A New Method to Optimize Stent Deployment by High-Definition Intravascular Ultrasound

Objectives. Optimal stent deployment by intravascular ultrasound (IVUS) improves outcome, but it can only be achieved in 50% of patients. We investigated the feasibility and effect of a new method of stent optimization on optimal stent deployment. Methods. IVUS analyses of 168 coronary segments were performed after angiography-guided stenting (AGS) and stent optimization in 29 patients (30 lesions). Minimum stent area (MSA), stent volume index (SVI), lumen area, external elastic membrane (EEM), and plaque burden (PB) were measured.  Stent optimization included post-stent dilation with a balloon sized by high-definition (HD)-IVUS to the distal reference EEM diameter for stent underexpansion or malapposition, and stenting of PB >50% or edge dissection. Results. After AGS, stent deployment was suboptimal in 77% of patients. After stent optimization, MSA and SVI were significantly larger than AGS. Adequate stent expansion — defined as MSA ≥5.4 mm² or ≥90% of distal reference lumen area — was significantly higher after stent optimization vs AGS (87% vs 56%, respectively; P=.02).  Optimal stent deployment — a composite of adequate stent expansion, no malapposition, PB <50% at the stent edges, and no edge dissection — was markedly higher after stent optimization vs AGS (87% vs 35%, respectively; P<.01). Conclusion. After stent deployment and postdilation, stent results were suboptimal in two-thirds of patients. This simple online stent optimization by HD-IVUS was feasible and resulted in optimal stent deployment in the majority of patients. Randomized studies are warranted to compare the rate of optimal stent deployment and outcomes of this strategy vs other techniques.

J INVASIVE CARDIOL 2021;33(7):E532-E539.

Key words: angiography-guided stenting, high-definition intravascular ultrasound, optimal stent deployment, stent optimization

Intravascular ultrasound (IVUS) has a pivotal role in detecting suboptimal stent results, including stent underexpansion and malapposition, plaque burden (PB) >50% at the stent edges, and major dissection.^1-7

Optimal stent deployment with IVUS improves outcomes, and can be achieved in approximately 50% of patients. Recently, a number of strategies for stent optimization with IVUS have been proposed, including stent and post-stent balloon dilation sized to the reference segment’s lumen dimensions;⁸ post-stent dilation with a balloon sized to the distal reference lumen diameter with expansion of proximal and mid stent by 0.25 mm;⁹ and post-stent dilation with a balloon sized to the external elastic membrane (EEM) diameters of the proximal reference, distal reference, or lesion site, rounded down by 0.5 mm.¹⁰ In a large, randomized IVUS study,¹¹ the stent optimization strategy was left to the operator’s discretion.

Given that optimal stent deployment can be achieved in 50% of patients, we investigated the feasibility and effect of a new method of stent optimization on the rate of optimal stent deployment.

Study population. Between August 2018 and January 2019, 55 serial high-definition (HD)-IVUS examinations were performed after successful angiography-guided stenting (AGS) and stent optimization in 29 patients (30 lesions). Inclusion criteria consisted of patients with stable and/or unstable angina, or patients with non-ST segment elevation myocardial infarction undergoing percutaneous coronary intervention (PCI). Exclusion criteria were as follows: previous history of coronary artery bypass graft surgery; in-stent restenosis; renal failure (serum creatinine level >2.5 mg/dL); chronic total occlusion; acute myocardial infarction; coronary bifurcation lesions requiring 2-stent strategy; ostial lesions; the presence of heavy calcification by angiography; and left main coronary artery stenosis. The protocol was reviewed and approved by the institutional review board of the University of Alabama-Birmingham and informed consent was obtained from all patients.

