Intravascular MRI for the Evaluation of Lipid-Rich Vulnerable Plaques in Coronary and Peripheral Vessels

Friday, 09/05/08 | 9794 reads
Author(s): 

1Robert L. Wilensky, MD and 2Jacob Schneiderman, MD

Acute coronary syndromes (ACS) and ischemic strokes develop suddenly and often unpredictably in patients with vascular disease. In the majority of patients, these clinical scenarios result from either rupture of a thin-cap fibroatheroma or superficial erosion of an atheroma.1 An intraluminal thrombus then forms on the damaged lesion, possibly embolizing and resulting in decreased blood flow within the artery leading to ischemia and clinical instability. In the coronary circulation, the result of thrombus and embolus formation is myocardial ischemia, infarction, or death, while in the carotid or aortic circulation, an ischemic stroke is the consequence. In the peripheral vasculature, distal ischemia may result. Such rupture-prone lesions, mostly non flow-limiting plaques, have been observed in coronary arteries, the aortic arch, carotid artery bifurcation, proximal segment of the renal, superior mesenteric, and celiac arteries, the aortic bifurcation, and iliac and femoral arteries.2

At the current time, there is no available diagnostic approach to predict the presence of unstable atherosclerotic lesions. Angiography, although excellent for demonstrating the presence of flow-limiting lesions or moderate atherosclerosis, is limited, as it demonstrates only the arterial lumen and not the arterial wall. Also, early atherosclerotic lesions or lesions with expansive remodeling cannot be visualized by angiography. These lesions are more likely to be vulnerable, as several studies have shown that the majority of myocardial infarctions (MI) are caused not by the most angiographic stenotic lesion, but lesions intermediate in severity (e.g., 50–70% in diameter reduction).

Intravascular ultrasonography is the standard in determining the presence and extent of atherosclerotic disease, but has generally proven inadequate in determining lesion composition and predicting future clinical developments. As a result, new diagnostic approaches have been developed to locate such potentially vulnerable plaques prior to plaque instability and a subsequent clinical event. These technologies, which can identify the underlying pathophysiologic substrate, may be more helpful in correlating plaque biology with future clinical developments. Intravascular magnetic resonance imaging (IVMRI) is one such device.

Attraction of MRI for Evaluation of Plaque Composition
Magnetic resonance imaging (MRI) interrogates the differential biophysical and biochemical response of protons following application of a transient electromagnetic radiofrequency (RF) pulse in the setting of a strong static magnetic field. MRI is a powerful tool to determine the presence of lipids within the arterial wall or, when used in combination with local delivery of contrast agents, to determine the presence of specific cell types associated with plaque instability, such as thrombi,3,4 activated macrophages,5 or tissue factor.6 Noninvasive, high-resolution, multicontrast MRI has been used to assess atherosclerotic lesion composition, as well as determine fibrous cap thickness, necrotic core size and fissures within the fibrous cap of atherosclerotic lesions7–10 and has been used clinically to identify the presence of ruptured fibrous caps that are associated with transient ischemic attacks or stroke.11

However, noninvasive MRI has not been shown to effectively and reproducibly evaluate coronary artery lesion composition. The small volume of the typical coronary plaque and the tortuous and irregular course of the vessels make imaging of coronary arteries difficult. In order to prevent motion artifacts, MRI requires both respiratory and cardiac gating. Furthermore, as the distance increases between a MRI coil and the interrogated structure (in this case an artery), there is a corresponding decrease in the signal to noise ratio (SNR), and a consequent reduction in the image resolution. The deeper location of the coronary arteries compared to the surface of the chest (4–10 cm), and the difficulty of optimal receiver coil placement makes attaining a sufficient SNR a major challenge for coronary MRI. To provide a possible solution, a novel intravascular MRI catheter has been developed within which the magnets, RF transmitters, and receivers are miniaturized. The IVMRI holds promise in the in vivo evaluation of lipid-rich, potentially unstable vascular lesions.

Technical Properties of the Intravascular MRI Catheter
The intravascular MRI system is an integrated, self-contained MRI probe positioned on the tip of a vascular catheter attached to a portable control unit. There are no external magnets or coils. Hence, the catheter can be used in the catheterization suite rather than within a MRI magnet. Local static magnetic field gradients are generated at the site of measurement, which is responsive to the diffusion properties of the analyzed vascular tissue.

