Author + information
- Published online May 7, 2018.
- Joaquin E. Cigarroa, MD∗ ()
- ↵∗Address for correspondence:
Dr. Joaquin E. Cigarroa, Cardiovascular Division, Knight Cardiovascular Institute, UHN-62, Oregon Health & Sciences University, 3181 SouthWest Sam Jackson Park Road, Portland, Oregon 97239.
Coronary computed tomography angiography (CTA) has been used to assess patients with chest pain as well as patients with coronary artery disease who have been treated with percutaneous coronary intervention with coronary stents (1). In patients with native coronary artery disease, factors which adversely impact interpretation include increasing body mass index, relative tachycardia, arrhythmias, extensive coronary artery calcification, and small vessel diameter. The presence of coronary artery stents increases the technical challenges in visualization of the in-stent lumen due to blooming artifacts. The magnitude of blooming artifacts is impacted by the metal type, strut thickness, and stent design. As the metal content increases, so does the degree of difficulty in visualizing the in-stent lumen. Using 64-slice coronary CTA, prior investigators have demonstrated a sensitivity of 90.5% and a specificity of 93.0% with a positive predictive value of 63.3% and a negative predictive value of 99.3% (2). The development of bioresorbable scaffolds such as the one used in this trial are composed of an absorbable polymeric scaffold coupled with a scaffold coating containing everolimus. The scaffold consists of a series of circumferential sinusoidal rings connected by 3 linear links with a platinum marker at each end to facilitate visualization during stent implantation. The fully bioresorbable stent provides initial arterial mechanical support to prevent recoil coupled with an antiproliferative drug to prevent restenosis followed by bioresorption, which is often complete by 3 years. Theoretically, this technology promotes favorable arterial remodeling and normalization of vasoreactivity and may allow for better visualization of coronary arteries because the scaffold, other than the platinum markers, disappears over time (3).
In this issue of iJACC, Collet et al. (4) assess the diagnostic accuracy of coronary CTA of 258 lesions in 238 patients enrolled in the ABSORB II (A Bioresorbable Everolimus-Eluting Scaffold Versus a Metallic Everolimus-Eluting Stent II) study at the 3-year follow-up. The primary endpoint was coronary CTA diagnostic accuracy for detecting in-scaffold obstruction and luminal dimensions, as compared with invasive coronary angiography (ICA) and intravascular ultrasound (IVUS) examination. All patients underwent ICA and IVUS during the 3-year follow-up period. For data analyses, independent core labs were used for coronary CTA, quantitative coronary angiography, and IVUS. Restenosis was defined as a percentage area stenosis of >75%, which equals a 50% diameter stenosis for coronary CTA, 50% diameter stenosis by ICA, and a minimal lumen area <2.5 mm2 by IVUS which has been demonstrated to correlate with an abnormal fractional flow reserve in prior studies.
For comparison of diagnostic accuracy, receiver-operating characteristics curves were used for assessing the accuracy of the diagnostic test with test agreement between the 3 imaging tests assessed by the Bland-Altman and Passing Bablock methods.
Of the 238 patients enrolled who had 258 lesions, 189 lesions in 180 patients had all 3 imaging tests performed. In comparison with IVUS, both coronary CTA and ICA underestimated the in-scaffold minimal lumen diameter and minimal lumen area. For the overall group, coronary CTA criteria to predict an ICA stenosis of >50% had an area under the curve of 0.88 for a sensitivity of 75% and a specificity of 100%. Compared with IVUS, coronary CTA resulted in an area under the curve of 0.82, a sensitivity of 77%, and a specificity of 82%. The authors conclude that coronary CTA has a diagnostic accuracy similar to ICA for the detection of obstructive disease and that it might become the imaging modality of choice for assessment in patients treated with bioresorbable vascular scaffold (BVS). The comparison of coronary CTA with IVUS is interesting, given that the IVUS value used for significance is the value which correlates with an abnormal fractional flow reserve rather than an anatomic comparison of area or minimal luminal diameter obtained by coronary CTA and IVUS.
