Central nervous system tumors are one of the most lethal cancer types, especially among children. Every year, 150 children are diagnosed with a tumor in the brain or spinal cord in the Netherlands. For many, surgery is the first step taken in their treatment journey.
However, neurosurgeons often do not know the precise type and aggressiveness of the brain tumor they are contending with during surgery. This presents a considerable challenge to strike a balance between removing as much of the tumor as possible while minimizing neurological damage.
Tumor diagnosis typically takes one week after the operation following histological and molecular analysis.
Professor Dr. Eelco Hoving, a pediatric neurosurgeon and clinical director of neuro-oncology at Máxima Hospital, explains, "During surgery, there are instances where a small remnant of tumor tissue may intentionally be left behind to prevent neurological damage. However, if the diagnosis indicates that the tumor is highly aggressive, a secondary operation may still be necessary to remove that last remnant. This will again create risks and emotional stress for the patient and their family.”
Sturgeon
Researchers at UMC Utrecht have developed a new deep-learning algorithm, a form of artificial intelligence, called Sturgeon. Using ultra-fast nanopore sequencing to obtain a limited DNA methylation profile, Sturgeon then analyzes this information to identify the molecular subclassification of the brain tumor.
Dr. Jeroen de Ridder, group leader at UMC Utrecht and Oncode Institute researcher, says, “The Nanopore sequencing technology helps to read DNA in real time. Our algorithm has been trained on millions of simulated realistic ‘DNA snapshots’. Combined, these technologies enable us to determine the tumor type very quickly, within a turnaround time of less than 90 minutes during an operation. That is fast enough to directly adjust the surgical strategy, if necessary.”
Biobank Propels Innovation
Crucially, Sturgeon was trained using data from the extensive biobank maintained by the Princess Máxima Center, housing tissue samples from children with brain tumors.
"The ability to determine the brain tumor type during surgery underscores how technology accelerates diagnostics. It also demonstrates how we can leverage an existing biobank to propel novel technology forward," says Dr. Bastiaan Tops. Dr. Tops is head of the Pediatric Oncology Laboratory at the Princess Máxima Center and played a pivotal role in bridging the gap between surgical needs and emerging technology.
Niels Verburg, neurosurgeon at Amsterdam UMC - Cancer Center Amsterdam: "The possibility of a molecular diagnosis during surgery offers tremendous opportunities in adult neuro-oncology as well, which we are further delving into. We are already implementing the technique in our daily practice to significantly reduce the time patients wait for a diagnosis."
The “step into clinical practice”
While the technique has already shown remarkable results, further research is necessary to expand its scope and robustness. Additional tumor types will be integrated into the algorithm to meet international standards and allow data comparison. Comparative research between the new and traditional methods will also be conducted in partnership with other national and international centers to gauge the long-term impact on patients' quality of life.
“It is wonderful that we have been able to actually make the step into clinical practice by combining all areas of expertise, from basic researchers to pathologists and surgeons. By doing so, we can help surgeons to optimize the outcome of brain tumor surgery,” says Jeroen de Ridder.
For more information, contact Cancer Center Amsterdam, or read the scientific publication in Nature. https://www.nature.com/articles/s41586-023-06615-2
Researchers involved from Amsterdam UMC - Cancer Center Amsterdam:
Luuk Dankmeijer
This article was created for Cancer Center Amsterdam.