Lead Investigator: Lena Friberg, Uppsala University
Title of Proposal Research: Evaluation of dose individualization strategies in optimizing survival and patient-reported outcomes of sunitinib treatment in patients with metastatic renal cell carcinoma
Vivli Data Request: 10039, 8807
Funding Source: None
Potential Conflicts of Interest: None
Summary of the Proposed Research:
Renal cell carcinoma (RCC) is a type of cancer that affects the kidney. According to Global Cancer Observatory (GLOBOCAN) data, in 2020, over 400,000 new cases of RCC were diagnosed, and around 180,000 people died from the disease. People with metastatic RCC (mRCC, when tumor has spread to other parts of their body) used to have a low chance of surviving for 5 years of 7.3% (during 1992-1995), but now, thanks to new medicines, the survival rate has improved to around 14% (during 2011-2017).
Sunitinib is a drug that has been approved for treating mRCC for more than 15 years. It is currently the most cost-effective initial (first-line) treatment for this type of cancer. However, how sunitinib is processed by the body can vary greatly among different patients, and we do not fully understand why. In a study of 60 mRCC patients taking a fixed dose of 50mg of sunitinib every day, the amount of the drug found in their blood plasma varied widely, ranging from 24.5 to 209.4 nanograms per milliliter. This big variability in blood concentration is problematic, since studies have shown that patients with lower levels of sunitinib in their blood may have a lower chance of responding to the treatment and thus lower chance of survival, and that giving them a higher dose may help. However, if the dose is too high, it can cause serious side effects that not only make the treatment less effective but also lower the patient’s quality of life.
Doctors and researchers have been looking into ways to personalize the dose of sunitinib for patients with mRCC instead of giving the same dose to everyone. But there isn’t much clear evidence on how to do this yet, and there aren’t many guidelines or tools to help making these personalized dosing decisions. Therefore, this study aims to gather and analyse patient level data collected across different previously conducted clinical trials to see if there are certain factors, such as the amount of sunitinib in plasma or other indicators in the blood (biomarkers), that can help predict how well a patient will respond to sunitinib treatment.
In this study, we want to explore the inclusion of the perspective and experiences of patients as an important factor when evaluating predictors of treatment outcomes, alongside side effects reported by doctors and survival. Patient-reported outcomes (PROs) are data collected from patients about how they feel during treatment, such as improvement in symptoms related to the disease, any side effects they experience that is hard to tolerate, and their quality of life. By looking at patient level data that includes information about how much of the drug is in their system, any changes in their biomarkers, and their PROs over time, we hope to better understand how we can personalize sunitinib treatment to fit the needs and experiences of each mRCC patient.
Statistical Analysis Plan:
The intended study population includes patients with mRCC who received at least one dose of sunitinib as first-line treatment and who participated in phase I-III clinical trials. Trials investigated sunitinib as a single-arm treatment or compared sunitinib versus interferon-alfa as a first-line treatment for mRCC patients, with data available for survival and patient-reported outcomes (PROs) that are available in Vivli were identified.
This work will be conducted in the Uppsala Pharmacokinetic (PK) and Pharmacodynamic (PD) group by applying pharmacometrics analysis methods in the following steps:
1. A dataset suitable for non-linear mixed effects (NLME) and time-to-event modeling will be prepared.
2. A graphical analysis will be performed and guide the model-building process.
3. To maintain the structure and the independence of the studies being requested, exploratory analysis of data will be conducted to explore if the same model structure could be applied to different studies. For studies combined for analysis, the study as a categorical covariate will be tested on key model parameters.
4. Handling missing values
• Missing dose information:
i. Actual time of dosing will be used in the analysis. If the dosing time or dose amount is unknown, the nominal time or dose amount will be imputed.
ii. When dosing information is missing and if it is clear from the data that the dose was not administered, then the dose record will be excluded in the analysis.
• Missing PK and PD observations:
i. Actual time of measurements will be used in the analysis, in cases of missing, the default rule is not to include the observation.
• Missing outcomes:
i. Patients with missing survival outcome measurements will not be included in the analysis set.
ii. Missing safety outcomes or patients’ reported outcomes will not be included in the analysis.
