OPT: a package of computer programs for parameter optimisation in clinical pharmacokinetics

Do Vancomycin Pharmacokinetics Differ Between Obese and Non-obese Patients? Comparison of a General-Purpose and Four Obesity-Specific Pharmacokinetic Models

Over the past decade, numerous obesity-specific pharmacokinetic (PK) models and dosage regimens have been developed. However, it is unclear whether vancomycin PKs differ between obese and other patients after accounting for weight, age, and kidney function. In this study, the authors investigated whether using obesity-specific population PK models for vancomycin offers any advantage in accuracy and precision over using a recently developed general-purpose model.

Population pharmacokinetic evaluation and optimization of amikacin dosage regimens for the management of mycobacterial infections

Background

There is limited information on amikacin pharmacokinetics (PK) and dose requirements in patients with mycobacterial infections.

Objectives

To conduct a population PK analysis of amikacin data from patients with mycobacterial infections and compare predicted concentrations from standard and modified dosage guidelines with recommended target ranges.

Methods

A population PK model was developed using NONMEM. Cmax, Cmin, concentration 1 h post-infusion (C1h) and AUC0–24 using 15 mg/kg daily (once daily), the WHO table, 25 mg/kg three times weekly (TTW) and modified guidelines were compared using Monte Carlo simulations of 1000 patients.

Results

Data were available from 124 patients (684 concentrations) aged 16–92 years. CL was 4.64 L/h per 100 mL/min CLCR; V was 0.344 L/kg. With once-daily regimens, Cmax was 35–45 mg/L in 30%–35% of patients and 35–50 mg/L in 46%–48%; C1h was 25–40 mg/L in 53%–59%. The WHO table produced high Cmax values in patients <60 kg and low in patients >75 kg. With TTW dosing, around 30% of Cmax values were 65–80 mg/L, 40% were 60–80 mg/L, and 48% of C1h were 45–65 mg/L. Increasing the dosage interval for patients with CLCR <50 mL/min reduced Cmin values >2 mg/L from 34% to 25% for once-daily dosing and from 18% to 13% for TTW. In patients whose Cmin was <2 mg/L, 82% of AUC0–24 values were 100–300 mg.h/L.

Conclusions

Standard amikacin dosing guidelines achieve low percentages of target concentrations for mycobacterial infections. Extending the dosing interval in renal impairment and widening target ranges would reduce the need for dose adjustment.

Pharmacokinetics and dose requirements of disopyramide in neonates and children

The pharmacokinetics and dose requirements of disopyramide were evaluated in 10 infants and children during the course of routine therapeutic drug monitoring. Drug clearance was estimated using a Bayesian parameter estimation program and found to range from 0.09-0.31L/h/kg. Concentrations associated with a satisfactory response were achieved with doses of 15-20mg/kg/day in most patients, but some required higher doses of 25-30mg/kg/day.

Pharmacokinetic optimization of immunosuppressive therapy in thoracic transplantation: part I

Although immunosuppressive treatments and therapeutic drug monitoring (TDM) have significantly contributed to the increased success of thoracic transplantation, there is currently no consensus on the best immunosuppressive strategies. Maintenance therapy typically consists of a triple-drug regimen including corticosteroids, a calcineurin inhibitor (ciclosporin or tacrolimus) and either a purine synthesis antagonist (mycophenolate mofetil or azathioprine) or a mammalian target of rapamycin inhibitor (sirolimus or everolimus). The incidence of acute and chronic rejection and of mortality after thoracic transplantation is still high compared with other types of solid organ transplantation. The high allogenicity and immunogenicity of the lungs justify the use of higher doses of immunosuppressants, putting lung transplant recipients at a higher risk of drug-induced toxicities. All immunosuppressants are characterized by large intra- and interindividual variability of their pharmacokinetics and by a narrow therapeutic index. It is essential to know their pharmacokinetic properties and to use them for treatment individualization through TDM in order to improve the treatment outcome. Unlike the kidneys and the liver, the heart and the lungs are not directly involved in drug metabolism and elimination, which may be the cause of pharmacokinetic differences between patients from all of these transplant groups. TDM is mandatory for most immunosuppressants and has become an integral part of immunosuppressive drug therapy. It is usually based on trough concentration (C(0)) monitoring, but other TDM tools include the area under the concentration-time curve (AUC) over the (12-hour) dosage interval or the AUC over the first 4 hours post-dose, as well as other single concentration-time points such as the concentration at 2 hours. Given the peculiarities of thoracic transplantation, a review of the pharmacokinetics and TDM of the main immunosuppressants used in thoracic transplantation is presented in this article. Even more so than in other solid organ transplant populations, their pharmacokinetics are characterized by wide intra- and interindividual variability in thoracic transplant recipients. The pharmacokinetics of ciclosporin in heart and lung transplant recipients have been explored in a number of studies, but less is known about the pharmacokinetics of mycophenolate mofetil and tacrolimus in these populations, and there are hardly any studies on the pharmacokinetics of sirolimus and everolimus. Given the increased use of these molecules in thoracic transplant recipients, their pharmacokinetics deserve to be explored in depth. There are very few data, some of which are conflicting, on the practices and outcomes of TDM of immunosuppressants after thoracic transplantation. The development of sophisticated TDM tools dedicated to thoracic transplantation are awaited in order to accurately evaluate the patients' exposure to drugs in general and, in particular, to immunosuppressants. Finally, large cohort TDM studies need to be conducted in thoracic transplant patients in order to identify the most predictive exposure indices and their target values, and to validate the clinical usefulness of improved TDM in these conditions. In part I of the article, we review the pharmacokinetics and TDM of calcineurin inhibitors. In part II, we will review the pharmacokinetics and TDM of mycophenolate and mammalian target of rapamycin inhibitors, and provide an overall discussion along with perspectives.

Effect of various estimates of renal function on prediction of vancomycin concentration by the population mean and Bayesian methods

OBJECTIVE

Renal function was estimated in 129 elderly patients with methicillin-resistant Staphylococcus aureus (MRSA) who were treated with vancomycin (VCM). The estimation was performed by substituting serum creatinine (SCR) measured enzymatically and a value converted using the Jaffe method into the Cockcroft-Gault and Modification of Diet in Renal Disease (MDRD) equations. The serum trough level was predicted from three estimates of renal function by the population mean (PM) and Bayesian methods and the predictability was assessed.

