Two-dimensional (2D) frequency domain optical coherence tomography (FD-OCT) has enhanced our understanding of coronary atherosclerotic disease and is increasingly being used in conventional percutaneous coronary intervention (PCI) to elucidate mechanisms of disease and improve our understanding of complex coronary anatomy. Since the first report of three-dimensional (3D) OCT applied in human coronary vessels,1 the technology has rapidly progressed.2–10 Currently, the main limitation of this technology is the need for off-line creation of 3D reconstructions—prototypes of current generation ‘real time’ (i.e. available peri-procedurally at the ‘push-of-a-button’) remain experimental, work in progress, and are limited by relatively poor image quality/resolution.4 As of now, the potential clinical application of 3D FD-OCT remains undefined. Recently, the application of this emerging technology to the coronary bifurcation has allowed visualization and assessment of jailed side branches (SideBs) at a level of detail not previously reported.2–7 The assessment of a jailed SideB, after implantation of a bioresorbable scaffold in the main branch (MainB) of a bifurcation, lead to the proposal of a new classification system based on the assessment of the number of compartments the SideB ostium was divided into, with examples of how this may potentially effect the neointimal response and subsequent coverage of the struts.2 More recently, the application of this technology to the coronary bifurcation in patients implanted with conventional metallic stents, utilizing the Terumo optical frequency domain imaging (OFDI) system, was described for the first time.3 Hypotheses related to types of coronary bifurcation (‘parallel’ and ‘perpendicular’ bifurcations) based on the bifurcation angle, and how this leads to certain specific characteristics of the carina, which potentially made the SideB more vulnerable to the effects of carina shift and potential SideB closure, were described. Furthermore, the potential practical application of 3D FD-OCT in guiding the rewiring of the distal compartment of the SideB ostium—jailed with stent struts after MainB stenting—to minimize the risk of floating struts was demonstrated, something not easily achievable with conventional 2D FD-OCT systems or other intravascular imaging modalities.3,4,6 The potential for jailed stent struts at the SideB ostium, to act as a focus for neointimal ‘bridge’ formation and focal restenosis warranting further intervention, has also recently been demonstrated with 3D FD-OCT7—this was not so apparent on the corresponding 2D images; the practical suggestion from these findings was that final kissing balloon post-dilatation should be performed to clear any jailed struts at the SideB ostium. In addition, the use of 3D FD-OCT in potentially guiding the management of acute myocardial infarction has been reported.11 The identification of thrombus and stent malapposition to guide subsequent further aspiration thrombectomy, post-dilatation, and concomitant use of drugs was proposed. Three-dimensional reconstructions of other intravascular coronary imaging modalities have previously been described; they have however failed to find a useful clinical role.12–18 One of the reasons that potentially makes 3D FD-OCT more clinically applicable is the unrivalled resolution of OCT technology (10–15 µm) compared with other intravascular imaging modalities—coronary angiography: 100–200 µm, computed tomography (CT): 300–500 µm, intravascular ultrasound (IVUS): 100–150 µm, coronary angioscopy: <200 µm19—and consequent ability to visualize intraluminal structures in unmatched detail. Conversely, with, for example, IVUS, this cannot reliably visualize intraluminal structures/detail such as thrombus or stent apposition; furthermore, post-processing of 2D IVUS images is required to allow 3D IVUS reconstructions, thus limiting its application in the catheterization laboratory.12,13 Despite the appeal of fusion of other intravascular imaging modalities, such as CT and IVUS,20,21 these are still limited by the resolution of the images compared with OCT. Through a series of off-line 3D reconstructions performed with the Terumo OFDI system, in patients undergoing conventional PCI, followed by technical issues performed in a porcine model, the rapid progression of this emerging technology and the potential for clinical application are proposed. Opinion on the future development of 3D FD-OCT is also discussed. Three-dimensional frequency domain optical coherence tomography reconstructions of patients who underwent conventional PCI from the original First-In-Man study intracoronary Terumo OFDI system (Terumo Corporation, Tokyo, Japan),22 and of a single patient from a further ongoing study,23 are presented. All 3D reconstructions were performed at baseline as per the study protocol. The high-speed Terumo OFDI system is capable of acquiring 160 frames/s during the catheter pull-back, to a maximum speed of 40 mm/s; all images were acquired with a motorized pull-back of 