Angiographic Radial Wall Strain

Corresponding Author

Although percutaneous coronary intervention has significantly improved outcomes for patients with coronary artery disease, the late stent-related adverse events associated with drug-eluting stents (DES) and the bleeding risks tied to dual antiplatelet therapy persist.¹ These limitations sustain interest in drug-coated balloon (DCB) angioplasty as a "leave nothing behind" strategy.^2,3 Currently, the safety and efficacy of DCBs are widely validated for treating in-stent restenosis (ISR) and selected small-vessel disease4, 5, 6, 7; however, their role in de novo nonsmall-vessel lesions remains unsettled. In the REC-CAGEFREE I (Paclitaxel-Coated Balloon for Treatment of De-Novo Non-Complex Coronary Artery Lesions) trial, a DCB strategy with bailout stenting did not achieve noninferiority to intended DES implantation for the 2-year device-oriented composite endpoint in a broad population with de novo, noncomplex coronary lesions.⁸ The practical challenge is therefore not simply whether a stent can be avoided, but whether a lesion can be prepared safely, achieve adequate lumen gain and flow, and remain clinically durable without a metal scaffold. Dissection is central to that decision, but angiographic assessment alone has notable limitations; integrating plaque composition and biomechanical characteristics into predictive models holds promise for providing unique value in selecting the optimal candidates for DCB therapy. An objective preprocedural marker of vessel response would be attractive, but its intended use and evidentiary threshold must be clearly defined.

In this issue of JACC: Asia, Shi et al⁹ evaluated maximum radial wall strain (RWS_max), an angiography-derived measure of cyclic luminal deformation, as a predictor of acute dissection during a DCB-based treatment strategy for de novo nonsmall-vessel lesions. Overall, 62 patients (32.0%) met the study definition of severe dissection (National Heart, Lung, and Blood Institute type C-F), and 52 of these 62 events (83.9%) were type C. Median RWS_max was higher with severe than with nonsevere dissection (14.70% [Q1-Q3: 12.80%-16.65%] vs 12.20% [Q1-Q3: 11.00%-13.30%]; P < 0.001). Each 1% increase in RWS_max was associated with higher odds of severe dissection in the principal adjusted model (OR: 1.55; 95% CI: 1.31-1.90). RWS_max demonstrated good predictive ability for severe dissection, with an area under the curve of 0.75. At a cutoff of 13.7%, sensitivity was 61.3% and specificity was 81.1%. Adding RWS_max to a model containing lesion and procedural variables increased the area under the curve from 0.76 to 0.85. These findings suggest that preprocedural RWS_max may help identify high-risk lesions prone to dissection before DCB treatment.

Vulnerable plaques, typically characterized by a large lipid core and a thin fibrous cap on intravascular ultrasound, optical coherence tomography, or near-infrared spectroscopy, are identifiable.¹⁰ The appeal of RWS is understandable. Unlike intracoronary imaging assessments, it can be calculated from routine angiographic images without introducing another intracoronary imaging catheter, offering a practical, relatively noninvasive alternative for plaque biomechanical analysis.¹¹ Previous studies have shown that RWS correlates with vulnerable morphological features such as lipid burden, fibrous-cap thickness, and the lipid-to-cap ratio identified by optical coherence tomography, although its discrimination of thin-cap fibroatheroma has been moderate rather than definitive¹¹; higher RWS has also been associated with subsequent vessel-oriented events in deferred nonflow-limiting lesions and with adverse angiographic or clinical outcomes after DCB treatment of small vessels.12, 13, 14 These data support RWS as a marker of an adverse biomechanical phenotype. However, the current study did not systematically perform intracoronary imaging and therefore cannot demonstrate that the lesions with higher RWS_max were thin-cap fibroatheromas or that plaque rupture caused the observed dissections. A mechanistic pathway from plaque composition to high RWS and then to procedural wall injury is plausible and interesting, but remains an inference that requires paired imaging and procedural validation.

This study has several limitations that warrant consideration. First, the investigators classified all type C-F dissections as severe, but the study did not report long-term target lesion failure, periprocedural myocardial infarction linked to dissection, or healing on follow-up imaging. The analysis therefore addresses the probability of an acute angiographic finding during lesion preparation or DCB treatment, not the long-term safety of a stentless strategy. Recent literature suggests that not all dissections are harmful; many nonflow-limiting dissections heal over time through positive vessel remodeling, a phenomenon often termed "therapeutic dissection."^3,15,16 These studies should not be interpreted as evidence that every type C dissection is benign, but rather that an acute angiographic grade is an imperfect surrogate for clinical harm. Because the study does not distinguish between dissections that cause periprocedural myocardial infarctions and those that heal benignly, it cannot definitively predict the long-term clinical success of DCB therapy in nonsmall vessels. Second, the predictive model may not have fully accounted for the confounding effects of procedural techniques. Adequate lesion preparation (eg, using scoring balloons, cutting balloons, or specific inflation strategies) heavily influences the success of DCB angioplasty. The retrospective design makes it difficult to control these unmeasured variables. Finally, as a single-center, retrospective study with a relatively small sample size, the generalizability of the findings is limited. External validation in independent, multicenter cohorts is essential before RWS can be widely adopted in routine clinical decision-making.

Although these findings are not yet sufficient to change current clinical guidelines, they highlight an important physiological concept: selecting nonsmall-vessel lesions for DCB therapy should go beyond traditional anatomical parameters. It must also consider underlying plaque vulnerability and localized biomechanics. At present, RWS_max could be viewed as a hypothesis-generating adjunct that may enrich preprocedural assessment, not as a binary gatekeeper for DCB vs DES. Supporting this shift, a recent study on small-vessel DCB therapy found that risk stratification based on RWS could effectively predict adverse clinical outcomes at 1 year.¹⁴ A high RWS_max could prompt closer attention to lesion morphology, more deliberate preparation, the necessity of intravascular imaging, and a lower threshold for a bailout stent. Conversely, a lower value should not override an unfavorable preparation result. The final device decision should still incorporate the postpreparation result, residual stenosis, distal flow, recoil, dissection morphology, symptoms, and physiologic or intracoronary imaging information when indicated.

Although the present study moves the field toward a more quantitative understanding of procedural vessel response by linking high RWS_max to an increased risk of dissection, current evidence is not yet sufficient to determine whether vulnerable plaques can be safely treated without a stent. Ongoing studies, such as the FAVOR V AMI (Functional and Angiography-Derived Strain Guided Multi-Vessel/Lesion Revascularization Strategy in Patients with Acute Myocardial Infarction; NCT05669222) trial and the NASCENT (Natural History of Coronary Atherosclerosis; NCT06040073) study,¹⁷ may both strengthen the validation of angiography-derived biomechanics, but neither directly proves that RWS-guided device selection improves DCB outcomes. Future prospective, multicenter randomized trials are needed to validate this biomechanical approach. They should also test whether a strategy incorporating RWS information changes operator behavior and improves patient outcomes compared with a standardized strategy guided by anatomy and lesion preparation. Prediction models require internal resampling and independent external validation before a threshold can be used clinically.

Funding Support and Author Disclosures

The authors have reported that they have no relationships relevant to the contents of this paper to disclose.