Author + information
- Received May 5, 2017
- Revision received May 23, 2017
- Accepted May 27, 2017
- Published online August 7, 2017.
- Xiao Wang, MDa,b,c,
- Mitsuaki Matsumura, BSa,
- Gary S. Mintz, MDa,
- Tetsumin Lee, MDa,b,
- Wenbin Zhang, MDa,b,d,
- Yang Cao, MDa,b,e,
- Akiko Fujino, MDa,b,
- Yongqing Lin, MDa,b,
- Eisuke Usui, MDf,
- Yoshihisa Kanaji, MDf,
- Tadashi Murai, MDf,
- Taishi Yonetsu, MDf,
- Tsunekazu Kakuta, MD, PhDf and
- Akiko Maehara, MDa,b,∗ ()
- aClinical Trials Center, Cardiovascular Research Foundation, New York, New York
- bNewYork-Presbyterian Hospital/Columbia University Medical Center, New York, New York
- cBeijing Anzhen Hospital, Capital Medical University, Beijing, China
- dDepartment of Cardiology, Sir Run Run Shaw Hospital, College of Medicine, Zhejiang University, Hangzhou, China
- eThe First Affiliated Hospital of Harbin Medical University, Harbin, China
- fTsuchiura Kyodo General Hospital, Ibaraki, Japan
- ↵∗Address for correspondence:
Dr. Akiko Maehara, Cardiovascular Research Foundation, 1700 Broadway, 9th Floor, New York, New York 10019.
Objectives The aim of this study was to evaluate optical coherence tomography (OCT) and intravascular ultrasound (IVUS) versus coronary angiography in the assessment of target lesion calcification and its effect on stent expansion.
Background IVUS is more sensitive than angiography in the detection of coronary artery calcium, but the relationship among IVUS, OCT, and angiography has not been studied.
Methods Overall, 440 lesions (440 patients with stable angina) underwent OCT- and IVUS-guided stent implantation. Coronary calcification was evaluated using: 1) angiography; 2) IVUS (maximum calcium angle and the surface pattern); and 3) OCT (mean and maximum calcium angle, calcium length, and maximum calcium thickness).
Results Median patient age was 66 years, and 82.5% were men. Among 440 lesions, calcium was detected by angiography in 40.2%, IVUS in 82.7%, and OCT in 76.8%. The maximum calcium angle, maximum calcium thickness, and calcium length by OCT or IVUS increased in relation to the increasing severity of angiographically visible calcium. In 13.2% of lesions with IVUS-detected calcium, calcium was either not visible or was underestimated (>90° smaller maximum arc) by OCT mostly due to superficial OCT plaque attenuation. In 21.6% of lesions with IVUS calcium angle >180°, angiography did not detect any calcium; these lesions had thinner and shorter calcium deposits as assessed using OCT, and final minimum stent area was larger compared to those with angiographically visible calcium. In lesions with thinner calcium deposits by OCT, IVUS detected a smooth surface with reverberations whereas thick calcium deposits were associated with an irregular surface without reverberations.
Conclusions Angiographic detection of target lesion coronary calcium (compared to intravascular imaging) has not changed in the past 2 decades, and angiographically invisible calcium (only detectable by IVUS or OCT) did not appear to inhibit stent expansion.
Coronary calcification is a well-known marker of atherosclerotic plaque burden (1–3) and a determinant of stent underexpansion during percutaneous coronary intervention (PCI) (4–8). Thus, accurate evaluation of coronary calcification is critical to planning a PCI strategy. Two decades ago, Mintz et al. (9) showed a much higher sensitivity of calcium detection by intravascular ultrasound (IVUS) compared to coronary angiography and significant disagreement in the assessment of calcium between the 2 techniques. There was little new in this area until the development of optical coherence tomography (OCT), which, unlike IVUS, can penetrate calcium and evaluate its thickness (5–7,10). Thus, our goal in this study was to revisit intravascular imaging (both IVUS and OCT and their direct comparisons) versus angiography in the assessment of target lesion calcification in patients undergoing PCI.
