Innovation at the Intersection of Clinical Trials and Real-World Data Science to Advance Patient Care.

While efficacy and safety data collected from randomized clinical trials are the evidentiary standard for determining market authorization, this alone may no longer be sufficient to address the needs of key stakeholders (regulators, providers, and payers) and guarantee long-term success of pharmaceutical products. There is a heightened interest from stakeholders on understanding the use of real-world evidence (RWE) to substantiate benefit-risk assessment and support the value of a new drug. This review provides an overview of real-world data (RWD) and related advances in the regulatory framework, and discusses their impact on clinical research and development. A framework for linking drug development decisions with the value proposition of the drug, utilizing pharmacokinetic-pharmacodynamic-pharmacoeconomic models, is introduced. The summary presented here is based on the presentations and discussion at the symposium entitled Innovation at the Intersection of Clinical Trials and Real-World Data to Advance Patient Care at the American Society for Clinical Pharmacology and Therapeutics (ASCPT) 2017 Annual Meeting.

Harnessing the Potential of Emerging Digital Health and Biological Sampling Technologies for Clinical Drug Development: Promise to Reality.

Advances in emerging innovative technologies have led to optimistic outlooks on their transformative potential for healthcare and clinical trials.1 Given the increased attention, this white paper by the International Consortium for Innovation and Quality in Pharmaceutical Development (IQ) presents perspectives on pharmaceutical and biotechnology industry trends for innovative digital health, adherence, and outpatient sampling technologies. As stimulus for cross-company scientific dialog points to consider for adoption, implementation, and recommendations to broaden uptake are proposed.

A proposal for scientific framework enabling specific population drug dosing recommendations.

Over the last 3 decades, there has been little change in the paradigm to derive dosing recommendations for specific populations (e.g., renal failure, elderly, or obese patients) despite better understanding of clearance pathways in these groups and availability of modeling and simulation tools. Dosing recommendations for specific populations are often incomplete or unavailable at the time of drug approval. Currently, there is no regulatory framework to incorporate model-based dosing recommendations for specific populations. This paper proposes a scientific framework for using modeling and simulation to support specific population dosing recommendations. This framework creates a knowledgebase of drug and population attributes where model-based approaches can be developed to inform dosing recommendations. The framework may benefit patients by having reliable dosing information at the time of drug approval. Patients with conditions where studies are difficult to perform would benefit from dosing based on state-of-the-art knowledge. Industry and regulators would benefit from a scientific and efficient approach to improve specific population prediction. A research approach to determine specific population dose prediction is discussed along with challenges and risks. We hope to initiate a dialogue to explore the role of modeling based on data for drugs with similar clearance mechanisms to predict drug dosing.

Earlier sustained virologic response end points for regulatory approval and dose selection of hepatitis C therapies.

BACKGROUND & AIMS: Trials of therapies for chronic hepatitis C have used detection of hepatitis C virus (HCV) at week 24 of follow-up (sustained virologic response [SVR] 24) as a primary end point. However, there is increasing evidence that most patients who have an SVR at earlier time points (such as SVR12) maintain it until week 24. Use of earlier time points for key regulatory decisions (SVR12) and dose selection (SVR4) could facilitate HCV drug development. METHODS: We assessed data from 15 phase II and III trials, 3 pediatric studies, and 5 drug-development programs to determine the concordance between SVR24 and SVR12 or SVR4. Data were analyzed from groups of subjects who received various combinations and regimens with interferon, pegylated-interferon, ribavirin, and direct-acting antivirals. RESULTS: The positive predictive value (PPV) of SVR12 was 98% and the negative predictive value (NPV) was 99% for SVR24 among subjects with genotype 1 HCV infection. A similar level of concordance was observed for subjects with HCV genotype 2 or 3 infections, as well as in pediatric studies. About 2% of subjects who achieved an SVR12 subsequently relapsed by week 24 (did not achieve an SVR24). Furthermore, the treatment effect size (difference between treatment and active control arms) was similar for subjects with SVR12 and SVR24. The PPV of SVR4 was 91% and the NPV was 98% for SVR24 in subjects with genotype 1 HCV infection. CONCLUSIONS: SVR12 and SVR24 measurements were concordant in a large population of subjects with HCV infection who participated in clinical trials with various treatment regimens and durations. SVR12 is suitable as a primary end point for regulatory approval. SVR4 might be used to guide dose and treatment strategies in trials.

