From Trial and Error to Trial Simulation. Part 2: An Appraisal of Current Beliefs in the Design and Analysis of Clinical Trials for Antidepressant Drugs

A Simulation Study of the Effect of Clinical Characteristics and Treatment Choice on Reliever Medication Use, Symptom Control and Exacerbation Risk in Moderate-Severe Asthma

INTRODUCTION

The relationship between immediate symptom control, reliever medication use and exacerbation risk on treatment response and factors that modify it have not been assessed in an integrated manner. Here we apply simulation scenarios to evaluate the effect of individual baseline characteristics on treatment response in patients with moderate-severe asthma on regular maintenance dosing monotherapy with fluticasone propionate (FP) or combination therapy with fluticasone propionate/salmeterol (FP/SAL) or budesonide/formoterol (BUD/FOR).

METHODS

Reduction in reliever medication use (puffs/24 h), change in symptom control scores (ACQ-5), and annualised exacerbation rate over 12 months were simulated in a cohort of patients with different baseline characteristics (e.g. time since diagnosis, asthma control questionnaire (ACQ-5) symptom score, smoking status, body mass index (BMI) and sex) using drug-disease models derived from large phase III/IV clinical studies.

RESULTS

Simulation scenarios show that being a smoker, having higher baseline ACQ-5 and BMI, and long asthma history is associated with increased reliever medication use (p < 0.01). This increase correlates with a higher exacerbation risk and higher ACQ-5 scores over the course of treatment, irrespective of the underlying maintenance therapy. Switching non-responders to ICS monotherapy to combination therapy after 3 months resulted in immediate reduction in reliever medication use (i.e. 1.3 vs. 1.0 puffs/24 h for FP/SAL and BUD/FOR, respectively). In addition, switching patients with ACQ-5 > 1.5 at baseline to FP/SAL resulted in 34% less exacerbations than those receiving regular dosing BUD/FOR (p < 0.01).

CONCLUSIONS

We have identified baseline characteristics of patients with moderate to severe asthma that are associated with greater reliever medication use, poor symptom control and higher exacerbation risk. Moreover, the effects of different inhaled corticosteroid (ICS)/long-acting beta agonist (LABA) combinations vary significantly when considering long-term treatment performance. These factors should be considered in clinical practice as a basis for personalised management of patients with moderate-severe asthma symptoms.

Understanding the Clinical Implications of Individual Patient Characteristics and Treatment Choice on the Risk of Exacerbation in Asthma Patients with Moderate-Severe Symptoms

INTRODUCTION

The assessment of future risk has become an important feature in the management of patients with asthma. However, the contribution of patient-specific characteristics and treatment choices to the risk of exacerbation is poorly understood. Here we evaluated the effect of interindividual baseline differences on the risk of exacerbation and treatment performance in patients receiving regular maintenance doses of inhaled corticosteroids (ICS) or ICS/long-acting beta-agonists (LABA) combination therapy.

METHODS

Exacerbations and changes to asthma symptoms 5-item Asthma Control Questionnaire (ACQ-5) were simulated over a 12-month period using a time-to-event and a longitudinal model developed from phase III/IV studies in patients with moderate-severe asthma (N = 16,282). Simulations were implemented to explore treatment performance across different scenarios, including randomised designs and real-world settings. Treatment options included regular dosing with ICS monotherapy [fluticasone propionate (FP)] and combination therapy [fluticasone propionate/salmeterol (FP/SAL) or budesonide/formoterol (BUD/FOR)]. Exacerbation rate was analysed using the log-rank test. The cumulative incidence of events was summarised stratified by treatment.

RESULTS

Being a woman, smoker, having higher baseline ACQ-5 and body mass index (BMI) and lower forced expiratory volume in the first second (FEV1) are associated with increased exacerbation risk (p < 0.01). This risk is bigger in winter because of the seasonal variation effect. Across the different scenarios, the use of FP/SAL resulted in a 10% lower annual incidence of exacerbations relative to FP or regular dosing BUD/FOR, independently of baseline characteristics. Similar differences in the annual incidence of exacerbations were also observed between treatments in obese patients (BMI ≥ 25-35 kg/m2) (p < 0.01) and in patients who do not achieve symptom control on FP monotherapy.

CONCLUSIONS

Individual baseline characteristics and treatment choices affect future risk. Achieving comparable levels of symptom control whilst on treatment does not imply comparable risk reduction, as shown by the lower exacerbation rates in FP/SAL vs. BUD/FOR-treated patients. These factors should be considered as a basis for personalised clinical management of patients with moderate-severe asthma.