This was a retrospective, single-center, observational study to evaluate the ability to detect coronary calcification and its unique imaging appearance by comparing OCT, IVUS, and coronary angiography as well as the relationship between calcification and acute stent expansion after PCI. From October 2008 to May 2015 (6.5 years), a total of 2,252 lesions in 1,711 patients were treated with PCI at Tsuchiura Kyodo General Hospital. Patients were excluded if they did not receive both pre-PCI OCT and IVUS imaging. Exclusions included only IVUS- or OCT-guided PCI, cardiogenic shock, congestive heart failure, renal insufficiency, a suboptimal result after thrombectomy with Thrombolysis In Myocardial Infarction flow grade 0 to 2, or the imaging device could not pass across the lesion. This left 1,272 lesions in 1,007 patients who underwent both OCT and IVUS imaging before PCI. Patients were then excluded from the analysis if the lesions required balloon angioplasty before imaging or if the OCT and IVUS images could not be coregistered. Next, lesions with stent thrombosis, in-stent restenosis, bypass graft failure, or insufficient image quality or lesions in patients with acute coronary syndromes were also excluded. Finally, in 51 patients with multiple lesions, we included only the lesion with the most severe stenosis. Finally, 440 lesions in 440 patients with stable angina pectoris were included in the analysis (Online Figure 1). Imaging and PCI procedures were performed per local standards and each operator’s discretion. The study was approved by the institutional review board, and written informed consent was obtained from all patients.
Coronary angiography analysis
Coronary angiograms were reviewed and analyzed at the Cardiovascular Research Foundation (New York, New York). Off-line analysis of the pre-PCI angiogram was performed using the QCA-CMS system (Medis medical imaging systems bv, Leiden, the Netherlands). Pre-PCI minimum lumen diameter, interpolated reference vessel diameter, and lesion length were analyzed using standard methods (11). Moderate calcification was defined as radio-opacities noted only during the cardiac cycle before contrast injection, whereas severe calcification was defined as radio-opacities observed without cardiac motion, usually affecting both sides of the arterial lumen (11–14).
OCT and IVUS image acquisition
Both OCT and IVUS were performed pre- and post-PCI after intracoronary nitroglycerin (100 to 200 μg).
For OCT, either frequency-domain OCT (ILUMIEN OPTIS or C7-XR [St. Jude Medical, St. Paul, Minnesota] or Lunawave [Terumo Corporation, Tokyo, Japan]) or time-domain OCT (M2/M3, LightLab Imaging, Inc., Westford, Massachusetts) was used (15). For frequency-domain OCT, a 2.7-F (Dragonfly Duo or Dragonfly OPTIS, St. Jude Medical) or 2.6-F (Fastview, Terumo Corporation) catheter was advanced across the lesion over an angioplasty guidewire, and automatic pullback was performed with a frame interval of 0.1 to 0.2 mm (Dragonfly) or 0.20 to 0.25 mm (Fastview) with continuous contrast injection (4 ml/s, 14 to 18 ml total). For time-domain OCT, the ImageWire (LightLab Imaging) was advanced across the lesion, an occlusion balloon (Helios, LightLab Imaging) was inflated proximal to the lesion, and automated pullback with a frame interval of 0.05 to 0.10 mm was performed during saline infusion through the tip of the occlusion balloon.
For IVUS, a commercially available IVUS system (iLAB, Boston Scientific Corporation, Marlborough, Massachusetts) was used. A 40-MHz, 2.6F imaging catheter (Atlantis SR Pro or Pro 2, Boston Scientific) was advanced distal to the lesion, and automated pullback was performed at a speed of 0.5 mm/s until the aorto-ostium.
OCT and IVUS image analysis
Off-line OCT images were analyzed using LightLab ORW software version E.0.2 (St. Jude Medical) or Lunawave version 1.1.0 (Terumo) at the Cardiovascular Research Foundation. Plaque morphology was assessed using previously established criteria (10). Calcified plaque was a signal-poor region with sharply delineated borders. The maximum thickness of calcified plaque was analyzed. The angles of calcified plaque were analyzed every 1 mm using the center of mass of the lumen; both maximum and mean (total calcium angle/lesion length) angles were calculated. Maximum calcium angle was further divided into the angle of calcium with <0.5 mm, 0.5 to 1.0 mm, or ≥1.0 mm in thickness. Lengths of calcified plaque were calculated by the total slice number multiplied by the frame interval.
