Adaptive Dose-finding Designs Research Articles

Dose Finding in Oncology Trials Guided by Ordinal Toxicity Grades Using Continuous Dose Levels.

We present a Bayesian adaptive design for dose finding in oncology trials with application to a first-in-human trial. The design is based on the escalation with overdose control principle and uses an intermediate grade 2 toxicity in addition to the traditional binary indicator of dose-limiting toxicity (DLT) to guide the dose escalation and de-escalation. We model the dose-toxicity relationship using the proportional odds model. This assumption satisfies an important ethical concern when a potentially toxic drug is first introduced in the clinic; if a patient experiences grade 2 toxicity at the most, then the amount of dose escalation is lower relative to that wherein if this patient experienced a maximum of grade 1 toxicity. This results in a more careful dose escalation. The performance of the design was assessed by deriving the operating characteristics under several scenarios for the true MTD and expected proportions of grade 2 toxicities. In general, the trial design is safe and achieves acceptable efficiency of the estimated MTD for a planned sample size of twenty patients. At the time of writing this manuscript, twelve patients have been enrolled to the trial.

Adaptive designs were primarily used but inadequately reported in early phase drug trials

BackgroundFaced with the high cost and limited efficiency of classical randomized controlled trials, researchers are increasingly applying adaptive designs to speed up the development of new drugs. However, the application of adaptive design to drug randomized controlled trials (RCTs) and whether the reporting is adequate are unclear. Thus, this study aimed to summarize the epidemiological characteristics of the relevant trials and assess their reporting quality by the Adaptive designs CONSORT Extension (ACE) checklist.MethodsWe searched MEDLINE, EMBASE, Cochrane Central Register of Controlled Trials (CENTRAL) and ClinicalTrials.gov from inception to January 2020. We included drug RCTs that explicitly claimed to be adaptive trials or used any type of adaptative design. We extracted the epidemiological characteristics of included studies to summarize their adaptive design application. We assessed the reporting quality of the trials by Adaptive designs CONSORT Extension (ACE) checklist. Univariable and multivariable linear regression models were used to the association of four prespecified factors with the quality of reporting.ResultsOur survey included 108 adaptive trials. We found that adaptive design has been increasingly applied over the years, and was commonly used in phase II trials (n = 45, 41.7%). The primary reasons for using adaptive design were to speed the trial and facilitate decision-making (n = 24, 22.2%), maximize the benefit of participants (n = 21, 19.4%), and reduce the total sample size (n = 15, 13.9%). Group sequential design (n = 63, 58.3%) was the most frequently applied method, followed by adaptive randomization design (n = 26, 24.1%), and adaptive dose-finding design (n = 24, 22.2%). The proportion of adherence to the ACE checklist of 26 topics ranged from 7.4 to 99.1%, with eight topics being adequately reported (i.e., level of adherence ≥ 80%), and eight others being poorly reported (i.e., level of adherence ≤ 30%). In addition, among the seven items specific for adaptive trials, three were poorly reported: accessibility to statistical analysis plan (n = 8, 7.4%), measures for confidentiality (n = 14, 13.0%), and assessments of similarity between interim stages (n = 25, 23.1%). The mean score of the ACE checklist was 13.9 (standard deviation [SD], 3.5) out of 26. According to our multivariable regression analysis, later published trials (estimated β = 0.14, p < 0.01) and the multicenter trials (estimated β = 2.22, p < 0.01) were associated with better reporting.ConclusionAdaptive design has shown an increasing use over the years, and was primarily applied to early phase drug trials. However, the reporting quality of adaptive trials is suboptimal, and substantial efforts are needed to improve the reporting.

