Using adaptive genetic algorithms combined with high sensitivity single cell-based technology to detect bladder cancer in urine and provide a potential noninvasive marker for response to anti-PD1 immunotherapy
Gardado en:
| Autores: | , , , , , , |
|---|---|
| Formato: | artículo original |
| Data de Publicación: | 2020 |
| Descripción: | Objective: To use adaptive genetic algorithms (AGA) in combination with single-cell flow cytometry technology to develop a noninva- sive test to detect bladder cancer. Materials and Methods: Fifty high grade, cystoscopy confirmed, superficial bladder cancer patients, and 15 healthy donor early morn- ing urine samples were collected in an optimized urine collection media. These samples were then used to develop an assay to distinguish healthy from cancer patients’ urine using AGA in combination with single-cell flow cytometry technology. Cell recovery and test perfor- mance were verified based on cystoscopy and histology for both bladder cancer determination and PD-L1 status. Results: Bladder cancer patients had a significantly higher percentage of white blood cells with substantial PD-L1 expression (P< 0.0001), significantly increased post-G1 epithelial cells (P < 0.005) and a significantly higher DNA index above 1.05 (P < 0.05). AGA allowed parameter optimization to differentiate normal from malignant cells with high accuracy. The resulting prediction model showed 98% sensitivity and 87% specificity with a high area under the ROC value (90%). Conclusions: Using single-cell technology and machine learning; we developed a new assay to distinguish bladder cancer from healthy patients. Future studies are planned to validate this assay. 2019 Elsevier Inc. All rights reserved. |
| País: | Kérwá |
| Institución: | Universidad de Costa Rica |
| Repositorio: | Kérwá |
| Idioma: | Inglés |
| OAI Identifier: | oai:kerwa.ucr.ac.cr:10669/102779 |
| Acceso en liña: | https://hdl.handle.net/10669/102779 https://doi.org/10.1016/j.urolonc.2019.08.019 |
| Palabra crave: | Bladder cancer PD-L1 status Assay Diagnosis Machine learning Single-cell Technology |