A clinically applicable mathematical model was developed by researchers to predict outcomes to immunotherapy for patients with cancer previously treated with anti–CTLA-4 or anti–PD-1/PD-L1 antibodies.
Researchers developed a clinically applicable mathematical model that can predict patient outcomes to immunotherapy, while also acting as a tool for personalized oncology and for engineering proper immunotherapy regimens, according to a study published in Science Advances.1-2
This model calculates a predicted response on a per-patient basis, enabling it to provide a framework for individualized treatment strategies.2 More, the parameters of the model successfully differentiate between pseudo-progression from true progression, ultimately providing unidentified insights into the characteristics of pseudo-progression.
“This paper applied a refined mechanistic mathematical model of immunotherapy response to a prospective clinical trial of checkpoint inhibitor therapy, validated with two separate cohorts representing various immunotherapy drugs, drug mechanisms, and cancer types, to predict the outcomes of patients,” wrote the researchers.1“In doing so, we demonstrated that our model can accurately describe the variable response patterns of patients and discern pseudo-progressors from traditional responders and nonresponders.”
The model parameters identified 18 patients with common and 10 patients with rare malignancy types who benefitted or did not benefit from these monotherapy types. The accuracy at first restaging was as high as 88%.
The researchers analyzed the CT-scan imaging data of tumors from before, during, and after immunotherapy of 121 patients previously treated with checkpoint inhibitor immunotherapy. Then, the data was validated with an additional 124 patients who were previously treated with common checkpoint inhibitor immunotherapy.
In total, the study retrospectively applied this model to 245 patients across multiple clinical trials who were treated with anti–CTLA-4 or anti–PD-1/PD-L1 antibodies.
“We use tumor growth rate obtained from standard-of-care CT imaging, and model-derived quantifications of tumor cell kill rate and antitumor immune state to predict treatment outcomes for a mixed population of patients undergoing checkpoint inhibitor treatment for metastatic disease of various primary tumor types, and demonstrate methodologies whereby the model may be implemented prospectively as a predictor of patient response and survival at times as early as first restaging,” wrote the researchers.
The research had a number of limitations starting with its approach as a retrospective review of a prospective study. More, the cohort was heterogeneous including only patients with liver and/or lung metastasis. Also, different institutions and radiologists had varying tumor boundary identification and variations in CT imaging protocol, both of which have not been investigated at this stage.
Some of the included patients also underwent different combinations of radiation therapy, which were removed from the overall calculation of tumor burden. The overall abscopal effects remain unclear, and a small number of patients could not remain with the study through first restaging. The researchers were unable the model to these patients because of a lack of data.
“This early predictor of immunotherapy response may allow for critical adjustments to individual patient treatments and is one step closer to the promise of personalized medicine,” wrote the researchers. “Future investigations will focus on expanding the scope of this model to other clinical datasets and further refinement of the complex interactions between the tumor and its milieu.”
1. Butler JD, Elganainy D, Wang CX, et al. Mathematical prediction of clinical outcomes in advanced cancer patients treated with checkpoint inhibitor immunotherapy. Science Advances. DOI: 10.1126/sciadv.aay6298.
2. Clinically applicable math model predicts patient outcomes to cancer immunotherapy [news release]. Houston Methodist. Published April 30, 2020. https://www.eurekalert.org/pub_releases/2020-04/hm-cmm042920.php. Accessed May 7, 2020.
Related Content:Immuno Oncology