MicroRNA: A Breakthrough for Early Multi-Cancer Detection and Localization

Dr Ben Miles

Summary:

MicroRNAs are emerging as a pivotal tool in the early detection of cancer, offering a potential breakthrough to save millions of lives.

• The Problem: Traditional cancer detection methods (symptom-based, imaging, tumor markers, circulating tumor DNA) often catch cancer too late, resulting in low survival rates. Early detection is critical for better outcomes.
• The Discovery: MicroRNAs are tiny RNA molecules that regulate gene expression and cellular processes. Cancer cells release altered microRNAs, and these signals are amplified throughout the body via exosomes, making them a strong indicator of early-stage cancer.
• The Breakthrough Technology: Xgenera has developed a novel blood test, mIONCO-Dx, that uses a modified qPCR technique to detect these minute microRNA changes. They extended the microRNA length to make it compatible with qPCR, enabling scalable and affordable testing.
• AI for Precision: Machine learning models (Support Vector Machine and Neural Network) are then applied to the detected microRNA patterns to not only identify the presence of cancer with high accuracy but also pinpoint its exact location in the body.
• Promising Results & Impact: Initial case-control studies show over 99% sensitivity and specificity, and 96% accuracy in identifying tumor origin. This technology promises less invasive treatments, reduced healthcare burdens, and significant cost savings, with trials now moving into real-world asymptomatic populations.

The Critical Need for Early Cancer Detection [0:00:00]

MicroRNA is presented as a potentially crucial tool in the fight against cancer, capable of acting as an early alarm system due to its ability to signal cellular changes.

• The Importance of Timing:
  • Late detection of cancer leads to survival odds under 10%.

    

  • Early detection dramatically increases survival rates to over 90%.

    

• Introducing MicroRNA as a Cancer Indicator:
  • MicroRNAs are small molecules (about 100 times the size of a single atom) detected in the blood.
  • Their altered presence can indicate cancer, attracting significant media attention (e.g., "AI blood test to speed up bowel cancer diagnosis").

    

• The Vision: To develop a single, comprehensive test that not only identifies the presence of a problem but also its specific location in the body.
  • This aspiration drives the focus on microRNA research and development.

    

  • The video introduces Andy, a scientist working on this "impossible" goal, whom the speaker (Dr. Ben Miles) met in 2024.

    

Evolution of Cancer Detection Methods [0:01:56]

Cancer remains challenging to detect early, despite being one of the most studied diseases, often growing for years without symptoms.

• The Persistent Challenge of Early Detection:
  • News headlines emphasize rising cancer rates, especially among younger populations, and increasing diagnoses.

    

    

  • Cancer still claims nearly 10 million lives annually, making early detection critical.

    

• Historical Detection Methods:
  • Symptom-based diagnosis [0:02:30]:
    • Cancer was typically found by accident or far too late (e.g., noticeable lumps or persistent pain).
    • This approach was primitive and reactive.

      

      

  • Imaging technologies [0:02:34]:
    • X-rays (1890s), CT scans (1970s), and MRI (1980s) allowed visualization of tumors before symptoms.
    • Limitation: Required prior knowledge of where to look, making widespread early screening difficult.

      

      

  • Biopsies and Tumor Markers [0:02:51]:
    • More specific tests for markers like PSA (prostate cancer) and CA125 (ovarian cancer) were developed.
    • Limitation: Signals could be ambiguous and often only appeared strongly in advanced stages.

      

• The Promise and Pitfalls of Circulating Tumor DNA (ctDNA) [0:03:07]:
  • Genetic Signatures: Recent decades shifted focus to genetic signatures, leading to the discovery of ctDNA – tiny DNA fragments shed by tumors into the bloodstream.

    

  • Advancements: ctDNA tests (e.g., Grail's Galleri Test) promised multi-cancer detection and tumor origin prediction from a single blood sample.

    

  • Challenges:
    • Not all mutations in ctDNA are cancer-specific.
    • Signal variability between different cancers.
    • High cost and complexity limit scalability for routine population screening.
    • Early detection sensitivity can be as low as 16.8% [0:04:32], meaning the signal is weakest when most critical.

      

      

MicroRNA: A New Frontier in Early Detection [0:05:46]

The breakthrough in early detection lies in understanding and leveraging microRNAs.

• Discovery of MicroRNA [0:05:46]:
  • In the 1990s, Victor Ambrose and Gary Ruvkun discovered microRNAs while studying the roundworm C. elegans.

    

  • MicroRNAs are very small RNAs (about 20 bases long).
  • They bind to messenger RNA (mRNA) to regulate gene expression, thus having significant downstream effects on cell function.

    

    

• Role in Cell Regulation and Cancer [0:06:35]:
  • MicroRNAs regulate key cellular processes like growth, division, and death.
  • Cancer cells exhibit altered production of various microRNAs, reflecting a breakdown in normal cellular processes.
  • A key insight: simply detecting altered microRNA levels from the tumor might be too weak a signal, similar to ctDNA.
• The Endocrine Effect and Exosomes [0:06:59]:
  • Cells constantly release tiny packages called exosomes, filled with microRNAs and other biological cargo.

    

  • Exosomes enable cell-to-cell communication.
  • When cells turn cancerous, the mix of microRNA they release changes.
  • These altered signals affect surrounding healthy cells, which, in turn, change their own microRNA outputs.

    

  • Crucially, these microRNAs enter the bloodstream, traveling to distant tissues (liver, bone marrow, and blood cells), which also respond by altering their microRNA outputs.

    

  • This systemic change means the entire body effectively "screams" that something is wrong, providing an amplified signal for detection, even when the primary tumor is small.

