Author + information
- Received April 4, 2015
- Revision received May 1, 2015
- Accepted May 5, 2015
- Published online July 1, 2015.
- ∗Department of Interventional Cardiology and Angiology, Institute of Cardiology, Warsaw, Poland
- †Department of Internal Medicine 2 (Cardiology), University of Erlangen, Erlangen, Germany
- ↵∗Reprint requests and correspondence:
Dr. Maksymilian P. Opolski, Department of Interventional Cardiology and Angiology, Institute of Cardiology, Alpejska 42, 04-628 Warsaw, Poland.
Coronary computed tomography angiography (CTA) is increasingly used to diagnose and rule out coronary artery disease. Beyond stenosis detection, the ability of CTA to visualize and characterize coronary atherosclerotic plaque, as well as to obtain 3-dimensional coronary vessel trajectories, has generated considerable interest in the context of pre-procedural planning for revascularization of chronic total occlusions (CTOs). Coronary CTA can characterize features that influence the success rate of percutaneous coronary intervention (PCI) for CTOs such as the extent of calcification, vessel tortuosity, stump morphology, presence of multiple occlusions, and lesion length. Single features and combined scoring systems based on CTA may be used to grade the level of difficulty of the CTOs before PCI and have been shown to predict procedural success rates in several trials. In addition, the procedure itself may be facilitated by real-time integration of 3-dimensional CTA data and fluoroscopic images in the catheterization laboratory. Finally, the ability of coronary CTA to assess anatomy, perfusion, and viability in 1 single examination makes it a potential “one stop shop” that predicts not only the likelihood of successful PCI but also the clinical benefit of CTO revascularization. Further research is clearly needed, but many experienced sites have already integrated coronary CTA into the routine planning and guiding of CTO procedures.
- chronic total occlusion
- coronary angiography
- coronary computed tomography angiography
- percutaneous coronary intervention
Chronic total occlusions (CTOs), sometimes referred to as the “final frontier in interventional cardiology,” represent a major challenge for percutaneous coronary intervention (PCI) (1). Although leaving CTOs untreated is associated with poor prognosis (2,3), current PCI attempt rates for CTOs are surprisingly low, and most patients are managed medically (4). To further complicate matters, success rates of PCI in CTOs are usually moderate (with higher success in some luminary expert centers) (5,6), and failed revascularization increases the risk of adverse outcomes (7). Therefore, careful selection of patients in which successful PCI for CTOs can be expected is of high clinical relevance. Furthermore, a variety of modern imaging modalities, dedicated devices, and techniques have been developed in an effort to overcome difficulties encountered in percutaneous treatment of CTOs (8–10). Coronary computed tomography angiography (CTA) can characterize features that influence the success rates of PCI for CTOs (most of which are not readily seen on invasive coronary angiography), and it is therefore increasingly used to predict the likelihood of successful recanalization (10). The present review outlines established and potential applications of coronary CTA for the pre-procedural planning and guidance of PCI for CTOs.
Definition, Prevalence, and Clinical Characteristics of CTOs
A CTO of a coronary artery is defined as the total occlusion of the vessel on invasive angiography with complete interruption of antegrade blood flow (Thrombolysis In Myocardial Infarction flow grade 0) and an age of the occlusion ≥3 months proven by prior angiography or estimated from the clinical course (11). Although the true prevalence of CTOs in the general population is unknown, chronic occlusions are common among patients with known or suspected coronary artery disease, and they are found in 15% to 30% of all subjects undergoing coronary angiography (1,4). CTOs are associated with advancing age and cardiovascular risk factors, and they are most prevalent in the right coronary artery, followed by the left anterior descending coronary artery (4).
Rationale for CTO Revascularization
The aim of CTO revascularization is the same as for coronary stenoses, namely improvement in symptoms or prognosis, which requires that a sufficiently large and CTO-specific territory of viable and ischemic myocardium is present. There is conclusive evidence indicating substantial symptomatic improvement, recovery of left ventricular function, and a reduced need for bypass surgery among CTO patients after successful PCI (1). In addition, survival is better after successful (compared with failed) PCI of CTO lesions, which may, however, be partly attributed to the deleterious effects of failed PCI (2,7). For these reasons, CTO revascularization is fully sanctioned according to current guidelines when persistent symptoms and/or objective evidence of viability/ischemia exist (12).