Off-line IVUS images were analyzed using echoPlaque 3.0 (INDEC Medical Systems, Inc., Mountain View, California). Calcified plaque was defined as a region having a hyperechoic leading edge compared to the adventitia with acoustic shadowing; the location was categorized as superficial when the leading edge appeared within the most shallow 50% of plaque plus media thickness and was otherwise considered as deep (16). The angle of calcium was measured using the center of mass of the lumen. Reverberation was an artifact represented by secondary, false echoes of the same structure, and reverberation from the leading edge of calcium was evaluated. The surface shape of calcium by IVUS was categorized as: 1) smooth with reverberation; 2) irregular without reverberation; or 3) mixed.
Coregistration of OCT, IVUS, and angiography
OCT and IVUS images were coregistered using fiduciary side branches and known pullback speeds. The lesion segment was defined as the segment where the stent was implanted. After coregistration, the lesion segment in each OCT and IVUS image was analyzed independently by individuals who were blinded to the other analyses as well as to the clinical information. Finally, OCT and IVUS were compared side-by-side to confirm the relationship of each morphology to each other and to the angiograms.
Intraobserver and interobserver variability
Intraobserver and interobserver variability of the image analysis were assessed by evaluating 30 to 40 randomly selected images for angiographic calcification (moderate, severe), IVUS surface shape of calcium (smooth with reverberation, irregular without reverberation), IVUS and OCT maximum calcium angle, and OCT maximum calcium thickness by 2 independent readers and comparison by the same reader 2 weeks after the initial evaluation. Interobserver and intraobserver reproducibility of image analyses was assessed by Kappa statistics for categorical variable or intraclass correlation coefficients (ICC) for continuous variable.
Categorical variables are presented as frequency and were compared using chi square statistics or Fisher exact test, as appropriate. The normality of continuous variable was tested using the Kolmogorov-Smirnov test. Because most values were not normally distributed, continuous variables are shown as median and first and third quartiles and compared using the Mann-Whitney U or Kruskal-Wallis test with post hoc analysis by Dunn-Bonferroni test. Bland-Altman plot was used to show the difference between OCT and IVUS measurement for maximum calcium angle and the differences compared by paired t test. Passing-Bablok regression was used to show the relationship between OCT and IVUS and conducted using R version 3.2.1 (R Foundation for Statistical Computing, Vienna, Austria). Receiver operating characteristic analysis was used to determine the discriminatory capability as an area under the curve (AUC) with the optimal cutoff value using Youden’s index (the maximum value of [sensitivity + specificity − 1]). All tests were 2-tailed with a 0.05 significance level. All statistical analyses, except Passing-Bablok regression, were performed with SPSS 18.0 (IBM, Armonk, New York).
There was good interobserver and intraobserver agreement for the assessment of angiographic moderate calcium (Kappa: 0.81, 0.86) and severe calcium (Kappa: 0.87, 0.91), IVUS surface shape of calcium (Kappa: 0.87, 1.00) and maximum calcium angle (ICC: 0.99, 0.99), and OCT maximum calcium angle (ICC: 0.98, 0.98) and maximum calcium thickness (ICC: 0.99, 0.99).
Median patient age was 66 years (first quartile 60 years, third quartile 73 years), and 82.5% were men. The prevalence of chronic kidney disease (estimated glomerular filtration rate <60 ml/min/1.73 m2) was 32.5%, and 3.4% of patients were on hemodialysis (Table 1).
