Can we do better with Mylotarg? Model-based assessment of opportunities to improve therapeutic index.

Late-stage clinical trial failures increase the overall cost and risk of bringing new drugs to market. Determining the pharmacokinetic (PK) drivers of toxicity and efficacy in preclinical studies and early clinical trials supports quantitative optimization of drug schedule and dose through computational modeling. Additionally, this approach permits prioritization of lead candidates with better PK properties early in development. Mylotarg is an antibody-drug conjugate (ADC) that attained U.S. Food and Drug Administration (FDA) approval under a fractionated dosing schedule after 17 years of clinical trials, including a 10-year period on the market resulting in hundreds of fatal adverse events. Although ADCs are often considered lower risk for toxicity due to their targeted nature, off-target activity and liberated payload can still constrain dosing and drive clinical failure. Under its original schedule, Mylotarg was dosed infrequently at high levels, which is typical for ADCs because of their long half-lives. However, our PK modeling suggests that these regimens increase maximum plasma concentration (Cmax)-related toxicities while producing suboptimal exposures to the target receptor. Our analysis demonstrates that the benefits of dose fractionation for Mylotarg tolerability should have been obvious early in the drug's clinical development and could have curtailed the proliferation of ineffective Phase III studies. We also identify schedules likely to be even more efficacious without compromising on tolerability. Alternatively, a longer-circulating Mylotarg formulation could obviate the need for dose fractionation, allowing superior patient convenience. Early-stage PK optimization through quantitative modeling methods can accelerate clinical development and prevent late-stage failures.

An optimization model for monthly time-step drilling schedule under planned field production.

The field production profile over the yearly horizon is planned for a balance between economy, security, and sustainability of energy. An optimal drilling schedule is required to achieve the planned production profile with minimized drilling frequency and summation. In this study, we treat each possible production process of each well as a dependent time series and the basic unit. Then we ensemble all of them into a tensor. Based on formulated tensor calculation and Lasso regularization, a linear mathematical optimization model for well drilling schedule was developed. The model is aimed at minimizing production profile error while optimizing drilling frequency and summation. Although the model proposed in this work requires more memory consumption to be solved using a computer, it is assured as a linear model and could be numerically globally solved in a stable and efficient way using gradient descent, avoiding complex nonlinear programming problems. Main input data and parameters involved in the model are analyzed in detail to understand the effects of different production parameters on the drilling schedule and production profile. The proposed model in this work can evaluate the manual drilling schedule and automatically generate an optimized drilling schedule for the gas field, significantly reducing development plan formulation time.

A Variable-Length Chromosome Genetic Algorithm for Time-Based Sensor Network Schedule Optimization.

Scheduling sensor nodes has an important role in real monitoring applications using sensor networks, lowering the power consumption and maximizing the network lifetime, while maintaining the satisfaction to application requirements. Nevertheless, this problem is usually very complex and not easily resolved by analytical methods. In a different manner, genetic algorithms (GAs) are heuristic search strategies that help to find the exact or approximate global optimal solution efficiently with a stochastic approach. Genetic algorithms are advantageous for their robustness to discrete and noisy objective functions, as they are only evaluated at independent points without requirements of continuity or differentiability. However, as explained in this paper, a time-based sensor network schedule cannot be represented by a chromosome with fixed length that is required in traditional genetic algorithms. Therefore, an extended genetic algorithm is introduced with variable-length chromosome (VLC) along with mutation and crossover operations in order to address this problem. Simulation results show that, with help of carefully defined fitness functions, the proposed scheme is able to evolve the individuals in the population effectively and consistently from generation to generation towards optimal ones, and the obtained network schedules are better optimized in comparison with the result of algorithms employing a fixed-length chromosome.

Light electric vehicle charging strategy for low impact on the grid.

The alarming increase in the average temperature of the planet due to the massive emission of greenhouse gases has stimulated the introduction of electric vehicles (EV), given transport sector is responsible for more than 25% of the total global CO2 emissions. EV penetration will substantially increase electricity demand and, therefore, an optimization of the EV recharging scenario is needed to make full use of the existing electricity generation system without upgrading requirements. In this paper, a methodology based on the use of the temporal valleys in the daily electricity demand is developed for EV recharge, avoiding the peak demand hours to minimize the impact on the grid. The methodology assumes three different strategies for the recharge activities: home, public buildings, and electrical stations. It has been applied to the case of Spain in the year 2030, assuming three different scenarios for the growth of the total fleet: low, medium, and high. For each of them, three different levels for the EV penetration by the year 2030 are considered: 25%, 50%, and 75%, respectively. Only light electric vehicles (LEV), cars and motorcycles, are taken into account given the fact that batteries are not yet able to provide the full autonomy desired by heavy vehicles. Moreover, heavy vehicles have different travel uses that should be separately considered. Results for the fraction of the total recharge to be made in each of the different recharge modes are deduced with indication of the time intervals to be used in each of them. For the higher penetration scenario, 75% of the total park, an almost flat electricity demand curve is obtained. Studies are made for working days and for non-working days.

Algorithm comparison for schedule optimization in MR fingerprinting.

In MR Fingerprinting, the flip angles and repetition times are chosen according to a pseudorandom schedule. In previous work, we have shown that maximizing the discrimination between different tissue types by optimizing the acquisition schedule allows reductions in the number of measurements required. The ideal optimization algorithm for this application remains unknown, however. In this work we examine several different optimization algorithms to determine the one best suited for optimizing MR Fingerprinting acquisition schedules.

Aapplied analysis of radiotherapy apparatus schedule optimization for tumor patients / 中华放射肿瘤学杂志

Objective To design and develop a treatment unit selecting system in aim of enhancing work efficiency and safety,adjusting treatment unit workload,improving quality of medical care.Methods Various treatment techniques,immobilization devices and setup verification devices were modeled in software.Workload of treatment units were extract from the Record and Verify System.These two types of information were then combined with the unit's workload capability to calculate the optimal radiotherapy apparatus for tumor patient.Results The system had finished selecting radiotherapy apparatusv for more than 5 000 patients since its taking place of the old patient selecting methods.Maximum variation of daily treatment duration between treatment units had decreased from 4-5 hours (mean 4.22 hours) to 1-2 hours (mean 1.84 hours) since the system have been put into operation.Workload and device configuration of various units could be controlled by easily editing of the system configuration file.Conclusions The system developed not only accomplish patient selecting in an optimal and safe way,but also provide a way of easily control the treatment unit workload.