Cardiac risk assessment based on early Phase I data and PK-QTc analysis is concordant with the outcome of thorough QTc trials: an assessment based on eleven drug candidates.

Cardiac safety assessment is a key regulatory requirement for almost all new drugs. Until recently, one evaluation aspect was via a specifically designated, expensive, and resource intensive thorough QTc study, and a by-time-point analysis using an intersection-union test (IUT). ICH E14 Q&A (R3) (http://www.ich.org/fileadmin/Public_Web_Site/ICH_Products/Guidelines/Efficacy/E14/E14_Q_As_R3__Step4.pdf) allows for analysis of the PK-QTc relationship using early Phase I data to assess QTc liability. In this paper, we compared the cardiac risk assessment based on the early Phase I analysis with that from a thorough QTc study across eleven drug candidate programs, and demonstrate that the conclusions are largely the same. The early Phase I analysis is based upon a linear mixed effect model with known covariance structure (Dosne et al. in Stat Med 36(24):3844-3857, 2017). The treatment effect was evaluated at the supratherapeutic Cmax as observed in the thorough QTc study using a non-parametric bootstrap analysis to generate 90% confidence intervals for the treatment effect, and implementation of the standardized methodology in R and SAS software yielded consistent results. The risk assessment based on the concentration-response analysis on the early Phase I data was concordant with that based on the standard analysis of the thorough QTc study for nine out of the eleven drug candidates. This retrospective analysis is consistent with and supportive of the conclusion of a previous prospective analysis by Darpo et al. (Clin Pharmacol Ther 97(4):326-335, 2015) to evaluate whether C-QTc analysis can detect QTc effects in a small study with healthy subjects.

A nonparametric method to detect increased frequencies of adverse drug reactions over time.

Signal detection is routinely applied to spontaneous report safety databases in the pharmaceutical industry and by regulators. As an example, methods that search for increases in the frequencies of known adverse drug reactions for a given drug are routinely applied, and the results are reported to the health authorities on a regular basis. Such methods need to be sensitive to detect true signals even when some of the adverse drug reactions are rare. The methods need to be specific and account for multiplicity to avoid false positive signals when the list of known adverse drug reactions is long. To apply them as part of a routine process, the methods also have to cope with very diverse drugs (increasing or decreasing number of cases over time, seasonal patterns, very safe drugs versus drugs for life-threatening diseases). In this paper, we develop new nonparametric signal detection methods, directed at detecting differences between a reporting and a reference period, or trends within a reporting period. These methods are based on bootstrap and permutation distributions, and they combine statistical significance with clinical relevance. We conducted a large simulation study to understand the operating characteristics of the methods. Our simulations show that the new methods have good power and control the family-wise error rate at the specified level. Overall, in all scenarios that we explored, the method performs much better than our current standard in terms of power, and it generates considerably less false positive signals as compared to the current standard.

Exact and asymptotic inference in clinical trials with small event rates under inverse sampling.

In this paper, we discuss statistical inference for a 2 × 2 table under inverse sampling, where the total number of cases is fixed by design. We demonstrate that the exact unconditional distributions of some relevant statistics differ from the distributions under conventional sampling, where the sample size is fixed by design. This permits us to define a simple unconditional alternative to Fisher's exact test. We provide an asymptotic argument including simulations to demonstrate that there is little power loss associated with the alternative test when the expected event rates are very small. We then apply the method to design a clinical trial in cataract surgery, where a rare side effect occurs in one in 1000 patients. The objective of the trial is to demonstrate that adjuvant treatment with an antibiotic will reduce this risk to one in 2000. We use an inverse sampling design and demonstrate how to set this up in a sequential manner. Particularly simple stopping rules can be defined when using the unconditional alternative to Fisher's exact test.

Platform trial design for neurofibromatosis type 1, NF2-related schwannomatosis and non-NF2-related schwannomatosis: A potential model for rare diseases.

Background: Neurofibromatosis type 1, NF2-related schwannomatosis and non-NF2-related schwannomatosis (grouped under the abbreviation "NF") are rare hereditary tumor predisposition syndromes. Due to the low prevalence, variability in the range, and severity of manifestations, as well as limited treatment options, these conditions require innovative trial designs to accelerate the development of new treatments. Methods: Within European Patient-Centric Clinical Trial Platforms (EU-PEARL), we designed 2 platform-basket trials in NF. The trials were designed by a team of multidisciplinary NF experts and trial methodology experts. Results: The trial will consist of an observational and a treatment period. The observational period will serve as a longitudinal natural history study. The platform trial design and randomization to a sequence of available interventions allow for the addition of interventions during the trial. If a drug does not meet the predetermined efficacy endpoint or reveals unacceptable toxicities, participants may stop treatment on that arm and re-enter the observational period, where they can be re-randomized to a different treatment arm if eligible. Intervention-specific eligibility criteria and endpoints are listed in intervention-specific-appendices, allowing the flexibility and adaptability needed for highly variable and rare conditions like NF. Conclusions: These innovative platform-basket trials for NF may serve as a model for other rare diseases, as they will enhance the chance of identifying beneficial treatments through optimal learning from a small number of patients. The goal of these trials is to identify beneficial treatments for NF more rapidly and at a lower cost than traditional, single-agent clinical trials.

Exposure-response modeling for extrapolation from adult to pediatric patients who differ with respect to prognostic factors: Application to everolimus.

Pediatric extrapolation is essential for bringing treatments to the pediatric population, especially for indications where the recruitment of pediatric patients into clinical trials is difficult and where fully powered trials are impossible. Often a similar exposure-response relationship between adult and pediatric patients can be assumed, but just matching exposures can be misleading when some prognostic factors for efficacy differ between those two patient populations. We present an example in liver transplantation where different study designs led to different (time-dependent) hazards between populations. Only after accounting for this difference an apparent mismatch between the extrapolation from adults and the pediatric study could be resolved. This article also exemplifies a clear scientific, methodological approach of pediatric extrapolation, including model building in adults, extrapolation to pediatrics, qualification of the extrapolation, and derivation of the actual pediatric efficacy.

Non-parametric estimators of a monotonic dose-response curve and bootstrap confidence intervals.

In this paper we consider study designs which include a placebo and an active control group as well as several dose groups of a new drug. A monotonically increasing dose-response function is assumed, and the objective is to estimate a dose with equivalent response to the active control group, including a confidence interval for this dose. We present different non-parametric methods to estimate the monotonic dose-response curve. These are derived from the isotonic regression estimator, a non-negative least squares estimator, and a bias adjusted non-negative least squares estimator using linear interpolation. The different confidence intervals are based upon an approach described by Korn, and upon two different bootstrap approaches. One of these bootstrap approaches is standard, and the second ensures that resampling is done from empiric distributions which comply with the order restrictions imposed. In our simulations we did not find any differences between the two bootstrap methods, and both clearly outperform Korn's confidence intervals. The non-negative least squares estimator yields biased results for moderate sample sizes. The bias adjustment for this estimator works well, even for small and moderate sample sizes, and surprisingly outperforms the isotonic regression method in certain situations.