Two Australian studies published in Cell and Nature Communications have uncovered new pathways to precision cancer care.
Australian researchers have uncovered new clues to why cancers respond differently to treatment, with studies published in two of the world’s leading scientific journalshighlighting the growing role of tumour metabolism, artificial intelligence, and spatial biology in advancing precision oncology.
The research, led by two Australian teams supported by Cure Cancer at pivotal stages of their careers, provides fresh insight into cancer vulnerabilities and treatment resistance, while demonstrating how increasingly sophisticated technologies are reshaping cancer research and clinical decision-making.
The findings, published in Cell and Nature Communications, reflect a broader shift away from one-size-fits-all biomarkers towards integrated approaches that combine molecular profiling, tissue imaging, and computational analysis to better predict treatment response and identify new therapeutic targets.
In the Cell article, Melbourne-based researcher Dr Emily Gruber and collaborators reported that acute myeloid leukaemia (AML) cells were selectively dependent on de novo haem biosynthesis, revealing a previously underappreciated metabolic vulnerability in the disease.
The researchers found that inhibiting key enzymes involved in haem production could induce cuproptosis, a recently described copper-dependent form of programmed cell death.
The study also identified glycolysis-related pathways that may offer opportunities for combination treatment strategies.
The findings suggest that targeting haem metabolism could represent a novel therapeutic approach for AML, a blood cancer that remains associated with high rates of relapse and poor outcomes for many patients.
“In conclusion, our study identifies heme depletion as a cuproptosis trigger, strengthening the notion that it is a bona fide cell death pathway that may operate in a host of physiological and pathophysiological contexts,” they wrote.
“We also identify HBEs as promising drug targets in AML, with the availability of HBE crystal structures and tool compounds providing a starting point for drug development.”
Dr Gruber said early-career research support had been instrumental in enabling the work.
“Without this early support, we wouldn’t have generated such significant findings that have the real potential to improve outcomes for patients living with AML,” she said.
In a separate study published in Nature Communications, Brisbane-based researcher Associate Professor Arutha Kulasinghe and colleagues applied multiplexed immunofluorescence and deep-learning analysis to pre-treatment biopsies from patients with non-small cell lung cancer (NSCLC) receiving immune checkpoint inhibitors.
“Underlying mechanisms of resistance to immune checkpoint inhibitors (ICI) are poorly understood, despite representing most clinical outcomes for patients with non-small cell lung cancer (NSCLC) (70–80%),” the researchers wrote.
“Treatment resistance, either primary or acquired, to standard of care therapies poses a major barrier, making lung cancer the predominant cause of cancer related deaths worldwide.”
The researchers used advanced computational methods to classify cellular phenotypes, assess functional and metabolic states, and map spatial interactions between cells within the tumour microenvironment.
Their multivariate model predicted progression-free survival over 24 months with an area under the curve (AUC) of 0.8, highlighting the potential of AI-enabled spatial biology to improve patient stratification and treatment selection.
The study adds to growing evidence that the organisation and interactions of cells within tumours may provide clinically relevant information beyond conventional biomarkers.
“In this work, we sought to combine robust computational methods with meaningful statistical reasoning to understand the biologically relevant measurements from a custom-designed mIF panel,” the researchers wrote.
Related
“In doing so we develop an analytical pipeline for the evaluation of putative predictive biomarkers for ICI treatment that can be tailored and applied to diverse cohorts.
“We demonstrate the predictive power of a multivariable model dominated by metabolic and interaction features, which yields high accuracy in predicting progression free survival over 24 months.”
Professor Kulasinghe said philanthropic funding had played a critical role in building the laboratory capability needed to undertake this type of research.
“My lab wouldn’t be where it is today without the funding from Cure Cancer,” he said.
Cure Cancer Chief Executive Officer Vanessa Barry said the publications demonstrate the value of supporting emerging researchers and investing in innovative approaches to cancer research.
“These studies show what becomes possible when emerging researchers have early backing to test bold ideas,” Barry said.
“By investing at the start, we help build the evidence and capability that can later shape more personalised cancer care, and bring us closer to a future safe from cancer.”
Together, the studies underscore how advances in cancer metabolism, high-dimensional tissue profiling and machine learning are converging to deliver more precise insights into tumour behaviour.
As researchers continue to integrate biological and computational approaches, such discoveries may help pave the way for more personalised treatment strategies across a range of cancer types.
