Near-Infrared Spectroscopy Intravascular Ultrasound Imaging: State of the Art

Acute coronary syndromes (ACS) secondary to coronary vessel plaques represent a major cause of cardiovascular morbidity and mortality worldwide. Advancements in imaging technology over the last 3 decades have continuously enabled the study of coronary plaques via invasive imaging methods like intravascular ultrasound (IVUS) and optical coherence tomography (OCT). The introduction of near-infrared spectroscopy (NIRS) as a modality that could detect the lipid (cholesterol) content of atherosclerotic plaques in the early nineties, opened the potential of studying “vulnerable” or rupture-prone, lipid-rich coronary plaques in ACS patients. Most recently, the ability of NIRS-IVUS to identify patients at risk of future adverse events was shown in a prospective multicenter trial, the Lipid-Rich-plaque Study. Intracoronary NIRS-IVUS imaging offers a unique method of coronary lipid-plaque characterization and could become a valuable clinical diagnostic and treatment monitoring tool.


INTRODUCTION
Coronary artery disease (CAD) has continued to be a major cause of morbidity and mortality worldwide, despite recent advances in medical and interventional therapies (1). The pathogenesis of acute coronary events involves the development of an early core with the accumulation of lipid-rich free cholesterol which then progresses to the formation of a fibroatheroma (2). Coronary angiography has been a crucial tool for detecting the gross presence of disease of coronary lesions, it however underestimates the magnitude of atherosclerosis in non-culprit arteries particularly in the early stages of disease process and provides no information regarding the composition of the plaque responsible for the lesion. Intracoronary imaging including intravascular ultrasound (IVUS) and near infrared spectroscopy (NIRS) have been studied to determine the plaque burden (PB) and plaque composition, respectively.
Intravascular Ultrasound provides 2-dimensional cross-sections of arterial vessels which enables the visualization and characterization of not just lumen and stent struts, but the vessel wall dimensions and the presence of plaques within it. In spite of the usefulness of IVUS in the study of plaques and vessel remodeling, it remains limited in the visualization and quantification of certain plaque characteristics mainly due to the inherent properties of sound waves (3). NIRS imaging offers the ability to penetrate blood and tissue to detect lipid core containing coronary plaques. The technology is based on the property of different organic molecules to scatter and absorb light at different intensities and wavelengths (4,5). The integration of NIRS and IVUS systems has provided a hybrid imaging modality which combines the penetration and high resolution of IVUS with the lipid core quantification and characterization of NIRS which has been demonstrated to correlate with the lipid detection (6,7).
We will review the present and potential clinical and research utility of NIRS-IVUS imaging in the study of coronary lipid plaques in the context of CAD.

NEAR INFRARED SPECTROSCOPY TECHNIQUE
Near-infrared diffuse reflectance spectroscopy is a technique that relies on the property of substances to absorb and scatter NIR light (wavelengths from 800 to 2,500 nm) at different intensities as a function of wavelengths (8)(9)(10). NIRS as a technique has been used in various science fields, including chemistry and pharmaceuticals, for the determination of the chemical composition of substances.
Prior to the development of NIRS, various spectroscopy techniques (nuclear magnetic resonance spectroscopy, Raman spectroscopy, and fluorescence spectroscopy) had been studied for possible intravascular applications toward the study of atherosclerotic plaques. The Raman near-infrared spectroscopy (NIRS) which was widely used several disciplines was based on the inelastic scattering of photons following collision with molecules and while it was suggested to have the potential to identify vulnerable plaques, was limited by signal-to-noise problems (11)(12)(13)(14). The NIRS technology was first used in 1993 by Cassis and Lodder in animal experiments in which they sought the characterization of low-density lipoprotein cholesterol accumulation in the aortas of hypercholesterolaemic rabbits. This was followed by the in-vivo use of diffuse reflectance NIRS in imaging the lipid content in human carotid plaques exposed during surgery (15,16)

