Personalized drug administrations using Support Vector Machine

The decision-making process regarding drug dose, regularly used in everyday medical practice, is critical to patients' health and recovery. It is a challenging process, especially for a drug with narrow therapeutic ranges, in which a medical doctor decides the quantity (dose amount) and frequency (dose interval) on the basis of a set of available patient features and doctor's clinical experience (a priori adaptation). Computer support in drug dose administration makes the prescription procedure faster, more accurate, objective, and less expensive, with a tendency to reduce the number of invasive procedures. This paper presents an advanced integrated Drug Administration Decision Support System (DADSS) to help clinicians/patients with the dose computing. Based on a support vector machine (SVM) algorithm, enhanced with the random sample consensus technique, this system is able to predict the drug concentration values and computes the ideal dose amount and dose interval for a new patient. With an extension to combine the SVM method and the explicit analytical model, the advanced integrated DADSS system is able to compute drug concentration-to-time curves for a patient under different conditions. A feedback loop is enabled to update the curve with a new measured concentration value to make it more personalized (a posteriori adaptation).

Monte Carlo derivation of filtered tungsten anode X-ray spectra for dose computation in digital mammography

Abstract Objective: Derive filtered tungsten X-ray spectra used in digital mammography systems by means of Monte Carlo simulations. Materials and Methods: Filtered spectra for rhodium filter were obtained for tube potentials between 26 and 32 kV. The half-value layer (HVL) of simulated filtered spectra were compared with those obtained experimentally with a solid state detector Unfors model 8202031-H Xi R/F & MAM Detector Platinum and 8201023-C Xi Base unit Platinum Plus w mAs in a Hologic Selenia Dimensions system using a direct radiography mode. Results: Calculated HVL values showed good agreement as compared with those obtained experimentally. The greatest relative difference between the Monte Carlo calculated HVL values and experimental HVL values was 4%. Conclusion: The results show that the filtered tungsten anode X-ray spectra and the EGSnrc Monte Carlo code can be used for mean glandular dose determination in mammography.

Higher evening antiepileptic drug dose for nocturnal and early-morning seizures

We describe 17 children with nocturnal or early-morning seizures who were switched to a proportionally higher evening dose of antiepileptic drugs and were retrospectively reviewed for seizure outcome and side effects. Of 10 children with unknown etiology, clinical presentation was consistent with nocturnal frontal lobe epilepsy (NFLE) in 5 and benign epilepsy with centrotemporal spikes (BECTS) in 3. After a mean follow-up of 5.3 months, 15 patients were classified as responders: 11 of these became seizure free (5 NFLE, 1 BECTS, 5 with structural lesions) and 4 (2 BECTS, 2 with structural lesions) experienced 75-90% reductions in seizures. Among two nonresponders, seizures in one had failed to resolve with epilepsy surgery. Nine subjects (53%) received monotherapy after dose modification, and none presented with worsening of seizures. Two complained of transient side effects (fatigue/somnolence). Differential dosing led to seizure freedom in 64.7% (11/17) of patients, and 88.2% (15/17) experienced >= 50% reductions in seizures. (C) 2010 Elsevier Inc. All rights reserved.

Competition Between Novelty and Cocaine Conditioned Reward Is Sensitive to Drug Dose and Retention Interval

The conditioned rewarding effects of novelty compete with those of cocaine for control over choice behavior using a place conditioning task. The purpose of the present study was to use multiple doses of cocaine to determine the extent of this competition and to determine whether novelty’s impact on cocaine reward was maintained over an abstinence period. In Experiment 1, rats were conditioned with cocaine (7.5, 20, or 30 mg/kg ip) to prefer one side of an unbiased place conditioning apparatus relative to the other. In a subsequent phase, all rats received alternating daily confinements to the previously cocaine paired and unpaired sides of the apparatus. During this phase, half the rats had access to a novel object on their initially unpaired side; the remaining rats did not receive objects. The ability of novelty to compete with cocaine in a drug free and cocaine challenge test was sensitive to cocaine dose. In Experiment 2, a place preference was established with 10 mg/kg cocaine and testing occurred after 1, 14, or 28 day retention intervals. Findings indicate that choice behaviors mediated by cocaine conditioning are reduced with the passing of time. Taken together, competition between cocaine and novelty conditioned rewards are sensitive to drug dose and retention interval.

