Following are public and technical abstracts for the Biomarkers for Nivolumab project funded by the Department of Defense Kidney Cancer Research Program (KCRP) for 2017.
Principal Investigators: Sabina Signoretti and Toni Choueiri (research partners)
Institutions: Brigham and Women’s Hospital and Dana-Farber Cancer Institute
Funding Mechanism: Translational Research Partnership
Award Amounts: $392,010 and $672,464
Public Abstract
On November 23, 2015, the US Food and Drug Administration (FDA) approved nivolumab to treat advanced kidney cancer. The approval came following the results of a trial, called CheckMate 025, that compared the survival of patients who received nivolumab to those who received everolimus, which was a standard of care treatment modality. The study was done among individuals who were diagnosed with advanced kidney cancer, and who had previously been treated with one or more drugs. In that study, the investigators found that the survival of patients treated with nivolumab was better than those treated with everolimus. Despite these benefits, only a subset of patients responded to the nivolumab treatment in that trial. Within the study, that proportion was determined to be around 25%. In an attempt to get more individuals to respond to immunotherapy, investigators tested the combination of ipilimumab (another type of immunotherapy) and nivolumab vs. sunitinib in patients with advanced kidney cancer. As hoped, the combination therapy resulted in more patients responding to the combination, but not without the occurrence of significant serious side effects. Consequently, clinicians are hoping to find ways to be able to identify patients (before therapy) that are more likely to respond to just nivolumab, and thereby sparing patients who are unlikely to respond to go through the unnecessary treatment and potential toxic side effects of the combination of nivolumab+ipilimumab.
The advances of biomarker technologies have made cancer treatment evolve from the classic “one-drug-fits-all” to a more personalized strategy where treatment regimen is driven by individual biomarker expression profiles. Right now, there is no biomarker used in routine clinical practice for selecting patients who are likely to benefit from nivolumab for advanced kidney cancer. Working towards this goal, our team previously identified several biomarkers found in the blood and in the tumor tissues. However, these studies were performed in smaller clinical datasets with a limited number of patients. Moreover, biomarker status was not always available in the control population (i.e., patients not treated with the experiment of interest), which is an important detail in determining the real clinical utility of these biomarkers. Hence, the goal of the current proposal is to validate our initial findings by using samples obtained from the CheckMate 025 trial, a randomized phase 3 trial that contains a control arm (i.e., those treated with everolimus). We will test the relationship between the biomarkers we studied before with the time to progression of disease and time to death. Moreover, we will be measuring how each patient responds to the drug through a set of established rules that define when cancer patients respond, stay stable, or worsen during treatment.
From a worldwide perspective, both the European Medicines Agency (EMA) and the FDA are encouraging the development of biomarkers as companion diagnostics in general practice. From the economical burden and cost-effectiveness point of view, there is a clear compelling argument to develop and test biomarkers that can be accurately and safely used in clinical practice to identify the responsive subpopulation. While a handful of biomarkers have reached clinical practice in other cancers, there is currently no predictive biomarker during patient counselling and treatment management decision in kidney cancer. In other cancers, the identification, development/identification, and validation/reconfirmation of biomarkers based on rigorous scientific studies and tested in focused, well-designed clinical trials allowed more efficient clinical development, as well as an associated reduced failure rate of drug development. For the kidney cancer patient, the benefits of completing this proposal will be enormous, reflecting the concept of the “right drug for the right patient.”
Technical Abstract
Background: On November 23, 2015, the US Food and Drug Administration (FDA) approved nivolumab, a fully human monoclonal antibody (mAb) directed against PD-1, for patients with advanced renal cell carcinoma (RCC). In a randomized, controlled phase III trial, nivolumab demonstrated a clinically significant overall survival (OS) improvement compared to everolimus for patients with advanced RCC in the post-anti-angiogenic setting with an acceptable toxicity profile. Despite the OS benefit, only a subset of patients (between 20%-25%) experienced a durable response and benefited from nivolumab. While an even better response rate can be expected through the combination of nivolumab with the anti-CTLA 4 agent ipilimumab, as shown recently, significant toxicities can also be anticipated for those patients. Given these considerations, the selection of patients most likely to respond to nivolumab alone thus represents an urgent priority.
Objective and Hypothesis: There are currently no predictive biomarkers used in clinical practice for selecting patients who are likely to benefit from immunotherapy in metastatic RCC. The use of biomarkers would allow clinicians to be able to identify patients who will derive a benefit from nivolumab, sparing the ones unlikely to benefit from unneeded toxicities and additional costs. Working towards this goal, we have previously identified both blood-based and tissue-based biomarkers that demonstrated a strong correlation with response following checkpoint inhibitors. Building upon our preliminary findings, our objective is to validate these biomarkers using unique patient resources from the CheckMate 025 phase III trial, which also includes a control, non-nivolumab arm (everolimus). Subsequent to validating our biomarkers, we will also develop a multi-biomarker model for prediction of benefit from nivolumab in metastatic ccRCC.
Specific Aims
Aim 1 will focus on the validation of blood-based biomarkers of response to nivolumab in metastatic ccRCC. Aim 1a: We plan on validating the impact of nivolumab on the metabolite kynurenine and to examine its relationship with our clinical endpoints. Aim 1b: We plan on validating the predictive value of baseline serum adenosine levels in patients treated with nivolumab versus everolimus.
Aim 2 will focus on the validation of tissue-based biomarkers of response to nivolumab in metastatic ccRCC. Aim 2a: We plan on validating the predictive value of tumor cell PD-L1 expression combined with high percentage of CD8+ tumor infiltrating cells expressing PD-1 (but not TIM-3 or LAG-3) in patients treated with nivolumab versus everolimus. Aim 2b: We plan on validating the predictive value of somatic loss-of-function PBRM1 alterations in patients treated with nivolumab versus everolimus. Aim 2c (exploratory): We plan on investigating the predictive value of T-effector and myeloid-associated gene signatures in patients treated with nivolumab versus everolimus.
Study Design: The current proposal represents a post-hoc analysis of the CheckMate 025 phase III trial, a study comparing nivolumab vs. everolimus for advanced RCC. Specimens from 575 patients, including 324 treated with nivolumab and 251 treated with everolimus (controls) are available. The current proposal will assess whether a biomarker is predictive of treatment benefit, defined as progression-free survival (PFS)/immune-related PFS, which represents our primary endpoint. Secondary endpoints will consist of OS and objective response rate (ORR)/immune-related ORR. Finally, we will develop a multi-biomarker model for prediction of PFS by exploring the performance of the model with and without established clinical factors.
Impact: The current proposal possesses both short- and long-term impact. In the short-term, the completion of the current proposal will achieve the establishment of the clinical validity (i.e., the validation of biomarkers as predictive of treatment response in a prospective-retrospective analysis) and clinical utility (i.e., the prediction of response to treatment that can result in actual changes in clinical management) of the tested biomarkers. In the long-term, the proposal can enhance existing risk stratification and targeting of interventions in order to maximize benefits and minimize harms within the current treatment management of metastatic RCC. We hope that the current proposal may eventually serve to develop an all-comers, prospective biomarker-stratified design feasibility trial.