Multislice Spiral Computed Tomography for the Evaluation of Stent Patency After Left Main Coronary Artery Stenting
A Comparison With Conventional Coronary Angiography and Intravascular Ultrasound
Background— Surveillance conventional coronary angiography (CCA) is recommended 2 to 6 months after stent-supported left main coronary artery (LMCA) percutaneous coronary intervention due to the unpredictable occurrence of in-stent restenosis (ISR), with its attendant risks. Multislice computed tomography (MSCT) is a promising technique for noninvasive coronary evaluation. We evaluated the diagnostic performance of high-resolution MSCT to detect ISR after stenting of the LMCA.
Methods and Results— Seventy-four patients were prospectively identified from a consecutive patient population scheduled for follow-up CCA after LMCA stenting and underwent MSCT before CCA. Until August 2004, a 16-slice scanner was used (n=27), but we switched to the 64-slice scanner after that period (n=43). Patients with initial heart rates >65 bpm received β-blockers, which resulted in a mean periscan heart rate of 57±7 bpm. Among patients with technically adequate scans (n=70), MSCT correctly identified all patients with ISR (10 of 70) but misclassified 5 patients without ISR (false-positives). Overall, the accuracy of MSCT for detection of angiographic ISR was 93%. The sensitivity, specificity, and positive and negative predictive values were 100%, 91%, 67%, and 100%, respectively. When analysis was restricted to patients with stenting of the LMCA with or without extension into a single major side branch, accuracy was 98%. When both branches of the LMCA bifurcation were stented, accuracy was 83%. For the assessment of stent diameter and area, MSCT showed good correlation with intravascular ultrasound (r=0.78 and 0.73, respectively). An intravascular ultrasound threshold value ≥1 mm was identified to reliably detect in-stent neointima hyperplasia with MSCT.
Conclusions— Current MSCT technology, in combination with optimal heart rate control, allows reliable noninvasive evaluation of selected patients after LMCA stenting. MSCT is safe to exclude left main ISR and may therefore be an acceptable first-line alternative to CCA.
Received December 17, 2005; revision received June 7, 2006; accepted June 8, 2006.
Multislice computed tomography coronary angiography (MSCTA) is a promising noninvasive coronary imaging modality.1,2 Although several reports have shown that it may be used to evaluate stent patency, more precise evaluation of the lumen within the stent is markedly hindered by artificial enlargement of the metallic stent struts caused by blooming artifact.3 The impact of blooming artifact on the evaluation of structures inside stents is inversely related to stent diameter.4 In large-diameter coronary stents, such as those implanted in the left main coronary artery (LMCA) or proximal left anterior descending artery (LAD), neointimal hyperplasia (NIH) within the stent can be visualized on multislice computed tomography (MSCT), demonstrating its potential for the detection of in-stent restenosis (ISR), in addition to stent patency, in specific lesion subsets.5
Editorial p 616
Clinical Perspective p 653
Although CABG surgery is still the recommended treatment for significant left main (LM) disease, the introduction of drug-eluting stents (DES) with much lower restenosis rates has increasingly resulted in the alternative use of percutaneous coronary intervention (PCI).6,7 However, LM ISR still occurs with DES and may result in fatal myocardial infarction or sudden death.8–10 Careful surveillance, including routine coronary angiography at 2 to 6 months after PCI, is therefore strongly recommended.11 A noninvasive method to detect ISR and to potentially preempt its clinical consequences would be of evident clinical value. In the present study, we evaluated the potential of MSCT for detecting LM ISR.
