Steep Rise in PCI for Ischemic Heart Failure Seen in Sweden, but Survival Favors CABG
Observational study limitations aside, the findings should serve as a warning: there’s no evidence to support PCI in these patients.
ATHENS, Greece—Swedish patients with stable ischemic heart failure who have undergone revascularization over the better part of the last two decades were more likely to survive if treated with CABG rather than PCI, a new retrospective analysis from the Swedish Coronary Angiography and Angioplasty Register (SCAAR) shows.
Elmir Omerovic MD, PhD (Sahlgrenska University Hospital, Gothenburg, Sweden), who presented the study during a late-breaking clinical trial session of the European Society of Cardiology 2019 Heart Failure Congress, also pointed to the wide variation between centers. Some, he noted, are clearly following guideline recommendations that CABG be the default strategy in this patient group, while others have clearly decided that PCI can be used as an alternate—even preferred—treatment strategy.
The 2018 European revascularization guidelines were updated last year to make myocardial revascularization with CABG in coronary artery disease patients with heart failure and LVEF ≤ 35% a class I recommendation. That change comes in the wake of the STICH 10-year results, as reported by TCTMD.
“Our study supports the current European and American guidelines, in which CABG is a preferred method for revascularization of patients with congestive heart failure due to ischemic heart disease,” Omerovic said.
This latest SCAAR analysis looked at all patients with congestive heart failure (CHF) but angina-free who underwent coronary angiography between 2000 and 2018 in Sweden, restricting the analysis to 2,509 patients with multivessel or left main disease. To compare outcomes between patients revascularized with PCI versus surgery, investigators adjusted for a wide range of covariates including age, sex, diabetes, severity of coronary disease, smoking status, hypertension, hyperlipidemia, creatine clearance, body mass index, previous myocardial infarction, previous PCI, previous CABG, and calendar year.
Rising Tide of Revascularization for CHF
In all, 1,409 (56.2%) patients underwent PCI over the study period and 1,100 (43.8%) had CABG. Patients undergoing PCI were younger, more likely to have hypertension, hyperlipidemia, previous MI, and previous revascularization (both PCI and CABG).
Over the 19 years of study, the number of revascularized CHF patients increased steeply at a rate of 7.5% per year. The rise in PCIs was even more dramatic, increasing at a rate of 13.6% per year. CABG rates, by contrast, have remained relatively static over the same period.
Strikingly, not all hospitals show the same patterns. While the vast majority of hospitals in Sweden that offer both procedures perform more PCIs than CABG in this patient population, eight hospitals have CABG as their default procedure and in half of these the number of CABG surgeries is double the number of PCIs.
Over a median follow-up of 1,429 days, just over 1,000 patients died, with more deaths occurring among the PCI-treated patients. Compared to PCI over the whole period, bypass surgery was associated with a lower adjusted 10-year risk of mortality (HR 0.79; 95% CI 0.68-0.93). Curves began to diverge at 3 years, Omerovic noted, prompting investigators to do a landmark analysis that confirmed no difference in event rates during the first 3 years. There was, however, a swift separation at around the 3-year mark, such that the difference in risk between years 3 and 10 was even more stark than seen in the overall analysis (HR 0.68; 95% CI 0.53-0.87).
Limitations and Patience
Speaking with TCTMD, Omerovic stressed the many limitations of using registry data to look at this issue. “This data happens to support the STICH trial,” he said, by showing a long-term mortality benefit from CABG. “We have numbers, and we have events.” But what appears to be happening is that operators are extrapolating from STICH and using PCI as a stand-in for surgery in CHF patients, Omerovic said. “In the US, studies suggest they are respecting guideline recommendations for STICH, and CABG is a default strategy. But in Sweden, PCI is obviously being used more and more,” and there is currently no comparative data to support that growth.
STICH trial investigator Christopher O’Connor, MD (Duke Clinical Research Institute, Durham, NC), the discussant for Omerovic’s SCAAR analysis, called it “an outstanding contribution to the literature,” given the dearth of clinical trial data in this space. This is true, he said, despite suffering from all of the shortcomings of observational studies.
The impossibility of controlling for all of the reasons patients in a registry might have been treated with PCI is the pivotal issue, he noted. “You saw a number of differences that suggested that the PCI group was a sicker group: it was older, it had more risk factors, and it had more previous revascularization with coronary artery bypass grafting.” O’Connor also pointed out that another key covariate to control for would be the indication for angiography, given that patients analyzed were angina-free—data not provided in Omerovic’s presentation.
These might be patients needing perfusion studies, O’Connor told TCTMD, or they might diabetic—more than one-third of the patients in this SCAAR snapshot had diabetes, so they might have not had classic chest pain symptoms yet still had ischemia. Moreover, “it’s relatively common to evaluate new onset heart failure and worsening heart failure with angiography,” O’Connor explained.
Like the US, Sweden is adopting PCI as a strategy for revascularization in heart failure without the evidence, and I think we are doing harm to our patients with that. Christopher O’Connor
In contrast with Omerovic’s remarks about adherence to guidelines among US operators, O’Connor said he’s seen for himself the rising use of PCI for patients with CHF and reduced LVEF. “The real danger in my view is that like the US, Sweden is adopting PCI as a strategy for revascularization in heart failure without the evidence,” he commented, “and I think we are doing harm to our patients with that.”
Instead, he urged, physicians should wait for the results of the ongoing randomized trial of PCI versus medical therapy in heart failure known as REVIVED-BCIS2, out of the United Kingdom. That trial will finish enrolling patients March 2020 and should report out in 2022, O’Connor said. “Until we have the randomized trial data, we should make the commitment to continue to recommend bypass surgery as the preferred revascularization strategy.”
Interventionalists should take note, O’Connor emphasized to TCTMD. “I know that the interventional cardiologists feel comfortable getting patients through the acute setting, but it’s not the acute setting that we’re worried about, it’s the long-term advantage of CABG.”
Omerovic, E. Comparative effectiveness of CABG and PCI in patients with ischemic heart failure. Presented at: European Society of Cardiology 2019 Heart Failure Congress. May 27, 2019. Athens, Greece.
- Omerovic reports having no conflicts.