Developing a Nanoparticle Delivered High-efficacy Treatment for Infantile Hemangiomas Using a Mouse Hemangioendothelioma Model

Abstract

Background

The current treatments for infantile hemangiomas (IHs) have unpredictable outcomes. The aim of this study is to develop a nanoporphyrin (NP) delivered, high-efficacy treatment for IHs using a mouse hemangioendothelioma (HT) model.

Methods

We injected mouse hemangioendothelioma cells intradermally to axillary regions of five-week-old, female, nude mice (n=19) to induce HT growth. We documented NP accumulation in HTs using positron emission tomography (PET). For the treatment study, we randomized HT-bearing nude mice (n=9) into three groups (n=3, each). Animals in group I received only saline injections. Animals in group II received only laser treatment after saline injection, and animals in group III received laser treatment after NP injection via tail vein. We followed up the treatment response with digital caliper measurements.

Results

HTs started to grow approximately one week after inoculation, and resembled IHs histologically. NP uptake in HTs was 19.7±2.2, 16.7±2.02, 8.4±0.3, and 4.9±0.6 %ID/g at 3, 6, 24, 48 hours postinjection. NP uptake in HTs was significantly higher than blood at 24 and 48 hours postinjection (p<0.05). Results of ex vivo biodistribution study were consistent with PET imaging. HTs in group III started to regress one day after the treatment and disappeared totally by day 21. The difference between tumor volumes in group III and other groups was significant on days 17, and 21 (p<0.05).

Conclusion

NP accumulated in HTs at high concentrations enabling a high-efficacy photodynamic therapy. Based on the similarities between HTs and IHs, this treatment potentially can be a high-efficacy treatment for IHs.

Introduction

Infantile hemangiomas (IHs) are the most common tumors of childhood with a reported incidence between 4-10 %. ^1-4 Tumors typically progress through two phases of growth: an initial “proliferative phase” during the first several months of life in which the tumor grows rapidly, followed by an “involution phase”, with slow, spontaneous regression. ^{3, 5} Although most lesions proliferate and subsequently regress with minimal consequences, central facial lesions usually result in disfigurement. The residual fibro-fatty masses force patients to live with unsightly scars for the remainder of their lives, causing lifelong psychological and physiological problems. Preventing rapid growth of IHs early on can eliminate potential complications and dramatically improve quality of life for patients. Unfortunately, current treatment methods for IHs face multiple challenges including poor overall efficacy and serious side effects. ^6-8 The goal of this study is to develop a novel, locally active, non-toxic treatment method for IHs using photodynamic therapy (PDT) delivered via nanoparticles.

PDT uses a photosensitizing drug in combination with laser light to kill target cells. First a photosensitizer is administered then the target region is irradiated with laser light at a wavelength that matches the absorption spectrum of the photosensitizer. The photosensitizer absorbs a photon of laser light and then transfers most of the absorbed energy to a molecule of oxygen. This converts the oxygen molecule into a relatively strong subtype of reactive oxygen species (ROS) known as singlet oxygen. ^9-11 As a consequence, in the tissues that have accumulated the sensitizer, light-induced ROS exerts a cytotoxic effect by causing lethal oxidative damage to biologically important structures. PDT can eliminate tumors with minimal risk of fibrosis or scarring and it is becoming an increasingly accepted therapeutic modality, either alone or in combination with other treatments for various malignant and nonmalignant conditions. ^{11, 12}

We used nanoporphyrin nanoparticle (NP) as a photosensitizer for PDT in this study. NP is a porphyrin based nanocompound that has been recently developed and used as a photosensitizer for the PDT treatment of ovarian and breast cancer xenografts in nude mice. ¹³ Similar to other nanoparticles, NP selectively accumulates in tumors due to its small size (25 nm). This phenomenon is known as the “enhanced permeability and retention” (EPR) effect. The general explanation for this phenomenon is that, in order for tumor cells to grow quickly, they must stimulate the production of blood vessels. These newly formed tumor vessels are usually abnormal in that they have poorly aligned defective endothelial cells with wide fenestrations, and lack a smooth muscle layer, or innervation. These vessels are also named as “leaky vessels”. ^14-16 IHs are also fast growing vascular tumors; therefore, we hypothesize that IH vasculature is leaky and intravenously administered NP will preferentially accumulate at the IH lesion, which can then be destroyed through PDT without affecting surrounding normal tissues.

Materials and Methods

All the animal experiments were approved by Institutional Animal Care and Use Committee (protocol # 17821).