As with all clinical studies, one must question the generalizability of the results. To answer this question, one must delve deeper into the patients studied, lesions treated, and the imaging performed. The ABSORB II trial is a prospective, multicenter, single blind trial that randomized patients 2:1 with up to 2 de novo native coronary lesions with a minimal lumen diameter of between 2.25 and 3.80 mm with a lesion length not exceeding 48 mm to a BVS or everolimus-eluting metallic stent (EES) (4). The BVS contains zigzag hoops linked by 3 longitudinal bridges whereas the everolimus-eluting metallic stent is made of cobalt chromium alloy. At implantation, the BVS device volume is 3 times greater and the strut thickness is 150 μm versus 81 μm. The clinical trial included 501 patients of which 364 lesions involving 335 patients were treated with BVS. Patients were young (61.5 years of age), primarily male (76%), with a body mass index of 27.9 kg/m2, and with diabetes present in 24%. With regard to lesion type, only 2% of lesions were type C, only 13% had moderate or severe calcification, and the average lesion length was 13.8 mm. In addition, patients with ostial or bifurcation lesions were excluded. The majority of treated patients (81%) had only 1 study device implanted.
By 1 year, 72 BVS patients (22%) had new or worsening angina, and 3 had definite or probable scaffold thromboses. Of these patients, 7% (24) experienced death, myocardial infarction, or revascularization. The overall 3-year ABSORB II trial reported an in-device binary restenosis rate of 7.0% for the 298 lesions with 3-year invasive coronary angiographic analysis. As a result of recurrent clinical events and patients lost to follow-up, of the initial 335 patients with 364 lesions, only 238 patients (71% of enrolled patients) with 258 lesions (70% of lesions) were available for the analysis at 3 years and were included in this report. As a consequence, there is a bias of excluding the patients who are most apt to be of interest, those who develop recurrent angina and/or restenosis, which almost always occurs in the first 6 to 12 months. As a result of not assessing patients at the time they developed recurrent symptoms with probable loss of these key patients for the analysis in this report, the generalizability of their conclusion is limited, because patients with significant symptoms and/or symptomatic restenosis would have already been treated with recurrent revascularization and excluded from this report.
So why did the investigators use the 3-year time period for assessing the co-primary endpoints of vasomotion and angiographic late luminal loss? The premise of the study was to assess the potential benefit of a fully BVS to provide the mechanical support of a scaffold and the drug delivery during the first year with subsequent resorption during the following 2 years, with a goal of restoring vascular function and improving patient outcomes. By 3 to 5 years, the polymer should be replaced by a de novo connective tissue with associated vessel remodeling (5,6). As such, the ABSORB II design used a 3-year endpoint with ICA, IVUS, and coronary CTA designed to assess coronary vasoreactivity, resorption of the BVS, and coronary artery remodeling. Therefore, the use of coronary CTA at 3 years does not address the fundamental question of determining the presence or absence of binary restenosis during the first 12 months, when the majority of patients develop angiographic and clinical restenosis, a time when the backbone of the BVS with their associated thicker struts (relative to an everolimus-eluting metallic stent) and greater mass are present. The conclusion, therefore, that coronary CTA has good diagnostic accuracy to detect in-scaffold luminal obstruction is incorrect as most of the scaffold no longer exists at 3 years. However, coronary CTA at 3 years is an effective tool to assess the presence of a lesion and the remodeling associated with the prior BVS percutaneous coronary intervention.
To address the question of diagnostic accuracy of coronary CTA to detect in-scaffold luminal obstruction in patients treated with BVS, one would have to perform a comparison of coronary CTA and ICA at 6 to 12 months, a more clinically relevant time frame for our patients and our clinicians. Finally, whether one could use this technique reliably in patients who have comorbidities, including obesity and more complex lesions including extensive calcification, bifurcation lesions, and multiple overlapping BVS, also remains a question. Future studies performed within the first year after BVS implantation would be required to answer whether coronary CTA can be a useful tool in assessing in-scaffold lesions.
The author thanks Dr. Michael Shapiro for his review of this editorial.
↵∗ Editorials published in JACC: Cardiovascular Imaging reflect the views of the authors and do not necessarily represent the views of JACC: Cardiovascular Imaging or the American College of Cardiology.
Dr. Cigarroa has reported that he has no relationships relevant to the contents of this paper to disclose.
- 2018 American College of Cardiology Foundation
- Meijboom W.B.,
- van Mieghem C.A.,
- Mollet N.R.,
- et al.
- Onuma Y.,
- Dudek D.,
- Thuesen L.,
- et al.
- Collet C.,
- Chevalier B.,
- Cequier A.,
- et al.
- Serruys P.W.,
- Chevalier B.,
- Dudek D.,
- et al.
- Karanasos A.,
- Simsek C.,
- Gnanadesigan M.,
- et al.