Abstract: Objectives. The effect of intraprocedural coronary computed tomography angiography (coronary CTA) guidance on percutaneous coronary intervention (PCI) is unknown. We sought to determine the influence of CTA guidance on procedural strategies and immediate angiographic outcomes of PCI. Methods. Sixty patients were randomized to CTA-guided PCI (29 patients, 36 lesions) or angiography-guided PCI (31 patients, 39 lesions). To enable hands-free manipulation of CTA images by the interventional cardiologist during PCI, we developed an onsite augmented-reality (AR) system comprising a mobile application and AR glass. The primary endpoints were defined as: (1) stent length; and (2) largest stent diameter according to compliance chart. Procedural strategies, two-dimensional (2D) and three-dimensional (3D) quantitative coronary angiography (QCA), and safety outcomes were compared. Results. Whereas CTA guidance resulted in significantly higher frequency of stent postdilation using non-compliant (67% vs 31%; P<.01) and shorter balloons (16.6 ± 5.4 mm vs 20.5 ± 9.4 mm; P=.04) with numerically larger diameter (3.50 ± 0.63 mm vs 3.28 ± 0.45 mm; P=.10), it did not differ from angiography guidance with respect to lesion predilation, stent length, largest stent diameter according to compliance chart, and nominal stent diameter. The results of 2D- and 3D-QCA and safety outcomes were similar between groups. Neither death nor stroke occurred in either group. Conclusions. PCI under intraprocedural CTA guidance is associated with similar stent size selection and more frequent stent postdilation, resulting in comparable immediate angiographic and safety outcomes as compared with PCI under angiographic guidance alone.
J INVASIVE CARDIOL 2020;32(11):E268-E276. Epub 2020 September 10.
Key words: 2D and 3D quantitative coronary angiography, coronary computed tomography angiography, PCI
Planning and procedural guidance of percutaneous coronary intervention (PCI) is routinely based on visual estimation of invasive coronary angiography (ICA).1,2 This approach, however, has well-known limitations, including assessment of luminal dimensions only, and lack of information about plaque morphology and coronary vessel foreshortening, all of which may increase the risk of stent underestimation and greater residual reference segment disease.3 These limitations are partially overcome by intravascular imaging, which allows cross-sectional measurements of the vessel wall and potentially leads to larger postprocedural coronary lumen and lower rates of subsequent stent-edge restenosis and thrombosis as compared with angiographic guidance alone.4,5 The routine adoption of intravascular imaging is, however, hampered by longer procedural time, higher cost, and slightly increased risk of intraprocedural complications.6
Non-invasive coronary computed tomography angiography (coronary CTA) is currently gaining recognition as the first-line diagnostic test in patients with suspected coronary artery disease (CAD).7 Given its unique ability for preprocedural visualization of coronary plaque, it seems reasonable to incorporate CTA information into the process of PCI planning.2,8 To date, only two prospective studies assessed the impact of preprocedural CTA on changing PCI strategy.9,10 Nevertheless, the influence of intraprocedural CTA guidance on PCI outcomes is unknown. We therefore developed an augmented reality (AR) environment for display and review of CTA datasets in the catheterization laboratory, and performed a two-arm randomized trial to establish whether intraprocedural analysis of coronary CTA would affect the PCI treatment strategy and its immediate results as compared with standard PCI under angiographic guidance alone.
Study design and population. The AR-PCI trial was a single-center, investigator-initiated, unblinded, randomized, controlled, prospective trial on the effect of coronary CTA-guided PCI vs angiography-guided PCI (www.clinicaltrials.gov identifier, NCT03531424). Between April 2018 and May 2019, consecutive stable patients with documented obstructive CAD by either non-invasive coronary CTA or ICA (defined as the presence of at least 1 stenosis ≥70% in a native coronary artery) in whom PCI was considered by the local heart team were screened for inclusion. Subjects with obstructive CAD diagnosed by ICA underwent additional CTA. Eligible patients had ≥1 target lesion located in a native coronary artery with a visually estimated reference vessel diameter of at least 2.25 mm without any upper limits of the lesion length. We excluded patients with insufficient quality of coronary CTA, chronic renal failure (estimated glomerular filtration rate <30 mL/min), known allergy to contrast medium, PCI of the left main coronary artery, PCI of bifurcation lesions in which strategies other than a single crossover stent technique were anticipated, PCI of in-stent restenosis, PCI of bypass grafts, and PCI of chronically occluded vessels. The study was approved by the institutional ethics committee and complied with the declaration of Helsinki. Written informed consent was obtained from all participants.