METHODS

Two-compartment model-based Japanese population parameters for VCM were used, and the mean prediction error (ME) and root mean squared error (RMSE) were calculated as indices of bias and accuracy, respectively, for predictions by the PM and Bayesian methods.

RESULTS

The PM method gave the highest correlation with the measured value using the estimate of renal function obtained by substituting the Jaffe-converted SCR into the Cockcroft-Gault equation. There was no positive or negative bias in the ME and the value was significantly smaller than for other predicted data (P < 0.05). RMSE was also the smallest, indicating that this method increases the predictability of the serum VCM trough level. While, ME showed a negative bias for all values predicted by the Bayesian method, both the ME and RMSE were very small.

CONCLUSION

In the application of the PM method for VCM treatment of elderly patients with MRSA, substitution of SCR based on the Jaffe method into the Cockcroft-Gault equation increases the predictability of the serum VCM trough level. The Bayesian method predicted the serum VCM trough level with high accuracy using any of the estimates of renal function.

Therapeutic drug monitoring of gentamicin: evaluation of five nomograms for initial dosing at Al-Amiri Hospital in Kuwait

OBJECTIVE

To compare five published nomograms (Thomson guidelines, Mawer nomogram, rule of eights, Hull-Sarubbi table and Dettli method) for calculating the initial gentamicin dosage regimen in a Kuwaiti population.

MATERIALS AND METHODS

Based on measured peak and trough gentamicin concentrations, the elimination rate constant and volume of distribution of gentamicin were calculated for each patient (n = 56), using a modified two-point Sawchuk-Zaske method. The calculated individual set of pharmacokinetic parameters and the initial dose regimen recommended by each of the five methods were used to predict the steady-state peak and trough of gentamicin concentrations.

RESULTS

The Thomson guidelines produced consistent results in predicting gentamicin concentrations within the target ranges of peak plus trough, peak only and trough only (63, 75 and 75%, respectively). The Mawer nomogram, Hull-Sarubbi table and Dettli methods achieved similar percentages of patients (46-50%) within the target ranges (5-10 mg x l(-1) for peak and 0.5-2 for trough), whereas empirical dosing and the rule of eights showed the lowest percentages of patients within the peak plus trough target range (25 and 37%, respectively). However, with respect to the underdosing target range (predicted concentration <5 mg x l(-1)), the Thomson guidelines showed that 21% of patients were underdosed.

CONCLUSION

The results show that a large number of patients (37-63%) were outside the target ranges in all initial gentamicin dosing methods evaluated in this study. Therefore, serum concentration measurement can be advised to assist in the optimization of gentamicin dose selection.

Population pharmacokinetic modelling of gentamicin and vancomycin in patients with unstable renal function following cardiothoracic surgery

AIMS

To describe the population pharmacokinetics of gentamicin and vancomycin in cardiothoracic surgery patients with unstable renal function.

METHODS

Data collected during routine care were analyzed using NONMEM. Linear relationships between creatinine clearance (CL(Cr)) and drug clearance (CL) were identified, and two approaches to modelling changing CL(Cr) were examined. The first included baseline (BCOV) and difference from baseline (DCOV) effects and the second allowed the influence of CL(Cr) to vary between individuals. Final model predictive performance was evaluated using independent data. The data sets were then combined and parameters re-estimated.

RESULTS

Model building was performed using data from 96 (gentamicin) and 102 (vancomycin) patients, aged 17-87 years. CL(Cr) ranged from 9 to 172 ml min(-1) and changes varied from -76 to 58 ml min(-1) (gentamicin) and -86 to 93 ml min(-1) (vancomycin). Inclusion of BCOV and DCOV improved the fit of the gentamicin data but had little effect on that for vancomycin. Inclusion of interindividual variability (IIV) in the influence of CL(cr) resulted in a poorly characterized model for gentamicin and had no effect on vancomycin modelling. No bias was seen in population compared with individual CL estimates in independent data from 39 (gentamicin) and 37 (vancomycin) patients. Mean (95% CI) differences were 4% (-3, 11%) and 2% (-2, 6%), respectively. Final estimates were: CL(Gent) (l h(-1)) = 2.81 x (1 + 0.015 x (BCOV(CLCr)-BCOV(CLCr Median)) + 0.0174 x DCOV(CLCr)); CL(Vanc) (l h(-1)) = 2.97 x (1 + 0.0205 x (CL(Cr)-CL(Cr Median))). IIV in CL was 27% for both drugs.

CONCLUSIONS

A parameter describing individual changes in CL(cr) with time improves population pharmacokinetic modelling of gentamicin but not vancomycin in clinically unstable patients.

Evaluation of Bayesian predictability of vancomycin concentration using population pharmacokinetic parameters in pediatric patients

The objective of this study was to evaluate the Bayesian predictability of vancomycin (VCM) pharmacokinetics in Japanese pediatric patients using one-compartment population pharmacokinetic (PPK) parameters, which we reported previously. The validity of the PPK model was evaluated by bootstrap method and cross validation method, and the Bayesian predictive performance was examined. The predictive performance of the PPK model for premature patients was also examined. The cross validation method showed the predictability to be acceptable for practical use, especially for predicting trough concentration using other trough data. However, for the external premature patient data, this PPK model did not seem to be adequate. A theoretical approach using a simulation technique was also examined to evaluate the predictive performance. The results suggested that the predictability at the peak was not necessarily good at all sampling times and the predictability at the trough was better when a later time point was used. The optimal sampling time for prediction of VCM concentration in pediatric patients is discussed.

Evaluation of predictability for vancomycin dosage regimens by the Bayesian method with Japanese population pharmacokinetic parameters

The predictability of serum vancomycin (VCM) concentrations by means of the Bayesian method was evaluated to establish whether the method can be used to select safe and effective VCM treatment regimens for individual patients. Serum VCM concentrations at the trough and 2 h after the end of infusion (peak) were measured. Pharmacokinetic parameters were calculated for VCM dosage regimens based on a two-compartment model with the Bayesian method, using the Japanese population pharmacokinetic parameters estimated by Yasuhara et al. (1998). The predictive performance for serum VCM concentrations and the dosage regimens were analyzed using two points of serum VCM concentration in 41 patients whose serum creatinine and age were in the ranges of 0.4-4.6 mg/dl and 24-92 years, respectively. Although the predicted values for trough and peak VCM concentrations were slightly lower than measured VCM concentrations, the predictive performance was generally good. There were no differences among the groups classified by serum creatinine or age. An examination of predicted data that differed markedly from the measured serum VCM concentrations indicated that a larger difference in volume of distribution at the steady state (Vdss) calculated from serum VCM concentrations at the beginning and revision of dosage regimens resulted in a poorer correlation of predicted values and measured values. This finding indicates that therapeutic drug monitoring should be conducted frequently, and the dosage regimen revised accordingly, in the case of patients who may have a change of Vdss of VCM, for example, due to a complication such as heart failure or edema.

