Optical coherence tomography in coronary atherosclerosis assessment and intervention

Abstract

Since optical coherence tomography (OCT) was first performed in humans two decades ago, this imaging modality has been widely adopted in research on coronary atherosclerosis and adopted clinically for the optimization of percutaneous coronary intervention. In the past 10 years, substantial advances have been made in the understanding of in vivo vascular biology using OCT. Identification by OCT of culprit plaque pathology could potentially lead to a major shift in the management of patients with acute coronary syndromes. Detection by OCT of healed coronary plaque has been important in our understanding of the mechanisms involved in plaque destabilization and healing with the rapid progression of atherosclerosis. Accurate detection by OCT of sequelae from percutaneous coronary interventions that might be missed by angiography could improve clinical outcomes. In addition, OCT has become an essential diagnostic modality for myocardial infarction with non-obstructive coronary arteries. Insight into neoatherosclerosis from OCT could improve our understanding of the mechanisms of very late stent thrombosis.

The appropriate use of OCT depends on accurate interpretation and understanding of the clinical significance of OCT findings. In this Review, we summarize the state of the art in cardiac OCT and facilitate the uniform use of this modality in coronary atherosclerosis. Contributions have been made by clinicians and investigators worldwide with extensive experience in OCT, with the aim that this document will serve as a standard reference for future research and clinical application.

Graphical Abstract

Appropriate use of optical coherence tomography (OCT) for optimization of therapies for coronary atherosclerosis is based on proper knowledge of the methodology, terminology, and clinical relevance of OCT. Global input from clinicians and investigators with extensive OCT experience from the United States, Asia, Canada, India, and Australia was incorporated, with the aim of promoting the uniform use of OCT in coronary atherosclerosis. ACS = acute coronary syndrome

Introduction

Optical coherence tomography (OCT) is an imaging technology for evaluating the microstructure of coronary arteries at an axial resolution of approximately 10–15 μm.¹ Due to its higher axial resolution than intravascular ultrasound (IVUS), OCT can characterize the superficial structure of the vessel wall in greater detail.^2–4 Over the past two decades, a large number of OCT studies have been reported.

The utility of OCT imaging is based on correct acquisition and interpretation of images as well as knowledge of the currently available evidence. This document aims to facilitate the uniform use of OCT in coronary atherosclerosis. Global input from clinicians and investigators with extensive OCT experience was incorporated into the original draft prepared at the Massachusetts General Hospital (MGH) on the methodology, terminology, and clinical relevance of OCT. More than one hundred OCT experts from Europe, the United States, Asia, Canada, India, and Australia endorsed this review document (listed at the end of the text). The strategy for intracoronary imaging-guided optimization of percutaneous coronary intervention (PCI) largely followed the recommendations of the European Association of PCI (EAPCI).^{5, 6}

At present, there are two major commercially available intracoronary optical-based imaging systems: OCT made by Abbott Vascular (Santa Clara, CA, USA) and optical frequency domain imaging (OFDI) made by Terumo Corporation (Tokyo, Japan). Since there is no difference in the image interpretation between the two systems, the following descriptions about OCT also apply to OFDI.

Physical principles of OCT imaging

OCT is a near-infrared, light-based imaging modality that generates high-resolution cross-sectional images of tissue microstructure.¹ The penetration depth depends on the tissue type and ranges from 0.1 to 2.0 mm, which is less than IVUS. Physical principles of OCT imaging are detailed in the Supplementary Material. OCT equipment is also described in the Supplementary Material.

OCT image acquisition

OCT should be used with caution in patients with a single remaining vessel, markedly impaired renal function, or known allergy to contrast media. In lesions with near-complete or total occlusion of the artery, it is recommended that antegrade blood flow is restored before OCT examination to allow the flushing media to adequately clear blood from the artery. This can usually be achieved by pre-dilatation using an under-sized balloon or by aspiration thrombectomy. If subintimal wiring is suspected in cases of chronic total occlusion, OCT should be spared because injection of flushing media may force progression of coronary dissections. OCT catheters are sometimes unable to pass through severely narrowed, calcified, or tortuous lesions or may be unable to follow the guide wire in arteries such as the left circumflex with a sharp angle from the left main coronary artery. The use of a guide extension catheter may help to deliver the catheter into the left circumflex without prolapse.

If not contraindicated, image acquisition should be conducted after the administration of intracoronary nitroglycerin to minimize catheter-induced vasospasm with appropriate attention to anticoagulation. OCT imaging catheters require careful purging with viscous undiluted contrast to remove air or blood before the interrogation. The OCT imaging catheter is advanced into the distal coronary artery over a standard angioplasty guide wire (0.014-inch). Automated OCT pullback is performed during either manual contrast injection through the guiding catheter using a syringe or automatically using a power injector (typical flush rate 4.0–5.5 ml/s for the left coronary artery and 3.0–4.0 ml/s for the right coronary artery at 300–400 psi).

Coaxial positioning of the guiding catheter into the coronary ostium is important for obtaining optimal blood clearance. A 6-Fr or larger diameter guiding catheter is recommended but OCT imaging catheters also fit 5-Fr guiding catheters or 6-Fr guide extensions. A guiding catheter without side holes may facilitate blood clearance. OCT acquisition can be performed through guide extension catheters. This approach may allow better target vessel intubation, providing high quality images for coronary ostial lesions and reducing contrast volume.

