Considerations for Evaluating Treatment Effects From Randomized Clinical Trials

Functional loss and mortality in randomized clinical trials for amyotrophic lateral sclerosis: to combine, or not to combine – that is the estimand

Amyotrophic lateral sclerosis is a rapidly progressive disease leading to death in, on average, 3–5 years after first symptom onset. Consequently, there are frequently a non‐negligible number of patients who die during the course of a clinical trial. This introduces bias in end points such as daily functioning, muscle strength, and quality of life. In this paper, we outline how the choice of strategy to handle death affects the interpretation of the trial results. We provide a general overview of the considerations, positioned in the estimand framework, and discuss the possibility that not every strategy provides a clinically relevant answer in each setting. The relevance of a strategy changes as a function of the intended trial duration, hypothesized treatment effect, and population included. It is important to consider this trade‐off at the design stage of a clinical trial, as this will clarify the exact research question that is being answered, and better guide the planning, design, and analysis of the study.

Estimating the treatment effect for adherers using multiple imputation

Randomized controlled trials are considered the gold standard to evaluate the treatment effect (estimand) for efficacy and safety. According to the recent International Council on Harmonization (ICH)-E9 addendum (R1), intercurrent events (ICEs) need to be considered when defining an estimand, and principal stratum is one of the five strategies to handle ICEs. Qu et al. (2020, Statistics in Biopharmaceutical Research 12:1-18) proposed estimators for the adherer average causal effect (AdACE) for estimating the treatment difference for those who adhere to one or both treatments based on the causal-inference framework, and demonstrated the consistency of those estimators; however, this method requires complex custom programming related to high-dimensional numeric integrations. In this article, we implemented the AdACE estimators using multiple imputation (MI) and constructed confidence intervals (CIs) through bootstrapping. A simulation study showed that the MI-based estimators provided consistent estimators with the nominal coverage probabilities of CIs for the treatment difference for the adherent populations of interest. As an illustrative example, the new method was applied to data from a real clinical trial comparing two types of basal insulin for patients with type 1 diabetes.

Coronavirus Disease 2019 Related Clinical Studies: A Cross-Sectional Analysis

OBJECTIVE

The quality and rationality of many recently registered clinical studies related to coronavirus disease 2019 (COVID-19) needs to be assessed. Hence, this study aims to evaluate the current status of COVID-19 related registered clinical trial.

METHODS

We did an electronic search of COVID-19 related clinical studies registered between December 1, 2019 and February 21, 2020 (updated to May 28, 2020) from the ClinicalTrials.gov, and collected registration information, study details, recruitment status, characteristics of the subjects, and relevant information about the trial implementation process.

RESULTS

A total of 1,706 studies were included 10.0% of which (n=171) were from France, 943 (55.3%) used an interventional design, and 600 (35.2%) used an observational design. Most of studies (73.6%) aimed to recruit fewer than 500 people. Interferon was the main prevention program, and antiviral drugs were the main treatment program. Hydroxychloroquine and chloroquine (230/943, 24.4%) were widely studied. Some registered clinical trials are incomplete in content, and 37.4% of the 1,706 studies may have had insufficient sample size.

CONCLUSION

The quality of COVID-19 related studies needs to be improved by strengthening the registration process and improving the quality of clinical study protocols so that these clinical studies can provide high-quality clinical evidence related to COVID-19.

Comprehensive Survey of Clinical Trials Registration for Melanoma Immunotherapy in the ClinicalTrials.gov

Objective: Comprehensively evaluate the immunotherapeutic clinical trials and provide reference for melanoma treatment and research.

Methods: The website of ClinicalTrials.gov was searched to retrieve and download all registered clinical trials for melanoma immunotherapy on August 1 (updated on August 25), 2019. All registration trials met the inclusion criteria were collected regardless of the type of study, the status of recruitment, and the results of the study. The general characteristics, methodological characteristics, and the types of immunotherapeutic drugs included of these trials were analyzed.