Experimental protocol. The study flow chart is shown in Figure 1. All patients underwent predilation with a non-compliant balloon and stent deployment was guided by angiography. Stent diameter and length were selected by visual estimation, with a ratio of stent/vessel diameter of 1.1:1.0. Postdilation was performed with a non-compliant balloon (balloon/stent diameter = 1.0:1.0) using high-pressure balloon inflation. After satisfactory AGS, defined as TIMI 3 flow with no dissection or perforation, and a diameter stenosis <20%, a 60 MHz HD-IVUS imaging catheter (Kodama HD-IVUS system; ACIST Medical) was advanced >10 mm distal to the stent and pulled back to a point >5 mm proximal to the stent using motorized automatic pullback at a speed of 1 mm/s. After IVUS pullback, online IVUS measurements were performed. The orthogonal measurements of the minimum and maximum EEM diameters were averaged at the distal reference segment using the frame with the largest lumen and smallest plaque burden. We did not perform baseline IVUS imaging because we used the distal reference EEM diameter for postdilation balloon sizing, which does not change after stenting vs baseline. In patients with stent underexpansion (defined as MSA <5.4 mm² or <90% of distal reference lumen area; DRLA)^1,3 or malapposition (defined as separation of at least 1 stent strut from the vessel wall),^4,5 post-stent dilation was performed with a non-compliant balloon sized by the HD-IVUS to the distal reference average EEM diameter rounded down to the nearest 0.25 mm, as we previously reported in an optical coherence tomography (OCT) study.¹² We sized the postdilation balloon to the average distal reference EEM diameter by HD-IVUS because it is smaller than in the proximal reference segment. In patients with a plaque burden >50% at the 5 mm proximal or distal stent edges,⁶ or significant edge dissection (defined as the lumen area within the dissection <5.1 mm² or the presence of intramural hematoma),⁷ an additional stent was deployed that was sized to the distal reference lumen diameter and expanded with a balloon sized to the distal reference EEM diameter.

After stent optimization, HD-IVUS examination was repeated and images were assessed for stent underexpansion, PB >50%, edge dissection, and stent malapposition. After postdilation guided by HD-IVUS, if the stent was still malapposed, postdilation with a non-compliant balloon sized to the distal reference EEM diameter was repeated at the malapposition site using higher inflation pressure. Based on the protocol, if the stent was still underexpanded after postdilation based on the above protocol, no further postdilation was performed because of risk of vessel perforation.

Quantitative IVUS analysis. Quantitative analysis of the IVUS images was performed offline at the University of Alabama IVUS core lab using computerized planimetry software (EchoPlaque 2.5; Indec Systems), as previously described.^13,14 Quantitative offline IVUS analysis of 168 segments was performed after AGS vs HD-IVUS optimization. For quantitative IVUS analysis, the lumen, stent, and EEM areas were analyzed every 1.0 mm throughout the stent and within the 5.0 mm of the proximal and distal stent edges before a major sidebranch. To standardize for different lengths, stent volume index (SVI), plaque plus media (P+M) volume index, and EEM volume index were calculated as their volume divided by length, as previously described.^13,14P+M area was defined as EEM – lumen or stent cross-sectional area (CSA). PB was calculated as (EEM CSA – lumen CSA)/EEM CSA. Minimum lumen area (MLA) and minimum stent area (MSA) were defined as the smallest lumen area within the length of the lesion or stent, respectively.

Endpoints. The primary endpoint of the study is optimal stent deployment, defined as a composite of 4 IVUS criteria as follows: (1) adequate stent expansion (MSA ≥5.4 mm² or ≥90% of DRLA), as previously reported;^1,3 (2) complete apposition of the stent struts to the vessel wall;^4,5 (3) PB <50% at the 5 mm proximal or distal stent edges;⁶ and (4) no major edge dissection.⁷ The above measurements were performed offline using the quantitative IVUS analysis. The secondary endpoints were as follows: feasibility of the study; major adverse events at 30 days; stent diameter; final post stent balloon diameter; and post-stent balloon inflation pressure.