20 mm/s. Comparatively, the current generation LightLab DragonFly C7 system (LightLab Imaging Inc., Westford, MA, USA) is capable of 100 frames/s. The higher frame rate of the Terumo OFDI system appears to be the primary reason why it is capable of producing superior 3D FD-OCT reconstructions compared with the current generation LightLab C7 system.3,22 The methodology for off-line 3D reconstructions has previously been described.2 In brief, manual detection of every strut in each cross-section, using bitmap sequences (704 × 704 pixels) generated from prior 2D OFDI frames, was performed and 3D reconstructions were generated utilizing volume-rendering software (INTAGE Realia, KGT, Tokyo, Japan). Technical issues relating to image quality are also demonstrated in a porcine model. ‘Fly-through’ views indicate a selected still image of an internal view of a vessel looking either downstream (proximal to distal vessel) or upstream (distal to proximal vessel). The fly-through (internal) view of the vessel is akin to the view obtained during endoscopy or angioscopy showing the internal lumen of the vessel. Longitudinal views are a cut-away view of the vessel down the longitudinal axis with the internal lumen viewed externally—wherever possible a view without the guide-wire shadow is shown. External views are taken from outside the vessel with views to show the internal vessel at the region of interest. A severe proximal right coronary artery (RCA) lesion treated with a 3.0 × 18 mm Xience V drug-eluting stent (Figure 1) is demonstrated on coronary angiography (left images). Three-dimensional longitudinal reconstructions pre- and post-intervention are illustrated. The longitudinal 3D reconstructions represent the horizontal segment of the RCA before the first curvature—the use of the white arrows, identifying the jailed SideBs, will help the reader to co-register the 3D reconstruction with the coronary angiogram. Note the almost identical characteristics of the minimum lumen area on coronary angiography and 3D reconstruction (yellow arrows), indicative of the high resolution of OCT. Also present are one non-jailed SideB (striped white arrow) proximal to the implanted stent, and two jailed SideBs (white arrows) within the implanted stent, in both the coronary angiogram and 3D reconstruction. Simple lesion: pre- and post-coronary angiography and corresponding three-dimensional longitudinal reconstructions. A significant RCA aorto-ostial lesion (not illustrated) was directly stented with a 3 × 15 mm Xience V stent (Figure 2, upper left image). Post-procedural longitudinal (right image) and downstream fly-through (lower left image) 3D reconstructions are demonstrated. Note the jailed SideB in the downstream fly-through view (lower left image), and the guide catheter tip (white arrow), with some of the same render as the vessel, applied by the volume-rendering software. Yellow asterisks highlight the ostium of the right ventricular (RV) branch seen on coronary angiography and longitudinal and fly-through 3D reconstructions. From a technical perspective—one of the difficulties with imaging the aorto-ostial lesions is ensuring that the catheter tip is sufficiently disengaged from the coronary ostium to allow visualization with the OCT imaging wire during the pull-back, but close enough to allow injection of contrast for blood clearance—with inevitable over-spilling of contrast in the aortic cusps—and thus allow appropriate imaging.24 Ostial lesion: post-procedural three-dimensional reconstructions for treatment of an ostial lesion; RV, right ventricular. As the OCT light cannot penetrate metallic struts, OCT can only image the endoluminal strut border. Consequently, a ‘shadow’ is cast behind the metallic strut, and the adjustment for the thickness of the strut and polymer is therefore required to determine the apposition of the stent strut. Based on this phenomenon, stent strut apposition by post-implantation OCT has been defined as embedded, protruding, or malapposed.25 With an embedded strut, the strut is buried in the intima for more than half its thickness; consequently, the shadow the strut casts is projected through the vessel wall, with no shadowing visible on the endoluminal vessel surface on post-implantation 2D and 3D FD-OCT imaging (Figure 3A). With a protruding strut, the stent strut is apposed to the intima but not embedded; consequently, a shadow appears to be ‘cast’ on the vessel surface and wall (due to the thickness of the metallic strut and polymer which the OCT light cannot penetrate)—this represents a potential limitation of 3D FD-OCT, as the protruding struts erroneously gives the impression of strut malapposition, when it is in actual intimal contact (Figure 3B). With a malapposed strut, there is no intimal contact with the stent strut and vessel wall and appropriately leads to a shadow on the vessel surface and wall (Figure 3C).25 (A) Embedded struts (i.e. strut buried in the intima for more than half its thickness): three-dimensional reconstruction with corresponding