Calcium detection among angiography, OCT, and IVUS
Among 440 lesions, any amount of calcium was detected by coronary angiography in 40.2% (177 of 440) of lesions, by IVUS in 82.7% (364 of 440) of lesions, and by OCT in 76.8% (338 of 440) of lesions (Figure 1). The sensitivity and specificity of angiography to detect any OCT calcium were 50.9% and 95.1%, and the sensitivity and specificity of angiography to detect any IVUS calcium were 48.4% and 98.7%, respectively. By receiver operating characteristic analysis, the IVUS maximum calcium angle that predicted angiographically visible calcium was a cutoff value of 110° (AUC 0.80, 95% confidence interval [CI]: 0.76 to 0.83; p < 0.0001). Similarly, the OCT maximum calcium angle, length, and thickness that predicted angiographically visible calcium was a cutoff value of 101° (AUC 0.78, 95% CI: 0.73 to 0.81; p < 0.0001), 4.0 mm (AUC 0.81, 95% CI: 0.77 to 0.84; p < 0.0001), and 0.57 mm (AUC 0.80, 95% CI: 0.76 to 0.84; p < 0.0001), respectively.
In this cohort almost all calcium deposits were superficial. By OCT, 97.9% of calcium deposits (331 of 338) were located within 0.5 mm of the plaque surface, and by IVUS 91.2% of calcium deposits (332 of 364) were also located within 0.5 mm of the plaque surface.
The maximum calcium angle by IVUS was compared to the maximum calcium angle by OCT using Passing-Bablok regression (Figure 2A) along with the corresponding Bland-Altman plot (Figure 2B). The Passing-Bablok intercept estimate was 0.00 (95% CI: −1.83 to 0.00); and the slope estimate was 1.00 (95% CI: 0.97 to 1.05). IVUS measurements of maximum calcium angle were significantly larger than by OCT (mean difference 7.7°, 95% CI: 3.3 to 12.0°; p = 0.001). Twenty-six lesions with IVUS visible calcium had no calcium detected by OCT; these included 15 IVUS superficial calcifications in which OCT superficial plaque attenuation masked the calcium and 11 cases of deep IVUS calcium (Figure 3). Additionally, there were 22 cases in which the IVUS maximum calcium was >90° larger than that measured by OCT; all underestimations of the maximum calcium angle by OCT were due to superficial plaque attenuation. In 9 cases in which OCT maximum calcium was >90° larger than that measured by IVUS, 5 had an “overestimation” of the maximum OCT calcium angle because noncalcified, hypointensity plaque was thought to represent calcium; and 4 “underestimations” by IVUS were due to an imaging artifact (IVUS catheter close to the surface of the plaque resulted in a reduced echogenicity of the calcium leading edge or the guidewire masked part of the calcium).
Figure 4 shows angiographic calcium detection in relation to the maximum calcium angle either by OCT or IVUS, and Table 2 summarizes the OCT and IVUS findings stratified by angiographic calcium severity. OCT calcium angle (maximum, mean), IVUS maximum calcium angle, OCT maximum calcium thickness, and OCT calcium length increased with greater semiquantitative amounts of angiographic calcium resulting in smaller stent expansion in lesions with angiographically severe calcification.
Overall, 23.9% (17 of 71) of lesions with an OCT calcium angle >180° and 21.6% (16 of 74) of lesions with IVUS calcium angle >180° did not show angiographically visible calcium; we assessed OCT and IVUS findings in lesions with IVUS maximum calcium angle >180° stratified by angiographic visibility (Table 3). In this cohort, lesions without angiographically visible calcium had thinner calcium by OCT; the majority were <0.5-mm thick and had a shorter calcium length. In lesions with an IVUS maximum calcium angle >180°, the final minimum stent area was greater in the absence of angiographically visible calcium than in the presence of angiographically visible calcium.
Overall, 4.9% (5 of 102) of lesions without any OCT calcium had angiographic calcium; and 1.3% (1 of 76) of lesions without any IVUS calcium (also not visible by OCT) had angiographic calcium because IVUS superficial attenuated plaque and OCT superficial lipidic plaque masked the underlying calcium. The other 4 cases (OCT invisible, IVUS and angiographically visible calcium) include 2 cases with superficial OCT attenuation and 2 cases of deep calcium confirmed by IVUS.