Clinical Trial Diversity: An Opportunity for Improved Insight into the Determinants of Variability in Drug Response

Although the number of countries participating in pivotal trials submitted to enable drug registration has nearly doubled over the past 25 years, there has not been a substantial increase in the diversity of clinical trial populations. In parallel, our understanding of factors that influence medicine response and variability has continued to evolve. The notion of intrinsic and extrinsic sources of variability has been embedded into different regulatory guidelines, including the recent guideline on the importance of enhancing the diversity of clinical trial populations.

In addition to presenting the clinical and scientific reasons for ensuring that clinical trial populations represent the demographics of patient populations, this overview outlines the efforts of regulatory agencies, patient advocacy groups and clinical researchers to attain this goal through strategies to meet representation in recruitment targets and broaden eligibility criteria. Despite these efforts, challenges to participation in clinical trials remain, and certain groups continue to be underrepresented in development programmes. These challenges are amplified when the representativeness of specific groups may vary across countries and regions in a global clinical programme.

Whilst enhanced trial diversity is a critical step towards ensuring that results will be representative of patient populations, a concerted effort is required to characterise further the factors influencing interindividual and regional differences in response for global populations. Quantitative clinical pharmacology principles should be applied to allow extrapolation of data across groups or regions as well as provide insight into the effect of patient‐specific characteristics on a medicine's dose rationale and efficacy and safety profiles.

Extrapolation of Drug Clearance in Children ≤ 2 Years of Age from Empirical Models Using Data from Children (> 2 Years) and Adults

BACKGROUND

The application of modeling and simulation approaches in clinical pharmacology studies has gained momentum over the last 20 years.

OBJECTIVES

The objective of this study was to develop six empirical models from clearance data obtained from children aged > 2 years and adults to evaluate the suitability of the models to predict drug clearance in children aged ≤ 2 years (preterm, term, and infants).

METHODS

Ten drugs were included in this study and administered intravenously: alfentanil, amikacin, busulfan, cefetamet, meperidine, oxycodone, propofol, sufentanil, theophylline, and tobramycin. These drugs were selected according to the availability of individual subjects' weight, age, and clearance data (concentration-time data for these drugs were not available to the author). The chosen drugs are eliminated by extensive metabolism by either the renal route or both the renal and hepatic routes. The six empirical models were (1) age and body weight-dependent sigmoidal maximum possible effect (E_max) maturation model, (2) body weight-dependent sigmoidal E_max model, (3) uridine 5'-diphospho [body weight-dependent allometric exponent model (BDE)], (4) age-dependent allometric exponent model (ADE), (5) a semi-physiological model, and (6) an allometric model developed from children aged > 2 years to adults. The model-predicted clearance values were compared with observed clearance values in an individual child. In this analysis, a prediction error of ≤ 50% for mean or individual clearance values was considered acceptable.

RESULTS

Across all age groups and the ten drugs, data for 282 children were compared between observed and model-predicted clearance values. The validation data consisted of 33 observations (sum of different age groups for ten drugs). Only three of the six models (body weight-dependent sigmoidal E_max model, ADE, and semi-physiological model) provided reasonably accurate predictions of clearance (> 80% observation with ≤ 50% prediction error) in children aged ≤ 2 years. In most instances, individual predicted clearance values were erratic (as indicated by % error) and were not in agreement with the observed clearance values.

CONCLUSIONS

The study indicated that simple empirical models can provide more accurate results than complex empirical models.

Power behind the throne: A clinical trial simulation study evaluating the impact of controllable design factors on the power of antidepressant trials

OBJECTIVE

To evaluate the impact of controllable design factors on the power of antidepressants trials.

METHODS

Using clinical trial simulation (CTS), we analyzed the combined impact on the power of trials of controllable design factors (sample size, outcome metrics, and disease severity at inclusion) and uncontrollable parameters (heterogeneity of diseases labeled "depression" in the source population and selective effects of drugs on items of the Hamilton Depression Rating Scale [HDRS], the most used outcome measurement tool). We elaborated 3,840 scenarios calibrated with real data, particularly the publication bias-corrected effect size.