Coronary CTA. Patients were scanned with a 256-slice CT scanner (Philips Brilliance iCT; Philips Healthcare) with a collimation of 128 x 0.625 mm, tube rotation time of 270 ms, and tube current between 200 and 360 mAs at 120 kV. Prior to the scan, patients with heart rate >65 beats/min without contraindications to beta-blockers received 50-150 mg of metoprolol orally, followed by intravenous boluses of 5 mg metoprolol if necessary. All patients received 0.8 mg of nitroglycerin sublingually immediately before CT acquisition. For each examination, a bolus of 100 mL iobitridol was injected intravenously followed by a 50 mL saline chaser. Prospective electrocardiography gating at 75% of the R-R interval was performed in order to minimize radiation exposure (Step & Shoot Cardiac; Philips Healthcare), and the mean dose-length product was 448 ± 196 mGy•cm. All CTA datasets were reconstructed and analyzed on a dedicated workstation (IntelliSpace Portal; Philips Healthcare) by an experienced reader (MO) blinded to all clinical information before patient randomization. Lesion characteristics were assessed using axial, longitudinal, and curved multiplanar reconstructions for the measurements of lesion length, minimal lumen diameter (MLD), and minimal lumen area (MLA), as well as maximal and minimal lumen diameters and lumen areas at the proximal and distal reference sites, as previously described.9
AR system for display of coronary CTA in the catheterization laboratory, We developed an onsite AR system consisting of a commercially available AR glass (Moverio BT-350; Epson) and an in-house built mobile application (cARdio; Amsterdam UMC). Selected CT curved multiplanar reconstructions of the target lesions were extracted and transmitted for display in a mobile application of AR glass. Specifically, at least three CT images depicting the morphology and anatomical measurements (lesion length and the luminal diameters of the proximal and distal reference sites) were used in each case (Figure 1). For sterile control of CTA datasets in the catheterization laboratory, the mobile application communicated with the gyroscopic sensors of the AR glass featuring a simple, head-controlled user interface.
Coronary angiography analysis. Pre- and postintervention angiograms were analyzed offline using a two-dimensional (2D) quantitative coronary angiography (QCA) software tool (CAAS II; Pie Medical) by an experienced reader (SPS) blinded to all other test results. In addition, postintervention angiograms underwent three-dimensional (3D)-QCA analysis with the QAngio XA 3D (Medis Medical Imaging Systems) by two experienced analysts (PAD, SPS). For the purpose of 3D-QCA analysis, an end-diastolic frame of two projections of at least 25° apart was used. Reference vessel diameter and area, as well as in-stent MLD, MLA, percentage diameter stenosis, and percentage area stenosis, were calculated using automated contour analysis with manual corrections if required. Reference segments were defined as the sites with the largest lumen proximal and distal to a stenosis but within the same segment (usually within 10 mm of the lesion with no intervening branches). The same prespecified projections were used in all cases, and only end-diastolic frames with minimal foreshortening and overlap were selected for analysis.
Randomization and study groups. Randomization was performed in the catheterization laboratory only in patients with a definitive decision on PCI and directly before coronary revascularization. Patients were randomized in a 1:1 fashion to CTA-guided PCI or angiographically guided PCI.