IVMRI is currently designed to determine the presence of lipids with the arterial wall, since MRI is particularly adept at differentiating between fibrous and lipid-laden tissue. Insofar that the fibrous cap and the normal medial layer possess similar biophysical properties, IVMRI cannot differentiate the two. However, lipid-laden tissue can easily be differentiated from normal, fibrous, and calcific tissue.

After stabilization of the IVMRI probe within an artery, a field of view (FOV) 2-mm in length, 60 degrees along the circumference, radial depth of penetration of 250 µ, is acquired. Several acquisitions can be obtained by rotating the catheter. The image obtained by the intravascular MRI probe is not a picture of the actual morphology of the plaque, but provides a simplified spatial representation of its lipid-rich component. In order to allow “online” interpretation of the lipid composition in each measurement zone, color-coding is used, where yellow represents the presence of lipid, while blue denotes non-lipidemic tissue, whether fibrous tissue or normal vascular smooth muscle cells. Color-coded data from each measurement sector are displayed separately with a numerical lipid-fraction index. Depending on the imaging protocol, each sector can display lipid fraction in various measurement zones, based on the zone depth.

Preclinical Data
Proof of concept studies were performed in a series of studies. An initial in vivo study made use of a patch of subcutaneous fat sewn on either the left anterior descending (LAD) coronary artery through open chest surgery or the femoral artery (Figure 1). The IVMRI catheter was introduced into the LAD via a femoral approach and MRI measurements were performed in 4 quadrants. The same animal was used for MRI examination in the noninstrumented femoral artery. The vessel was dissected and a subcutaneous fat patch was tightly wrapped around half of the vessel’s circumference. The IVMRI catheter was introduced directly into the femoral artery, and measurements were performed at wrapped and unwrapped segments of the artery. The MRI displays obtained in both arterial locations demonstrated the capability of the MRI probe to detect lipid-rich tissue within the range of its radial penetration. The thin (100µ) femoral artery allowed diagnosis of lipid at the superficial corresponding sector, whereas the rather thick (300 µ) coronary artery left the wrapped fat beyond the functional diagnostic capabilities of the IVMRI probe. Subsequent studies performed in atherosclerotic and nonatherosclerotic swine demonstrated the ability of the IVMRI catheter to locate increased intra-arterial lipid concentrations, as well as the safety of the device within coronary and peripheral vessels.

Studies were then performed to determine the efficacy of IVMRI to determine the presence of lipid-rich vulnerable plaques in human coronary arteries. In this ex vivo study, in situ coronary arteries with atherosclerosis from 14 hearts were evaluated.12 Post-mortem coronary angiography identified moderately stenotic atherosclerotic lesions, 30–60% in severity. Circumferential IVMRI acquisition was performed and lesion composition characterized by a blinded observer. The IVMRI diagnosis of lesion characteristics correlated with histologic diagnosis in 16 out of the interrogated 18 lesions (89%), with correct assessment of all 3 thin-cap fibroatheromas. These studies demonstrated that the intravascular MRI catheter could potentially differentiate vulnerable plaques from stable atherosclerotic lesions in human coronary arteries.

Evaluation of aortic lesions was also performed in this study.12 Interrogation of selected intimal sites, extending from macroscopically normal appearing luminal surface to complex protruding or ulcerated lesions was performed. The IVMRI was mechanically applied to the surface within a saline bath at 37°C. One sector measurement was performed, obtaining an online MRI display, showing the lipid content within the wall to a depth of 250 µ. The site undergoing MRI examination was analyzed histologically for correlative validation. The IVMRI scans correctly predicted the histologic diagnosis in 15 of 16 lesions (94%). Figure 2 depicts typical photomicrographs of relatively normal aortic segments devoid of significant lipid component, compared to lesions with high lipid content.

Clinical Data
Two studies have been completed. The First-in-Man study enrolled 29 patients in 4 European centers and was designed to demonstrate safety and feasibility of the self-contained IVMRI system during a diagnostic or therapeutic cardiac catheterization.13 A single nonobstructive plaque with a minimal arterial luminal diameter between 2 and 4 mm in diameter was interrogated. The study demonstrated that the IVMRI catheter was safe, as no catheter-related complications at 30-day follow-up were observed, (absent of a composite of cardiac death, MI (Q wave and non-Q wave). The IVMRI data suggested that the plaque lipid fraction in the study population showed a frequency distribution similar to that found in the ex vivo study of aortic plaques.

