Randomized phase II trials: inevitable or inadvisable?

Two-stage response adaptive randomization designs for multi-arm trials with binary outcome

In recent years, adaptive randomization methods have gained significant popularity in clinical research and trial design due to their ability to provide both efficiency and flexibility in adjusting the statistical procedures of ongoing clinical trials. For a study to compare multiple treatments, a multi-arm two-stage design could be utilized to select the best treatment from the first stage and further compare that treatment with control in the second stage. The traditional design used equal randomization in both stages. To better utilize the interim results from the first stage, we propose to develop response adaptive randomization two-stage designs for a multi-arm clinical trial with binary outcome. Two allocation methods are considered: (1) an optimal allocation based on a sequential design; (2) the play-the-winner rule. Optimal multi-arm two-stage designs are obtained under three criteria: minimizing the expected number of failures, minimizing the average expected sample size, and minimizing the expected sample size under the null hypothesis. Simulation studies show that the proposed adaptive design based on the play-the-winner rule has good performance. A phase II trial for patients with pancreas adenocarcinoma and a germline BRCA/ PALB2 mutation was used to illustrate the application of the proposed response adaptive randomization designs.

An evolutionary algorithm for the direct optimization of covariate balance between nonrandomized populations

Matching reduces confounding bias in comparing the outcomes of nonrandomized patient populations by removing systematic differences between them. Under very basic assumptions, propensity score (PS) matching can be shown to eliminate bias entirely in estimating the average treatment effect on the treated. In practice, misspecification of the PS model leads to deviations from theory and matching quality is ultimately judged by the observed post-matching balance in baseline covariates. Since covariate balance is the ultimate arbiter of successful matching, we argue for an approach to matching in which the success criterion is explicitly specified and describe an evolutionary algorithm to directly optimize an arbitrary metric of covariate balance. We demonstrate the performance of the proposed method using a simulated dataset of 275,000 patients and 10 matching covariates. We further apply the method to match 250 patients from a recently completed clinical trial to a pool of more than 160,000 patients identified from electronic health records on 101 covariates. In all cases, we find that the proposed method outperforms PS matching as measured by the specified balance criterion. We additionally find that the evolutionary approach can perform comparably to another popular direct optimization technique based on linear integer programming, while having the additional advantage of supporting arbitrary balance metrics. We demonstrate how the chosen balance metric impacts the statistical properties of the resulting matched populations, emphasizing the potential impact of using nonlinear balance functions in constructing an external control arm. We release our implementation of the considered algorithms in Python.

Estimation of treatment effects in early-phase randomized clinical trials involving external control data

There are good reasons to perform a randomized controlled trial (RCT) even in early phases of clinical development. However, the low sample sizes in those settings lead to high variability of the treatment effect estimate. The variability could be reduced by adding external control data if available. For the common setting of suitable subject-level control group data only available from one external (clinical trial or real-world) data source, we evaluate different analysis options for estimating the treatment effect via hazard ratios. The impact of the external control data is usually guided by the level of similarity with the current RCT data. Such level of similarity can be determined via outcome and/or baseline covariate data comparisons. We provide an overview over existing methods, propose a novel option for a combined assessment of outcome and baseline data, and compare a selected set of approaches in a simulation study under varying assumptions regarding observable and unobservable confounder distributions using a time-to-event model. Our various simulation scenarios also reflect the differences between external clinical trial and real-world data. Data combinations via simple outcome-based borrowing or simple propensity score weighting with baseline covariate data are not recommended. Analysis options which conflate outcome and baseline covariate data perform best in our simulation study.

Q-HAM: a multicenter upfront randomized phase II trial of quizartinib and high-dose Ara-C plus mitoxantrone in relapsed/refractory AML with FLT3-ITD

BACKGROUND

About 50% of older patients with acute myeloid leukemia (AML) fail to attain complete remission (CR) following cytarabine plus anthracycline-based induction therapy. Salvage chemotherapy regimens are based on high-dose cytarabine (HiDAC), which is frequently combined with mitoxantrone (HAM regimen). However, CR rates remain low, with less than one-third of the patients achieving a CR. FLT3-ITD has consistently been identified as an unfavorable molecular marker in both relapsed and refractory (r/r)-AML. One-quarter of patients who received midostaurin are refractory to induction therapy and relapse rate at 2 years exceeds 40%. The oral second-generation bis-aryl urea tyrosine kinase inhibitor quizartinib is a very selective FLT3 inhibitor, has a high capacity for sustained FLT3 inhibition, and has an acceptable toxicity profile.

METHODS

In this multicenter, upfront randomized phase II trial, all patients receive quizartinib combined with HAM (cytarabine 3g/m2 bidaily day one to day three, mitoxantrone 10mg/m2 days two and three) during salvage therapy. Efficacy is assessed by comparison to historical controls based on the matched threshold crossing approach with achievement of CR, complete remission with incomplete hematologic recovery (CRi), or complete remission with partial recovery of peripheral blood counts (CRh) as primary endpoint. During consolidation therapy (chemotherapy and allogeneic hematopoietic cell transplantation), patients receive either prophylactic quizartinib therapy or measurable residual disease (MRD)-triggered preemptive continuation therapy with quizartinib according to up-front randomization. The matched threshold crossing approach is a novel study-design to enhance the classic single-arm trial design by including matched historical controls from previous clinical studies. It overcomes common disadvantages of single-armed and small randomized studies, since the expected outcome of the observed study population can be adjusted based on the matched controls with a comparable distribution of known prognostic and predictive factors. Furthermore, balanced treatment groups lead to stable statistical models. However, one of the limitations of our study is the inability to adjust for unobserved or unknown confounders. Addressing the primary endpoint, CR/CRi/CRh after salvage therapy, the maximal sample size of 80 patients is assessed generating a desirable power of the used adaptive design, assuming a logistic regression is performed at a one-sided significance level α=0.05, the aspired power is 0.8, and the number of matching partners per intervention patient is at least 1. After enrolling 20 patients, the trial sample size will be recalculated in an interim analysis based on a conditional power argument.

CONCLUSION

Currently, there is no commonly accepted standard for salvage chemotherapy treatment. The objective of the salvage therapy is to reduce leukemic burden, achieve the best possible remission, and perform a hemopoietic stem-cell transplantation. Thus, in patients with FLT3-ITD mutation, the comparison of quizartinib with intensive salvage therapy versus chemotherapy alone appears as a logical consequence in terms of efficacy and safety.

ETHICS AND DISSEMINATION

Ethical approval and approvals from the local and federal competent authorities were granted. Trial results will be reported via peer-reviewed journals and presented at conferences and scientific meetings.

TRIAL REGISTRATION

ClinicalTrials.gov NCT03989713; EudraCT Number: 2018-002675-17.