RESULTS

For an effect size of 0.26, simulations revealed that in trials with ≤650 participants, power was less than 80%. Among the tested outcome metrics, the "remission" outcome provided more robustness for sample heterogeneity, whereas the continuous outcome "HDRS changes" provided more robustness when investigating drugs with a selective effect on the HDRS items. For the "remission" outcome, the power of trials increased with increasing HDRS threshold at inclusion but decreased with the outcomes "response" and "HDRS changes. Drugs with a selective effect on the HDRS items could not reach the same power as for the reference drug.

CONCLUSION

Our study allows for drawing recommendations to avoid underpowered trials of antidepressants.

Impact of disease progression on individual IPSS trajectories and consequences of immediate versus delayed start of treatment in patients with moderate or severe LUTS associated with BPH

Purpose

Despite superiority of tamsulosin–dutasteride combination therapy versus monotherapy for lower urinary tract symptoms due to benign prostatic hyperplasia (LUTS/BPH), patients at risk of disease progression are often initiated on α-blockers. This study evaluated the impact of initiating tamsulosin monotherapy prior to switching to tamsulosin–dutasteride combination therapy versus immediate combination therapy using a longitudinal model describing International Prostate Symptom Score (IPSS) trajectories in moderate/severe LUTS/BPH patients at risk of disease progression.

Methods

Clinical trial simulations (CTS) were performed using data from 10,238 patients from Phase III/IV dutasteride trials. The effect of varying disease progression rates was explored by comparing profiles on- and off-treatment. CTS scenarios were investigated, including a reference (immediate combination therapy) and six alternative virtual treatment arms (delayed combination therapy of 1–24 months). Clinical response (≥ 25% IPSS reduction relative to baseline) was analysed using log-rank test. Differences in IPSS relative to baseline at various on-treatment time points were assessed by t tests.

Results

Delayed combination therapy initiation led to significant (p < 0.01) decreases in clinical response. At month 48, clinical response rate was 79.7% versus 74.1%, 70.3% and 71.0% and IPSS was 6.3 versus 7.6, 8.1 and 8.0 (switchers from tamsulosin monotherapy after 6, 12 and 24 months, respectively) with immediate combination therapy. More patients transitioned from severe/moderate to mild severity scores by month 48.

Conclusions

CTS allows systematic evaluation of immediate versus delayed combination therapy. Immediate response to α-blockers is not predictive of long-term symptom improvement. Observed IPSS differences between immediate and delayed combination therapy (6–24 months) are statistically significant.

Electronic supplementary material

The online version of this article (10.1007/s00345-019-02783-x) contains supplementary material, which is available to authorized users.

Challenges in translational drug research in neuropathic and inflammatory pain: the prerequisites for a new paradigm

AIM

Despite an improved understanding of the molecular mechanisms of nociception, existing analgesic drugs remain limited in terms of efficacy in chronic conditions, such as neuropathic pain. Here, we explore the underlying pathophysiological mechanisms of neuropathic and inflammatory pain and discuss the prerequisites and opportunities to reduce attrition and high-failure rate in the development of analgesic drugs.

METHODS

A literature search was performed on preclinical and clinical publications aimed at the evaluation of analgesic compounds using MESH terms in PubMed. Publications were selected, which focused on (1) disease mechanisms leading to chronic/neuropathic pain and (2) druggable targets which are currently under evaluation in drug development. Attention was also given to the role of biomarkers and pharmacokinetic-pharmacodynamic modelling.

RESULTS

Multiple mechanisms act concurrently to produce pain, which is a non-specific manifestation of underlying nociceptive pathways. Whereas these manifestations can be divided into neuropathic and inflammatory pain, it is now clear that inflammatory mechanisms are a common trigger for both types of pain. This has implications for drug development, as the assessment of drug effects in experimental models of neuropathic and chronic pain is driven by overt behavioural measures. By contrast, the use of mechanistic biomarkers in inflammatory pain has provided the pharmacological basis for dose selection and evaluation of non-steroidal anti-inflammatory drugs (NSAIDs).

CONCLUSION

A different paradigm is required for the identification of relevant targets and candidate molecules whereby pain is coupled to the cause of sensorial signal processing dysfunction rather than clinical symptoms. Biomarkers which enable the characterisation of drug binding and target activity are needed for a more robust dose rationale in early clinical development. Such an approach may be facilitated by quantitative clinical pharmacology and evolving technologies in brain imaging, allowing accurate assessment of target engagement, and prediction of treatment effects before embarking on large clinical trials.