Between March 2004 and November 2005, we screened 91 consecutive patients scheduled for follow-up coronary angiography after LMCA stenting for inclusion in a protocol to compare MSCT with conventional angiography. In addition, most patients also underwent intravascular ultrasound (IVUS) evaluation of the LMCA. All patients in sinus rhythm who were able to hold their breath for 20 seconds were eligible for inclusion. We excluded patients with the following: previous allergic reaction to contrast, impaired renal function (serum creatinine >1.6 mg/dL), contraindication to β-blockers (high-degree heart block, poor left ventricular function, asthma, or severe chronic obstructive pulmonary disease), obesity (body mass index >30 kg/m2 for patients scanned with the 16-row MSCT scanner), and those who had an acute coronary syndrome at the time of scheduled angiography. The institutional review board of Erasmus MC Rotterdam approved the study, and all subjects gave informed consent.
MSCT Protocol and Image Acquisition
MSCT was performed in the fortnight before conventional angiography. Patients with a heart rate >65 bpm received 100 mg of metoprolol orally 1 hour before the scan. Up to July 2004, MSCT data were acquired with a 16-slice MSCT scanner (Sensation 16, Siemens, Germany). From August 2004 on, all scans were performed with a 64-slice MSCT scanner (Sensation 64, Siemens). A bolus of 100 mL of contrast (iodixanol, 320 mg of iodine per 1 mL [mgI/mL]; Visipaque, Nycomed, Inc, Princeton, NJ) was injected intravenously at 4 to 5 mL/s. MSCT data were acquired during a single breath hold once the contrast material in the ascending aorta reached a predefined threshold of 100 Hounsfield units. The 16-slice MSCT scanner had the following scan parameters: detector collimation 16×0.75 mm, table feed 3.0 mm per rotation, gantry rotation time 420 ms, tube voltage 120 kV, and tube current 400 to 450 mA. The parameters for the 64-slice MSCT scanner were: detector collimation 64×0.6 mm, table feed 3.8 mm per rotation, gantry rotation time 330 ms, tube voltage 120 kV, and tube current of 900 mA. With dedicated software (WinDose, Institute of Medical Physics, Erlangen, Germany), the average radiation exposure was calculated as 11.8 to 16.3 mSv for the 16-slice MSCT scanner and 15.2 to 21.4 mSv for the 64-slice MSCT scanner. All data were reconstructed with a field of view of 50×50 mm, image matrix of 512×512 pixels, and a sharp heart view (B46f) convolution kernel. Image reconstruction was retrospectively gated to the ECG. The position of the reconstructed window within the cardiac cycle was individually optimized to minimize motion artifacts. Data sets containing no or minimal motion artifacts were transferred to a remote workstation for further evaluation.
MSCT Data Analysis
Two experienced observers, unaware of the results of conventional angiography, evaluated the MSCT data sets on both the original axial images and on multiplanar reformatted reconstructions rendered orthogonal and perpendicular to the vessel course. In the case of LM bifurcation stenting, each of the 3 segments (LM, LAD, and circumflex artery [CX]) was evaluated individually. To assess significant restenosis (≥50% decrease of lumen diameter as defined by angiography), we evaluated the stent in the LMCA and, where applicable, in related branches, including the 5-mm borders proximal and distal to the stent. The stent lumen was visually evaluated as (1) patent with no visible NIH, (2) patent with nonobstructive (<50% lumen diameter) NIH, (3) patent with obstructive (>50% lumen diameter) NIH, ie, restenosis, and (4) occluded. Disagreements were resolved by consensus. Where a single stent was placed across the LM bifurcation into the LAD or CX, with no intervention, or solely with balloon dilatation in the other branch, we also evaluated the proximal 5 mm of the untreated side branch unless a functional bypass graft was attached to the nonstented branch.
For comparison with IVUS, stent area and diameter at the proximal and distal stent edge and at the ostium of LAD and CX were measured in cross-sectional orthogonal image planes. The degree of luminal narrowing was quantified in the cross section with maximum stenosis as percent diameter and area stenosis by calculating the ratio of the minimal lumen diameter or area by the stent diameter or area.
