ISCHEMIA: Invasive Strategy No Better Than Meds for CV Events

PHILADELPHIA, PA (UPDATED)—After more than 10 years and $100 million, the International Study of Comparative Health Effectiveness With Medical And Invasive Approaches (ISCHEMIA) has delivered a decisive result: an invasive strategy on top of optimal medical therapy (OMT) offers no benefit beyond OMT alone in preventing a range of major cardiovascular events in patients with stable, moderate-to-severe coronary artery disease.

That was true for the ISCHEMIA trial’s “expanded” primary endpoint of cardiovascular death, MI, hospitalization for unstable angina, hospitalization for heart failure, or resuscitation due to cardiac arrest—rates were essentially equal between groups out to 4 years. It was also true for the ISCHEMIA study’s original primary endpoint of death or MI.

Where angiography followed by PCI or CABG did have the edge over medical care in this hotly anticipated study was angina relief. Among trial subjects with daily or weekly angina, half of the invasively treated patients were angina-free at 1 year, compared to just 20% in the OMT group.

“What do patients care about when they are seeking treatment? They care about living longer and feeling better,” study chair Judith Hochman, MD (NYU Langone Medical Center, New York, NY), told TCTMD. “In the median 3.3 years of follow-up—and up to 5 years—overall, there was no difference in their survival in this trial, but they felt better with the invasive strategy if they had angina at baseline. So this is really a perfect example of where we need to do shared decision-making between patients and their physicians.”

Hochman, who led the study with primary investigator David J. Maron, MD (Stanford University School of Medicine, CA), presented the main trial results here today at the American Heart Association 2019 Scientific Sessions today; the quality-of-life outcomes (QoL) were presented by John Spertus, MD (Saint Luke’s Mid America Heart Institute, Kansas City, MO). Separately, Sripal Bangalore, MD (NYU Langone Medical Center), presented the ancillary ISCHEMIA-CKD trial results, followed by the QoL analyses for this prespecified ISCHEMIA study. Like the main trial findings, ISCHEMIA-CKD also found no benefit of an invasive approach over OMT in a population of patients with advanced chronic kidney disease and significant, stable coronary artery disease.

All of the experts who spoke with TCTMD said they were not particularly surprised by the results, but all of them had caveats.

Pamela Douglas, MD (Duke University School of Medicine, Durham, NC), joked that despite what people say: “Anything that ISCHEMIA found would have been a bit of a surprise because there truly is equipoise on this one, with people feeling ‘in their bones’ about one thing or the other.” That said, she continued, it’s not “terrifically” a surprise that there was no difference in the major outcomes, nor that there were more events early on in the invasive group that were then matched by later events in the conservative arm—details also presented by Hochman in her presentation today.

ISCHEMIA at Last

ISCHEMIA, conducted at 320 sites in 37 countries, enrolled 5,179 patients with stable CAD, preserved ejection fraction, and moderate-to-severe ischemia based on either stress imaging or exercise tolerance test (ETT). In all, more than 50% of patients in the trial had severe inducible ischemia at baseline, 33% had moderate, and 12% had mild.

I would think that the number of those procedures would decrease. That in and of itself we estimate could save over 500 million dollars a year, which is five times the cost of the trial over 8 years. Judith Hochman

Patients were randomized to a strategy of invasive coronary angiography followed by revascularization, if needed, on top of OMT or to an initial conservative strategy of OMT alone. Randomization was done prior to angiography and blinded CT angiography (CTA) was done in approximately two-thirds of the enrolled patient cohort to either exclude life-threatening left main disease (or disease deemed of comparable severity) and other acute problems such as aortic dissection or, in a smaller proportion of patients, to verify that CAD was in fact present. Ultimately, 73% of randomized subjects had a study CTA.

At a median of 3.3 years (range 2.2-4.4 years), rates of the primary endpoint were no different between groups: 13.3% in the invasive group and 15.5% in the OMT group (adjusted HR 0.93; 95% CI 0.80-1.08). Event curves out to 5 years show that the conservative strategy had fewer CV events occurring in the first 2 years, while the invasive strategy held the edge between years 3 and 5. The absolute difference between groups during both periods was roughly similar, leading to the equivalent results at final follow-up. Investigators are hoping to get funding to continue to follow patients for an additional 5 years.

For the major secondary endpoint of CV death or MI, the event curves followed a similar pattern, crossing at around the 2-year mark but again did not significantly differ at 4 years: 13.9% in the conservative group versus 11.7% in the invasive group (HR 0.90; 95% CI 0.77-1.06).