Animal model for IH

We used mouse hemangioendothelioma (EOMA) cells (ATCC^®, Manassas, VA) to establish an easily reproducible animal model for IH based on a previous paper. ¹⁷ We intradermally (id.) injected 1.5 × 10⁶ EOMA cells to bilateral dorsal axillary regions of 5-week old, female nude mice (n=7) and followed up the tumor growth by digital caliper measurements every other day. We calculated the tumor volume based on these measurements using the formula “Volume = (Width²×Length)/2”. We euthanized the mice when the largest diameter of the tumor reached to 1 cm. We also performed Hematoxylin and Eosin (HE) and immunofluorescence (IF) staining to demonstrate the highly vascular structure of hemangioendothelioma tumors (HTs). Briefly, HTs were fixed with 4% paraformaldehyde, embedded in paraffin blocks and cut to 5 μm sections. The sections were incubated with a primary antibody for CD31 (Novus Biologicals, San Diego, CA) overnight at 4°C. Anti-rat Texas red (Invitrogen, Eugene, OR) was used as a secondary antibody. The nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI) (Vector Laboratories, Burlingame, CA) and the images were captured under a fluorescence microscope.

Positron emission tomography (PET) imaging of HTs using NP

In order to obtain maximum efficacy from PDT, high levels of photosensitizer accumulation in the target tissue is crucial. To demonstrate NP accumulation in HTs, we performed PET imaging on HT bearing nude mice (n=3). We induced HT formation via id. injection of EOMA cells, and NPs were synthesized as described previously. ¹³ When HTs reached to a volume of 500 mm³ we injected the mice with copper 64 (⁶⁴Cu) labeled NP (⁶⁴Cu-NP) (800μCi / 2mg) via tail vein and obtained images using a PET/CT scanner (Inveon DPET, Siemens, Knoxville, TN) at 3, 6, 24 and 48 hours postinjection (pi.). A CT image was also acquired at 24 hours pi. for registration to the corresponding PET image. We performed an ex vivo biodistribution study at 48 hours pi. after the final PET scan. We euthanized the animals (n=3) and excised the HTs and major organs to measure the radioactivity level of the organs using a gamma counter (PerkinElmer, Waltham,MA). We quantified PET images by region of interest (ROI) analysis and expressed the results as percent injected dose per gram of tissue (%ID/g).

Treatment of HTs with PDT

We randomized female nude mice bearing HTs (n=9) into three groups (n=3, each) (Table 1). Animals in control group (group I) received only PBS injections via tail vein. Animals in group II received only near-infrared laser (NIRL) treatment after PBS injection via tail vein, and animals in group III received NIRL treatment after NP injection via tail vein (NP delivered PDT) on day 11 after tumor inoculation. For PDT, we illuminated HTs with NIRL for 2 minutes at 0.8 Watt power following injection of 200 μl NP solution at a concentration of 10mg/ml. We followed up the treatment response by digital caliper measurements.

Table 1. Study groups.

Group I	No treatment.
Group II	NIRL treatment
Group III	NP injection followed by NIRL treatment (PDT)

Results

Animal model for IH

We were able to induce a reliable HT growth by id. injection of mouse hemangioendothelioma cells (Figure 1A). HT growth commenced one week subsequent to injection, and the tumors continued growth until the day 21 pi., when animals were humanely euthanized due to increased tumor size. Tumors showed accelerated growth rates after day 19 that may correlate with the “proliferative phase” noted in human IH (Figure 1B). HTs were located superficially and untreated tumors led to ulcer formation like human IH (supplemental Figure 1). We did not observe tumor invasion to the underlying tissue layers in any of the animals. CD31 IF and HE staining demonstrated the highly vascular characteristic structure of the tumor resembling human IH (Figure 1C).

Figure 1.

(A) The HTs in the bilateral axillary region of nude mice 14 days pi. (B) The growth curve of the tumors in the axillary region. Errors bars represent SD. (C) CD31 IF and HE staining showing the vessels within the HT. Microbars 50 μm and 200 μm respectively.

