2023-04-28 02:17:06
28.04.2023
Since going online in 2016, the Heidelberg Brain Tumor Classifier AI has analyzed molecular data from more than 100,000 brain tumors. A current study now proves the benefit of the method for cancer diagnosis developed at the Hopp Children’s Cancer Center in Heidelberg, the German Cancer Research Center and the Heidelberg University Hospital.
© Deminos – Fotolia.com
Through the AI-supported analysis of so-called methylation patterns, brain tumors in children and adolescents can not only be classified more precisely, but also more reliably in certain tumor groups than with the microscope. Patients with rare types of tumors and particularly aggressive brain tumors can particularly benefit from this.
The “Hopp Children’s Cancer Center Heidelberg” (KiTZ) is a joint institution of the German Cancer Research Center (DKFZ), the Heidelberg University Hospital (UKHD) and the University of Heidelberg (Uni HD).
In children, more subtypes in tumors of the central nervous system
Cancer in children and adolescents is different from cancer in adults, which is also reflected in its diversity. There are more than 150 subtypes of tumors of the central nervous system in childhood – many times more than in adults. “Depending on the tumor class, radiotherapy and chemotherapy also have very different effects. Classifying the tumors as precisely as possible is therefore crucial for effective treatment,” emphasizes the leader of the study, David Jones, head of department at the Hopp Children’s Cancer Center Heidelberg (KiTZ) and the German Cancer Research Center (DKFZ).
For a long time, looking through the microscope was crucial for diagnosing cancer. According to the classification of the World Health Organization (WHO), most brain tumors were until recently classified into tumor groups mainly based on their tissue properties. “This expert knowledge is still indispensable for making a diagnosis. However, it is not possible to classify all tumor types more precisely based on their tissue structure alone. In addition, some tumor types are so rare that even experienced pathologists hardly ever see them,” says Jones.
Refined diagnosis and more personalized treatment using AI
In 2018, a research team led by Stefan Pfister, director at the KiTZ, head of department at the DKFZ and pediatric oncologist at the Heidelberg University Hospital (UKHD), in close cooperation with the Department of Neuropathology at the UKHD with Andreas von Deimling and Felix Sahm, developed a new AI-based method in the published in the journal Nature and made available worldwide. The algorithm known as the “Heidelberg Brain Tumor Classifier” evaluates so-called DNA methylation in the genome of the tumor. The complex pattern of methyl markings that our DNA is provided with forms a second level of information – in addition to the genetic information that is defined in the base sequence of the DNA. The methylation marks genes and the cell can use them to control their activity. A large number of studies have already shown that not only cancer cells and healthy cells differ in their methylation pattern, but also different types of tumors.
The task of the AI is to use the methylation data to identify a unique fingerprint for each tumor group in order to refine the diagnosis. Since it first went online, the Brain Tumor Classifier has evaluated more than 100,000 tumor samples that were uploaded to the www.molecularneuropathology.org platform for research purposes worldwide.
The current study comes to the conclusion that the method significantly improves the accuracy of the previously established diagnostic methods and thus enables even more personalized treatment.
It is the first study that was able to verify the reliability of methylation profiles for the diagnosis of cancer in children by observing the course of the disease over a longer period of time. For 1,200 newly diagnosed brain tumors in children and adolescents, the research team compared the diagnosis made using previous WHO criteria with the result of the AI. In half of the patients, the diagnosis basically corresponded to the original WHO classification, but the AI analysis enabled a more precise classification of the tumor into certain subgroups. “Some of the identified methylation patterns are so specific that the AI can even use them to make statements about the estimated age and sex of the child and the location of the tumor,” explains Dominik Sturm, lead author of the study and pediatrician at KiTZ and UKHD.
Genetic cancer risk in families can be determined
In addition to the methylation data, the research team, in close cooperation with the Human Genetics department at the UKHD, also analyzed the genetic information of certain cancer-associated genes in order to refine the diagnosis. In almost 50% of those affected, the scientists came across genetic changes that are crucial for the diagnosis or can be used therapeutically. In 10% of cases, the research team also discovered a hereditary cancer risk. “Recognizing hereditary causes of cancer when making the diagnosis can help to make the right therapy decision for cancer treatment,” explains Sturm. “Affected families can get advice on their genetic cancer risk and, for example, certain check-ups for siblings and other affected family members.”
There were deviations from the diagnosis made according to WHO criteria, particularly in young patients with high-grade gliomas – brain tumors that grow particularly aggressively. In about 15% of these patients, the AI found that they were not high-risk tumors, but lower-grade gliomas with a significantly better prognosis. In fact, follow-up over several years confirmed that these patients had a significantly better disease course and better survival than would have been the case with high-grade glioma. “This patient group in particular could therefore benefit particularly from the new procedure,” says David Jones.
Help for countries with few experts
A new edition of the WHO classification of tumors of the central nervous system was recently published, which was developed jointly by scientists at the KiTZ, the UKHD and the DKFZ and numerous other international experts. For the first time, it is based on a modern, multi-layered approach in which methylation patterns are now also firmly anchored. “Our study shows that the combination with AI-supported methods can significantly improve precision diagnostics for children and adolescents with brain tumors,” says Stefan Pfister. “Even in countries where there are often too few specialized pathologists to evaluate tumor samples, these methods could help to use more precise diagnostic approaches as standard, especially for children with cancer. We are just starting to explore this with partner institutions in Africa and Asia.”
______________
dr Sibylle Kohlstädt, Press and Public Relations, German Cancer Research Center
______________
Sources: idw-online.de, Nature Medicine, MolecularNeuropathology.org
#AIsupported #cancer #diagnosis #children #adolescents #www.kinderaerzteimnetz.de