TheraCluster: personalized medicine for brain metastasis patients

Cancer is a leading cause of global mortality, both in wealthy countries and increasingly in developing nations. Brain metastases develop in approximately 20% of all cancer patients, significantly contributing to overall cancer mortality, as the mean survival rate remains below 5% after the first year. Importantly, highly prevalent primary tumors, such as lung and breast cancers, as well as melanoma, frequently disseminate to the brain. Conventional oncologic treatments are mainly designed to target features of the primary tumor, often disregarding their effectiveness against brain metastatic cells. Moreover, patients with active central nervous system (CNS) progression have historically been underrepresented in clinical trials, posing a persistent challenge in the field. Our team has developed TheraCluster, an in-silico prediction model that analyzes patient- derived brain metastases, classifying them into three different genetic fingerprints. Each genetic signature corresponds to a specific therapeutic vulnerability identified through PANDRUGS. We have found precise biomarkers that define each therapeutic cluster and facilitate drug-response prediction by determining the most effective pharmacological agents for each group. By integrating TheraCluster with our clinically compatible drug-screening platform, we are validating pharmacological susceptibility predictions in live human brain metastases. This innovative approach has the potential to revolutionize the diagnosis, prognosis and treatment of cancer patients with advanced-stage disease that has spread to the brain.

Brain metastasis occurs when cancer spreads to the brain, affecting around 20% of oncological patients and leading to poor survival rates. Common cancers, such as lung and breast cancer or melanoma, frequently disseminate to the brain, but current treatments mainly target the original tumor rather than advanced stages. Additionally, patients with brain metastasis are often excluded from clinical trials, limiting their treatment options. TheraCluster is a groundbreaking approach that identifies three distinct genetic "fingerprints” in brain metastasis patients, each linked to specific therapeutic vulnerabilities. This allows us to predict which therapies will be most effective for each patient. By detecting the presence of specific biomarkers in patients, we can match each scenario with the most successful treatments. In conclusion, TheraCluster paves the way for more precise and personalized medicine to improve the outcome of patients with brain metastasis.

The incidence of brain metastases is rising, yet therapies remain ineffective, failing to improve survival. Standard treatments—surgery, radiotherapy, and systemic therapies—rarely cure, while chemotherapy is hindered by poor blood-brain barrier penetration. Targeted therapies for NSCLC, melanoma, and breast cancer improve survival but offer limited intracranial control. Clinical trials often exclude patients with active brain disease, preventing drug efficacy assessment. Our group has consistently highlighted the importance of patient-derived samples in tackling this need. We have now integrated omics analysis of human brain metastases with our drug-screening platform, enabling precise drug-response predictions. While other molecular tools exist in oncology, brain metastasis research remains largely preclinical, with limited real-world use. This innovation enhances patient stratification, optimizes trial recruitment, and accelerates therapy identification, driving clinical impact.