PCI procedure. All interventional procedures were performed on a monoplane cardiovascular x-ray system (Allura Xper FD 10/10; Philips Healthcare) by three highly experienced PCI operators (AN, NV, PK) trained on the use of AR technology and interpretation of CTA information. The training on the use of AR technology and CTA lasted approximately 5 minutes, and involved activation of the AR application as well as gesture-controlled review and reading of CTA images displayed in the AR glass (including identification of calcified plaque and interpretation of annotated CTA measurements). In the angiography-guided group, the operator was blinded to CTA information, and all PCI steps (including planned and performed strategies) were recorded and undertaken based on visual estimation under standard angiographic guidance. In the CTA-guided group, the operator browsed CTA images of all significant lesions using a wearable AR glass during PCI to develop the CTA-assisted treatment strategy (including predilation and postdilation, as well as stent length and diameter). Suggested stent length was selected as the distance from distal to proximal reference site with lumen area >4.5 mm2 and minimal amount of plaque as defined by coronary CTA, whereas the nominal stent diameter was sized according to the mean of the distal reference lumen diameters.9 In lesions with significant discrepancy between proximal and distal reference diameters and/or high calcium load on CTA, high-pressure balloon postdilation was advocated. To ensure sterile conditions in the catheterization laboratory, the review of CTA images was controlled by head movements of the PCI operator. Each operator was advised to abide by the preliminary PCI strategy, and any change between planned and performed PCI was noted in both groups. All patients received dual-antiplatelet therapy, and PCI was performed via radial or femoral access.
Study endpoints. The primary endpoints were defined as: (1) stent length; and (2) largest stent diameter according to compliance chart. The secondary endpoints included procedural characteristics (nominal stent diameter, number of stents, predilation and postdilation, maximal balloon pressure), results of postprocedural 2D and 3D QCA, and stent-edge dissection as assessed by visual angiography.
Statistical analysis. Data are presented as mean ± standard deviation or median (interquartile range [IQR]) for continuous variables and frequency (percentage) for categorical variables. The Shapiro-Wilk normality test was used to evaluate the distribution of the data. Comparison of continuous variables was performed using the Student’s t-test or non-parametric Mann-Whitney U-test, as appropriate. Categorical variables were compared using the Fisher exact test. Following the rules established for superiority trials, the sample size was calculated to be 60 patients based on the prior data,9 adopting a power of 80% and an alpha of 5% to detect at least 10% and 25% differences in stent diameter and stent length, respectively. Randomization was performed in a 1:1 ratio using a computerized program (Castor EDC) with block stratification according to age and gender. A P-value of <.05 was considered statistically significant. Analyses were performed using SPSS, version 20 software (SPSS).
Baseline characteristics. Out of 88 enrolled patients, we excluded patients with deferred PCI based on negative results of ICA or fractional flow reserve (n = 25), patients with insufficient CTA quality (n = 2), and a patient with extracardiac finding on CTA (n = 1). Thus, randomization was performed among 60 patients (48 men; mean age, 65.6 ± 9.6 years) who were allocated to the CTA-guided PCI (29 patients with 36 lesions) or ICA-guided PCI (31 patients with 39 lesions). The study flow chart is shown in Figure 2.
Median time from coronary CTA to PCI was 12.5 days (IQR, 6-25.7 days). The target lesions were most frequently located in the left anterior descending coronary artery (n = 37), followed by right coronary artery (n = 22) and left circumflex artery (n = 17). The mean lesion length was 20 ± 11.9 mm by ICA and 32.3 ± 15 mm by coronary CTA (P<.001), whereas MLD at the reference segment was 2.53 ± 0.49 mm by ICA and 3.11 ± 0.4 mm by coronary CTA (P<.001). Most lesions had mixed computed tomographic plaque morphology (68%), whereas non-calcified and calcified plaques were present in 21% and 11% of cases, respectively.
Baseline demographic and clinical characteristics of randomized patients were similar in both groups (Table 1). Similarly, no significant between-group differences were found with regard to preprocedural angiographic and CTA characteristics (Table 2).
Planned PCI strategy. In the all-lesion analysis comparing preprocedural planned PCI strategies, the use of CTA was associated with more frequent postdilation (89% vs 69%; P<.01) using larger balloons (3.51 ± 0.55 mm vs 3.30 ± 0.55 mm; P=.04) as compared with ICA (Table 3). Similarly, in the analysis restricted to lesions stratified according to the randomization arm, postdilation was more common in the CTA group than in the ICA group (92% vs 73%, respectively; P=.04). The two groups did not differ regarding planned balloon predilation, stent number, stent length, or nominal stent diameter (Table 4).