Quantitative coronary angiography (QCA). QCA was performed offline at baseline, post angiography-guided stenting, and post HD-IVUS optimization using validated commercially available edge-detection software (QCACMS, version 5.2; Medis Medical Imaging Systems), as previously described.¹³

Statistical analysis. This was an exploratory study to investigate the feasibility of the stent optimization strategy with the HD-IVUS system. Since there are no published data, the sample size of the study was arbitrarily determined. Based on the Shapiro-Wilk test, clinical characteristics and IVUS data were not normally distributed. Therefore, the above data were compared before and after optimization using the Wilcoxon signed-rank test and are presented as median (interquartile range [IQR]). The QCA data were compared using the Friedman test and significance was determined after multiple pairwise comparisons with the Dunn’s post test. QCA data are reported as median (IQR). Categorical variables were compared using the McNamara’s test. P-value <.05 was considered statistically significant. Statistical analysis was performed with SPSS, version 24 (IBM) and MedCalc, version 17.6 (MedCalc).

Baseline patient characteristics. Clinical and angiographic characteristics of all patients are displayed in Table 1. The majority of lesions were American Heart Association/American College of Cardiology lesion types B1 and B2.

Stent results after AGS. After AGS, stent deployment was suboptimal in 23 patients (77%), including stent underexpansion in 10 patients (44%), malapposition in 7 patients (30%), PB >50% in 5 patients (22%), and edge dissection in 1 patient. After stent optimization, optimal stent deployment was achieved in all patients, except in 2 patients in whom stents remained underexpanded.

Procedural characteristics. The procedural characteristics of patients are shown in Table 2. The final post-stent dilation balloon diameter was significantly larger after HD-IVUS optimization vs AGS. The final post-stent balloon inflation pressure was significantly higher after HD-IVUS optimization vs AGS. After HD-IVUS optimizations, 5 additional stents were deployed for PB >50% or edge dissection.

Results of quantitative IVUS analysis. Table 3 displays quantitative offline IVUS analysis after AGS vs stent optimization. In the stented segment, MSA and SVI were significantly larger, but P+M area and P+M volume index were significantly smaller after stent optimization vs AGS. Likewise, in the stented segment, average stent area (ASA) and EEM volume index were significantly larger after stent optimization vs AGS. In the distal reference segments, PB, EEM area, P+M area, lumen area, and EEM diameter were not significantly different between the groups. In the proximal reference segment, plaque burden and P+M area were significantly smaller, but lumen area was larger post stent optimization vs post AGS.

Results of QCA. The results of QCA analysis at baseline, post AGS, and post stent optimization are shown in Table 4. At baseline, minimum luminal diameter and reference vessel diameter were significantly smaller, diameter stenosis was higher, and lesion or stent length was shorter vs measurements after AGS and stent optimization. After stent optimization, minimum luminal diameter was larger, diameter stenosis was lower, and lesion or stent length was longer vs measurements at baseline and post AGS.

Case examples of stent optimization after AGS

Case 1. After stenting of the mid left circumflex (LCX) coronary artery and postdilation with a 3.0 mm non-compliant balloon, the angiogram shows satisfactory result (Figure 2A and Figure 2B). HD-IVUS shows that the stent is underexpanded, with MSA of 4.66 mm² (Figure 2C) and malapposed struts (Figure 2D). Distal reference EEM diameter measured 3.61 mm (Figure 2E). After repeat postdilation with a 3.5 mm non-compliant balloon sized to the distal EEM diameter, the MSA significantly increased to 6.36 mm² (Figure 2F) and malapposition resolved (Figure 2G and Figure 2H).

Case 2. After stenting of the proximal and mid left anterior descending (LAD) coronary artery with a 3.0 x 34 mm stent and postdilation with a 3.5 non-compliant balloon, the angiogram shows satisfactory result (Figure 3A and Figure 3B). HD-IVUS shows large intramural hematoma proximal to the stent edge and malapposition at the stent edge (Figure 3C and Figure 3D). After deploying a 4.0 x 8 mm stent to the proximal stent edge, HD-IVUS shows resolution of intramural hematoma and malapposition (Figure 3E). Final angiogram shows an excellent result (Figure 3F). PB at the distal edge of the stent was 42%, which is not significant and was not stented (Figure 3G).