Conventional x-ray coronary angiography was performed with standard techniques after intracoronary injection of 2 mg of isosorbide dinitrate. Care was taken to ensure that the same views of the LM were obtained as at the time of intervention. An experienced cardiologist, unaware of the results of MSCT, analyzed all angiographic data quantitatively, using validated, automated, edge-detection software (CAAS II, Pie Medical, Maastricht, Netherlands). With the outer diameter of the catheter tip, not filled with contrast, as the calibration standard, we measured from multiple projections in diastole the minimal lumen diameter, the reference diameter, and the percentage diameter stenosis of the LMCA and, where indicated, the LAD and CX; the averaged results were recorded. Binary angiographic restenosis was defined as percentage diameter stenosis ≥50% anywhere within the stent or within the 5-mm segments proximal or distal to the stent margins. This definition also applied to the proximal 5-mm vessel segment of the untreated LAD or CX, where a stent was placed over the distal LM bifurcation.
IVUS was performed with a 2.9F, 40-MHz or a 3.2F, 30-MHz single-element mechanical transducer (Boston Scientific, Natick, Mass). After intracoronary injection of 2 mg of isosorbide dinitrate, the IVUS catheter was positioned at least 1 cm distal to the stent in the LMCA; for patients with bifurcation stenting, a sufficiently distal starting point in LAD or CX was chosen. IVUS images were recorded after initiation of automated pullback at 0.5 mm/s. In case of bifurcation stenting, only 1 of the 2 major branches, preferentially the LAD, was investigated. IVUS images of the entire pullback were recorded on both S-VHS videotape and CD-ROM for offline analysis.
In our institution, all IVUS pullbacks are retrospectively gated with a program that selects images recorded in the end-diastolic phase.12 Two experienced observers, blinded to the angiography and MSCT results, analyzed and reviewed the IVUS image data sets using an offline semiautomated software package (Curad, version 3.1, Wijk bij Duurstede, the Netherlands).13 The stented portion of the LMCA, including the stented parts of the side branches, was evaluated for the presence of NIH. For each of the 3 coronary segments (ie, stented part of LM, LAD, and CX), the following IVUS parameters were measured: (1) minimal lumen cross-sectional area (CSA, mm2), (2) mean stent CSA (mm2), (3) minimal lumen diameter (mm), (4) mean stent diameter (mm), and (5) maximal NIH thickness (mm). Similar to the MSCT measurements, stent area and diameter at the proximal and distal stent edges and at the ostium of LAD or CX were determined. The percent diameter and area stenosis were comparable to MSCT determined in cross sections with maximal in-stent lumen obstruction.
Continuous variables are presented as mean±SD or median with 25% and 75% interquartile ranges. Categorical variables are presented as counts and percentages. Continuous variables were compared by Student t test for normally distributed values; otherwise the Mann-Whitney U test was used. All tests were 2-tailed, and a probability value of <0.05 was considered significant. Accuracy (percentage of patients correctly classified), sensitivity, specificity, and positive and negative predictive values of MSCT for the detection of ≥50% ISR, as determined by quantitative coronary angiography, was calculated on a per-patient basis. Diagnostic test results are reported with associated 95% CIs based on binomial probabilities. Quantitative IVUS and MSCT data were correlated by means of Bland-Altman and linear regression analysis and by calculating the Pearson correlation coefficient.14 Mean values were compared with the 2-tailed t test. The degree of agreement between IVUS and MSCT was also measured by the κ-statistic and expressed as a function of NIH thickness, as determined by IVUS. The intraobserver and interobserver variability for the detection of ISR and NIH was determined with the κ-statistic. Statistical analysis was performed with SPSS, version 12.1 (SPSS Inc, Chicago, Ill).
The authors had full access to the data and take full responsibility for its integrity. All authors have read and agree to the manuscript as written.