Since higher viscosity solutions provide superior results, iso-osmolar contrast media is used for flushing. In patients with impaired renal function, low molecular weight dextran, normal saline, or diluted contrast media may substitute for contrast media. It should be noted that low molecular weight dextran also has renal toxicity.

Understanding potential complications related to OCT imaging is important. In a prospective registry, catheter-related complications occurred in 0.6% and they were either self-limiting after retrieval of the imaging catheter or easily treatable in the catheterization laboratory.⁷ Contrast injection should be terminated as soon as imaging of the target segment has been completed because contrast volume during pullback was reported to be a risk factor for malignant arrhythmias.⁸ Safety data are detailed in the Supplementary Material.

OCT image display

There are three modes for displaying OCT images: the traditional cross-sectional image (orthogonal to the vessel), longitudinal view, and 3-dimensional visualization. These are detailed in the Supplementary Material. Contemporary OCT systems provide co-registration with angiography to allow matching cross-sectional OCT images with their position on angiograms, which facilitates plaque assessment and optimization of PCI.

OCT artifacts

The recognition of artifacts is critical for the correct interpretation of images. The OCT artifacts are summarized in Table 1 and representative images are shown in Figure 1.

Table 1.

Summary of OCT artifacts

Names	Characteristics
Guide wire	An artifact that is present in all current clinical OCT images is a sharp shadow caused by the guide wire. The guide wire will block the OCT light beam, so that tissue and stents behind the wire cannot be imaged (Figure 1A).
Gas bubbles	Gas bubbles can be present inside fluid-filled catheters. They can be recognized as distinct focal bright regions inside the catheter and a diminished signal in the part of the OCT image in the path of the affected region of the catheter (Figure 1B). These gas bubbles can form inside the OCT catheter if the operator does not adequately flush the catheter before introduction into the guide. Gas bubbles can also appear in the lumen of the vessel, injected through
Suboptimal vessel flushing	When blood is inadequately cleared from the field of the view, amorphous bright features can be present in the OCT image, because erythrocytes scatter light (Figure 1C). When mixed with optically transparent flushing media, blood may form many different patterns within the artery lumen. Blood can be confused with red thrombus or vessel wall dissections. Care should be taken when performing measurements and interpreting plaque components and stents through blood, as blood attenuates the OCT signal and therefore
Ghost lines	Ghost lines appear as circular features around the OCT catheter in the OCT image (Figure 1D). These lines appear in the image but are caused by light reflections from interfaces within the catheter or optical connections between OCT instrument and catheter. These artifacts should be identified and not interpreted as the OCT catheter sheath. If the ghost lines are misinterpreted and used for a calibration, measurements will be artifactually reduced. For
Non-uniform rotational distortion (NURD)	NURD is due to variations in the rotation rate of the catheter torque cable and distal optics and appears in OCT images as a blurring or smearing in the lateral (or rotational) direction (Figure 1E). NURD is normally caused by mechanical rotational resistance in the catheter due to either a tortuous or narrow vessel, a
Fold-over artifact	When the vessel diameter is larger than the ranging depth, a portion of the vessel may appear to fold over in the image (Figure 1F). These portions of the image data that contain fold-over artifacts should not be interpreted.
Tangential signal dropout	When the catheter is located very eccentrically, near the vessel wall, the optical beam may be directed almost parallel to the tissue surface. In these situations, the light beam is attenuated as it passes along the surface of the artery wall and the resultant artery wall may appear signal-poor below the luminal surface (Figure 1G). This artifact may be confused with the appearance of a thin-capped fibroatheroma (TCFA), intimal disruption, or
Seam artifact	If the catheter has moved with respect to the vessel during the time of single cross-sectional image acquisition, an axial discontinuity termed a seam artifact (Figure 1H), may appear at the location of the transition between the first and
Blooming	Blooming is the increase in strut reflection thickness in the axial direction of OCT images (Figure 1I), causing the small bright-line segment that represents the leading edge of a stent strut to appear thicker than normal, making it
Saturation	When a high reflector is encountered by OCT light, it may be backscattered at too high an intensity to be accurately detected by the detector, thereby causing saturation artifact (Figure 1J). Structures that exhibit high backscattering commonly include the guide wire, the tissue surface, and metallic stent struts.
Merry-go-round artifact	Merry-go-round artifact occurs when the small bright-line segments that represent a stent strut appear elongated in the lateral (rotational) direction (Figure 1K). Merry-go-rounding has also been described as an artifactual stretching of the stent strut. It is proposed that optical scattering in the lumen (primarily due to inadequate blood flushing) between the catheter and arterial
Sunflower artifact	When the OCT catheter is eccentrically located in the presence of metallic stent struts, a sunflower artifact can occur. Sunflower artifact occurs when stent struts appear to curve towards the OCT catheter, like sunflowers growing toward the sun (Figure 1L). This artifact may cause a well-apposed stent to appear malapposed in a vessel. This occurs because incident OCT light only

Figure 1. Artifacts.