Results: Finally, 242 eligible trials were included and evaluated. Of them, 30.6% were completed, 16.9% were terminated, and two were withdrawn; 77.7% recruited less than 100 participants; 30.5% were randomized; 45.5% was single group assignment; 88.8% were not masked; the primary purpose was treatment; 44.2% had data on monitoring committees; 27.7% used US FDA-regulated immunization drugs; 78.5% without results posted; 43.0% were sponsored by the industry. Immunological checkpoint inhibitors were most often studied, with 53.6% of the trials involving PD-1, the most commonly studied was Nivolumab.

Conclusions: Currently, most of the registered clinical trials for melanoma immunotherapy were interventional open-label trials. Most immunotherapy research hotspots were in the FDA-regulated drug product, and a few trials reported available test results. It is necessary to strengthen the supervision of results and explore and disseminate more effective and safe immunotherapy methods.

Clinical Trial Registration and Reporting: Drug Therapy and Prevention of Cardiac-Related Infections

Objective: Clinical trials are the source of evidence. ClinicalTrials.gov is valuable for analyzing current conditions. Until now, the state of drug interventions for heart infections is unknown. The purpose of this study was to comprehensively assess the characteristics of trials on cardiac-related infections and the status of drug interventions.

Methods: The website ClinicalTrials.gov was used to obtain all registered clinical trials on drug interventions for cardiac-related infections as of February 16, 2019. All registration studies were collected, regardless of their recruitment status, research results, and research type. Registration information, results, and weblink-publications of those trials were analyzed.

Results: A total of 45 eligible trials were evaluated and 86.7% of them began from or after 2008 while 91.1% of them adopted interventional study design. Of all trials, 35.6% were completed and 15.6% terminated. Besides, 62.2% of interventional clinical trials recruited more than 100 subjects. Meanwhile, 86.7% of the eligible trials included adult subjects only. Of intervention trials, 65.8% were in the third or fourth phase; 78.1% adopted randomized parallel assignment, containing two groups; 53.6% were masking, and 61.0% described treatment. Moreover, 41.5% of the trials were conducted in North America while 29.3% in Europe. Sponsors for 40.0% of the studies were from the industry. Furthermore, 48.9% of the trials mentioned information on monitoring committees, 24.4% have been published online, and 13.3% have uploaded their results. Drugs for treatments mainly contained antibiotics, among which glycopeptides, β-lactams, and lipopeptides were the most commonly studied ones in experimental group, with the former ones more common. Additionally, 16.2% of the trials evaluated new antimicrobials.

Conclusions: Most clinical trials on cardiac-related infections registered at ClinicalTrials.gov were interventional randomized controlled trials (RCTs) for treatment. Most drugs focused in trials were old antibiotics, and few trials reported valid results. It is necessary to strengthen supervision over improvements in results, and to combine antibacterial activity with drug delivery regimens to achieve optimal clinical outcomes.

The remarkable therapeutic potential of response-based dose individualisation in drug trials and patient care

The FDA reported that most drugs are effective in only 25-62% of patients. Although many drugs require dose individualisation in clinical practice, dose-finding trials usually aim to identify an optimal dose for the patient population. Such a dose would be suboptimal for many patients. Simulations show that individualised dose titration, balancing efficacy against toxicity, can remarkably increase the response rate - doubling it in some situations. Dose titration in a clinical trial can efficiently establish the realistic expectations for the drug's true utility in a trial setting that reflects clinical practice, as well as generate important knowledge to guide patient care through informative drug labels. This design answers key questions truly relevant to patient care that other designs cannot - will a patient benefit from a given therapy, to what extent and at what dose? Therefore, response-based dose titration should be considered for dose-finding trials, where appropriate, for drugs that will eventually be used this way in the clinic.

Estimands: A More Strategic Approach to Study Design and Analysis

Do we always precisely define the treatment effects that our clinical trial will estimate? Our tenet is that this is not always done, or is done inadequately. This lack of clarity can result in a misalignment among trial objectives, trial design, and statistical methods. We will discuss these challenges and present an improved framework using estimands that is proposed in a draft International Council for Harmonization (ICH) E9 addendum.