Modeling tumor size dynamics based on real‐world electronic health records and image data in advanced melanoma patients receiving immunotherapy

Abstract

The development of immune checkpoint inhibitors (ICIs) has revolutionized cancer therapy but only a fraction of patients benefits from this therapy. Model‐informed drug development can be used to assess prognostic and predictive clinical factors or biomarkers associated with treatment response. Most pharmacometric models have thus far been developed using data from randomized clinical trials, and further studies are needed to translate their findings into the real‐world setting. We developed a tumor growth inhibition model based on real‐world clinical and imaging data in a population of 91 advanced melanoma patients receiving ICIs (i.e., ipilimumab, nivolumab, and pembrolizumab). Drug effect was modeled as an ON/OFF treatment effect, with a tumor killing rate constant identical for the three drugs. Significant and clinically relevant covariate effects of albumin, neutrophil to lymphocyte ratio, and Eastern Cooperative Oncology Group (ECOG) performance status were identified on the baseline tumor volume parameter, as well as NRAS mutation on tumor growth rate constant using standard pharmacometric approaches. In a population subgroup (n = 38), we had the opportunity to conduct an exploratory analysis of image‐based covariates (i.e., radiomics features), by combining machine learning and conventional pharmacometric covariate selection approaches. Overall, we demonstrated an innovative pipeline for longitudinal analyses of clinical and imaging RWD with a high‐dimensional covariate selection method that enabled the identification of factors associated with tumor dynamics. This study also provides a proof of concept for using radiomics features as model covariates.

Current Status and Future Directions: The Application of Artificial Intelligence/Machine Learning for Precision Medicine

“…Furthermore, the synergy between ML and the integration of high‐dimensional datasets presents an innovative avenue for leveraging novel biomarkers such as liquid biopsy circulating tumor DNA and image‐based radiomics. For the latter, it first attempts to integrate radiomic features into models of tumor growth dynamics report the use of ML‐based radiomics feature selection in the real‐world setting 31 . Leveraging tumor phenotype features extracted from images into longitudinal disease models has a great potential to deepen the comprehension of tumor evolution and progression dynamics, thereby contributing to informed clinical decision making and to advance precision medicine.…”

Machine Learning in Modeling Disease Trajectory and Treatment Outcomes: An Emerging Enabler for Model‐Informed Precision Medicine

“…Advances in liquid biopsy are enabling deep molecular characterization of cancer evolution and molecular response dynamics via circulating tumor DNA (ctDNA) measurements in clinical trials, whereas advances in radiomics have exponentially increased the information content in radiographic imaging data. Although linkage between changes in ctDNA or radiomic signatures and survival outcomes have been described 16 and progress has been made in applying pharmacometric methodologies to modeling tumor size with integration of ctDNA and radiomics‐based covariates, 17,18 disease progression modeling of longitudinal high‐dimensional ctDNA or radiomics profiles for elucidating POC and for identifying patient subgroups with differential treatment response remain largely untapped opportunities in drug development settings. In a study involving 466 patients with non‐small cell lung cancer from a clinical trial that evaluated atezolizumab‐based treatments, ML was applied to jointly model the dynamics of multiple ctDNA features as predictors of overall survival 19 .…”