The subsequent Phase II study has completed the enrollment of 131 patients and was designed to evaluate the IVMRI catheter in more unstable, higher-risk patients, including patients with ACS, as well as patients undergoing percutaneous coronary interventions. The results are currently being evaluated, but no safety concerns have been raised. Correlations of lipid fraction index and clinical parameters are anticipated in the near future. Based on the safety results, the European CE mark has been obtained, and approval from the FDA is being sought. Figure 3 shows an example of results from a patient with hyperlipidemia and a family history of coronary artery disease (CAD), demonstrating heterogeneity of the atherosclerotic plaque with regard to lipid concentration. Two of the 6 interrogated sections had an increased lipid fraction index, while the other 4 segments did not.

Use of IVMRI in peripheral vessels. The relevance of the IVMRI to vulnerable plaque diagnosis in peripheral arteries has been explored in a few clinical experiments, performed to assess the efficacy and safety of the IVMRI in the lower extremity arteries. The superficial femoral artery (SFA) was elected as the target vessel because of its straight course and accessibility. Moderate atherosclerotic lesions, defined as a 25–50% stenosis, were chosen for investigation. Figure 4 demonstrates a typical SFA fibrotic lesion, devoid of significant lipid content, as diagnosed by the IVMRI. The temporary short occlusion of the artery at this location, (as seen by angiography), to achieve approximation of the probe to the luminal aspect of the lesion during MRI acquisition caused no ischemic consequences. It is anticipated that IVMRI analysis of atherosclerotic plaques in peripheral vessels will allow better understanding of the disease process and more appropriate therapy, whether medical, and/or device implantation or surgery.

Another arterial territory where the IVMRI may serve a vital role in the future is the carotid bifurcation, used in the diagnosis of the carotid vulnerable plaque. A direct linkage has been established between clinical symptoms of brain ischemia and carotid plaque vulnerability.14 MRI has been demonstrated effective in identifying those symptomatic plaques preoperatively,11 suggesting a modality for accurate diagnosis of patients at risk, and allowing for future potential assignment of patients to carotid endarterectomy surgery (CEA) versus carotid artery stenting (CAS). It has been noted that shallow asymptomatic carotid plaques may, in fact, harbor sizable necrotic cores that will eventually progress and become symptomatic. It is, therefore, important to identify those plaques and guide preventive therapy. Regarding carotid stenting, the data obtained from potential MRI assessment should assist the physician in choosing the right stent configuration, adding tissue diagnosis to anatomical considerations. The relevance of intraprocedural plaque assessment has been recognized by interventionalists that utilize IVUS to guide their CAS procedures.15

Use of IVUS. The use of IVMRI is complementary to IVUS. IVUS has become the method of choice to determine plaque volume and extent of arterial remodeling, although it is limited in plaque composition evaluation, given its limited resolution. However, IVUS is able to interrogate the entire length of the coronary artery, thereby providing a roadmap that can be exploited by the IVMRI interrogation. Once a lesion with increased plaque volume, lack of calcification, and positive arterial remodeling is identified, further evaluation with regard to the lesion’s lipid content can be performed in order to determine the potential instability of the lesion (i.e., thin fibrous cap and a large lipid content).

Limitations
The IVMRI catheter interrogates a focal area of the arterial wall. Insofar that the graphic interface projects a representation of the lipid content within the superficial 250 µ, rather than a true anatomic picture, the severity of the atherosclerotic plaque cannot be determined with IVMRI. Hence, localization of the lesion to be interrogated with IVMRI requires the use of either angiography or IVUS. Also, the use of the IVMRI within stented segments of the artery are unknown, and currently, the use of the IVMRI catheter is limited to patients that have not undergone stent placement in the artery to be interrogated.

Summary
Given that vulnerable plaques or thin-fibrous cap atheromas are lipid-laden, they can be visualized, and the lipid content quantified by IVMRI. Animal and human clinical trials have demonstrated the safety and feasibility of catheter-based MRI coronary interrogation. The vulnerable plaque in the peripheral vasculature can be effectively diagnosed by the IVMRI in many arterial territories.