Analysis on efficacy of magnetic resonance lymphangiography using INV-001 in healthy beagle dogs

We aimed to conduct a proof-of-concept study of INV-001 in visualizing lymphatic vessels and nodes without venous contamination and to determine the optimal dose condition of INV-001 for magnetic resonance lymphangiography (MRL) in healthy beagles. MRL was performed using a 3.0-Tesla (T) whole body clinical magnetic resonance imaging (MRI) scanner. A dose-finding study of INV-001 for MRL in beagles (N = 6) was carried out according to an adaptive optimal dose finding design. For the reproducibility study (N = 6), MRL was conducted at selected INV-001 doses (0.056 and 0.112 mg Fe/kg) with a 15 mM concentration. Additionally, an excretion study (N = 3) of INV-001 was conducted by analyzing T1, T2, and T2* maps of the liver and kidney 48 h post-administration. INV-001 administration at doses of 0.056 and 0.112 mg Fe/kg (concentration: 15 mM) consistently demonstrated the visualization of contrast-enhanced lymphatic vessels and nodes without venous contamination in the beagles. The contrast enhancement effect was highest at 30 min after INV-001 administration, then gradually decreasing. No toxicity-related issues were identified during the study. After 48 h, the T1, T2, and T2* values in the liver and both kidneys were found to be comparable to the pre-administration values, indicating thorough INV-001 excretion. The optimal dosing conditions of INV-001 for MRL for contrast-enhanced visualization of lymphatic vessels and nodes exclusively with no venous contamination in beagles was determined to be 0.056 mg Fe/kg with a 15 mM concentration.

A systematic survey of adaptive trials shows substantial improvement in methods is needed

ObjectivesTo investigate the design, conduct, and analysis of adaptive trials through a systematic survey and provide recommendations for future adaptive trials. Study Design and SettingWe systematically searched MEDLINE, EMBASE, Cochrane Central Register of Controlled Trials, and ClinicalTrials.gov databases up to January 2020. We included trials that were self-described as adaptive trials or applied adaptive designs. We identified three frequently used adaptive designs and summarized their methodological details in terms of design, conduct, and analysis. Lastly, we provided recommendations for future adaptive trials. ResultsWe included a total of 128 trials in this study. The primary motivations for using adaptive design were to speed up the trials and facilitate decision-making (n = 29, 31.5%). The three most frequently used methods were group sequential design (GSD) (n = 71, 55.5%), adaptive dose-finding design (ADFD) (n = 35, 27.3%), and adaptive randomization design (ARD) (n = 26, 20.3%). The timing and frequency of interim analysis were detailed in three-fourths of the GSD trials (n = 55, 77.5%) and in half of the ADFD trials (n = 19, 54.3%); however, more than half of the ARD trials (n = 15, 57.7%) did not provide this information. Some trials selected a different outcome than the primary outcome for interim analysis (GSD: n = 7, 12.7%; ADFD: n = 8, 27.6%; ARD: n = 7, 50.0%), but the majority of these trials did not provide explicit reasons for this choice (GSD: n = 7, 100.0%; ADFD: n = 7, 87.5%; ARD: n = 5, 71.4%). More than half (n = 76, 59.4%) of trials did not mention the accessibility of supporting documents, and two-thirds (n = 86, 67.2%) did not state the establishment of independent data monitoring committees (IDMCs). Moreover, unplanned adjustments were observed during the conduct of one-sixth adaptive trials (n = 22, 17.2%). Based on our findings, we provide 14 recommendations for improving adaptive trials in the future. ConclusionSubstantial improvements were needed in methods of adaptive trials, particularly in the areas of interim analysis, the establishment of independent data monitoring committees, and unplanned adjustments. In this study, we offer recommendations from both general and specific aspects for researchers to carefully design, conduct, and analyze adaptive trials.

Understanding Research Methods: Up-and-down Designs for Dose-finding

For the task of estimating a target benchmark dose such as the ED50 (the dose that would be effective for half the population), an adaptive dose-finding design is more effective than the standard approach of treating equal numbers of patients at a set of equally spaced doses. Up-and-down is the most popular family of dose-finding designs and is in common use in anesthesiology. Despite its widespread use, many aspects of up-and-down are not well known, implementation is often misguided, and standard, up-to-date reference material about the design is very limited. This article provides an overview of up-and-down properties, recent methodologic developments, and practical recommendations, illustrated with the help of simulated examples. Additional reference material is offered in the Supplemental Digital Content.

Lessons Learned From Implementing a Novel Bayesian Adaptive Dose-Finding Design in Advanced Pancreatic Cancer.