Performed PCI strategy. All lesions were treated with everolimus-eluting coronary stent systems using Xience stents (Abbott Vascular) or Promus stents (Boston Scientific). As presented in Table 4, change of planned PCI strategy was noted in numerically fewer lesions in the CTA group vs the ICA group (33% vs 51%, respectively; P=.16), with significantly less frequent change in stent length in the CTA group vs the ICA group (19% vs 41%, respectively; P=.049).
Whereas CTA guidance resulted in lower maximal pressure of stent implantation (12.3 ± 3.2 atm vs 13.6 ± 2.9 atm; P=.04), it was associated with significantly higher frequency of postdilation using non-compliant (67% vs 31%; P<.01) and shorter balloons (16.6 ± 5.4 mm vs 20.5 ± 9.4 mm; P=.04), with numerically larger diameters (3.50 ± 0.63 mm vs 3.28 ± 0.45; P=.10) than in the ICA group, respectively (Figure 3). The groups did not differ regarding other procedural characteristics, including the primary endpoints (stent length and largest stent diameter according to compliance chart), predilation, stent number, and nominal stent diameter (Table 4).
Angiographic and safety outcomes. Although CTA-guided PCI resulted in numerically larger 2D-QCA reference segment MLD and MLA (P=.10 and P=.09, respectively) compared with ICA-guided PCI, we did not find any significant between-group differences with regard to 2D-QCA and 3D-QCA results or angiographically defined stent edge dissections (Table 5).
Similarly, there were no significant differences in periprocedural myocardial infarction (7% vs 6%; P>.99), coronary perforation (0% vs 3%; P>.99), or contrast use during PCI (206 ± 65 mL vs 217 ± 83 mL; P=.54) between the CTA group and ICA group, respectively. Conversely, CTA guidance was associated with lower procedural radiation dose vs ICA alone (1.17 ± 0.60 Gy vs 1.48 ± 0.53 Gy, respectively; P=.02). No death or stroke occurred in either group.
The present study evaluated the impact of intraprocedural CTA guidance on changing PCI strategy and its immediate angiographic results. Notably, we show that intraprocedural guidance of PCI using CTA displayed in the catheterization laboratory is feasible and safe. In addition, whereas CTA-guided PCI resulted in more liberal use of stent postdilation using non-compliant balloons, we did not observe any significant differences with regard to other procedural strategies (including lesion predilation and stent size) or final immediate QCA results as compared with PCI under angiographic guidance alone.
While stand-alone angiography is an approved and primary imaging modality for guidance of PCI, it has some inherent limitations that might result in suboptimal postprocedural stent dimensions along with increased residual peristent disease, and thus higher rate of target-vessel revascularization.3 Although intravascular ultrasound achieves larger luminal dimensions and reduces major adverse cardiovascular events as compared with ICA,4,5 it is still infrequently used in most countries.1,6 Recently, coronary CTA has emerged as a valuable non-invasive tool for visualization and quantification of coronary plaque.2 Specifically, several studies have repeatedly indicated that lesion length and lumen diameter are underestimated on ICA compared with CTA,10,11 and CTA-derived lumen dimensions are well correlated with intravascular ultrasound.12 Furthermore, whereas CTA is gradually being adopted as the first-line diagnostic test in patients with CAD,7 it seems reasonable to incorporate the readily available CTA information into procedural planning for PCI. In this context, the current investigation underscores the feasibility and non-inferiority of CTA-guided versus ICA-guided PCI. Specifically, whereas PCI under CTA guidance did not differ from angiography-guided intervention regarding predilation and stent size parameters, it was associated with significantly higher frequency of postdilation using non-compliant and shorter balloons with numerically larger diameters. Of note, similar results were encountered in the ILLUMIEN III study, wherein both optical coherence tomography and intravascular ultrasound-guided PCI led to more postdilation balloon inflations, with greater maximal balloon diameters and pressure as compared with angiographic guidance alone.1 We assume that more profound stent postdilation in the CTA group might have translated into numerically larger maximal stent diameter according to the compliance chart, as well as luminal dimensions at the reference segments in both 2D- and 3D-QCA in our study. Nevertheless, none of the post-PCI QCA parameters reached statistical significance. Of note, the more aggressive postdilation in the CTA group must be balanced against the penalty of peristent dissections. However, in the present study, we noted numerically fewer stent edge dissections in the CTA group than in the ICA group (8% vs 20%, respectively; P=.20), potentially reflecting more stent-oriented postdilation and/or less peristent reference segment disease under CTA guidance.
