Diabetes still carries excess risk after PCI, even with OCT guidance

Design

Prespecified subgroup analysis of a randomised trial (OCT- vs angiography-guided PCI)

Population

2,487 patients undergoing PCI; 1,044 with diabetes, 1,443 without

Primary outcome

Target vessel failure and serious MACE at 2 years

Effect

TVF 9.9% vs 6.3%, adjusted HR 1.48 (95% CI 1.09-2.02); serious MACE 5.3% vs 2.7%, adjusted HR 1.77 (1.14-2.76)

ILUMIEN IV randomised patients to optical coherence tomography (OCT)-guided or angiography-guided PCI. This analysis split the 2,487 patients by diabetes status: 1,044 with diabetes and 1,443 without, and followed them for two years.

After adjustment, people with diabetes had more target vessel failure (9.9% vs 6.3%; adjusted HR 1.48, 95% CI 1.09 to 2.02) and more serious major adverse cardiovascular events (5.3% vs 2.7%; adjusted HR 1.77, 1.14 to 2.76). The excess was present whether or not the lesions treated were complex, though risk was highest when both applied.

OCT guidance did not reduce either endpoint in the trial as a whole, and that held in both groups. Better stent optimisation in the lab did not remove the diabetes penalty. This analysis did not test what would lower the residual risk, but care after discharge, where the diabetologist has most influence, is a reasonable focus.

• Treat a patient with diabetes who has just had PCI as high risk for repeat events for at least two years.
• Review glucose-lowering therapy after PCI with cardiovascular benefit in mind, including SGLT2 inhibitors or GLP-1 receptor agonists where indicated.
• Check LDL cholesterol, blood pressure and smoking status at the first diabetes visit after the procedure.
• Ask whether complex lesions were treated; that combination carried the highest two-year event rates.

Why it matters

Better intracoronary imaging did not close the outcome gap, so attention after discharge to modifiable risk factors is a reasonable focus, though this analysis did not test it.

Don't overread it

The diabetes comparison is not randomised, so it describes risk rather than showing what would lower it.

The statistics, in plain English

A hazard ratio of 1.48 means that, at any point in follow-up, the rate of target vessel failure was about half as high again in people with diabetes. The confidence interval (1.09 to 2.02) stays above 1, so the difference is unlikely to be chance. The comparison between diabetes and no diabetes is observational within a trial, because diabetes was not randomised; the adjustment reduces but cannot remove confounding.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.