Can you tell us about University Hospitals (UH) Health System?
We are a 17-hospital, four billion-dollar health system in northeastern Ohio. We treat 1.3 million unique patients a year and employ nearly 2,500 physicians and surgeons. Our heart programs are organized around the Harrington Heart & Vascular Institute under the leadership of Marco Costa, MD, PhD. It is a comprehensive institute model including cardiology, cardiac surgery, vascular surgery, and cardiovascular imaging and anesthesia. The Harrington Heart & Vascular Institute oversees nearly 20,000 cath and electrophysiology lab procedures annually across the health system. There are ten hospitals with cath labs and two no on-site surgical backup percutaneous coronary intervention (PCI) programs in rural community medical centers. The five cardiac surgery programs have a total heart surgery volume of about 1500 pumps.
How is UH incorporating coronary computed tomography (CT) angiography and HeartFlow into the clinical workflow?
As an American College of Cardiology (ACC) National Cardiovascular Data Registry (NCDR) quality measure, we follow our “no significant coronary artery angiographic disease” rate when we take patients to the cath lab.
In a very large Medicare database study, the rate of “normal” coronary arteries at angiography in the United States exceeds 50% utilizing current diagnostic modalities, including stress testing with and without echo and nuclear imaging.1 In a pilot program incorporating HeartFlow FFRCT as the upfront diagnostic test, our “normal” coronary angiogram rate fell to less than 10%.
UH has an accountable care organization (ACO) that manages 520,000 covered lives. We are very focused on value, which may be defined as quality and patient experience divided by cost. If you take a patient to the cath lab and do a procedure based on what has been best-in-class diagnostic testing, i.e., stress echo and a nuclear stress test, and the patient has no significant disease, that is a failure. It is an unnecessary invasive procedure with higher costs.
We were a very early adopter of HeartFlow FFRCT and performed the first three HeartFlow cases when it was FDA cleared. We had been following the technology and participated in one of the clinical trials as a core lab for invasive fractional flow reserve. We have been very impressed with the combination of anatomical information from CT angiography and per vessel functional information from FFRCT, which offers us the ability to direct patients to the cath lab accordingly. It has been a journey through multiple different clinical trial sets.2-4 Data has shown that the knowledge provided through CT angiography affects treatment decisions, as well as the intensity of optimal medical therapy and compliance. Five-year results from the SCOT-HEART CT study5 demonstrated tremendous value in the CT angiography group. Heart attack and revascularization were significantly reduced in the group that underwent CT angiography. Adding function to anatomy with HeartFlow FFRCT only enhances that. The U.K.’s NHS is one of the best quality arbiters of sensitivity, specificity, positive/negative predictive value, accuracy of testing, and cost effectiveness. It is not a coincidence that the NHS came forward in support of FFRCT, noting that all else being equal, if you have state-of-the-art CT angiography, a sufficient number of cardiologists or radiologists who can read the studies, and if it is accessible and available to patients, FFRCT is the most accurate diagnostic test. We chose to go down that path.
However, just because it was cleared by the FDA didn’t mean we had reimbursement codes from the Centers for Medicare and Medicaid Services (CMS) or coverage from commercial payors. We essentially invested in the technology, in the sense that we were offering it to patients only for reimbursement of the cardiac CTA. We were not getting reimbursed for the HeartFlow portion, but we thought it was the right thing to do for our patients. Since we have 28,000 employees and insure all their dependents, we made the case to our ACO that FFRCT is the best diagnostic test to use for our patients. We were able to show that even in our own hands, outside of a clinical trial setting, the “normal” coronary angiography rate was reduced to less than 10%. That certainly attracted the interest of our ACO, because invasive angiography is expensive. FFRCT was approved for use with all of our employees and their dependents. We then worked, with the help of other centers and HeartFlow, on getting reimbursement from CMS, and then on getting insurers in our region on board. Anthem BlueCross Blue Shield became very supportive as the data came forward. We are now working with other commercial payers. It has been a step-by-step process.
