Search PubMed⌕ Search

Biomedical subjects

L B Sheiner

Publications and source records attributed to L B Sheiner.

At least 19 recordsLinked to original sources
PubMed1992

Bioequivalence revisited.

The FDA permits marketing of a generic formulation of a drug G for the same indications as a standard preparation S if one can show that G is bioequivalent to S. Present implementation requires convincing evidence that the population mean difference in bioavailability (drug exposure) between the two preparations lies within specified bounds. The basis for this standard does not appear to involve a comprehensive model for the dose-response relationship, or consideration of clinical issues, notably (i) whether a patient is to commence on the drug or to switch from an established regimen to a new one; or (ii) that the risk of inequivalence relates to uncertainty of outcome. In this paper, I propose a comprehensive model for dose response and a tentative model for risk that addresses these issues. Specifically, I propose two new measures of bioequivalence which are based on these models, which differ in the two clinical circumstances above, and which respond to both bias and variance of outcome. I present two examples, and some simulations of the application of the new measures.

Biological Availability↗
PubMed1992

Building population pharmacokinetic--pharmacodynamic models. I. Models for covariate effects.

One major task in clinical pharmacology is to determine the pharmacokinetic-pharmacodynamic (PK-PD) parameters of a drug in a patient population. NONMEM is a program commonly used to build population PK-PD models, that is, models that characterize the relationship between a patient's PK-PD parameters and other patient specific covariates such as the patient's (patho) physiological condition, concomitant drug therapy, etc. This paper extends a previously described approach to efficiently find the relationships between the PK-PD parameters and covariates. In a first step, individual estimates of the PK-PD parameters are obtained as empirical Bayes estimates, based on a prior NONMEN fit using no covariates. In a second step, the individual PK-PD parameter estimates are regressed on the covariates using a generalized additive model. In a third and final step, NONMEM is used to optimize and finalize the population model. Four real-data examples are used to demonstrate the effectiveness of the approach. The examples show that the generalized additive model for the individual parameter estimates is a good initial guess for the NONMEM population model. In all four examples, the approach successfully selects the most important covariates and their functional representation. The great advantage of this approach is speed. The time required to derive a population model is markedly reduced because the number of necessary NONMEM runs is reduced. Furthermore, the approach provides a nice graphical representation of the relationships between the PK-PD parameters and covariates.

Adult↗
PubMed1992

Advanced computer programs for drug dosing that combine pharmacokinetic and symbolic modeling of patients.

In this paper, we describe our design for advanced drug dosing programs that "reason" using a combination of Bayesian pharmacokinetic modeling and symbolic modeling of patient status and drug response. Our design is similar to the design of the Digitalis Therapy Advisor program, but extends this previous work by incorporating a Bayesian pharmacokinetic model, performing a "meta-level" analysis of drug concentrations to identify sampling errors and changes in pharmacokinetics, and including the results of this analysis in reasoning for dosing and therapeutic monitoring recommendations. The design has been implemented in a program for aminoglycoside antibiotics called Aminoglycoside Therapy Manager. The program is user-friendly and runs on low-cost general-purpose hardware. The initial validation study showed that the program was as accurate in predicting future drug concentrations as an expert using commercial Bayesian forecasting software and that its dosing recommendations were similar to those of an expert.

Aminoglycosides↗
PubMed1992

A pharmacodynamic model of erythropoietin therapy for uremic anemia.

Fifty-seven patients receiving chronic high-flux hemodialysis began receiving recombinant alpha-human erythropoietin (rHuEPO). The mean initial rHuEPO dose used in 54 evaluable patients was 9963 +/- 4364 U/week; the final dose was 8972 +/- 4058 U/week. Treatment over a mean period of 154 +/- 40 days (84 to 224 days) resulted in an average increase in hematocrit from 24.7% +/- 3.7% to 32.5% +/- 4.4%. We present a model for these data that describes changes in hematocrit during rHuEPO therapy and that allows simultaneous estimation of red blood cell lifespan and rHuEPO-induced increases in red blood cell production rate. Analysis of the hematocrit values of the patients with the model, by use of NONMEM, a computer program for analysis of population data, reveals a nonlinear dose-response relationship with large interindividual variability (coefficient of variation) of about 50%. The estimated mean red blood cell lifespan is 64 days, with interindividual variability of about 30% (coefficient of variation). The intraindividual random variability in hematocrit about its prediction is +/- 5% of the prediction. For clinical dose adjustment, we present a method that uses only simple calculations.

Anemia↗
PubMed1992

Comparing responses when each response is a curve.