Clinical Review

Add new comment

Back to top

Intravascular MRI for the Evaluation of Lipid-Rich Vulnerable Plaques in Coronary and Peripheral Vessels

Friday, 09/05/08 | 9794 reads
Author(s): 

1Robert L. Wilensky, MD and 2Jacob Schneiderman, MD

Acute coronary syndromes (ACS) and ischemic strokes develop suddenly and often unpredictably in patients with vascular disease. In the majority of patients, these clinical scenarios result from either rupture of a thin-cap fibroatheroma or superficial erosion of an atheroma.1 An intraluminal thrombus then forms on the damaged lesion, possibly embolizing and resulting in decreased blood flow within the artery leading to ischemia and clinical instability. In the coronary circulation, the result of thrombus and embolus formation is myocardial ischemia, infarction, or death, while in the carotid or aortic circulation, an ischemic stroke is the consequence. In the peripheral vasculature, distal ischemia may result. Such rupture-prone lesions, mostly non flow-limiting plaques, have been observed in coronary arteries, the aortic arch, carotid artery bifurcation, proximal segment of the renal, superior mesenteric, and celiac arteries, the aortic bifurcation, and iliac and femoral arteries.2

At the current time, there is no available diagnostic approach to predict the presence of unstable atherosclerotic lesions. Angiography, although excellent for demonstrating the presence of flow-limiting lesions or moderate atherosclerosis, is limited, as it demonstrates only the arterial lumen and not the arterial wall. Also, early atherosclerotic lesions or lesions with expansive remodeling cannot be visualized by angiography. These lesions are more likely to be vulnerable, as several studies have shown that the majority of myocardial infarctions (MI) are caused not by the most angiographic stenotic lesion, but lesions intermediate in severity (e.g., 50–70% in diameter reduction).

Intravascular ultrasonography is the standard in determining the presence and extent of atherosclerotic disease, but has generally proven inadequate in determining lesion composition and predicting future clinical developments. As a result, new diagnostic approaches have been developed to locate such potentially vulnerable plaques prior to plaque instability and a subsequent clinical event. These technologies, which can identify the underlying pathophysiologic substrate, may be more helpful in correlating plaque biology with future clinical developments. Intravascular magnetic resonance imaging (IVMRI) is one such device.

Attraction of MRI for Evaluation of Plaque Composition
Magnetic resonance imaging (MRI) interrogates the differential biophysical and biochemical response of protons following application of a transient electromagnetic radiofrequency (RF) pulse in the setting of a strong static magnetic field. MRI is a powerful tool to determine the presence of lipids within the arterial wall or, when used in combination with local delivery of contrast agents, to determine the presence of specific cell types associated with plaque instability, such as thrombi,3,4 activated macrophages,5 or tissue factor.6 Noninvasive, high-resolution, multicontrast MRI has been used to assess atherosclerotic lesion composition, as well as determine fibrous cap thickness, necrotic core size and fissures within the fibrous cap of atherosclerotic lesions7–10 and has been used clinically to identify the presence of ruptured fibrous caps that are associated with transient ischemic attacks or stroke.11

However, noninvasive MRI has not been shown to effectively and reproducibly evaluate coronary artery lesion composition. The small volume of the typical coronary plaque and the tortuous and irregular course of the vessels make imaging of coronary arteries difficult. In order to prevent motion artifacts, MRI requires both respiratory and cardiac gating. Furthermore, as the distance increases between a MRI coil and the interrogated structure (in this case an artery), there is a corresponding decrease in the signal to noise ratio (SNR), and a consequent reduction in the image resolution. The deeper location of the coronary arteries compared to the surface of the chest (4–10 cm), and the difficulty of optimal receiver coil placement makes attaining a sufficient SNR a major challenge for coronary MRI. To provide a possible solution, a novel intravascular MRI catheter has been developed within which the magnets, RF transmitters, and receivers are miniaturized. The IVMRI holds promise in the in vivo evaluation of lipid-rich, potentially unstable vascular lesions.

Technical Properties of the Intravascular MRI Catheter
The intravascular MRI system is an integrated, self-contained MRI probe positioned on the tip of a vascular catheter attached to a portable control unit. There are no external magnets or coils. Hence, the catheter can be used in the catheterization suite rather than within a MRI magnet. Local static magnetic field gradients are generated at the site of measurement, which is responsive to the diffusion properties of the analyzed vascular tissue.