Validation of NIRS
In the early days of NIRS, 2 pivotal studies were carried out to validate its accuracy for the detection of lipid core plaques (LCPs) in human vessels. The first study by Gardner  have a quantitative target for constructing the algorithm and validating the findings, an LCP of interest was defined as a "fibroatheroma" (FA) with a lipid core >60 • in circumferential extent, >200 µm thickness, and with a fibrous cap of mean thickness <450 µm." The primary analysis which was done by comparing NIRS information presented on block chemogram readings vs. the classified histologic findings showed a "receiver operating characteristic (ROC) area under the curve (AUC) of 0.80 (95 % CI: 0.76-0.85), " confirming the ability of the NIRS system to accurately identify the LCPs (4).
Secondly, the Spectroscopic Assessment of Coronary Lipid (SPECTACL) study which was the first catheter-based technique to use NIRS in humans for percutaneous application was performed to validate the applicability of the autopsy-based LCP detection algorithm in patients. The study, in addition to showing that the NIRS imaging catheter had a similar safety profile to that of IVUS, demonstrated that the spectra obtained from imaging the epicardial vessels of living patients were similar to those from previously validated spectra from autopsy specimens, thereby supporting the use of NIRS for detection of LCPs in human patients (9).
Earlier, several ex-vivo studies had examined the ability of NIRS to identify histological features of lipid-rich atherosclerotic plaques in human blood vessels obtained at autopsy. These studies reported >90% sensitivity and specificity for the identification of characteristic features suggesting lipid-rich plaques including the rupture-prone thin-cap fibroatheromas (TCFAs) seen in ACS patients. More recent studies have corroborated these findings as well as pointing to the additive value of NIRS to IVUS-derived PB in detecting vulnerable plaques (17)(18)(19)(20)(21).
Intra-and Inter-catheter reproducibility of the NIRS catheter has also been validated in a number of independent studies (22,23).

Principles of Near Infrared Spectroscopy-Intravascular Ultrasound (NIRS-IVUS) Catheter
A little over a decade ago, a single modality NIRS system was originally developed for the invasive detection of lipid core plaques (LipiScan TM , Infraredx Inc., Bedford, MA, USA). In later years, a dual modality system which combined IVUS with NIRS was developed to provide in a single catheter information on both vessel structure and plaque composition in a single acquisition. The NIRS-IVUS systems have continued to evolve and now exist in the form of a dual frequency, dual modality system. (TVC Imaging System TM and Makoto Intravascular Imaging System TM , Infraredx Inc.) Table 1.
The NIRS-IVUS system comprises a scanning NIR laser, a pullback and rotation unit, and a traditional IVUS-sized catheter. The 3.2F rapid exchange catheter has an entry profile of 2.4F and a shaft profile of 3.6F and is compatible with 6F guiding catheters. The NIRS-IVUS can be inserted over a 0.014-inch guide wire while its passage through the lesion is facilitated by the hydrophilic coating present on the flexible distal 50 cm end. The IVUS images are acquired during an automated pullback with simultaneous co-registered NIRS measurements. In the latest generation, the Makoto system, the catheter's imaging core pulls back at speeds of 0.5, 1.0, or 2.0 mm/s and rotates at 1800 rpm with a maximum imaging length of 15 cm, acquiring up to ∼130,000 NIRS per 100 mm (4, 9, 24) (Figure 1). NIRS lipid core data are automatically displayed on a "chemogram" which displays the probability of the presence of a lipid rich plaque with the millimeters of pull-back on the x-axis and the circumferential position on the y-axis. Areas containing lipid core are displayed as yellow and those without any significant lipid core as red. A quantitative image metric is automatically reported as a numerical lipid-core burden index (LCBI), which represents the fraction of the chemogram yellow pixels whose probability of lipid exceeds 0.6, per 1,000. The LCBI provides a quantitative metric of the lipid core plaque present in a scanned vessel and can be computed over the entire length of a vessel scan, segments of scans, or defined width windows within segments (such as the 4 mm sliding window with the maximum LCBI, maxLCBI4mm) (9) (Figure 2).