We describe, generalize, and demonstrate the application of a method (W. H. Lawton, A. Sylvestre, and M. S. Maggio, Technometrics 14: 513-532, 1972) that can be used to partially analyze population data. The data from each subject consist of a series of responses observed at distinct values of a predictor variable. The method assumes that all subjects' data originate from a common process but differ because the "units" of predictor and response variables differ among subjects. For example, if the predictor variable is time, time can be "faster" or "slower" from subject to subject. We deal with two different problems. In the first one the response at x for the ith subject is of the form beta 1i + beta 2iG[(x - beta 3i)/beta 4i] + epsilon, where G(x) is a mathematical "shape" function (of the predictor variable x) representing the process and epsilon is observation error. The units of observed and predictor variable are then defined by the values of beta 1i, beta 2i and beta 3i, beta 4i, respectively. In particular beta 1i and beta 3i express shifts and beta 2i, beta 4i express scales of the observed and predictor variables, respectively. In the second problem the response at x for the ith subject is of the form beta 1i + beta 2i integral of x0 G[(s - beta 3i)/beta 4i].Hi(x - s) ds + epsilon, where Hi(x) is a known function. In both problems, the method estimates the common "shape" function G(x) nonparametrically and the parameters beta 1i, beta 2i, beta 3i, and beta 4i for each subject.(ABSTRACT TRUNCATED AT 250 WORDS)

Animals↗
PubMed1991

A simulation study comparing designs for dose ranging.

Only with knowledge of the (prior) distribution of dose-response parameters in a population, can one determine both the initial dose of a drug for chronic administration to an individual (such as the dose producing a fixed degree of response in a fixed proportion of the population) and an appropriate subsequent (adjusted) dose (such as the dose yielding a desirable response according to the posterior parameter distribution, given an observed response to an initial dose). The currently FDA-sanctioned design for a dose-ranging study, the parallel-dose design, assigns just one of several doses to each patient. It does not provide good information on the distribution of individual dose-response parameters. A cross-over design assigns several dose levels to each patient. It therefore can provide better information, but does not resemble clinical practice. Consequently, study participants must be restricted to patients who can tolerate such non-therapeutic drug exposure, posing problems in extrapolation of study results to other types of patients. A titration or dose-escalation design begins all patients on placebo and, except for those patients assigned to a placebo-only group, escalates the dose for a patient at preset intervals only when clinical response at lower doses is inadequate. It both exposes patients to several dose levels and resembles good clinical practice, allowing study of a representative patient sample. We report here the simulation results of parameter estimation for the three designs when the data arise from complex and realistic dose-response models and/or with certain complications in study execution. The dose-escalation design clearly performs better overall than the parallel-dose design for the models considered here, and generally, just a little worse than the cross-over design. These results support the conclusion that for dose ranging, depending on the demands of the clinical situation, one should use either the cross-over or the dose-escalation design.

Algorithms↗
PubMed1991

Disposition of prednisone and prednisolone in the perfused rabbit liver: modeling hepatic metabolic processes.

The livers of 15 rabbits were perfused in situ with prednisone (PO) or prednisolone (POH) over a wide range of steady state concentrations, resulting in multiple experimental measurements per organ. Linearity of extraction, an apparent lack of oxidative conversion, and marked preference for the reduction of PO to POH was observed. Predictions of hepatic tissue concentrations were made using both the well-stirred and parallel-tube model approximations. Glucocorticoid disposition across the liver was described by a series of differential equations. Discrimination between the two models was accomplished by examining the effects of changes in flow rate upon the availability of the highly extracted drug PO. The well-stirred model very closely predicted the observed changes in availability of PO, whereas the parallel-tube model provided poor predictions. The intrinsic clearances of interconversion and elimination of PO and POH were subsequently calculated by population analysis using NONMEM. This method assumed the well-stirred model and resulted in intrinsic clearance estimates of 26 ml/min for the elimination of POH, 157 ml/min for reductive conversion of PO to POH, and 205 ml/min for the irreversible elimination of PO. A mechanism of intrahepatic disposition of these glucocorticoids was proposed using well-stirred model predictions of hepatic drug concentrations, the perfusion rate limitation to drug transport, and the assumption of no oxidative interconversion of POH to PO. In this case, the capacity for reduction of PO to POH approaches the elimination clearance of PO and the elimination of PO is about 13 times greater than the elimination clearance of POH.

Animals↗
PubMed1991

Semiparametric analysis of non-steady-state pharmacodynamic data.