Population Pharmacokinetics: Some Observations in Pediatric Modeling for Drug Clearance

The objective of this study is to evaluate the predictive performance of several models to predict drug clearance in preterm and term neonates. Five models using different types of allometric and linear models were developed. Two sets of data were used to develop these models (data from preterm neonates to adults and data from preterm and term neonates). Models were also developed with (normalized to 70 kg) or without body weight normalization (body weight 1 kg). From the literature, clearance values for four drugs from neonates to adults were obtained. External data were used to evaluate the predictive performance of these models in preterm and term neonates. The results of the study indicated that (1) normalization to a standard body weight had no impact on the predictive performance of the models, (2) the model developed from preterm neonates to adults using fixed exponent 0.75 provided inaccurate estimate (overestimation) of drug clearance in neonates, (3) a far superior prediction of clearance was observed with the model when the exponents of allometry were estimated than the model using exponent 0.75, (4) linear models with the exception of the model with intercept provided comparable results to the estimated exponent model and were superior in their predictive performance to the model using exponent 0.75, and (5) when the models were developed from neonate data, the predictive performance of all models were similar. Overall, the study indicated that body weight normalization had no impact on the performance of model prediction, the exponents of allometry in pharmacostatistical models should be estimated rather than fixed, and more studies are needed to evaluate the suitability of linear models for the prediction of drug clearance in neonates.

How to optimise drug study design: pharmacokinetics and pharmacodynamics studies introduced to paediatricians

OBJECTIVES

In children, there is often lack of sufficient information concerning the pharmacokinetics (PK) and pharmacodynamics (PD) of a study drug to support dose selection and effective evaluation of efficacy in a randomised clinical trial (RCT). Therefore, one should consider the relevance of relatively small PKPD studies, which can provide the appropriate data to optimise the design of an RCT.

METHODS

Based on the experience of experts collaborating in the EU-funded Global Research in Paediatrics consortium, we aimed to inform clinician-scientists working with children on the design of investigator-initiated PKPD studies.

KEY FINDINGS

The importance of the identification of an optimal dose for the paediatric population is explained, followed by the differences and similarities of dose-ranging and efficacy studies. The input of clinical pharmacologists with modelling expertise is essential for an efficient dose-finding study.

CONCLUSIONS

The emergence of new laboratory techniques and statistical tools allows for the collection and analysis of sparse and unbalanced data, enabling the implementation of (observational) PKPD studies in the paediatric clinic. Understanding of the principles and methods discussed in this study is essential to improve the quality of paediatric PKPD investigations, and to prevent the conduct of paediatric RCTs that fail because of inadequate dosing.

Stratification, Hypothesis Testing, and Clinical Trial Simulation in Pediatric Drug Development

BACKGROUND

Pediatric drug development is plagued by small sample sizes, unvalidated clinical endpoints, and limited studies.

OBJECTIVES

The objective of this study was to determine whether age stratification within the pediatric population could be used to (1) assess response to a pharmacologic intervention and to (2) design future trials based upon published stratified disease data using clinical trial simulation (CTS).

METHODS

Data available from the literature for Kawasaki disease (KD) was used in the model. Age-stratified CTS for a theoretical new drug was conducted.

RESULTS

Population-specific differences due to age might affect trial success if not taken into account. CTS predicted inflammatory indices, and inclusion cutoff significantly altered the trial outcome. Finally, altered pharmacokinetics/pharmacodynamics in varying age groups of KD patients may alter drug exposure and response.

CONCLUSIONS

If assumptions regarding a pediatric disease process, such as KD, do not include age stratification with inclusion or response, then the wrong decision could result with regard to age-appropriateness or approval of a drug.

Prediction of Human Glomerular Filtration Rate from Preterm Neonates to Adults: Evaluation of Predictive Performance of Several Empirical Models

The objective of this study was to evaluate the predictive performance of several allometric empirical models (body weight dependent, age dependent, fixed exponent 0.75, a data-dependent single exponent, and maturation models) to predict glomerular filtration rate (GFR) in preterm and term neonates, infants, children, and adults without any renal disease. In this analysis, the models were developed from GFR data obtained from inulin clearance (preterm neonates to adults; n = 93) and the predictive performance of these models were evaluated in 335 subjects (preterm neonates to adults). The primary end point was the prediction of GFR from the empirical allometric models and the comparison of the predicted GFR with measured GFR. A prediction error within ±30% was considered acceptable. Overall, the predictive performance of the four models (BDE, ADE, and two maturation models) for the prediction of mean GFR was good across all age groups but the prediction of GFR in individual healthy subjects especially in neonates and infants was erratic and may be clinically unacceptable.

