Management of unprotected left main coronary disease with PCI versus CABG: a single-center retrospective study
T Canova, B Lewis, B Johnson, A Arghami, M Singh, B IngrahamAbstract
Background
Randomized trials have demonstrated similar long-term mortality with percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG) in unprotected left main coronary artery disease (ULMCAD). However, contemporary real-world outcomes remain unclear.
Purpose
To compare in-hospital mortality (IHM) and 5-year mortality (5-yr) between PCI and CABG in a contemporary ULMCAD cohort, including patients with ejection fraction ≤35% (HFrEF), diabetes, and surgical turndown status, and to evaluate outcomes among PCI patients treated with IVUS guidance or Impella support, as well as temporal trends in revascularization.
Methods
We performed a retrospective study of patients with ULMCAD treated at our clinic from 2009–2024. Baseline demographic, comorbidity, laboratory, procedural, and outcome data were collected. Primary outcomes were IHM and 5-yr mortality. Univariate analyses compared PCI and CABG overall and among patients with HFrEF or diabetes. Exploratory subgroup analyses assessed 5-yr by surgical candidacy (PCI despite surgical acceptance, surgical turndown, and CABG). Additional univariate analyses evaluated PCI patients with HFrEF treated with versus without Impella and PCI with versus without IVUS guidance. Temporal trends were assessed using annual PCI and CABG utilization rates (per 100 patients) and the Mann–Kendall test.
Results
A total of 4,366 patients met inclusion criteria, of whom 727 (16.7%) underwent PCI and 3,639 (83.3%) underwent CABG. Compared to CABG, PCI showed lower IHM (OR 0.54, 95% CI 0.27–0.97; p=0.04) but a significant increase in 5-yr mortality (OR 2.94, 2.45–3.53; p<0.01; Pic. 1A). 5-yr was higher in the PCI subgroup compared with CABG among patients with HFrEF (OR 2.71, 95% CI 1.89–3.89, p<0.001) and diabetes (OR 3.06, 95% CI 2.31–4.06, p<0.001). Among PCI patients, 489 (56.3%) underwent surgical evaluation; 29.3% underwent PCI despite CABG recommendation, while 70.7% were turned down for surgery. Picture 1B demonstrates 5-yr mortality in these subgroups compared with CABG (p<0.0001). In the PCI subset, IVUS demonstrated a trend toward reduced 5-yr (OR 0.75, 95% CI 0.54–1.04; p=0.079; Pic. 1C), whereas no significant difference was observed with Impella use among patients with HFrEF (OR 0.74, 95% CI 0.36–1.56; p=0.432; Pic. 1D). Over the study period, PCI utilization increased from 0.45 to 7.51 per 100 patients (p=0.022), while CABG declined from 50.0 to 17.5 per 100 patients (p<0.001; Pic. 2).
Conclusions
In this retrospective cohort, PCI utilization increased while CABG declined. Although CABG was associated with improved 5-yr overall, IHM was lower with PCI. Among PCI patients who underwent surgical evaluation, exclusion of those deemed unsuitable for surgery (70.7%) resulted in similar outcomes through three years, with late separation favoring CABG. A trend toward reduced 5-yr was observed with IVUS, with no significant difference seen with Impella.Picture 1For image description, please refer to the figure legend and surrounding text.Picture 2For image description, please refer to the figure legend and surrounding text.