Early and Long-Term Outcomes of Minimally Invasive Direct Coronary Artery Bypass in Elderly Patients: A Propensity...

Background: Advancing age is incorporated as a strong risk variable in EuroSCORE II and is consistently associated with adverse outcomes after conventional coronary artery bypass grafting (CABG). Whether minimally invasive direct coronary artery bypass (MIDCAB)-which avoids both sternotomy and cardiopulmonary bypass-modifies this age-related risk in patients with single-vessel or LAD-predominant coronary artery disease remains insufficiently characterised. Methods: We retrospectively analysed 350 consecutive patients who underwent MIDCAB at Hannover Medical School between July 1999 and April 2025 (follow-up to April 2025). Elderly patients (age ≥70 years; n = 117) were compared with younger patients (age <70 years; n = 233) before and after 1:1 propensity score matching using greedy nearest-neighbour matching with a caliper of 0.2 × SD of the logit propensity score; age was excluded from the propensity model as it represented the exposure variable. A pre-specified sensitivity propensity model that additionally excluded EuroSCORE II (because EuroSCORE II contains an age component) was also evaluated. The primary endpoint was all-cause long-term mortality; secondary endpoints included perioperative complications and in-hospital outcomes. Long-term survival was assessed by Kaplan-Meier analysis and multivariable Cox proportional hazards regression, with a pre-specified parsimonious Cox model (age, LVEF, renal impairment) and cluster-robust standard errors on matched-pair identifiers. Results: Matching produced 109 pairs with excellent covariate balance (all standardized mean differences < 0.20). MIDCAB was completed without intraoperative conversion in all patients. No 30-day mortality, perioperative stroke, or new postoperative dialysis was observed in either age stratum (Clopper-Pearson 95% CI 0.00-3.33% for each zero-event outcome). Perioperative complications-including new-onset atrial fibrillation, re-exploration for bleeding, and intensive care unit and hospital length of stay-did not differ significantly between elderly and younger patients in the matched cohort (Newcombe 95% CI for risk differences all crossing zero; McNemar's exact tests non-significant for all matched-pair binary endpoints; Hodges-Lehmann median difference for hospital length of stay +1.0 day, bootstrap 95% CI 0.0-1.0 days). At a median follow-up of 19.0 years (IQR 11.5-24.3; reverse Kaplan-Meier potential median 19.7 years), all-cause mortality was higher in the elderly (20.2% vs. 5.5%, p = 0.002; log-rank p = 0.001; exact McNemar's p = 0.0025 for the matched-pair mortality endpoint). After multivariable adjustment, elderly age (≥70 years) was independently associated with long-term mortality (adjusted hazard ratio 4.48, 95% CI 1.79-11.20, p = 0.001), as was EuroSCORE II (HR 2.40 per unit, p = 0.034); a pre-specified parsimonious model (age, LVEF, renal impairment) with pair-cluster robust standard errors yielded an essentially identical adjusted HR for elderly age of 4.08 (95% CI 1.65-10.04, p = 0.002), and a sensitivity propensity model without EuroSCORE II yielded HR 3.95 (95% CI 1.68-9.29, p = 0.002). Conclusions: In this propensity-matched analysis with a median follow-up of ~19 years, MIDCAB was associated with excellent observed perioperative outcomes in appropriately selected elderly patients (no 30-day mortality, stroke, or new dialysis observed; upper 95% CI 3.3%) and no evidence of an excess of major in-hospital complications compared with younger patients within the statistical resolution of the cohort. The long-term mortality excess in the elderly is consistent with age-related life-expectancy curves in the source population; cause-specific mortality was not available in this cohort. External benchmarks from large MIDCAB cohorts in which long-term survival approximates that of the age-matched general population support this interpretation indirectly. These findings support MIDCAB as a feasible revascularization strategy associated with favourable observed early outcomes and long-term survival consistent with published MIDCAB literature, in appropriately selected elderly patients with single-vessel or LAD-predominant coronary artery disease treated at experienced centres.