HeartFlow Analysis: a better way to diagnose heart disease?

Abi Millar 24 October 2019 (Last Updated December 23rd, 2019 10:21)

A technology called HeartFlow Analysis has been rolled out by the NHS to aid the diagnosis of coronary heart disease. Using data captured by a CT scan, it generates a 3D model of a patient’s heart, and applies deep learning techniques to predict the impact of any blockages. We spoke to HeartFlow’s founder to find out its benefits.

HeartFlow Analysis: a better way to diagnose heart disease?
HeartFlow Analysis changes the very nature of cardiac CT. Credit: HeartFlow

For patients with symptoms of heart disease, getting an accurate diagnosis can be challenging. Not only is the process quite protracted – hospitals can take several weeks to make their diagnosis – it can also be clouded with uncertainty. In some cases, patients’ heart disease is missed, while in other cases they undergo invasive tests that ultimately prove unnecessary.

According to data from the American College of Cardiology, around 60% of coronary angiograms (an invasive procedure) found no obstructive coronary heart disease. This suggests the test is being overused – a real problem considering the cost to health services and the risk of complications for patients.

In recent years, there has been much discussion around reducing unnecessary angiography, as well as ensuring all patients receive an accurate diagnosis. Since coronary heart disease is the leading cause of death worldwide, killing approximately 17.3 million people a year, it’s crucial that the signs are identified and that patients are given an appropriate course of treatment.

It’s for this reason that a technology called HeartFlow Analysis is being rolled out across the NHS. Developed by Silicon Valley med-tech company HeartFlow, the technology was recommended by NICE in 2017 and is now being used in more than 40 hospitals across NHS England.

“The HeartFlow Analysis has demonstrated higher diagnostic performance compared to other non-invasive cardiac tests and gives cardiologists a level of examination only otherwise possible through an invasive procedure,” explains Charles Taylor, HeartFlow’s founder and CTO. “It can help physicians understand the impact of any blockages on blood flow to the heart and decide what action needs to be taken.”

How it works

The technology works by taking data captured during a CT scan, and generating a 3D digital model of a patient’s heart. Using deep learning techniques, it can home in on any arterial blockages and accurately predict the impact on blood flow.

“When doctors want to understand if a patient has coronary heart disease, or how serious the condition might be, they send them for a cardiac CT angiogram (CTA),” says Taylor. “The image data from the CTA is then uploaded securely from the hospital to HeartFlow’s cloud-based system. We’re then able to take the data and create a model of the patient’s coronary arteries. It applies powerful algorithms to solve millions of complex equations that simulate blood flow in the modeled arteries.”

Once the analysis is complete, it is securely transferred back to the hospital, where cardiologists can interact with the model and learn where blood flow is most compromised.

“If the HeartFlow Analysis indicates that blood flow may be severely restricted, doctors can decide how best to treat the patient and where to focus any invasive procedures such as stenting,” says Taylor.

In practice, most patients can be diagnosed on this basis alone, without needing any further invasive tests. One trial found that 61% of patients were able to safely avoid a coronary angiogram, and that the number of angiograms showing no significant heart disease (i.e., unnecessary angiograms) was reduced by 83%.

In many cases, using HeartFlow also impacts on the patient’s course of treatment. Because the technology offers cardiologists a much clearer view of what’s going on, it enables them to strategically target interventions.

“Research from the ADVANCE Registry, which looked at more than 5,000 patients across the world, found that using HeartFlow changed a doctor’s recommended treatment plan in two-thirds of cases,” says Taylor. “This means that some patients who were due to receive stenting or a bypass didn’t need this and could be treated with medication. Meanwhile, others who were due to be treated with medication were found to need invasive management to optimise blood flow to the heart.”

How it’s benefitting the NHS and patients

In essence, the HeartFlow Analysis changes the very nature of cardiac CT. Once viewed mostly as a way of ‘ruling out’ low-risk patients (with higher risk patients required to take further tests), HeartFlow turns the scan into a sophisticated triaging service. According to Dr Derek Connolly of Birmingham Hospitals NHS Trust, incorporating the technology has had a meaningful impact on his practice.

“For every five patients who have a cardiac CT and a HeartFlow Analysis, four patients go home knowing they don’t need anything else. Half of those patients will be on cholesterol tablets because they have early disease, and the other half will have normal coronary arteries,” he said.

From the patient’s point of view, getting a HeartFlow Analysis has many benefits – it does not require any additional visits or exposure to radiation, and gives them some much-needed diagnostic clarity. It’s good news for the NHS too, with estimated cost savings of £9.1m every year.

“These savings come from being able to avoid costly, unnecessary procedures and focusing time in the catheterisation lab on those who really need it,” says Taylor. “In short, it allows clinicians to be more efficient with their resources.”

The HeartFlow technology began to appear in NHS hospitals last year, under the Innovation and Technology Payment (ITP) programme. This June, HeartFlow announced that the funding had been extended through to March 2020, meaning an additional 30 hospitals would benefit.

“Thanks to our latest funding under the ITP programme, we’re able to continue building links within the healthcare system and are aiming to be in more than 60 hospitals by end of year,” says Taylor.

In the meantime, HeartFlow is working to enhance its product and develop new services. The team is also working on a tool called HeartFlow Planner – a real-time, non-invasive interactive planning tool that will help physicians explore different treatment options by virtually modifying the coronary arteries.

“It’s a really exciting advancement and something we are anticipating being able to launch later this year,” says Taylor. “In the longer term, we want to continue building the network of hospitals that we work with – we now have a presence in the US, UK and Japan and it would be fantastic to continue expanding our reach.”