Acute coronary syndrome (ACS) is a life-threatening condition resulting from rupture of vulnerable atherosclerotic plaque (Gotlieb, 2005). The optimal medical therapy currently available with statins prevents only 20–30% of cardiovascular events (Kapur and Musunuru, 2008). Dual anti platelet therapy (DAPT) with aspirin plus clopidogrel, prasugrel or ticagrelor, for up to 12 months after ACS has been shown to improve outcomes after the cardiovascular event (Yusuf et al., 2001; Davis et al., 2015). However, DAPT is often associated with increased risk of bleeding (Clappers et al., 2007; Fanari et al., 2016). Percutaneous coronary intervention (PCI) with balloon angioplasty, often followed by stent implantation is the most commonly performed procedure for the reopening of an occluded artery. Besides restenosis and stent related problems (Kuchulakanti et al., 2006; Gogas et al., 2014), this procedure is mainly used on highly stenosed plaques. Less stenosed plaques, diffused over long segments of artery, are often left untreated. These untreated plaques continue to evolve and may eventually rupture (Skowasch et al., 2005; Moulias and Alexopoulos, 2011). Further interventions, either standalone or in conjunction with existing procedures, are emerging for management of atherosclerosis (Kalanuria et al., 2012; Riccioni and Sblendorio, 2012). Among them, photodynamic therapy (PDT) has been shown to prevent atherosclerotic plaque progression (Usui et al., 1999; Peng et al., 2011). PDT generally involves either topical or, mostly, systemic administration of a photosensitizer which is allowed to circulate for a specific period of time to accumulation, in our case, in atherosclerotic lesions. This time period is often referred to as drug-light interval (DLI). Activation of the photosensitizer with visible light of a specific wavelength generates cytotoxic reactive oxygen species (ROS; Zhu et al., 2015). These ROS oxidize essential tissue components and can eventually induce cell death in tissue. Application of PDT for atherosclerotic plaque treatment has faced some drawbacks due to the non-specific accumulation of the photosensitizer in the skin, leading to cutaneous photosensitivity (Wooten et al., 1988; Dougherty et al., 1990; Wagnieres et al., 1998). Another major hurdle for the clinical use of PDT for cancerous tissue (Braichotte D. et al., 1995; Braichotte D. R. et al., 1995; Zellweger et al., 2000) or particularly for the management of atherosclerosis is long DLI (from 3 to 24 h) after systemic injection with most of the tested PSs (Kereiakes et al., 2003; Tawakol et al., 2006). This puts PDT out of the scope for cases where it is used alongside the treatment of acute myocardial infarction. A short DLI is therefore critically important for its applicability to treat those cases.

Besides DLI, accurate and efficient delivery of light is also important for the activation of photosensitizers. Previous reports indicated that external illumination of the artery using Talaporfin sodium as a photosensitizer resulted into photo toxicity of the irradiated site, however the contralateral walls show limited photo toxicity (Wakamatsu et al., 2005). Intra vascular irradiation circumvents this shortcoming. Intra-arterial photo activation enables circumferential, controlled, and homogeneous illumination of the atherosclerotic artery. In addition, illumination length with our present setup can vary from 1 to 7 cm and even longer light/drug applications are possible. This permits to treat much longer segments of the artery as compared to stents. We hypothesized that local delivery of photosensitizer using an intravascular catheter followed by intravascular illumination may be performed during or just after a PCI procedure and stent implantation. Thus, this “intra-arterial PDT” would be expected to have fewer side effects and due to the short time involved would be compatible with PCI procedures. Combining PCI with PDT may reduce restenosis and may permit the preventive treatment of non-stented vulnerable plaque in order to reduce further coronary events. Furthermore, as the main parameter of PDT is the selective drug uptake in sufficient quantity in given location of the lesion, this can likely be improved by local drug delivery.