Response-guided telaprevir therapy in prior relapsers? The role of bridging data from treatment-naïve and experienced subjects.

The purpose of this report is to illustrate the US Food and Drug Administration's rationale for approving response-guided therapy (RGT) for telaprevir (TVR) in combination with pegylated interferon-α and ribavirin (P/R) for the treatment of adults with genotype 1 chronic hepatitis C who were prior relapsers. RGT was prospectively evaluated in two registration trials of treatment-naïve subjects. In these studies, RGT allowed subjects who achieved undetectable hepatitis C virus RNA from weeks 4 and 12, known as extended rapid virologic response (eRVR), to stop all treatments at 24 weeks. A patient without eRVR received an additional 36 weeks of P/R after 12 weeks of a TVR triple regimen (total of 48 weeks). However, RGT in prior P/R relapsers was not prospectively evaluated. Empirical cross-trial data indicated high sustained virologic response rates (>90%) in prior relapsers achieving eRVR, irrespective of P/R duration (24 or 48 weeks). Further analyses demonstrated that interferon responsiveness does not change in P/R-experienced subjects with a second round of P/R. The comparability in interferon responsiveness across treatment courses allowed us to bridge data between treatment-naïve and P/R-experienced subjects to support the approval of RGT in prior relapse subjects.

From in vitro EC50 to in vivo dose-response for antiretrovirals using an HIV disease model. Part II: application to drug development.

The purpose of this research was to qualify a previously derived quantitative model-based framework that proposed an in vitro-in vivo linkage to predict the dose-response relationship of an antiretroviral (ARV) new molecular entity (NME) in a monotherapy trial. Human immunodeficiency virus (HIV)-1 ribonucleic acid (RNA) data from monotherapy trials in ARV naïve/experienced HIV-infected subjects for eight drugs (i.e. application case, a representative drug for a given class used for external validation) across four distinct classes of ARV agents (co-receptor antagonists; non-nucleoside reverse transcriptase inhibitors; nucleotide reverse transcriptase inhibitors; and integrase strand transfer inhibitors) were obtained. Using the in vitro EC50 (protein binding corrected) and a class-specific scaling factor (SF), the in vivo IC50 was calculated for each drug. The integrated pharmacokinetic (PK)-pharmacodynamic (PD) disease model used the predicted in vivo IC50 to simulate the HIV-1 RNA-time profiles for dosing regimens that were studied in the monotherapy trials for each drug. The simulated HIV-1 RNA time profiles were then compared to the observed data. The simulated HIV-1 RNA-time profiles matched well with those observed in the monotherapy trials except for one drug in the INSTIs class. The derived SF represents a useful in vitro-in vivo linkage to predict the dose-response relationship for a NME using in vitro data. The mechanistic PK-PD disease model-based framework is useful to assist the dose selection for monotherapy trials and comparator modeling approaches.

From in vitro EC50 to in vivo dose-response for antiretrovirals using an HIV disease model. Part I: a framework.