Novel Bayesian adaptive designs provide an effective way to improve clinical trial efficiency. These designs are superior to conventional methods, but implementing them can be challenging. The aim of this article was to describe what we learned while applying a novel Bayesian phase I-II design in a recent trial. The primary goal of the trial was to optimize radiation therapy (RT) dose among three levels (low, standard, and high), given either with placebo (P) or an investigational agent (A), for treating locally advanced, radiation-naive pancreatic cancer, deemed appropriate for RT rather than surgery. Up to 48 patients were randomly assigned fairly between RT plus P and RT plus A, with RT dose-finding done within each arm using the late-onset efficacy-toxicity design on the basis of two coprimary end points, tumor response and dose-limiting toxicity, both evaluated at up to 90 days. The random assignment was blinded, but within each arm, unblinded RT doses were chosen adaptively using software developed within the institution. Implementing the design involved double-blind balance-restricted random assignment, real-time assessment of patient outcomes to evaluate the efficacy-toxicity trade-off for each RT dose in each arm to optimize each patient's RT dose adaptively, and transition from a single-center trial to a multicenter trial. We present lessons learned and illustrative documentation. Implementing novel Bayesian adaptive trial designs requires close collaborations between physicians, pharmacists, statisticians, data managers, and sponsors. The process is difficult but manageable and essential for efficient trial conduct. Close collaboration during trial conduct is a key component of any trial that includes real-time adaptive decision rules.

Adaptive Designs in Clinical Trials

The traditional clinical trial is a rigid channel of drug development process. After successful Phase 1 trials and Phase 2 trial with sufficient efficacy and safety, the drugs goes into Phase 3 trials, where it is compared with a placebo or standard treatment (control). Performing this for each new drug separately need more manpower (patients/participant), money (financial resources), and time. On November 2019, the Food and Drug Administration issued a final guidance for the industries entitled, “Adaptive Designs for Clinical Trials of Drugs and Biologics.” They defined adaptive design as “a clinical trial design that allows for prospectively planned modifications to one or more aspects of the design based on accumulating data from subjects in the trial.” In simple words, adaptive design means a design that permits modifications of “the trial procedures and/or statistical procedures of the trial” after its beginning without compromising the validity and the integrity of a trial. There are nine types of adaptive designs that includes adaptive randomization design, group sequential design, sample size re-estimation design, drop-the-losers design, adaptive dose finding design, adaptive treatment-switching design, biomarker-adaptive design, adaptive-hypotheses design, and adaptive seamless Phase II/III designs which has been explained in detail, in this review article. Although the concept of adaptive design has been there from long before but one of the major disadvantages it faces is the lack of uniformity.

Bias induced by adaptive dose-finding designs

There is a long literature on bias in maximum likelihood estimators. Here we demonstrate that adaptive dose-finding procedures (such as Continual Reassessment Methods, Up-and-Down and Interval Designs) themselves induce bias. In particular, with Bernoulli responses and dose assignments that depend on prior responses, we provide an explicit formula for the bias of observed response rates. We illustrate the patterns of bias for designs that aim to concentrate dose allocations around a target dose, which represents a specific quantile of a cumulative response-threshold distribution. For such designs, bias tends to be positive above the target dose and negative below it. To our knowledge, this property of dose-finding designs has not previously been recognized by design developers. We discuss the implications of this bias and suggest a simple shrinkage mitigation formula that improves estimation at doses away from the target.

A Bayesian Adaptive Design in Cancer Phase I Trials using Dose Combinations in the Presence of a Baseline Covariate.

A Bayesian adaptive design for dose finding of a combination of two drugs in cancer phase I clinical trials that takes into account patients heterogeneity thought to be related to treatment susceptibility is described. The estimation of the maximum tolerated dose (MTD) curves is a function of a baseline covariate using two cytotoxic agents. A logistic model is used to describe the relationship between the doses, baseline covariate, and the probability of dose limiting toxicity (DLT). Trial design proceeds by treating cohorts of two patients simultaneously using escalation with overdose control (EWOC), where at each stage of the trial, the next dose combination corresponds to the α quantile of the current posterior distribution of the MTD of one of two agents at the current dose of the other agent and the next patient's baseline covariate value. The MTD curves are estimated as function of Bayes estimates of the model parameters at the end of trial. Average DLT, pointwise average bias, and percent of dose recommendation at dose combination neighborhoods around the true MTD are compared to the design that uses the covariate to the one that ignores the baseline characteristic. We also examine the performance of the approach under model misspecifications for the true dose-toxicity relationship. The methodology is further illustrated by the case of a pre-specified discrete set of dose combinations.