Verteporfin, also known as benzoporphyrin derivative monoacid ring A is a second generation photosensitizer. Visudyne® (Novartis) is a liposomal formulation of Verteporfin with a short DLI and is clinically utilized for the treatment of age related macular degeneration and polypoidal choroidal vasculopathy. Previous reports highlight that Verteporfin can accumulate in plaque (Hsiang et al., 1993; Allison et al., 1997). However, studies demonstrating the therapeutic potential of Verteporfin upon photo activation for the treatment of atherosclerotic plaque are still lacking. The present study aims to examine the effect of Visudyne® mediated PDT on atherosclerotic plaque as well as optimization of the local photosensitizer and light delivery within atherosclerotic plaque to reduce the side effects of treatment.

We report here that local arterial perfusion for a short period of time can lead to a high accumulation of Visudyne® in atherosclerotic plaque. Subsequent intra-arterial photo-activation of Visudyne® induces significant apoptosis of plaque macrophages. Moreover, Visudyne®-PDT attenuates the vascular functionality of smooth muscle cells. Therefore local Visudyne®-PDT could be effective for management of atherosclerosis and restenosis.

The role of PDT in the prevention of catastrophic events in atherosclerosis is well established. Previous studies utilizing Photofrin (Hsiang et al., 1995) or 5-aminolevulinic acid (Peng et al., 2011) have demonstrated that PDT can inhibit atherosclerotic plaque progression. An in vivo study conducted by Hayase et al. showed that Motexafin lutetium based PDT promotes stabilization by decreasing plaque macrophage content (Hayase et al., 2001). Photosensitizers are generally administered via systemic injection that may lead to inconvenient cutaneous photosensitivity such as erythema and oedema lasting up to 3 months, in some cases (Braathen, 2014), as well as limited accumulation in the specific region of interest. Selective accumulation of the photosensitizer is critically important for the success of PDT. As a result, the intra-arterial, local drug delivery approach has been chosen to minimize systemic exposure to the drug and to attain high therapeutic concentration of the photosensitizer directly on the plaque within short treatment time compatible with the surgical interventions.

Verteporfin is a potent, second-generation, FDA approved photosensitizer with minimal skin photosensitization. Verteporfin is activated by 690 nm photons, allowing for sufficient light penetration and drug activation. It binds with endogenous low density lipoproteins (LDL) and is significantly more cytotoxic against non-adherent cells as compared to Hematoporphyrin derivative (Richter et al., 1987). Previous reports suggest that Verteporfin-based PDT might prevent the development of intimal hyperplasia (Adili et al., 1998). However, studies evaluating anti-atherosclerotic efficacy were still lacking. In the present work, local delivery of Visudyne® into the plaque of atherosclerotic aorta has been achieved via an ex vivo approach based on a modified system described previously (Zulliger et al., 2002; Prasad et al., 2014). The ex vivo approach reduces the complexity of in vivo system and provides a near physiological environment. This ex vivo approach also enabled us to perform intra vascular light activation of the photosensitizer localized in the plaque. Since Visudyne® accumulation reaches saturation after only 15 min of treatment. We have therefore selected this relatively short time as optimal treatment time for the laser irradiation. DLI is an important parameter that governs the efficacy and selectivity of vascular damage that occurs during the treatment. A short DLI is clinically desirable to facilitate a single treatment in the patient or day-patient clinic that otherwise is limited to treating patients at some interval after angioplasty with possible undesired complications of repetitive invasive injury. Previously, P. Gonschior et al. demonstrated the possibility of reducing DLI using a local delivery approach (Gonschior et al., 1996). Local delivery of Photofrin (5 mg in 2 mL over 30 s) was performed in porcine arteries using a macro-porous balloon catheter. Intra-vascular illumination with an argon-dye laser reduces myointimal hyperplasia after arterial injury (Gonschior et al., 1996). Likewise Hayase M, et al. showed that activation of Motexafin lutetium after 15 min of local infusion reduces plaque burden. Motexafin lutetium was applied by using a drug delivery balloon catheter where it was infused for 15 min directly over the plaque in rabbit atherosclerotic artery (Hayase et al., 2001). Local delivery also enabled to cut the total dose required to about one tenth to that of systemic dose. The findings of the present study showed that local delivery can result in high accumulation of Visudyne® in the plaque within a short application time, thus supporting the feasibility of Visudyne® mediated PDT under clinical conditions.