Seventeen of the 91 potentially eligible patients (ie, those scheduled for conventional angiography during the inclusion period) had exclusion criteria as outlined in Table 1; 4 of the 74 patients included had a technically inadequate scan. One of those 4 patients had ISR on conventional angiography. The remaining 70 patients constitute the study population. The baseline clinical characteristics of these patients are summarized in Table 2. The median interval between stent placement and computed tomography (CT) imaging was 259 days (range 89 to 758 days), and the average interval between MSCT and conventional coronary angiography (CCA) was 14±16 days. Mean basal heart rate was 68±10 bpm; 49 patients (70%) received additional β-blockers, which resulted in a mean periscan heart rate of 57±7 bpm. Mean scan time with the 64-row detector MSCT (n=43, 11.1±1.2 seconds) was significantly (P<0.001) shorter than with the 16-row detector MSCT (n=27, 15.9±1.5 seconds).
Anatomic Characteristics and Procedural Details of Index Interventions
Table 3 presents the anatomic characteristics of the lesions treated and the procedural details of the index interventions. The vast majority of patients (83%, 58/70) underwent PCI of unprotected LM stem lesions, and DES were used in all but 3 patients (96%). Two groups of patients were defined, according to the characteristics of the PCI procedure. The first group (n=46) was defined as a “simple stenting group” and comprised patients in whom single or overlapping stents were implanted in the LM stem alone (n=14) or in whom they extended from the LM stem into only the LAD or the CX (n=32). The second group was defined as a “complex bifurcation stenting group” and comprised patients in whom the LM stem and both the LAD and CX were stented (n=24). Further details are presented in Table 3.
Overall MSCT Results
The diagnostic accuracy of MSCT is summarized in Table 4. On the basis of conventional angiography, ISR occurred in 14% of the patients (10/70). In the overall population, MSCT identified all patients with restenosis and correctly classified 55 patients as having no ISR (Figures 1 and 2⇓). In the group of patients (n=46) in whom simple stenting was performed, only 1 patient was incorrectly scored (false-positive) on MSCT. By contrast, in the complex bifurcation stenting group (n=24), 4 patients who did not have restenosis were scored as having ISR on MSCT. In 3 of the 5 false-positive scans, stent-related high-density artifacts at the level of the CX ostium, due to the complex stenting strategy used, precluded correct assessment of the lumen within the stent. Three false-positives occurred among those evaluated with the 16-slice scanner; the other 2 occurred with the 64-slice scanner. Numbers were too small to support a statistical comparison among the 2 scanner types. The characteristics of the patients with restenosis and the 5 patients with a false-positive scan are reported in Table 5. Seven of the 10 patients presented with restenosis in the ostium of the CX or intermediate branch. Three patients were revascularized; 2 underwent re-PCI, and the third patient underwent CABG.
Ten patients presented with 11 de novo lesions in a previously untreated vessel segment, which required revascularization by PCI in 2 patients. All 11 lesions were correctly identified on MSCT; overestimation of lesion severity by MSCT occurred in 2 additional calcified lesions. In 12 patients, the presence of nonsignificant (<50% stenosis) NIH in the LM (n=5), LAD (n=5), and CX (n=2) on diagnostic coronary angiography was correctly classified as nonobstructive on MSCT (Figure 3).
Quantitative Coronary Angiography, IVUS, and MSCT
Quantitative coronary angiography (QCA) and IVUS data are shown in Table 6. The reference diameter and minimal lumen diameter of the patients who were wrongly classified did not differ from those who were correctly analyzed by MSCT (data not shown). Fifty-six nonconsecutive patients (90 segments) had technically adequate IVUS examinations. NIH was present in 31 patients (38 segments). On the basis of a κ-value >0.8, which indicates very good agreement between the 2 imaging techniques, ie, MSCT and IVUS, a cutoff value of 1 mm was identified in which NIH could be identified with good confidence by MSCT (Figure 4). This corresponded to a 30% diameter obstruction as determined by QCA.
