In a groundbreaking study led by the University of South Australia, a host of metabolic biomarkers that could be harbingers of cancer risk have been identified. Utilising machine learning technologies to scrutinise data from an expansive cohort of 459,169 UK Biobank participants, the study has pinpointed 84 features that could serve as early warning signs for cancer susceptibility.
This venture into hypothesis-free discovery used artificial intelligence to isolate these risk factors from a pool of over 2800 features. Dr Iqbal Madakkatel, one of the key researchers on the study, explained, “More than 40% of the features identified by the model were biomarkers—biological molecules that can indicate health or disease states—and several were co-related to cancer risk as well as kidney or liver disease.”
The study’s findings extend beyond the realm of cancer. They also shed light on chronic conditions affecting the kidneys and liver, thus highlighting the importance of understanding the pathological connections between these diseases and cancer. “After age, high levels of urinary microalbumin was the highest predictor of cancer risk. Albumin is a serum protein necessary for tissue growth and healing, but when it shows up in your urine, it’s a red flag for both kidney disease and cancer,” said Dr Amanda Lumsden, another researcher involved in the study.
Another intriguing finding was the association of greater red cell distribution width (RDW)—the variation in the size of red blood cells—with an elevated risk of cancer. Generally, uniformity in red blood cell size is indicative of good health, but discrepancies could be correlated with higher inflammation, impaired renal function, and as this study suggests, an increased likelihood of cancer.
High levels of C-reactive protein, a marker for systemic inflammation, and high levels of the enzyme gamma glutamyl transferase, a biomarker related to liver stress, were also linked to cancer risk.
Professor Elina Hyppönen, Centre Director of the Australian Centre for Precision Health at UniSA and chief investigator of the study, emphasised the power of machine learning in this research. “Our model could incorporate and cross-reference thousands of features to identify relevant risk predictors that might otherwise remain concealed,” she said.
While the study opens new avenues for future research and highlights the possibility of detecting cancer risk through relatively straightforward blood tests, it also beckons further studies for confirming causality and clinical relevance. Nevertheless, the implications are clear: early detection may offer a critical window for prevention, and artificial intelligence could be a valuable tool in this life-saving endeavour.
Support independent community journalism. Support The Indian Sun.
A study led by the Uni of South Australia has identified 84 metabolic biomarkers that could indicate an individual's susceptibility to #cancer using machine learning technology; may also offer insights into kidney & liver diseases. #TheIndianSun #AIhttps://t.co/tLzLeIqdzL
— The Indian Sun (@The_Indian_Sun) August 31, 2023