(A) Guide wire (arrow) causes a shadow (asterisk). (B) Gas bubbles in the catheter (arrow) cause a shadow (asterisk). (C) Suboptimal vessel flushing leads to residual blood in the lumen (asterisk), hampering the assessment of the vessel (D) Ghost lines (arrow). (E) Non-uniform rotational distortion (NURD) (asterisk). (F) Fold-over artifact (arrow). (G) Tangential signal dropout. Blue lines indicate the direction of light beams tangential to the tissue that appears to have an artifactual light dropout. (H) Seam artifact (arrow). (I) Blooming. Stent struts (arrows) appear thicker than the actual thickness due to blooming artifact. (J) Saturation. Arrows point to high signal intensity artifact that extends along the axial dimension. (K) Merry-go-round artifact is artifactual stretching in the lateral (rotational) direction of stent struts (arrows), which occurs due to scatters in the lumen such as residual blood, clots and neointima. (L) Sunflower artifact may cause a well-apposed stent to appear malapposed in a vessel (white circles).

Basic measurements

It has been reported that reproducibility of OCT measurements was high.^10–14 Meaningful measurements require good-quality images that contain no or minimal artifacts. Although OCT systems provide automatic lumen detection, manual correction is required especially when the image quality is suboptimal. To make accurate measurements, the image should be correctly calibrated for z-offset. Although OCT systems offer an automated calibration, manual correction is sometimes required. Since calibration can vary depending on flushing medium, this needs to be adjusted. Proximal or distal reference is defined as the site with the largest lumen proximal or distal to a stenosis but within the same segment (usually within 10 mm of the stenosis, with no major intervening branches).⁹ This may not be the site with the least plaque.

The definition of measurements are as follows:

Arc. Arc measured using the center of the lumen as the vertex.

Depth. The distance between the luminal border and the adluminal surface of the plaque feature.

Thickness. The thickest distance between the inner and outer border of the plaque component (valid only if the deep boundary can be identified).

Area. The cross-sectional area (CSA) of the plaque component (valid only if the deep boundary can be identified).

Volume. The sum of areas of each cross-section multiplied by the slice thickness over the segment of interest (Simpson’s rule).

Cap thickness. The thickness of a cap present over calcium or lipid. Minimum fibrous cap thickness is conventionally measured 3 times at the thinnest point in magnified views, and the average value is calculated.¹⁵

Lesion length. The distance between the proximal and distal reference. Length measurement is detailed in the Supplementary Material.

Vessel and plaque morphology

Since OCT acquires multiple sequential images during the pullback, confirming a finding in more than one frame helps to improve diagnostic confidence. The following image interpretations are widely accepted.

Normal vessel wall

The media of the normal vessel appears as a lower signal intensity band relative to that of the intima and adventitia, providing a three-layered appearance (bright-dark-bright) (Figure 2A). The internal elastic membrane (IEM), or internal elastic lamina is defined as the border between the intima and media. The external elastic membrane (EEM), or external elastic lamina is defined as the border between the media and the adventitia. These membranes may be visualized as highly backscattering thin structures. Intimal thickening occurs during the early phase of coronary atherosclerosis¹⁶ and is represented as a signal-rich homogeneous thick intimal band in OCT images (Figure 2B). The definition of intimal thickening is limited by the lack of an established cutoff value.

Figure 2. Normal vessel and plaques.

(A) Normal vessel wall is characterized by 3-layered architecture, comprising a high backscattering, thin intima (i), a low backscattering media (m), a high backscattering adventitia (a), internal elastic membrane as the border between the intima and media (blue arrow), and external elastic membrane as the border between the media and the adventitia (yellow arrow). (B) Intimal thickening occurs during the early phase of atherosclerosis in the coronary artery, which is represented as a signal-rich thick intimal band (double-headed arrow). (C) Fibrous plaque exhibits homogeneous, signal-rich (highly backscattering) regions (asterisk). (D) Calcified plaque exhibits signal-poor regions with sharply delineated borders and limited shadowing (asterisk). (E) Lipid plaque exhibits signal-poor regions (asterisk) with diffuse borders and overlying signal-rich bands (double-headed arrow). (F) Thin-cap fibroatheroma (TCFA) is defined as a lipid plaque in which the minimum thickness of the fibrous cap (arrows) is less than a predetermined threshold and lipid occupies >90° in circumference.

Atherosclerotic plaque

An atherosclerotic plaque is defined as a mass lesion (focal thickening) or loss of a layered structure of the vessel wall.

Fibrous plaque

Fibrous plaque exhibits homogeneous, signal-rich (highly backscattering) regions (Figure 2C). Sometimes, the limited penetration depth of OCT precludes the accurate detection of signal-poor regions located deep within the vessel wall thereby limiting the assessment of lipid or calcium behind fibrous tissue.

Calcified plaque

Calcified plaque exhibits signal-poor regions with sharply delineated borders and limited shadowing (Figure 2D). Unlike IVUS, OCT can finely delineate calcification as OCT light penetrates calcium and it permits the quantification of both the arc and thickness¹⁷ if calcium is not too thick¹⁸. Calcification thickness can be measured only when both leading and trailing edges are visible. Calcium length is measured on the longitudinal view. Calcium index is calculated as the product of calcium arc and calcium length. These measurements can have a bearing on how to best prepare a calcified lesion before stent implantation. The EAPCI document recognized that a calcium arc >180° and calcium thickness >0.5 mm are risk factors for stent underexpansion.⁵

Lipid plaque

Lipid plaque exhibits signal-poor regions with diffuse borders (lipid pool) and overlying signal-rich bands (fibrous caps), accompanied by high signal attenuation (Figure 2E). Histopathologic validation studies show a good correlation between signal-poor OCT regions with diffuse borders and lipid pools on histology, corresponding to either a necrotic core or a region within pathological intimal thickening that contains extracellular lipid and/or proteoglycans.^{4, 19, 20}