Adequate stent expansion. As shown in Figure 4, the percentage of patients with an MSA ≥5.4 mm² was significantly higher post stent optimization vs post AGS (78% vs 39%, respectively; P<.01); the percentage of patients with an MSA ≥90% of the DRLA was significantly higher post stent optimization vs post AGS (43% vs 17%, respectively; P=.03); and the percentage of patients with adequate stent expansion (MSA ≥5.4 mm² or ≥90% of DRLA) was significantly higher post stent optimization vs post AGS (87% vs 56%, respectively; P=.02).

Primary outcome, optimal stent deployment. The primary outcome of the study is shown in Figure 5. The percentage of optimal stent deployment, defined as a composite of MSA ≥5.4 mm² or ≥90% of the DRLA, complete strut apposition, PB <50%, and no edge dissection, was significantly higher post stent optimization vs post AGS (87% vs 35%, respectively; P<.01). After stent optimization, optimal stent deployment was achieved in all patients, except 2 patients in whom stents remained underexpanded.

Secondary outcomes. After stent optimization, there was no vessel perforation, slow flow, or acute closure. All patients were discharged the day after the procedure. There was no major or minor bleeding. There were no deaths, readmissions, or myocardial infarctions at 30-day follow-up, except 1 patient who did not take clopidogrel after discharge and was readmitted with subacute stent thrombosis 72 hour post discharge. She underwent balloon angioplasty with successful results.

The main findings of this study are summarized as follows: (1) after stent deployment and postdilation, stent results were suboptimal in the majority of patients; (2) MSA and SVI were significantly larger after stent optimization vs AGS; (3) adequate stent expansion, defined as MSA ≥5.4 mm² or ≥90% of DRLA, was significantly higher post HD-IVUS optimization vs post AGS; and (4) optimal stent deployment, defined as a composite of adequate stent expansion, no malapposition, PB <50% at the stent edges, and no edge dissection, was markedly higher post  stent optimization vs post AGS.

Stent optimization strategies. A number of stent optimization techniques have been recommended. Song et al⁸ recommended that IVUS-guided stenting and post-stent dilation be sized to the reference segment lumen dimensions. The expert consensus document of the European Association of Percutaneous Cardiovascular Interventions⁹ recommended that stent sizing be based on the distal reference lumen diameter with subsequent expansion of the mid and proximal stent segments by 0.25 mm. Maehara et al¹⁰ recommended that IVUS-guided stent or post-stent balloon dilation sizing be based on either: (1) EEM diameters of the proximal reference, distal reference, or lesion site, rounded down by 0.5 mm; or (2) reference lumen diameters. In a number of large IVUS trials,^1,11 stent optimization was left to the discretion of the operators and optimal stent deployment was achieved in approximately 50% of patients. In a recent randomized IVUS study, the ULTIMATE (Intravascular Ultrasound Guided Drug Eluting Stents Implantation in “All-Comers” Coronary Lesions) trial,³ the authors used specific IVUS-guided strategies for stent sizing, as the ratio of 0.8 to media diameter or 1:1 to the distal reference lumen diameter. We have shown that stent sizing to the proximal or distal reference EEM diameter by optical coherence tomography (OCT) resulted in safe stent expansion.¹² However, in patients randomized to IVUS strategy, given there was no standard strategy for stent optimization by IVUS, stent expansion was left to the operator’s discretion. We have also shown that post-stent dilation in patients with positive lesion site remodeling significantly improved stent expansion with no complications.¹⁵ In the present study, we did not assess the lesion site remodeling because the assessment of remodeling at the lesion site after stent deployment is not accurate.