A single-arm study design with non-inferiority and superiority time-to-event endpoints: a tool for proof-of-concept and de-intensification strategies in breast cancer

De-escalation trials in oncology evaluate therapies that aim to improve the quality of life of patients with low-risk cancer by avoiding overtreatment. Non-inferiority randomized trials are commonly used to investigate de-intensified regimens with similar efficacy to that of standard regimens but with fewer adverse effects (ESMO evidence tier A). In cases where it is not feasible to recruit the number of patients needed for a randomized trial, single-arm prospective studies with a hypothesis of non-inferiority can be conducted as an alternative. Single-arm studies are also commonly used to evaluate novel treatment strategies (ESMO evidence tier B). A single-arm design that includes both non-inferiority and superiority primary objectives will enable the ranking of clinical activity and other parameters such as safety, pharmacokinetics, and pharmacodynamics data. Here, we describe the statistical principles and procedures to support such a strategy. The non-inferiority margin is calculated using the fixed margin method. Sample size and statistical analyses are based on the maximum likelihood method for exponential distributions. We present example analyses in metastatic and adjuvant settings to illustrate the usefulness of our methodology. We also explain its implementation with nonparametric methods. Single-arm designs with non-inferiority and superiority analyses are optimal for proof-of-concept and de-escalation studies in oncology.

The effect of RCTs on drug demand: Evidence from off-label cancer drugs

This paper investigates the effect of scientific information from randomized controlled clinical trials (RCTs) on the demand for off-label uses of cancer drugs. This is a unique setting where demand for a drug for a specific use is observable both before and after the first RCT results are released. Using variation in the timing of RCTs across off-label uses of drugs, I find that demand responds asymmetrically to the trial results based on the statistical significance of the clinically relevant endpoint. When this endpoint is statistically significant, there is a large and immediate increase in demand. When this end point is not statistically significant, physicians are relatively slow to abandon use of the drug.

The Trial within Cohorts (TwiCs) study design in oncology: experience and methodological reflections

A Trial within Cohorts (TwiCs) study design is a trial design that uses the infrastructure of an observational cohort study to initiate a randomized trial. Upon cohort enrollment, the participants provide consent for being randomized in future studies without being informed. Once a new treatment is available, eligible cohort participants are randomly assigned to the treatment or standard of care. Patients randomized to the treatment arm are offered the new treatment, which they can choose to refuse. Patients who refuse will receive standard of care instead. Patients randomized to the standard of care arm receive no information about the trial and continue receiving standard of care as part of the cohort study. Standard cohort measures are used for outcome comparisons. The TwiCs study design aims to overcome some issues encountered in standard Randomized Controlled Trials (RCTs). An example of an issue in standard RCTs is the slow patient accrual. A TwiCs study aims to improve this by selecting patients using a cohort and only offering the intervention to patients in the intervention arm. In oncology, the TwiCs study design has gained increasing interest during the last decade. Despite its potential advantages over RCTs, the TwiCs study design has several methodological challenges that need careful consideration when planning a TwiCs study. In this article, we focus on these challenges and reflect on them using experiences from TwiCs studies initiated in oncology. Important methodological challenges that are discussed are the timing of randomization, the issue of non-compliance (refusal) after randomization in the intervention arm, and the definition of the intention-to-treat effect in a TwiCs study and how this effect is related to its counterpart in standard RCTs.

Optimal predictive probability designs for randomized biomarker-guided oncology trials

Introduction

Efforts to develop biomarker-targeted anti-cancer therapies have progressed rapidly in recent years. With efforts to expedite regulatory reviews of promising therapies, several targeted cancer therapies have been granted accelerated approval on the basis of evidence acquired in single-arm phase II clinical trials. And yet, in the absence of randomization, patient prognosis for progression-free survival and overall survival may not have been studied under standard of care chemotherapies for emerging biomarker subpopulations prior to the submission of an accelerated approval application. Historical control rates used to design and evaluate emerging targeted therapies often arise as population averages, lacking specificity to the targeted genetic or immunophenotypic profile. Thus, historical trial results are inherently limited for inferring the potential "comparative efficacy" of novel targeted therapies. Consequently, randomization may be unavoidable in this setting. Innovations in design methodology are needed, however, to enable efficient implementation of randomized trials for agents that target biomarker subpopulations.

Methods

This article proposes three randomized designs for early phase biomarker-guided oncology clinical trials. Each design utilizes the optimal efficiency predictive probability method to monitor multiple biomarker subpopulations for futility. Only designs with type I error between 0.05 and 0.1 and power of at least 0.8 were considered when selecting an optimal efficiency design from among the candidate designs formed by different combinations of posterior and predictive threshold. A simulation study motivated by the results reported in a recent clinical trial studying atezolizumab treatment in patients with locally advanced or metastatic urothelial carcinoma is used to evaluate the operating characteristics of the various designs.

Results

Out of a maximum of 300 total patients, we find that the enrichment design has an average total sample size under the null of 101.0 and a total average sample size under the alternative of 218.0, as compared to 144.8 and 213.8 under the null and alternative, respectively, for the stratified control arm design. The pooled control arm design enrolled a total of 113.2 patients under the null and 159.6 under the alternative, out of a maximum of 200. These average sample sizes that are 23-48% smaller under the alternative and 47-64% smaller under the null, as compared to the realized sample size of 310 patients in the phase II study of atezolizumab.

Discussion

Our findings suggest that potentially smaller phase II trials to those used in practice can be designed using randomization and futility stopping to efficiently obtain more information about both the treatment and control groups prior to phase III study.

Early-phase clinical trial eligibility and response evaluation criteria for refractory, relapsed, or progressive neuroblastoma: A consensus statement from the National Cancer Institute Clinical Trials Planning Meeting

BACKGROUND

International standardized criteria for eligibility, evaluable disease sites, and disease response assessment in patients with refractory, progressive, or relapsed high-risk neuroblastoma enrolled in early-phase clinical trials are lacking.

METHODS

A National Cancer Institute-sponsored Clinical Trials Planning Meeting was convened to develop an international consensus to refine the tumor site eligibility criteria and evaluation of disease response for early-phase clinical trials in children with high-risk neuroblastoma.

RESULTS

Standardized data collection of patient and disease characteristics (including specified genomic data), eligibility criteria, a definition of evaluable disease, and response evaluations for primary and metastatic sites of disease were developed. Eligibility included two distinct patient groups: progressive disease and refractory disease. The refractory disease group was subdivided into responding persistent disease and stable persistent disease to better capture the clinical heterogeneity of refractory neuroblastoma. Requirements for defining disease evaluable for a response assessment were provided; they included requirements for biopsy to confirm viable neuroblastoma and/or ganglioneuroblastoma in those patients with soft tissue or bone disease not avid for iodine-123 meta-iodobenzylguanidine. Standardized evaluations for response components and time intervals for response evaluations were established.

