Stent Evaluation by Coronary Computed Tomographic Angiography and Expert Consensus Documents

As coronary computed tomographic angiography (CCTA) has rapidly evolved, its extension to more complex scenarios has kept pace. Stent evaluation has intrinsic technical difficulties. However, since left main disease, in particular, is not reliably addressed by myocardial perfusion imaging, the potential importance of alternate technologies such as CCTA is increased. The paper by Veselka et al in the current issue of the Journal of Invasive Cardiology¹ evaluated the accuracy of CCTA to detect left main in-stent restenosis (ISR) in 34 patients, 3 (8.8%) with intravascular-ultrasound documented ISR. They reported sensitivity, specificity, and positive and negative predictive accuracies of 100%, 74%, 18%, and 100%, respectively. Despite the very high negative predictive value, the poor positive predictive accuracy and systematic underestimation of minimum luminal area suggested that CCTA was not an adequate tool for left main ISR evaluation.

These results differ from those of Van Mieghen et al² in an earlier series of 70 patients with left main stents, 10 with invasive-angiography documented ISR (14%), in which the sensitivity, specificity, and positive and negative predictive values were 100%, 91%, 67%, and 100%, respectively. The obvious difference between the 2 studies is the number of patients with ISR. In the current study, it is difficult to accept any conclusions based upon only 3 cases of ISR. Van Mieghen et al² had 10 patients with ISR, still inadequate for broad conclusions but more believable than a study group of 3.

However, there is a larger question: If the best-case, large stent, left main scenario yields poor positive predictive accuracy, is CCTA an appropriate tool to rule in ISR under any circumstances? The Executive Summary of the recently published 2010 Expert Consensus Document on Coronary CT Angiography (CTA) provides a negative (by omission) response: ‘‘The literature suggests that in patients who have large-diameter stents, good image quality, and whose clinical presentation suggests low to intermediate probability for restenosis, 64-channel coronary CTA can be used to rule out severe in-stent restenosis (ISR).’’³

In reality, the literature suggests quite the opposite. The accuracy of 64-slice CTA for ISR in a recent meta-analysis of 18 studies involving 1300 patients and 2003 stents⁴ yielded the following results: sensitivity 89.7%, specificity 92.2%, positive predictive accuracy 72.5%, negative predictive accuracy 97.4%, total accuracy 91.9%, and unevaluable stents 9.6%. Evaluability increased with larger diameter stents, thinner struts, cobalt chromium, and nonoverlapping stents. Moreover, the sensitivity and specificity for ISR detection, as well as for detection of obstructive disease in general, are superior to those for noninvasive stress testing, including stress electrocardiography, myocardial perfusion imaging, and stress echocardiography, both in comparison with meta-analyses of the technologies individually⁵ as well as in head-to-head comparisons of CTA with nuclear testing^6-8 and electrocardiography⁹ in the same individuals.

So, what is the CCTA practitioner to do? Listen to the consensus document or forge out on his own? The answer is in the rarely acknowledged disconnect that almost universally exists in all imaging modalities between published results from academic centers and community practice. It is fair to assume that only the more favorable results are submitted for publication from the high-volume university centers, which possess the latest state of the art scanners; these results may not resemble those obtained from lower-volume community hospitals with less sophisticated equipment and less experienced readers. More than any other imaging modality, the technical issues of CCTA are daunting, particularly for stent evaluation.

To deal with these issues, the following are recommended:

1. A scanner with a minimum of 64 slices is required. Studies may be acquired either retrospectively or prospectively.
2. Because stents accentuate blurring, motion artifact must be avoided as much as possible; adequate heart rate control with beta blockers to <60 beats/minute is imperative.
3. Iterative reconstruction algorithms that maximize spatial resolution and minimize blooming artifacts should be standard, using the same kV and mAs employed without these algorithms. The emphasis is on improving quality rather than reducing radiation. 
4. Sharp filters and wide window settings maximize intra-stent visualization and should be routinely used.
5. Multiplanar reconstruction and cross-sectional analysis are critical, as they are for routine CTA evaluation; maximum intensity projections alone are inadequate.
6. Hounsfield unit (HU) measurement should be routinely obtained for evaluating stent, contrast, and calcium densities.
7. Recognition of the critical role of overlying calcified plaque is absolutely critical to avoid false-positive ISR diagnoses. Because calcified lesions are often stented, the problem is a common one. The most problematic artifacts result from the higher-density calcium overlaying the stents rather than from the stents themselves, which are in the range of 300-350 HU. Because ISR is diagnosed by hypodensity within the stent, adjacent densely calcified plaque, generally >1000 HU, may ‘‘shadow’’ contiguous contrast, and the resulting photopenia may be misdiagnosed as the hypodensity diagnostic of ISR. The shadowing may occur both proximal and distal to the dense calcium as well as immediately adjacent in the same slice. Excessive broadening of the window settings to shrink the calcium ‘‘blooming’’ should be avoided, because the HU remain the same despite the change in appearance.  
8. Since densely calcified plaque may be so extensive that it may obscure stents, cross-sectional images are required to increase the likelihood of detecting the underlying struts, and facilitate identification of the heavily calcified stent. 
9. On occasion, calcium deposition in a non-stented vessel may be so uniformly severe that it mimics a stent. Reliance on the patient certitude that no stent has  been inserted may be the only way to avoid this mistake.
10. Appropriate patient selection will improve positive predictive accuracy. If patients with a low likelihood of ISR are not scanned, accuracy will improve. In this group, characterized by atypical pain or absence of symptoms more than one year after drug-eluting stent deployment, the expected low ISR rate will increase the false-positive rate to unacceptable levels.
11. Calculation of stent luminal areas, as performed in the present paper, should be avoided because of blooming artifacts and their sequelae, as discussed above. 
12. Angiographic verification of the diagnosis is crucial to improving interpretive skills. Feedback from the catheterization laboratory will help identify the interpretive patterns that result in false positives and negatives.

Expert Consensus Documents, by their very nature, are outdated by the time of publication, without access to the data and technologies introduced after the paper was submitted. Consequently, the 2010 Document³ may have little bearing on what should or should not be done currently, particularly with the recent addition of iterative reconstruction algorithms. Attention to the above recommendations will facilitate accurate evaluation of the most complicated stent pathologies.¹⁰

References

1. Veselka J, Cadova P, Tomasov, P, Theodor A, Zemanek D. Dual-source CT angiography for detection and quantification of in-stent restenosis in the left main coronary artery: comparison with intracoronary ultrasound and coronary angiography. J Invasive Cardiol. 2011;23(11):460-464.
2. Van Mieghem CAG, Cademartiri F, Mollet NR, et al. 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. Circulation. 2006;114(7):645-653.
3. Mark DB, Berman DS, Budoff MJ, et al. ACCF/ACR/AHA/NASCI/SAIP/SCAI/SCCT 2010 expert consensus document on coronary computed tomographic angiography. J Am Coll Cardiol. 2010;55(23):2663-2699. 
4. Andreini D, Pontone G, Mushtaq S, Pepi M, Bartorelli AL. Multidetector computed tomography coronary angiography for the assessment of coronary in-stent restenosis. Am J Cardiol. 2010;105(5):645-655.
5. Hecht HS. A paradigm shift: coronary computed tomographic angiography before stress testing. Am J Cardiol. 2009;104(4):613-618. Epub 2009 May 21.
6. Schuijf JD, Wijns W, Ravipati JWJ, et al. Relationship between noninvasive coronary angiography with multi-slice computed tomography and myocardial perfusion imaging. J Am Coll Cardiol. 2006;48(12):2508-2514. Epub 2006 Nov 28.
7. Ravipati G, Aronow WS, Lai H, et al. Comparison of sensitivity, specificity, positive predictive value, and negative predictive value of stress testing versus 64-multislice coronary computed tomography angiography in predicting obstructive coronary artery disease diagnosed by coronary angiography. Am J Cardiol. 2008;101(6):774-775.
8. Schuijf JD, van Werkhoven JM, Gabija Pundziute G, et al. Invasive versus noninvasive evaluation of coronary artery disease. JACC Cardiovasc Imaging. 2008;1(2):190-199.
9. Kristian A, Øvrehus KA, Jensen JK, et al. Comparison of usefulness of exercise testing versus coronary computed tomographic angiography for evaluation of patients suspected of having coronary artery disease. Am J Cardiol. 2010;105(6):773-779.
10. Hecht HS, Gade C. Current and evolving stent evaluation by coronary computed tomographic angiography. Cathet Cardiovasc Interv. 2011;77(6):843-889.

________________________________

From Lenox Hill Heart and Vascular Institute, New York, New York.
Disclosure: The author has completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr. Hecht has received grant support from Philips Medical Systems and participates on their speakers’ bureau.
Address for correspondence: Harvey S. Hecht, MD, FACC, FSCCT, Lenox Hill Heart and Vascular Institute, 130 E. 77th St., New York, NY 10021. Email: hhecht@aol.com