Artificial intelligence will be used to try and figure out which cancer patients to treat with immunotherapies and potentially help crack the mystery of why only 20 per cent of patients currently respond.
The research project entitled Assessment of predictors of response to checkpoint blockade has the potential to help not only patients with melanoma but several other forms of cancer including lung cancer and head and neck cancer.
After getting training on high-dimensional data integration from overseas, project lead Dr Ahmed Mehdi realised that he could use his knowledge and expertise in proteomics to help clinicians and oncologists better determine who will respond to immunotherapies.
Dr Ahmed Mehdi
"In late 2019, before coronavirus started, I was sent to UK Kings College in London to get training on developing automated approaches on multiparameter flow cytometry. This project is using a technology which is called Immunophenotyping using flow cytometry," he said.
"After coming back to Australia from this training, I developed more connections and developed a proposal to predict patient-specific immunotherapy response using artificial intelligence approaches.
"We developed a world-class team consisting of clinicians, immunologists and biostatisticians who are contributing to this research. Our team includes Prof Euan Walpole, Prof Ian Frazer, Prof Ken O'Byrne, Dr Janin Chandra, Dr Elizabeth McCaffrey, Dr Rahul Ladwa, A/Prof Emma Hamilton-Williams, A/Prof Victoria Atkinson, and Dr Tasafin Hossain.
Made possible by the supporters of the PA Research Foundation and based at the PA Hospital, donated funds will be used to drive the project across several areas including the use and development of computer algorithms using computer programming.
Dr Ahmed Mehdi alongside colleague Prof Ian Frazer
"The funds are being used for taking blood from the patients and processing it at TRI (Translational Research Institute) in Professor Ian Frazer's lab and experimenting on it. Using flow cytometry, we can look into the cells, each individual cell of the patient so that we can see what are the markers that are present in those cells," Dr Mehdi said.
"From those markers, we are designing algorithms and computer programs, so funds are also going towards how the blood is being taken from the patients, going to the lab, and people working in the lab to process that blood, and people working on algorithms to design computer projects to help us advance the research.
"We have the expertise of using high performance computers at UQ and TRI, this is important because this is a huge database because we expect to get over 10,000 cells per patient and for each cell the expression of several proteins will be measured. In the future, we will have up to 150 patients, there are so many data points and for each data point, there are so many biomarkers. So, if you count the number of data points, it's increasing exponentially, so we need high performance computing, efficient programming when we are processing the data."
"The research will examine blood samples from recruited patients at two time points with an initial aim of developing a defined way of informing clinicians which patients will respond to immunotherapy to save them having to have immunotherapy and not responding, with a second more long term aim of trying to identify why some patients fail to respond to immunotherapy and why.
"In this project, we have asked for two time points, the first time point is before the immunotherapy is given, and the second time point is when the therapy is given and after six weeks, this is where we are trying to answer the questions of why those who have responded what is happening with their immune system, how is their immune system behaving, and those who didn't respond what happened to the immune system, we are interested in knowing why is it some patients are not responding," he said.
"This is the first year of the project, we were more interested to tell the clinicians, who is going to respond at this moment."
Dr Mehdi said he hopes his research will provide cancer patients with a belief that no matter their diagnosis, that their cancer can be targeted and managed with the right treatment for their circumstances.
"If someone has cancer, it's very painful and stressful for the patient and we want to increase hope for the patient, that if they have any type of cancer, it is number one predictable and it is curable.
"We want to clearly tell the patient what the success rate for the specific treatment is for them, and for that treatment we know who is going to respond. We want to tell the patient that according to our computer programs you are likely to respond to this treatment.
"We want to also give some predictions to those patients who aren't going to respond, so they know that they're not going to respond, and the clinicians will use an alternative approach. But the outcome of the project would hopefully also tell, why some patients didn't respond."
Dr Mehdi said he sees the PA Research Foundation's donors as vital to creating better outcomes for people with cancer.
"They are doing great work, without their generosity the scientists cannot work without their contribution, they are the first people on the front line to work on generating the money we need to do our work.
"They have the confidence in the Foundation and the scientists that their money is being used to help humanity. What we are doing it is not specific to Brisbane, it will of course be spread worldwide, they will see how the people around the world are benefiting from their contribution."
Donate to this important research here.