Current Trends in the Management of CTOs
Despite marked advances in endovascular techniques and device technology for PCI, most patients with CTO are managed medically (4). The barriers to more frequent application of PCI in CTOs are numerous. First, percutaneous recanalization of CTOs requires specific expertise and, even then, it remains technically challenging with lower procedural success rates (60% to 85%) compared with subtotal stenoses (98%) (13). Second, the unpredictable and often long procedural times complicate logistics in the catheterization laboratories and may limit physician willingness to attempt CTO recanalization (14). Third, the incomplete visualization of CTO on invasive angiography, with no objective means to precisely identify the vessel trajectory during the “hit-or-miss” wire manipulation, increases the risk of serious complications (10). Thus, there is substantial interest to use imaging for pre-procedural CTO characterization, and coronary CTA is the prime candidate.
Diagnosis of CTO on Coronary CTA
A CTO may be incidentally identified on a diagnostic coronary CTA, or a coronary CTA may be specifically performed before attempting CTO recanalization. Although CTOs are defined based on the angiographic criterion of antegrade blood flow interruption, coronary CTA typically replicates angiographic findings by demonstrating a complete lack of contrast opacification within the coronary occlusion on multiplanar reformations (Figure 1) (10). Until now, no systematic computed tomography (CT) comparison between the characteristics of acute versus chronic total coronary occlusions has been available, and reliable identification of CTOs relies more on clinical history than on specific CT morphological findings. However, typically, the distal vessel lumen of a CTO is opacified with contrast on coronary CTA, and acute occlusions may exhibit a particularly pronounced degree of positive remodeling (10).
In fact, coronary CTA may have difficulty in differentiating high-grade stenoses from complete occlusions, which is a consequence of its limited spatial resolution (Figure 2). Lesion length can be an indicator of total occlusion, with lesions ≥9 mm regarded as likely to represent complete occlusion (15). Furthermore, the recently described “reverse attenuation gradient sign” has been proposed as a highly specific feature of CTO on coronary CTA (16).
Characterization of CTO on Coronary CTA
Invasive coronary angiography provides incomplete information on vessel morphology in occluded coronary artery segments. Coronary CTA, on the other hand, reliably visualizes the occluded coronary segment and is able to identify and quantitatively evaluate morphological (e.g., calcification) and anatomical (e.g., tortuosity, length) features of the occlusion (10). Similarly, the distal vessel territory is often more clearly depicted on coronary CTA than in invasive angiography (Figure 3) (17). Two- and 3-dimensional reconstructions enable visualization of the exact vessel trajectory with accurate measurement of the occlusion length and precise mapping of vessel tortuosity (10). Coronary CTA can also be used to identify ostial occlusions, a feature that may be missed in invasive angiography (Figure 4). Advantages of coronary CTA over standard, catheter-based angiography are most apparent in long and tortuous CTO lesions or when the distal CTO segment cannot be visualized in invasive angiography (10,17).
Prediction of Procedural Outcome of PCI by Coronary CTA
Angiographic studies have consistently reported that previously failed recanalization, increasing age of the occlusion, presence of marked calcification, long occlusion length, a blunt proximal stump, excessive vessel tortuosity, presence of a side branch at the occlusion site, and antegrade bridging collaterals decrease the likelihood of successful guidewire crossing in CTO (18). Recently, the J-CTO (Multicenter CTO Registry of Japan) score has been introduced as an accurate angiographic prediction rule for time-efficient guidewire crossing that determines the difficulty level of CTOs (14). Importantly, with the availability of improved techniques and devices for PCI, some of the historical predictors of failed revascularization (e.g., side branch at the occlusion entry, presence of antegrade bridging collaterals) no longer seem valid. Consequently, a number of CT features of CTO have been identified for the prediction of PCI success by using contemporary percutaneous techniques (Table 1).
Calcification is the hallmark of a high difficulty level of CTOs. It poses difficulties at all steps of the procedure, hampering successful guidewire passage, lesion pre-dilation, and adequate stent expansion (18). In this regard, an increase in calcium content over time may explain the progressive difficulty in successfully treating older occlusions (1). Of note, several studies have consistently indicated that coronary CTA is more sensitive to detect, quantify, and localize calcification compared with invasive angiography (19–21).
The influence of CTA-derived calcification on PCI outcome in CTO was first reported by Mollet et al. (22). They used 16-slice CT scan and demonstrated that calcification involving >50% of the vessel cross-sectional area along with an occlusion length >15 mm were independent predictors of guidewire crossing failure in a series of 47 lesions. Another small CT study conducted by Soon et al. (19) included 39 patients with 43 chronically occluded vessels; they found that calcification ≥50% of vessel cross-section was the only independent correlate of unsuccessful PCI, conferring a 10-fold higher risk for procedural failure.