CONCLUSIONS

The use of international consensus eligibility, evaluability, and response criteria for early-phase clinical studies will facilitate the collection of comparable data across international trials and promote more rapid identification of effective treatment regimens for high-risk neuroblastoma.

The design and evaluation of hybrid controlled trials that leverage external data and randomization

Patient-level data from completed clinical studies or electronic health records can be used in the design and analysis of clinical trials. However, these external data can bias the evaluation of the experimental treatment when the statistical design does not appropriately account for potential confounders. In this work, we introduce a hybrid clinical trial design that combines the use of external control datasets and randomization to experimental and control arms, with the aim of producing efficient inference on the experimental treatment effects. Our analysis of the hybrid trial design includes scenarios where the distributions of measured and unmeasured prognostic patient characteristics differ across studies. Using simulations and datasets from clinical studies in extensive-stage small cell lung cancer and glioblastoma, we illustrate the potential advantages of hybrid trial designs compared to externally controlled trials and randomized trial designs.

Patient-level external control data from prior clinical studies or electronic health records can be used in the design and analysis of clinical trials. Here the authors report a hybrid trial design combining the use of external control data and randomization to test experimental treatments, using small cell lung cancer and glioblastoma datasets as examples.

IMbrave 151: a randomized phase II trial of atezolizumab combined with bevacizumab and chemotherapy in patients with advanced biliary tract cancer

Background:

Biliary tract cancers (BTCs) are heterogenous, highly aggressive tumors that harbor a dismal prognosis for which more effective treatments are needed. The role of cancer immunotherapy in BTC remains to be characterized. The tumor microenvironment (TME) of BTC is highly immunosuppressed and combination treatments are needed to promote effective anticancer immunity. Vascular endothelial growth factor (VEGF) drives immunosuppression in the TME by disrupting antigen presentation, limiting T-cell infiltration, or potentiating immune-suppressive cells. Many VEGF-regulated mechanisms are thought to be relevant to repressed antitumor immunity in BTC, making dual targeting of VEGF and programmed cell death protein 1 (PD-1)/PD-L1 pathways a rational approach. Gemcitabine and Cisplatin (Gem/Cis) can also modulate anticancer immunity through overlapping and complementary mechanisms to those regulated by VEGF. Anti-PD-L1/VEGF inhibition, coupled with chemotherapy, may potentiate antitumor immunity leading to enhanced clinical benefit.

Methods:

IMbrave 151 is a randomized, double-blind, placebo-controlled, multicenter, international phase II study to evaluate atezolizumab (a PD-L1 inhibitor) in combination with chemotherapy (gemcitabine and cisplatin) and bevacizumab (an anti-VEGF monoclonal antibody) as a first-line treatment for advanced BTC. Approximately 150 patients with previously untreated, advanced BTC will be randomized to either Arm A (atezolizumab + bevacizumab + Gem/Cis) or Arm B (atezolizumab + placebo + Gem/Cis). Randomization is stratified by the presence of metastatic disease, primary tumor location, and geographic region. The primary efficacy endpoint is investigator-assessed progression-free survival (PFS) per RECIST 1.1. Secondary endpoints include objective response rate (ORR), duration of response (DoR), disease control rate (DCR), overall survival (OS), and safety and patient reported outcomes (PROs). Tissue, blood, and stool samples will be collected at baseline and on-treatment in order to perform correlative biomarker analyses.

Discussion:

IMbrave 151 represents the first randomized study to evaluate combined PD-L1/VEGF blockade on a chemotherapy backbone in BTC.

Trial registration:

NCT identifier: NCT04677504; EUDRACT number: 2020-003759-14

Robustness and pragmatism of the evidence supporting the European Society for Medical Oncology guidelines for the diagnosis, treatment, and follow-up of follicular lymphoma

Background: Results of randomized clinical trials may not be entirely applicable to clinical practice. The present manuscript aims to explore the pragmatism and robustness of the evidence that supports the European Society for Medical Oncology (ESMO) follicular lymphoma (FL) guidelines.Methods & design: Analysis of all trials used to support positive, therapeutic, oncological recommendations in the 2020 ESMO FL guidelines. Predefined data points were extracted from each trial. Pragmatism was assessed by means of the PRECIS-2 tool, the difference in overall survival in the interventions compared and the source of funding. Robustness was assessed by means of the fragility index and the p value.Results: 28 trials were included. The full protocol or a protocol summary was provided for 12 (43%). Based on the PRECIS-2 domains, trials were considered pragmatic in organization, analysis and flexibility and explanatory in eligibility. Robustness was high, with 4/24 (17%) trials with p values between 0.05 and 0.005 and a median fragility index of 18.Conclusions: Results of trials to support ESMO recommendations in FL were robust. Pragmatism was high in some domains but modest to low in others and the pattern was similar across trials. Transparency in the publication of trial protocols was suboptimal.

OLIGOPELVIS GETUG P07, a Multicenter Phase II Trial of Combined High-dose Salvage Radiotherapy and Hormone Therapy in Oligorecurrent Pelvic Node Relapses in Prostate Cancer

BACKGROUND

Oligorecurrent pelvic nodal relapse in prostatic cancer is a challenge for regional salvage treatments. Androgen depriving therapies (ADTs) are a mainstay in metastatic prostate cancer, and salvage pelvic radiotherapy may offer long ADT-free intervals for patients harboring regional nodal relapses.

OBJECTIVE

To assess the efficacy of the combination of ADT and salvage radiotherapy in men with oligorecurrent pelvic node relapses of prostate cancer.

DESIGN, SETTING, AND PARTICIPANTS

We performed an open-label, phase II trial of combined high-dose intensity-modulated radiotherapy and ADT (6 mo) in oligorecurrent (five or fewer) pelvic node relapses in prostate cancer, detected by fluorocholine positron-emission tomography computed tomography imaging.

OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS

The primary endpoint was 2-yr progression-free survival defined as two consecutive prostate-specific antigen levels above the level at inclusion and/or clinical evidence of progression as per RECIST 1.1 and/or death from any cause.

RESULTS AND LIMITATIONS

Between August 2014 and July 2016, 67 patients were recruited in 15 centers. Half of the patients had received prior prostatic irradiation. The median age was 67.7 yr. After a median follow-up of 49.4 mo, 2- and 3-yr progression-free survival rates were 81% and 58%, respectively. Median progression-free survival was 45.3 mo. The median biochemical relapse-free survival (BRFS) was 25.9 mo. At 2 and 3 yr, the BRFS rates were 58% and 46%, respectively. Grade 2 + 2-yr genitourinary and gastrointestinal toxicities were 10% and 2%, respectively.

CONCLUSIONS

Combined high-dose salvage pelvic radiotherapy and ADT appeared to prolong tumor control in oligorecurrent pelvic node relapses in prostate cancer with limited toxicity. After 3 yr, nearly half of patients were in complete remission. Our study showed initial evidence of benefit, but a randomized trial is required to confirm this result.