Pilot study to determine the ability of health-care professionals to undertake drug dose calculations

Background: It is essential for health-care professionals to calculate drug doses accurately. Previous studies have demonstrated that many hospital doctors were unable to accurately convert dilutions (e.g. 1:1000) or percentages (e.g. percentage w/v) of drug concentrations into mass concentrations (e.g. mg/mL). Aims: The aims of the present study were to evaluate the ability of health-care professionals to perform drug dose calculations accurately and to determine their preferred concentration convention when calculating drug doses. Methods: A selection of nurses, medical students, house surgeons, registrars and pharmacists undertook a written survey to assess their ability to perform five drug dose calculations. Participants were also asked which concentration convention they preferred when calculating drug doses. The surveys were marked then analysed for health-care professionals as a whole and then by subgroup analysis to assess the performance of each health-care-professional group. Results: Overall, less than 14% of the surveyed health-care professionals could answer all five questions correctly. Subgroup analysis revealed that health-care pro-fessionals' ability to calculate drug doses were ranked in the following order: registrars approximate to pharmacists > house surgeons > medical students >> nurses. Ninety per cent of health-care professionals preferred to calculate drug doses using the mass concentration convention. Conclusions: Overall, drug dose calculations were performed poorly. Mass concentration was clearly indicated as the preferred convention for calculating drug doses.

Nonlinear dynamics of the patient’s response to drug effect during general anesthesia

In today’s healthcare paradigm, optimal sedation during anesthesia plays an important role both in patient welfare and in the socio-economic context. For the closed-loop control of general anesthesia, two drugs have proven to have stable, rapid onset times: propofol and remifentanil. These drugs are related to their effect in the bispectral index, a measure of EEG signal. In this paper wavelet time–frequency analysis is used to extract useful information from the clinical signals, since they are time-varying and mark important changes in patient’s response to drug dose. Model based predictive control algorithms are employed to regulate the depth of sedation by manipulating these two drugs. The results of identification from real data and the simulation of the closed loop control performance suggest that the proposed approach can bring an improvement of 9% in overall robustness and may be suitable for clinical practice.

Nonlinear dynamics of the patient’s response to drug effect during general anesthesia

In today’s healthcare paradigm, optimal sedation during anesthesia plays an important role both in patient welfare and in the socio-economic context. For the closed-loop control of general anesthesia, two drugs have proven to have stable, rapid onset times: propofol and remifentanil. These drugs are related to their effect in the bispectral index, a measure of EEG signal. In this paper wavelet time–frequency analysis is used to extract useful information from the clinical signals, since they are time-varying and mark important changes in patient’s response to drug dose. Model based predictive control algorithms are employed to regulate the depth of sedation by manipulating these two drugs. The results of identification from real data and the simulation of the closed loop control performance suggest that the proposed approach can bring an improvement of 9% in overall robustness and may be suitable for clinical practice.

A control flow prototype for a dose recommending device for chronic myeloid leukemia patients

In this paper we present a prototype of a control flow for an a posteriori drug dose adaptation for Chronic Myelogenous Leukemia (CML) patients. The control flow is modeled using Timed Automata extended with Tasks (TAT) model. The feedback loop of the control flow includes the decision-making process for drug dose adaptation. This is based on the outputs of the body response model represented by the Support Vector Machine (SVM) algorithm for drug concentration prediction. The decision is further checked for conformity with the dose level rules of a medical guideline. We also have developed an automatic code synthesizer for the icycom platform as an extension of the TIMES tool.

Safe and effective variability-a criterion for dose individualization.

BACKGROUND: : A primary goal of clinical pharmacology is to understand the factors that determine the dose-effect relationship and to use this knowledge to individualize drug dose. METHODS: : A principle-based criterion is proposed for deciding among alternative individualization methods. RESULTS: : Safe and effective variability defines the maximum acceptable population variability in drug concentration around the population average. CONCLUSIONS: : A decision on whether patient covariates alone are sufficient, or whether therapeutic drug monitoring in combination with target concentration intervention is needed, can be made by comparing the remaining population variability after a particular dosing method with the safe and effective variability.

