Off-pump and on-pump bypass surgery had similar outcomes in frail patients
In the 508-patient FRAGILE trial, the 30-day composite of death and major complications did not differ significantly between the two techniques; wound infections were less frequent with off-pump surgery.
| Group | Value (%) |
|---|---|
| Off-pump CABG | 10.9 |
| On-pump CABG | 16 |
Coronary artery bypass grafting can be done in two ways: on a heart-lung machine that stops the heart (on-pump), or on the beating heart without it (off-pump). The debate over which is better is decades old, and it sharpens in older, frailer patients, who face the most to lose from the machine's circulatory stress but also from a technically harder operation. A randomised trial in Circulation now offers a direct comparison in exactly that group — and finds, for its main outcome, no clear winner [s1].
What the trial did
FRAGILE was a multicentre, randomised, open-label trial that enrolled patients aged 60 or older classified as prefrail or frail and scheduled for bypass surgery, between January 2018 and December 2024 [s1]. They were randomly assigned 1:1 to off-pump or on-pump CABG [s1]. The primary outcome was a 30-day composite of death, myocardial infarction, stroke, renal failure, acute respiratory distress syndrome, or reoperation for bleeding [s1]. Secondary outcomes included ventilation time, intensive-care and hospital stay, transfusion needs, completeness of revascularisation, surgical wound infection at 30 days, and neurocognitive function before surgery and at one year [s1]. Analyses followed the intention-to-treat principle [s1]. The trial is registered as NCT02338947 [s2].
A total of 508 patients were included — 258 assigned to off-pump and 250 to on-pump surgery [s1]. Their median age was 68 years (interquartile range 65.0 to 73), and 30.5% were women [s1].
What it found
The primary composite event occurred in 28 off-pump patients (10.9%) and 40 on-pump patients (16.0%) [s1]. Numerically that favours off-pump, but the difference did not reach statistical significance: the relative risk was 1.47, with a 95% confidence interval of 0.94 to 2.31 and a P value of 0.09 [s1]. A confidence interval that crosses 1.0 means the trial could not rule out chance as the explanation for the gap.
One secondary outcome did separate the groups. Surgical wound infection was more frequent after on-pump surgery — 20 patients (8.0%) versus 9 (3.5%); P=0.04 [s1]. There were no significant differences in neurocognitive or quality-of-life outcomes at one year [s1]. Estimated one-year survival was 96.1% in the off-pump group and 92.8% in the on-pump group, a difference that again fell short of significance (P=0.099) [s1].
How to read it
The clean reading is a null result on the primary endpoint, with a hint worth noting rather than banking on. For the composite of death and major 30-day complications, the two techniques performed similarly, and the trial was not able to show that off-pump surgery was better [s1]. The lower wound-infection rate with off-pump is a genuine, statistically significant secondary finding, but secondary outcomes are hypothesis-generating — one positive result among several comparisons can arise by chance, and the trial's own authors frame the overall picture as similar rates on the primary outcome [s1].
A few limits frame it. The trial was open-label, so surgeons and patients knew which operation was performed, which can colour the handling of softer outcomes [s1]. It enrolled a specific population — prefrail and frail patients aged 60 and over — so it does not speak to younger or more robust patients, where the balance may differ [s1]. And at 508 patients, it was powered to detect a sizeable difference in a composite of uncommon events; a smaller but real difference could be present and undetected, which is one way to read the non-significant 10.9%-versus-16.0% gap [s1].
What to watch
The practical question is whether surgical teams read FRAGILE as licence to choose off-pump surgery for frail patients on the strength of the wound-infection signal, or as confirmation that, for the outcomes that matter most, the choice can rest on a surgeon's own skill and judgement. Longer follow-up and any pooling of FRAGILE with earlier off-pump-versus-on-pump trials will matter more than the headline numbers. For now, the finding is specific: in prefrail and frail bypass patients, the two techniques produced similar rates of death and major 30-day complications, with fewer wound infections after off-pump surgery [s1].
This article describes clinical-trial results and is not medical advice. Decisions about surgical technique should be made with the treating surgical team.
Sources
- Off-Pump versus On-Pump Coronary Artery Bypass Grafting in Prefrail and Frail Patients: 30-Day and 1-Year Outcomes From the FRAGILE Randomized Trial — Circulation , November 10, 2026
- Off-pump Versus On-pump Coronary Artery Bypass Grafting in Frail Patients (FRAGILE) (NCT02338947) — ClinicalTrials.gov
More on
A gentler way to harvest bypass veins did not cut heart events at seven years
In the Nordic SWEDEGRAFT trial, the no-touch vein-harvesting technique left the rate of death, heart attack, or repeat procedures at 19.8% versus 20.4% through seven years — a difference indistinguishable from chance.
Double artery grafts show no survival edge at 15 years, trial follow-up finds
In a randomised trial of 3,102 bypass patients across seven countries, all-cause mortality was near-identical between bilateral and single internal thoracic artery grafting.
Does colchicine, an old gout drug, prevent heart attacks?
Two trials found low-dose colchicine cut cardiovascular events, and the FDA approved it for heart risk in 2023. Then the largest trial found nothing, leaving the question genuinely unsettled.
Does cardiac rehabilitation after a heart attack work?
Pooled across 85 trials, exercise-based cardiac rehab cut later heart attacks and hospital admissions but barely moved death rates within a year — a real benefit, and a narrower one than the usual pitch.