We embraced HeartFlow technology for two reasons: one is its accuracy as a go-to diagnostic test, but the other is the power it holds in helping to plan interventions for patients before getting to the cath lab. What looks angiographically to be two-vessel disease but by FFRCT is only single-vessel disease allows the operator to know that they are going to stent the left anterior descending coronary artery (LAD) and do optimal medical therapy for the other vessel. It is hard to bring a patient in, cath them, and then say, we have great news for you — you have no significant disease. Invariably, a spouse or relative will ask, then why did you do the cath? It is always a difficult conversation to have and explain that the sensitivity, specificity and positive-negative predictive value of our existing technologies isn’t good enough for us to be definitive, and that is why we did a cath. Usually the good news will override that, but people are smart, and will say, this was an unnecessary test and my husband/wife shouldn’t have had it done.
Since we have implemented FFRCT, that conversation is now less common. That being said, we are not at 100% implementation. We haven’t given up on stress testing and stress echo, because not only do we have a lot of hospitals in the UH system, but importantly, you need access to high quality CT angiography. That in itself can be a journey, because patients need to be beta blocked with an ideal heart rate of 65 or less. When you start out, of course, some cases are rejected for technical insufficiency for upload and to get information to calculate FFRCT. High quality scans are essential, but if you work with HeartFlow and your radiology technicians, you should be able to get the acceptance rate up into the 90%-plus range. HeartFlow studies and the recent PROTECTION VI study show concerns about contrast use and radiation dose should not be a determinant. Modern CT angiograms are now well below 3 millisieverts and even the dye load with spiral CT is lower. Large amounts of contrast are no longer necessary with the new algorithms and scanner sequences. We are doing CT angiography with 40-50 ccs of contrast.
We have also presented data on our aortic stenosis patients being planned for transcatheter aortic valve replacement (TAVR). Patients undergoing TAVR get what we call a “TAVR CT”, basically a gated cardiac CT that gives you structural and coronary artery disease information. Approximately 25% of patients with aortic stenosis have concomitant coronary artery disease. We have reported our preliminary experience in a Transcatheter Cardiovascular Therapeutics (TCT) abstract that FFRCT can be useful in making a determination as to whether these patients need to be revascularized prior to TAVR. We are watching with great interest the ongoing studies regarding FFRCT use in patients who are recommended for coronary artery bypass graft (CABG) surgery, and how this anatomic and functional information may modify revascularization strategy.
How do you determine which patients undergo CT angiography with HeartFlow?
First, a patient with a high pretest probability of coronary artery disease shouldn’t undergo stress echo and nuclear stress testing, because with a high pretest probability of coronary disease, a negative stress test is more likely to be a false negative. The sensitivity and specificity of stress echo and nuclear stress testing are not sufficient to rule out coronary artery disease in someone with high pretest probability. In terms of pretest probabilities, these imaging modalities are optimized for intermediate risk patients. If a patient has a low pretest probability of disease, a stress test is more likely to be false positive than true positive. All of this means that you are going to cath people who don’t really have coronary artery disease. One area where HeartFlow may be useful is if someone with atypical chest pain keeps coming to the emergency room and says, I keep doing stress tests and they are negative. If you definitively want to rule out disease, then it is very useful in the low pretest probability patients who keep bouncing back. This is also true for the high pretest probability patients, for example, the diabetic with exertional chest pain and hypertension risk factors, where stress tests keep being negative, and you are wondering, what should I do here? FFRCT is particularly useful in these scenarios, because it offers both anatomy and function. However, all else being equal, if I have a patient in whom I want a diagnostic test, I use FFRCT as my frontline test, very similar to NHS in the UK. Cardiology guidelines have been built on nuclear stress testing and stress echo. It is hard to change clinical practice patterns overnight. Guidelines practice patterns change more slowly as well. It has been a journey for us over the past three years with respect to reimbursement, but now commercial reimbursement has followed CMS and the environment is much better. Let’s say I have a brother with an intermediate or high pretest probability for coronary disease. If he asks me, what is the one test you think I should have? I would say, FFRCT.
