Stephanie Kollmann 1, Chloe Bracis 1, Geraldine Celliere 1
1 Simulations Plus, Product & Technology division (Antony, France)
INTRODUCTION
Clinical reporting of Tables, Figures and Listings (TFLs) is often time-consuming and error-prone due to repetitive post-processing and formatting steps. We present monolixSuiteTFL, an R package designed to harmonize and automate the generation of reporting-ready tables and listings. It retrieves analysis outputs from Monolix/PKanalix projects via the MonolixSuite API, and applies standardized, customizable formatting. By reducing manual formatting the package enables pharmacometricians, statisticians, and clinicians to focus on interpretation rather than presentation while improving reproducibility and consistency across studies. In addition, it supports reporting elements aligned with regulatory expectations, including diagnostics not available as standard Monolix plots, such as random-effects distribution checks (Q–Q), |IWRES| versus IPRED/time, and simulation-based forest plots for communicating covariate effects on exposure. [1]
METHODS
The package is organized into three complementary layers.
Data extraction and table/listings formatting.
Functions retrieve analysis outputs from loaded Monolix and PKanalix projects via the MonolixSuite API and transform them into reporting-ready R data frames. Flexible user arguments enable multiple table or listing variants from the same underlying results. It supports further inspection, elements combination, and post-processing if needed.
Reusable styling engine.
The package applies consistent, report-wide visual standards and returns flextable objects. Because it operates on generic R data frames, it ensures harmonized formatting even for clinical study report (CSR) components that are not directly tied to modeling outputs (e.g., exclusion listings or covariate statistics tables), or generated entirely outside MonolixSuite.
Extended diagnostic plotting.
monolixSuiteTFL package provides diagnostic plots not available through the MonolixSuite API. It creates them as ggplot2 objects and uses data directly from the Monolix and PKanalix project. [2]
The implementation includes a minimal set of widely adopted R packages to support a broad range of formatting needs while keeping workflows lightweight and reproducible in scripted reporting pipelines (R/Quarto). [3]
RESULTS
The package was evaluated on a set of internal Monolix and PKanalix projects covering representative use cases in population PK and non-compartmental analysis (NCA) workflows. Using a single function call per target output, the package retrieved the relevant results via the MonolixSuite API and produced report-ready tables and listings with standardized styling and user-configurable formats. The generated tables and listings reproduced the values displayed in the Monolix/PKanalix graphical interface.
Across projects, table/listing were generated with minimal code (often 1-2 lines), eliminating the need for report-specific formatting script that is hard to reuse across different studies.
Reports included both harmonized default outputs and custom variants of the same table/listing through argument-based settings, enabling reuse across studies while maintaining flexibility for study-specific requirements. Outputs were consistent with the results available in the Monolix/PKanalix user interface. Because they are returned as standard R data frames or flextable objects, they can be easily integrated into scripted pipelines (e.g., R/Quarto), reducing manual post-processing steps and formatting errors.
Beyond core modeling outputs, the package generated complementary CSR components such as analysis-independent listings (e.g., exclusion listings; covariate summary tables), facilitating comprehensive clinical reporting workflows. Where additional refinements were needed, outputs were standard R data frames, making post-processing straightforward and reproducible.
Separately, CSR-ready outputs that are computed entirely outside MonolixSuite were produced in R and integrated into the same report without breaking formatting consistency. The package’s table/listing styling functions accept any R data frame, so externally generated tables and listings can be rendered with the same visual standards as MonolixSuite-derived outputs in a single consolidated CSR report.
CONCLUSIONS
By combining direct MonolixSuite result retrieval with standardized yet highly customizable TFL styling, monolixSuiteTFL enables reproducible, harmonized clinical reporting workflows while minimizing error-prone manual post-processing.The package complements existing reporting workflows and supports comprehensive CSR deliverables, including analysis outputs, study listings, and regulatory-aligned diagnostic visualizations. This approach promotes consistency across studies and facilitates scalable, automated reporting pipelines for pharmacometric analysis.
References:
[1] U.S. Food and Drug Administration. Population Pharmacokinetics: Guidance for Industry. 2022; (Final guidance; FDA)
[3] https://cran.r-project.org/web/packages/ggplot2/index.html
[2] https://quarto.org
Reference: PAGE 34 (2026) Abstr 12227 [www.page-meeting.org/?abstract=12227]
Poster: Methodology - Other topics