An Adaptive Dose-Finding Design Based on Both Safety and Immunologic Responses in Cancer Clinical Trials

ABSTRACTDose-finding in cancer clinical trials has been dominated by algorithmic designs on the principle that the highest tolerable dose is also the most effective dose. This assumption no longer applies to the biologic treatments that are characterized by different toxicity and/or efficacy profiles to the extent that the best therapeutic dose might be well below any dose that produces serious toxicity. As such, we propose a two-stage design with focus on immunotherapy trials, incorporating both safety and efficacy information. The first stage establishes the safety profile of each dose, with escalation decisions based on likelihood principles. Continuous immunologic outcomes are used to evaluate the relative efficacy of the doses. The second stage employs an adaptive randomization to assign patients to doses showing higher efficacy. Safety is being continuously monitored throughout Stage 2, where some doses may be ‘closed’ due to unacceptable toxicity. The proposed design is compared to the modified toxicity probability interval (mTPI) design using percent dose allocation and estimation of outcomes under different scenarios. We show that by using an efficacy-driven adaptive randomization with safety constraints, the allocation distribution is skewed towards more efficacious doses, and thus limit the number of patients exposed to toxic or non-therapeutic doses. Supplementary materials for this article are available online.

Dose-finding design with repeated continuous toxicity and efficacy data

Introduction Despite the characteristics of molecularly targeted agents (MTA) in oncology, the recommended dose remains typically based on the binary dose-limiting toxicity (DLT) defined as the dose associated with a certain probability of DLT during the first cycle of treatment. This conventional definition may then not be appropriate for MTAs with prolonged administration, for which late and multiple moderate toxicities also deserve attention. Different authors explored more appropriate and richer toxicity endpoints than DLT. The idea of toxicity score accounting for several toxicity types and grades seems interesting for MTA. Methods In this study, we developed a novel adaptive dose-finding design for inclusion of a single continuous biomarker outcome in addition to repeated quasi-continuous toxicity score. A joint multivariate Gaussian model was used for toxicity and efficacy outcomes, considering that toxicity score follows a truncated positive Gaussian distribution. The correlation between toxicity and efficacy outcomes is modelled via a matrix of variance-covariance. Maximum likelihood estimators (MLE) were used to estimate the optimal dose for further trials. This dose was defined as the dose associated with a maximum efficacy and with a toxicity score at each cycle inferior or equal to a given toxicity-score threshold. First, the robustness of the MLE approach was assessed by using different scenarios of toxicity-dose and efficacy-dose relationships. The number of patients was fixed at 10, 20, 30, and 50 in order to study the convergence of MLE. We replicated each scenario 1000 times. Different variance values for toxicity score have been investigated. We used the bias defined by the difference between true and estimated values as metrics of robustness. Secondly, the performance of the dose-finding design is assessed using simulation study. Different scenarios of toxicity-dose and efficacy-dose relationships were used. The performance of the design was measured using the capacity of the design to identify correctly the optimal dose. Results Concerning the robustness of the approach : The bias of all estimators depends on the variance value and the number of patients included. It converges to 0 when the number of subject increases ; for low variance values. The performance of the design presents good results for the overall of scenarios. Conclusion Using repeated toxicity score and efficacy data in dose-finding trials provide information to estimate the optimal dose.

Adaptive design clinical trials: a review of the literature and ClinicalTrials.gov

ObjectivesThis review investigates characteristics of implemented adaptive design clinical trials and provides examples of regulatory experience with such trials.DesignReview of adaptive design clinical trials in EMBASE, PubMed, Cochrane Registry of...

