CHICAGO — When imagining the role of AI quantitation of coronary plaque, a possible starting point would be understanding the distribution and calcification of plaque in individuals with premature coronary artery disease, according to expert commentary presented on December 3 at RSNA 2025.
Jonathon Leipsic, MD, is looking at the potential of CT beyond stenosis.
“To understand the maturity of plaque and then eventually, potentially … maybe one day we’ll actually be able to track treatment response, although I think we’re a ways off from that on a personalized level,” Leipsic said during his talk.
The professor of radiology and cardiology at the University of British Columbia in Vancouver, Canada, and past president of the Society of Cardiovascular Computed Tomography (SCCT), reflected on his 20 years of practice evaluating coronary atherosclerosis on CT and highlighted how key research is guiding the path forward.
“Four years ago, one of the most important days in my clinical practice took place, and that was the updated chest pain guidelines,” Leipsic explained. “They provided a class 1A indication for the evaluation of symptomatic patients with suspected coronary disease in both the acute and stable setting, I think for a whole host of reasons.”
“I think what we’ve shown for the last 20 years is that CT is anatomically more accurate and more predictive of anatomical disease as it relates to invasive angiography,” he said.
Further, CT is prognostic beyond traditional functional stress testing, and CT informs medical management in a fashion that improves outcomes, Leipsic added.
What we’ve learned most recently is that millions of Americans are presenting with acute coronary syndrome, Leipsic noted. Furthermore, prior to a myocardial infarction (MI), about half of MI patients had no documented symptoms.
“By no means am I suggesting this is evidence that we should be screening people, but I do think at very least it gives us a pause to really understand how we can glean more from CT visually and where, if at all, quantitative atherosclerosis may play a role as it relates to informing risk and guiding perhaps a more bespoken approach to medical management,” Leipsic stressed.
Tools are limited, according to Leipsic, who also consults for HeartFlow. The segment involvement score is limited; the calcium score is limited in patients already on a statin, he said, adding that growing evidence suggests that lower calcium score may reflect less dense calcium. He also noted that in a multicultural, diverse population, calcium may underpredict risk in certain populations.
Next, Leipsic discussed high-risk, vulnerable plaque features, saying the maturation of atherosclerosis is inconsistent. Plaques evolve.
“Identifying a potentially vulnerable lesion doesn’t necessarily mean that it will remain vulnerable, or even that if it’s quote-unquote vulnerable or has a thin cap, that it will actually cause atherothrombotic occlusion because there’s so many other factors at play,” he said. It is rarely a lesion at baseline that goes on to cause incidental MI, according to Leipsic.
Jonathon Leipsic, MD, professor of radiology and cardiology at the University of British Columbia, explained what research is revealing about plaque progression in relation to myocardial infarction.
Which patients are really at higher risk? It’s a question left unanswered. Ultimately, tracking coronary artery calcium (CAC) score alone is ineffective because you can’t distinguish plaque progression, Leipsic said.
The bottom line is that there is still a lot of debate about when (or if) to use AI quantitation of plaque. It’s a point particularly important for younger people, according to Leipsic.
To see full coverage of RSNA 2025, visit our RADCast.