For people with left main coronary artery disease (LMCAD) of low or intermediate complexity there is no significant difference in major outcomes at five years between PCI and CABG, latest findings from the EXCEL trial show.

Composite rates of death, stroke and myocardial infarction were similar at five years for patients who were randomised to revascularisation with either PCI or CABG, according to results published in the NEJM.

In the study that involved 1905 patients with LMCAD  (SYNTAX score less than or equal to 32) there was no significant difference in the primary outcome of death, stroke or myocardial infarction for those who everolimus-eluting stents (22.0%) or CABG (19.2%).

This finding was consistent across subgroups including patients with diabetes and patients with lower and higher SYNTAX scores.

The study investigators said the findings showed there was an early benefit of PCI but a later crossover with CABG after about a year, resulting in similar long term outcomes

“By 5 years, the early benefit of PCI due to reduced periprocedural risk was attenuated by the greater number of events that occurred during follow-up than with CABG, such that the cumulative mean time free from adverse events was similar in the two treatment groups,” they noted.

For secondary outcomes, there was a slightly higher risk of death from any cause in the PCI group than in the CABG group (13.0% vs. 9.9%) but not in cardiovascular death (5.0% vs 4.5%).

Cerebrovascular events were more frequent after CABG than PCI  (5.2% vs 3.3%), driven more by transient ischaemic attacks, and the incidence of stroke was not significantly different between the two groups (2.9% and 3.7%). Conversely, ischaemia-driven revascularisation was more frequent after PCI than after CABG (16.9% vs. 10.0%), but only 1 of 25 patients initially treated with everolimus-eluting stents underwent CABG within 5 years.

Presenting the results simultaneously at the Transcatheter Cardiovascular Therapeutics (TCT) scientific symposium in San Francisco, lead investigator said the study follow up would be extended to ten years.

“Ten-year or longer follow-up is required to characterise the very late safety profile of PCI and CABG as both stents and bypass grafts progressively fail over time,” he said

Co-investigator Professor Tony Gershlick from the University of Leicester, UK, said the five year data could be used in multidisciplinary team meetings and in discussions with patients about long term outcomes

“You can say that in terms of cardiovascular death, MI and stroke, there are no differences between the two [but] you might want to mention the little signal of the overall mortality,” he said.