Lipid Rich Plaque Characterization
Prior autopsy studies have shown that lesions with a thin fibrous cap(<65 mn of cap thickness) overlying a large necrotic core are most frequently prone to rupture (4,25,26). These "vulnerable" plaques have therefore become a target for identification by novel intracoronary imaging modalities. Several studies in ACS patients have since shown the presence of LCPs in the culprit arterial segments to be major precursors of the disease and the possible association of the lipid core burden with adverse cardiovascular outcomes in CAD patients overtime (27-34) ( Table 2).
The PROSPECT (Providing Regional Observations to Study Predictors of Events in the Coronary Tree) Trial, a landmark study which included 697 acute coronary syndrome (ACS) patients was a natural history study that sought to provide in vivo evidence of the hypothesis that the histopathological characteristic of plaques and not the degree of stenosis on angiography was responsible for the development of ACS. The study investigated the non-culprit coronary lesions in patients with ACS who underwent three-vessel gray scale and VH-IVUS imaging after successful PCI of culprit lesion demonstrated that "independent predictors of non-culprit-related events were the presence of VH-IVUS thin-cap fibroatheroma (hazard ratio 3.35; 95% confidence intervals 1.77-6.36), a PB ≥70% (hazard ratio 5.03; 95% confidence intervals 2.51-10.11), and a minimum lumen area ≤4.0 mm 2 (hazard ratio 3.21; 95% confidence intervals 1.61-6.42)" (44).
Following the results from the PROSPECT Study, The COLOR registry prospectively observed a positive association between plaque morphology evaluated by NIRS and the degree of coronary artery stenosis. It demonstrated that increasing degree of stenosis seen by angiography was associated with more vulnerable plaque morphology as assessed by  In the most recent prospective multicenter natural history intracoronary imaging study-The LRP (Lipid Rich Plaque) identified patients and coronary segments at risk of future major  adverse coronary events using NIRS-IVUS system. Patients with known or suspected CAD undergoing cardiac catherization with possible PCI were examined via NIRS-IVUS imaging in nonculprit arteries when possible. The study showed that 9% of the patients had subsequent non-culprit-major adverse cardiac events (NC-MACE) and out of these patients, higher event rate was associated with analyzable maxLCBI 4mm > 400 (HR 3.39; 95% confidence interval 1.85-6.20), pointing to the diagnostic value of the NIRS-IVUS (42,43). Following these results, the FDA granted a label claim for NIRS detection and identification of patients at increased risk of major adverse cardiac events (MACE) Table 2.

Clinical Applications of Intracoronary Near-Infrared Spectroscopy
In general, the early evidence from NIRS-IVUS use points to multiple potential clinical applications of this imaging system. The identification and localization of vulnerable atherosclerotic plaques offers a clinical tool that could help assess precise lesion lengths including segments with a high lipid core burden toward optimal stenting (48). Secondly, NIRS-IVUS offers information on the necrotic core of atherosclerotic plaques and can predict the embolization of highly thrombogenic lipid depositions. The release of these elements into the blood stream which results in distal embolization is a known culprit in peri-procedural myonecrosis (35,38,(49)(50)(51). Similarly, the utility of NIRS-IVUS in optimizing carotid artery stenting toward preventing periprocedural stroke is a subject of ongoing research (52,53). Furthermore, NIRS-IVUS has been employed in the study of plaque morphology and composition in the peripheral arteriessuperficial femoral arteries in the setting of severe stenoses and symptomatic peripheral artery disease (PAD) (54,55).
Several other studies (37,43,56,57) which show the association of either max4mmLCBI or high LCBI (in the case of Danek et al. and Oemrawsingh et al.) in non-target/nonculprit vessel segment with increased incidence of MACCE and the risk of future coronary events subsequent to the presence of vulnerable plaques in non-culprit segments has given credence to retrospective autopsy studies regarding the assessment of plaque vulnerability. The ability of NIRS-IVUS to assess the possibility of future coronary events based on the LCBI values in nonculprit segments offers a unique diagnostic utility and a riskstratification tool. The predictive role of NIRS-IVUS shown in these studies is being further studied in ongoing studies and raises the question of the need for treating non-culprit vulnerable plaques in vulnerable patients.
With the evolution of pharmacological therapies in addition to the performance of the present lipid-lowering therapies, NIRS-IVUS offers a tool for monitoring the effects of medications such as statins and PCSK9-inhibitors on the coronary vasculature lipid burden. Already, studies have investigated the effect of lipid-lowering medications on reducing necrotic-core containing plaques. The YELLOW (Reduction in Yellow Plaque by Aggressive Lipid-Lowering therapy) trial was a randomized clinical trial (rosuvastatin 40 mg daily vs. the standard-ofcare lipid-lowering therapy) which recruited patients with multi-vessel CAD (including at least 1 severely obstructive [FFR≤0.8] non-culprit) undergoing revascularization by PCI. The non-target lesions in both groups were evaluated at baseline and following 7 weeks of therapy with fractional flow reserve (FFR), and NIRS-IVUS, comparison between baseline and follow up results for NTLs demonstrated that reduction in LCBI4mm max was significantly higher in intensive statin group as compared to standard group (36) In a follow up study -the YELLOW II, the effects of high-dose statin therapy on changes in plaque morphology (evaluated by OCT) and plaque lipid content by NIRS were assessed in obstructive non-culprit lesions (NCLs) in addition to a comprehensive assessment of cholesterol efflux capacity (CEC) and peripheral blood mononuclear cell (PBMC) transcriptomics. The mean baseline NIRS-derived maxLCBI4mm was over 400 (416.6 ± 172.9) and the study reported no significant change in plaque lipid content by NIRS albeit a significant change in fibrous cap thickness after about 12 weeks of statin therapy. (41). Almost similarly, the IBIS-3 (Integrated Biomarker and Imaging Study 3) trial demonstrated that "high dose rosuvastatin therapy had a neutral effect on LCP (assessed by NIRS LCBI) in non-stenotic NCLs in the coronary vessels after 6-and 12-month follow-up intervals despite a relatively lower mean baseline maxLCBI4mm (201.9 ± 1623.8), in the cohort of 103 patients" (39,40). While differences in LCBI cut-offs of the various studies could be a possible reason for the discrepancy in findings, the mean baseline maxLCBI4mm in the YELLOW I and YELLOW II trials were similar, even though the baseline LCBI was significantly higher in patients randomly allocated to the intensive vs. standard therapy group in the YELLOW I and there was no comparative standard therapy group in the YELLOW II. There is also the question of short term LRP/LCBI regression occurring mostly in plaques with a large PB and large LRPs reported in a YELLOW trial sub-study evaluating the relationship between LRPs, plaque morphology and lesion progression or regression. These findings point to the need for further exploration of the effect of statins as well as novel lipid-lowering therapies on NIRS-derived LCPs and LCBI in conjunction with plaque morphology analysis via other intracoronary modalities (58).

