Lennart Kester

Molecular pediatric cancer classification 

 Molecular diagnostics are becoming an increasingly important aspect of the diagnostic process. We have recently developed M&M, a pan-cancer pediatric classification algorithm that uses RNA sequencing data to diagnose pediatric cancers. With RNA sequencing we measure the gene expression profile of a sample and what we have learned is that these gene expression profiles are highly tumor-type specific. We exploit this by comparing every new patients’ gene expression profile to all gene expression profiles that we measured in the past and based on this comparison predict the diagnosis for every new patient. This technique has already been implemented in the routine diagnostic process of the Máxima Center. 

 The next step is to dramatically reduce the time it takes to acquire these molecular diagnoses. In collaboration with the group of Prof. dr. Jeroen de Ridder in the University Medical Center Utrecht we’re utilizing novel sequencing techniques that allow data generation within an hour. The prime example of this is the development of an algorithm based on nanopore sequencing data can classify brain tumors within 90 minutes from acquisition of the tumor material (Vermeulen et al., Nature 2023). This means that we can acquire a molecular diagnosis for brain tumors while a patient is in surgery, allowing the surgeon to adapt the surgical strategy based on the acquired molecular diagnosis. We’re now working on expanding this beyond brain tumors, such that in the future all patients can benefit from this technology. 

'We use state-of-the-art techniques to achieve faster and more accurate diagnoses for children with cancer' Dr. Lennart Kester - Associate group leader

 Improving patient stratification for optimal treatment 

Upon diagnosis of the tumor, a patient starts treatment. However, we know that for every tumor type and treatment plan, some patients get cured, while others eventually relapse. In the Kester group we’re trying to use molecular data to predict whether patients have a good or a poor prognosis. For instance, in Osteosarcoma we have validated certain gene expression profiles that predict a high probability of relapse. In the future this might be used to better guide treatment and adapt treatment for patients with a high probability of relapse. 

 We’re now exploring similar approaches in low grade gliomas, where we’re specifically interested in the influence of the immune cells in the tumor on the chance of relapse or progression. In a different project we’re developing patient specific molecular assays that allow us to track therapy response in patients with acute myeloid leukemia. These assays are highly specific for the patients’ leukemia amd allow us to accurately follow the disease load in these patients. We hope that in the future approaches like this will help in selecting the optimal treatment for each individual patient.