An Interactive Tool for Efficacy Analyses
Efficacy analyses for new targeted treatment in early phase usually require close collaboration between the biometric team and the clinical development team to identify delicate signals hidden in patient subgroups defined by biomarkers. The conventional communication paradigm, based on pre-planned TLF (tables, listings, and figures) deliverables, is often inefficient.
A flexible presentation of PFS/OS/ORR/DoR data is very much desired to facilitate communication among a drug development team. Below is a demonstration of a dynamic tool to allow highly customizable statistical analyses for efficacy endpoints commonly seen in an oncology trial.
Beginning with a set of validated ADaM data (simulated in this instance), what types of analyses would you be interested in exploring? Feel free to utilize the various analytical tools provided below for your investigation.
Kaplan-Meier Curves (for PFS/OS/DoR)
Two biomarkers are simulated: FLG1, a binary marker in ADSL, and FLG2, another binary marker for the targeted treatment, defined by applying a cut-off value to BMI. You can adjust the cut-off value here:
Count by Flags
FLG1=No | FLG1=Yes | TOTAL | |
BMI |
|||
BMI |
|||
TOTAL |
Follow each patient for a maximum of
Number of subjects still at risk
Median and Hazard Ratio
Event and Censored
Objective Response Rate (ORR)
The bar chart displays the ORR results for each subgroup based on The Response Evaluation Criteria in Solid Tumors (RECIST), including Complete Response (CR), Partial Response (PR), Stable Disease (SD), Progressive Disease (PD), and Not Evaluable (NE). The chart also presents the count and rate of ORR, calculated as the sum of complete and partial responses.
Since the chart is dynamic, it can be applied to any subgroup analysis, providing flexibility for both the biometrics and clinical development teams to efficiently explore and interpret data across different patient populations..
Two More Simulated ADaM Datasets
ADTTE
- USUBJID: dynamic ITT with default sample size =
- AVAL: integer samples from a uniform distribution
- CNSR: samples from a binomial distribution (n = 1, p = 0.2)
- PARAMCD: flag for endpoint
ADRS
- USUBJID: dynamic ITT with default sample size =
- AVALC: 'SD', 'CR', 'PR', 'PD', 'NE' simulated by using multinomial distribution with prob = [0.2, 0.4, 0.2, 0.1, 0.1]
- PARAMCD: flag for response criteria