Prediction of drug clearance in children: an evaluation of the predictive performance of several models

The objective of this study is to evaluate the predictive performance of several models to predict drug clearance in children ≤5 years of age. Six models (allometric model (data-dependent exponent), fixed exponent of 0.75 model, maturation model, body weight-dependent model, segmented allometric model, and age-dependent exponent model) were evaluated in this study. From the literature, the clearance values for six drugs from neonates to adults were obtained. External data were used to evaluate the predictive performance of these models in children ≤5 years of age. With the exception of a fixed exponent of 0.75, the mean predicted clearance in most of the age groups was within ≤50% prediction error. Individual clearance prediction was erratic by all models and cannot be used reliably to predict individual clearance. Maturation, body weight-dependent, and segmented allometric models to predict clearances of drugs in children ≤5 years of age are of limited practical value during drug development due to the lack of availability of data. Age-dependent exponent model can be used for the selection of first-in-children dose during drug development.

A population pharmacokinetic model of remifentanil in pediatric patients using body-weight-dependent allometric exponents

BACKGROUND

Allometric exponents in population pharmacokinetic analysis are regularly used but the issue of fixing or estimating an allometric exponent remains controversial. The objective of the current analysis is to evaluate the performance of a body-weight-dependent allometric exponent (BDE) model of remifentanil.

METHODS

The study was conducted in 34 patients (neonates to 17 years and 2.5 to 97 kg body weight) following a single intravenous (IV) infusion of remifentanil (5 μg/kg). A population pharmacokinetic approach was taken to describe drug clearance by the following BDE equation: CL=CLpop(BW/14.6 kg)L×BW(-M). Three allometric models were used to explore the impact of allometric exponents on the total clearance of remifentanil.

RESULTS

All model-fitted structural, covariate, and statistical parameters were estimated with good to excellent precision (%RSE). However, on the basis of calculated Akaike weights (0.000 for model 1, 0.004 for model 2, and 0.996 for model 3), model 3 is the most robust model to describe individual clearance estimates.

CONCLUSIONS

The BDE model performed best for the estimation of remifentanil clearance and is realistic and of practical value. Further investigation should be conducted for such models.

Duloxetine-bupropion combination for treatment-resistant atypical depression: a double-blind, randomized, placebo-controlled trial

The efficacy, safety, and tolerability of combined bupropion versus placebo using duloxetine as active reference drug, in patients with a DSM-IV diagnosis of major depression with atypical features and a history of treatment resistance, were evaluated in this preliminary six-week study. Patients (n=46) had a baseline Hamilton Depression Scale (HAM-D) ≥14 and were randomly assigned to 150/300 mg/day bupropion vs. placebo, which was added to 60 to 120 mg/day duloxetine depending on baseline depression severity. Atypical features of depression were assessed using the additional eight-item module of the Structured Interview Guide for the HAM-D with the Atypical Depression Supplement. By week 6, only five (21.7%) patients receiving duloxetine+placebo vs. six (26.1%) patients on the bupropion combination achieved response. No significant difference in final HAM-D scores between the two groups was observed between those patients achieving response. The presence of a higher number of atypical features significantly predicted non-response, with the relevant binary logistic regression model correctly classifying 17 out 22 (77.3%) of non-responders [Exp(B)=0.294; p=0.016] vs. 17 out 23 (73.9%) [Exp(B)=0.353; p=0.028] non-responder cases in the "+placebo" and "+bupropion" groups, respectively. In those patients receiving bupropion, treatment-emergent adverse events leading to withdrawal were more common among those receiving lower doses of the combination drug, and no life-threating dangers were noted. Additional studies, including an adequate course of duloxetine trial, are nonetheless aimed to allow a firm conclusion about the usefulness of the combination of duloxetine and bupropion for treatment-resistant cases of major depression with atypical features.