To date, only two prospective randomized trials investigated the role of preprocedural CTA on changing PCI strategy (among which 1 study focused on coronary bifurcation lesions only).9,10 In a study by Pregowski et al,9 including 64 native coronary lesions, analysis of CTA before PCI resulted in the selection of significantly longer stents with a trend toward larger nominal stent diameter when compared with angiography-guided strategy alone. In contrast, the present study of 76 coronary lesions did not note significant differences in stent length and stent diameter between the CTA and ICA groups. The disparity between the current study and prior study may represent a function of differing clinical practices regarding selection of stent size under angiographic guidance, variability in operator experience, and different scenarios for planning PCI (intraprocedural vs preprocedural). Furthermore, although statistically non-significant we noted numerically longer lesion length at baseline in the ICA-guided group than in the CTA-guided group, which might have potentially counteracted the selection of longer stents under CTA guidance. Conversely, both our report and the study by Pregowski et al9 consistently suggested less residual disease with potentially more adequate lesion coverage (as confirmed by larger luminal areas in reference segments) after PCI under CTA guidance. These findings may offer clues as to the potential benefit of CTA-guided PCI, and large-scale clinical trials with intracoronary imaging and long-term clinical outcomes (preferably encompassing a large number of PCI operators) appear warranted.
The analysis of the relationship between planned and performed PCI revealed numerically fewer changes to the prespecified PCI protocol (with significantly less conversions from the planned stent length) under CTA guidance than angiographic guidance alone. These findings are in line with prior data suggesting rare modifications of the predefined PCI strategy under CTA guidance,10 and thus substantiate the reliability of CTA-guided approach among PCI operators.
Recently, the concept of AR, in which the operator is supplemented with additional visual information generated by a head-mounted computer, has been recognized as a safe and reliable carrier for coronary CTA to the catheterization laboratory.13,14 To the best of our knowledge, this is the first randomized study to employ AR technology in the catheterization laboratory for clinical decision-making during PCI. We substantiate the feasibility and safety features of AR for intraprocedural guidance of CTA-assisted PCI. Importantly, to ensure sterile control of CTA datasets by the operator while maintaining the ability to interact with the patient and support personnel during PCI, we developed the mobile application that communicated with the gyroscopic sensors of the AR glass featuring a simple, head-controlled user interface. We believe that AR technology offers an easily accessible, miniaturized, and cost-effective alternative to most advanced angiography systems with sizable monitor displays for projection of CTA to the catheterization laboratory.
Study limitations. This study had several limitations. First, it was a relatively small, single-center study reflecting PCI treatment strategies of only three operators. Second, we excluded left main coronary arteries, bifurcation lesions requiring a two-stent approach, in-stent restenosis, bypass grafts, and chronically occluded vessels. Nevertheless, with the median coronary artery calcium score of 513 (IQR, 191-1190) and high prevalence of calcified or mixed coronary plaques (79%), few of our patients had simple coronary lesions. Third, our study was not powered to assess clinical outcomes. Finally, the results for contrast use and radiation dose were limited to PCI (excluding preprocedural CTA) and should be interpreted with caution, provided that all of our subjects underwent additional angiographic interrogation for 2D- and 3D-QCA analyses (with at least two x-ray projections >25° apart).
PCI under intraprocedural CTA guidance is associated with selection of similar stent dimensions along with more frequent stent postdilation, resulting in comparable immediate angiographic and safety outcomes as compared with PCI under angiographic guidance alone.
*Joint first authors.
From the 1Department of Interventional Cardiology and Angiology, National Institute of Cardiology, Warsaw, Poland; 2Department of Cardiology, Amsterdam UMC, Vrije Universiteit Amsterdam, Amsterdam, the Netherlands; 3Interdisciplinary Centre for Mathematical and Computational Modelling, University of Warsaw, Poland; and the 4Department of Radiology, Amsterdam UMC, Vrije Universiteit Amsterdam, Amsterdam, the Netherlands.