Quantitative IVUS and MSCT stent data were available for comparison in 50 patients (Figures 5 and 6⇓). The mean stent diameter was 3.4±0.70 mm with MSCT compared with 3.6±0.6 mm with IVUS. The correlation coefficient was significant (r=0.78, P<0.001), but the Bland-Altman analysis showed a systematic overestimation of the stent diameter by MSCT (mean difference 0.22±0.44 mm, P<0.0001). Mean stent area of MSCT compared with IVUS was 10±3.3 versus 10.3±3.5 mm2. Bland-Altman analysis for stent area demonstrated good correlation between the 2 imaging techniques (mean difference 0.35±2.53 mm2, P=0.13; r=0.73, P<0.001). Only 11 patients presented sufficient lumen narrowing to compare IVUS with MSCT. As shown in Figure 7, correlations were moderate for quantifying the degree of diameter and area stenosis (r=0.65 and 0.55, respectively).
To assess internal validity, all MSCT data sets were analyzed twice by 2 observers. Interobserver agreement for the detection of ISR or NIH was good to moderate (κ-value of 0.74 and 0.6, respectively), whereas intraobserver agreement was very good in both situations (κ-value 0.82).
Stent implantation in the LM and proximal LAD/CX provides the “best case scenario” for the use of MSCT in the detection of ISR for several reasons. First, stents implanted in the LM and proximal LAD/CX are relatively large; second, the LM and proximal LAD are usually running in an axial plane that corresponds to the scan direction, whereas the proximal CX generally runs in a slightly less favorable craniocaudal longitudinal plane; finally, this part of the coronary tree is relatively protected from motion artifact.
Recent technical improvements, including improved z-resolution, faster tube rotation, and the development of dedicated filters for image reconstruction in the presence of stents, have also significantly improved the potential of MSCT to assess coronary stent patency. Furthermore, preliminary experience with 16-detector-row CT scanners showed promise for the evaluation of stents in the LMCA.4,15
The present study demonstrates that evaluation of stent patency in the LMCA is feasible with 16-slice and the latest 64-slice MSCT scanners. The technique is reliable (accuracy of 98%) for the evaluation of patients with stenting of the LM and for those patients with distal LM bifurcation lesions, in whom only 1 of the major branch vessels (ie, the LAD or the CX) is stented. However, in patients who required complex bifurcation stenting, ie, treatment of the LM and both major side branches, reliability was significantly less (accuracy of 83%). The most obvious explanation is the large quantity of metal, involving up to 3 layers of struts for crush bifurcation stenting, at and around the ostium of the branch vessels, which is a major source of artifact on MCST.
A few studies reported reasonable accuracy of MSCT to quantify the degree of coronary artery stenosis in untreated coronary arteries.16,17 The present study reports on the quantification of stent dimensions and the amount of lumen narrowing within LMCA stents. Quantitative assessment by MSCT of stent diameter and area showed good correlation with IVUS. Not unexpectedly, metal-related artifacts, referred to as “blooming” effect, impair optimal visualization of the stent lumen and thus accurate quantification of lumen narrowing within the stent, especially for minor to moderate degrees of NIH. Nevertheless, NIH of >1 mm thickness is recognized with sufficient sensitivity to allow the detection of clinically relevant restenosis in patients after LMCA stenting.
LMCA stent implantation traditionally represented a relatively small proportion of PCIs. Until recently, more widespread applicability of this revascularization modality had been hampered by the high occurrence of ISR.9,18,19 The advent of the DES has both significantly reduced restenosis rates and improved long-term clinical outcome; this has already resulted in an increasing proportion of patients with LM disease being treated percutaneously.6,7 However, restenosis still occurs and when located in the LMCA may cause severe myocardial ischemia with potentially fatal consequences.19,20 A noninvasive method for its detection, before clinical events occur, would be of major clinical value. The results of the present study indicate that current MSCT scanners may provide a reliable alternative to CCA. The low number of false-positive scans that lead to “unnecessary” diagnostic coronary angiograms should be acceptable in the light of the potential serious consequences of LM ISR.