Due to the limited penetration depth of OCT for lipid-containing tissues, OCT is not capable of measuring the thickness, area, or volume of lipids. Instead, the circumferential arc is used to semi-quantify lipids. Lipid-rich plaque is defined as a lipid plaque that has a lipid arc of >90°.²¹ Lipid length is measured on the longitudinal view. Lipid index is calculated as the product of mean lipid arc and lipid length.²²

Thin-cap fibroatheroma (TCFA)

TCFA is defined as a lipid plaque in which the minimum thickness of the fibrous cap is less than a predetermined threshold and lipid occupies >90° in circumference (Figure 2F). The most frequently adopted cutoff minimal cap thickness is 65 μm, derived from histology studies.²³ However, caution should be taken when a cutoff is directly applied to OCT because tissue shrinkage can occur during histopathologic tissue processing and only fatal cases were assessed. Thus, the true minimal fibrous cap thickness that may predispose to plaque rupture may be >65 μm, although it is presently uncertain. A cutoff thickness of 80 μm is sometimes used.²⁴ Semi-automated assessment of cap thickness²⁵ and 3-dimensional measurement of the surface area of the thin fibrous cap²⁶ have been reported. It should be noted that tangential signal dropout may falsely create the appearance of a TCFA (Table 1) (Figure 1G).

Macrophages

Macrophages are inflammatory cells located in the fibrous tissue covering a lipid plaque and occasionally in fibrous plaque. Macrophages appear in OCT as signal-rich, distinct, or confluent punctate regions that exceed the intensity of background speckle noise²⁷ (Figure 3A). Due to their high backscatter and attenuation characteristics, macrophages cast a dark shadow behind their aggregates. Macrophages frequently cluster together, forming a band of highly reflective tissue, which causes a plaque to appear as if it is a TCFA (Supplementary Figure 1). The strong attenuating properties of macrophage aggregates cause the casting of a laterally sharp shadow and rapid change in appearance from frame to frame. This typical characteristic helps to differentiate a band of macrophages overlying a fibroatheroma from a true TCFA. A detailed classification of macrophages into lines or dots has been reported.²⁸ Quantification of macrophages has been attempted using a parameter termed the normalized standard deviation.^{27, 29} Semi-quantification (grade 0–4) has also been reported.^{30, 31} It is important to note that rigorous histology validation of macrophage on OCT has not been yet reported.

Figure 3. Qualitative findings.

(A) Macrophages are seen as signal-rich, distinct, or confluent punctate regions that exceed the intensity of background speckle noise and cast a dark shadow (arrows). (B) Microvessels within the intima can appear as signal-poor voids that are sharply delineated and can usually be followed in multiple contiguous frames (white circle). (C) Cholesterol crystals appear as thin, linear regions of high intensity (arrow), usually in proximity to a lipid-rich plaque. (D) Thrombus is seen as an irregular mass protruding into the lumen. Red thrombus is seen as a high backscattering protrusion with signal-free shadowing (asterisk). (E) White thrombus is seen as a signal-rich, low backscattering projection, which does not obscure the vessel (asterisk). (F) Layered plaque. The double-headed arrow indicates a layer of different optical densities.

Microvessels

With the gradually increasing size of the plaque, nurturing vessels, known as microvessels (or microchannels) can be observed. Microvessels promote the influx of lipid materials, and the infiltration of inflammatory or red blood cells within coronary plaques.³² Microvessels within the intima can appear in OCT as signal-poor voids that are sharply delineated and can usually be followed in multiple contiguous frames (Figure 3B).³³ A cutoff value of 50–300 μm in diameter and a requirement of the extension in at least 3 consecutive frames are sometimes adopted to define microvessels.³³ As these microvessels communicate with the periadventitial vasculature and eventually with the lumen of the coronary artery, their blood content is flushed during an OCT pullback. To discriminate larger microvessels from small side branches, an inspection of adjacent frames is helpful. The accuracy of microvessel detection decreases in lipid plaques in which the penetration of OCT light is hampered. It has not been fully clarified if these microvessels communicate with the luminal surface or emanate from the vasa vasorum. Detection and quantification of adventitial vasa vasorum have been attempted.^{34, 35} Further studies are required to explore this issue.

Cholesterol crystals

Cholesterol expands in volume when crystallizing from a liquid to a solid state and may cause plaque volume expansion.^{36, 37} Both ex vivo human studies, as well as animal studies, suggest that sharp-tipped cholesterol crystals may perforate the fibrous cap.³⁸ Cholesterol crystals appear in OCT as thin, linear regions of high intensity, usually in proximity of a lipid plaque (Figure 3C). In one histology study, the sensitivity of OCT for detecting cholesterol crystals was only 68%.³⁹ In another histology study, the sensitivity of OCT for detecting cholesterol crystals was 26% even after a modification to the definition of cholesterol crystals.⁴⁰ It should be noted that some cholesterol crystals visualized in histology are not visualized by OCT, particularly when cholesterol crystals are aligned tangential to the OCT light beam.