Subsequently, a larger study using 64-detector row CT scan in a series of 142 CTO lesions corroborated the decisive role of calcification in reducing the probability of successful PCI (17). Of note, Cho et al. (23) performed a comprehensive calcium quantification assessment in a group of 64 patients with 72 CTO lesions and found that only the relative percentage of calcium area divided by the vessel area in the cross-section with the highest calcium content was an independent determinant of PCI failure. Interestingly, they identified a cross-sectional calcium area of 54% as the best cutoff value for predicting PCI failure. Hence, a simplified approach of visual calcium assessment that uses a 50% cutoff value makes intuitive sense. Consequently, it is not the length of calcium but rather the extent of cross-sectional vessel calcification that indicates crossing difficulty. Two additional studies have suggested that calcification at the entry (but not at the exit) site of CTO constitutes the major hurdle of successful guidewire crossing when using the antegrade approach (17,24). It should be noted that Hsu et al. (25), in a group of 67 patients with 82 de novo CTO lesions, found that a calcification length ratio >0.5 (defined as calcification length to occlusion length) was an independent predictor of failed PCI. However, they did not account for the extent of cross-sectional calcification. Finally, although several different approaches for calcium quantification have been reported thus far, most of the CT studies consistently suggest the practical utility of severe calcification involving ≥50% of the vessel cross-sectional area for prediction of PCI failure in CTO (17,19,21,22).
CTO interventions require negotiating the wire along the correct track of the occluded segment. Marked vessel tortuosity (often referred to as bending) substantially increases the difficulty of this task and is associated with a greater risk of coronary injury, especially if stiffer wires are selected (18).
Yokoyama et al. (26) were the first to highlight the ability of coronary CTA to depict marked CTO bending that could not be identified on coronary angiography. In a study by Ehara et al. (27) of 110 patients with the same number of CTOs, bending (defined as an angle >45° either at the occlusion site or proximal to the occluded segment) was the most prominent independent predictor of wiring failure (followed by vessel shrinkage and severe calcium). Subsequently, García-García et al. (17) showed that vessel tortuosity throughout the occluded segment (but not proximal to the lesion) was associated with procedural failure. In their study, however, tortuosity failed to independently predict PCI success beyond severe calcium.
Occlusion length and multiple occlusion sites
Historically, the Euro CTO Club considered an occlusion length >20 mm as an important predictor of failed recanalization (18). Consequently, multiple reports showed that accurate measurement of occlusion length is easily possible on coronary CTA and that lesion length may be significantly underestimated in invasive angiography (19,20,22). However, the relevance of CTA-derived occlusion length as an independent predictor of PCI failure is unclear. In some studies, the presence of a long occluded segment on coronary CTA (defined as either >15 mm or >18 mm or ≥32 mm or as a continuous variable) was found to be an independent predictor of procedural failure (22,28–30), whereas other reports failed to demonstrate any relevant relationship between CTA-derived occlusion length and PCI outcome (17,19,21,24,27). These findings suggest that contemporary PCI techniques may be less affected by lesion length when attempting CTO recanalization.
Recently, it was shown that up to one-fourth of all CTO lesions include multiple occlusion sites, a characteristic clearly seen on coronary CTA but usually missed in invasive angiography (31). The influence of the presence of multiple occlusions on recanalization success was tested in the CT-RECTOR (CT-Registry of Chronic Total Occlusion Revascularization) registry. The presence of multiple occlusions showed a robust and independent association with guidewire crossing, while occlusion length did not (21). The underlying reason for the negative effect of multiple occlusions on recanalization success is likely due to the fact that multiple entry and exit sites increase the difficulty in achieving successful guidewire passage into the distal true lumen.
In invasive coronary angiography, the presence of a blunt stump is related to failed guidewire crossing, whereas a tapered shape of the proximal stump is considered a key marker for a potentially patent residual lumen (“microchannel”) that facilitates guidewire negotiation across the proximal cap (18).
CTA studies have provided conflicting results regarding the influence of stump morphology on PCI success rates in CTO. Two reports demonstrated a significant influence of stump morphology on PCI success based on multivariable regression analyses of CTO parameters (21,22). Conversely, several other CTA studies failed to show any relevant relationship between stump appearance and the procedural success of PCI (17,23,24,27). Interestingly, in a study by Li et al. (32) including 88 CTO lesions, a CTA-visible linear intrathrombus enhancement within the occluded segment (reflecting the presence of “microchannels”) has been suggested as a robust predictor of PCI success. Of note, according to Rolf et al. (20), coronary CTA seems to be more sensitive for the detection of a blunt stump than invasive angiography.
