Go beyond the limits of genetic algorithm in daily covariate selection practice.

Covariate identification is an important step in the development of a population pharmacokinetic/pharmacodynamic model. Among the different available approaches, the stepwise covariate model (SCM) is the most used. However, SCM is based on a local search strategy, in which the model-building process iteratively tests the addition or elimination of a single covariate at a time given all the others. This introduces a heuristic to limit the searching space and then the computational complexity, but, at the same time, can lead to a suboptimal solution. The application of genetic algorithms (GAs) for covariate selection has been proposed as a possible solution to overcome these limitations. However, their actual use during model building is limited by the extremely high computational costs and convergence issues, both related to the number of models being tested. In this paper, we proposed a new GA for covariate selection to address these challenges. The GA was first developed on a simulated case study where the heuristics introduced to overcome the limitations affecting currently available GA approaches resulted able to limit the selection of redundant covariates, increase replicability of results and reduce convergence times. Then, we tested the proposed GA on a real-world problem related to remifentanil. It obtained good results both in terms of selected covariates and fitness optimization, outperforming the SCM.

A translational model-based approach to inform the choice of the dose in phase 1 oncology trials: the case study of erdafitinib.

PURPOSE: Erdafitinib (JNJ-42756493, BALVERSA) is a tyrosine kinase inhibitor indicated for the treatment of advanced urothelial carcinoma. In this work, a translational model-based approach to inform the choice of the doses in phase 1 trials is illustrated. METHODS: A pharmacokinetic (PK) model was developed to describe the time course of erdafitinib plasma concentrations in mice and rats. Data from multiple xenograft studies in mice and rats were analyzed using the Simeoni tumor growth inhibition (TGI) model. The model parameters were used to derive a range of erdafitinib exposures that might inform the choice of the doses in oncology phase 1 trials. Conversion of exposures to doses was based on preliminary PK assessments from the first-in human (FIH) study. RESULTS: A one-compartment PK disposition model, with linear absorption and dose-dependent clearance, adequately described the PK data in both mice and rats via an allometric scaling approach. The TGI model was able to describe tumor growth dynamics, providing quantitative measurements of erdafitinib antitumor potency in mice and rats. Based on these estimates, ranges of efficacious unbound concentration were identified for erdafitinib in mice (0.642-5.364 µg/L) and rats (0.782-2.565 µg/L). Based on the FIH data, it was possible to transpose exposures into doses and doses of above 4 mg/day provided erdafitinib exposures associated with significant TGI in animals. The findings were in agreement with the results of the FIH trial, in which the first hints of clinical activities were observed at 6 mg. CONCLUSION: The successful modeling exercise of erdafitinib preclinical data showed how translational PK-PD modeling might be a tool to help to inform the choice of the doses in FIH studies.

A Population Dynamic Energy Budget-Based Tumor Growth Inhibition Model for Etoposide Effects on Wistar Rats.

PURPOSE: This work aimed to develop a population PK/PD tumor-in-host model able to describe etoposide effects on both tumor cells and host in Walker-256 tumor-bearing rats. METHODS: Etoposide was investigated on thirty-eight Wistar rats randomized in five arms: two groups of tumor-free animals receiving either placebo or etoposide (10 mg/kg bolus for 4 days) and three groups of tumor-bearing animals receiving either placebo or etoposide (5 or 10 mg/kg bolus for 8 or 4 days, respectively). To analyze experimental data, a tumor-in-host growth inhibition (TGI) model, based on the Dynamic Energy Budget (DEB) theory, was developed. Total plasma and free-interstitial tumor etoposide concentrations were assessed as driver of tumor kinetics. RESULTS: The model simultaneously describes tumor and host growths, etoposide antitumor effect as well as cachexia phenomena related to both the tumor and the drug treatment. The schedule-dependent inhibitory effect of etoposide is also well captured when the intratumoral drug concentration is considered as the driver of the tumor kinetics. CONCLUSIONS: The DEB-based TGI model capabilities, up to now assessed only in mice, are fully confirmed in this study involving rats. Results suggest that well designed experiments combined with a mechanistic modeling approach could be extremely useful to understand drug effects and to describe all the dynamics characterizing in vivo tumor growth studies.

Modeling tumor growth inhibition and toxicity outcome after administration of anticancer agents in xenograft mice: A Dynamic Energy Budget (DEB) approach.

Host features, such as cell proliferation rates, caloric intake, metabolism and energetic conditions, significantly influence tumor growth; at the same time, tumor growth may have a dramatic impact on the host conditions. For example, in clinics, at certain stages of the tumor growth, cachexia (body weight reduction) may become so relevant to be considered as responsible for around 20% of cancer deaths. Unfortunately, anticancer therapies may also contribute to the development of cachexia due to reduced food intake (anorexia), commonly observed during the treatment periods. For this reason, cachexia is considered one of the major toxicity findings to be evaluated also in preclinical studies. However, although various pharmacokinetic-pharmacodynamic (PK-PD) tumor growth inhibition (TGI) models are currently available, the mathematical modeling of cachexia onset and TGI after an anticancer administration in preclinical experiments is still an open issue. To cope with this, a new PK-PD model, based on a set of tumor-host interaction rules taken from Dynamic Energy Budget (DEB) theory and a set of drug tumor inhibition equations taken from the well-known Simeoni TGI model, was developed. The model is able to describe the body weight reduction, splitting the cachexia directly induced by tumor and that caused by the drug treatment under study. It was tested in typical preclinical studies, essentially designed for efficacy evaluation and routinely performed as a part of the industrial drug development plans. For the first time, both the dynamics of tumor and host growth could be predicted in xenograft mice untreated or treated with different anticancer agents and following different schedules. The model code is freely available for downloading at http://repository.ddmore.eu (model number DDMODEL00000274).