A Bayesian adaptive design for cancer phase I trials using a flexible range of doses

ABSTRACTWe present a Bayesian adaptive design for dose finding in cancer phase I clinical trials. The goal is to estimate the maximum tolerated dose (MTD) after possible modification of the dose range during the trial. Parametric models are used to describe the relationship between the dose and the probability of dose-limiting toxicity (DLT). We investigate model reparameterization in terms of the probabilities of DLT at the minimum and maximum available doses at the start of the trial. Trial design proceeds using escalation with overdose control (EWOC), where at each stage of the trial we seek the dose of the agent such that the posterior probability of exceeding the MTD of this agent is bounded by a feasibility bound. At any time during the trial, we test whether the MTD is below or above the minimum and maximum doses, respectively. If during the trial there is evidence that the MTD is outside the range of doses, we extend the range of doses and complete the trial with the planned sample size. At the end of the trial, a Bayes estimate of the MTD is proposed. We evaluate design operating characteristics in terms of safety of the trial design and efficiency of the MTD estimate under various scenarios and model misspecification. The methodology is further compared to the original EWOC design. We showed by comprehensive simulation studies that the proposed method is safe and can estimate the MTD more efficiently than the original EWOC design.

A Versatile Adaptive Dose-Finding Design Based on Multiple Endpoints

ABSTRACTIn drug development, assessment of the dose–response relationship gives essential insight into which doses are efficacious, yet safe. These relationships are often estimated by parametric, nonlinear models. A hurdle for trials that aim to estimate these models, is that the optimal study design depends on the unknown model parameters. A solution is to use adaptive designs. Ideally, such procedures evaluate a safety and an efficacy endpoint to get the best understanding of the therapeutic window. Many such designs have already been proposed. However, the outcomes are typically required to be of a specific type and the correlation between them has to be modeled explicitly. We propose a more general and simpler approach. Separate models can be specified for each endpoint, without restrictions on the type of model or outcome. The dependence between the outcomes is considered, without having to specify its structure. A new penalty function is introduced to balance the need for optimal precision and safe dose assignments. The method is illustrated by an example study in depression. Simulation results indicate that the adaptive design is more efficient than a fixed design. Use of the penalty function decreases estimation efficiency, but increases dose assignment to favorable doses. Supplementary materials for this article are available online.

Novel dose-finding designs and considerations on practical implementations in oncology clinical trials

ABSTRACTOne of the main objectives in phase I oncology trials is to evaluate safety and tolerability of an experimental treatment by estimating the maximum tolerated dose (MTD) based on the rate of dose-limiting toxicities (DLT). To meet emerging challenges in dose-finding studies, over the past two decades, extensive research has been conducted by statistical and medical researchers to create innovative dose finding designs that perform better than the standard 3 + 3 design, which often exhibits undesirable statistical and operational properties. However, clinical implementation and practical usage of these new designs have been limited. This article begins with a review of the most recent literature and then provides some perspectives on implementing novel adaptive dose finding designs in oncology phase I trials from a pharmaceutical industry perspective. Statistical planning and logistical considerations on how to effectively execute such designs in multi-center clinical trials are discussed using two recent case studies.

Simulation work in adaptive dose-finding designs to identify dose-intervals that achieve targets in multiple endpoints

The DIL-frequency study aims to establish the optimal dose and repeat dosing schedule (interval) to administer ultra-low dose Interlekin-2 to participants with type 1 diabetes in order to target a desired regulatory T cell (Treg) increase and increased CD25 expression on Tregs and a minimal increase of CD4 T effector cells (Teffs). Before the start of the trial, the available doses are chosen within a safe range. The interval is chosen to be within the range 2 to 14 days between doses. The study design has an initial learning phase with pairs of patients assigned to six pre-assigned dose-intervals. There are three further phases of eight patients, participants are given dose-interval combinations that give an expected Treg and CD25 increase closest to the targets of 20% and 25% respectively, provided that there is not an increase in Teffs above a specified level. The primary endpoints are assessed when steady state is achieved. A variety of strategies were considered to select the dose-interval combinations for the next cohort of patients and to identify the recommended dose-interval. The method chosen was to select the dose-interval that would be closest to the target increases of Treg and CD25. The estimated coefficients from the multivariate regression analysis gave equations to predict the estimated increase in Treg and CD25 for any given dose and interval. The targets are substituted into the equations and solved to find the estimated best dose-interval. The results of the simulations are presented.

Exposure Adjusted Continuous Reassessment Method (EACRM) - an Adaptive Design Incorporating Time to Toxicity Event for Phase I Dose Finding Studies.