Rates of all-cause death were nearly superimposable over the years studied, reaching 6.5% and 6.4% at 4 years for the invasive and conservative groups, respectively. MIs, meanwhile, were more common among the invasively treated subjects in the first 2 years then more common among the conservative therapy group patients in subsequent follow-up, but the rate was ultimately not significantly different between groups. All other secondary endpoints were also neutral.

Take COURAGE With ISCHEMIA

The ISCHEMIA trial was designed to answer key questions left over from the 2007 COURAGE trial, which found no benefit with revascularization over OMT in stable CAD. At the time of ISCHEMIA’s launch in 2011, Hochman herself acknowledged that the medical community had been having “a very hard time accepting the [COURAGE] results.”

Some dismissed the findings, noting that COURAGE used outdated stent technology (mostly bare-metal stents) or didn’t truly optimize patients on the best-available medication. Others pointed to the fact that patients were randomized after angiography, potentially biasing the results; there were considerable crossovers, and some feared that the trial didn’t truly enroll patients with the most severe ischemia. In-depth analyses of COURAGE hinted that patients with the most ischemia appeared to benefit more from revascularization.

But despite the unknowns, ISCHEMIA investigators struggled for years to convince doctors to enroll patients. Maron famously performed a song modelled on an Elvis Presley hit, uploaded to YouTube, to cajole people to enroll one patient that very night. The trial was also tweaked, sparking controversy, to shore up the numbers: the primary endpoint was expanded beyond cardiovascular death and MI to include resuscitated cardiac arrest, hospitalization for unstable angina, and hospitalization for heart failure. The researchers also chose to extend the definition of ischemia from 10% or more on nuclear perfusion imaging to also include patients with 5% ischemic burden at low levels of exertion (≤ 7 METS) as well as those with ECG changes during exercise tolerance testing (ETT) without imaging. Most recently, the quality-of-life endpoint was changed to the shortened, seven-item version of the Seattle Angina Questionnaire (SAQ) that was introduced after the trial was funded.

Those changes, to varying degrees, prompted an outcry from experts asking if the changes will prevent ISCHEMIA from answering key questions that have plagued this field since COURAGE.

But now, with these results in hand, the cardiology community should accept that ISCHEMIA did, indeed, overcome many of the issues identified with COURAGE, session moderator Elliott Antman, MD (Brigham and Women’s Hospital, Boston, MA), told TCTMD.

“This is the very best medical therapy that we know how to deliver now, with the latest aggressive targets, and the very best cobalt-chromium, thin-strut stents that elute very effective limus derivatives that reduce restenosis, and we have a modern-day approach to CABG surgery. In that context,” he said, “this is a very powerful study and it does answer a lot of questions.”

Those answers should now inform decision-making and, critically, discussions with patients, Antman continued.

“To those who might say, ‘But I thought I knew about all of this from the results of COURAGE and I haven’t changed what I do,’ I would recommend that those individuals think about the OMT that was deployed in this trial, and the very latest stent technology, and the very latest surgical approaches,” he stressed. “And if they thought the COURAGE trial was not sufficient to change their practice because their technology and cath lab had advanced and they felt that that would have changed the results of COURAGE, this is the test of that.”

Hochman agreed, telling TCTMD: “For people without symptoms, who’ve either never had symptoms, which was 10% of our cohort, or who had well-controlled symptoms within the prior month, which was 36% of our cohort, there was no benefit. So, I really can’t see why people would still be recommending an invasive strategy, stenting or bypass surgery, and I would think that the number of those procedures would decrease. That in and of itself we estimate could save over 500 million dollars a year, which is five times the cost of the trial over 8 years.”

Interpreting Trial Results

Also commenting on the study for TCTMD, Gregg Stone, MD (Icahn School of Medicine at Mount Sinai, New York, NY), an ISCHEMIA investigator, stressed that the trial results do not apply to patients with ACS, patients with class III or IV angina, patients with heart failure or reduced ejection fraction, or those with left main disease.

“I think this trial supports either an invasive or a conservative approach for patients with stable disease [who have] either no symptoms or mild symptoms that can be controlled with medication and moderate-to-severe ischemia,” Stone said. “It tells you that you can safely revascularize patients with moderate or severe ischemia and they will feel substantially better if they had angina at baseline. Then, an absolutely reasonable approach—probably the appropriate approach—is the conservative strategy if they’re really having no symptoms at baseline. And finally, if a patient who has angina at baseline would prefer medical therapy then that would be also a very reasonably approach once left main disease is ruled out, because the risk of sudden death seems to be extremely low and there is overall no difference in survival with an early invasive versus a conservative approach.”