Evaluation of Bayesian predictability of vancomycin concentration in patients with various degrees of renal function

To assess the usefulness of the population pharmacokinetic parameters of vancomycin (VCM) based on a two-compartment model in Japanese adult patients, predictability by a Bayesian method was evaluated using a concentration time course after single dosing to 22 patients with various degrees of renal function. Using one or two points from the observed data for each patient, the concentrations predicted by a Bayesian method were compared with the observed data for each sampling time. The patients were separated into five groups based on their renal functions indicated by creatinine clearance, and the mean prediction error (MPE) and root mean squared error (RMSE) were calculated for each group as measures of accuracy and precision, respectively. In both one- and two-point methods, the absolute MPE values at each sampling time in the elimination phase were less than 2.5 microg/ml, and the RMSE values were also small. No clear differences were found in MPE and RMSE among the groups. In the distribution phase, the MPE and RMSE were somewhat greater, and RMSE in some groups was around 15 microg/ml when trough data was used to predict the peak concentration. Also, the theoretical RMSE using this population parameter setting could well explain the observed RMSE. These results confirmed this population parameter setting is useful for at least predicting concentration in the elimination phase after single dosing, and the predictability was independent of renal function.

Optimization of phenytoin therapy in adults with epilepsy in the Western Cape, South Africa

OBJECTIVE

To assess the extent to which adults with epilepsy were optimized and individualized on phenytoin monotherapy in the Western Cape, South Africa and to estimate the average optimized dose and serum phenytoin concentration, and the therapeutic range for this patient group.

METHODS

Patients were considered to be optimized on phenytoin if they were seizure-free or the best compromise was achieved between seizure reduction and side-effects.

RESULTS

538 (233 black and 305 coloured) adult people with epilepsy were treated at nine epilepsy clinics as outpatients. Of these patients, 332 (226 male and 106 female, 149 black and 183 coloured) were included in the data analysis as they were considered to have reliable phenytoin levels. Phenytoin doses and steady-state serum concentrations were predicted using the Michaelis-Menten equation. Patients attended a clinical pharmacokinetic service for 7.7+/-5.3 (range 1-22) months. The average optimized dose was 305.8 (range 100-500) mg/day and the average optimized level was 62.7+/-23.9 (range 15-133) micromol/l. Most patients (61.9%) were optimized in the therapeutic range 40-79 micromol/l; 21.1% were optimized above and 17% below this range. In 1.6% of patients serum concentrations above 120 micromol/l were required. Dosage adjustments were made in 47.0% of patients, increased in 31.9% and reduced in 15.1%.

CONCLUSION

These findings indicate that many patients (47%) attending outpatient clinics were not optimized on phenytoin therapy.

Population pharmacokinetics of aminoglycoside antibiotics in patients with cystic fibrosis

The population pharmacokinetics of gentamicin and tobramycin were investigated in a group of 51 young adults with cystic fibrosis. Their ages ranged from 14-35 years, weights from 38-82 kg, and 27 of the patients were female. None of the patients had renal impairment, but 3 patients were treated in the intensive therapy unit (ITU) during one of their courses of therapy. Data comprised 219 courses of therapy and 544 concentrations (mean: 11 per patient). Concentration-time data were analyzed using a nonlinear mixed-effects model package (NONMEM) and were best described by a one-compartment model. Factors identified as potentially influencing aminoglycoside pharmacokinetics were added in a stepwise fashion and the best model found that drug clearance and volume of distribution were related to body surface area and admission to ITU. The mean population estimates were a clearance of 2.89 L/hr/m2 and a volume of distribution of 9.21 L/m2 with a 60% increase in patients who were admitted to ITU. Interpatient variability in clearance and volume were 14% and 8%, respectively. The results suggested that a dose of 120 mg/m2 should achieve an average 1 hour postdose peak of 10 mg/L and trough of <1 mg/L and that higher doses might be required in ITU patients.

Effect of the number of samples on Bayesian and non-linear least-squares individualization: a study of cyclosporin treatment of haematological patients with multidrug resistance

We have studied whether the prediction of drug concentrations improves as the number of samples used for individualization is increased, and whether the Bayesian method of individualization is superior to the non-linear least-squares method. Data were obtained from ten adult haematological patients with multidrug resistance who were treated with cyclosporin. The predictions of blood-cyclosporin concentrations were made using the Abbott PKS program. The number of samples used for individualization was increased from 1 to 30 for the Bayesian method and from 4 to 30 for the non-linear least-squares method. Linear regression, percentage prediction error, and absolute and relative predictive performance were used to evaluate the predictions. The results show that the Bayesian method affords greater precision than the non-linear least-squares method, but that the non-linear least-squares method is more accurate and results in less bias. Whereas for linear regression predictions improve as the number of samples is increased, other evaluations show improvement in the range from 5 to 11 samples; linear regression, percentage prediction errors and prediction bias support the opinion that the Bayesian method progressively becomes the non-linear least-squares method as the number of samples used for individualization is increased, but the accuracy and precision of prediction do not support this opinion. The study supports the statement that Bayes' law requires parameters from an infinite population, otherwise the advantage of the Bayesian method might be marginal.

Bayesian forecasting in paediatric populations

Bayesian forecasting offers several important advantages for dosage individualisation in children, although, unlike for adults, its use in this population is much lower. Indeed, currently Bayesian methods are underused in this patient population. The paucity of paediatric population pharmacokinetic parameters, and the unavailability of specific clinical pharmacokinetic software for the whole paediatric population, are the main limitations to the application of Bayesian methods in these patients. When these problems have been overcome, this approach will allow clinicians to achieve therapeutic concentrations more readily, faster and more precisely, thus making the methodology highly attractive in the paediatric setting.