Shrinkage is defined as an abrupt narrowing of the occluded segment of CTO to <1 mm in vessel cross-sectional diameter and has been postulated to reflect the age of the occlusion (27).
The influence of vessel shrinkage (as defined earlier) on successful guidewire passage has thus far only been investigated in a single CTA study. In a series of 110 CTO lesions, shrinkage was the second most prominent and independent predictor of wiring failure (27). Of note, a recent study using a 256-detector CT in a series of 108 CTO lesions found that negative remodeling (defined as the ratio of the diameter of the occluded vessel to the diameter of the adjacent normal vessel <1), reflecting vessel shrinkage, was the strongest predictor of failed antegrade PCI (29). As a potential limitation, it may be difficult to reliably measure the vessel diameter of the occluded segment in CTA, given its typically low-contrast attenuation relative to the surrounding tissue.
The CT-RECTOR score was developed as a noninvasive prediction tool that combines several CT-derived parameters to predict recanalization success in CTO (21). It was derived from 240 consecutive CTO lesions included in a multicenter registry, and the outcome variable was the ability to successfully cross the CTO with a guidewire within 30 min. The advantage of this endpoint (as opposed to any success) is the fact that it provides some independence from individual operator skills (14). Multivariable analysis yielded 6 independent and straightforward CTA-based and clinical variables that were associated with guidewire crossing: the presence of multiple occlusions, a blunt stump, severe calcification, bending, a CTO age ≥12 months, and previously failed PCI. The CT-RECTOR score was calculated by assigning 1 point for each independent predictor and adding them per CTO lesion (Figure 5). With an increasing score, the likelihood of successful guidewire crossing within 30 min decreased, ranging from 95% for a CT-RECTOR score of 0 to 22% for a CT-RECTOR score of ≥3 (Table 2). Of note, the CT-RECTOR score exceeded the discriminatory performance of the angiography-based J-CTO score. In contrast to the J-CTO score, the CT-RECTOR score consists of 6 variables, including multiple occlusions and CTO duration, and excluding the lesion length. Although the relatively simple CT-RECTOR score gave equal weight to all predictors of recanalization failure, a more refined stratification may be achieved by putting more weight on strong predictors such as the cross-sectional degree of calcification, but this remains to be determined in the future. Clinically, CT-based findings such as the CT-RECTOR score may be used to assess CTO difficulty level before PCI (Figures 6 and 7⇓⇓), both to help decide whether recanalization should be attempted and to allocate resources (e.g., particularly experienced operators, CTO-dedicated devices, procedural time) specifically to complex CTO lesions.
Planning the Procedure With Coronary CTA
The availability of detailed anatomical information of coronary arteries before PCI is essential, and coronary CTA has been shown to aid in procedural planning of CTOs. Indeed, the most optimal PCI techniques and devices can be selected based on the morphological and anatomical features of CTOs on coronary CTA. For example, the depiction of severe calcium in the proximal CTO cap, together with a good retrograde collateral flow, offers a reasonable justification for choosing the retrograde approach, especially after previously failed PCI. In addition, focally distributed severe calcium can lead to selection of stiff flat or tapered wires, with a subsequent exchange to soft polymeric wires once calcification is crossed. Similarly, for heavily calcified lesions, the need for additional debulking devices (e.g., high-pressure balloons or rotablation) can be anticipated. Finally, 3-dimensional coronary CTA may aid in the pre-procedural selection of the most suitable fluoroscopic projection angles, thus minimizing additional contrast use during invasive angiography (10).
Of particular interest, coronary CTA can be directly used in the catheterization laboratory during PCI. One approach has been to import the CTA datasets directly to the catheterization laboratory with a possibility of automatic alignment according to the angulation of the C-arm (10). This method allows the operator to verify the direction of the guidewire advancement relative to the course of the occluded vessel segment and to identify angulations of the C-arm that avoid foreshortening of the treated vessel segment. This has been recently corroborated in a study by Rolf et al. (20), who reported significantly higher success rates for PCI with pre-procedural CTA displayed in the catheterization laboratory compared with PCI without CTA. Alternatively, a more complex approach of real-time integration of 3-dimensional coronary CTA data and fluoroscopic images in the catheterization laboratory has been proposed (Figure 8) (10). However, while recanalization of CTO based on CTA datasets in combination with magnetic navigation has already been reported in a small series of 43 patients (33), such an approach demonstrated only a moderate success rate (44.2%). Indeed, fusion techniques are hampered by patient, cardiac, and respiratory motion artifacts and thus do not provide the sub-millimeter accuracy required to determine the exact wire position relative to the distal vessel lumen (e.g., an intraluminal versus subintimal position of the guidewire).