PATIENT SUMMARY

In this report, we looked at the outcomes of combined high-dose salvage pelvic radiotherapy and 6-mo-long hormone therapy in oligorecurrent pelvic nodal relapse in prostatic cancer. We found that 46% of patients presenting with oligorecurrent pelvic node relapses in prostate cancer were in complete remission after 3 yr following combined treatment at the cost of limited toxicity.

The Use of External Control Data for Predictions and Futility Interim Analyses in Clinical Trials

PURPOSE

External control data from completed clinical trials and electronic health records can be valuable for the design and analysis of future clinical trials. We discuss the use of external control data for early stopping decisions in randomized clinical trials (RCTs).

METHODS

We specify interim analyses (IAs) approaches for RCTs, which allow investigators to integrate external data into early futility stopping decisions. IAs utilize predictions based on early data from the RCT, possibly combined with external data. These predictions at IAs express the probability that the trial will generate significant evidence of positive treatment effects. The trial is discontinued if this predictive probability becomes smaller than a pre-specified threshold. We quantify efficiency gains and risks associated with the integration of external data into interim decisions. We then analyze a collection of glioblastoma (GBM) datasets, to investigate if the balance of efficiency gains and risks justify the integration of external data into the IAs of future GBM RCTs.

RESULTS

Our analyses illustrate the importance of accounting for potential differences between the distributions of prognostic variables in the RCT and in the external data to effectively leverage external data for interim decisions. Using GBM datasets, we estimate that the integration of external data increases the probability of early stopping of ineffective experimental treatments by up to 25% compared to IAs that don't leverage external data. Additionally, we observe a reduction of the probability of early discontinuation for effective experimental treatments, which improves the RCT power.

CONCLUSION

Leveraging external data for IAs in RCTs can support early stopping decisions and reduce the number of enrolled patients when the experimental treatment is ineffective.

Optimal, minimax and admissible two-stage design for phase II oncology clinical trials

Background

The article aims to compare the efficiency of minimax, optimal and admissible criteria in Simon’s and Fleming’s two-stage design.

Methods

Three parameter settings (p₁-p₀ = 0.25–0.05, 0.30–0.10, 0.50–0.30) are designed to compare the maximum sample size, the critical values and the expected sample size for minimax, optimal and admissible designs. Type I & II error constraints (α, β) vary across (0.10, 0.10), (0.05, 0.20) and (0.05, 0.10), respectively.

Results

In both Simon’s and Fleming’s two-stage designs, the maximum sample size of admissible design is smaller than optimal design but larger than minimax design. Meanwhile, the expected samples size of admissible design is smaller than minimax design but larger than optimal design. Mostly, the maximum sample size and expected sample size in Fleming’s designs are considerably smaller than that of Simon’s designs.

Conclusions

Whenever (p₀, p₁) is pre-specified, it is better to explore in the range of probability q, based on relative importance between maximum sample size and expected sample size, and determine which design to choose. When q is unknown, optimal design may be more favorable for drugs with limited efficacy. Contrarily, minimax design is recommended if treatment demonstrates impressive efficacy.

A Review of Perspectives on the Use of Randomization in Phase II Oncology Trials

Historically, phase II oncology trials assessed a treatment's efficacy by examining its tumor response rate in a single-arm trial. Then, approximately 25 years ago, certain statistical and pharmacological considerations ignited a debate around whether randomized designs should be used instead. Here, based on an extensive literature review, we review the arguments on either side of this debate. In particular, we describe the numerous factors that relate to the reliance of single-arm trials on historical control data and detail the trial scenarios in which there was general agreement on preferential utilization of single-arm or randomized design frameworks, such as the use of single-arm designs when investigating treatments for rare cancers. We then summarize the latest figures on phase II oncology trial design, contrasting current design choices against historical recommendations on best practice. Ultimately, we find several ways in which the design of recently completed phase II trials does not appear to align with said recommendations. For example, despite advice to the contrary, only 66.2% of the assessed trials that employed progression-free survival as a primary or coprimary outcome used a randomized comparative design. In addition, we identify that just 28.2% of the considered randomized comparative trials came to a positive conclusion as opposed to 72.7% of the single-arm trials. We conclude by describing a selection of important issues influencing contemporary design, framing this discourse in light of current trends in phase II, such as the increased use of biomarkers and recent interest in novel adaptive designs.

To randomize, or not to randomize, that is the question: using data from prior clinical trials to guide future designs

BACKGROUND

Understanding the value of randomization is critical in designing clinical trials. Here, we introduce a simple and interpretable quantitative method to compare randomized designs versus single-arm designs using indication-specific parameters derived from the literature. We demonstrate the approach through application to phase II trials in newly diagnosed glioblastoma (ndGBM).

METHODS

We abstracted data from prior ndGBM trials and derived relevant parameters to compare phase II randomized controlled trials (RCTs) and single-arm designs within a quantitative framework. Parameters included in our model were (i) the variability of the primary endpoint distributions across studies, (ii) potential for incorrectly specifying the single-arm trial's benchmark, and (iii) the hypothesized effect size. Strengths and weaknesses of RCT and single-arm designs were quantified by various metrics, including power and false positive error rates.

RESULTS

We applied our method to show that RCTs should be preferred to single-arm trials for evaluating overall survival in ndGBM patients based on parameters estimated from prior trials. More generally, for a given effect size, the utility of randomization compared with single-arm designs is highly dependent on (i) interstudy variability of the outcome distributions and (ii) potential errors in selecting standard of care efficacy estimates for single-arm studies.

CONCLUSIONS

A quantitative framework using historical data is useful in understanding the utility of randomization in designing prospective trials. For typical phase II ndGBM trials using overall survival as the primary endpoint, randomization should be preferred over single-arm designs.

Sample size computation in phase II designs combining the A'Hern design and the Sargent and Goldberg design

This work focuses on the modification of two classical phase II trials designs, the A'Hern design, a single-arm single-stage design, and the Sargent and Goldberg design introduced in the context of flexible screening designs. In the first part of the paper, we have proposed a drift-adjusted A'Hern design, a hybrid design combining the A'Hern design and the Sargent and Goldberg design. Indeed, classical single-arm phase II designs such as the A'Hern design are still widely used in oncology. Conducting randomized comparative phase II trials may be challenging in many settings due to the increased sample size and this despite larger type 1 errors. Randomized non-comparative phase II designs first introduced by Herson and Carter include a simultaneous randomized standard-treatment reference arm to detect any drift in the reference arm assumption, but the trial is analyzed against historical controls as if it were a single-arm study. However, not incorporating at all an internal control arm in the trial design has been criticized in the literature. Our new design takes into account the observed response rate in a non-comparative reference arm to reduce the trial's risk of a false-positive conclusion. In the second part, we have proposed an alternative strategy to determining the sample size of the screened selection design. The latter, introduced in recent years by Yap et al. and Wu et al., extended the Sargent and Goldberg design to include a comparison to a historical control. However, their sample size computations may have potential weaknesses, which motivated us to revisit the existing approaches. A detailed simulation study has been carried out to evaluate the operating characteristics of the drift-adjusted A'Hern design and the different sample size strategies of the screened selection designs.