    

    

Xgenera's Breakthrough: Scalable MicroRNA Detection [0:08:24]

The challenge shifted from identifying a strong signal to making its detection scalable, affordable, and accessible.

• The Problem with Existing MicroRNA Detection [0:08:14]:
  • Despite their abundance compared to ctDNA, microRNAs are present in low concentrations relative to other blood components, making them difficult to detect.
  • Traditional lab detection methods are too slow, complex, and costly for widespread clinical use.
• Andy's Vision: Democratizing Early Detection [0:08:31]:
  • Andy (from Xgenera) aimed to create a test that is cheap, scalable, and easily accessible for routine population screening (e.g., annual or biennial tests for individuals aged 50-79).
• Leveraging qPCR Technology [0:10:12]:
  • Quantitative Polymerase Chain Reaction (qPCR): This technique, made widely available post-COVID, can amplify and quantify even minute amounts of specific DNA sequences. It is considered a diagnostic "gold standard."

    

    

    

  • The qPCR Process:
    1. Isolation [0:10:47]: Genetic material (RNA) is isolated from the blood sample.
    2. Reverse Transcription [0:10:51]: RNA is converted into a complementary DNA (cDNA) sequence using reverse transcriptase.
    3. Cyclical Amplification (30-40 cycles) [0:11:01]:
       • Denaturation: Heating separates the DNA strands.
       • Annealing: Short DNA primers (designed for the target sequence) bind to the separated strands.
       • Extension: A DNA polymerase enzyme builds new DNA strands from the primers, using free nucleotides and incorporating a fluorescent tag.

         

    4. Detection and Quantification [0:11:28]: The fluorescent signal, increasing with each cycle, allows for counting and quantifying the amplified DNA copies, which can number in the billions.
• The "Secret Sauce": Overcoming MicroRNA's Size Limitation [0:11:43]:
  • The Fundamental Problem: MicroRNAs are only about 20 base pairs long, which is too short for standard qPCR, as the primers and probes themselves require similar lengths, making effective amplification historically impossible.

    

  • Xgenera's Innovation: Xgenera's breakthrough involves a proprietary "target-specific extension" protocol.
    • This process extends the microRNA sequence.
    • It introduces site-specific adaptations that enable clear distinction between target and non-target sequences, making qPCR viable for microRNAs.

      

      

AI and Machine Learning for Cancer Prediction and Localization [0:12:30]

With a method to detect microRNA, the next step was to identify which microRNAs were most predictive of cancer and to pinpoint their origin.

• Data Training and Feature Selection [0:12:42]:
  • Xgenera's test was trained on data from approximately 21,000 patients, covering 12 different cancer types and non-cancerous conditions.
  • The data was divided into training, test, and validation cohorts.
  • Feature selection algorithms identified the 50 most predictive microRNAs.

    

• Visualizing Data with Principal Component Analysis (PCA) [0:13:06]:
  • PCA was used to simplify the multi-dimensional data (50 microRNA dimensions) into two dimensions for visualization.
  • A scatter plot showed a clear, though not perfectly distinct, separation between cancer (red) and non-cancer (blue) samples.

    

  • This visualization confirmed the potential for differentiation but also highlighted that using all 50 dimensions would improve accuracy.
• Advanced Machine Learning Models [0:14:04]:
  • To maximize the predictive power of the 50 microRNA dimensions, Xgenera developed two machine learning models:
    1. Support Vector Machine (SVM): Designed to classify whether cancer is present (yes/no).
    2. Neural Network: Used to identify the specific location of the cancer in the body if it is detected.

       

       

  • Example: A demonstration showed a neural network effectively separating healthy patients from those with lung cancer and esophageal cancer using only 8 microRNA targets, indicating strong potential for precise localization.

    

Promising Results and Real-World Impact [0:14:54]

The robust performance in clinical trials points to a significant shift in cancer diagnostics and treatment.

• Astounding Performance in Case-Control Studies [0:14:54]:
  • The mIONCO-Dx test achieved >99% sensitivity and specificity across all stages for predicting cancer presence (yes/no) for 12 different cancers.
  • It demonstrated 96% accuracy in identifying the tumor's site of origin.

    

  • The next critical phase is testing in asymptomatic populations with long-term follow-up.

    

• Transition to Real-World Application [0:15:26]:
  • The NHS is launching a new trial for Xgenera's blood test, focusing on bowel cancer and 11 other forms of the disease.
  • The UK Health Secretary has committed £2.4 million in funding, underscoring confidence in the technology.

    

• Broader Societal and Economic Benefits [0:15:41]:
  • Improved Patient Outcomes: Early detection enables less invasive and more effective treatments.
  • Reduced Healthcare Burden: Less advanced disease means simpler, less resource-intensive care.
  • Significant Cost Savings: Widespread adoption of this test could save an estimated $11 billion annually in the US by reducing late-stage treatments, unnecessary follow-up tests, and false positives from current screening methods.
• Commitment to Transparency and Impact [0:16:07]:
  • Xgenera is dedicated to developing the test with utmost transparency, adhering to stringent regulatory standards (MHRA, FDA, EMA).
  • The driving motivation for the team, including founder Andy, is the genuine positive impact the technology can have on people's lives.

Conclusion: A New Era in Cancer Diagnostics [0:16:56]

This development signifies a shift towards scientists taking discoveries from the lab directly into real-world applications.

• The video emphasizes the importance of looking beyond superficial headlines to understand the deep scientific work behind breakthroughs.
• Andy and his team at Xgenera embody this new generation of scientists who are driving their discoveries forward to create tangible impact.