two-dimensional optical frequency domain imaging frames—no shadow is visible on the vessel wall. Actual Xience V stent with an interductile hinge is illustrated for comparison (yellow arrows). (B) Protruding struts (i.e. strut in intimal contact): three-dimensional reconstruction—note the shadow on the vessel wall (yellow arrows) can erroneously give the impression of stent malapposition due to the thickness of the metallic strut and polymer, which the optical coherence tomography light cannot penetrate. Corresponding protruding struts on two-dimensional optical frequency domain imaging frames are marked with an asterisk. (C) Malapposed struts: three-dimensional reconstruction with corresponding two-dimensional optical frequency domain imaging frames—true stent malapposition is illustrated (yellow arrows) with corresponding shadow on the vessel wall. Corresponding protruding struts on two-dimensional optical frequency domain imaging frames are marked with an asterisk. Overlapping Xience V stents, used to treat a long segment of disease consisting of severe tandem lesions arising from the proximal first diagonal, are demonstrated (Figure 4). Downstream fly-through 3D reconstructions demonstrate embedded, protruding (yellow circles), and occasional malapposed struts (yellow arrow). Note the malapposed ‘interductile hinge’ and the shadow it casts on the vessel wall visible in the 3D reconstruction (Figure 4)—for comparison, an interductile hinge in an actual Xience V stent is illustrated in Figure 3A (yellow arrows). Corresponding 2D OFDI frames of the overlap are shown (top left image). Overlapping Xience V stents: corresponding two-dimensional optical frequency domain imaging frames and downstream fly-through views are illustrated. Three-dimensional reconstructions are illustrated, post-implantation of a 3.0 × 24 Xience V drug-eluting stent with prior aspiration thrombectomy, following an inferior wall ST elevation myocardial infarction with occlusion of the mid-RCA. Downstream fly-through views (Figure 5A), post-stent implantation, demonstrate a large volume of thrombus (white arrows) and consequential stent malapposition at the proximal stent edges (yellow arrows)—note the shadow the malapposed struts cause on the proximal vessel wall (white arrows) and thrombus adhering to the stent strut at the coronary bifurcation. A longitudinal view of the same vessel (Figure 5B) demonstrates the extent of the thrombus (white arrows) and consequential stent malapposition at both stent edges (yellow arrows); also evident are over-hanging struts, with thrombus attached, at the bifurcation (upper white arrow). If the interventional cardiologist saw the 3D reconstruction after stent implantation, this is likely to have been crucial in the subsequent decision-making process in the use of further aspiration thrombectomy, post-dilatation, and concomitant use of drugs such as glycoprotein IIb/IIIa inhibitors. Further aspiration thrombectomy and post-dilation of the stent were performed with angiographic resolution of the thrombus (not illustrated). (A) Primary percutaneous coronary intervention—thrombus and malapposition: downstream fly-through view, post-stent implantation, in the context of an acute ST elevation myocardial infarction is illustrated. White arrows represent thrombus and yellow arrows consequential stent malapposition. Note the thrombus adhered to the stent at the SideB further downstream (close-up view is illustrated in Figure 6). (B) Longitudinal three-dimensional reconstruction demonstrating the extent of thrombus (lower white arrows), over-hanging struts and thrombus at the side branch (upper white arrow), and stent malapposition (yellow arrows). The promising potential clinical application of coronary bifurcations with 3D FD-OCT has recently been described by our group2,3 and is beyond the scope of this paper. Examples of 3D FD-OCT in the coronary bifurcation to demonstrate its potential clinical application are illustrated below. Close-up views of the RV branch of the RCA (Figure 6) from the previous study (Figure 5A and B) demonstrate the thrombus adhering to the over-hanging struts of the SideB. The principle of a ‘parallel bifurcation’ is demonstrated3—note the corresponding 2D OFDI frames on the left—with the parallel origins of the MainB and SideB at their respective point of take-off. The carina (labelled) appears interposed between the parallel origins of the MainB and SideB at their respective point of take-off; if further MainB post-dilatation was undertaken alone with larger angioplasty it may be that this lead to carina shift and potential SideB of a parallel Close-up view of SideB from Figure with thrombus to stent struts at the SideB ostium, in a three-dimensional frequency domain optical coherence tomography downstream fly-through Note how the MainB and SideB to parallel to each other at their respective point of take-off. Corresponding two-dimensional optical frequency domain imaging frames are illustrated on the A bifurcation lesion two branches of the with severe