New cardiac prevention paradigm explored in TRANSFORM trial using AI and CCTA

Transforming cardiac prevention screenings

The TRANSFORM trial will enroll more than 7,000 patients classified as moderate to high risk for cardiovascular events. Participants will be randomized into two groups. The control group will receive guideline-based care, while the investigational group will undergo CCTA scans at baseline and two years later. These scans, enhanced by AI assessment and detailed quantification of the various types of coronary plaques, will classify coronary plaque into stages of severity, dictating the intensity of medical intervention.

"So if they've got stage zero plaque, meaning they don't have any coronary plaque, they will get treated according to guidelines. So if the LDL cholesterol is super high, it'll still get treated. For example, on the other hand, if they had stage three plaque, well then we would throw the proverbial kitchen sink at them medically speaking, in terms of tiered medical therapy going above and beyond what the guidelines recommend using drugs that are FDA approved, but in ways that haven't been on label so to speak," Bhatt explained.

He said this may offer a better way to not only to catch plaque formation earlier when it is easier to treat, but also show patients images of the plaques to educate them on the developing disease and what they need to do to fix it.

"We're looking to see whether we can actually improve cardiovascular outcomes, reduce cardiovascular events with this approach of letting the patient know and their doctor know, 'Hey, you've got plaque and it's looking pretty bad. You've got stage three.' That might motivate the patient to actually take their medicines," Bhatt explained.

The trial will track patients with CCTA scans over time see whether intense medical therapy in the investigational arm keeps plaque from progressing over a period of two years. If the study shows the approach can transform outcomes, it could be a paradigm shift in early coronary disease detection, monitoring and treatment for the future, he said.

The current guidelines and way of doing things to manage patients can be improved, Bhatt said, and AI and CT screenings appears to be a very good way to do this. He said if a patients has a 70% blockage of the proximal LAD, they are often referred for an angiogram or to get a stent or bypass surgery. He said invasive angiograms carry some risk of complications, as do the suggested procedures.

"So bottom line is we need to figure out whether to do this or not. A lot of patients would love to just get a CT angio. It's a popular request in New York among the rich and famous, but the real question is: Is it the right thing to do? So if the trial is positive, it'll transform primary prevention. And if it's negative, then it means don't waste a lot of money on getting CT coronary angios," Bhatt said.

Key goals of the TRANSFORM trial

Bhatt said the TRANSFORM trial has three main objectives:
  • Risk Prediction: Evaluating whether CCTA with AI can outperform current risk calculators.
  • Plaque Progression: Assessing whether targeted medical therapy can prevent plaque progression between initial and follow-up scans.
  • Outcomes Improvement: Investigating whether this approach reduces cardiovascular events like heart attacks and strokes.

He emphasized that while coronary calcium scoring is now a widely used and inexpensive method for screening patients for coronary disease, it only identifies calcified, and often older, and now stabilized plaque. While an indicator or coronary disease, it does not show active disease or if lipid lowering treatments are working. In contrast, CCTA can detect softer, lipid-rich plaques that are more prone to rupture and cause heart attacks, particularly in younger patients.

AI may be a game-changer in cardiac prevention

AI in the trial plays a pivotal role in analyzing complex 3D images of coronary arteries, rapidly identifying high-risk plaque features such as necrotic cores. This capability allows for personalized risk assessment by comparing individual scans to vast databases of CCTA scans and patient outcomes.

While Cleerly is sponsoring this trial, he said other vendors such as Elucid and HeartFlow are also working on AI to enhance and automate analysis of CCTA.

"I think it's good that there are these three companies, and probably several others as well in startup mode, that are trying to figure out can we do more than just look at the pictures to gauge risk and hopefully then to help guide physicians and patients with what to do. And I think the AI is going to be a big part of it because right now, our ability to predict risk, especially in the primary prevention settings, is really poor. I mean, the risk calculators aren't really that great," Bhatt said.

He said the AI helps by performing a much deeper analysis than is possible by manual measurements or eye-balling assessments currently used. This type of quantification can develop personalized risk assessments for individual patients by comparing their pixels to the millions of pixels and images that other patients have gotten where event rates are already known.

The TRANSFORM trial has the potential to redefine how physicians approach cardiovascular prevention. Bhatt used the example of an asymptomatic patient who has a normal stress echo, a really low calcium score, but not a zero calcium score. He said one could argue whether any of that testing was actually totally warranted, but it does not end up predicting anything. He believes CCTA with AI assessments will offer a much clearer picture of what is going in inside a patient and help revise the guidelines, so he is excited about what the trial will show in the coming years.