Design and Evaluation of an External Control Arm Using Prior Clinical Trials and Real-World Data

PURPOSE

We discuss designs and interpretable metrics of bias and statistical efficiency of "externally controlled" trials (ECT) and compare ECT performance to randomized and single-arm designs.

EXPERIMENTAL DESIGN

We specify an ECT design that leverages information from real-world data (RWD) and prior clinical trials to reduce bias associated with interstudy variations of the enrolled populations. We then used a collection of clinical studies in glioblastoma (GBM) and RWD from patients treated with the current standard of care to evaluate ECTs. Validation is based on a "leave one out" scheme, with iterative selection of a single-arm from one of the studies, for which we estimate treatment effects using the remaining studies as external control. This produces interpretable and robust estimates on ECT bias and type I errors.

RESULTS

We developed a model-free approach to evaluate ECTs based on collections of clinical trials and RWD. For GBM, we verified that inflated false positive error rates of standard single-arm trials can be considerably reduced (up to 30%) by using external control data.

CONCLUSIONS

The use of ECT designs in GBM, with adjustments for the clinical profiles of the enrolled patients, should be preferred to single-arm studies with fixed efficacy thresholds extracted from published results on the current standard of care.

Phase II Trials in Drug Development and Adaptive Trial Design

Phase II clinical studies represent a critical point in determining drug costs, and phase II is a poor predictor of drug success: >30% of drugs entering phase II studies fail to progress, and >58% of drugs go on to fail in phase III. Adaptive clinical trial design has been proposed as a way to reduce the costs of phase II testing by providing earlier determination of futility and prediction of phase III success, reducing overall phase II and III trial sizes, and shortening overall drug development time. This review examines issues in phase II testing and adaptive trial design.

Re-formulating Gehan's design as a flexible two-stage single-arm trial

BACKGROUND

Gehan's two-stage design was historically the design of choice for phase II oncology trials. One of the reasons it is less frequently used today is that it does not allow for a formal test of treatment efficacy, and therefore does not control conventional type-I and type-II error-rates.

METHODS

We describe how recently developed methodology for flexible two-stage single-arm trials can be used to incorporate the hypothesis test commonly associated with phase II trials in to Gehan's design. We additionally detail how this hypothesis test can be optimised in order to maximise its power, and describe how the second stage sample sizes can be chosen to more readily provide the operating characteristics that were originally envisioned by Gehan. Finally, we contrast our modified Gehan designs to Simon's designs, based on two examples motivated by real clinical trials.

RESULTS

Gehan's original designs are often greatly under- or over-powered when compared to type-II error-rates typically used in phase II. However, we demonstrate that the control parameters of his design can be chosen to resolve this problem. With this, though, the modified Gehan designs have operating characteristics similar to the more familiar Simon designs.

CONCLUSIONS

The trial design settings in which Gehan's design will be preferable over Simon's designs are likely limited. Provided the second stage sample sizes are chosen carefully, however, one scenario of potential utility is when the trial's primary goal is to ascertain the treatment response rate to a certain precision.

Varenicline for tobacco-dependence treatment in alcohol-dependent smokers: A randomized controlled trial

BACKGROUND

Tobacco use is prevalent among persons with alcohol abuse and dependence. Varenicline has been shown to be the most effective pharmacotherapy for smoking cessation and may decrease alcohol consumption. The purpose of this study was to evaluate the efficacy of 12 weeks of varenicline for increasing smoking abstinence rates in smokers with alcohol abuse or dependence.

METHODS

Participants were eligible for enrollment if they were 18 years or older, smoked 10 or more cigarettes per day for at least 6 months, had current alcohol abuse or dependence, and were interested in quitting smoking. Participants were randomly assigned to receive 12 weeks of varenicline 1 mg twice daily or matching placebo. The primary end point was 7-day point prevalence smoking abstinence at week 12.

RESULTS

The 7-day point prevalence smoking abstinence rate at 12 weeks was significantly higher with varenicline (n = 16) than placebo (n = 17) (43.8% vs 5.9%; P = .01). At 24 weeks, the 7-day point prevalence smoking abstinence rate was still significantly higher with varenicline than placebo (31.3% vs 0%; P = .02). At 12 weeks, mean (SD) drinks per drinking day was significantly lower with varenicline than placebo (5.7 [3.9] vs 9.0 [5.3] drinks; treatment effect estimate, -2.8 [90% CI, -6.6 to -1.0]). Adverse events were minor and comparable to varenicline clinical trials.

CONCLUSIONS

Varenicline is safe and efficacious for increasing smoking abstinence rates in smokers with alcohol abuse or dependence. Varenicline may decrease alcohol consumption in this population of smokers.

Phase 2 Neoadjuvant Treatment Intensification Trials in Rectal Cancer: A Systematic Review

PURPOSE

Multiple phase 2 trials of neoadjuvant treatment intensification in locally advanced rectal cancer have reported promising efficacy signals, but these have not translated into improved cancer outcomes in phase 3 trials. Improvements in phase 2 trial design are needed to reduce these false-positive signals. This systematic review evaluated the design of phase 2 trials of neoadjuvant long-course radiation or chemoradiation therapy treatment intensification in locally advanced rectal cancer.

METHODS AND MATERIALS

The PubMed, EMBASE, MEDLINE, and Cochrane Library databases were searched for published phase 2 trials of neoadjuvant treatment intensification from 2004 to 2016. Trial clinical design and outcomes were assessed, with statistical design and compliance rated using a previously published system. Multivariable meta-regression analysis of pathologic complete response (pCR) was conducted.

RESULTS

We identified 92 eligible trials. Patients with American Joint Committee on Cancer stage II and III equivalent disease were eligible in 87 trials (94.6%). In 43 trials (46.7%), local staging on magnetic resonance imaging was mandated. Only 12 trials (13.0%) were randomized, with 8 having a standard-treatment control arm. Just 51 trials (55.4%) described their statistical design, with 21 trials (22.8%) failing to report their sample size derivation. Most trials (n=84, 91.3%) defined a primary endpoint, but 15 different primary endpoints were used. All trials reported pCR rates. Only 38 trials (41.3%) adequately reported trial statistical design and compliance. Meta-analysis revealed a pooled pCR rate of 17.5% (95% confidence interval, 15.7%-19.4%) across treatment arms of neoadjuvant long-course radiation or chemoradiation therapy treatment intensification and substantial heterogeneity among the reported effect sizes (I² = 55.3%, P<.001). Multivariable meta-regression analysis suggested increased pCR rates with higher radiation therapy doses (adjusted P=.025).

