Biology subjects

Masoumi, E.

Publications and source records attributed to Masoumi, E..

2 recordsLinked to original sources

bioRxiv2024

A new universal chimeric-antigen receptor (CAR)- fragment antibody binder (FAB) split system for cancer immunotherapy

Chimeric antigen receptor T (CAR-T) cell therapy has shown extraordinary results in treating hematological cancer. However, many patients relapse because of heterogeneous antigen expression and outgrowth of antigen lost variants. Other problems include on-target-off-tumor toxicity and the requirement for manufacturing of complex cellular products. Universal and modular CAR constructs offer significantly improved flexibility, safety and cost-effectiveness over conventional CAR constructs. Here we present a new chimeric-antigen receptor (CAR)-fragment antibody binder (Fab) platform based on an engineered protein G variant (GA1) and Fab scaffolds that present exquisite specificity and selectivity on antibody capture. The expression of GA1CAR on human CD8+T cells leads to antigen recognition and T cell effector function that can be modulated according to the affinity of the CAR for the Fab scaffold and of the Fab for the target. GA1CAR-T cells can recognize multiple Fab-antigen pairs on breast and ovarian cancer cell lines. Adoptive transfer of GA1CAR-T cells/Fabs in breast cancer xenograft models leads to effective tumor control. Rapid re-direction of the CAR-T cells to a new target can be achieved by using different Fabs. GA1CAR expression confers favorable phenotypic properties to T cells including a higher effector function upon exposure to antigen as compared to conventional scFv CAR-T cells. This highly versatile "plug and play" CAR-T platform has potential for application in personalized therapy, preventing antigen loss variant escape, decreasing toxicity and increasing access.

bioengineering↗

bioRxiv2019

A Machine Learning Approach to Predict Hypotensive Events in ICU Settings

PurposePredicting hypotension well in advance provides physicians with enough time to respond with proper therapeutic measures. However, the real-time prediction of hypotension with high positive predictive value (PPV) is a challenge due to the dynamic changes in patients physiological status under the drug administration which is limiting the amount of useful data available for the algorithm.\n\nMethodsTo mimic real-time monitoring, we developed a machine learning algorithm that uses most of the available data points from patients record to train and test the algorithm. The algorithm predicts hypotension up to 30 minutes in advance based on only 5 minutes of patients physiological history. A novel evaluation method is proposed to assess the algorithm performance as a function of time at every timestamp within 30 minutes prior to hypotension. This evaluation approach provides statistical tools to find the best possible prediction window.\n\nResultsDuring 181,000 minutes of monitoring of about 400 patients, the algorithm demonstrated 94% accuracy, 85% sensitivity and 96% specificity in predicting hypotension within 30 minutes of the events. A high PPV of 81% obtained and the algorithm predicted 80% of the events 25 minutes prior to their onsets. It was shown that choosing a classification threshold that maximizes the F1 score during the training phase contributes to a high PPV and sensitivity.\n\nConclusionThis study reveals the promising potential of the machine learning algorithms in real-time prediction of hypotensive events in ICU setting based on short-term physiological history.

bioengineering↗