Complete Revascularization Analysis of ISCHEMIA Opens ‘Pandora’s Box’

Gregg Stone calls the data hypothesis-generating, but William Boden disagrees, calling them “interventional science fiction.”

Complete Revascularization Analysis of ISCHEMIA Opens ‘Pandora’s Box’

Patients with stable coronary artery disease who received “complete” revascularization in the ISCHEMIA trial appear to have derived more clinical benefit with the invasive strategy than medical therapy over follow-up, a finding at odds with what was observed for the overall trial. The implication, say investigators, is that revascularization might offer clinical advantages over conservative medical therapy if all anatomically relevant lesions are treated.

This analysis, which focused only on the patients who had complete revascularization defined anatomically, showed that the primary composite endpoint occurred in 11.9% of patients treated invasively as opposed to 15.4% of patients who were treated medically.

The clinical take-away, said Gregg Stone, MD (Icahn School of Medicine at Mount Sinai, New York, NY), who presented this exploratory analysis at the American College of Cardiology (ACC) 2021 Scientific Session, is that if imaging suggests complete revascularization can be achieved, it might tip the balance in favor of the invasive approach with PCI or CABG surgery.

“For hard events, we saw that anatomic complete revascularization is important, particularly when we decide on an invasive versus a conservative strategy, whether it’s by a CT scan where we can get some information on the likelihood of achieving complete revascularization or if it’s from an angiogram,” said Stone. “We must be honest with ourselves. If we think we can achieve complete revascularization, we might push the patient a little bit more, inform them a little bit more, that their outcomes will be slightly better with the invasive approach. We also have to appropriately use bypass surgery if we say we can’t achieve complete revascularization with PCI.”

If we think we can achieve complete revascularization, we might push the patient a little bit more. Gregg Stone

William Boden, MD (Boston University School of Medicine/VA New England Healthcare System, MA), who led the COURAGE trial in patients with stable ischemic heart disease and was an ISCHEMIA investigator, said he is unimpressed by the new data. He has specific criticisms of the analysis—including questioning how PCI could ever achieve complete revascularization when operators are treating focal lesions—but even larger objections to what he sees as a reluctance on the part of interventional cardiologists to accept the ISCHEMIA trial results.

“ISCHEMIA was a neutral/negative trial for both the five-component primary endpoint and the two-component endpoint of cardiovascular death/MI,” Boden told TCTMD. “That is the inescapable scientific fact of the trial. Even if we prespecify various subgroup analyses, they must be interpreted in the context of the overall trial’s proper outcomes. That’s what bothers me about this. . . . This is interventional science fiction. It’s making up a story that you want to portray an outcome that really wasn’t there to begin with.”

Kevin Bainey, MD (Mazankowski Alberta Heart Institute, Edmonton, Canada), on the other hand, took a more upbeat, albeit cautious, view of the findings, but he acknowledged that the new analysis open a “Pandora’s box” with respect to complete revascularization in patients with stable ischemic heart disease. He also emphasized that while these data are provocative, they should not take away from the overall ISCHEMIA findings.   

“It’s a prespecified subgroup analysis from the ISCHEMIA trial, so the association of angiographic complete revascularization with an improvement in CV death/MI compared with the conservative approach is only an association or hypothesis-generating, but I do think it’s the best evidence to date,” Bainey told TCTMD. “It shouldn’t, in my opinion, detract from the main ISCHEMIA trial findings.”    

In the ACS arena, particularly in STEMI, complete revascularization makes clear sense, Bainey added. The COMPLETE trial showed that complete revascularization as opposed to treating the culprit-lesion only led to a reduction in the risk of CV death or MI, and subsequent meta-analyses, including one he led, showed there was an improvement in clinical outcomes, including a reduction in CV death. ISCHEMIA, by contrast, was undertaken in patients with stable, ischemic coronary disease.


In ISCHEMIA, published in the New England Journal of Medicine last year, the primary endpoint at 4 years had occurred in 15.5% of patients managed medically versus 13.3% of those randomized to PCI or CABG surgery on top of medical therapy, a nonsignificant difference of 2.2%.