FUTURE DIRECTIONS
The findings of Waksman et al. in the recent LRP study (43) strongly suggest the viability of NIRS-IVUS as a diagnostic and risk-stratifying modality and opens the door to further trials to evaluate therapeutic strategies against high lipid core burden and/or vulnerable plaques/patients in addition to studies which could explore the value of LCBI as a surrogate marker for the checking the effectiveness of new therapeutic interventions.
Notably, the effects of newer LDL-lowering medications are currently being studied in ongoing trials. The PACMAN AMI (Vascular Effects of Alirocumab in Acute MI-Patients) trial (ClinicalTrials.gov Identifier: NCT03067844) is examining the effects of PCSK9-inhibiting monoclonal antibody, alirocumab on coronary atherosclerosis in acute MI patients via NIRS, IVUS and OCT. Also, the FITTER (Functional Improvement of Coronary Artery Narrowing by Cholesterol Reduction With a PCSK9 Antibody) Trial (ClinicalTrials.gov Identifier: NCT04141579), will be enrolling patients to investigate the impact of evolocumab plus statins on FFR of non-infarct related arteries (non-IRAs) in multivessel disease patients and correlating baseline NIRSderived lipid core burden with changes in FFR in the non-IRAs.
In a study design which integrates a natural history study and a randomized trial, the PROSPECT II and PROSPECT ABSORB, ClinicalTrials.gov Identifier: NCT02171065), patients with high risk plaque were randomized to receive Absorb BVS alongside the standard of care, optimal medical therapy (OMT) or vs. OMT alone and changes in the plaque will be evaluated by IVUS and NIRS at follow up. Lastly, The Preventive Coronary Intervention on Stenosis With Functionally Insignificant Vulnerable Plaque (PREVENT) study, is presently recruiting patients to determine the effect of preventive PCI on functionally insignificant coronary lesions with vulnerable plaque characteristics using NIRS, IVUS, OCT and virtual histology-IVUS modalities (ClinicalTrials.gov Identifier: NCT02316886).
While the value of NIRS-IVUS and other intracoronary imaging techniques are being shown in various imaging trials, there are still limitations in assessing plaque characteristics which could explain some of the conflicting study results. Multimodality imaging may be critical in overcoming these limitations, thus emphasizing the need for co-registration of NIRS-IVUS with other imaging including angiography and as well as OCT or in the form of hybrid imaging catheters (59). Finally, the ability of NIRS-IVUS to predict the location of histologically-confirmed TCFA and the possibility of NIRS-guided cap thickness detection, and collagen content analysis which are already being explored in ex-vivo studies could represent a huge breakthrough in vulnerable plaque studies in the near future.

LIMITATIONS
One of the main limitations of NIRS-IVUS is the invasiveness of the technique which precludes its use in primary prevention in symptomatic patients with subclinical disease. Presently, NIRS imaging does not have the capability to detect the depth of the lipid core or vulnerable plaque features such as the thinness of the fibrous cap.
CONCLUSION NIRS in conjunction with IVUS is certainly a diagnostically useful tool for the detection of vulnerable plaques and can help identify patients at risk of future coronary events. While there is a wide spectrum of clinical applications for this technology, its viability for demonstrating the effects of present and forthcoming lipid-lowering therapies could significantly influence clinical perspectives and practice in the years to come.