In the field of cancer treatment, efforts to develop precision oncology treatments face challenges due to the vast array of cancer types, each with its own unique characteristics. Most of the focus has been on identifying mutations in cancer driver genes through genetic sequencing assays and matching treatments accordingly. However, many cancer patients do not respond to these targeted therapies effectively. A new study published in the journal Nature Cancer introduces a groundbreaking computational pipeline called PERCEPTION, which aims to predict patient response to cancer drugs at the single-cell level. This approach utilizes transcriptomics, the study of transcription factors that carry out DNA instructions, to better understand tumor complexity and evolution.
PERCEPTION, spearheaded by Sanju Sinha and colleagues, employs artificial intelligence and transfer learning to delve into single-cell omics data. By analyzing the clonal architecture of tumors and monitoring resistance emergence, this innovative approach offers valuable insights into cancer treatment strategies. The ability to adapt to cancer cell evolution and modify treatment plans accordingly is a significant advantage of PERCEPTION. Despite the challenges posed by limited single-cell data, the success of PERCEPTION in predicting responses to monotherapy and combination treatments in clinical trials for multiple myeloma, breast, and lung cancer underscores its potential as a valuable tool in precision oncology.
While PERCEPTION is not yet ready for clinical use, its success in predicting treatment responses in various cancer types demonstrates the promise of single-cell information in guiding therapy decisions. Sinha emphasizes the importance of generating more data to enhance the accuracy and reliability of predictions, ultimately leading to the development of a clinical tool that can systematically predict individual patient responses. As the quality and quantity of data play a crucial role in the effectiveness of this approach, efforts to encourage the adoption of technology in clinical settings are vital for further refinement and development.
The research team involved in the study, supported by the Intramural Research Program of the NIH, NCI, and various NIH grants, collaborated to validate the effectiveness of PERCEPTION in stratifying patients into responder and non-responder categories across different clinical trials. Notably, the ability of PERCEPTION to capture the development of drug resistance in lung cancer progression highlights its potential impact in predicting treatment outcomes. With a strong emphasis on generating more data and conducting additional studies, the researchers hope to accelerate the adoption of this technology in clinics to improve patient outcomes and drive advancements in precision oncology.