We present an approach to the analysis of pharmacodynamic (PD) data arising from non-steady-state experiments, meant to be used when only PD data, not pharmacokinetic (PK) data, are available. The approach allows estimation of the steady-state relationship between drug input and effect. The analysis is based on a model describing the time dependence of drug effect (E) on (unobserved) drug concentration (Ce) in an hypothetical effect compartment. The model consists of (i) a known model for the input rate of drug I(t), (ii) a parametric model; L(t, alpha) (a function of time t, and vector of parameters alpha), relating I to an observed variable X, (iii) a nonparametric model relating X to E. Ce is proportional to X. X (t) is given by I(t) * L(t, alpha)/AL, where L(t, alpha) = e-alpha 1t * sigma k m = 1 alpha 2k e-alpha 2k + 1t, sigma k m = 1 alpha 2k = 1, AL = integral of 0 infinity L(t) dt, and * indicates convolution. The nonparametric model relating X to E is a cubic spline, a function of X and a vector of (linear) parameters beta. The values of alpha and beta are chosen to minimize the sum of squared residuals between predicted and observed E. We also describe a similar model, generalizing a previously described one, to analyze PK/PD data. Applications of the approach to different drug-effect relationships (verapamil-PR interval, hydroxazine-wheal and flare, flecainide and/or verapamil-PR, and left ventricular ejection fraction) are reported.

Computer Simulation↗
PubMed1991

Reversal of neuromuscular blockade in humans by neostigmine and edrophonium: a mathematical model.

Generalizations of the integrated model describing the interaction of nondepolarizing neuromuscular blocking drugs with reversible anticholinesterase drugs described in Unadkat et al. (1) are reported. The models can deal with possible incomplete reversal (irreversible block) and/or noninstantaneous anticholinesterase kinetics. Experimental data were obtained from 22 human volunteers. Different levels of steady-state vecuronium block were induced in each volunteer (in the range of 50% to 95%), and reversed by short infusions of edrophonium (10 volunteers) or neostigmine (12 volunteers). Edrophonium or neostigmine concentrations and twitch tension (measured as the force of thumb adduction) were measured. The generalized integrated models fit the data well. In the case of neostigmine we find a nondistributional delay in its action. We relate this delay to the slow decarbamylation rate of the (neostigmine-induced) carbamylated anticholinesterase observed in vitro, and are able to model such noninstantaneous anticholinesterase kinetic processes. For both edrophonium and neostigmine we detect an inverse relationship between the (induced) level of initial block and maximal percentage recovery.

Adult↗
PubMed1991

Mean time parameters for generalized physiological flow models (semihomogeneous linear systems).

This note gives expressions for recirculation mean time parameters of the disposition kinetics of particles in a semihomogeneous stationary linear system. In such a system each compartment may have an arbitrary single-pass disposition function, rather than a known parametric (usually monoexponential) one. Such systems provide a generalization of physiological flow models. Given observations of arterial blood concentrations and tissue amounts, and making the additional assumptions that (i) the fraction of total blood flow exiting each tissue that goes to each other tissue is constant and known, and (ii) the fraction of drug entering each tissue that is eliminated to the outside is constant and known, the input to each tissue can be known, and therefore both its total blood flow and its single-pass disposition function can be estimated. Recirculation mean time parameters can be computed from these estimates. Application to real thiopental data is presented as an example.

Mathematical Computing↗
PubMed1991

Designs for population pharmacodynamics: value of pharmacokinetic data and population analysis.

Analyses of simulated data from pharmacokinetic/pharmacodynamic (PK/PD) studies varying with respect to the amount and timing of observations were undertaken to assess the value of these design choices. The simulation models assume mono- or biexponential drug disposition, and Emax-type pharmacodynamics. Data analysis uses a combined PK/PD population analysis or a hybrid, individual-PK/population-PD analysis. Assuming that the goal of the PK/PD studies is to estimate population PD, performance of designs is judged by comparing the precision of estimates of population mean PD parameters and of their interindividual variability. The simulations reveal that (i) PK data, even in small number (2 points per person from as few as 25-50% of persons) are very valuable for estimating population PD; (ii) designs involving more individuals, even if many are sparsely sampled, dominate designs calling for more complete study of fewer persons; (iii) the population analysis is generally superior to the hybrid analysis, especially when the PK model is misspecified (biexponential assumed to be monoexponential for analysis); (iv) varying sampling times and doses among subjects protects against the ill effects of model misspecification. In general, the results are quite encouraging about the usefulness of sparse data designs to estimate population dose response.

Computer Simulation↗
PubMed1991

Laboratory data predicts survival post hospitalization.

From a database of 93,077 in-patient admissions, patients assigned to catastrophic, very severe, moderately severe, and average 30-day mortality risk categories (as defined in Medicare Hospital Mortality Information, 1989 release, from the Health Care Financing Administration (HCFA] were selected for study. These admissions account for 30% of all admissions, but 70% of all deaths up to 1 year post admission. To determine whether laboratory information adds to the predictive power of the information used by HCFA, we compare the performance of 1 year survival predictors (Cox model) that use only diagnostic, demographic, and comorbidity information, with the performance of predictors that also include laboratory information. Using a separate set of patients not used for model definition, we find that laboratory data contain significant prognostic information independent of that already available in non-laboratory data. In HCFA's catastrophic disorders for example, non-laboratory information reduces the average risk of predicting a wrong outcome by 17% relative to considering only catastrophic group membership, and adding laboratory data reduces this risk by a further 21%. These improvements result primarily from considering the outcomes of a small set of routine laboratory tests (maximum BUN, AST, and WBC, and minimum CO2, hematocrit, and sodium).