CONCLUSIONS

Improvement in the design of future phase 2 rectal cancer trials is urgently required. A significant increase in randomized trials is essential to overcome selection bias and determine novel schedules suitable for phase 3 testing. This systematic review provides key recommendations to guide future treatment intensification trial design in rectal cancer.

Accounting for uncertainty in the historical response rate of the standard treatment in single-arm two-stage designs based on Bayesian power functions

In phase II single-arm studies, the response rate of the experimental treatment is typically compared with a fixed target value that should ideally represent the true response rate for the standard of care therapy. Generally, this target value is estimated through previous data, but the inherent variability in the historical response rate is not taken into account. In this paper, we present a Bayesian procedure to construct single-arm two-stage designs that allows to incorporate uncertainty in the response rate of the standard treatment. In both stages, the sample size determination criterion is based on the concepts of conditional and predictive Bayesian power functions. Different kinds of prior distributions, which play different roles in the designs, are introduced, and some guidelines for their elicitation are described. Finally, some numerical results about the performance of the designs are provided and a real data example is illustrated. Copyright © 2016 John Wiley & Sons, Ltd.

ESMO - Magnitude of Clinical Benefit Scale V.1.0 questions and answers

The ESMO Magnitude of Clinical Benefit Scale (ESMO-MCBS) is a standardised, generic, validated tool to stratify the magnitude of clinical benefit that can be anticipated from anticancer therapies. The ESMO-MCBS is intended to both assist oncologists in explaining the likely benefits of a particular treatment to their patients as well as to aid public health decision makers' prioritise therapies for reimbursement. From its inception the ESMO-MCBS Working Group has invited questions and critiques to promote understanding and to address misunderstandings regarding the nuanced use of the scale, and to identify shortcomings in the scale to be addressed in future planned revisions and updates. The ESMO-MCBS V.1.0 has attracted many questions regarding its development, structure and potential applications. These questions, together with responses from the ESMO-MCBS Working Group, have been edited and collated, and are herein presented as a supplementary resource.

Failures in Phase III: Causes and Consequences

Phase III randomized controlled trials (RCT) in oncology fail to lead to registration of new therapies more often than RCTs in other medical disciplines. Most RCTs are sponsored by the pharmaceutical industry, which reflects industry's increasing responsibility in cancer drug development. Many preclinical models are unreliable for evaluation of new anticancer agents, and stronger evidence of biologic effect should be required before a new agent enters the clinical development pathway. Whenever possible, early-phase clinical trials should include pharmacodynamic studies to demonstrate that new agents inhibit their molecular targets and demonstrate substantial antitumor activity at tolerated doses in an enriched population of patients. Here, we review recent RCTs and found that these conditions were not met for most of the targeted anticancer agents, which failed in recent RCTs. Many recent phase III RCTs were initiated without sufficient evidence of activity from early-phase clinical trials. Because patients treated within such trials can be harmed, they should not be undertaken. The bar should also be raised when making decisions to proceed from phase II to III and from phase III to marketing approval. Many approved agents showed only better progression-free survival than standard treatment in phase III trials and were not shown to improve survival or its quality. Introduction of value-based pricing of new anticancer agents would dissuade the continued development of agents with borderline activity in early-phase clinical trials. When collaborating with industry, oncologists should be more critical and better advocates for cancer patients.

Do single-arm trials have a role in drug development plans incorporating randomised trials?

Often, single-arm trials are used in phase II to gather the first evidence of an oncological drug's efficacy, with drug activity determined through tumour response using the RECIST criterion. Provided the null hypothesis of 'insufficient drug activity' is rejected, the next step could be a randomised two-arm trial. However, single-arm trials may provide a biased treatment effect because of patient selection, and thus, this development plan may not be an efficient use of resources. Therefore, we compare the performance of development plans consisting of single-arm trials followed by randomised two-arm trials with stand-alone single-stage or group sequential randomised two-arm trials. Through this, we are able to investigate the utility of single-arm trials and determine the most efficient drug development plans, setting our work in the context of a published single-arm non-small-cell lung cancer trial. Reference priors, reflecting the opinions of 'sceptical' and 'enthusiastic' investigators, are used to quantify and guide the suitability of single-arm trials in this setting. We observe that the explored development plans incorporating single-arm trials are often non-optimal. Moreover, even the most pessimistic reference priors have a considerable probability in favour of alternative plans. Analysis suggests expected sample size savings of up to 25% could have been made, and the issues associated with single-arm trials avoided, for the non-small-cell lung cancer treatment through direct progression to a group sequential randomised two-arm trial. Careful consideration should thus be given to the use of single-arm trials in oncological drug development when a randomised trial will follow.

Comparison of single-arm vs. randomized phase II clinical trials: a Bayesian approach

Single-arm studies are typically used in phase II of clinical trials, whose main objective is to determine whether a new treatment warrants further testing in a randomized phase III trial. The introduction of randomization in phase II, to avoid the limits of studies based on historical controls, is a critical issue widely debated in the recent literature. We use a Bayesian approach to compare single-arm and randomized studies, based on a binary response variable, in terms of their abilities of reaching the correct decision about the new treatment, both when it performs better than the standard one and when it is less effective. We evaluate how the historical control rate, the total sample size, and the elicitation of the prior distributions affect the decision about which study performs better.

Correlation of single arm versus randomised phase 2 oncology trial characteristics with phase 3 outcome

BACKGROUND AND AIM

The primary aim of this study was to determine whether randomised phase 2 (RP2) trials predict phase 3 trial outcome better than single arm phase 2 (SAP2) studies. Although theoretical superiority of RP2 trials has been postulated, no empiric studies have been conducted.

METHODS

Published phase 3 trials testing systemic cancer therapy were identified through a Medline search. Those of superiority design, which cited phase 2 trials supporting the experimental arm, were included. Trial design and outcome details were extracted. Statistical analysis was performed using the Generalized Estimating Equation method correlating phase 2 features with phase 3 outcome, accounting for any phase 3 duplication.

RESULTS

Of 189 eligible phase 3 trials, 18.5% were in haematological malignancies and 81.5% in solid tumors. The primary outcome was positive in 79 (41.8%). These were supported by 336 phase 2 trials (range 1-9 per phase 3 trial) including 66 RP2 trials. Positive phase 2 outcome, randomised or not, correlated with positive phase 3 outcome (p=0.03). RP2 studies were not superior to SAP2 studies at predicting phase 3 study success. Phase 2 trial features not predictive of phase 3 outcome included primary endpoint, sponsorship, sample size, similarity in patient population and therapy.

CONCLUSIONS

RP2 studies were not superior to SAP2 trials at predicting phase 3 study success. Further research into phase 2 trial design is required given the added resources required to conduct RP2 studies and the lack of empiric evidence supporting superiority over single arm studies.