Asked if this interpretation put too much emphasis on the secondary health-status outcomes rather than the primary endpoint findings, not to mention the concerns in the cardiology community that too many patients are typically treated with an invasive approach, Stone said: “The trial results are the trial results, and what I’ve just told you is the proper scientific interpretation of results and how to apply them to patient care. . . . The reductions in angina and the improvements in quality of life were substantial and in the United States we have tended to pay for therapies that make people feel better.”

Feeling better or living longer are “what patients want,” Stone continued. “What we’ve shown in this trial, with a high precision, is that we’re not going to make you live longer by undergoing an invasive strategy, but we will most likely make you feel better. And with a relatively small number of patients needed to treat to make you angina-free. I think that’s a substantial and important finding—it was the major secondary endpoint of the trial and it was a whole separate study done for that reason because those are our two priorities.”

Sanjay Kaul, MD (Cedars-Sinai Medical Center, Los Angeles, CA), who also reviewed the trial results for TCTMD, gave investigators an “A+” for ISCHEMIA’s design and conduct, adding that the “quality of the data is excellent.”

Otherwise, he continued, “nothing surprised me. I think the lack of a death or MI benefit is consistent with prior evidence and the improved frequency of no angina is also consistent with prior evidence.”

The “catch,” Kaul continued, “is that the functional outcome was a secondary endpoint and so the key question is, how do you interpret a favorable secondary endpoint when the trial fails to win on the primary endpoint? And the answer is: with circumspection. Basically you cannot draw any meaningful inferences, but that’s a statistical caveat. The fact that the invasive strategy had an advantage in reducing angina is consistent with prior evidence. . . . These results may be disappointing for some, but they are not surprising to me.”

Asked about the problem of overinterpreting the secondary health outcomes, Hochman had this to say: “This is not a novel drug [for which] we’re are seeking approval from the FDA, where you live and die by the primary endpoint and whether it reaches ‘statistical significance.’ We specifically set out to see what the hazard ratio—the point estimate—was for multiple endpoints, with precision around those endpoints. So, for example, stroke is very important for patients, and we showed that the stroke rate was very low and similar between two groups. You don’t look at this in the same way you look at a drug registration pathway trial, and clinical events are important to patients, particularly survival, and quality of life is important, too.”

Spertus stressed the same thing in his presentation today. “In patients with angina, shared decision-making should occur to align treatment with patients’ goals and preferences,” he emphasized.

But speaking TCTMD last month, Spertus predicted that Medicare and payers will be paying close attention to the ISCHEMIA results.  

Antman, when asked whether he anticipated changes to reimbursement following ISCHEMIA, said it was “too soon to say.”

“There are many more things we need to know, detailed analyses, and the paper which hasn’t been submitted yet—we’re all going to want to read that paper,” he commented, adding, “We’re going to want to ask detailed questions like—tell me what it was like in the ‘ideal’ groups, those who got complete revascularization versus those who truly achieved an LDL below 70 mg/dL and a BP below 130 mm Hg in the conservative group. That’s going to be a very important thing.”

Even more important will be longer-term insights, Antman continued. “Imagine that you’re a patient or it’s a relative of yours, a patient with stable exertional angina. The question is not ‘what do the next 4 years bring to me?’ but ‘in the next 10, 15, or 20 years, [what happens] if I go with a conservative approach or an early invasive approach? And I hope the investigators are successful in obtaining the funding to do that, because that’s what we really want to know.”

 ‘A Big Win for CT’

One group of cardiologists who may be celebrating the ISCHEMIA results tonight are advocates of CT angiography, who point to the 434 patients excluded from the trial on the basis of CT findings showing left main disease, as well as the 1,218 patients excluded for having no obstructive CAD despite chest pain symptoms.

Commenting on the results for TCTMD, Jonathon Leipsic, MD ((St. Paul's Hospital, Vancouver, Canada), called the trial “a big win for CT.”

“Real-world stress testing does not confidently identify epicardial coronary disease,” he said. “[Cardiac] CTA served as a good gatekeeper to the cath lab with over 85% participants screened by CT being revascularized, highlighting that prior to conservative initial management a CTA is needed to confirm the extent of disease and to exclude [left main] disease. Perhaps, then, all that is needed is lesion-specific physiological evaluation to assess which lesions need revascularization based on recent data such as FAME 2 and SWEDEHEART. This approach may in fact result in an even more profound angina relief and MI reduction in the interventional arm.”

