Meta-analysis supports use of all-cause mortality in PCI versus CABG trials

Mario Gaudino

A meta-analysis of 23 randomised clinical trials comparing percutaneous coronary intervention (PCI) with coronary artery bypass grafting (CABG), published in JAMA Internal Medicine, has concluded that PCI is associated with a higher rate of all-cause, cardiac and non-cardiac mortality compared with CABG at five years. According to the study’s authors, Mario Gaudino (New York-Presbyterian and Weill Cornell Medicine, New York, USA) et al, the significantly higher non-cardiac mortality associated with PCI suggests that even non-cardiac deaths observed after PCI may be procedure related, and the analysis therefore supports the use of all-cause mortality as the endpoint for myocardial revascularisation trials.

Gaudino and colleagues note that during the course of the last three decades, several randomised clinical trials have compared the results of PCI and CABG in patients with ischaemic heart disease and recent coronary syndromes. Individual trials were underpowered to detect differences in mortality, and all used a composite of major adverse cardiac or cardiovascular events as the primary  outcome. “Although mortality was included in the primary composite outcome of all the trials, some used all-cause mortality and others used cardiac mortality,” Gaudino et al note, adding that the use of all-cause mortality reduces the risk of adjudication bias due to incomplete, skewed, or inadequate supporting evidence.

The study team notes that the controversy has been “ignited” by the recent publication of the five-year results of the EXCEL trial, in which PCI was associated with higher all-cause mortality, but the difference between the two groups was not observed when considering cardiac mortality alone. Through the meta analysis, they have sought to evaluate the difference in all-cause and cause-specific mortality in the randomised trials that have compared PCI and CABG for the treatment of patients with coronary artery disease.

The primary outcomes used in the analysis were all-cause and cause-specific—cardiac versus noncardiac—mortality. Subgroup analyses were performed for trials comparing PCI using bare-metal or drug-eluting stents versus CABG and for trials comparing PCI with CABG in patients with left main disease.

Applying exclusion criteria, a total of 425 citations were evaluated, yielding 23 trials from 24 studies. Of these, 18 used all-cause mortality in their composite primary endpoint. In total, the studies included 13,620 patients, 6,829 undergoing PCI and 6,791 undergoing CABG. The mean follow-up duration of the individual studies was 4.5 years (range, 0.5‒11.4 years); the mean age of patients ranged from 60‒71 years; women constituted 1‒40% of the study populations; men, 39.9‒99% and the prevalence of diabetes ranged from 6‒100%.

Gaudino et al report that the pooled results of the studies showed that compared with CABG, PCI was associated with a higher rate of all-cause mortality (IRR, 1.17; 95% CI, 1.05‒1.29), cardiac mortality (IRR, 1.24; 95% CI, 1.05‒1.45) and non-cardiac mortality (IRR, 1.19; 95% CI, 1.00‒1.41).

Discussing the findings, they write: “The outcomes of PCI and CABG have been extensively evaluated, but comparative data on the cause of mortality after these revascularisation procedures are limited. Our meta-analysis of 23 RCTs (13,620 patients) is the first, to our knowledge, to compare all-cause and cause-specific mortality between the two revascularization modalities. Among the included patients, compared with CABG, PCI was associated with higher all-cause, cardiac, and non-cardiac mortality at a mean follow-up of 5.3 years. On subgroup analysis, PCI with drug-eluting stents was associated with higher all-cause, cardiac, and non-cardiac mortality compared with bare metal stents, although the test for interaction did not reach statistical significance.”

The study team adds that observational evidence shows that the causes of mortality after PCI and CABG are predominantly cardiac in the first year after the procedure and non-cardiac in the following years, with the common causes of cardiac mortality including: cardiogenic shock, heart failure, stent thrombosis, bleeding, coronary dissection, malignant arrhythmia, and sudden death; whereas cancer, sepsis, bleeding, and vascular, pulmonary, and/or renal disease are among the most frequent causes of non-cardiac mortality.

They note: “Our finding of higher all-cause mortality with PCI is consistent with the most recent individual patient data meta-analysis of 11 RCTs and 11,518 patients. Compared with PCI, CABG offers additional protection against the evolution of lesions that were non-flow limiting at the time of the procedure, and this has been proposed as a potential mechanism for the observed survival benefit in the surgical arm.”

Gaudino and colleagues found PCI to be associated with significantly higher non-cardiac mortality compared with CABG, with all six revascularisation randomised trials during the last decade having shown PCI to be associated with an increase in the rate of non-cardiac mortality compared with CABG. “Large observational studies have shown an increased risk of non-cardiac mortality in the late post-PCI period and independent of patients’ characteristics,” they write, adding: “This finding may have several explanations. Dual antiplatelet therapy has been linked to non-cardiac-related deaths in a large trial, but not in an individual data meta-analysis. Evidence suggests that longer duration of dual antiplatelet therapy is associated with an increased risk of non-cardiac mortality. The reasons for these associations are not fully understood, but may include deaths due to major bleeding events (that are often coded as non-cardiac) or a higher bleeding-related mortality in case of trauma or other acute events in patients receiving dual antiplatelet therapy.”

A more likely explanation for this finding, they add, is that cardiac deaths were coded as non-cardiac owing to bias or errors. “The risk that insufficient or inadequate supporting data and/or assessors bias may lead to misclassification in the adjudication of cause-specific mortality is well known and has been described in detail previously. This may be particularly true in case of sudden cardiac death, a particularly frequent cause of death in patients who presented with acute coronary syndromes.”

They go on to add that the findings will likely have implications for future randomised trials comparing PCI and CABG, and that based on the outcomes of their analysis, the use of cardiac mortality may exclude deaths that are related to the procedure, either through non-cardiac mechanisms or because of misclassification.


Please enter your comment!
Please enter your name here