Estimate of pharmacokinetic parameters. From non-Compartmental data using -Microsoft Excel-24

Estimate of pharmacokinetic parameters (CDISC) using Microsoft Excel-estimation , Pharmacokinetic parameters, such as clearance (CL) and volume of distribution (V), represent constants that vary based on the model used, illustrating the relationship between drug input (dose, dosing interval, dosing time, sampling times) and output (drug concentrations). Understanding these parameters is crucial for tasks like providing initial estimates for Bayesian population analysis, designing clinical trials, and predicting concentration-time profiles for future studies and clinical practice. Two main approaches, compartmental and non-compartmental, are employed to estimate these pharmacokinetic parameter values. Excel PK calculation download Free

Table of Contents

Compliant datasets-CDISC-Estimate of pharmacokinetic parameters (CDISC)

Numerous sponsors can confirm that preparing datasets compliant with CDISC standards is both time-intensive and expensive. It’s crucial to factor in existing and forthcoming data regulations when strategizing studies for your development program.
Two significant SDTM domains for PK analyses and pharmacometrics are the PC domain, housing PK concentrations, and the PP domain, containing PK parameters data computed by PK scientists. At PhinC, we recognize that adopting CDISC standards may entail additional expenses. However, leveraging these standards can benefit sponsors in several ways:
Standardizing the analysis and reporting of data. Streamlining the pooling of data for future analyses such as PK population modeling. Simplifying the review of study data by regulatory authorities. Enhancing the applicability of study findings (including PK) across different contexts.

Implement standards PK CDISC

A demonstration of incorporating SDTM datasets into a typical CDISC compliant PK workflow is depicted in the flowchart provided. Adhering to PK CDISC standards poses a significant challenge initially, primarily due to the need for efficient coordination among all stakeholders involved in the CDISC process:
data management, bioanalysis, and regulatory submission. Consequently, it is highly recommended to engage experienced PK scientists early in the process. This ensures effective communication among team members and prevents working in isolation, which is commonly referred to as operating “in silos.” How PhinC Can Offer Assistance:

By integrating pharmacometricians and PK experts into the data transfer and analysis dataset creation processes, traditionally handled by data managers and clinical programmers, PhinC ensures that transferred data encompass all essential elements required for analyses and modeling. This approach minimizes delays post-analysis commencement by streamlining communication and enhances compliance with CDISC standards. Notably, it’s been observed that few CDISC/data management teams possess the internal resources to guarantee the delivery of high-quality PK datasets in CDISC format, particularly for intricate PK studies.
By employing a native CDISC process for PK analyses, PhinC ensures the seamless generation of CDISC compliant results (e.g., SDTM datasets) that are easily reproducible and encompass all analysis aspects (e.g., reasons for uncalculable PK parameters, flags, populations). Both the source (bioanalysis) data and analysis results (pharmacokinetics) are produced directly in CDISC format, as opposed to retroactively converting data to CDISC format (mapping), a procedure susceptible to errors that may diminish the reusability of PK results.

Excel formula for PK-Equation
Compartmental approaches necessitate specialized software for parameter estimation through nonlinear regression analysis within a specified model. Conversely, non-compartmental approaches can be executed using general spreadsheet programs to compute values such as maximum drug concentration (Cmax), time to reach maximum concentration (tmax), and area under the concentration-time curve (AUC). These variables are then used to calculate pharmacokinetic parameters using standard formulas. Calculate Using SAS

PK SOLVER for Excel add-in

Non-compartmental analysis is often preferred in many pharmacokinetic studies due to its seemingly less restrictive assumptions and its provision of variables relevant to certain study types. However, deriving pharmacokinetic parameter values from non-compartmental variables (NCVs) can sometimes be straightforward (e.g., for CL from AUC), while for others, like absorption rate constant (ka), it may pose more challenges. Nevertheless, such computations inherently imply specific compartmental assumptions, thereby mitigating some advantages purported by non-compartmental analysis. Despite being commonly reported in literature, NCVs hold limited predictive value in clinical applications as their accuracy heavily depends on factors like dose, dosing interval, and the timing of blood samples, compared to compartmental analysis.
The motivation behind our study stems from the widespread availability of non-compartmental variables (NCVs) in pharmacokinetic literature, and the significant relevance of understanding pharmacokinetic parameters for both clinical and research purposes. A previous study introduced a technique for estimating the absorption rate constant (ka) using SolverJ based on the observed value of tmax. However, their approach necessitated prior estimation of certain pharmacokinetic parameters from intravenous data. Presently, there exists no research outlining a method for concurrently estimating all model-dependent pharmacokinetic parameters solely from NCVs. To address this gap, we have devised a technique termed ‘back analysis’ (BA), enabling the simultaneous estimation of pharmacokinetic parameter values from NCVs when original concentration-time data are unavailable. It’s important to note that the BA method is not recommended when concentration-time data are accessible, as conventional modeling techniques are preferred in such cases.
This paper presents a motivating example in Section 2 to illustrate the rationale behind the development of the BA method. Sections 3 and 4 detail the implementation and evaluation of the BA method, respectively, utilizing simulated data in both sections. Furthermore, Section 5 demonstrates the application of the BA method using our motivating example.
BA method implementation: The BA technique aims to translate a series of Non-compartmental Analysis (NCV) results into pharmacokinetic parameters for both one- and two-compartment models. This is achieved through a conventional two-stage process: initially, the NCV data is utilized to calculate individual pharmacokinetic parameters for a specific model; subsequently, in the second stage, these parameters from multiple individuals are aggregated to determine the population’s mean and variance.

Excels:Estimate/Derivation of pharmacokinetic parameters (CDISC)

Excel PK calculation download Free
Microsoft Excel was originally designed to cater to the demands of the competitive business software market. This might lead one to question its suitability for pharmacokinetic (PK) and pharmacodynamic (PD) simulation and modeling. Can a software primarily intended for business applications handle the complex calculations required in PK/PD simulation and modeling? The answer is unequivocally ‘yes!’ While many software programs operate by inputting data, executing algorithms, and producing results, users are often left unaware of the intricate processes occurring in between. However, with a spreadsheet like Excel, users have the advantage of observing the entire calculation process from start to finish.
For instance, during one instance, we encountered difficulties understanding why NONMEM would crash with specific initial estimates. By employing Excel, we were able to pinpoint the root cause effortlessly. Initially, we developed a custom ordinary differential equation solver in C to implement the intricate PK model within Excel. Subsequently, we configured a mixed-effects analysis utilizing our XLMEM code. This enabled us to visually inspect the predicted concentrations for various initial estimates and scrutinize each value within the matrices involved in the calculation of the extended least squares objective function during optimization. Once we gained a comprehensive view of the issue, the solution became apparent.