Designing Dose-Finding Phase I Clinical Trials: Top 10 Questions That Should Be Discussed With Your Statistician

Meta-analyses of phase I dose-finding studies: Application for the development of protein kinase inhibitors in oncology

This study aimed to assess the feasibility of applying two recent phase I meta-analyses methods to protein kinase inhibitors (PKIs) developed in oncology and to identify situations where these methods could be both feasible and useful. This ancillary study used data from a systematic review conducted to identify dose-finding studies for PKIs. PKIs selected for meta-analyses were required to have at least five completed dose-finding studies involving cancer patients, with available results, and dose escalation guided by toxicity assessment. To account for heterogeneity caused by various administration schedules, some studies were divided into study parts, considered as separate entities in the meta-analyses. For each PKI, two Bayesian random-effects meta-analysis methods were applied to model the toxicity probability distribution of the recommended dose and to estimate the maximum tolerated dose (MTD). Meta-analyses were performed for 20 PKIs including 96 studies corresponding to 115 study parts. The median posterior probability of toxicity probability was below the toxicity thresholds of 0.20 for 70% of the PKIs, even if the resulting credible intervals were very wide. All approved doses were below the MTD estimated for the minimum toxicity threshold, except for one, for which the approved dose was above the MTD estimated for the maximal threshold. The application of phase I meta-analysis methods has been feasible for the majority of PKI; nevertheless, their implementation requires multiple conditions. However, meta-analyses resulted in estimates with large uncertainty, probably due to limited patient numbers and/or between-study variability. This calls into question the reliability of the recommended doses.

A Phase 1 Study of Cabozantinib and Trifluridine/Tipiracil in Metastatic Colorectal Adenocarcinoma

INTRODUCTION

This study determined the safety and recommended phase 2 dose (RP2D) of the multikinase inhibitor cabozantinib in combination with trifluridine/tipiracil (FTD/TPI) in refractory metastatic colorectal carcinoma (mCRC).

PATIENTS AND METHODS

Single institution investigator-initiated phase 1 study using 3+3 design. Eligible mCRC patients had received prior standard regimens. Cabozantinib was given orally (p.o.) at 20 mg (dose level [DL] 0) or 40 mg (DL 1) daily on days 1-28, and FTD/TPI p.o. at 35 mg/m2 on days 1-5 and 8-12 every 28 days. Prophylactic growth-factor support was allowed.

RESULTS

Fifteen patients were enrolled. Median age 56 years (31-80), male (12/15), ECOG 0/1 = 9/6. Three patients were treated at DL 0 and another nine were treated at DL 1, none exhibiting a DLT. Most common any grade (G) treatment related adverse events (TRAE) were diarrhea (50%), nausea (42%), neutropenia (42%), fatigue (33%), and rash (25%). G3-4 TRAE were neutropenia (25%) and thrombocytopenia, hypokalemia, and weight loss (each 8%). No serious TRAE or G5 were reported. The RP2D was determined to be DL 1. Median PFS was 3.8 months (95% CI 1.9-6.8) and disease control rate was 86.7%.

CONCLUSION

The combination of cabozantinib and FTD/TPI is feasible and tolerable at standard doses with the use of growth factors and showed encouraging clinical activity in refractory mCRC.

CLINICALTRIALS

GOV: NCT04868773.

Application of multi-armed bandits to dose-finding clinical designs

Multi-armed bandits are very simple and powerful methods to determine actions to maximize a reward in a limited number of trials. An early phase in dose-finding clinical trials needs to identify the maximum tolerated dose among multiple doses by repeating the dose-assignment. We consider applying the superior selection performance of multi-armed bandits to dose-finding clinical designs. Among the multi-armed bandits, we first consider the use of Thompson sampling which determines actions based on random samples from a posterior distribution. In the small sample size, as shown in dose-finding trials, because the tails of posterior distribution are heavier and random samples are too much variability, we also consider an application of regularized Thompson sampling and greedy algorithm. The greedy algorithm determines a dose based on a posterior mean. In addition, we also propose a method to determine a dose based on a posterior mode. We evaluate the performance of our proposed designs for nine scenarios via simulation studies.

CrossFuse-XGBoost: accurate prediction of the maximum recommended daily dose through multi-feature fusion, cross-validation screening and extreme gradient boosting

In the drug development process, approximately 30% of failures are attributed to drug safety issues. In particular, the first-in-human (FIH) trial of a new drug represents one of the highest safety risks, and initial dose selection is crucial for ensuring safety in clinical trials. With traditional dose estimation methods, which extrapolate data from animals to humans, catastrophic events have occurred during Phase I clinical trials due to interspecies differences in compound sensitivity and unknown molecular mechanisms. To address this issue, this study proposes a CrossFuse-extreme gradient boosting (XGBoost) method that can directly predict the maximum recommended daily dose of a compound based on existing human research data, providing a reference for FIH dose selection. This method not only integrates multiple features, including molecular representations, physicochemical properties and compound-protein interactions, but also improves feature selection based on cross-validation. The results demonstrate that the CrossFuse-XGBoost method not only improves prediction accuracy compared to that of existing local weighted methods [k-nearest neighbor (k-NN) and variable k-NN (v-NN)] but also solves the low prediction coverage issue of v-NN, achieving full coverage of the external validation set and enabling more reliable predictions. Furthermore, this study offers a high level of interpretability by identifying the importance of different features in model construction. The 241 features with the most significant impact on the maximum recommended daily dose were selected, providing references for optimizing the structure of new compounds and guiding experimental research. The datasets and source code are freely available at https://github.com/cqmu-lq/CrossFuse-XGBoost.

CRISPR-Cas System Is an Effective Tool for Identifying Drug Combinations That Provide Synergistic Therapeutic Potential in Cancers

Despite numerous efforts, the therapeutic advancement for neuroblastoma and other cancer treatments is still ongoing due to multiple challenges, such as the increasing prevalence of cancers and therapy resistance development in tumors. To overcome such obstacles, drug combinations are one of the promising applications. However, identifying and implementing effective drug combinations are critical for achieving favorable treatment outcomes. Given the enormous possibilities of combinations, a rational approach is required to predict the impact of drug combinations. Thus, CRISPR-Cas-based and other approaches, such as high-throughput pharmacological and genetic screening approaches, have been used to identify possible drug combinations. In particular, the CRISPR-Cas system (Clustered Regularly Interspaced Short Palindromic Repeats) is a powerful tool that enables us to efficiently identify possible drug combinations that can improve treatment outcomes by reducing the total search space. In this review, we discuss the rational approaches to identifying, examining, and predicting drug combinations and their impact.

Considerations for designing trials targeting muscle dysfunction in exercise oncology

Individuals diagnosed with cancer commonly experience a significant decline in muscle mass and physical function collectively referred to as cancer related muscle dysfunction. This is concerning because impairments in functional capacity are associated with an increased risk for the development of disability and subsequent mortality. Notably, exercise offers a potential intervention to combat cancer related muscle dysfunction. Despite this, research is limited on the efficacy of exercise when implemented in such a population. Thus, the purpose of this mini review is to offer critical considerations for researchers seeking to design studies pertaining to cancer related muscle dysfunction. Namely, 1) defining the condition of interest, 2) determining the most appropriate outcome and methods of assessment, 3) establishing the best timepoint (along the cancer continuum) to intervene, and 4) understanding how exercise prescription can be configured to optimize outcomes.