• Missing covariates
i. Missing baseline covariates will be replaced by a screening value if available.
ii. Missing continuous covariates will be replaced with the median value in the population. Body weight will be age and sex-specific if replaced. Missing categorical covariates will be replaced by the most common category.
ii. If a covariate is missing in more than 15% of the subjects, more sophisticated methods may be used to impute the covariate based on regression models between covariates.
iv. If a covariate is missing in more than 25% of the subjects, it will be excluded from the analysis.
5. Models for outcomes:
• Parametric time-to-event model will be used to describe the survival endpoints PFS and OS.
• Longitudinal PROs will be described with the item response theory (IRT) pharmacometrics model.
6. Models for biomarkers/predictors:
• The previously developed population PK model will be applied to describe the observed concentrations of sunitinib and SU-012662. Model-based individual estimates of clearance, area under the curve (AUC), and/or Chi-square value (Cmin) will be derived.
• The previously developed population PD models will be applied to describe the longitudinal measurements of VEGF, sVEGFR‐2, sVEGFR‐3, and tumor sizes.
• A population model describing the time course of PIGF will be developed.
7. Integration of PK, PD, and clinical outcomes
• Possible connections between the PK and PD models described above will be investigated to build a model framework. For example, the plasma concentrations of sunitinib and SU-012662 may drive the changes of sVEGFR-3 or PIGF, which in turn may drive changes in tumor size.
• The ability of covariates and model-derived individual PK metrics (e.g. clearance, AUC, Cmin) and PD metrics (e.g. tumor growth rate, VEGF change from baseline) in predicting efficacy and quality of life will be tested by linking to the outcome models.
8. The goodness-of-fit of the models to observed data will be evaluated using statistical methods, diagnostic plots, and visual predictive checks (VPC). The Wald statistics test was used for model evaluations, and the likelihood ratio test was used to compare hierarchical models
9. Create a simulation framework for the evaluation of a dose individualization algorithm based on integrated models (this step does not require access to additional data).
For data cleaning and management, the statistical software R will be used, and for non-linear mixed effects (NLME) modeling the software NONMEM will be used, peral-speaks-NONMEM (PsN) and Pirana will be used to manage NONMEM runs.
Requested Studies:
A Phase 3, Randomized Study Of SU011248 Versus Interferon-Alfa As First-Line Systemic Therapy For Patients With Metastatic Renal Cell Carcinoma
Data Contributor: Pfizer Inc.
Study ID: NCT00083889
Sponsor ID: A6181034
A Phase II Efficacy And Safety Study Of Sunitinib Malate (SU011248) Administered In A Continuous Daily Regimen In Patients With Advanced (First-Line) Renal Cell Cancer
Data Contributor: Pfizer Inc.
Study ID: NCT00338884
Sponsor ID: A6181110
Phase II Study Of Single-Agent SU011248 In The Treatment Of Patients With Renal Cell Carcinoma
Data Contributor: Pfizer Inc.
Study ID: NCT00254540
Sponsor ID: A6181072
A Randomized Phase II Study Of The Efficacy And Safety Of Sunitinib Malate Schedule 4/2 vs. Sunitinib Malate Continuous Dosing As First-Line Therapy For Metastatic Renal Cell Cancer (Renal EFFECT Trial)
Data Contributor: Pfizer Inc.
Study ID: NCT00267748
Sponsor ID: A6181065
A Phase 2 Efficacy And Safety Study Of SU011248 Administered In A Continuous Daily Regimen In Patients With Cytokine-Refractory Metastatic Renal Cell Carcinoma
Data Contributor: Pfizer Inc.
Study ID: NCT00137423
Sponsor ID: A6181061
Phase II Study Of Single-Agent SU011248 In The Second-Line Treatment Of Patients With Metastatic Renal Cell Carcinoma
Data Contributor: Pfizer Inc.
Study ID: NCT00054886
Sponsor ID: RTKC-0511-014
A Pivotal Study Of SU011248 In The Treatment Of Patients With Cytokine-Refractory Metastatic Renal Cell Carcinoma
Data Contributor: Pfizer Inc.
Study ID: NCT00077974
Sponsor ID: A6181006
Public Disclosures:
Liu H., Ray T., Friberg LE. Comparison between multistate and parametric time-to-event models for analysis of progression-free-survival (PFS) and overall survival (OS) in oncology trials. 2024. Population Approach Group People, PAGE 32. Abstract 10980.