Efficient and accurate delivery of light to the target tissues is particularly important. The best way to achieve suitable illumination is through endovascular light delivery. Endo-luminal photo activation of Visudyne® was achieved using a cylindrical light diffuser which being at the center of the vessel's lumen, thus providing a homogenous and precisely controlled light dosimetry. It should be noted that the distribution of the light throughout the thickness of the plaque is not fully uniform. This could also account for more apoptosis of plaque cells when the photosensitizer is activated via an inside-out approach.

Upon illumination, photosensitizer molecules absorb light and start reactions eventually generating hydroxyl radical, superoxide anion, hydrogen peroxide (Type I PDT), and singlet oxygen (Type II PDT). Singlet oxygen can further react with nearby molecules to induce the formation of other ROS species that could be detected via chemical sensors like DHE (Satoh et al., 2010; He et al., 2016). When ROS concentration surpasses a certain level it leads to apoptosis via loss of mitochondrial membrane potential, lipid peroxidation, or protein denaturation of cellular membranes and organelles (Chen et al., 2001; Wang and Yi, 2008). Increasing ROS levels is widely utilized in multiple therapeutic approaches like radiotherapy, chemotherapy and PDT for the destruction of leukemia cells (Chou et al., 2004; Wang and Yi, 2008) or macrophages (Zhu et al., 2015). The amount and micro-localization of ROS is dependent on the concentration and supply of oxygen in the tissue and possibly also on the irradiation intensity and fluence rate. This indicates that it is important to assess light dose effects for effective treatment. A light dose dependent increased ROS level leading to 50–60% apoptosis was noted following photo activation of Visudyne®. Our results demonstrate that locally delivered Visudyne® in the absence of light does not alter the apoptosis, indicating no dark toxicity of the photosensitizer. These findings are in agreement with previous reports where Verteporfin alone below 100 ng/ml concentration does not exerts any cytotoxic effects in Retinoblastoma Cell Lines (Stephan et al., 2008). Furthermore, our data show that Visudyne®-PDT resulted in a higher apoptosis of plaque cells as compared to part of the media which was covered by plaque. A possible explanation for such a differential effect could be made on the basis of inherent property of Verteporfin to preferentially localize in plaque (Hsiang et al., 1993; Allison et al., 1997). BA Allison et al. used a mutant fibroblast cell line which has no LDL receptors to demonstrate that LDL receptors are responsible for the accumulation of lipoprotein-associated Verteporfin (Allison et al., 1994). LDL receptors being present on macrophages (Hansson and Hermansson, 2011) may be involved in the uptake of Visudyne® which is a liposomal formulation of Verteporfin. PDT using porphyrins or texaphyrin derivatives like Lutetium texaphyrin, or 5-aminolevulinic acid, have previously been shown to reduce plaque macrophages resulting in stabilization of atherosclerotic plaques and reduction of plaque inflammation (Hayase et al., 2001; Peng et al., 2011). We therefore further investigated the effect of PDT on plaque macrophages. Our results demonstrated that the bulk of apoptotic cells were macrophages. Phagocytic cells like macrophages possess a powerful ROS generating system in response to stimulation with various agents (Forman and Torres, 2001; Swindle et al., 2002; Italiani and Boraschi, 2014). It could thus be assumed that macrophages were more efficient in PDT induced ROS production and were therefore more susceptible to apoptosis. Another possible explanation could in relation with the plaque composition. Atherosclerotic plaques in the present study were characterized by thin fibrous cap with high macrophage population (63% of total cells). Since PDT induces inflammation, which in turn causes the tight junctions between endothelial cells to let leukocytes pass, hence Visudyne® can easily get to the macrophages (with high probability to come in contact with the photosensitizer) in the plaque. Since the ROS produced by activated photosensitizer have very short diffusion distance (<100 nm), PDT induced apoptosis is restricted to cells with adequate presence of photosensitizer like the macrophages in this case.