Application of population pharmacokinetics to the optimization of theophylline therapy

The aim of this study was to determine theophylline clearance (Cl) values in adult patients using serum concentrations gathered from routine clinical care. This information was used to estimate an a priori dosing regimen that would permit steady-state concentrations of 5-15 mg/litre, now recommended for the treatment of chronic asthma, and to evaluate the need to establish monitoring strategies when theophylline is given at these lower doses and when it can be expected that almost no adverse effects are likely. Retrospective data from 204 asthmatic and COPD patients, with a total of 517 serum concentrations, were studied. Population pharmacokinetic analysis was performed with the MULTI(ELS) computer program according to a one-compartment model. The influence of the following factors on theophylline Cl were investigated: body weight (TBW, IBW and LBM) and age as continuous variables, and gender, smoking habit and the presence of congestive heart failure (CHF) as indicator variables. To validate the results of the population pharmacokinetic analysis, a second independent group of 63 patients was studied prospectively. Hypothesis testing to evaluate potentially significant factors produced a final model in which Cl was based on IBW (kg) and age (years), and was reduced by 25% in patients with moderate CHF and increased by 28% in patients who smoked (Cl (litres/h) = (0.037 IBW-0.006 age) x 1.284 smoke x 0.751 CHF). The variability in Cl, expressed as the coefficient of variation, was 36%. In adult non-smoker and non-CHF patients, application of a maintenance dosing regimen calculated from IBW and age using the final model for Cl would theoretically afford only 1.5% of patients with potentially toxic concentrations. Thus, measurement of serum theophylline concentrations (STC) would only be required when other conditions known to alter theophylline metabolism exist, such as smoking or disease factors. These appreciations could have important clinical implications at a time when the potential immunomodulatory activity of theophylline is being emphasized and health resources should be allocated properly.

Bayesian individualization via sampling-based methods

We consider the situation where we wish to adjust the dosage regimen of a patient based on (in general) sparse concentration measurements taken on-line. A Bayesian decision theory approach is taken which requires the specification of an appropriate prior distribution and loss function. A simple method for obtaining samples from the posterior distribution of the pharmacokinetic parameters of the patient is described. In general, these samples are used to obtain a Monte Carlo estimate of the expected loss which is then minimized with respect to the dosage regimen. Some special cases which yield analytic solutions are described. When the prior distribution is based on a population analysis then a method of accounting for the uncertainty in the population parameters is described. Two simulation studies showing how the methods work in practice are presented.

Pharmacokinetics of phenytoin in routine clinic patients in Malaysia

We estimated individual and population Michaelis-Menten pharmacokinetic parameters for phenytoin (DPH) in epileptic patients attending our neurology clinic using the computer programme. OPT. Our results agreed well with literature values but were lower than those we obtained earlier in a smaller number of patients. The Km was independent of age, weight and sex but there was a weak, correlation between Vm and body weight. We conclude that the use of population Vm and Km in normograms could lead to errors in DPH dose estimations as they correlated very poorly with patient characteristics. OPT was easy to use and sufficiently accurate for deriving dose estimates in routine patients. Its use would enable practitioners to generate their patients' own parameters for use in individual dosage adjustments. The estimates can subsequently be updated as more data become available.

Value of theophylline treatment in patients handicapped by chronic obstructive lung disease

BACKGROUND

It is still not certain whether it is worth using theophylline in addition to inhaled bronchodilators and corticosteroids to treat obstructive airways disease. This trial was designed to test whether the addition of prescribed theophylline in doses sufficient for sustained optimal steady state plasma concentrations would produce any detectable additional advantage in spirometric or functional variables in these handicapped patients.

METHODS

A randomised, double blind, placebo controlled, crossover study of added theophylline treatment was aimed at steady state plasma concentrations of 10 and 17 mg/l, the dose being calculated individually by Bayesian parameter estimation and maintained for six weeks along with the patient's previously prescribed bronchodilators and steroids. Of 20 patients sequentially recruited, 15 provided data that could be analysed. All had chronic obstructive lung disease with a mean forced expiratory volume in the first second (FEV1) up to about 30% of the predicted value and gave no history of being treated with theophylline. The protocol included spirometry, whole body plethysmography, and treadmill exercise. Measurements also included steady state plasma theophylline concentrations and trapped gas volume. Quality of life was assessed by an established questionnaire method covering breathlessness in everyday activities, fatigue, emotional function, and control over the disease.

RESULTS

Both target plasma concentrations were achieved. Improvements in peak flow (PEF; mean 20%), trapped gas volumes (38%), two stage vital capacity (15%), distances walked (48%), breathlessness in everyday activities (32%), and fatigue (18%) were found at the higher plasma concentration only. FEV1, forced vital capacity (FVC), emotional function, and control did not change.

CONCLUSION

Theophylline treatment with sustained steady state concentrations about 17 mg/l provides worthwhile objective and subjective further benefits for patients handicapped by chronic obstructive lung disease when it is added to bronchodilators and corticosteroids.

Estimation of population pharmacokinetics for carbamazepine in Malaysian patients using the OPT computer program

We used OPT to estimate individual and population pharmacokinetics for carbamazepine (CBZ) in Malaysian epileptic patients attending our Neurology Clinic. We noted that plasma CBZ concentrations and clearances correlated poorly with daily doses and body weights respectively but we found the values for clearance, volumes of distribution, elimination rate constants and half lives to be in good agreement with earlier reports. We conclude that OPT is a simple yet useful program to derive individual and population pharmacokinetic parameters for CBZ for use in dosage adjustments. We also conclude that although the Malaysian population do not differ substantially in handling CBZ, available data for the pharmacokinetic parameters must be used cautiously in applying it to the therapeutic drug monitoring for CBZ in our patients.