This is interventional science fiction. It’s making up a story that you want to portray an outcome that really wasn’t there to begin with. William Boden

In this latest analysis of the trial, investigators assessed the anatomic and functional completeness of revascularization after PCI and CABG. With PCI, completeness was assessed with a review of pre- and postprocedure angiograms and included staged procedures, while complete revascularization of surgery was assessed by a review of the preprocedure angiograms and CABG procedure reports.

Among 1,825 patients randomized to invasive therapy, 43.3% had complete anatomic revascularization, including 46.2% treated with PCI and 35.8% treated with surgery. Based on functional assessment, 58.3% of all invasively managed patients were completely revascularized, including 61.1% of patients treated with PCI and 50.6% treated with surgery.

In Stone’s analysis, the primary endpoint occurred in 13.8% of patients with incomplete revascularization versus 9.3% with anatomic complete revascularization (P = 0.001). For those with a functional assessment, the primary endpoint rates with incomplete and complete revascularization were 14.0% and 10.3% (P = 0.02), respectively, and those differences held up in adjusted models.

Researchers also compared the clinical outcomes of patients who were completely revascularized with either PCI or CABG surgery against patients treated with medical therapy. Here, in an inverse probability-weighted analysis, the primary endpoint at 4 years occurred in 11.9% of patients with anatomic complete revascularization versus 15.4% of ISCHEMIA patients treated with medical therapy, a -3.5% difference (-7.4% to 0.1%). For those with functional complete revascularization, the primary endpoint occurred in 13.1% of patients, a -2.3% difference (-5.4% to 0.8%) versus conservative treatment.  

“Anatomic complete revascularization was more important than functional complete revascularization and that’s actually interesting,” said Stone.

Anatomic Complete Revascularization: Covariate-Adjusted Cox Model



Conservative Medical Therapy

Adjusted Difference (95% CI)

Primary Endpoint*



-3.6% (-6.9% to -0.7%)

CV Death or MI



-3.0% (-6.5% to -0.1%)

CV death



-2.2% (-4.5% to -0.5%)




-1.0% (-4.2% to 1.7%)

All-Cause Mortality



-0.8% (-3.9% to 1.5%)

* CV death, MI, or hospitalization for cardiac arrest, heart failure, or unstable angina

Regarding the comparison of complete revascularization versus conservative medical therapy, Boden said this is a best-case scenario that isn’t a fair fight between the two strategies. By removing patients with incomplete revascularization, the invasive approach is optimized and then compared with all-comers in the conservative arm, a group that would include a range of patients, including those with single lesions as well as those with more extensive, diffuse disease.

“It’s not an apples-to-apples comparison,” said Boden. “It’s comparing the best-of-the-best in the invasive group versus all-comers in the conservative group. It’s just not balanced.”

Moreover, Boden said these new analyses focus specifically on spontaneous (type 1) MIs but ignore/downplay other MIs, such as those related to PCI, stent thrombosis, restenosis, or CABG (types 4a, 4b, 4c, and 5). There is an early hazard to PCI and CABG, he said, and pointed to a recent analysis from ISCHEMIA by Bernard Chaitman, MD (Saint Louis University School of Medicine, MO), showing that 4b and 4c MIs were strongly associated with adverse clinical events, including all-cause mortality and cardiac mortality.

To TCTMD, Bainey stressed that the ISCHEMIA trial is the gold standard right now when it comes to clinical decision-making for patients with stable CAD. Like others, he noted that the event curves cross at 2 years and said the community is eagerly awaiting the long-term results from the trial. As part of ISCHEMIA-EXTEND, which now includes more than 5,000 participants enrolled in the follow-up study, investigators say they plan to track patients for approximately 10 years.

“The main trial shows no difference, and we have to remember that,” said Bainey. “Where I do think this analysis becomes critically important would be in patients that we see electively, who for whatever reason find their way to the cath lab and are now contemplating a revascularization strategy. This analysis really pushes us towards trying to achieve complete revascularization in these patients, whether it be PCI or CABG.”

For Bainey, one of the interesting findings was that just roughly half of patients treated invasively were completely revascularized. “I would have thought more people would have tried to go for complete revascularization, but a little less than 50% makes me think there is equipoise in the community,” he said.