Thrombus

Destabilization of plaque leading to subsequent thrombotic occlusion of the coronary artery accounts for the majority of acute coronary events. Thrombus by OCT appears as a protruding mass attached to the luminal surface or floating within the lumen. OCT can discriminate between 2 types of thrombus: red (erythrocyte-rich) thrombus (Figure 3D), which is highly backscattering and has a high attenuation (resembles blood), and white (platelet-rich) thrombus (Figure 3E), which is less backscattering, is homogeneous, and has lower attenuation. Since red blood cells reflect light, red thrombus typically shadows or obscures underlying structures, while platelet thrombus does not. Spontaneous recanalized coronary thrombus may appear as signal-rich, high backscattered septa dividing the lumen into multiple small cavities with smooth inner borders, which is known as “honeycomb” appearance.^{41, 42} However, this finding requires further histological validation. The thrombus score is calculated by summing the number of involved quadrants (0, 1, 2, 3, or 4) in each cross-sectional image through the longitudinal length of the thrombus.⁴³ The thrombus area is measured by tracing outlines of the thrombus on cross-sectional images. The thrombus volume is calculated by mean thrombus area multiplied by thrombus length.^{44, 45}

Layered plaque

Pathology studies have shown that atherosclerotic plaques might destabilize without clinical consequences.^{46, 47} Whether an acute coronary syndrome (ACS) develops following disruption of a plaque depends on the severity of stenosis and the balance between systemic/local thrombogenicity and endogenous anti-thrombotic/thrombolysis mechanisms. Thrombus becomes organized, with connective tissue deposition of predominantly proteoglycans and type III collagen.^{48, 49} During the healing process, type III collagen is gradually replaced by type I collagen, which appears as a band of high backscattering signal on OCT.^{48, 49} A histological validation study reported excellent agreement between healed plaques on pathology and layered plaques on OCT.⁵⁰ Layered plaque is defined as plaque with one or more layers of different optical densities and a clear demarcation from underlying components (Figure 3F).⁴⁸ Serial OCT assessments in large and well-characterized study populations are required to better understand this process.

Intraplaque hemorrhage

OCT has never been validated for the diagnosis of intraplaque hemorrhage. Only two case reports have been published so far, in which OCT images showed a clearly bordered, crescent-shaped low-signal region.^{51, 52} Histological validation studies are required for this finding.

Other lesion morphologies

Aneurysms and transplant vasculopathy are described in the Supplementary Material.

Lumen measurements

Lumen area. The area bounded by the luminal border.

Minimum lumen diameter. The shortest diameter through the center of mass of the lumen.

Maximum lumen diameter. The longest diameter through the center of mass of the lumen.

Lumen eccentricity. (Maximum lumen diameter minus minimum lumen diameter) divided by maximum lumen diameter.

Minimum lumen area (MLA). The smallest lumen area along the length of the target lesion.

Percent area stenosis (%AS). (Reference lumen area minus MLA) divided by reference lumen area multiplied by 100. The reference segment used should be specified (proximal, distal, largest, or average).

IEM measurements

For plaques in which the IEM can be identified, area and diameter can be measured for the IEM. In the presence of a large plaque burden or lipid, OCT fails to visualize the entire circumference of the IEM. Consequently, IEM measurements can be made only for those plaques in which the IEM can be identified for ≥180 degrees. IEM measurements are detailed in the Supplementary Material. These can also be applied for the EEM measurements. Attenuation compensation technique, a post-processing methodology to enhance the EEM has been reported.^{53, 54}

OCT for mechanisms of ACS

OCT can accurately assess vessel and lumen geometry and identify the hallmark of a culprit lesion including plaque disruption and thrombus.⁵⁵ A better understanding of mechanisms of ACS has a significant impact on patient management.

Plaque rupture

Plaque rupture occurs in the context of TCFAs and shows a disruption of the fibrous cap⁵⁶ (Figure 4A). When injected with optically transparent contrast media, the disruption may appear as a hollow cavity. Thrombi are often found overlying the ruptured cap. However, thrombus may be absent at the site of an old plaque rupture or with acute rupture treated with anti-thrombotic or thrombolytic therapies. It should be noted that the presence of a large luminal red thrombus obscures the underlying component, rendering the reliable diagnosis of plaque rupture difficult. Assessing contiguous frames rather than a single frame helps to differentiate plaque rupture from other components such as side branches and artifacts.

Figure 4. Culprit lesions in ACS patients.

(A) Plaque rupture is characterized by the presence of fibrous cap discontinuity with a cavity formation (asterisk) within the plaque. (B) Definite erosion is characterized by the presence of attached thrombus (arrows) overlying an intact and visualized plaque. (C) Probable erosion is characterized by luminal surface irregularity (arrows) at the culprit lesion in the absence of thrombus. (D) Eruptive calcified nodule appears in OCT as single or multiple regions of calcium that protrude into the lumen with fibrous cap disruption, frequently forming sharp, protruding edges, and the presence of substantive calcium proximal and/or distal to the lesion (asterisk). (E) Spontaneous coronary artery dissection (SCAD) is seen as a separation of the intima and media from the adventitia (intramural hematoma) (asterisk), with or without communication with the vessel lumen (intimal tear). ACS = acute coronary syndrome