Limitations of the Study
The use of 2 different MSCT scanners in the present study might be criticized; however, the spatial resolution of the 16- and 64-slice scanners in the x- and y-axes has remained identical (0.4 mm), and only the resolution in the z-axis has improved with the 64-slice scanner. Because the LM and proximal LAD generally course in an axial plane, only the resolution in the x- and y-axes plays a major role in the high-resolution visualization of these segments.
The rather high radiation dose of MSCT is a general limitation of the technique; however, new developments, such as the recent introduction of dual-source CT scanners, systematically allow the use of ECG pulsing, thereby significantly reducing patient dose. On the other hand, many physicians underestimate the radiation exposure of widely accepted imaging techniques such as CCA or technetium sestamibi scans that produce a dose as high as 5 or 20 mSv, respectively.21,22
The rather low positive predictive value (67%) is related to the low prevalence of ISR in patients treated with DES and reflects the limitations of current CT technology. New technical advances, such as the development of flat panel detectors and the general trend in interventional cardiology to use stents with thinner struts, will be beneficial for CT imaging of stents and reduce the number of false-positive scans.
The present study was limited by the relatively low number of patients with complex bifurcation stenting, because this group importantly affects the reported diagnostic accuracy data. Nonetheless, it is unlikely that this proportion will increase in the near future given the tendency to treat bifurcation lesions through a simpler approach and to use a second stent toward the side branch only when strictly necessary.23
Finally, only 2 stent types with similar strut dimensions were used in the present study. These results do not necessarily apply to different stent types, because parameters such as strut thickness and stent design importantly influence the amount of blooming effect and thus assessability of the lumen inside.
In combination with optimal heart rate control, current MSCT technology allows reliable noninvasive evaluation of selected patients after LMCA stenting. A negative MSCT scan virtually rules out the presence of LM ISR and may therefore be an acceptable first-line alternative to CCA.
Nieman K, Cademartiri F, Lemos PA, Raaijmakers R, Pattynama PM, de Feyter PJ. Reliable noninvasive coronary angiography with fast submillimeter multislice spiral computed tomography. Circulation. 2002; 106: 2051–2054.
Ropers D, Baum U, Pohle K, Anders K, Ulzheimer S, Ohnesorge B, Schlundt C, Bautz W, Daniel WG, Achenbach S. Detection of coronary artery stenoses with thin-slice multi-detector row spiral computed tomography and multiplanar reconstruction. Circulation. 2003; 107: 664–666.
Mollet NR, Cademartiri F. Images in cardiovascular medicine: in-stent neointimal hyperplasia with 16-row multislice computed tomography coronary angiography. Circulation. 2004; 110: e514.
Chieffo A, Stankovic G, Bonizzoni E, Tsagalou E, Iakovou I, Montorfano M, Airoldi F, Michev I, Sangiorgi MG, Carlino M, Vitrella G, Colombo A. Early and mid-term results of drug-eluting stent implantation in unprotected left main. Circulation. 2005; 111: 791–795.
Valgimigli M, van Mieghem CA, Ong AT, Aoki J, Granillo GA, McFadden EP, Kappetein AP, de Feyter PJ, Smits PC, Regar E, Van der Giessen WJ, Sianos G, de Jaegere P, Van Domburg RT, Serruys PW. Short- and long-term clinical outcome after drug-eluting stent implantation for the percutaneous treatment of left main coronary artery disease: insights from the Rapamycin-Eluting and Taxus Stent Evaluated At Rotterdam Cardiology Hospital registries (RESEARCH and T-SEARCH). Circulation. 2005; 111: 1383–1389.