Plasma drug concentrations and human immunodeficiency virus (HIV)-1 ribonucleic acid (RNA) data from short-term monotherapy trials in antiretroviral (ARV) naïve/experienced HIV-infected patients for one drug (i.e., lead case, a representative drug for a given class) from four distinct classes of ARV drugs [non-nucleoside reverse transcriptase inhibitors (NNRTIs); nucleotide reverse transcriptase inhibitors (NtRTIs); co-receptor antagonists (CRAs); and integrase strand transfer inhibitors (INSTIs)] were obtained. For each drug (the lead case; for example, etravirine for NNRTIs), the pharmacokinetic (PK)-pharmacodynamic (PD) disease model was used to estimate the in vivo IC50 value based on the relationship between plasma drug concentrations and HIV-1 RNA decline. The disease model was a mechanistic viral dynamic model with the addition of a transduction delay for HIV-1 RNA decay. The model characterized the observed time-course of plasma drug concentrations and HIV-1 RNA in each lead case trial reasonably well. The estimated viral dynamic parameters were in good agreement with the literature values. The scaling factor (SF) was calculated as a ratio using the estimated in vivo IC50 value and protein binding corrected EC50 value for the lead case. For NNRTIs, the calculated SF was found to be 9.47 based on etravirine analysis. Using the same approach, the SF for CRAs (maraviroc), INSTIs (GSK1349572), and NtRTIs (tenofovir) were 4.35, 1.63, and 0.27. The SF is reflective of the uncertainty between in vitro measurement of drug potency and the specific model-based drug potency parameter (IC50). The use of SF and PK-PD disease models can be a valuable tool to predict dose-response of NMEs and support rational dose selection for monotherapy trials.

Interferon responsiveness does not change in treatment-experienced hepatitis C subjects: implications for drug development and clinical decisions.

BACKGROUND: The purpose of this research was to compare interferon (IFN) responsiveness in treatment-naive and pegylated interferon α-ribavirin (P/R)-experienced subjects and to understand the implications of comparability in IFN responsiveness across treatment courses on drug development and clinical decision making. METHODS: Data from 3750 subjects treated with P/R in 8 trials were reviewed. The change in hepatitis C virus (HCV) RNA at week 4 in response to P/R was compared according to end-of-study (EOS) status (responder, relapser, partial and null responder) for treatment-naive subjects and the previous P/R response status (known as prior relapsers, prior partial responders, and prior null responders at the baseline) for P/R-experienced subjects. RESULTS: In subjects receiving a first course of P/R treatment (treatment-naive subjects), HCV RNA change after 4 weeks of P/R was correlated with EOS status on a P/R regimen. Importantly, for the first time, we have quantitatively demonstrated that IFN responsiveness in P/R-experienced subjects administered a second course of P/R treatment was similar to the IFN responsiveness in the treatment-naive subjects with corresponding EOS status. CONCLUSIONS: We contend that P/R-experienced subjects are represented within treatment-naive subjects. There are 2 important implications of this finding: (1) from a drug development perspective, a successful direct antiviral plus P/R therapy (IFN-based triple therapy) trial in P/R-experienced subjects may serve as supportive evidence in treatment-naive subjects; and (2) from a clinical decision perspective, previous P/R exposure should not alter new treatment decisions involving IFN-based triple therapy, as the IFN responsiveness to a second course of IFN is comparable.

Impact of pharmacometric analyses on new drug approval and labelling decisions: a review of 198 submissions between 2000 and 2008.

Pharmacometric analyses have become an increasingly important component of New Drug Application (NDA) and Biological License Application (BLA) submissions to the US FDA to support drug approval, labelling and trial design decisions. Pharmacometrics is defined as a science that quantifies drug, disease and trial information to aid drug development, therapeutic decisions and/or regulatory decisions. In this report, we present the results of a survey evaluating the impact of pharmacometric analyses on regulatory decisions for 198 submissions during the period from 2000 to 2008. Pharmacometric review of NDAs included independent, quantitative analyses by FDA pharmacometricians, even when such analysis was not conducted by the sponsor, as well as evaluation of the sponsor's report. During 2000-2008, the number of reviews with pharmacometric analyses increased dramatically and the number of reviews with an impact on approval and labelling also increased in a similar fashion. We also present the impact of pharmacometric analyses on selection of paediatric dosing regimens, approval of regimens that had not been directly studied in clinical trials and provision of evidence of effectiveness to support a single pivotal trial. Case studies are presented to better illustrate the role of pharmacometric analyses in regulatory decision making.

Antiviral Information Management System (AIMS): a prototype for operational innovation in drug development.