Plaque erosion

The pathology definition of plaque erosion is endothelial denudation with thrombosis without evidence of plaque rupture.⁵⁶ As the detection of an endothelial monolayer (i.e., 1–5 μm) is below the resolution of the current OCT systems, plaque erosion remains a diagnosis of exclusion in vivo. In the criteria developed for the diagnosis of plaque erosion by OCT,⁵⁷ definite erosion was defined as the presence of attached thrombus overlying an intact and visualized plaque (Figure 4B). Probable erosion was defined by luminal surface irregularity at the culprit lesion in the absence of thrombus (Figure 4C) or attenuation of underlying plaque by thrombus without superficial lipid or calcification immediately proximal or distal to the site of thrombus. An algorithm for the diagnosis of plaque erosion is presented in Supplementary Figure 2. As these criteria lack the precision of the pathology definition of erosion, some studies call this entity an intact fibrous cap.⁵⁸ Plaque erosion is predominantly accompanied by white thrombus rather than red thrombus. When a large red thrombus is present, which attenuates the OCT light, the diagnosis of plaque erosion becomes less definite as plaque rupture cannot be excluded. Identification of plaque erosion may facilitate tailoring of patient treatment. Proof-of-concept studies have demonstrated that plaque erosion associated with a residual diameter stenosis <70% may be treated with anti-thrombotic therapy without stenting.^59–62 Larger randomized controlled studies are needed to further explore this concept.

Eruptive calcified nodule

Eruptive calcified nodule is characterized by an underlying heavily calcified plaque with a distinct nodular mass of calcium that protrudes into the lumen and causes dysfunction or loss of the overlying endothelial cells.⁶³ Eruptive calcified nodules appear in OCT as single or multiple regions of calcium that protrude into the lumen with fibrous cap disruption, frequently forming sharp, protruding edges, and the presence of substantive calcium proximal and/or distal to the lesion¹⁸ (Figure 4D). It should be noted that the attenuation of OCT light from protruding calcium conceals deeper structures, rendering the differentiation from red thrombus difficult. Eruptive calcified nodule is typically found in elderly or dialysis patients with heavily calcified vessels at hinge points. Eruptive calcified nodule is associated with a higher rate of target vessel failure after PCI, highlighting the complex nature of this lesion.^{64, 65} In vivo classification of calcific plaque-related ACS has been proposed, which included eruptive calcified nodule, superficial calcific sheet (Supplementary Figure 3A), and calcified protrusion (Supplementary Figure 3B).⁶⁶ In this study, superficial calcific sheet, which was frequently located in the left anterior descending artery, was the most prevalent type and associated with greatest post-PCI myocardial damage. However, this classification lacked histological validations.

Spontaneous coronary artery dissection (SCAD)

SCAD is observed in 2–4% of angiograms performed for ACS^{67, 68} and is more frequent in young to middle-aged women⁶⁹. The hallmark appearance in OCT is an accumulation of blood or flushing media within the medial space, displacing the IEM inward and EEM outward (intramural hematoma), with or without communication with the lumen (intimal tear) (Figure 4E). It is rare to observe overlying thrombus in SCAD. OCT may provide greater diagnostic clarity than IVUS in assessing the distinctive features of intramural hematoma and/or intimal tear. However, the instrumentation and injection of contrast media into a dissected vessel carries the risk of propagation of dissection and vessel closure. Therefore, OCT imaging should be reserved for cases where diagnostic uncertainty persists, particularly angiographic SCAD type 3 and 4.⁷⁰

Myocardial infarction with nonobstructive coronary arteries (MINOCA)

MINOCA accounts for 6–15% of cases of spontaneous MI and frequently affects females.⁷¹ Pathogenesis of MINOCA is varied and may include vascular etiologies such as plaque rupture, erosion, vasospasm, embolization, SCAD, or microvascular disease, of which findings are often angiographically inapparent. Incomplete understanding of mechanisms of ACS leads to variable use of diagnostic testing and medical therapies for secondary prevention with potentially higher costs and lesser efficacy. In the absence of significant obstructive lesions on angiography, the physician should consider non-atherosclerotic etiologies, particularly if patients present with atypical symptoms or unusual demographic/clinical risk profiles. OCT is an invaluable tool to detect atherosclerotic plaque disruption or non-atherosclerotic arterial pathology.^72–74

OCT for detection of vulnerable plaque

Detection of vulnerable plaque has been one of the main focuses of OCT research. There is ongoing controversy on detection of vulnerable plaque and preemptive local treatment.⁷⁵ Clinical outcomes of preemptive stenting are being investigated in the PREVENT study, in which patients with high-risk features identified by IVUS, near-infrared spectroscopy (NIRS), or OCT are randomized to PCI plus medication vs. medication alone (NCT02316886). Detection of vulnerable plaque may help to identify patients who benefit from aggressive prevention treatments such as PCSK9 inhibitors or anti-inflammatory therapies. This topic is detailed in the Supplementary Material.

PCI-related OCT findings

Basic stent measurements

Descriptions are provided in the Supplementary Material.

Stent underexpansion

Stent expansion describes the minimum stent area (MSA) either as an absolute expansion (MSA itself) or relative expansion (MSA divided by the mean of the proximal and distal reference lumen areas). Small MSA is known as a predictor of suboptimal post-PCI fractional flow reserve (FFR)⁷⁶ and adverse outcomes^{77, 78}. Different targets of stent expansion have been adopted in clinical trials that assessed the impact of OCT-guided PCI.^79–82 The EAPCI document recommended a relative expansion >80% as a goal of optimization.⁵ It should be noted that a relative expansion >80% may result in a small MSA in small vessels. There is ongoing controversies on this recommendation.