ostial disease in the first was treated with a × 18 mm bifurcation stent implanted in the with the SideB of the stent the other branch of the of the with no angiographic of the ostium, Longitudinal 3D FD-OCT reconstructions of the pre- and post-intervention bifurcation are demonstrated with corresponding 2D coronary (Figure ostium of the is not in the longitudinal 3D reconstructions illustrated, as it in a to the of the comparison, a bifurcation stent in a of a bifurcation is illustrated of right image). Note the area of malapposition with the corresponding shadowing on the vessel wall in the proximal stent (yellow corresponding 2D OFDI frames are illustrated (Figure The of the stent to be relatively there was however of (Figure white arrow) when compared with the model. (A) bifurcation longitudinal three-dimensional reconstructions of the pre- and post-intervention bifurcation are demonstrated with corresponding two-dimensional coronary (B) corresponding two-dimensional optical frequency domain imaging frames of the stent seen at the proximal of the bifurcation stent on three-dimensional frequency domain optical coherence tomography reconstructions (Figure (C) Downstream fly-through of the treated vessel pre- and post-implantation with the bifurcation Corresponding optical frequency domain imaging frames are illustrated below. downstream views of the treated vessel, pre- and are illustrated (Figure this view, no malapposition or floating struts are evident at the bifurcation. comparison, corresponding 2D OFDI frames post-implantation are illustrated below. proximal coronary disease was directly stented with a × 15 mm Xience V stent to the ostium of the no post-dilatation was performed (Figure and (A) and malapposition at the coronary longitudinal three-dimensional reconstructions, pre- and are illustrated with the respective coronary (B) at the coronary vessel three-dimensional reconstructions with downstream (upper and upstream (upper fly-through views and corresponding two-dimensional optical frequency domain imaging frames (lower demonstrating a single malapposed strut at the ostium (yellow arrows). Figure demonstrates pre- and 3D reconstructions with corresponding White in the post-implantation longitudinal 3D reconstruction is of stent (lower right image), as also seen in the corresponding coronary angiogram left Note the in the post-implantation longitudinal 3D reconstruction with an to the of the struts is in Figure and the malapposed strut at the ostium, with the shadow it casts on the vessel wall (yellow right image). the endoluminal point of view, for with views in the longitudinal (not illustrated) or with fly-through views looking downstream (Figure upper left yellow arrow) or upstream (Figure upper right yellow arrow), can help visualize stent which can be on corresponding 2D OFDI frames if (lower images). The corresponding 2D OFDI frames the of malapposed struts at the arrows indicate the same malapposed strut seen in the 3D reconstructions. coronary angiography was performed in a patient with a of risk and of the distal left main on coronary angiography possible as by (Figure upper right yellow arrow), with no lesion The area of was not easily on 2D FD-OCT only a high of that an was of the 2D OFDI frames and 2D longitudinal views was If 3D FD-OCT was this have potentially the area of and allowed assessment with the corresponding 2D Two-dimensional OFDI frames were of on 2D and longitudinal views (lower yellow arrows). Three-dimensional reconstructions demonstrated the visible in the distal left main arrows in downstream fly-through (upper left image) and longitudinal (upper right image) The patient was demonstrated on three-dimensional longitudinal and downstream fly-through views with corresponding two-dimensional optical frequency domain imaging frames and coronary angiogram. Two-dimensional OFDI were undertaken in a porcine implanted with a Xience V drug-eluting stent, to allow for the assessment of the technical issues relating to 3D FD-OCT reconstructions. The porcine study has previously been described and was by the and the study was in to the for and the for the and of the the speed of the is at the rapid and to the pull-back speed of the OFDI during image can potentially cause a as illustrated (Figure (A) The of in three-dimensional frequency domain optical coherence tomography the three-dimensional reconstructions were performed in a porcine model. (B) The types of in three-dimensional frequency domain optical coherence tomography and are illustrated. (C) from a human an upstream fly-through three-dimensional reconstruction of the left artery is illustrated. Since the with longitudinal the longitudinal axis of the vessel) and this can lead to types of such as or as illustrated in Figure white an white of same struts due to the and of the vessel the imaging wire during imaging in the white due to the of the coronary artery the imaging from a human study is shown in the fly-through view upstream in the of an (Figure pull-back were used to image quality and