Pharmacokinetic optimisation of anticonvulsant therapy

Changing attitudes towards the use of antiepileptic drugs have led to an emphasis on monotherapy with serum concentration measurement coupled with standard, weight-adjusted starting and maintenance regimens to guide initial therapy and subsequent dosage titration. Currently, the established anticonvulsants are carbamazepine, valproic acid (sodium valproate) and phenytoin. Phenobarbital is now less commonly prescribed due to its propensity to produce sedation and impair cognitive function. The value of pharmacokinetic optimisation with valproic acid is limited by its wide therapeutic index, large fluctuations in the concentration-time profile and concentration-dependent protein binding. Thus, although serum concentrations are often measured, they are rarely subjected to pharmacokinetic interpretation. Carbamazepine has a flatter concentration-time profile than valproic acid. Its target range is more clearly defined and it undergoes autoinduction of metabolism and interacts with other drugs. Pharmacokinetic principles can, therefore, be more readily applied to carbamazepine, although, in general, a simple clinical approach to its use is usually satisfactory. Phenytoin has required the greatest pharmacokinetic input due to its nonlinear pharmacokinetics and narrow target range. Many different graphical methods, equations and computer programs have been reported, some of which demand 2 steady-state, dose-concentration pairs; others function satisfactorily with only 1. Recent attempts have been made to interpret non-steady-state data. In addition, a number of workers have demonstrated the value of altering the population parameter estimates to account for ethnic differences. A pharmacokinetic approach can also be used to tailor the use of phenytoin in the treatment of status epilepticus. Dosage alterations may be needed for specific patient groups. In particular, children generally require higher dosages on a weight-for-weight basis than adults, while equivalently lower dosages should be given to neonates. Most anticonvulsants are principally cleared by hepatic mechanisms, so dosage adjustment is not usually required in renal disease, although care must be taken in interpreting serum concentrations because of changes in protein binding. Close monitoring is required in the elderly and patients with hepatic impairment, while increased dosages may be needed in critically ill patients and during pregnancy. Pharmacokinetic principles can be used in the treatment of treat self-poisoning with anticonvulsants. There are few data available on the pharmacokinetics of vigabatrin, lamotrigine, oxcarbazepine and gabapentin in patients. Due to its mode of action in binding irreversibly to its target enzyme, serum concentration monitoring of vigabatrin plays no role in optimising therapy. The value of applying pharmacokinetic principles with the other 3 drugs remains to be investigated.(ABSTRACT TRUNCATED AT 400 WORDS)

Practical computer-assisted dosing for aminoglycoside antibiotics

In principle, computer-assisted individualization of antibiotic dosing offers the prospect of better patient outcomes through improved dosing precision. In practice, however, the expertise in pharmacokinetics required to operate these programs has precluded their use by most physicians and pharmacists. We developed a computer program for individualization of dosing of aminoglycoside antibiotics under conditions in which access to experts in pharmacokinetics is impractical. The program is accurate, yet it requires less effort for data collection than previous drug dosing programs did. The program generates advice on a broad spectrum of topics, including dose adjustment, interpretation of measured drug concentrations in blood, and recommendations for monitoring drug concentrations. We tested its performance by prospectively comparing it with a clinical pharmacokinetic consultation service in a series of 78 consecutive patients. There were no differences in accuracy or bias in the prediction of drug concentrations. The rate of agreement between the program's dosing recommendations and those of the consultation service was 67 percent. This rate of agreement is typical of interexpert variation. In a stratified set of 24 of the 41 instances with significant disagreement regarding the recommended dose, experts ranked the program's recommendations as highly as those of the consultation service (95% confidence interval for difference in rank, -0.30 less than chi less than 0.47). The results suggest that expert systems can be coupled with pharmacokinetic dosing programs to deliver high-quality clinical recommendations for administration of antimicrobial agents.

Bayesian parameter estimation and population pharmacokinetics

The widespread application of Bayesian parameter estimation in the area of therapeutic drug monitoring (TDM) has prompted the need for well conducted population studies to obtain relevant prior pharmacokinetic parameter estimates. In many cases the population has consisted of a relatively small number of subjects. This may be unavoidable for drugs used in cancer chemotherapy or in small, specific populations of patients. In contrast, information about drugs which are used extensively, such as the aminoglycosides, can be obtained by population studies which involve a large number of individuals. Indeed, this technique has proved particularly useful for determining parameter estimates which can be employed in neonatal TDM. Bayesian parameter estimation has been most frequently used for drugs with narrow therapeutic ranges such as the aminoglycosides, cyclosporin, digoxin, anticonvulsants (especially phenytoin), lithium and theophylline. However, the technique has now been extended to cytotoxic drugs, Factor VIII and warfarin. Bayesian methods have also been used to limit the number of samples required in more conventional pharmacokinetic studies with new drugs. Further advances in the use of these methods are likely to include measures of drug response and toxicity requiring population studies which also include relevant pharmacodynamic information.

An updated comparison of drug dosing methods. Part III: Aminoglycoside antibiotics

Aminoglycoside antibiotics continue to be useful for the treatment of Gram-negative infections. Available dosing methods include predictive algorithms and nomograms, pharmacokinetics-based dosing methods, and methods that incorporate Bayesian forecasting. The individualised Sawchuk-Zaske and Bayesian methods have been extensively evaluated since the previous review in the Journal. Both methods continue to be rapid and accurate means of individualising dosage requirements for patients with diverse pharmacokinetic profiles. The predictive performance of the Bayesian method can be further enhanced when population-based parameters reflect the patient population being monitored. There are now several cost-effectiveness studies that demonstrate that pharmacokinetic dosing services for aminoglycosides result in cost savings, better therapeutic concentrations, fewer toxic serum concentrations, and shorter mean durations of hospital stay and aminoglycoside therapy. Further studies are needed for cost-effectiveness and comparison of various dosing methods in paediatric and neonatal patients.

Gentamicin and vancomycin removal by continuous venovenous hemofiltration

The dispositions and dose requirements for vancomycin and gentamicin were investigated in a 58-year-old man who was receiving long-term continuous venovenous hemofiltration. Estimates of clearance were obtained using a Bayesian parameter estimation program and stayed remarkably consistent throughout the therapy. Single daily doses of both vancomycin and gentamicin generally maintained the profiles for both drugs around the target ranges of peak 5-10 mg/L (gentamicin) and 25-40 mg/L (vancomycin) and trough less than 2 mg/L (gentamicin) and less than 10 mg/L (vancomycin).

Assessment of theophylline compliance in general practice using a 'compliance index' based on apparent drug clearance

The monitoring of compliance in outpatients using serum drug levels suffers from the difficulty of relating concentrations to changes in dosage. Clearance is independent of dose for most drugs, and may be predicted using a Bayesian based forecasting program. This paper describes how the apparent clearance obtained following a single drug analysis can be compared with the Bayesian population value to derive a Compliance Index. In compliant individuals with stable hepatic and renal function this index should remain constant. We evaluated the usefulness of the Compliance Index in general practice with patients prescribed theophylline. Of 54 patients examined on up to five occasions, only 12% were deemed to be fully compliant at all times. Compliance improved during the study, but worsened again once regular consultation ceased. The results indicate that such a Compliance Index is a viable method of assessing compliance with the aid of concentration measurements.

