TLR Rates Rise Steadily After Left Main PCI, but Deaths Stay Stable

The need for repeat revascularization following PCI in the left main coronary artery is more likely to occur early after the index procedure, but TLR does not appear to be linked to worse survival in long-term follow-up, according to results from a pooled registry analysis.

When four independent observational studies were analyzed together, the cumulative incidence of target lesion revascularization was 10.8% over roughly a decade, with more than 60% of these procedures taking place within the first 2 years after PCI for left main disease. However, no matter whether TLR was performed early or late, it was not associated with higher mortality.    

“Most of the patients with TLR were re-treated well with repeat PCI or CABG,” senior investigator Duk-Woo Park, MD, PhD (Asan Medical Center/University of Ulsan College of Medicine, Republic of Korea), told TCTMD in an email. “Given that such patients were optimally revascularized with PCI or CABG, the occurrence of TLR was not significantly associated with an increased risk of all-cause or cardiac mortality.”  

Current US guidelines recommend surgery to improve survival in patients with stable ischemic heart disease with significant left main stenosis (class 1, level of evidence B), but PCI is a class 2a indication (level of evidence B) for those with left main disease of low-to-medium anatomical complexity. In Europe, it’s been recently proposed that PCI should be downgraded to a class IIa (level of evidence A) indication for stable patients with a low-to-intermediate SYNTAX score, but that change has not yet been formally adopted.

Park said that data from randomized trials and observational registries have consistently shown PCI to be associated with higher risks of restenosis when compared with CABG surgery. Given the amount of myocardium at risk with left main coronary artery disease, it’s critical to understand the long-term outcomes associated with TLR after left main PCI, he said.

The occurrence of TLR was not significantly associated with an increased risk of all-cause or cardiac mortality. Duk-Woo Park

Other studies have reached different conclusions regarding the impact of TLR on mortality. In 2020, an analysis of patients enrolled in two ISAR-LEFT-MAIN studies showed there was a significant difference in mortality rates between patients who did and did not require TLR following PCI for left main coronary artery disease. At 5 years, mortality was 30.2% for those with restenosis requiring revascularization compared with 17.3% for those who didn’t.

Similarly, data from the randomized EXCEL trial showed that TLR following PCI for left main coronary artery disease was associated with worse outcomes. In that study, the need for repeat revascularization tied to with a twofold higher risk of all-cause mortality and a fourfold higher risk of cardiovascular mortality at 3 years. PCI was associated with a higher rate of repeat revascularization, but the link between TLR and mortality was seen in those who initially underwent PCI and CABG surgery.    

The reasons for the discrepancy between this latest analysis and EXCEL is unclear, say investigators.

“These findings may be due to differences in patient selection in a real-world registry versus a randomized controlled trial, procedural characteristics, or duration of follow-up,” said Park. However, in both the EXCEL and current study, the exact relationship between TLR and mortality is difficult to clearly determine because TLR could also be influenced by multiple confounders that affect mortality, he added.

TLR Climbs Steadily Over Time

For the study, which was led by Tae Oh Kim, MD (Asan Medical Center/University of Ulsan College of Medicine), and published online January 8, 2024, in JACC: Cardiovascular Interventions, the researchers pooled patients treated at Asan Medical Center who were included in four multicenter observational registries, including IRIS-DES, IRIS-MAIN, MAIN-COMPARE, and PRECOMBAT. In total, they analyzed data on 1,397 patients with left main coronary artery disease treated with a drug-eluting stent, of whom 118 required TLR. The median follow-up was 6.8 years (9.6 years for those who underwent TLR and 6.4 years for those who did not).

TLR occurred a median of 1.1 years after PCI. The cumulative incidence of TLR followed a steep early time course—73 of the 118 patients (61.9%) underwent TLR during the first 2 years—followed by a more linear time course over the next 10 years. The incidence of early-stage TLR (within first 2 years) was 2.5 per 100 person-years while the incidence of late-stage TLR (between 2 and 10 years) was 0.60 events per 100 person-years. The most common cause of TLR—as documented by intravascular imaging—was neointimal hyperplasia (69.5%), with a combination of neointimal hyperplasia and stent underexpansion making up the rest.

