Cellworks Biosimulation Platform Predicts Chemotherapy Benefit in NSCLC Patients Beyond PD-L1 Status
核心洞察
Cellworks (搜索)' biosimulation platform accurately predicted overall survival and chemotherapy benefit in two real-world NSCLC (搜索) patient cohorts, advancing personalized cancer treatment beyond traditional PD-L1 (搜索) biomarkers.
The Therapy Response Index (TRI) was significantly associated with overall survival (p = 0.036), with patients in the low TRI group receiving approximately 3 months additional median survival benefit from chemotherapy plus immunotherapy.
The study validated the platform's ability to identify NSCLC (搜索) patients most likely to benefit from combining immunotherapy and chemotherapy, regardless of PD-L1 (搜索) levels, potentially refining treatment strategies for patients without targetable mutations.
Cellworks (搜索) Group Inc. has demonstrated that its biosimulation platform can accurately predict overall survival and chemotherapy benefit in non-small cell lung cancer (NSCLC (搜索)) patients, offering a new approach to treatment selection that goes beyond traditional PD-L1 (搜索) status assessment. The study results were presented at the IASLC 2024 World Conference on Lung Cancer in San Diego, showcasing the potential to optimize treatment strategies for patients with advanced NSCLC.
Addressing Current Treatment Selection Limitations
The majority of patients with advanced NSCLC (搜索) do not have targetable mutations and rely largely on PD-L1 (搜索) status to guide treatment decisions. However, no reliable biomarkers currently exist to identify patients who would benefit from the addition of chemotherapy to immunotherapy regimens.
"A majority of patients with advanced NSCLC (搜索) do not have a targetable mutation and rely largely on PD-L1 (搜索) status to guide treatment decisions – however, no reliable biomarkers currently exist to identify patients who would benefit from the addition of chemotherapy," said Dr. Charu Aggarwal, Professor of Medicine at the University of Pennsylvania and Principal Investigator for the study.
Study Design and Methodology
The research employed a two-phase approach using real-world patient data. The Cellworks (搜索) Therapy Response Index (TRI) algorithm was initially trained on a retrospective cohort of 553 NSCLC (搜索) patients from the U.S. Veterans Health Administration who had undergone tissue-based comprehensive genomic profiling conducted by Foundation Medicine.
The TRI, Cellworks (搜索) Computational Biology Model (CBM), and clinical thresholds were then locked and prospectively validated in an independent set of 710 advanced NSCLC (搜索) front-line patients obtained from the Flatiron Health (搜索)-Foundation Medicine NSCLC clinico-genomic database.
The biosimulation was performed using the Cellworks (搜索) CBM, which predicts changes in key biomarker concentrations and hallmark cancer phenotypes triggered by genomic aberrations and therapeutic intervention. The hallmark phenotypes were combined using statistical and machine learning methods to generate the TRI, which was trained to predict outcomes.
Significant Clinical Outcomes
The study demonstrated that the Cellworks (搜索) TRI was significantly associated with overall survival above and beyond other clinical factors, including PD-L1 (搜索) levels (p = 0.036). The results revealed distinct treatment benefits based on TRI scores:
Patients in the high TRI group (≥ 32) showed no benefit from the addition of chemotherapy to their treatment regimen. In contrast, patients in the low TRI group (< 32) received an estimated incremental benefit in median overall survival of approximately 3 months with the addition of chemotherapy and immunotherapy.
Technology Platform Details
The Cellworks (搜索) Platform utilizes multi-omic data from individual patients or cohorts as input to generate personalized or cohort-specific disease models. The CBM represents a highly curated mechanistic network of over 6,000 human genes, 30,000 molecular species, and 600,000 molecular interactions.
This comprehensive model, along with associated drug models, biosimulates the impact of specific compounds or combinations of drugs on patients and produces therapy response predictions. These predictions are statistically modeled to produce a qualitative TRI score, scaled from 0 (unfavorable outcome) to 100 (favorable outcome) for specific therapies.
Clinical Implications
"While PD-L1 (搜索) has been the standard biomarker for selecting NSCLC (搜索) patients for immunotherapy, it is not predictive of chemotherapy benefit," said Dr. Michael Castro, Cellworks (搜索) Chief Medical Officer. "This study shows that by using a tumor's genomic profile and Cellworks personalized therapy biosimulation to understand signaling pathway dysregulation and drug response, we can better predict which NSCLC patients may benefit from chemotherapy and which patients will not."
The research suggests that biosimulation has the potential to refine patient management for NSCLC (搜索) patients above and beyond generalized guidelines using PD-L1 (搜索) only, advancing the precision oncology mission of delivering the right treatment to the right patient.
Platform Validation
The Cellworks (搜索) CBM has been tested and applied against various clinical datasets, with results provided in over 125 presentations and publications with global collaborators. The company, backed by Artiman Ventures, Bering Capital, Sequoia Capital, UnitedHealth Group, and Agilent Ventures, operates from South San Francisco, California, with a CLIA-certified computational laboratory in Franklin, Tennessee, and a research and development facility in Bangalore, India.