Screening for aminoglycoside auditory toxicity in the old

1. We have investigated 22 patients receiving gentamicin, mean (s.d.) age 78 (6) years for auditory toxicity, using a standard audiometric technique in a sound-treated room (Study 1). 2. Use of a portable audiometer might allow a larger and more representative proportion of patients treated with aminoglycosides to be screened for ototoxicity. A method for detecting high frequency loss suitable for use in the ward was evaluated in 12 volunteers aged 27 (4) years (Study 2). 3. The error inherent in taking hearing at the start of treatment as a reference point was measured in 16 patients, aged 81 (8) years, prescribed non-ototoxic antibacterials (Study 3). 4. A significant (P = 0.05) reduction in hearing threshold was detected in Study 1, although psychometric tests revealed unchanged or improved ability to co-operate. This occurred only at 4000 Hz, the highest frequency used. The magnitude of this loss, mean 2.5 dB, was similar to that of the improvement in threshold detected (P = 0.0004) early in the course of treatment in Study 3. Thus, underestimation of ototoxicity is likely. 5. If a change of threshold of 10 dB or more is taken arbitrarily to represent a real change in hearing, then there was a significant excess of patients in Study 1 with losses at 4000 Hz only (P = 0.032). The six with such losses at this frequency were older than the rest. However, there was a significant (P less than 0.02) positive correlation between log mean predose serum gentamicin concentration and age. Thus, it remains to be determined whether presbyacusis sensitizes those hair cells which it does not destroy to toxic damage. 6. The cumulative dose of gentamicin (for a course of the duration given) was calculated according to published prescribing aids. There was no systematic reduction in the ratio of the dose recommended by a given aid to the dose prescribed in the six with hearing losses as defined above. 7. In Study 2, thresholds obtained at 6000 Hz in the open ward were, on average, 0.9 dB higher than in the sound treated room, but the effect of venue did not reach statistical significance. In the morning thresholds were marginally, but significantly (P = 0.04), lower than in the afternoon. Precision, as measured by the standard deviation of replicate determination, was independent of test conditions. Using multiple (ten) threshold determinations appeared to improve resolution.(ABSTRACT TRUNCATED AT 400 WORDS)

Application of Bayesian estimation for the prediction of an appropriate dosage regimen of amikacin

A Bayesian approach was developed to determine an amikacin dosage regimen to achieve the desired plasma concentrations for each patient. Statistical characteristics of pharmacokinetic parameters were first evaluated in a group of patients (reference population), which when combined with three individual plasma concentrations of drug led to a Bayesian estimation of individual pharmacokinetic parameters. By using these parameters, an individual dosage regimen was then established to avoid residual and peak amikacin concentrations of up to 3 and 25 micrograms/mL, respectively. In a test group of 33 patients, adapted amikacin dosage regimens ranged from 4 to 43 mg/kg/d, with schedules requiring up to four infusions per day. Infusion time varied from 40 min to 4 h. These differences in drug administration protocol result from the wide interindividual variability of amikacin pharmacokinetic parameters. Performance of the developed methodology was evaluated by computing bias and precision of the estimated total body clearance and of the trough and peak amikacin concentrations that were reached after dosage regimen determinations.

High-dose melphalan dosage adjustment: possibility of using a test-dose

Previous pharmacokinetic studies of i.v. high-dose melphalan (HDM) have demonstrated large interindividual variations in the pharmacokinetic parameters. We therefore studied the possibility of using a test dose of the drug to determine the level of a subsequent therapeutic dose. This study was undertaken to establish whether the pharmacokinetics of melphalan were linear and reproducible within the same patient and determine whether a linear extrapolation could be carried out from the test dose. The first eight patients were studied on two occasions separated by 2 hours (repeatability stage). Although reasonable evidence for linear pharmacokinetics was obtained from these patients, the data suggested a number of factors that might have introduced errors. Therefore, the second group of ten patients were treated on a slightly different protocol on two occasions 24 h apart (linearity stage). The ratios of the two doses ranged from 1 to 8 (repeatability stage) and from 2.6 to 10 (linearity stage). During both stages there was a good correlation between the AUC measured for the second infusion and that predicted from the first (r = 0.929 and r = 0.943, respectively). We conclude that a test dose can be used to determine the subsequent dose of melphalan necessary to produce a desired AUC.

Dynamical dosage regimen calculations in linear pharmacokinetics

The objective of this analysis of a linear compartment system is to compute drug input functions that are optimal in producing nontoxic pharmacological responses of maximal therapeutic efficacy. Pharmacokinetics should underlie the rational use of drugs and when a therapeutic range is known, the achievement of safe and effective target concentrations may be assured by a dosage regimen computed for a given administration schedule. The method developed herein is based on linearity and superimposition principles applicable to the class of systems considered. This method requires estimated values of model individual parameters and computes optimum dosage regimens in an iterative scheme, corresponding to a real time dynamical context. An interactive computer program has been developed to perform dosage regimen calculations.

Nonparametric maximum likelihood estimation for population pharmacokinetics, with application to cyclosporine

A new method, nonparametric maximum likelihood (NPML), for statistical analysis of population kinetic data is proposed. NPML provides a discrete estimate of the whole probability density function of the pharmacokinetic parameters. This permits a straightforward derivation of usual population characteristics. To illustrate the application of the NPML method, a population analysis of cyclosporine RIA measured plasma levels in 188 bone marrow transplant patients after intravenous infusion, is presented. The capability of NPML to extract population information from sparse individual data is also outlined.

Evaluation of a microcomputer program (OPT) for parameter optimisation in clinical pharmacokinetics: gentamicin and tobramycin

1. Several nomograms and computer programs are available to aid in aminoglycoside dosing. 2. Due to the variability in the relationship between dosage and serum drug levels, monitoring through the acquisition of serum drug levels is mandatory. 3. All clinical data, including serum drug levels, are subject to errors. 4. The program we have evaluated, OPT, calculates the most likely set of pharmacokinetic parameter estimates for individual patients by applying Bayes' theorem and the principle of Maximum Likelihood Estimation. Through a feedback process all available data are used, taking possible errors into account. 5. Our study shows that OPT is able to predict serum aminoglycoside levels accurately in the routine clinical setting. It may thus contribute to the quality of aminoglycoside therapy.