This article presents a prototype for an operational innovation in knowledge management (KM). These operational innovations are geared toward managing knowledge efficiently and accessing all available information by embracing advances in bioinformatics and allied fields. The specific components of the proposed KM system are (1) a database to archive hepatitis C virus (HCV) treatment data in a structured format and retrieve information in a query-capable manner and (2) an automated analysis tool to inform trial design elements for HCV drug development. The proposed framework is intended to benefit drug development by increasing efficiency of dose selection and improving the consistency of advice from US Food and Drug Administration (FDA). It is also hoped that the framework will encourage collaboration among FDA, industry, and academic scientists to guide the HCV drug development process using model-based quantitative analysis techniques.

The need for modeling and simulation to design clinical investigations in children.

Recent legislation in the United States and Europe has resulted in an increased number of clinical trials of pharmaceutical agents in children. Creating a well-designed clinical trial that can be successfully completed is a challenging task, particularly as the study population includes younger and smaller children. Although there are some established principles for initially estimating appropriate doses of pharmaceutical agents in children based on known effective doses in adults, these rules are inadequate as the sole basis for designing a clinical trial in children. Factors such as maturation of metabolizing enzymes, relative physical maturation of the child, and altered absorption because of physiological differences in adults and children may contribute to alterations in the dose-exposure relationship. To account for the impact of these potential factors on a clinical trial, the use of modeling and simulation is necessary to anticipate the influence these variables can have on the desired clinical question to be addressed. The examples presented in this article highlight the principle that modeling and simulation is critical for adequately designing pediatrics trials.

Leveraging prior quantitative knowledge in guiding pediatric drug development: a case study.

The manuscript presents the FDA's focus on leveraging prior knowledge in designing informative pediatric trial through this case study. In developing written request for Drug X, an anti-hypertensive for immediate blood pressure (BP) control, the sponsor and FDA conducted clinical trial simulations (CTS) to design trial with proper sample size and support the choice of dose range. The objective was to effectively use prior knowledge from adult patients for drug X, pediatric data from Corlopam (approved for a similar indication) trial and general experience in developing anti-hypertensive agents. Different scenarios governing the exposure response relationship in the pediatric population were simulated to perturb model assumptions. The choice of scenarios was based on the past observation that pediatric population is less responsive and sensitive compared with adults. The conceptual framework presented here should serve as an example on how the industry and FDA scientists can collaborate in designing the pediatric exclusivity trial. Using CTS, inter-disciplinary scientists with the sponsor and FDA can objectively discuss the choice of dose range, sample size, endpoints and other design elements. These efforts are believed to yield plausible trial design, qrational dosing recommendations and useful labeling information in pediatrics.

Leveraging prior quantitative knowledge to guide drug development decisions and regulatory science recommendations: impact of FDA pharmacometrics during 2004-2006.

The End-of-Phase 2A meetings are proposed to identify opportunities to make innovative medical products available sooner and to increase the quality of drug applications through early meetings between sponsors and the FDA. This article summarizes the overall experience across 11 pilot End-of-Phase 2A meetings since 2004. Four case studies are presented in more detail to demonstrate the various issues and methods encountered at these meetings. Overall, industry and FDA scientists ranked these meetings to be "very helpful" (average score of 4 on a scale of 1 to 5). In almost all the instances the sponsors changed their drug development plans subsequent to these extensive quantitative analyses-based meetings. A draft Guidance is being developed to be issued in 2008, and we hope this initiative will be resourced by then.

Concentration-QT relationships play a key role in the evaluation of proarrhythmic risk during regulatory review.

The criterion for assessing whether a drug prolongs QT as described in the International Conference on Harmonization topic E14 guideline does not explicitly account for individual drug concentrations. The authors' experience with reviewing QT studies indicates that understanding the relationship, if any, between individual drug concentration and QT change provides important additional information to support regulatory decision making. Therefore, regulatory reviews of "thorough QT" studies routinely include a characterization of the concentration-QT relationship. The authors provide examples to illustrate how the concentration-QT relationship has been used to plan and interpret the thorough QT study, to evaluate QT risk for drugs that have no thorough QT studies, to assess QT risk in subpopulations, to make dose adjustments, and to write informative drug labels.