I wouldn’t be comfortable, as a physician, with a patient that has significant ischemia, simply assuming that there was no left main disease in this population. Pamela Douglas

Douglas made a similar point, noting: “Over a third of the patient population were screen fails and so the only way to know about those [cases] . . . is from the CT. I’m not sure that most of us are willing to take a 5% risk of not intervening in the left main. If you were to act on ISCHEMIA and say, “I’ll treat all of my patients with medical management, and I’ll intervene if their angina doesn’t go away or it accelerates or they can’t manage their meds,’ then [you risk missing] the 5% of those patients in this population who have left main disease, and we know that revascularization is lifesaving in those people. I wouldn’t be comfortable, as a physician, with a patient that has significant ischemia, simply assuming that there was no left main disease in this population. This says you need to do some kind of angiographic second test, either using CT or just going to a cath lab for a look-see.”

Starting the Discussion

There’s more work to be done in this contentious field. Both Hochman and Douglas pointed to the high proportion of patients in the trial with chest pain but less ischemia and no identifiable obstructive disease, who are more likely to be women. The appropriate drug regimen in this group is unknown; more details may emerge from the CIAO-ISCHEMIA trial, which Douglas noted is looking at individuals screened out of ISCHEMI due to having no obstructive coronary artery disease on coronary CT angiography.

Hochman also stressed to TCTMD that the level of complications in the trial was exceedingly low, in part a reflection of the rigorous quality-control measures instituted in this trial. The results therefore do not extend to lower-volume centers or for that matter to patients who are not being maximized on guideline-directed medical therapy (GDMT).

Alice K. Jacobs, MD (Boston Medical Center, MA), picked up on this point as the formal discussant for the trial following its main arena presentation. “Adherence to GDMT is critically important and challenging, even in the context of a randomized controlled trial and certainly in our clinical practice,” she said.

Maximizing patient therapy, if a conservative approach is chosen, will be mandatory to achieving the same results seen in ISCHEMIA, Jacobs noted. Moreover, “it is incumbent upon all of us as clinicians to discuss the early and late risks and benefits, and what to expect based on the degree of angina. . . . The patient point of view cannot be overstated.”

There’s also the reality that ISCHEMIA was not a blinded study, a point raised by ORBITA principal investigator, Rasha Al-Lamee, MBBS (Imperial College London, England). She pointed out that the number needed to treat to be angina-free in ISCHEMIA was three, as compared with five in ORBITA. “We do need to be a bit cautious,” she said. “Obviously some of that impact may be the impact of placebo, and we can’t underestimate what we do as physicians in terms of making these patients feel better.”

Additional discussion zeroed in on the trade-off of early periprocedural MIs early and spontaneous MIs later on. Panel member Roxana Mehran, MD (Icahn School of Medicine at Mount Sinai), called the trial “practice-changing” for the many clinicians who refer these patients quickly for angiography. “Now we can relax—for at least the first 2 years anyway,” she quipped.

But a “very important question,” she continued, is what to tell patients to explain this trade-off. “The late events make me believe wholeheartedly that we can’t walk away from these patients without knowing what’s next, and the follow-up of ISCHEMIA is at least as important as what we learned today.”

Mehran also raised a question often posed, the possibility “we’re doing too many stress tests on these patients who have moderate or mild symptoms . . . . Maybe it’s more important to rule out left main using a CTA.”

On this point, Jacobs quickly said that “I think we will still need stress tests to make the diagnosis, because angina is a bedside diagnosis not a cath lab diagnosis, so we need stress tests to do that.”

But panelist James de Lemos, MD (UT Southwestern Medical Center, Dallas, TX), took a different tack, predicting, “The downstream effects of this for noninvasive testing are massive. . . . and that’s where we’ll see more of an impact on our practice, because many of us as a result of Rasha’s work [with ORBITA] have become more conservative, but this will have major effects on cardiac testing.”

Also serving as a panelist, Brahmajee K. Nallamothu, MD (University of Michigan, Ann Arbor), said ISCHEMIA will definitely change his practice, as did COURAGE. “This is teaching me that not every patient with moderate-to-severe ischemia needs to go right away to the cath lab,” he explained.

Spertus made a plea for an investment not only in tools that incorporate large data sets to help physicians make evidence-based decisions, but also the research to study how to implement such tools. “If we build the tools and nobody uses them,” he said, “then we’ve wasted our time.”

The last word went to Hochman, who noted that there had been a lot of talk on social media about the $100 million price tag of this trial, which she noted is more than offset by the savings that could be recouped if stable, asymptomatic patients are not sent to the cath lab. “I want you to know that this was not a pragmatic simple trial. Everything was read by the core lab—EKGs at baseline [and] over time, every single stress test, every [cardiac] CTA, the angiograms, the PCIs, the surgical reports; the clinical event committee adjudicated every single event. That costs money. I believe it was money well invested,” she concluded.

Note: Stone is Co-Director of Medical Research and Education at the Cardiovascular Research Foundation, the publisher of TCTMD.