The main findings of this post hoc QCA subanalysis of the AIDA trial, investigating the impact of lesion preparation and stent optimisation strategies on long-term LOCE with EES in non-complex coronary anatomy (SYNTAX 11) are:

1.

Lesions predilatated with oversized balloons relative to the RVD were more likely to experience LOCE. The risk increased up to fivefold with an oversized predilatation balloon relative to the RVD and also more than doubled with the LAD as the target vessel. Interestingly, this risk increase is mitigated in the presence of diabetes mellitus.

2.

Compared to non-LOCE lesions, predilatation was typically performed with larger balloons, while postdilatation was conducted with smaller balloons relative to the RVD and stent diameter, respectively, in lesions with LOCE. Smaller vessel diameters were linked to larger predilatation balloon-to-RVD ratios. The low variance suggests that other factors may influence the operators’ choices of balloon size.

Current observations are in line with clinical practice in PCI with current-generation DES [11]. A low LOCE rate of 7.7%, primarily driven by target vessel revascularisation and TV-MI with the majority of events occurring within the first 3 years post-PCI and the very low incidence of definite ST (0.8%), similar to previous reports using the same device, are reassuring observations reflecting good everyday PCI practice [12]. LOCE, driven by TLR in 74% of cases, persist even when the stent struts are anticipated to be fully covered by a new endothelial layer 1 year after initial device implantation, considerably reducing the risk of ST and eliminating the necessity for DAPT [13]. Lesion predilatation was common (92.3%), with 75% of these lesions predilatated with balloons of equal size to the RVD or up to 0.5 mm smaller. Stent sizing matched the Dmax at both edges in 42.8% of lesions, and oversizing (17.7%) or isolated oversizing at the distal stent edge (17.2%) were not uncommon. Nearly half of lesions (49%) were postdilatated. Predilatation and postdilatation were conducted using balloons with a median length of 15 mm at 10 atm and 16 atm, respectively, while the median stent length was 18 mm, implanted at 14 atm. Lesions predilatated with oversized balloons relative to the RVD were more likely to experience LOCE. Interestingly, this increase was partly mitigated in lesions of patients with diabetes mellitus. A different plaque composition likely contributes to different responses to barostressors applied during pre/postdilatation with a subsequent risk for LOCE [14].

The PSP technique for BVS involves predilatation using balloons oversized compared to the RVD and postdilatation with non-compliant balloons up to 0.5 mm larger than the scaffold’s nominal diameter at pressures ≥ 18 atm [15]. When applied to PCI with DES, the traditional PSP strategy elevates the risk of target lesion failure, primarily due to predilatation with oversized balloons [16, 17]. Further, optimal DES sizing (absolute difference between RVD and stent diameter ≤ 0.25 mm) may be a significant protective predictor [16]. Compared to our previous reports, we now adopted a more granular approach to analysing the components of meticulous DES implantation, treating them as continuous predictors rather than categorical variables [17, 18]. In complex CAD (e.g. left main artery, bifurcation lesions, diffuse lesions > 30 mm), IVUS-guided stent optimisation, especially post-dilatation, significantly reduces the risk of cardiac death and target vessel revascularisation [19, 20]. Reasonably, optimal implantation strategies may vary depending on disease complexity.

In lesions of low complexity, careful predilatation preventing intima barotrauma with a subsequent risk for (small) dissections and/or delayed vessel healing may enable effective plaque sealing with a properly sized DES, whereafter stent sizing and postdilatation have limited influence on LOCE risk. In drug-coated balloon treatment, small intima dissections are frequently captured with OCT, where the dissection volumes seem to correlate with late lumen loss [21]. Complex lesions in contrast, especially calcified lesions, may benefit from more aggressive predilatation to ensure optimal stent deployment in terms of strut apposition. Postdilatation using oversized non-compliant balloons at high pressures might improve clinical outcomes by ensuring optimal plaque sealing and stent strut apposition, though careful attention is needed to avoid iatrogenic vessel injury and edge dissection. Particularly in ST-elevation myocardial infarction patients and at distal stent edges in tapered vessels, where slow or no-flow complications can arise following aggressive dilation, caution is paramount [16]. Currently, Teeuwen et al. are performing a multicentre study on the impact of postdilatation in daily clinical practice using the NHR platform, which will provide us insights and further clarify this matter (personal communication).

Thus, improvements in outcomes can be made with relatively little effort during PCI procedures, but is angiography alone sufficient for optimal lesion selection, preparation and subsequent optimal device implantation?

As coronary CT angiography is very likely to become the primary screening tool for suspected CAD, and with computational fluid derived parameters (computed tomography-fractional flow reserve (FFR), angioFFR, quantitative flow ratio (QFR), Murray law-based QFR) derived from various coronary imaging modalities emerging at a rapid pace, our daily practice in treating CAD will very likely adapt to a more image-guided approach combining anatomical and physiological characteristics to improve outcomes [22, 23]. Therefore, results from ongoing clinical trials (P3 trial, P4 trial, FAVOR III Europe-Japan, PIONEER IV and MultiVessel Talent trials) incorporating these strategies in the PCI workflow are eagerly awaited [24,25,26,27].

Limitations

This study’s limitations include its post hoc nature, potential underpowering, the non-mandatory (intracoronary) imaging or implantation strategies in the AIDA trial’s all-comer design, operator-dependent lesion preparation and stent optimisation may have introduced indication bias. Furthermore, the inability to use QCA in 10% of lesions may indicate more complex lesions are underrepresented. Importantly, image quality and angulations were not predefined by protocol. This high exclusion rate is largely explainable, as QCA requires high-quality imaging, correctly angled with sufficient contrast to accurately assess luminal contours for the assessment of the three device-implantation steps.