Predictors of adverse events after stenting. Optimal stent deployment is the most consistent predictor of both stent restenosis and thrombosis.^1,3,16,17 Even in the contemporary era of drug-eluting stents, underexpansion is a common finding. In the present study, stent underexpansion was detected by HD-IVUS in 44% of patients after successful AGS. Likewise, we have shown that with OCT- or IVUS-guided stenting, MSA was <5.0 mm² in about one-third of patients.¹² It has been demonstrated that MSA cut-off points of 5.4 mm² and 5.3 mm² predicted stent restenosis after everolimus- and zotarolimus-eluting stent deployment, respectively.¹ Fujii et al¹⁶ reported that stent underexpansion and a significant residual reference segment stenosis were associated with stent thrombosis. Kang et al⁶ reported that PB >50% at the stent edges was associated with increased stent restenosis. Strut malapposition, as a predictor of stent thrombosis, has been debated. A number of IVUS studies^4,5 showed that acute stent malapposition did not emerge as an independent predictor of stent thrombosis. In contrast, 2 recent OCT studies^18,19 showed that malapposition was frequently observed in patients with acute, subacute, and late stent thrombosis. In line with these observations, an in vitro study showed that malapposition increased stent thrombogenicity.²⁰ Given the above evidence, further stent postdilation with a balloon sized to distal reference EEM diameter carries minimal risk and might mitigate the future risk of stent thrombosis.

The use of HD-IVUS for stent optimization. The HD-IVUS imaging system used in the present study has several advantages as compared with 40 MHz IVUS, which are as follows: (1) high frame rates with HD-IVUS result in more precise and accurate quantitative measurements than with a 40 MHz IVUS system; and (2) higher-speed pullback diminishes image artifacts caused by cardiac motion. Notably, Garcia-Guimaraes et al showed that MLA and MSA measurements with HD-IVUS had excellent concordance with OCT.²¹ They also showed that EEM visualization was better with HD-IVUS than with OCT.

Clinical implications. The ability to optimize stent deployment may not only improve clinical outcomes, but may also translate into an overall cost-saving benefit at 1 year (even with the “up-front” cost of an IVUS catheter).²² In this respect, Ellis et alshowed that a reduction in an absolute 10% of restenosis rate led to an overall savings of nearly a billion dollars in the United States alone.²²  This degree of savings would justify an inexpensive “online” stent optimization strategy with IVUS. The underutilization of IVUS information for stent expansion could be due to the unwillingness of operators to use larger balloons or higher pressures to expand the stent.²³

Study limitations. There are a number of limitations to the present study. First, this was a hypothesis-generating study and the primary endpoint of this feasibility study was to determine the role of this strategy on the rate of optimal stent deployment, not clinical outcomes. In this context, it has been shown that optimal stent deployment is the most consistent predictor of both stent restenosis and thrombosis.^1,3,16,17 Second, the study is not randomized, but each patient served as its own control where IVUS data were compared post AGS vs post HD-IVUS optimization. Third, this is a small feasibility study and we excluded patients with bifurcation lesions requiring 2 stents, ostial lesions, or lesions with heavy calcifications. Randomized studies are warranted to investigate the rate of optimal stent deployment of the present strategy vs other techniques in all comers. Finally, although MLA and MSA measurements by HD-IVUS MLA had excellent concordance with OCT,²¹ future randomized trials are needed to demonstrate the impact of enhanced image resolution by HD-IVUS vs 40 MHz IVUS on optimal stent deployment.

Acknowledgments. The authors thank ACIST Medical Systems; and Becky Rorke, RN, MPH, Jessica Rivard, RCIS, and Janet Larson, RN for their contributions and study monitoring.

We have shown that after stent deployment and postdilation, stent results were suboptimal in two-thirds of patients. This simple online stent-optimization method with HD-IVUS was feasible and resulted in optimal stent deployment in the majority of patients. Randomized studies are warranted to compare the rate of optimal stent deployment and outcomes of this strategy vs other techniques.