Initiation of warfarin therapy: comparison of physician dosing with computer-assisted dosing

In a prospective, randomized study at two university hospitals, the authors examined how effectively housestaff physicians (n = 36) managed the initiation of warfarin therapy compared with a computer-assisted dosing regimen (n = 39) using the software program Warfcalc, which was managed by one of the authors. Target prothrombin time ratios were selected by the physicians. Study endpoints included: the time to reach a therapeutic prothrombin ratio, the time to reach a stable therapeutic dose, the number of patients transiently overanticoagulated, the number of bleeding complications, and the accuracy of the predicted maintenance dose, which was assessed at steady-state 10-14 days later. Computer-assisted dosing consistently out-performed the physicians: a stable therapeutic dose was achieved 3.7 days earlier (p = 0.002), fewer patients were overanticoagulated (10% versus 41%), and the predicted maintenance dose was in the therapeutic range in 85% of the computer-dosed patients versus 42% of the physician group (p less than 0.002). For physicians who did not routinely manage warfarin therapy, computer-assisted dosing improved the accuracy of dosing and shortened the time required to achieve a stable therapeutic dose.

A clinically significant interaction between ciprofloxacin and theophylline

We report a case of theophylline toxicity following the co-administration of ciprofloxacin. Total theophylline clearance fell from 2.3 l.h-1 to 0.8 l.h-1 when ciprofloxacin was added to the treatment regimen and returned to 2.1 l.h-1 after ciprofloxacin was discontinued.

The prophylactic use of phenytoin during iopamidol contrast studies of the subarachnoid space

Contrast examinations of the subarachnoid space are associated with side effects including convulsions. Attention has been given to the prophylactic use of anticonvulsants. We describe a simple oral regimen using the established anticonvulsant phenytoin that can be administered to short-stay patients and that achieves effective serum and CSF concentrations. A preliminary account of this work was presented to the British Pharmacological Society in January 1984 in London.

Population pharmacokinetics. Theory and clinical application

Good therapeutic practice should always be based on an understanding of pharmacokinetic variability. This ensures that dosage adjustments can be made to accommodate differences in pharmacokinetics due to genetic, environmental, physiological or pathological factors. The identification of the circumstances in which these factors play a significant role depends on the conduct of pharmacokinetic studies throughout all stages of drug development. Advances in pharmacokinetic data analysis in the last 10 years have opened up a more comprehensive approach to this subject: early traditional small group studies may now be complemented by later population-based studies. This change in emphasis has been largely brought about by the development of appropriate computer software (NONMEM: Nonlinear Mixed Effects Model) and its successful application to the retrospective analysis of clinical data of a number of commonly used drugs, e.g. digoxin, phenytoin, gentamicin, procainamide, mexiletine and lignocaine (lidocaine). Success has been measured in terms of the provision of information which leads to increased efficiency in dosage adjustment, usually based on a subsequent Bayesian feedback procedure. The application of NONMEM to new drugs, however, raises a number of interesting questions, e.g. 'what experimental design strategies should be employed?' and 'can kinetic parameter distributions other than those which are unimodal and normal be identified?' An answer to the later question may be provided by an alternative non-parametric maximum likelihood (NPML) approach. Population kinetic studies generate a considerable amount of demographic and concentration-time data; the effort involved may be wasted unless sufficient attention is paid to the organisation and storage of such information. This is greatly facilitated by the creation of specially designed clinical pharmacokinetic data bases, conveniently stored on microcomputers. A move towards the adoption of population pharmacokinetics as a routine procedure during drug development should now be encouraged. A number of studies have shown that it is possible to organise existing, routine data in such a way that valuable information on pharmacokinetic variability can be obtained. It should be relatively easy to organise similar studies prospectively during drug development and, where appropriate, proceed to the establishment of control systems based on Bayesian feedback.

Steady-state dosage regimen calculations in linear pharmacokinetics

This paper describes a general approach to compute steady-state dosage regimens. The method deals with individual linear compartmental mammillary pharmacokinetic models and requires estimated values of model parameters. It is based on the linearity principle and computes dosage regimens to achieve and maintain specified steady-state plasma concentrations, e.g. mean value or minimum effective and maximum safe levels. An interactive computer program performing the calculations has been developed.

Disopyramide in acute myocardial infarction: problems with changing pharmacokinetics

In the acute phase of myocardial infarction it is recognized that serum disopyramide concentrations may be lower than expected. This has generally been attributed to reduced oral bioavailability. This report describes data obtained routinely from 6 patients with acute myocardial infarction and cardiac dysrhythmias treated initially with intravenous disopyramide. Serum disopyramide concentrations were consistently lower than expected, on average by 2.6 micrograms/ml. This was interpreted as being due to relatively high drug clearance, calculated as 6.7 +/- 1.5 l/h, compared to expected values of 3-4 l/h. Dosage schedules determined on the basis of the acute phase pharmacokinetics subsequently produced higher than predicted concentrations at later times on average by 2.8 micrograms/ml. Clearance at this time was calculated to be 3.1 +/- 0.6 l/h. Thus even with intravenous disopyramide therapy there are problems with changing pharmacokinetic parameters after myocardial infarction.

Feedback control methods for drug dosage optimisation. Concepts, classification and clinical application

The concept of feedback control methods for drug dosage optimisation is described from the viewpoint of control theory. The control system consists of 5 parts: (a) patient (the controlled process); (b) response (the measured feedback); (c) model (the mathematical description of the process); (d) adaptor (to update the parameters); and (e) controller (to determine optimum dosing strategy). In addition to the conventional distinction between open-loop and closed-loop control systems, a classification is proposed for dosage optimisation techniques which distinguishes between tight-loop and loose-loop methods depending on whether physician's interaction is absent or included as part of the control step. Unlike engineering problems where the process can usually be controlled by fully automated devices, therapeutic situations often require that the physician be included in the decision-making process to determine the 'optimal' dosing strategy. Tight-loop and loose-loop methods can be further divided into adaptive and non-adaptive, depending on the presence of the adaptor. The main application areas of tight-loop feedback control methods are general anaesthesia, control of blood pressure, and insulin delivery devices. Loose-loop feedback methods have been used for oral anticoagulation and in therapeutic drug monitoring. The methodology, advantages and limitations of the different approaches are reviewed. A general feature common to all application areas could be observed: to perform well under routine clinical conditions, which are characterised by large interpatient variability and sometimes also intrapatient changes, control systems should be adaptive. Apart from application in routine drug treatment, feedback control methods represent an important research tool. They can be applied for the investigation of pathophysiological and pharmacodynamic processes. A most promising application is the evaluation of the relationship between an intermediate response (e.g. drug level), which is often used as feedback for dosage adjustment, and the final therapeutic goal.