Funding: This work was supported by Research Grant “Mobilność Plus” from the Polish Ministry of Science and Higher Education (1629/MOB/V/17/2018/0).
Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Opolski reports grant support from National Institute of Cardiology, Warsaw, Poland. Dr Knaapen reports grant support from HeartFlow. The remaining authors report no conflicts of interest regarding the content herein.
The authors report that patient consent was provided for publication of the images used herein.
Manuscript accepted April 9, 2020.
Address for correspondence: Dr Maksymilian P. Opolski, Department of Interventional Cardiology and Angiology, National Institute of Cardiology, Alpejska 42, 04-628 Warsaw, Poland. Email: email@example.com
1. Ali ZA, Maehara A, Généreux P, et al; ILUMIEN III: OPTIMIZE PCI Investigators. Optical coherence tomography compared with intravascular ultrasound and with angiography to guide coronary stent implantation (ILUMIEN III: OPTIMIZE PCI): a randomised controlled trial. Lancet. 2016;388:2618-2628.
2. Opolski MP. Cardiac computed tomography for planning revascularization procedures. J Thorac Imaging. 2018;33:35-54.
3. Fujii K, Carlier SG, Mintz GS, et al. Stent underexpansion and residual reference segment stenosis are related to stent thrombosis after sirolimus-eluting stent implantation: an intravascular ultrasound study. J Am Coll Cardiol. 2005;45:995-998.
4. Witzenbichler B, Maehara A, Weisz G, et al. Relationship between intravascular ultrasound guidance and clinical outcomes after drug-eluting stents: the assessment of dual antiplatelet therapy with drug-eluting stents (ADAPT-DES) study. Circulation. 2014;129:463-470.
5. Casella G, Klauss V, Ottani F, Siebert U, Sangiorgio P, Bracchetti D. Impact of intravascular ultrasound-guided stenting on long-term clinical outcome: a meta-analysis of available studies comparing intravascular ultrasound-guided and angiographically guided stenting. Catheter Cardiovasc Interv. 2003;59:314-321.
6. Lee CH. Intravascular ultrasound guided percutaneous coronary intervention: a practical approach. J Interv Cardiol. 2012;25:86-94.
7. Knuuti J, Wijns W, Saraste A, et al; ESC Scientific Document Group. 2019 ESC guidelines for the diagnosis and management of chronic coronary syndromes. Eur Heart J. 2020;41:407-477.
8. Achenbach S. Coronary CTA and percutaneous coronary intervention - a symbiosis waiting to happen. J Cardiovasc Comput Tomogr. 2016;10:384-385.
9. Pregowski J, Kepka C, Kruk M, et al. Comparison of usefulness of percutaneous coronary intervention guided by angiography plus computed tomography versus angiography alone using intravascular ultrasound end points. Am J Cardiol. 2011;108:1728-1734.
10. Wolny R, Pregowski J, Kruk M, et al. Computed tomography angiography versus angiography for guiding percutaneous coronary interventions in bifurcation lesions - a prospective randomized pilot study. J Cardiovasc Comput Tomogr. 2017;11:119-128.
11. van Velzen JE, de Graaf MA, Ciarka A, et al. Non-invasive assessment of atherosclerotic coronary lesion length using multidetector computed tomography angiography: comparison to quantitative coronary angiography. Int J Cardiovasc Imaging. 2012;28:2065-2071.
12. Leber AW, Knez A, von Ziegler F, et al. Quantification of obstructive and nonobstructive coronary lesions by 64-slice computed tomography: a comparative study with quantitative coronary angiography and intravascular ultrasound. J Am Coll Cardiol. 2005;46:147-154.
13. Opolski MP, Debski A, Borucki BA, et al. First-in-man computed tomography-guided percutaneous revascularization of coronary chronic total occlusion using a wearable computer: proof of concept. Can J Cardiol. 2016;32:829.e11-829.e13.
14. Opolski MP, Debski A, Borucki BA, et al. Feasibility and safety of augmented-reality glass for computed tomography-assisted percutaneous revascularization of coronary chronic total occlusion: a single-center prospective pilot study. J Cardiovasc Comput Tomogr. 2017;11:489-496.