Doubling Down on SCOT-HEART: Mechanisms, Applicability, and Future Directions

BALTIMORE, MD—Almost a year after the release of the hotly debated results from SCOT-HEART, testing CT angiography (CTA) as a triage tool for chest pain, experts continue to scrutinize the findings for mechanisms of benefit and the potential human and economic impact of adopting CTA as a first-line approach.

SCOT-HEART was presented at the European Society of Cardiology (ESC) Congress 2018 and showed a 41% reduction in coronary heart disease death or nonfatal MI over 5 years if stable chest pain patients had been initially worked up using CTA as compared with standard care alone—predominantly exercise ECG. As reported by TCTMD at the time, the phenomenal results for CTA prompted both celebration and disbelief.

Last Sunday at the annual Society of Cardiovascular Computed Tomography (SCCT) meeting, SCOT-HEART primary investigator David Newby, MD, PhD (University of Edinburgh, Scotland), told the audience that in the wake of the presentation last August—at least in the United Kingdom—“what we’re seeing is SPECT is falling dramatically and CCTA is rising rapidly, and interestingly, there's a hint that invasive angiography is also falling because of CCTA.”

What is the Mechanism of Benefit?

Much of the criticism of the trial has centered around how a triage strategy could be responsible for such a dramatic reduction in late events. “Clearly we all recognize that CT isn’t a therapy in itself, so these benefits must have been derived from changes in patient management in response to the information that CT provides to clinicians,” said SCOT-HEART co-author Philip Adamson, MD (University of Otago, Christchurch, New Zealand), who spoke at SCCT about the potential mechanisms behind the study findings.

SEE ALSO: SCOT-HEART: A Win for CT Angiography in Chest Pain Workup Prompts Cacophony of Reactions

“First, we believe that CT improves the diagnostic accuracy for angina and importantly it improves risk stratification,” he explained. “Second, showing patients their CT scan could improve adherence to lifestyle advice and to medications. Third, the information that CT provides enables clinicians to be better targeted in their provision of preventive medications so [these] can be targeted to the presence or absence of coronary atheroma, rather than typing it towards a more nebulous clinical risk score. And finally, we believe that CT may benefit some patients by identifying those who benefit from early and appropriate coronary revascularization.”

This shows how important identifying that nonobstructive coronary disease is, because only around half of these patients were on statin therapy at baseline. This is a substantial treatment gap. Philip Adamson

Adamson provided several examples, some unpublished, of post hoc data from SCOT-HEART illustrating his arguments. Specifically, he noted, among patients who underwent exercise treadmill testing, even those with a normal result had a “significant” reduction in their 5-year MI rate of around 2% compared with the approximately 0.5% seen among patients who were originally deemed to have a normal coronary angiogram, demonstrating that “coronary CT angiography does a much better job of risk stratifying patients compared with standard care.”

In another analysis, he showed that among patients with a negative exercise tolerance test, more than half were subsequently seen on CT to have either nonobstructive or obstructive coronary artery disease.

“This shows how important identifying that nonobstructive coronary disease is, because only around half of these patients were on statin therapy at baseline. This is a substantial treatment gap,” Adamson said. Moreover, because “clinical risk scores do a poor job of identifying coronary atheroma . . . and clinicians clearly respond to the findings of the CT scan,” patients with CAD are much more likely to receive preventive medications compared with those with normal coronary arteries. “And that's irrespective of their clinical risk score.”

Lastly, looking at the timing of events according to randomization among patients with non-anginal chest pain, possible angina, and those with prior coronary heart disease, Adamson said each of these groups’ risk patterns “[correspond] to the benefit seen from the observed changes in treatments.” The first cohort “can be thought of as an asymptomatic or primary prevention population.” In this group, early revascularization rates were minimal in both arms, but use of preventive medications was markedly higher among patients who underwent CTA, and fatal or nonfatal MI rates were higher in those who received standard care.  “Statins are a long-term treatment,” he said, “and among this group, we can see that the majority of benefits were derived between 2 and 5 years.”

A similar pattern was seen in the “possible angina” group, a notable difference being a reduction in fatal and nonfatal MI following CTA both within and after the first year. According to Adamson, CT leads to increases in both early coronary revascularization and changes in preventive therapies, and “as a result, they benefit both early, within the first year, and late, out to 5 years.”