frequency of in the porcine model. Three-dimensional FD-OCT reconstructions were performed from 2D OFDI with pull-back of and 40 (Figure With pull-back the frequency of the due to a of during imaging for to however at the of a longitudinal resolution (due to a a more Conversely, with pull-back the image resolution of a longitudinal resolution (due to a this is however at the of an frequency of due to a imaging A between image resolution by the pull-back and the frequency of is therefore to the 3D reconstruction. The between pull-back and image resolution of the three-dimensional frequency domain optical coherence tomography as demonstrated in a porcine model. (A) represents 2D (OFDI) longitudinal represents corresponding 3D reconstructions with and consequential at the of a The for the Terumo OFDI system, in our is a pull-back speed of 20 mm/s; this will allow imaging limiting however will still as previously illustrated (Figure the LightLab C7 system, of the resolution of the 3D reconstructions, a pull-back speed of to improve the this however at the of frequency of with the risk of one of these at the region of interest. of 3D FD-OCT in our need to have higher frame frames/s to allow for rapid will potentially lead to higher resolution 3D reconstructions with limited to the unrivalled resolution of OCT compared with other intravascular imaging modalities, it is the that the potential clinical application of this as a to 2D FD-OCT is and if available at the when 2D intravascular imaging is will a with subsequent assessment with 2D FD-OCT imaging as Furthermore, the 3D reconstructions are to compared with the 2D which further to the appeal of this emerging The current of 3D FD-OCT systems have the to struts based on their specific optical such as the of the metallic The further from to potentially 3D FD-OCT a clinical the in strut higher systems with rapid pull-back to improve the resolution of the 3D reconstruction and the of volume of a to allow for software for and a to allow for of the vessel. With the as previously illustrated, as longitudinal and fly-through required to visualize the area of with the volume-rendering software in the in this any view of the vessel was achievable with should also be that the of of the coronary vessel is not previously been for in the catheterization The development of which is this technology is is of volume-rendering software to and application of volume has been with computed for the of the 3D are both crucial to allow for 3D reconstructions and which can have potential clinical and clinical of the 3D technology to allow for the assessment of of possible stent the of shadow on the vessel wall as seen with protruding and malapposed may subsequent with 2D imaging to help further if malapposition is on the 3D the may to further with this of protruding, or malapposed struts may be The application of this technology within the coronary bifurcation has previously been described and is a further promising area of the potential for this emerging to visualize the complex of the bifurcation and the effects of intervention is and not with 2D The of as previously may have a potential clinical application in the of the SideB ostial area after MainB to allow assessment as to the SideB is assessment of SideB after MainB has previously been shown to be with only of SideB with angiographic being to be significant on subsequent wire potential clinical previously demonstrated the identification and of the extent of thrombus and need for subsequent further aspiration thrombectomy, post-dilatation, and concomitant use of thrombus cannot be by the IVUS and if this imaging was the have to an as to malapposition is present or a understanding of stent in at the coronary the identification of other intravascular such as (Figure and stent OCT is not the intravascular imaging of for the assessment of stent due to the limited of OCT in intravascular IVUS be more in identifying the for appropriate vessel 3D reconstructions still an area stent was evident (Figure and may have the to further with or without IVUS Furthermore, the of with may allow of these of on longitudinal 3D views that may also guide subsequent The of within 3D FD-OCT reconstructions are further promising of in this these have previously been performed and have the potential to in the identification of clinically useful of such as and vulnerable imaging of 2D FD-OCT and IVUS has recently been described; this may be achievable with 3D Furthermore, the development of 2D and 3D a resolution of also recently been shown to clear of and with and in human coronary The of the potential of this emerging technology within interventional clinical appears to be the in the of with, for example, 3D visualization of the required from and in to allow the future development and of this to the of The has from the of All to their to and of the The of the who the original First-In-Man study of the intracoronary Terumo OFDI and of The and and of Terumo and Terumo for their technical
Farooq et al. (2011) studied this question.