Design: Meta-analysis of randomised trials with trial sequential and Bayesian analyses
Population: 28 trials, 24,634 patients undergoing drug-eluting stent PCI
Primary outcome: Target lesion failure; cardiac and all-cause death, MI, stent thrombosis
Effect: TLF RR 0.72 (0.61 to 0.85); cardiac death 0.73 (0.58 to 0.91); stent thrombosis 0.55 (0.40 to 0.76)

An updated meta-analysis pooled 28 randomised trials of intracoronary imaging (IVUS or OCT) against angiography guidance for drug-eluting stent PCI, covering 24,634 patients with a mean follow-up of 22 months. It added trial sequential, fragility and Bayesian analyses, and used meta-regression to explain why some recent trials were neutral.

Imaging guidance was associated with lower target lesion failure (RR 0.72), cardiac death (0.73), myocardial infarction (0.86), target vessel revascularisation (0.69), stent thrombosis (0.55) and all-cause death (0.82). Heterogeneity was moderate for target lesion failure (I² 61%) but absent for hard endpoints. Differences in geography and lesion characteristics accounted for most of the variation between trials.

This answers the doubt raised when several recent trials failed to show benefit: the pooled picture still favours imaging, most clearly in longer and more complex lesions. For the catheter laboratory, it supports using imaging routinely for left main and complex disease, where the Class I recommendation already sits. In India, cost and catheter availability remain the practical limit.

Use IVUS or OCT for left main, bifurcation, long-lesion and in-stent restenosis PCI
Act on imaging: optimise expansion and treat edge dissection rather than just recording the images
Longer stented segments appear to gain most
Document imaging findings in the procedure report for future reference

Why it matters

The neutral recent trials did not overturn the benefit of imaging; lesion complexity explains much of the difference.

The statistics, in plain English

A relative risk of 0.73 for cardiac death means about 27% fewer deaths with imaging, with the interval (0.58 to 0.91) excluding no effect. I² of 61% for target lesion failure means trials disagreed substantially on that outcome, which is why the meta-regression matters; I² of 0 to 5% for death and stent thrombosis means the trials agreed. A fragility index of 9 to 50 means that many patients would need to switch outcome before significance was lost.