What about the use of FFRCT for preprocedural planning?
HeartFlow’s preprocedural planning module is still not cleared by the FDA, but has been under development and investigation. It is amazing. The ability to say what happens with tandem lesions when picking a stent of x length, or if two stents vs one stent are placed, and then calculate the imputed FFRCT if only the proximal lesion is stented, is very exciting.
Would you ever need to use intraprocedural FFR after the use of FFRCT?
It is a rare situation. If an FFR is >0.85 (not ischemic) or <0.75 (ischemic), it is clear as to what to do. If you take someone to the lab and their LAD is 0.69 and you are going to fix it, but the right coronary is 0.82, you might measure an FFR at that point. There are still situations where we will do an invasive FFR to see about that intermediate region between 0.75 and 0.85. However, if I am taking a patient to the lab, and the LAD is 0.69 and the right coronary is 0.87, I don’t do an FFR on that vessel.
Once the CT angiography is uploaded and HeartFlow sends it back, how does that information reach the interventionalist?
It depends on the site. At UH, we incorporate the HeartFlow interpretation into our CT angiography reports as an addendum. In some hospitals, it can be uploaded as an addendum and the interventionalist interprets it. It just depends on the workflow.
Any advice for centers considering the use of CT angiography with HeartFlow?
You need a great champion for CT angiography and CT in general. UH is fortunate to have a large and high volume structural heart program, with interventionalists including Drs. Hiram Bezerra and Guilherme Attizzanni, trained not only in intervention, but CT and magnetic resonance (MR) imaging. If that is not available in your practice, then you need a radiology champion, because the core of this capability is high-quality CT angiography that will allow for interpretable HeartFlow data.
Disclosure: Dr. Simon reports honoraria for educational activities for Medtronic and that University Hospitals Cleveland Medical Center has received research grant support from HeartFlow.
Dr. Daniel Simon can be contacted via Diane Tumbry at email@example.com.
1. Patel MR, Dai D, Hernandez AF, et al. Prevalence and predictors of nonobstructive coronary artery disease identified with coronary angiography in contemporary clinical practice. Am Heart J. 2014 Jun; 167(6): 846-852.e2. doi: 10.1016/j.ahj.2014.03.001.
2. Douglas PS, Pontone G, Hlatky MA, et al; PLATFORM Investigators. Clinical outcomes of fractional flow reserve by computed tomographic angiography-guided diagnostic strategies vs. usual care in patients with suspected coronary artery disease: the prospective longitudinal trial of FFR(CT): outcome and resource impacts study. Eur Heart J. 2015 Dec 14; 36(47): 3359-3367. doi: 10.1093/eurheartj/ehv444.
3. Nørgaard BL, Leipsic J, Gaur S, et al; NXT Trial Study Group. Diagnostic performance of noninvasive fractional flow reserve derived from coronary computed tomography angiography in suspected coronary artery disease: the NXT trial (Analysis of Coronary Blood Flow Using CT Angiography: Next Steps). J Am Coll Cardiol. 2014 Apr 1;63(12):1145-1155. doi: 10.1016/j.jacc.2013.11.043.
4. Fairbairn TA, Nieman K, Akasaka T, et al. Real-world clinical utility and impact on clinical decision-making of coronary computed tomography angiography-derived fractional flow reserve: lessons from the ADVANCE Registry. Eur Heart J. 2018 Nov 1; 39(41): 3701-3711. doi: 10.1093/eurheartj/ehy530.
5. SCOT-HEART investigators. CT coronary angiography in patients with suspected angina due to coronary heart disease (SCOT-HEART): an open-label, parallel-group, multicentre trial. Lancet. 2015 Jun 13; 385(9985): 2383-2391. doi: 10.1016/S0140-6736(15)60291-4.