Feasibility of a fixed-dose regimen of pyrazinamide and its impact on systemic drug exposure and liver safety in patients with tuberculosis

Historically, dosing regimens for the treatment of tuberculosis (TB) have been proposed in an empirical manner. Dose selection has often been the result of efficacy trials in which drugs were administered regardless of the magnitude of the effect of demographic factors on drug disposition. This has created challenges for the prescription of fixed-dose combinations with novel therapeutic agents. The objectives of this investigation were to evaluate the impact of body weight on the overall systemic exposure to pyrazinamide (PZA) and to assess whether the use of one fixed dose, without adjustment according to weight, would ensure target exposure and safety requirements across the overall patient population. Using a population pharmacokinetic model, simulation scenarios were explored based on population demographics from clinical trials in TB patients and on historical hepatotoxicity data. The systemic drug exposure (area under the concentration-time curve [AUC]), peak concentrations (the maximum concentration of drug in serum [C(max)]), the time above the MIC (t > MIC), and the risk of hepatotoxicity were evaluated for the current weight-banded regimen and compared to fixed doses under the assumption that pharmacokinetic differences are the primary drivers of toxicity. Evaluation of the standard weight banding reveals that more than 50% of subjects in the weight range of 45 to 55 kg remain below the proposed target exposure to PZA. In contrast, the use of a fixed 1,500-mg dose resulted in a lower proportion of subjects under the target value, with a 0.2% average overall increase in the risk of hepatotoxicity. Our results strongly support the use of a fixed-dose regimen for PZA in coformulation or combination with novel therapeutic agents.

Mathematical modeling of tumor growth and tumor growth inhibition in oncology drug development

INTRODUCTION

Approaches aiming to model the time course of tumor growth and tumor growth inhibition following a therapeutic intervention have recently been proposed for supporting decision making in oncology drug development. When considered in a comprehensive model-based approach, tumor growth can be included in the cascade of quantitative and causally related markers that lead to the prediction of survival, the final clinical response.

AREAS COVERED

The authors examine articles dealing with the modeling of tumor growth and tumor growth inhibition in both preclinical and clinical settings. In addition, the authors review models describing how pharmacological markers can be used to predict tumor growth and models describing how tumor growth can be linked to survival endpoints.

EXPERT OPINION

Approaches and success stories of application of model-based drug development centered on tumor growth modeling are growing. It is also apparent that these approaches can answer practical questions on drug development more effectively than that in the past. For modeling purposes, some improvements are still needed related to study design and data quality. Further efforts are needed to encourage the mind shift from a simple description of data to the prediction of untested conditions that modeling approaches allow.

Paediatric drug development: are population models predictive of pharmacokinetics across paediatric populations?

AIMS

To assess the predictive value of a model-based approach for dose selection across paediatric populations in early clinical drug development.

METHODS

Abacavir was selected as a paradigm compound using data across a wide age range. Abacavir pharmacokinetics (PK) in children were analysed separately from infants and toddlers. Two independent models were obtained, and systemic exposure (AUC) was then simulated across populations based on the estimates from each model. Drug exposures in infants and toddlers were predicted using pharmacokinetic parameter distributions obtained from children, and the other way around.

RESULTS

The pharmacokinetic models (a two-compartment PK model for infants and toddlers and a one compartment PK model for children) accurately described the exposure in the population from which they were built. However, neither model predicted exposure in a different population: in infants, the median AUC (95%(-) CI) was estimated at 7.03 (6.72, 7.48) µg ml(-1) h, whilst it was predicted at 5.75 (4.82, 6.26) µg ml(-1) h; in children, the estimated median AUC was 6.96 (5.85, 7.91) µg ml(-1) h, whilst the predicted value was 6.45 (5.80, 7.01) µg ml(-1) h.

CONCLUSIONS

These findings suggest that the assumption of an identical (linear or nonlinear) correlation between pharmacokinetic parameters and demographic factors may not hold true across age groups. Whilst the use of modelling enables accurate characterization of pharmacokinetic properties, extrapolations based on such parameter estimates may have limited value due to differences in the impact of developmental growth across populations.

From trial and error to trial simulation III: a framework for interim analysis in efficacy trials with antidepressant drugs

Clinical trials with antidepressant drugs often fail to detect drug effect, even with drugs that are known to be efficacious. In a previous publication, we showed that a model-based approach is required to address some of the existing challenges in the design of clinical trial protocols. Here, we illustrate how the implementation of an interim analysis (IA) may help to identify studies that are headed for failure, early in the trial before completion of treatment. In contrast to traditional IA procedures, an adaptive Bayesian approach is proposed to optimize the timing of analysis and decision criteria for futility and efficacy, taking into account enrollment rate and treatment response at intermediate visits in the trial. Validation procedures involving re-enrollment of patients confirmed the performance of the method. Our findings reveal that optimization of the timing and decision criteria at the interim stage is critical for the accuracy of the conclusions about treatment efficacy or futility.