treatment. We then use a regression method to predict the best outcome of the next patient, using their biomarkers and the information from the previous patients. This estimated best treatment is assigned to the next patient with high probability. A completed clinical trial for the effect of catumaxomab on the survival of cancer patients is used as an example to demonstrate the use of the method and the differences to a controlled trial with equal allocation. Different regression procedures are investigated and compared to a randomized controlled trial, using efficacy and ethical measures.

A novel statistical test for treatment differences in clinical trials using a response-adaptive forward-looking Gittins Index Rule.

The most common objective for response-adaptive clinical trials is to seek to ensure that patients within a trial have a high chance of receiving the best treatment available by altering the chance of allocation on the basis of accumulating data. Approaches that yield good patient benefit properties suffer from low power from a frequentist perspective when testing for a treatment difference at the end of the study due to the high imbalance in treatment allocations. In this work we develop an alternative pairwise test for treatment difference on the basis of allocation probabilities of the covariate-adjusted response-adaptive randomization with forward-looking Gittins Index (CARA-FLGI) Rule for binary responses. The performance of the novel test is evaluated in simulations for two-armed studies and then its applications to multiarmed studies are illustrated. The proposed test has markedly improved power over the traditional Fisher exact test when this class of nonmyopic response adaptation is used. We also find that the test's power is close to the power of a Fisher exact test under equalrandomization.

The design of a Bayesian adaptive clinical trial of tranexamic acid in severely injured children

BackgroundTrauma is the leading cause of death and disability in children in the USA. Tranexamic acid (TXA) reduces the blood transfusion requirements in adults and children during surgery. Several studies have evaluated TXA in adults with hemorrhagic trauma, but no randomized controlled trials have occurred in children with trauma. We propose a Bayesian adaptive clinical trial to investigate TXA in children with brain and/or torso hemorrhagic trauma.Methods/designWe designed a double-blind, Bayesian adaptive clinical trial that will enroll up to 2000 patients. We extend the traditional Emax dose-response model to incorporate a hierarchical structure so multiple doses of TXA can be evaluated in different injury populations (isolated head injury, isolated torso injury, or both head and torso injury). Up to 3 doses of TXA (15 mg/kg, 30 mg/kg, and 45 mg/kg bolus doses) will be compared to placebo. Equal allocation between placebo, 15 mg/kg, and 30 mg/kg will be used for an initial period within each injury group. Depending on the dose-response curve, the 45 mg/kg arm may open in an injury group if there is a trend towards increasing efficacy based on the observed relationship using the data from the lower doses. Response-adaptive randomization allows each injury group to differ in allocation proportions of TXA so an optimal dose can be identified for each injury group. Frequent interim stopping periods are included to evaluate efficacy and futility. The statistical design is evaluated through extensive simulations to determine the operating characteristics in several plausible scenarios. This trial achieves adequate power in each injury group.DiscussionThis trial design evaluating TXA in pediatric hemorrhagic trauma allows for three separate injury populations to be analyzed and compared within a single study framework. Individual conclusions regarding optimal dosing of TXA can be made within each injury group. Identifying the optimal dose of TXA, if any, for various injury types in childhood may reduce death and disability.