On the other hand, for patients with established coronary heart disease, “we would expect nearly all of these patients to be on aggressive medical therapy at baseline, and indeed there was very little additional provision of preventive medications” in the CT arm, and almost none in the standard therapy arm, Adamson showed. “However, there was substantial coronary revascularization in this group, and if we look at the time course of observed benefits, we can see that essentially all the benefits in this cohort were in the first year, perhaps suggesting that they are benefiting from their early appropriate revascularization.”

“We hope all this data, in combination, will help reassure those who initially questioned the SCOT-HEART results,” he told TCTMD in an email, adding that he expects to publish a manuscript soon. “Our data is entirely consistent with other CT trials and across all patient subgroups. Although these analyses are post hoc and exploratory, we believe that we have demonstrated the plausibility of the SCOT-HEART findings.”

Examining the Exercise Test

In a separate presentation, Koen Nieman, MD, PhD (Stanford University School of Medicine, CA), homed in on the question that many have asked: did the use of exercise ECG—which he dubbed the “black sheep [of] diagnostic testing in cardiology”—as the comparator affect the ultimate SCOT-HEART results? PROMISE, the other large randomized trial of CTA as a decision-making tool, did not show a benefit of CTA over standard care, but in that earlier trial only 10% of patients underwent an exercise ECG and the majority had a nuclear stress test. By contrast, in SCOT-HEART, the standard workup consisted predominantly of exercise ECG, with only about 10% receiving a stress imaging test.

Nieman showed post hoc data demonstrating that use of coronary angioplasty in SCOT-HEART was “slightly more, but not significantly more, in the CT arm, which is of course in contrast to the revascularization numbers in PROMISE, where there was a significant difference. Also, in PROMISE there were more invasive angiograms [in the CTA group], which was not the case in SCOT-HEART. Those are very distinct differences.”

Nieman argued that critics who suggest that the high rate of ECG testing in SCOT-HEART biases its results aren’t appreciating the fact that patients underwent ECG before being randomized. Moreover, the same proportion of patients in both arms underwent stress testing—roughly 10% in each arm. In the standard care arm, however, most patients only had ECG testing, except for the approximately one-third who also underwent stress imaging or invasive coronary angiography. “It's not completely surprising that when you add an anatomical test to a functional test, you will change diagnosis, while if you add . . . a functional test to another functional test, that change will of course be less,” he said. “That illustrates the strength of combined anatomical and functional imaging to make us certain about our diagnosis.”

As to who were those patients identified as having CAD on CT who also received invasive angiography and why, Nieman said “it's hard to imagine that that did not have something to do with the exercise testing; that these physicians had to decide which of those patients who had a 50% lesion on their CT angiographs [should] go for revascularization or cardiovascular risk reduction based on statins and other therapies.”

He concluded that he looks forward to seeing future data from SCOT-HEART “investigating the role of stress testing in conjunction with CTA to guide catheterization and revascularization.”

SCOT-HEART Substudies

Reviewing the data from SCOT-HEART substudies, some unpublished, in his presentation, Newby said “we've really now got definitive evidence of how very, very sensitive CCTA is for defining obstructive disease. The specificity is as good as any of the other modalities, so surely we should be using that.”

As for cardiac biomarkers, he explained that a study in press of about 1,000 patients who gave blood samples at the time of CT found that “both BNP and troponin correlate beautifully with coronary disease and age. . . . LV mass is very particular for high-sensitivity troponin, so the bigger your heart, the higher your troponin. And BNP, the larger your heart, the larger your BNP.”

Additionally, in a gender analysis, there were no “big surprises.” Women in SCOT-HEART tended to have less typical angina and high rates of normal angiography, whereas men had more typical angina and more obstructive disease.

What really “struck” Newby, he said, was looking at the diagnostic power of CT with regard to gender, especially since it has been shown previously that among those who present with chest pain, women are more likely to be dismissed as having noncardiac symptoms compared with men. In SCOT-HEART, however, “women were certainly overdiagnosed and overtreated for coronary heart disease compared to men. . . . It really pulls out that for women it's actually CCTA [that] is having a big strength in terms of reducing overtreatment and misdiagnosis.” It’s notable that although both genders derived benefit, women had a lower rate of events and also seemed to be at an advantage more in terms of primary prevention compared with men.