Centers for Medicare and Medicaid Services, U.S.↗
PubMed1991

Population dose versus response of betaxolol and atenolol: a comparison of potency and variability.

The dose-response curves of betaxolol and atenolol were compared in 140 patients with mild to moderate essential hypertension. Patients with a supine diastolic blood pressure of 95 to 115 mm Hg at the end of a 4-week single-blind placebo washout phase were randomized (double-blind) to receive either betaxolol or atenolol in a dose-escalation manner. The dose (5 mg, 10 mg, and 20 mg betaxolol; 25 mg, 50 mg, and 100 mg atenolol) was increased if the supine diastolic blood pressure remained greater than 90 mm Hg after 4 weeks at each level. The final dose in the escalation phase was continued for an additional 12 weeks and then followed by a 2-week placebo phase. The data were analyzed with a population model using the program NONMEM (nonlinear mixed effects model). Atenolol exhibited a graded dose-response curve, whereas the lowest dose of betaxolol produced maximum or near-maximum effect. The estimated maximum effect (drug plus possibly unmeasured placebo effect) was similar for both treatments, about 13 mm Hg (95% confidence interval, 10 to 15 mm Hg). A trend toward less interindividual variability (coefficient of variation) was apparent for betaxolol compared to atenolol, 19% (95% confidence interval, 0% to 29%) versus 31% (95% confidence interval, 0% to 47%). The intraindividual variability (standard deviation) in supine diastolic blood pressure, 5.9 mm Hg (95% confidence interval, 5.2 to 6.5 mm Hg), did not differ significantly between drugs despite significantly greater intraindividual variability (coefficient of variation) in atenolol concentrations, 62% (95% confidence interval, 48% to 73%) versus 26% (95% confidence interval, 22% to 29%) for betaxolol.

Adolescent↗
PubMed1991

Combining physiologic models and symbolic methods to interpret time-varying patient data.

This paper describes a methodology for representing and using medical knowledge about temporal relationships to infer the presence of clinical events that evolve over time. The methodology consists of three steps: (1) the incorporation of patient observations into a generic physiologic model, (2) the conversion of model states and predictions into domain-specific temporal abstractions, and (3) the transformation of temporal abstractions into clinically meaningful descriptive text. The first step converts raw observations to underlying model concepts, the second step identifies temporal features of the fitted model that have clinical interest, and the third step replaces features represented by model parameters and predictions into concepts expressed in clinical language. We describe a program, called TOPAZ, that uses this three-step methodology. TOPAZ generates a narrative summary of the temporal events found in the electronic medical record of patients receiving cancer chemotherapy. A unique feature of TOPAZ is its use of numeric and symbolic techniques to perform different temporal reasoning tasks. Time is represented both as a continuous process and as a set of temporal intervals. These two temporal models differ in the temporal ontology they assume and in the temporal concepts they encode. Without multiple temporal models, this diversity of temporal knowledge could not be represented.

Adult↗
PubMed1989

A semiparametric approach to physiological flow models.

By regarding sampled tissues in a physiological model as linear subsystems, the usual advantages of flow models are preserved while mitigating two of their disadvantages, (i) the need for assumptions regarding intratissue kinetics, and (ii) the need to simultaneously fit data from several tissues. To apply the linear systems approach, both arterial blood and (interesting) tissue drug concentrations must be measured. The body is modeled as having an arterial compartment (A) distributing drug to different linear subsystems (tissues), connected in a specific way by blood flow. The response (CA, with dimensions of concentration) of A is measured. Tissues receive input from A (and optionally from other tissues), and send output to the outside or to other parts of the body. The response (CT, total amount of drug in the tissue (T) divided by the volume of T) from the T-th one, for example, of such tissues is also observed. From linear systems theory, CT can be expressed as the convolution of CA with a disposition function, F(t) (with dimensions 1/time). The function F(t) depends on the (unknown) structure of T, but has certain other constant properties: The integral integral infinity0 F(t) dt is the steady state ratio of CT to CA, and the point F(0) is the clearance rate of drug from A to T divided by the volume of T. A formula for the clearance rate of drug from T to outside T can be derived. To estimate F(t) empirically, and thus mitigate disadvantage (i), we suggest that, first, a nonparametric (or parametric) function be fitted to CA data yielding predicted values, CA, and, second, the convolution integral of CA with F(t) be fitted to CT data using a deconvolution method. By so doing, each tissue's data are analyzed separately, thus mitigating disadvantage (ii). A method for system simulation is also proposed. The results of applying the approach to simulated data and to real thiopental data are reported.

Models, Biological↗