Finding the maximum tolerated dose (MTD) is the most important primary objective for Phase I oncology studies. The drug exposure and the toxicity event are the two components required for characterization of the MTD. Phase I trials currently require multiple subjects treated at numerous dose levels, with many of these subjects receiving doses of drug at less than predicted efficacious exposures. Further, these studies employ predefined, step-wise increments in doses that can result in inaccurate predictions of the MTD. Improvements to conventional adaptive dose finding designs to attain more accurate information on the extent of exposure and the time to event, with a more dynamic enrollment to facility better decision making are needed and have been the focus of previous published efforts. The EACRM dynamically extends the conventional adaptive dose finding designs by incorporating dose limiting toxicity (DLT) events as well as at-the-event information from each patient. An Accelerated Failure Time model with some tweaking to update the time to DLT for subjects who did not complete the entire DLT assessment period at particular time point was developed. Simulations, under a variety of assumptions, have indicated that the EACRM performed equal or better than the traditional 3+3 design in identifying the MTD and also outperformed the 3+3 design in other important operating characteristics such as the length of study, the number of subjects treated at above the MTD, and the total number of subjects.

Adaptive designs for identifying optimal biological dose for molecularly targeted agents.

Background Traditionally, the purpose of a dose-finding design in cancer is to find the maximum tolerated dose based solely on toxicity. However, for molecularly targeted agents, little toxicity may arise within the therapeutic dose range and the dose-response curves may not be monotonic. This challenges the principle that more is better, which is widely accepted for conventional chemotherapy. Methods We propose three adaptive dose-finding designs for trials evaluating molecularly targeted agents, for which the dose-response curves are unimodal or plateaued. The goal of these designs is to find the optimal biological dose, which is defined as the lowest dose with the highest rate of efficacy while safe. The first proposed design is parametric and assumes a logistic dose-efficacy curve for dose finding, the second design is nonparametric and uses the isotonic regression to identify the optimal biological dose, and the third design has the spirit of a 'semiparametric' approach by assuming a logistic model only locally around the current dose. Results We conducted extensive simulation studies to investigate the operating characteristics of the proposed designs. Simulation studies show that the nonparametric and semiparametric designs have good operating characteristics for finding the optimal biological dose. Limitations The proposed designs assume a binary endpoint. Extension of the proposed designs to ordinal and time-to-event endpoints is worth further investigation. Conclusion Among the three proposed designs, the nonparametric and semiparametric designs yield consistently good operating characteristics and thus are recommended for practical use. The software to implement these two designs is available for free download at http://odin.mdacc.tmc.edu/~yyuan/ .

Dose finding with drug combinations in cancer phase I clinical trials using conditional escalation with overdose control.

We present a Bayesian adaptive design for dose finding of a combination of two drugs in cancer phase I clinical trials. The goal is to estimate the maximum tolerated dose (MTD) as a curve in the two-dimensional Cartesian plane. We use a logistic model to describe the relationship between the doses of the two agents and the probability of dose limiting toxicity. The model is re-parameterized in terms of parameters clinicians can easily interpret. Trial design proceeds using univariate escalation with overdose control, where at each stage of the trial, we seek a dose of one agent using the current posterior distribution of the MTD of this agent given the current dose of the other agent. At the end of the trial, an estimate of the MTD curve is proposed as a function of Bayes estimates of the model parameters. We evaluate design operating characteristics in terms of safety of the trial design and percent of dose recommendation at dose combination neighborhoods around the true MTD curve. We also examine the performance of the approach under model misspecifications for the true dose-toxicity relationship.

Bayesian data augmentation dose finding with continual reassessment method and delayed toxicity

A major practical impediment when implementing adaptive dose-finding designs is that the toxicity outcome used by the decision rules may not be observed shortly after the initiation of the treatment. To address this issue, we propose the data augmentation continual re-assessment method (DA-CRM) for dose finding. By naturally treating the unobserved toxicities as missing data, we show that such missing data are nonignorable in the sense that the missingness depends on the unobserved outcomes. The Bayesian data augmentation approach is used to sample both the missing data and model parameters from their posterior full conditional distributions. We evaluate the performance of the DA-CRM through extensive simulation studies, and also compare it with other existing methods. The results show that the proposed design satisfactorily resolves the issues related to late-onset toxicities and possesses desirable operating characteristics: treating patients more safely, and also selecting the maximum tolerated dose with a higher probability. The new DA-CRM is illustrated with two phase I cancer clinical trials.