IVMRI is currently designed to determine the presence of lipids with the arterial wall, since MRI is particularly adept at differentiating between fibrous and lipid-laden tissue. Insofar that the fibrous cap and the normal medial layer possess similar biophysical properties, IVMRI cannot differentiate the two. However, lipid-laden tissue can easily be differentiated from normal, fibrous, and calcific tissue.

After stabilization of the IVMRI probe within an artery, a field of view (FOV) 2-mm in length, 60 degrees along the circumference, radial depth of penetration of 250 µ, is acquired. Several acquisitions can be obtained by rotating the catheter. The image obtained by the intravascular MRI probe is not a picture of the actual morphology of the plaque, but provides a simplified spatial representation of its lipid-rich component. In order to allow “online” interpretation of the lipid composition in each measurement zone, color-coding is used, where yellow represents the presence of lipid, while blue denotes non-lipidemic tissue, whether fibrous tissue or normal vascular smooth muscle cells. Color-coded data from each measurement sector are displayed separately with a numerical lipid-fraction index. Depending on the imaging protocol, each sector can display lipid fraction in various measurement zones, based on the zone depth.

Preclinical Data
Proof of concept studies were performed in a series of studies. An initial in vivo study made use of a patch of subcutaneous fat sewn on either the left anterior descending (LAD) coronary artery through open chest surgery or the femoral artery (Figure 1). The IVMRI catheter was introduced into the LAD via a femoral approach and MRI measurements were performed in 4 quadrants. The same animal was used for MRI examination in the noninstrumented femoral artery. The vessel was dissected and a subcutaneous fat patch was tightly wrapped around half of the vessel’s circumference. The IVMRI catheter was introduced directly into the femoral artery, and measurements were performed at wrapped and unwrapped segments of the artery. The MRI displays obtained in both arterial locations demonstrated the capability of the MRI probe to detect lipid-rich tissue within the range of its radial penetration. The thin (100µ) femoral artery allowed diagnosis of lipid at the superficial corresponding sector, whereas the rather thick (300 µ) coronary artery left the wrapped fat beyond the functional diagnostic capabilities of the IVMRI probe. Subsequent studies performed in atherosclerotic and nonatherosclerotic swine demonstrated the ability of the IVMRI catheter to locate increased intra-arterial lipid concentrations, as well as the safety of the device within coronary and peripheral vessels.

Studies were then performed to determine the efficacy of IVMRI to determine the presence of lipid-rich vulnerable plaques in human coronary arteries. In this ex vivo study, in situ coronary arteries with atherosclerosis from 14 hearts were evaluated.12 Post-mortem coronary angiography identified moderately stenotic atherosclerotic lesions, 30–60% in severity. Circumferential IVMRI acquisition was performed and lesion composition characterized by a blinded observer. The IVMRI diagnosis of lesion characteristics correlated with histologic diagnosis in 16 out of the interrogated 18 lesions (89%), with correct assessment of all 3 thin-cap fibroatheromas. These studies demonstrated that the intravascular MRI catheter could potentially differentiate vulnerable plaques from stable atherosclerotic lesions in human coronary arteries.

Evaluation of aortic lesions was also performed in this study.12 Interrogation of selected intimal sites, extending from macroscopically normal appearing luminal surface to complex protruding or ulcerated lesions was performed. The IVMRI was mechanically applied to the surface within a saline bath at 37°C. One sector measurement was performed, obtaining an online MRI display, showing the lipid content within the wall to a depth of 250 µ. The site undergoing MRI examination was analyzed histologically for correlative validation. The IVMRI scans correctly predicted the histologic diagnosis in 15 of 16 lesions (94%). Figure 2 depicts typical photomicrographs of relatively normal aortic segments devoid of significant lipid component, compared to lesions with high lipid content.

Clinical Data
Two studies have been completed. The First-in-Man study enrolled 29 patients in 4 European centers and was designed to demonstrate safety and feasibility of the self-contained IVMRI system during a diagnostic or therapeutic cardiac catheterization.13 A single nonobstructive plaque with a minimal arterial luminal diameter between 2 and 4 mm in diameter was interrogated. The study demonstrated that the IVMRI catheter was safe, as no catheter-related complications at 30-day follow-up were observed, (absent of a composite of cardiac death, MI (Q wave and non-Q wave). The IVMRI data suggested that the plaque lipid fraction in the study population showed a frequency distribution similar to that found in the ex vivo study of aortic plaques.