ACC Session Questions

During the question-and-answer session following his presentation, Stone said that some of the major predictors of incomplete revascularization included the presence of extensive coronary artery disease, chronic total occlusions, and smaller vessels.

“We were fairly sensitive requiring complete revascularization and it might be surprising to some that, anatomically, only 43% of the patients [undergoing PCI] were completely revascularized,” said Stone. “Bypass surgery was better at achieving complete revascularization than PCI, not surprisingly, so [for] patients with extensive disease, if you do want to achieve complete revascularization, that would point to performing bypass surgery rather than PCI in many of these patients.”

To TCTMD, Boden noted that multiple clinical trials—BARI-2D, FREEDOM, and SYNTAX, for example—have shown that CABG surgery is better than PCI for reducing hard clinical outcomes. “We have data from three randomized trials showing the superiority of CABG over PCI, presumably as a consequence of being able to achieve complete revascularization,” he said. “I’m at a loss why anyone would think that PCI is going to result in improved complete revascularization.”

As for the reason anatomic complete revascularization appeared to be better than the functional approach, Bainey suggested that a functional assessment of an angiographically significant lesion might not reveal ischemia. In the COMPLETE and COMPLETE-OCT trials, almost 50% of the angiographically lesions were thincap fibroatheromas, which lead to CV death and MI, and these lesions would be “captured so to speak with [anatomic] complete revascularization,” said Bainey. A functional assessment, on the other hand, wouldn’t identify vulnerable plaques and they might be missed for revascularization.

Quality-of-Life Analysis

In a second presentation during the same session, Kreton Mavromatis, MD (Emory University School of Medicine, Atlanta, GA), focused on the quality-of-life (QoL) findings stratified by complete revascularization.

Overall, the adjusted QoL findings tended to favor complete over incomplete revascularization, more so for patients with functional complete revascularization, Mavromatis reported. There were improvements in the Seattle Angina Questionnaire 7 (SAQ-7) summary score and angina frequency score, as well as improvements in the Rose Dyspnea Scale among patients undergoing functional complete versus incomplete revascularization. The magnitude of improvement was modest, said Mavromatis, but larger improvements were seen in those with daily/weekly angina at baseline.

Those who underwent complete revascularization—assessed either anatomically or functionally—also tended to have better adjusted QoL scores compared with the conservatively managed patients. The difference in QoL scores favoring invasive therapy over medical therapy were larger than what was seen in the ISCHEMIA trial, Mavromatis reported.

Brahmajee Nallamothu, MD (University of Michigan, Ann Arbor), the discussant following the presentation, said the new analysis suggests that it didn’t matter how complete revascularization was assessed—either functional and anatomic—when it came to QoL measures, but anatomic complete revascularization looked better than the functional approach with respect to hard clinical events in the analysis by Stone.

“Anatomic revascularization may take care of not only functionally significant lesions that cause angina, but also vulnerable plaques that might not be causing angina,” said Mavromatis. “With functional revascularization, it might really focus on those lesions that are causing the angina.”

Note: Stone is a faculty member of the Cardiovascular Research Foundation, the publisher of TCTMD.

Michael O’Riordan is the Managing Editor for TCTMD. He completed his undergraduate degrees at Queen’s University in Kingston, ON, and…

Read Full Bio
  • Stone GW, on behalf of the ISCHEMIA investigators. Impact of completeness of revascularization on clinical outcomes with stable ischemic heart disease treated with an invasive versus conservative strategy: the ISCHEMIA trial. Presented at: ACC 2021. May 17, 2021.

  • Mavromatis K, on behalf of the ISCHEMIA investigators. Impact of completeness of revascularization on quality of life in patients with stable ischemic heart disease: insights from the ISCHEMIA trial. Presented at: ACC 2021. May 17, 2021.

  • Stone reports consulting to Valfix, TherOx, Robocath, HeartFlow, Ablative Solutions, Miracor, Neovasc, Abiomed, Ancora, Vectorius, Elucid Bio, Occlutech, CorFlow, Cardiomech, and Gore; he reports equity/options from Ancora, Cagent, Applied Therapeutics, Biostar family of funds, SpectraWave, Orchestra Biomed, Aria, Cardiac Success, and Valfix.
  • Bainey, Boden, and Mavromatis report no conflicts of interest.