Malapposition

OCT enables accurate assessment of apposition of stent struts to the arterial wall. If the axial distance between the strut’s surface to the luminal surface is greater than the strut thickness (including polymer, if present), the strut is considered malapposed (Figure 5A). Differences in strut thickness across different stents should be taken into consideration. OCT studies have reported that malapposition was associated with acute (<1 day), subacute (1–30 days), late (30 days–1 year), and very late (>1 year) stent thrombosis.^{83, 84} However, malapposition noted at the time of stent thrombosis may either have been present after the index PCI (persistent malapposition) or developed during follow-up (late-acquired malapposition) due to thrombus resolution or vascular toxicity from the stent. Further studies are required to determine if malapposition is an independent risk factor for stent thrombosis. Malapposition increases the risk of accidental abluminal rewiring and deformation of stent in aorto-ostial segment due to collision with the guiding catheter. The EAPCI document⁵ recommended not correcting acute malapposition of <0.4 mm with longitudinal extension <1 mm as spontaneous neointimal integration is anticipated.

Figure 5. Post-PCI findings.

(A) Strut malapposition. Stent struts (white arrows) are distant from the luminal surface of the artery wall (blue arrows). (B) Tissue prolapse is a tissue extrusion from inside the stent area (arrow). (C) Edge dissection (arrow).

Tissue prolapse

Tissue prolapse between stent struts toward the lumen may represent protrusion of non-calcified plaque or, in the context of ACS, of atherothrombotic material (Figure 5B). OCT enables clearer visualization of tissue prolapse than IVUS. Tissue prolapse is more frequently observed when the stent is placed over a lipid-rich plaque or TCFA. In patients with ACS, tissue prolapse has been reported to be a predictor of target lesion revascularization.^{78, 85} There is insufficient evidence to act with further intervention on tissue prolapse.

Dissection

OCT allows visualization of subtle stent edge dissections typically missed by IVUS (Figure 5C). ILUMIEN III classified edge dissections as major, when dissection extends ≥60° of the circumference of the vessel or ≥3 mm in length.⁸¹ In the CLI-OPCI II study⁷⁷, distal edge dissection with a width ≥200 μm was a predictor of adverse events. In contrast, in a multicenter OCT registry, stent edge dissections were not associated with adverse events.⁷⁸ Minor edge dissections are unlikely to be clinically relevant.^{86, 87} Dissections can lead to intramural hematoma that sometimes progresses and may result in acute vessel closure, particularly when the hematoma is at the distal stent edge. The EAPCI document recognized the presence of residual plaque burden, extensive lateral (>60°), and longitudinal extension (>2 mm), involvement of deeper layers (medial or adventitia) and localization distal to the stent increase the risk for adverse events.⁵

OCT for optimization of PCI

OCT provides information that can be used to optimize stent implantation and minimize stent-related complications. The size and length of balloon and stent are decided based on the pre-PCI OCT measurements, which are more accurate than those obtained by quantitative coronary angiography. Plaque morphology should be assessed before treatment to guide lesion preparation and stent selection. For a heavily calcified lesion, a debulking technique such as atherectomy or intravascular lithotripsy (IVL) may be considered for better lesion preparation to avoid underexpansion. OCT is valuable to confirm calcium fracture, which is associated with better stent expansion⁸⁸. Where pre-dilatation or debulking technique was performed, OCT imaging should be repeated to confirm sufficient lesion preparation and optimize the stent sizing.

The normal, or near normal segment in the case of a diffusely diseased artery, that is the closest to the target lesion can be used as a landing zone of the stent. Since the presence of lipid-rich plaque at the landing zone is associated with a higher risk of periprocedural MI⁸⁹ and stent edge restenosis⁹⁰, landing in a lipid-rich area should be avoided. The EAPCI document proposed the use of the mean distal lumen diameter with up rounding the stent diameter (e.g. 3.76→4.0 mm) or the mean EEM (2 orthogonal measurements) with down rounding the stent diameter (e.g. 3.76→3.5 mm).⁵ Practical strategies of OCT-based optimization of PCI are detailed in the EAPCI document.⁵ Additional details of OCT-guided PCI are summarized in the Supplementary Material.

OCT-guided PCI vs. physiological indices-guided PCI

PCI for stable angina is indicated for hemodynamically significant coronary stenoses.^{91, 92} Pressure-derived invasive indices have been accepted as the gold standard for invasive ischemia assessment.^{91, 93} A randomized trial that included angiographically intermediate stenosis demonstrated a lower occurrence of adverse events in OCT-guided PCI than in FFR-guided PCI.⁹⁴ The impact of the addition of OCT to FFR on clinical outcomes is being investigated in a randomized trial (NCT02989740).⁹⁵ Several studies have reported that lumen measurements and plaque morphology assessed by OCT correlate with FFR.^96–98 A slight superiority of OCT over IVUS for the detection of FFR-positive lesions has been reported.^{99, 100} Optical flow ratio (OFR) is an OCT-based method for functional assessment of coronary stenosis based on computational fluid dynamics. OFR showed an excellent agreement with FFR and was superior to MLA in the detection of physiologically significant stenosis.^{101, 102} Post-PCI OFR showed good diagnostic concordance with post-PCI FFR.¹⁰³ Further studies are warranted on this topic.