After independent OCT and IVUS analysis, we reviewed the slices with the maximum calcium angle by OCT and IVUS side-by-side. In 84.3% (285 of 338) of the lesions, the slices with the maximum calcium angle by both OCT and IVUS could be matched; in the rest of the lesions (15.7%), the slices with the maximum calcium angle did not colocate, mostly due to the underestimation of the maximum arc of OCT calcium. Among these 285 lesions, there were 179 (62.8%) that showed a consistent IVUS superficial calcium pattern: a smooth surface with reverberations (97 lesions, Figure 5A) or an irregular surface without reverberations (82 lesions, Figure 5B). The other 37.2% of the lesions had a mixed and inconsistent appearance such that a similar analysis was not performed. As shown in Figure 6, IVUS calcium with a smooth surface and reverberations was thinner by OCT (OCT maximum calcium thickness <0.50 mm) compared to IVUS calcium with an irregular surface without reverberations (p = 0.001). On the other hand, IVUS calcium with an irregular surface without reverberations appeared thicker by OCT (calcium thickness >1 mm; p = 0.03).
Our main findings were as follows: 1) detection of calcium by coronary angiography compared to IVUS was consistent and similar to the previous report by Mintz et al. (9) in 1995; 2) any disagreement between coronary angiography and IVUS/OCT was due to a thin calcium deposit that did not appear to inhibit stent expansion; therefore, angiographically visible calcium (i.e., thick calcium) seemed to be a good marker to predict stent underexpansion; 3) in 13.2% of IVUS-detected calcium, calcium was either not visible by OCT (n = 26) or was underestimated (>90° smaller) by OCT (n = 22) mostly due to superficial OCT plaque attenuation; and 4) IVUS detected a smooth surface with a reverberation pattern in association with thinner calcium by OCT compared to IVUS-detected irregular surface without reverberation that was observed in the presence of thick calcifications.
Using IVUS, calcified plaque is defined as a region having a hyperechoic leading edge (compared to the adventitia) with acoustic shadowing and is typically quantified by measuring its angle because IVUS cannot penetrate calcium to measure its thickness, area, or volume (16). Using pathology as a gold standard, Kostamaa et al. (17) reported high sensitivity (89%) and specificity (97%) of detecting calcium deposits using 25-MHz IVUS. Friedrich et al. (18) reported that 30-MHz IVUS was able to detect dense, coherent calcified deposits with high sensitivity (90%) and specificity (100%), but it was limited in visualizing accumulation of microcalcification (≤0.05 mm, sensitivity 64%). In the 1995 study by Mintz et al. (9), angiography detected calcium in 38% (440 of 1,155) of lesions: 306 (26%) had moderate calcium and 134 (12%) had severe calcium. In the current study, angiography detected calcium in 40% (177 of 440) of lesions: 133 (30%) had moderate calcium and 10% had severe calcium. Similarly, IVUS detected calcium in 73% (841 of 1,155) of lesions with a mean angle of calcium measuring 115°, whereas in the current study IVUS detected lesion-associated calcium in 83%, with a mean angle of calcium measuring 127°. Although the prevalence of angiographic calcium between the previous and the current report is similar, the higher prevalence of IVUS calcium in the current study (83% vs. 73%) could be due to more superficial calcium location (less possibility to be masked by superficial attenuation) and higher transducer frequencies currently in use. This could also explain the higher false positive rate of angiographically visible calcium in the previous report compared to the current study.
Unlike IVUS, OCT is able to assess calcium thickness and therefore area and volume. Mehanna et al. (19) evaluated the accuracy of OCT quantification of calcium using cryo-imaging as a gold standard. Based on each cross-sectional calcium area measurement by OCT, the calcium volume was accurately measured in the absence of superficial plaque attenuation; however, calcium volume was underestimated in the setting of superficial plaque attenuation, consistent with our current findings.