There was no significant demographic or clinical differences, including the presence of diabetes, tobacco use, chronic renal disease, and ACS presentation, between patients with and without restenosis following left main PCI, but the patients who underwent TLR were more likely to have had initial procedures involving distal bifurcations, use of two stents, and inflation of a final kissing balloon. There was also less use of intravascular ultrasound during the index PCI for those who required TLR. Importantly, there was no difference in anatomical complexity between those who underwent TLR and those who didn’t.

In the overall population, all-cause and cardiac mortality occurred in 19.5% and 16.6% of patients, respectively. The unadjusted rates of all-cause mortality in those with and without TLR were 15.9% and 19.9%, and for cardiac mortality, they were 12.1% and 17.1%. In a multivariate-adjusted model, there were no significant differences in all-cause or cardiac mortality between those with and without TLR. In landmark analyses at 2 years, which excluded the early-stage TLR cases, there was again no significant difference in mortality rates between those who underwent TLR and those who did not.

Not all restenosis is the same. Robert Byrne

Robert Byrne, MBBCh, PhD (Mater Private Network, Dublin, Ireland), who was involved in the ISAR-LEFT-MAIN analyses, said the assumption, generally speaking, is that restenosis in the left main area is more likely to be prognostically significant than restenosis outside of it given the large amount of myocardium in jeopardy. This new analysis, he told TCTMD, “speaks against the strong association [between TLR] with mortality.”

Byrne said there are likely a couple of reasons for the finding.

“One of the explanations is that there's a specific type of restenosis that you see in the left-main stem area, that being restenosis at the ostium of the circumflex,” he said. “If you look at this study, just over 50 percent of the cases had restenosis that seemed restricted only to the ostium of the circumflex. That might explain why there wasn't a strong association with increased mortality. Lesions in the circumflex territory often are of less prognostic relevance than lesions elsewhere.”

In contrast, just 15.4% of TLR cases involved treatment of coronary lesions in the ostium of the LAD, an area that would be more likely to be prognostically significant. “The point being that not all restenosis is the same,” said Byrne. “The majority of cases here were restricted to the ostium of the circumflex [artery] and that can dilute the impact on mortality.”

In an editorial, David Kandzari, MD, and Katherine Kunkel, MD (both Piedmont Heart Institute, Atlanta, GA), put forward another possibility, cautioning that there is a risk for confounding because only patients with angiographically documented TLR were included, potentially overlooking those with unrecognized left main stent failure. 

“Patients experiencing sudden death from left main stent thrombosis or ischemic arrhythmias would be attributed to the non-TLR group, falsely lowering TLR group mortality while raising the non-TLR group mortality,” they write. “This phenomenon could bias this study to finding no difference in mortality when one may exist.

Good Outcomes After TLR

In general, Byrne said the results in patients requiring TLR after an initial PCI for left main disease are quite good, noting there was a marked difference between outcomes in this analysis compared with ISAR-LEFT-MAIN. One explanation for the differing results might be because this latest study is based on registry data, the exception being PRECOMBAT, which was a randomized trial with extended follow-up. This inevitably creates the potential for selection bias. Like Park noted, Byrne also pointed out that this is not a natural history study looking at restenosis and mortality: if restenosis is seen, patients are treated with a stent.

In their editorial, Kandzari and Kunkel write that the present study suggests that TLR following left main PCI is likely more dependent on lesion rather than patient characteristics, which is an important finding because the risks and benefits of PCI are typically applied to individual patients by heart teams. They also hint that if TLR does not carry a negative prognostic significance, maybe its relative weight as an endpoint in clinical trials should be reconsidered.

Byrne agreed, adding that revascularization is a procedure rather than a clinical outcome, one that is more frequent with PCI than with CABG surgery. “The short answer is when we're looking at outcomes of patients with left main stem stenosis and comparing PCI to bypass surgery, for example, I don’t think revascularization should be a component of the primary endpoint,” he said.

Park took a similar view, noting that TLR is a soft endpoint—a point driven home by their finding that it does not impact mortality. As in EXCEL, TLR remains an important secondary endpoint to consider, however, just as stroke after CABG should be taken into account. “A higher TLR after PCI or higher stroke after CABG is [a] natural thing owing to inherent technology and procedures,” he said, adding that patient characteristics, anatomy, other prognostic variables could be used to decide which revascularization strategy is appropriate in left main disease.