OCT for bifurcation lesions

OCT provides useful guidance in the treatment of bifurcation lesions.¹⁰⁴ OCT depicts an ostial lesion in bifurcation without overlap and foreshortening, which frequently occur on angiograms. Three-dimensional reconstruction can be used to understand the geometry and morphology of bifurcation lesions. Following stent implantation, OCT can guide the re-crossing of guide wires through stent struts jailing the side branch ostium (Supplementary Figure 4), and confirm optimization of bifurcation segments, the clinical significance of which is being tested in the OCTOBER study (NCT03171311).¹⁰⁵ A randomized trial demonstrated that three-dimensional-OFDI-guidance was superior to angiography-guidance in terms of acute stent apposition at bifurcation.¹⁰⁶ The LEMON study reported the feasibility and performance of OCT-guided left main stem PCI according to a pre-specified protocol, in which OCT guidance modified operators’ strategy in 26% of the patients.¹⁰⁷ The use of OCT in bifurcation lesions is detailed in the document from European and Japanese bifurcation clubs.¹⁰⁸

Late stent change

Strut coverage

The low susceptibility of OCT to artifacts at the stent strut enables the visualization of tissue overlying struts and the evaluation of the response to stent implantation. Struts are termed “covered” if any tissue can be identified above the struts (Figure 6A) and “uncovered” if no evidence of tissue can be visualized above the struts (Figure 6B). The coverage thickness can be measured as the distance between the adluminal surface of the covered tissue and abluminal reflective edge of metallic struts. Whether the tissue is fibrin, endothelium, thrombus, mature neointima, or other cannot be ascertained at present because a single cell endothelial layer is beyond the limits of resolution of OCT and the nature of tissue cannot be differentiated. Tissue characteristics such as backscattering intensity may provide further discrimination of tissue type.¹⁰⁹ A threshold of tissue above struts ≥40 μm was reported to help to detect healthy strut coverage.¹¹⁰ However, this issue needs further investigation. In addition, whether the tissue is non- or less-thrombogenic is unknown. As with apposition, strut coverage can be reported per strut, per cross-section, or per stent. Percentage of uncovered struts can be measured as the number of struts without distinct overlying tissue, in which the luminal reflection of the strut surface is directly interfacing with the lumen, divided by a total number of analyzable struts × 100.

Figure 6. Vascular response after stenting.

(A) Covered struts. Stent struts are considered covered when tissue can be identified on the luminal side of the strut. (B) Uncovered struts. Struts are considered uncovered when tissue cannot be identified above the strut (struts in white circles). (C) Evagination. An outward bulge (arrows) in the luminal vessel contour between apposed struts. (D) Neoatherosclerosis. A signal-poor region with a diffuse border (asterisk) and an overlying signal-rich band (arrows), which correspond with lipid-laden tissue is seen within the stent.

Evagination

Stent implantation, and the ensuing mechanical trauma and inflammatory reactions, may lead to vascular positive remodeling, causing evagination. Evagination appears in OCT as an outward bulge in the luminal vessel contour between apposed struts (Figure 6C). In a retrospective study, evagination was more frequent in first-generation drug-eluting stents (DES).¹¹¹ Further studies are needed to better understand this finding.

Classification of neointima

OCT provides detailed morphological information about neointimal tissue within stents. Neointimal tissue has been historically subdivided into three different patterns by OCT: 1) homogeneous, 2) heterogeneous, and 3) layered, each of which is detailed in the Supplementary Material. Neointimal patterns might offer guidance to the treatment of in-stent restenosis.¹¹² However, this issue needs further investigation.

Neoatherosclerosis

Clinical trials have shown continuous target lesion revascularization over time, irrespective of the generation of the stent.¹¹³ Atherosclerotic changes within the neointima play an important role in the development of stent-related adverse events. Neoatherosclerosis is defined by the presence of at least one of the components of a mature atherosclerotic plaque such as lipid-laden tissue or calcification within the stent (Figure 6D). Neoatherosclerosis increases with time after stenting and appears in an earlier phase in DES than in bare-metal stent.^114–116 Neoatherosclerosis and subsequent disruption of lipid-laden neointima have been known to be associated with very late stent thrombosis.^{83, 117–119}

Serial OCT examination

Longitudinal studies (baseline versus follow-up) can be used to evaluate the progression or regression of coronary atherosclerosis and the response to stent implantation. This topic is described in the Supplementary Material.

Future directions

Under-utilization of OCT is persisting in most regions of the world due to hurdles such as increased procedure time, lack of reimbursement, and the need for education. Artificial intelligence may aid physicians with tissue characterization and guidance in stent sizing and optimization.^{121, 122}

Hybrid OCT-based imaging such as IVUS-OCT¹²³, OCT-NIRS¹²⁴, OCT-near-infrared fluorescence¹²⁵, and OCT-fluorescence lifetime imaging microscopy¹²⁶, higher resolution OCT (micro-OCT)¹²⁷, and polarization-sensitive OCT¹²⁸ are also being developed and may provide additional insights in the assessment of coronary atherosclerosis.

Conclusions

The past two decades have witnessed the generation of an enormous amount of data from cardiac OCT research. OCT has helped us to better understand the underlying mechanisms of ACS. Detection of plaque erosion in patients with ACS is one of the major advances in the study of in vivo vascular biology in recent years. Layered (healed) coronary plaque research provided insights into the mechanisms of plaque destabilization and healing. Detailed information obtained by OCT during PCI may potentially improve clinical outcomes. Further studies are needed to establish a clear role for OCT in cardiology.