During atherogenesis, SMC undergo phenotypic switching from the contractile to the synthetic state. SMC migrate from media to intima where they proliferate and contribute to other cellular processes such as inflammation, apoptosis, and matrix alterations (Dzau et al., 2002) that in turn aggravate atherosclerosis progression. Prominent vascular effects occur during and after PDT. Although there are no clear reports demonstrating the phenotypic modulation of SMC immediately after PDT, experimental evidence point to the fact that PDT by producing cytotoxic free radicals results in localized depletion of SMC (Granville et al., 2001) that eventually leads to initial reduction in plaque SMC content (Waksman et al., 2008; Peng et al., 2011). Such depletion of SMC in the medial wall is beneficial in preventing neointimal hyperplasia (Nyamekye et al., 1996; Nagae et al., 2001) or atherosclerosis (Peng et al., 2011). This is followed by repopulation of plaques with a non-proliferating quiescent SMC (Waksman et al., 2008). Moreover, PDT induced injury to SMC does not induce an inflammatory response in the vessel wall (Statius van Eps et al., 1997) but reduces basic fibroblast growth factor content thereby inhibiting invasive SMC mitogenesis to favorably alter the vascular injury response (Statius van Eps et al., 1997). We therefore examined the effect of Visudyne®-PDT on the viability and functionality of medial SMCs. Indeed, our data using Myogram demonstrated that the viability and contractility of SMCs was significantly reduced following Visudyne®-PDT. In the present study, we used Ach to study the influence of PDT on endothelium mediated relaxations. Ach receptors are expressed on endothelial cells of blood vessels and their activation results in increased intracellular calcium and generation of endothelium-derived relaxing factors (EDRF) such as nitric oxide which induce relaxation of the adjacent vascular SMC (Coats et al., 2001). Since phenylephrine induced smooth muscle contractions were substantially reduced following Visudyne-PDT, an expected reduction in Ach induced net relaxation was observed. However, it remains unclear whether such reduction in relaxation response is due to dysfunctional SMC or dysfunctional endothelium. We therefore calculated percentage relaxation by normalizing the net contraction induced by submaximal phenylephrine where a modest reduction in endothelium dependent vasorelaxation was observed. These results corroborated with previous reports where PDT using Photofrin as a photosensitizer impairs the production or release of EDRF by the endothelium (Gilissen et al., 1993). Previous reports also suggest that singlet molecular oxygen, superoxide anion radical, H₂O₂ and hydroxyl radical, generated during PDT are involved in not only supressing the noradrenaline induced contraction but also impairs endothelium-dependent relaxation evoked by acetylcholine (Mizukawa and Okabe, 1997). Our results indicate that Visudyne®-PDT could induce the apoptosis of plaque macrophages and at the same time could reduce the function of the vascular smooth muscle cells.

This study demonstrated that local drug delivery can result in high accumulation of Visudyne® within atherosclerotic plaque shortly after the drug administration. Intra-arterial light activation induces significant apoptosis of plaque macrophages. A short DLI offered by local Visudyne® application can be effectively utilized for treatment of atherosclerotic plaque in conjunction with PCI procedures.

The study was conducted in mouse model of atherosclerosis. Although the lesions of ApoE^−/− mice resemble many of the characteristic of human atheroma, yet the plaque stiffness, muscularity, and thickness vary to a great extent between mouse aorta and human coronary artery. Although previous reports documented the feasibility of intracoronary PDT in porcine model (Waksman et al., 2006) as well as in human subjects (Kereiakes et al., 2003), precautions must be taken to use drug delivery catheters and diffusers suitable to the geometry of the vessel particularly in the case of highly occluded vessels in order to allow smooth insertion and to avoid any mechanical damage to the vessel walls. Further studies in large animal models having plaque more similar to humans are therefore warranted to investigate long term effect in particular the late effects of Visudyne®-PDT in terms of plaque composition and plaque stabilization/regression.

MJ, MZ, AF, JV, HB, GW, SC, and MG contributed to design of the work. MJ performed the experiments. AF planned and provided experimental animals. JV helped with microscopic acquisition of the images. MJ and MG analyzed the data and drafted the manuscript. MZ and GW provided light source and contributed to light dosimetry calculations. MZ, HB, GW, and MG provided critical inputs for revising manuscript. All authors approve the final version of manuscript for publication and agree to be accountable for all aspects of the work.

This work was supported by the Swiss National Science Foundation Grant 150271.