Real-world intravascular ultrasound (IVUS) use in percutaneous intervention-naïve patients is determined predominantly by operator, patient, and lesion characteristics

BACKGROUND: Intravascular Ultrasound (IVUS) has been shown to improve clinical outcomes in patients undergoing percutaneous intervention (PCI) in numerous trials. However, it is still underutilized outside of trial settings, and most trials include a significant proportion of patients with prior PCI. The aim of this study is to look at real-world use and outcomes in PCI-naïve patients who undergo IVUS-guided intervention. METHODS AND RESULTS: Prospectively collected data from 10,574 consecutive patients undergoing their index PCI was retrospectively analyzed. 455 (4.3%) patients underwent IVUS, with a median follow-up of 4.6 years. Patients undergoing IVUS had higher levels of comorbidities including diabetes (27.5% vs. 19.7%, p < 0.001), hypertension (58.0% vs. 47.9%, p < 0.001), hypercholesterolemia (51.6% vs. 39.2%, p < 0.001) and were generally older (65.9 ± 14.5 vs. 64.5 ± 13.4 years, p = 0.031) with higher mean baseline creatinine levels (95.4 ± 63.3 vs. 87.8 ± 46.1 μmol/L). The strongest predictor of IVUS use was the operating consultant graduating from medical school after the year 2000 [OR 14.5 (3.5–59.8), p < 0.001] and the presence of calcific lesions [OR 5.2 (3.4–8.0) p < 0.001]. There was no significant difference in MACE nor 1-year mortality between patients undergoing IVUS-guided or angiography-only PCI on unadjusted analysis [OR 1.04 (0.73—1.5), p = 0.81, OR 1.055 (0.65–1.71) p = 0.828] nor mortality throughout the study period (HR 0.93 (0.69—1.26), p = 0.638). This held true for stents longer than 28 mm. Propensity matched analysis of patients similarly showed no mortality difference between arms for all patients and those with longer stents (p = 0.564 and p = 0.919). CONCLUSION: The strongest predictors of IVUS use in PCI-naïve patients are the operator’s year of graduation from medical school and proxy measures of calcific lesions. On both matched and adjusted analysis there was no evidence of improved mortality nor reduced MACE in this specific retrospective cohort, although this may well be explained by significant selection bias.