Cancer Stem Cells Responsible for Tumour Recurrence

Why in News?

A new artificial intelligence (AI) framework that identifies 3 distinct developmental states of cancer stem-like cells can help identify hidden cancer stem cells from thousands of patient samples.

• Cancer stem-like cells – These are a small and specialised population of cells within tumours.
• They possess stem-cell-like properties, including the ability to self-renew and generate different tumour cell types.
• They can contribute to
  • Tumour recurrence
  • Metastasis
  • Therapy resistance
  • Tumour progression
  • Treatment failure
  • Their rarity and ability to change their cellular identity make them difficult to detect using conventional approaches.
• ACSCeND - ACSCeND stands for AI-based Cancer Stem-like Cell Profiler and Neoplasm Deconvoluter.
• It is an AI-based framework designed to identify and quantify different cancer stem-like cell states within tumours.
• Key Features - Unlike conventional methods that assign a tumour a single “stemness” score, ACSCeND identifies three distinct CSC developmental states.

• Working - Single-cell RNA sequencing → Stemness-state learning → Deep learning → Bulk tumour RNA sequencing → CSC-state deconvolution → Clinical prediction
• It learns cellular characteristics from high-resolution single-cell RNA sequencing.
• It combines this information with deep learning.
• It can then analyse conventional bulk tumour RNA sequencing data.
• This allows researchers to estimate hidden CSC populations even in large patient datasets where single-cell sequencing is unavailable.
• Deconvolution - Deconvolution refers to computationally separating or estimating the different cell populations present within a mixed bulk tumour sample.
• Role of OncoMark - OncoMark was an earlier AI platform developed by the researchers.
• It was designed to identify biological hallmarks associated with cancer progression from large genomic datasets.
• It demonstrated the ability of AI to uncover complex biological patterns that are difficult to detect manually.
• ACSCeND builds upon this AI-based approach to specifically investigate cancer stem-like cells.

• Findings of the study - Researchers applied ACSCeND to more than 25,000 tumour samples from major international cancer databases, including:
  • TCGA – The Cancer Genome Atlas
  • PRECOG
• The analysis found that tumours with a higher abundance of highly potent, pluripotent-like CSCs were associated with:
  • Poorer patient survival
  • Higher probability of tumour recurrence
  • Reduced response to immunotherapy
• The framework also identified molecular programmes that may help CSCs survive, adapt and evade the immune system.

Reference

DST | AI uncovers hidden cancer stem cells