Introduction to admiralonco

admiralonco is an extension of the admiral package, designed specifically to handle oncology-specific data transformations in clinical trials. This package helps streamline the creation of CDISC ADaM datasets, ensuring compliance with regulatory standards.

In this tutorial, we will cover:

• The purpose and scope of admiralonco.
• Installation instructions.
• Overview of dependencies and compatibility.

1. Purpose and Scope

The admiralonco package provides functions tailored to the oncology domain, including:

• Handling tumor assessments (e.g., RECIST).
• Deriving progression-free survival (PFS) and overall survival (OS) endpoints.
• Managing longitudinal data from oncology studies.

By leveraging admiralonco, users can generate ADaM datasets efficiently while ensuring consistency with industry best practices.

2. Installation Instructions

To install admiralonco, first install the admiral package if it is not already available:

# Install admiral from CRAN
install.packages("admiral")

Next, install admiralonco from GitHub:

# Install devtools if not already installed
install.packages("devtools")

# Install admiralonco from GitHub
devtools::install_github("pharmaverse/admiralonco")

3. Overview of Dependencies and Compatibility

The admiralonco package is designed to work within the pharmaverse ecosystem. It has the following dependencies:

• admiral: Provides core ADaM dataset transformation functions.
• dplyr: Used for data manipulation.
• tidyr: Assists in data reshaping.
• lubridate: Handles date-time conversions in clinical datasets.

To check whether all dependencies are installed, use:

# Check for missing dependencies
required_packages <- c("admiral", "dplyr", "tidyr", "lubridate")
missing_packages <- required_packages[!(required_packages %in% installed.packages()[, "Package"])]
if(length(missing_packages)) install.packages(missing_packages)

After installation, you can load the package using:

# Load the admiralonco package
library(admiralonco)

4. Example Usage

Let's load a sample oncology dataset and apply a function from admiralonco:

# Load required libraries
library(admiral)
library(admiralonco)

# Example vector of tumor responses
responses <- c("SD", "PR", "CR", "PD", "NE")

# Convert response categories to numeric values
numeric_responses <- aval_resp(responses)

# Display the results
print(numeric_responses)
[1] 3 2 1 5 6

By running this example, we transform categorical response values (Progressive Disease, Stable Disease, Partial Response, Complete Response, Not Evaluable) into numerical scores for easier statistical analysis.

Conclusion

The admiralonco package is a powerful tool for managing oncology-specific ADaM datasets in R. By automating common transformations, it helps streamline clinical trial data processing and ensures compliance with CDISC standards.