Takagi T, Stankovic G, Finci L, Toutouzas K, Chieffo A, Spanos V, Liistro F, Briguori C, Corvaja N, Albero R, Sivieri G, Paloschi R, Di Mario C, Colombo A. Results and long-term predictors of adverse clinical events after elective percutaneous interventions on unprotected left main coronary artery. Circulation. 2002; 106: 698–702.
Smith SC Jr, Feldman TE, Hirshfeld JW Jr, Jacobs AK, Kern MJ, King SB III, Morrison DA, O’Neil WW, Schaff HV, Whitlow PL, Williams DO, Antman EM, Adams CD, Anderson JL, Faxon DP, Fuster V, Halperin JL, Hiratzka LF, Hunt SA, Nishimura R, Ornato JP, Page RL, Riegel B. ACC/AHA/SCAI 2005 guideline update for percutaneous coronary intervention: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (ACC/AHA/SCAI Writing Committee to Update 2001 Guidelines for Percutaneous Coronary Intervention). Circulation. 2006; 113: e166–e286.
Hamers RBN, Knook M, Sabate M, Roelandt JRTC. A novel approach to quantitative analysis of intravascular ultrasound images. Comput Cardiol. 2001; 28: 589–592.
Leber AW, Knez A, von Ziegler F, Becker A, Nikolaou K, Paul S, Wintersperger B, Reiser M, Becker CR, Steinbeck G, Boekstegers P. 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.
Kelley MP, Klugherz BD, Hashemi SM, Meneveau NF, Johnston JM, Matthai WH Jr, Banka VS, Herrmann HC, Hirshfeld JW Jr, Kimmel SE, Kolansky DM, Horwitz PA, Schiele F, Bassand JP, Wilensky RL. One-year clinical outcomes of protected and unprotected left main coronary artery stenting. Eur Heart J. 2003; 24: 1554–1559.
Tan WA, Tamai H, Park SJ, Plokker HW, Nobuyoshi M, Suzuki T, Colombo A, Macaya C, Holmes DR Jr, Cohen DJ, Whitlow PL, Ellis SG. Long-term clinical outcomes after unprotected left main trunk percutaneous revascularization in 279 patients. Circulation. 2001; 104: 1609–1614.
Morin RL, Gerber TC, McCollough CH. Radiation dose in computed tomography of the heart. Circulation. 2003; 107: 917–922.
Hoffmann U, Ferencik M, Cury RC, Pena AJ. Coronary CT angiography. J Nucl Med. 2006; 47: 797–806.
Surveillance conventional invasive coronary angiography is usually performed routinely 2 to 6 months after stent-supported percutaneous coronary intervention (PCI) of the left main coronary artery (LMCA), owing to the unpredictable occurrence of in-stent restenosis (ISR). Noninvasive evaluation of coronary artery stents with multislice computed tomographic coronary angiography (MSCTA) has to date been limited by the occurrence of metal-related artifacts that precluded accurate evaluation of the lumen inside the stent. This study evaluated the performance of 16- and 64-slice MSCTA to detect ISR after LMCA stenting. Seventy-four consecutive patients underwent MSCTA before invasive coronary angiography at a median of 8 months after PCI. Four patients were excluded because of a technically inadequate scan. MSCTA correctly identified all patients (n=10) with ISR and misclassified 5 patients without ISR. Overall, the sensitivity, specificity, and positive and negative predictive values were 100%, 91%, 67%, and 100%, respectively. MSCTA proved highly accurate (accuracy=98%) in the patient group (n=46) who had received a simple stenting procedure, with a false-positive rate of only 1 of 46. Patients with a complex true bifurcation stenting procedure showed a false-positive rate of 4 of 24 (accuracy=83%). Compared with IVUS, neointimal hyperplasia ≥1 mm in thickness was recognized with clinically relevant sensitivity. These data suggest that in contrast to the conventional thinking about the efficacy of MSCTA to detect in-stent abnormalities, in this population with LMCA stents, MSCTA may be effective to exclude ISR and potentially obviate the need for routine follow-up invasive angiography.