Obviously, pre-procedural coronary CTA adds contrast and radiation exposure, and there is yet no uniform policy among CTO operators (not to mention current guidelines) regarding whether coronary CTA should be routinely recommended before attempting CTO recanalization (11,34). Thus, for now, coronary CTA should be specifically considered for patients with challenging or ambiguous CTO lesions as evaluated by using invasive coronary angiography. However, strategies to limit radiation exposure are constantly being refined, and ultra-low-dose protocols with estimated effective radiation doses <1 mSv and preserved image quality have been reported (35). Moreover, the contrast volume required for coronary CTA can be effectively reduced (36), and scheduling the CT acquisition at least a few days before the PCI attempt should not increase the risk of contrast nephropathy. Finally, given the additional cost of pre-procedural CTA, an economic balance for noninvasive optimization of patient referral patterns and allocation of resources during percutaneous treatment of CTO should be addressed in future CT trials.
Coronary CTA is uniquely suited to visualize the anatomical features of both the occluded and distal vessel segments in chronically occluded coronary arteries. The value of coronary CTA to predict the success of percutaneous recanalization of CTOs has been thoroughly studied in numerous trials (including a total of 1,269 CTO lesions), and several relevant predictors of PCI failure have been identified. They include severe calcification, bending, blunt stump, the presence of multiple occlusions, and occlusion length. CTA-based scores that combine several such parameters may be particularly useful in differentiating between “easy” and “difficult” CTO lesions, and they further support treatment decisions as well as allocation of resources in the catheterization laboratory. Of particular interest, real-time integration of 3-dimensional CTA datasets and fluoroscopic images may modify and optimize recanalization strategy. The Central Illustration summarizes the established and developing roles of coronary CTA in the management of CTOs. Although more data are needed to identify the full potential of coronary CTA for planning percutaneous intervention in CTO, coronary CTA is rapidly gaining recognition as a highly useful tool in interventional cardiology.
Dr. Achenbach has received grant support and Speakers Bureau honoraria from Siemens Healthcare and Abbott Vascular. Dr. Opolski has reported that he has no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- computed tomography
- computed tomography angiography
- chronic total occlusion
- percutaneous coronary intervention
- Received April 4, 2015.
- Revision received May 1, 2015.
- Accepted May 5, 2015.
- American College of Cardiology Foundation
- Stone G.W.,
- Kandzari D.E.,
- Mehran R.,
- et al.
- George S.,
- Cockburn J.,
- Clayton T.C.,
- et al.
- Fefer P.,
- Knudtson M.L.,
- Cheema A.N.,
- et al.
- Galassi A.R.,
- Ganyukov V.,
- Tomasello S.D.,
- Haes B.,
- Leonid B.
- Khan M.F.,
- Brilakis E.S.,
- Wendel C.S.,
- Thai H.
- Montalescot G.,
- Sechtem U.,
- Achenbach S.,
- et al.
- Grantham J.A.,
- Marso S.P.,
- Spertus J.,
- House J.,
- Holmes D.R. Jr..,
- Rutherford B.D.
- Morino Y.,
- Abe M.,
- Morimoto T.,
- et al.
- Opolski M.P.,
- Achenbach S.,
- Schuhbäck A.,
- et al.
- Luo C.,
- Huang M.,
- Li J.,
- et al.
- Opolski M.P.,
- Kepka C.,
- Achenbach S.,
- et al.
- Werner G.S.,
- Hecht H.,
- Stone G.W.
- Achenbach S.,
- Marwan M.,
- Ropers D.,
- et al.
- Definition, Prevalence, and Clinical Characteristics of CTOs
- Rationale for CTO Revascularization
- Current Trends in the Management of CTOs
- Diagnosis of CTO on Coronary CTA
- Characterization of CTO on Coronary CTA
- Prediction of Procedural Outcome of PCI by Coronary CTA
- Planning the Procedure With Coronary CTA
- Study Limitations