Evaluation of calcium thickness as well as its angle is important to predict stent expansion (20). Hoffmann et al. (4) reported the impact of IVUS-detected calcium on stent expansion. In lesions with a maximum calcium >180°, a greater amount of calcium resulted in a smaller and more eccentric shaped stent area. Using IVUS, Vavarunakis et al. (20) studied lesions with moderately severe angiographic calcium and reported that stent expansion was inversely correlated to the arc of calcium, even after high-pressure balloon inflations. Using OCT, Kobayashi et al. (5) indicated that calcium area and angle were related to poor stent expansion; however, in other reports thinner calcium (<0.50 mm in thickness) was associated with calcium fracture irrespective of calcium angle, and calcium fracture was, in turn, associated with greater stent expansion versus the absence of calcium fracture (6,7). In our data, lesions with a large angle of calcium that were thin (<0.50 mm in thickness) did not appear to inhibit stent expansion. In the current study and in the previous study by Mintz et al. (3), angiography detected superficial calcium better angiographic visibility than deep calcium, presumably because it was thicker.
This was a retrospective observational study. The sample size was relatively small and may have reflected inclusion bias. Lesions with anticipated difficulty in advancing the OCT or IVUS catheter were excluded. Thus, the most severely calcified lesions were not included. However, the 40.2% prevalence of angiographic moderate/severe calcification was comparable to a previously published Resolute stent “all-comers” trial (21). The role of OCT or IVUS in detecting calcium and predicting stent underexpansion compared to angiography in patients with obesity needs further evaluation.
Angiographic detection of coronary calcium in the target lesion has remained similar in the past 2 decades and angiographic invisible calcium (only detectable by IVUS or OCT) did not appear to inhibit stent expansion.
COMPETENCY IN PATIENT CARE AND PROCEDURAL SKILLS: Any disagreement between coronary angiography and IVUS/OCT was due to a thin calcium deposit that did not appear to inhibit stent expansion; therefore, angiographically visible calcium (i.e., thick calcium) seemed to be a good marker to predict stent underexpansion. IVUS detected smooth surface with reverberation pattern was associated with the thinner calcium by OCT compared to IVUS-detected irregular surface without reverberation.
TRANSLATIONAL OUTLOOK: Further standardized, reproducible calcium measurements including detailed OCT or IVUS evaluation is warranted to evaluate the efficacy and safety for newer calcium debulking devices to include the appropriate homogeneous patient population.
The authors thank Dominic P. Francese, MPH, for assistance in preparing the manuscript.
For a supplemental figure, please see the online version of this paper.
Dr. Mintz has received grants from Volcano and St. Jude Medical; fellowships from Boston Scientific and St. Jude Medical; and honoraria from Volcano, Boston Scientific, and ACIST. Dr. Maehara has received grants from Boston Scientific and St. Jude Medical; and consulting fees from Boston Scientific and OCT Medical Imaging, Inc. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Dr. Wang and Mr. Matsumura are joint first authors. Deepak Bhatt, MD, MPH, served as the Guest Editor for this article.
- Abbreviations and Acronyms
- area under the curve
- intraclass correlation coefficient
- intravascular ultrasound
- optical coherence tomography
- percutaneous coronary intervention
- Received May 5, 2017.
- Revision received May 23, 2017.
- Accepted May 27, 2017.
- 2017 American College of Cardiology Foundation
- Otsuka F.,
- Sakakura K.,
- Yahagi K.,
- Joner M.,
- Virmani R.
- Sangiorgi G.,
- Rumberger J.A.,
- Severson A.,
- et al.
- Mintz G.S.,
- Pichard A.D.,
- Popma J.J.,
- et al.
- Maejima N.,
- Hibi K.,
- Saka K.,
- et al.
- Kubo T.,
- Shimamura K.,
- Ino Y.,
- et al.
- Madhavan M.V.,
- Tarigopula M.,
- Mintz G.S.,
- Maehara A.,
- Stone G.W.,
- Généreux P.
- Mintz G.S.,
- Popma J.J.,
- Pichard A.D.,
- et al.
- Tearney G.J.,
- Regar E.,
- Akasaka T.,
- et al.
- Pompa J.A.A.,
- Burke D.
- Abdel-Wahab M.,
- Richardt G.,
- Joachim Büttner H.,
- et al.
- Lee T.,
- Murai T.,
- Yonetsu T.,
- et al.
- Mintz G.S.,
- Nissen S.E.,
- Anderson W.D.,
- et al.
- Silber S.,
- Windecker S.,
- Vranckx P.,
- Serruys P.W.