Mixed response and time-to-event endpoints for multistage single-arm phase II design

BACKGROUND

The objective of phase II cancer clinical trials is to determine if a treatment has sufficient activity to warrant further study. The efficiency of a conventional phase II trial design has been the object of considerable debate, particularly when the study regimen is characteristically cytostatic. At the time of development of a phase II cancer trial, we accumulated clinical experience regarding the time to progression (TTP) for similar classes of drugs and for standard therapy. By considering the time to event (TTE) in addition to the tumor response endpoint, a mixed-endpoint phase II design may increase the efficiency and ability of selecting promising cytotoxic and cytostatic agents for further development.

METHODS

We proposed a single-arm phase II trial design by extending the Zee multinomial method to fully use mixed endpoints with tumor response and the TTE. In this design, the dependence between the probability of response and the TTE outcome is modeled through a Gaussian copula.

RESULTS

Given the type I and type II errors and the hypothesis as defined by the response rate (RR) and median TTE, such as median TTP, the decision rules for a two-stage phase II trial design can be generated. We demonstrated through simulation that the proposed design has a smaller expected sample size and higher early stopping probability under the null hypothesis than designs based on a single-response endpoint or a single TTE endpoint.

CONCLUSIONS

The proposed design is more efficient for screening new cytotoxic or cytostatic agents and less likely to miss an effective agent than the alternative single-arm design.

Two-stage phase II oncology designs using short-term endpoints for early stopping

Phase II oncology trials are conducted to evaluate whether the tumour activity of a new treatment is promising enough to warrant further investigation. The most commonly used approach in this context is a two-stage single-arm design with binary endpoint. As for all designs with interim analysis, its efficiency strongly depends on the relation between recruitment rate and follow-up time required to measure the patients’ outcomes. Usually, recruitment is postponed after the sample size of the first stage is achieved up until the outcomes of all patients are available. This may lead to a considerable increase of the trial length and with it to a delay in the drug development process. We propose a design where an intermediate endpoint is used in the interim analysis to decide whether or not the study is continued with a second stage. Optimal and minimax versions of this design are derived. The characteristics of the proposed design in terms of type I error rate, power, maximum and expected sample size as well as trial duration are investigated. Guidance is given on how to select the most appropriate design. Application is illustrated by a phase II oncology trial in patients with advanced angiosarcoma, which motivated this research.

A standardised, generic, validated approach to stratify the magnitude of clinical benefit that can be anticipated from anti-cancer therapies: the European Society for Medical Oncology Magnitude of Clinical Benefit Scale (ESMO-MCBS)

The value of any new therapeutic strategy or treatment is determined by the magnitude of its clinical benefit balanced against its cost. Evidence for clinical benefit from new treatment options is derived from clinical research, in particular phase III randomised trials, which generate unbiased data regarding the efficacy, benefit and safety of new therapeutic approaches. To date, there is no standard tool for grading the magnitude of clinical benefit of cancer therapies, which may range from trivial (median progression-free survival advantage of only a few weeks) to substantial (improved long-term survival). Indeed, in the absence of a standardised approach for grading the magnitude of clinical benefit, conclusions and recommendations derived from studies are often hotly disputed and very modest incremental advances have often been presented, discussed and promoted as major advances or 'breakthroughs'. Recognising the importance of presenting clear and unbiased statements regarding the magnitude of the clinical benefit from new therapeutic approaches derived from high-quality clinical trials, the European Society for Medical Oncology (ESMO) has developed a validated and reproducible tool to assess the magnitude of clinical benefit for cancer medicines, the ESMO Magnitude of Clinical Benefit Scale (ESMO-MCBS). This tool uses a rational, structured and consistent approach to derive a relative ranking of the magnitude of clinically meaningful benefit that can be expected from a new anti-cancer treatment. The ESMO-MCBS is an important first step to the critical public policy issue of value in cancer care, helping to frame the appropriate use of limited public and personal resources to deliver cost-effective and affordable cancer care. The ESMO-MCBS will be a dynamic tool and its criteria will be revised on a regular basis.

Randomized phase II clinical trials

Traditionally, Phase II trials have been conducted as single-arm trials to compare the response probabilities between an experimental therapy and a historical control. Historical control data, however, often have a small sample size, are collected from a different patient population, or use a different response assessment method, so that a direct comparison between a historical control and an experimental therapy may be severely biased. Randomized Phase II trials entering patients prospectively to both experimental and control arms have been proposed to avoid any bias in such cases. The small sample sizes for typical Phase II clinical trials imply that the use of exact statistical methods for their design and analysis is appropriate. In this article, we propose two-stage randomized Phase II trials based on Fisher's exact test, which does not require specification of the response probability of the control arm for testing. Through numerical studies, we observe that the proposed method controls the type I error accurately and maintains a high power. If we specify the response probabilities of the two arms under the alternative hypothesis, we can identify good randomized Phase II trial designs by adopting the Simon's minimax and optimal design concepts that were developed for single-arm Phase II trials.

Challenges of clinical trial design for targeted agents against pediatric leukemias

The past 40 years have seen significant improvements in both event-free and overall survival for children with acute lymphoblastic and acute myeloid leukemia (ALL and AML, respectively). Serial national and international clinical trials have optimized the use of conventional chemotherapeutic drugs and, along with improvements in supportive care that have enabled the delivery of more intensive regimens, have been responsible for the major improvements in patient outcome seen over the past few decades. However, the benefits of dose intensification have likely now been maximized, and over the same period, the identification of new cytotoxic drugs has been limited. Therefore, challenges remain if survival is to be improved further. In pediatric ALL, 5-year-survival rates of over 85% have been achieved with risk-stratified therapy, but a notable minority of patients will still not be cured. In pediatric AML, different challenges remain. A slower improvement in overall survival has taken place in this patient population. Despite the obvious morphological heterogeneity of AML blasts, biological stratification is comparatively limited, and translation into risk-stratified therapeutic approaches has only best characterized by the use of retinoic acid for t(15;17)-positive AML. Even where prognostic markers have been identified, limited therapeutic options or multi-drug resistance of AML blasts has limited the impact on patient benefit. For both, the acute morbidities of current treatment remain significant and may be life-threatening alone. In addition, the Childhood Cancer Survivor Study (CCSS) highlighted many leukemia survivors develop one or more chronic medical conditions attributable to treatment (1, 2). As the biology of leukemogenesis has become better understood, key molecules and intracellular pathways have been identified that offer the possibility of targeting directly the leukemia cells while sparing normal cells. Consequently, there is now a drive to develop novel leukemia-specific or "targeted" therapies. These new classes of drugs will have mechanisms of action, toxicities, and therapeutic indices quite different from conventional cytotoxic drugs previously encountered, thus rendering current clinical trial methodologies inappropriate. Clinical trial methods will need to be adapted to accommodate these features of these new classes of drugs. This review will address the challenges and some of the techniques for developing clinical trials for targeted therapies.