Prior studies looking at patients with nonanginal chest pain have found that they “still go on to have significant event rates and a third of subsequent MIs are in that group.” CT, however, can better define this patient cohort, which means that “you're more certain about treating people who do have disease,” Newby said.

With regard to plaque characteristics, he stressed that CT provides “quite a lot of exquisite detail” beyond simply telling if a patient has coronary disease. Specifically, one recent study showed that “adverse plaque characteristics do define those who go on to have events.”

It can be “messy” looking at all the permutations of having adverse plaque or obstruction in a vessel, but “you do get the sense that both adverse plaque and obstruction together add value,” Newby said. “So if you have an adverse plaque and no obstruction, it's as bad as having obstruction with no adverse plaque. The combination is the worst of them all. We are learning more and more about it. . . . Ultimately when we look at all these features, they are predicting and it's how we synthesize that down and take that forward in terms of giving a risk for our patients.”

Lastly, he agreed with much of the discussion regarding the mechanisms behind SCOT-HEART, namely the ‘seeing is believing’ argument. “I think it’s personally sort of intuitively obvious,” Newby said. “It doesn't matter what your cardiovascular risk score is, sometimes people just won’t take their medications until they really feel they’ve got the disease. The power of showing them that they've got coronary disease really does impact whether they take their prescribed therapies, and of course that's where CTCA comes in.”

I’m really looking forward to the 10-year results when hopefully I can prove that CT actually reduces overall long-term revascularization rates and indeed replace angiography. David Newby

As a final point, Newby urged the audience to stop apologizing for increasing revascularization rates with CCTA. “You found more disease and the normal rate is lower when you do take to cath lab, and the proof of the pudding is that beyond the year of when you actually change your treatment—you revascularize them— subsequent revascularization rates are half that in the standard care group because you found the people and you've treated them and they don’t then need further treatment downstream. I’m really looking forward to the 10-year results when hopefully I can prove that CT actually reduces overall long-term revascularization rates and indeed replace angiography.”

Applicable to the US

But can all of these results be applied to the United States? Ahmad Slim, MD (University of Washington, Seattle), argued that they can and should in his presentation.

Even if a CT-based strategy reduces the rate of MI in patients with nonobstructive disease solely by identifying patients who would benefit from statin therapy, he said the US healthcare system could save over $3 billion in a single year. “The cost of hospitalization in the US as it relates to MI in the first year [is] about $45,000 for each admission,” Slim explained. “If you look at about 5.9 million admissions for acute chest pain in the US annually, and about 175,000 of these are missed MIs in patients at low- to intermediate-risk and you halve it and apply $45,000 for admission, that is close to $3.5 billion dollars annually.”

Assuming these patients are put on statins (add $160 per person) and an “extremely high rate of revascularization,” that will still only cost $200 million he added.

Nevertheless, “it behooves us to not only just look at the benefit but also the challenges,” Slim said, showing a colored map of the UK designating areas with and without CTA capabilities. Per the NICE guidelines, while CT proved “very effective,” there were issues with the regional distribution of scanners and the workforce needed to operate them, which inevitably affected patient access.

“These are things we have to start planning for and anticipating in the US,” Slim said. “We have a clean map, and I stress that as a society, and as industry partners, and everyone in this room, it’s our job to make sure that this map is fully green, not sporadically green as [the] NICE guidelines highlighted in the first year. This is a little bit of a challenge, but I think this is a challenge that we all welcome.”

Slim told TCTMD in an email that down the road, he would like to see “analyses of cost savings, not just in the patients who had events, but in the patients who had no disease as compared to disease: how much cost savings were incurred over 5 years by reducing recurrent outpatient and inpatient evaluations, downstream utilization, etc., as compared to the nonobstructive and obstructive CAD group.”

Further, he would like to see SCCT convert the science generated by SCOT-HEART into a “simple, patient-based educational pamphlet and primary care provider simplified decision-making algorithm. The information that reflects the superiority of CCTA is there over other modalities, it just needs to be simplified and better disseminated.”

Ultimately, Slim said, those who maintain imaging tests can’t improve outcomes “are using a fallacy. . . . The false reassurance of a normal nuclear stress test leads to patients not being treated to goal with primary prevention therapy, [whereas] CCTA therapy identifies both obstructive and nonobstructive disease, which drives initiation of preventive therapies as translated into a lower risk of myocardial infarction as compared to physiologic testing upfront.”