The subsequent Phase II study has completed the enrollment of 131 patients and was designed to evaluate the IVMRI catheter in more unstable, higher-risk patients, including patients with ACS, as well as patients undergoing percutaneous coronary interventions. The results are currently being evaluated, but no safety concerns have been raised. Correlations of lipid fraction index and clinical parameters are anticipated in the near future. Based on the safety results, the European CE mark has been obtained, and approval from the FDA is being sought. Figure 3 shows an example of results from a patient with hyperlipidemia and a family history of coronary artery disease (CAD), demonstrating heterogeneity of the atherosclerotic plaque with regard to lipid concentration. Two of the 6 interrogated sections had an increased lipid fraction index, while the other 4 segments did not.

Use of IVMRI in peripheral vessels. The relevance of the IVMRI to vulnerable plaque diagnosis in peripheral arteries has been explored in a few clinical experiments, performed to assess the efficacy and safety of the IVMRI in the lower extremity arteries. The superficial femoral artery (SFA) was elected as the target vessel because of its straight course and accessibility. Moderate atherosclerotic lesions, defined as a 25–50% stenosis, were chosen for investigation. Figure 4 demonstrates a typical SFA fibrotic lesion, devoid of significant lipid content, as diagnosed by the IVMRI. The temporary short occlusion of the artery at this location, (as seen by angiography), to achieve approximation of the probe to the luminal aspect of the lesion during MRI acquisition caused no ischemic consequences. It is anticipated that IVMRI analysis of atherosclerotic plaques in peripheral vessels will allow better understanding of the disease process and more appropriate therapy, whether medical, and/or device implantation or surgery.

Another arterial territory where the IVMRI may serve a vital role in the future is the carotid bifurcation, used in the diagnosis of the carotid vulnerable plaque. A direct linkage has been established between clinical symptoms of brain ischemia and carotid plaque vulnerability.14 MRI has been demonstrated effective in identifying those symptomatic plaques preoperatively,11 suggesting a modality for accurate diagnosis of patients at risk, and allowing for future potential assignment of patients to carotid endarterectomy surgery (CEA) versus carotid artery stenting (CAS). It has been noted that shallow asymptomatic carotid plaques may, in fact, harbor sizable necrotic cores that will eventually progress and become symptomatic. It is, therefore, important to identify those plaques and guide preventive therapy. Regarding carotid stenting, the data obtained from potential MRI assessment should assist the physician in choosing the right stent configuration, adding tissue diagnosis to anatomical considerations. The relevance of intraprocedural plaque assessment has been recognized by interventionalists that utilize IVUS to guide their CAS procedures.15

Use of IVUS. The use of IVMRI is complementary to IVUS. IVUS has become the method of choice to determine plaque volume and extent of arterial remodeling, although it is limited in plaque composition evaluation, given its limited resolution. However, IVUS is able to interrogate the entire length of the coronary artery, thereby providing a roadmap that can be exploited by the IVMRI interrogation. Once a lesion with increased plaque volume, lack of calcification, and positive arterial remodeling is identified, further evaluation with regard to the lesion’s lipid content can be performed in order to determine the potential instability of the lesion (i.e., thin fibrous cap and a large lipid content).

Limitations
The IVMRI catheter interrogates a focal area of the arterial wall. Insofar that the graphic interface projects a representation of the lipid content within the superficial 250 µ, rather than a true anatomic picture, the severity of the atherosclerotic plaque cannot be determined with IVMRI. Hence, localization of the lesion to be interrogated with IVMRI requires the use of either angiography or IVUS. Also, the use of the IVMRI within stented segments of the artery are unknown, and currently, the use of the IVMRI catheter is limited to patients that have not undergone stent placement in the artery to be interrogated.

Summary
Given that vulnerable plaques or thin-fibrous cap atheromas are lipid-laden, they can be visualized, and the lipid content quantified by IVMRI. Animal and human clinical trials have demonstrated the safety and feasibility of catheter-based MRI coronary interrogation. The vulnerable plaque in the peripheral vasculature can be effectively diagnosed by the IVMRI in many arterial territories.

Clinical Review

Add new comment

Back to top