Therapeutic drug monitoring in cancer - Are we missing a trick?

Therapeutic drug monitoring (TDM) can be defined as the measurement of drug in biological samples to individualise treatment by adapting drug dose to improve efficacy and/or reduce toxicity. The cytotoxic drugs are characterised by steep dose-response relationships and narrow therapeutic windows. Inter-individual pharmacokinetic (PK) variability is often substantial. There are, however, a multitude of reasons why TDM has never been fully implemented in daily oncology practice. These include difficulties in establishing appropriate concentration target, common use of combination chemotherapies and the paucity of published data from pharmacological trials. The situation is different with targeted therapies. The large interindividual PK variability is influenced by the pharmacogenetic background of the patient (e.g. cytochrome P450 and ABC transporters polymorphisms), patient characteristics such as adherence to treatment and environmental factors (drug-drug interactions). Retrospective studies have shown that targeted drug exposure correlates with treatment response in various cancers. Evidence for imatinib currently exists, others are emerging for compounds including nilotinib, dasatinib, erlotinib, sunitinib, sorafenib and mammalian target of rapamycin (mTOR) inhibitors. Applications for TDM during oral targeted therapies may best be reserved for particular situations including lack of therapeutic response, severe or unexpected toxicities, anticipated drug-drug interactions and concerns over adherence treatment. There are still few data with monoclonal antibodies (mAbs) in favour of TDM approaches, even if data showed encouraging results with rituximab and cetuximab. TDM of mAbs is not yet supported by scientific evidence. Considerable effort should be made for targeted therapies to better define concentration-effect relationships and to perform comparative randomised trials of classic dosing versus pharmacokinetically-guided adaptive dosing.

Induction pegylated interferon Alfa-2a and high dose ribavirin do not increase SVR in heavy patients with HCV genotype 1 and high viral loads

Background & Aims Patients infected with hepatitis C virus (HCV) genotype 1, body weight <85 kg, and high baseline viral load respond poorly to standard doses of pegylated interferon (peginterferon) and ribavirin. We evaluated intensified therapy with peginterferon alfa-2a plus ribavirin. Methods This double-blind randomized trial included HCV genotype 1-infected outpatients from hepatology clinics with body weight <85 kg and HCV RNA titer <400,000 IU/mL. Patients were randomized to 180 μg/wk peginterferon alfa-2a for 48 weeks plus 1200 mg/day ribavirin (standard of care) (group A, n = 191) or 1400/1600 mg/day ribavirin (group B, n = 189). Additional groups included 360 μg/wk peginterferon alfa-2a for 12 weeks then 180 μg/wk peginterferon alfa-2a for 36 weeks plus 1200 mg/day ribavirin (group C, n = 382) or 1400/1600 mg/day ribavirin (group D, n = 383). Follow-up lasted 24 weeks after treatment. Results Sustained virologic response rates (HCV RNA level <15 IU/mL at end of follow-up) in groups A, B, C, and D were 38%, 43%, 44%, and 41%, respectively. There were no significant differences among the 4 groups or between pooled peginterferon alfa-2a regimens (A + B vs C + D: odds ratio [OR], 1.08; 95% confidence interval [CI], 0.831.39; P = .584) or pooled ribavirin regimens (A + C vs B + D: OR, 1.00; 95% CI, 0.791.28; P = .974). Conclusions In patients infected with HCV genotype 1 who are difficult to treat (high viral load, body weight <85 kg), a 12-week induction regimen of peginterferon alfa-2a and/or higher-dose ribavirin is not more effective than the standard regimen. © 2010 AGA Institute.

Advances in sickle cell disease treatment: From drug discovery until the patient monitoring

Sickle Cell Disease (SCD) is one of the most prevalent hematological diseases in the world. Despite the immense progress in molecular knowledge about SCD in last years few therapeutical sources are currently available. Nowadays the treatment is performed mainly with drugs such as hydroxyurea or other fetal hemoglobin inducers and chelating agents. This review summarizes current knowledge about the treatment and the advancements in drug design in order to discover more effective and safe drugs. Patient monitoring methods in SCD are also discussed. © 2011 Bentham Science Publishers Ltd.