Semi-mechanistic pharmacodynamic modeling for degarelix, a novel gonadotropin releasing hormone (GnRH) blocker.

An integrated semi-mechanistic pharmacodynamic (PD) model describing the relationship between luteinizing hormone (LH) and testosterone (T) after short-term administration of degarelix was developed. Data from three clinical studies involving, intravenous (IV) and subcutaneous (SC) dosing, in healthy male subjects were available. Degarelix pharmacokinetic (PK) data from all studies were modeled simultaneously. One intravenous study was used to develop the PD model and the two other studies (IV and SC dosing) were used to qualify the model. Degarelix PK follows a two-compartment model and exhibits flip-flop kinetics after subcutaneous dosing. Based on physiological mechanism, the gonadotropin releasing hormone (GnRH) time course was described using a pulsatile release model. A precursor-dependent pool model was used to describe the kinetics of LH in the pituitary and plasma compartment. In males, LH regulates T production in leydig cells. Degarelix inhibits the release of LH from the pool compartment to the plasma compartment leading to decreased T production. The plasma half-life of LH (2.6-3.3 hr) and T (2.7 hr) match well with the literature reports. The proposed PD model reasonably described the time course of LH and T including the LH rebound for short-term studies. The model predicted the time course of LH and T for the second IV and SC dosing studies very well. However, the long term simulations from the final model did not match with literature reports. A modification is suggested based on the physiological understanding of the system. The proposed novel modification to precursor models can be of general use for predicting long term responses.

A new equivalence based metric for predictive check to qualify mixed-effects models.

The main objective of any modeling exercise is to provide a rationale for effective decision making during drug development. The aim of the current simulation experiment was to evaluate the properties of predictive check as a covariate model qualification technique and, more importantly, to introduce and evaluate alternative criteria to qualify models.Original concentration-time profiles (yod) were simulated using a 1-compartment model for an intravenous drug administered to 25 men and 25 women. The typical clearance for male subjects (TVCLm) was assumed to be 5-fold higher than that for female subjects (TVCLf). Fifty such trials under the same design were generated randomly. Predictive check was used as the model qualification tool to study predictive performance of true (males not equal females) and false (males = females) models in the context of maximum likelihood estimation. For each yod, 200 replications were generated to study the properties of a discrepancy variable, a statistic that depends on the model, and a test statistic, a statistic that does not depend on the model. Several qualification criteria were evaluated in assessing predictive performance, such as, predictive p-value (Pp), probability of equivalence (peqv), and probability of rejecting the null hypothesis (data = model) using the Kolmogorov-Smirnov test (pks). The Pp value was calculated using sum of squared errors as a discrepancy variable. For both of the models, the Pp values uniformly ranged between 0 and 1. The pattern of Pp values suggests that qualification of the false model is unlikely. For both of the models, the range of peqv is about 0.95 to 1.0 for concentration at 0.5 hours. However, this is not the case for the concentration at 4 hours, which is primarily dependent on the clearance. The false model (0.35 to 0.50) has poor predictive performance compared with the true model (0.65 to 0.80) using peqv. The pks suggests no difference in the distributions of replicated and original concentrations at all of the time points for both of the models. Discrepancy variables cannot aid in rejecting false models, whereas the use of a test statistic can aid in rejecting false models. However, selection of an informative test statistic is challenging. As far as the qualification criteria are considered, the equivalence-based comparison of a test statistic is more informative than a significance-based comparison. No convincing evidence exists in the literature demonstrating the added advantages of predictive check as a routine model qualification tool over the existing tools, such as diagnostic plots or mechanistic reasoning. However, when a model is to be used for designing a trial, it should at least be able to regenerate the data used to build the model. In such cases, predictive check might offer insights into potential inconsistencies.