A two-stage patient enrichment adaptive design in phase II oncology trials

Illustrated is the use of a patient enrichment adaptive design in a randomized phase II trial which allows the evaluation of treatment benefits by the biomarker expression level and makes interim adjustment according to the pre-specified rules. The design was applied to an actual phase II metastatic hepatocellular carcinoma (HCC) trial in which progression-free survival (PFS) in two biomarker-defined populations is evaluated at both interim and final analyses. As an extension, a short-term biomarker is used to predict the long-term PFS in a Bayesian model in order to improve the precision of hazard ratio (HR) estimate at the interim analysis. The characteristics of the extended design are examined in a number of scenarios via simulations. The recommended adaptive design is shown to be useful in a phase II setting. When a short-term maker which correlates with the long-term PFS is available, the design can be applied in smaller early phase trials in which PFS requires longer follow-up. In summary, the adaptive design offers flexibility in randomized phase II patient enrichment trials and should be considered in an overall personalized healthcare (PHC) strategy.

Statistical evaluation of the use of concurrent controls in treatment screening studies

A well-designed pilot study can advance science by providing essential preliminary data to support or motivate further research, refine study logistics, and demonstrate proof of concept. Often, the outcomes of such studies can be quantified by a success/failure dichotomy. For example, a novel compound may show activation of a neural pathway, or it may not. When an intervention's efficacy is quantified using a dichotomous outcome, probability mass functions can be enumerated to determine the probability that the observed result from a pilot study supports further evaluation of the intervention since there is only a finite, and often small, number of sample configurations possible. The purpose of this research was to determine the probability of an "efficacy signal" for pilot studies using one- and two-sample pilot study designs. Efficacy signal was defined as the probability of observing a more favorable response proportion relative to a historical control (one-sample setting) or to a concurrent control (two-sample setting). An enumeration study (exact simulation) was conducted to calculate the efficacy signal probability. One-sample study designs yielded higher probability of determining an efficacy signal than the two-sample setting; however, sampling variation must be accounted for. A 68% score confidence interval is recommended for this purpose.

Estimation of renal cell carcinoma treatment effects from disease progression modeling

To improve future drug development efficiency in renal cell carcinoma (RCC), a disease-progression model was developed with longitudinal tumor size data from a phase III trial of sorafenib in RCC. The best-fit model was externally evaluated on 145 placebo-treated patients in a phase III trial of pazopanib; the model incorporated baseline tumor size, a linear disease-progression component, and an exponential drug effect (DE) parameter. With the model-estimated effect of sorafenib on RCC growth, we calculated the power of randomized phase II trials between sorafenib and hypothetical comparators over a range of effects. A hypothetical comparator with 80% greater DE than sorafenib would have 82% power (one-sided α = 0.1) with 50 patients per arm. Model-based quantitation of treatment effect with computed tomography (CT) imaging offers a scaffold on which to develop new, more efficient, phase II trial end points and analytic strategies for RCC.

Outcomes of phase II clinical trials with single-agent therapies in advanced/metastatic non-small cell lung cancer published between 2000 and 2009

PURPOSE

We analyzed the outcomes of single-agent phase II clinical trials in non-small cell lung cancer (NSCLC) to determine trial parameters that predicted clinical activity.

EXPERIMENTAL DESIGN

Data on response rate (RR), progression-free survival (PFS), and overall survival (OS) from all English language, single-agent phase II trials in advanced/metastatic NSCLC indexed by PubMed (January 2000 through December 2009) were abstracted.

RESULTS

A total of 143 single-agent phase II trials (7,701 patients) were identified. The median RR was 10%, PFS 2.8 months, and OS 7.6 months. RR and PFS correlated with OS (r = 0.46, P < 0.001, r = 0.52, P < 0.001, respectively) and RR correlated with PFS (r = 0.61, P < 0.001). Treatment arms enriched for patients with molecular targets had a higher median RR (48.8% vs. 9.7%, P = 0.005), longer median PFS (6 vs. 2.8 months, P = 0.005), and OS (11.3 vs. 7.5 months, P = 0.05) as compared with those of unselected patients. In multivariate analysis, only studies enriched for patients with molecular targets or including drugs that eventually gained FDA/EMA approval were associated with a higher RR, and longer PFS/OS.

CONCLUSIONS

In phase II trials in NSCLC, RR and PFS correlated with OS. Studies enriched for patients with putative molecular drug targets were associated with higher therapeutic benefit as compared with those of unselected populations.

Adaptive designs for single-arm phase II trials in oncology

Clinical phase II trials in oncology are conducted to determine whether the activity of a new anticancer treatment is promising enough to merit further investigation. Two-stage designs are commonly used for this situation to allow for early termination. Designs proposed in the literature so far have the common drawback that the sample sizes for the two stages have to be specified in the protocol and have to be adhered to strictly during the course of the trial. As a consequence, designs that allow a higher extent of flexibility are desirable. In this article, we propose a new adaptive method that allows an arbitrary modification of the sample size of the second stage using the results of the interim analysis or external information while controlling the type I error rate. If the sample size is not changed during the trial, the proposed design shows very similar characteristics to the optimal two-stage design proposed by Chang et al. (Biometrics 1987; 43:865-874). However, the new design allows the use of mid-course information for the planning of the second stage, thus meeting practical requirements when performing clinical phase II trials in oncology.

Resampling phase III data to assess phase II trial designs and endpoints

PURPOSE

The best phase II design and endpoint for growth inhibitory agents is controversial. We simulated phase II trials by resampling patients from a positive (sorafenib vs. placebo; TARGET) and a negative (AE941 vs. placebo) phase III trial in metastatic renal cancer to compare the ability of various designs and endpoints to predict the known results.

EXPERIMENTAL DESIGN

A total of 770 and 259 patients from TARGET and the AE 941 trial, respectively, were resampled (5,000 replicates) to simulate phase II trials with α = 0.10 (one-sided). Designs/endpoints: single arm, two-stage with response rate (RR) by Response Evaluation Criteria in Solid Tumors (RECIST; 37 patients); and randomized, two arm (20-35 patients per arm) with RR by RECIST, mean log ratio of tumor sizes (log ratio), progression-free survival (PFS) rate at 90 days (PFS-90), and overall PFS.

RESULTS

Single-arm trials were positive with RR by RECIST in 55% and 1% of replications for sorafenib and AE 941, respectively. Randomized trials versus placebo with 20 patients per arm were positive with RR by RECIST in 55% and 7%, log ratio in 88% and 25%, PFS-90 in 64% and 15%, and overall PFS in 69% and 9% of replications for sorafenib and AE 941, respectively.

CONCLUSIONS

Compared with the single-arm design and the randomized design comparing PFS, the randomized phase II design with the log ratio endpoint has greater power to predict the positive phase III result of sorafenib in renal cancer, but a higher false positive rate for the negative phase III result of AE 941.