Below the Belt Award
Mevan Ekanayake — 2026
PETomics for Precision Theranostics: Multimodal Deep Learning to Personalise Treatment Selection in Metastatic Prostate Cancer
Patients with prostate cancer that has spread to other organs (metastatic prostate cancer) have limited treatment options, and their disease often progresses rapidly, affecting quality of life. Each patient’s cancer is unique, making it difficult for clinicians to accurately predict whether a patient will respond successfully to a treatment. Radioligand therapy with Lutetium‑PSMA is an effective new treatment option for these patients. While one fifth of patients respond very well to Lutetium‑PSMA therapy, one fifth do not benefit from it, and the remaining patients benefit only for a short period of time. Developing strategies to better select patients for Lutetium‑PSMA therapy is therefore a critical unmet need.
The aim of this project is to help clinicians better decide which patients should receive Lutetium‑PSMA therapy. Currently, patients are recommended for Lutetium‑PSMA therapy based on limited characteristics derived from their medical scans, blood tests, and other clinical details. In this project, we will integrate all these types of information and analyse them together using artificial intelligence (AI) methods. AI methods can be developed to combine and learn from different types of information and identify patterns that are too complex for humans to see. We aim to better understand which combinations of cancer characteristics are most specifically associated with response to Lutetium‑PSMA therapy by comparing with standard chemotherapy with cabazitaxel. This will help identify which patients are most likely to benefit specifically from Lutetium‑PSMA therapy.
Regarding selecting participants for this research, we will use existing data from 200 men who took part in ANZUP’s landmark TheraP clinical trial. In this trial, half of the men received Lutetium‑PSMA, while the other half were treated with standard chemotherapy. The information available from the trial includes medical scans (such as PET/CT, SPECT/CT scans), blood test results (including cancer DNA found in the blood), and detailed clinical information.
This project is designed to benefit people with metastatic prostate cancer. Current clinical guidelines to predict whether Lutetium‑PSMA therapy will work well for an individual patient are based on what can be observed on a patient’s scans and blood tests. By developing AI methods that can better identify Lutetium‑PSMA‑specific characteristics of each individual cancer, this project aims to support better‑informed and more personalised treatment decisions for patients.
People with lived experience of cancer have been involved in this project from the beginning through bi‑directional engagement with the research team. Our consumer representatives have actively contributed their views to help shape the research question. They will continue to be involved throughout the project via regularly scheduled meetings with the team every 2 months. From these interactions it is clear that from a patient’s perspective, a major concern is not knowing beforehand whether a treatment will provide lasting benefit. For patients with metastatic prostate cancer and their families, this uncertainty can be stressful and emotionally challenging. Our goal is to provide more personalised insights into whether Lutetium‑PSMA therapy is more likely to benefit a particular patient, by identifying and analysing the key characteristics of their cancer.
The aim of this project is to help clinicians better decide which patients should receive Lutetium‑PSMA therapy. Currently, patients are recommended for Lutetium‑PSMA therapy based on limited characteristics derived from their medical scans, blood tests, and other clinical details. In this project, we will integrate all these types of information and analyse them together using artificial intelligence (AI) methods. AI methods can be developed to combine and learn from different types of information and identify patterns that are too complex for humans to see. We aim to better understand which combinations of cancer characteristics are most specifically associated with response to Lutetium‑PSMA therapy by comparing with standard chemotherapy with cabazitaxel. This will help identify which patients are most likely to benefit specifically from Lutetium‑PSMA therapy.
Regarding selecting participants for this research, we will use existing data from 200 men who took part in ANZUP’s landmark TheraP clinical trial. In this trial, half of the men received Lutetium‑PSMA, while the other half were treated with standard chemotherapy. The information available from the trial includes medical scans (such as PET/CT, SPECT/CT scans), blood test results (including cancer DNA found in the blood), and detailed clinical information.
This project is designed to benefit people with metastatic prostate cancer. Current clinical guidelines to predict whether Lutetium‑PSMA therapy will work well for an individual patient are based on what can be observed on a patient’s scans and blood tests. By developing AI methods that can better identify Lutetium‑PSMA‑specific characteristics of each individual cancer, this project aims to support better‑informed and more personalised treatment decisions for patients.
People with lived experience of cancer have been involved in this project from the beginning through bi‑directional engagement with the research team. Our consumer representatives have actively contributed their views to help shape the research question. They will continue to be involved throughout the project via regularly scheduled meetings with the team every 2 months. From these interactions it is clear that from a patient’s perspective, a major concern is not knowing beforehand whether a treatment will provide lasting benefit. For patients with metastatic prostate cancer and their families, this uncertainty can be stressful and emotionally challenging. Our goal is to provide more personalised insights into whether Lutetium‑PSMA therapy is more likely to benefit a particular patient, by identifying and analysing the key characteristics of their cancer.
