Magnetic resonance imaging (MRI) is a powerful tool that doctors use to visualize organs and tissues in three dimensions. While the images can provide key insights, the high cost and low availability of MRI systems has led to long wait times for patients.
Dr. Masoom Haider, an affiliate scientist with the Techna Institute and a clinician scientist in the Joint Department of Medical Imaging, has devised a method for reducing MRI demand by avoiding unnecessary scans for patients at risk of prostate cancer.
Dr. Haider says the new strategies, "may support clinical decisions for a more judicious application of MRI to further improve the cost-benefit ratio."
Dr. Haider, along with his research fellow, Dr. Dominik Deniffel, and research team, sought to find alternatives to MRI scans among other more readily available and inexpensive clinical parameters. They wondered whether factors such as age, prostate size or the presence of molecular markers could predict MRI results.
The team began by collecting data from hundreds of patients who were at risk of prostate cancer and who had undergone MRI screening. The researchers then built a statistical model for finding patterns and making predictions – to categorize patient risk. The model was then used to predict whether an MRI scan would reveal prostate cancer.
The researchers found that if doctors were to implement this model, 29 per cent fewer patients would need MRI scans. Skipping MRI for these patients would rarely lead to missed prostate cancer diagnosis.
As part of the model building, the researchers discovered that one patient factor was particularly informative in predicting risk: the density of a molecular marker known as prostate specific antigen (PSA).
Dr. Haider's team found that by applying a patient cut-off for this factor alone, the number of patients perceived as needing MRI scans could be reduced by 25 per cent.
Although the predictive power of PSA density is slightly weaker than that of the full model, a cut-off for PSA density could be easily implemented as a low-cost and routine filter for MRI testing.
This work was supported by the Ontario Institute for Cancer Research, a Deutsche Forschungsgemeinschaft (DFG) Fellowship and Sinai Health Foundation.