Therapeutic drug monitoring--the need for audit?

The experience of therapeutic drug monitoring (TDM) of seven drugs as part of the normal biochemistry services over the period April 1982-83 is reported. Reagent costs alone for these assays exceeded pounds 15,000. Of 1,841 digoxin estimations, 56 per cent fell within the therapeutic range whereas 25 per cent were potentially toxic. Fifty per cent of 968 lithium measurements were lower than 4 mg L-1, placing many of these patients at risk of treatment failure. Only 29 per cent and 32 per cent of 369 theophylline and 440 phenytoin concentrations respectively were within their well-established ranges. In only around 20 per cent of requests for theophylline and phenytoin assay was sufficient clinical information supplied to the laboratory. In the majority of patients with theophylline or phenytoin concentrations outwith the therapeutic range, further analysis was not requested and so optimisation of dosage cannot be assumed to have occurred. The wide therapeutic ranges of carbamazepine and phenobarbitone ensured that most patients attained acceptable concentrations. Sodium valproate analysis was requested on 160 occasions despite the poor correlation between the concentration of this drug and its therapeutic and toxic effects. TDM can be an expensive exercise which must be subjected to rigorous cost-benefit analysis. Requests should be made on a customised drug assay form and interpretative advice on individual patient problems made available. Recommendations for the organisation, daily running and clinical supervision of TDM are made.

Performance of Bayesian feedback to forecast lidocaine serum concentration: evaluation of the prediction error and the prediction interval

The prediction performance of the Bayesian feedback method was evaluated with respect to accuracy and precision, and efficacy and safety (width of the prediction interval) on the basis of 90 predictions in 30 patients treated with lidocaine. The mean of the prediction error (PE) and the root mean squared error (RMSE) served as a measure of accuracy and precision. The variance of the standardized prediction error (SPE) was used to evaluate the estimate of the standard deviation of the prediction error. SPE was defined as PE divided by the standard deviation of the predicted concentration. The standard error of RMSE and of the variance of SPE was determined by bootstrap. The results indicate that the lidocaine serum concentration at 12 hr (C2) after starting continuous infusion can be predicted with high accuracy and precision with a single feedback measurement obtained 2-4 hr (C1) after commencement of treatment: RMSE = 20.6%. Prediction at 24 hr (C3) was less accurate: RMSE = 31.4%. Using both C1 and C2 to predict C3 improved precision (RMSE = 23.4%). The evaluation of the prediction interval revealed that the current algorithm produces an upward biased estimate, probably due to a positive bias in the estimate of the covariance matrix of the parameter estimates. It is suggested that evaluation of prediction performance should include the estimate of the prediction interval.

A calculator program for clinical application of the Bayesian method of predicting plasma drug levels

A pharmacokinetic program that allows individualization of drug dosage regimens through the Bayesian method is described. The program, which is designed for the Hewlett-Packard HP-41 CV calculator, is based upon the one-compartment open model with either instantaneous or zero-order absorption. Individualized estimation of the patient's kinetic parameters (clearance and volume of distribution) is performed by analyzing the plasma levels measured in the patient as well as considering the population data of the drug. After estimating the individual kinetic parameters by the Bayesian method, the program predicts the dosage regimen that will elicit the desired peak and trough plasma levels at steady state. For comparison purposes, the least-squares estimates for clearance and volume of distribution are calculated, and dosage prediction can also be made on the basis of the least-squares estimates. The least-squares estimates can be used to calculate population pharmacokinetic parameters according to the Standard Two-Stage method. Several examples of clinical use of the program are presented. The examples refer to patients with classic hemophilia who were treated with Factor VIII concentrates. In these patients, the Bayesian kinetic parameters of Factor VIII have been estimated through the calculator program. The Bayesian parameter estimates generated by the HP-41 have been compared with those determined by a Bayesian program (ADVISE) designed for microcomputers.

Estimation of gentamicin clearance and volume of distribution in neonates and young children

Gentamicin therapy should be guided by serum level monitoring in all age groups, dosage adjustments depending on age related changes in pharmacokinetics. Population data analysed from two centres (43 infants from Glasgow and 100 infants and children from Manchester) by the computer program NONMEM showed that volume of distribution was related to body weight by a proportionality factor that decreased from the region of 0.41-0.46 l/kg in children less than 3 months to 0.25-0.32 l/kg in older children, a value which merges with that accepted for adults (0.25 l/kg). In both young and older children, clearance was also found to be dependent on body weight. Renal function (creatinine concentrations) provided no further explanatory power. When these results were used prospectively to forecast gentamicin concentrations with a Bayesian kinetic parameter estimation program, trough concentrations were more precisely predicted than peaks when a single concentration measurement was used. In clinical practice, however, two concentration measurements are usually routinely available and these should lead to greater precision of both peak and trough predictions. These results have been incorporated into a simple nomogram which can be used to determine a dose of gentamicin which will achieve target peak concentrations in infants, assuming that troughs should not exceed 2 micrograms/ml.

Clinical pharmacokinetics: a comprehensive system for therapeutic drug monitoring and prescribing

Clinical pharmacokinetics is an expanding scientific discipline which can make an impact on treatment in coronary care, intensive care, paediatrics, general medicine and surgery, and general practice. The aim of this study was to establish a rapid system of drug assay, to report the result, to assess the influence of pathological and clinical factors on the pharmacokinetics of certain drugs, and to use a computer to determine the optimum dosage of drugs. The clinical pharmacokinetics laboratory in Stobhill is available to all clinical departments and to general practitioners in the area. Digoxin, theophylline, and phenytoin have been assessed. Initial samples of these drugs showed that only about a third were in the therapeutic range; samples obtained after the issue of the laboratory report showed an improvement. The predictive performance of the computer program improved with feedback of one or two drug concentrations. Dosages of drugs chosen on an empirical basis may not lead to optimum treatment, and by testing samples early the dosage of the drug can be adjusted. It is hoped that the results achieved will encourage other clinical, pharmaceutical, and scientific colleagues to develop laboratories along similar lines.