PET imaging of HT using NP

A significant fraction of the injected dose of ⁶⁴Cu-NP accumulated in HTs. NP accumulation peaked as early as 3 hours pi. (19.8±2.2 %ID/g) with levels remaining high until 6 hours pi. (16.7±2.0 %ID/g). The uptake in HTs decreased to 8.4±0.3 at 24 hours pi. and 4.9±0.6 at 48 hours pi (Figures 2A&B). NP uptake in liver was 14.8±0.7, 13.0±1.0, 5.5±0.5, 1.4±0.1 %ID/g at 3, 6, 24 and 48 hours pi. respectively (Figures 2A&B). The accumulation in liver was expected because NP in blood circulation was carried to liver for metabolism. However, we believe in that the uptake in heart (23.6±1.2, 17.4±0.9, 2.1±0.2, 0.1±0.03 %ID/g at 3, 6, 24, and 48 hours respectively) was due to NP circulating in blood rather than myocardial uptake because a similar level of uptake was observed in aorta in PET images at 3 hours pi. The uptake in right HTs was significantly higher than blood at 24 and 48 hours pi (p<0.05, p<0.01 respectively), and liver at 48 hours pi. (p<0.05). The uptake in left HT was significantly higher than blood 24 and 48 hours pi. (p<0.05). NP uptake in other major organs was negligible. The uptake in PET imaging overlapped well with the location of HTs on CT images (Figure 2C). Ex vivo biodistribution study revealed an average uptake of 0.3±0.03, 1.9±2.28, 0.8±0.42, 0.1±0.01, and 1.00 %ID/g in left HT, right HT, liver, heart, and spleen respectively 48 hours pi. (Figure 3). The highest average uptake was in right HT, however the difference between the uptake in right HT and other organs was not significant most likely due to the wide range of values and resultant high standard deviation (SD). On the other hand there was a significant difference between the uptake in spleen and other organs, including left HT (p<0.01 for brain, heart, lung and muscle, and <0.05 for kidney, left HT, and small intestine). There was also a significant difference between the uptake in liver and brain, heart, lung, and muscle (p<0.05).

Figure 2.

(A) Representative PET images at different time points after the injection of ⁶⁴Cu-NP. The uptake was highest at 3 and 6 hours pi. White arrows point to the HTs. (B) The uptake of ⁶⁴Cu-NP in HTs, blood and liver (%ID/g). Errors bars represent SD. (C) PET signal overlapped perfectly with the HT in PET- CT images obtained 24 hours pi. * p<0.05, ^ψp<0.01

Figure 3.

Graph shows the result of ex vivo biodistribution study. Separate bars represent individual animals. Errors bars represent SD. * p<0.05, ^ψ p<0.01

Treatment of HTs with PDT

HTs in group III started to regress as early as the next day after the treatment and totally disappeared on day 21 after inoculation (Figure 4A&B). HTs in group II showed variable response to treatment. Smaller tumors regressed the next day after the treatment but started to regrow afterwards while the bigger tumors showed limited response and kept growing (Figure 4A&B). On the other hand the HTs in group I continued growth and the animals had to be euthanized on day 21 due to impairment of the vital functions by the growing HTs. The tumor volumes in group III was significantly less than groups I and II on day 21 (p<0.05) and significantly less than group I on day 17 (p<0.05).

Figure 4.

(A) HTs responded to treatment as early as day 1 after the treatment in group III, whereas the HTs in group I and II continued to grow. (B) The graph shows the volume changes of HTs in study groups after inoculation. Errors bars represent SD. * p<0.05, ↓ treatment

Discussion

The currently accepted approach to the treatment of IHs is to manage uncomplicated cases with expectant treatment. Treatment should be attempted if complications occur or if the IH grows substantially to block vital organs ^{18, 19}. Up to 38% of IH referred to tertiary care specialists require systemic treatment due to complications. ⁷ Systemic corticosteroids have been the historical mainstay of pharmacological treatment of persistent IHs. However, response to treatment is variable and adverse effects are common. ^{1, 7, 8} Conventional laser treatment (e.g., pulsed dye laser [PDL[) is a well-established treatment for other childhood vascular malformations. However, the use of PDL to treat proliferating IH remains controversial based on likelihood of adverse outcomes (e.g. ulceration, scarring), and due to limited depth of penetration (<1 mm). ^{3, 7} Surgical excision remains the treatment of choice for smaller IH, especially when there is threatened loss of function, life-threatening complications or when drug therapy fails or is not tolerated. ⁷

Propranolol - a β blocker typically used for cardiovascular disorders - has recently emerged as a useful therapeutic for management of IH as a first and/or second line treatment. ^{20, 21, 22} A recent multicenter trial showed a success rate of 88% with total cure in 60% of cases at doses of 3 mg per kilogram per day for 6 months. ²⁰ However, only 55 patients were assigned to the placebo group (vs. 405 patients enrolled in propranolol treatment groups) and out of 55, only 28 patients completed follow up (vs. 315 in treatment groups). Therefore, there is a possibility that this striking difference between the sample sizes in this study might have created a bias towards propranolol treated groups. Additionally, a recent paper reported a male infant who presented with IH even though his mother was receiving propranolol throughout pregnancy for the treatment of mild anxiety. ²³ The boy presented with several episodes of symptomatic hypoglycemia (0.6 g/L) at 3 days of age. Hypoglycemic episodes stopped after discontinuation of breastfeeding (at day 6 of age) and propranolol was considered to be responsible for the child's hypoglycemia in this case. It might have been hypothesized that treatment with propranolol, which crosses the placenta, in a pregnant woman could prevent the appearance of IH in her child but this appears not to be the case in this patient. Furthermore, propranolol can cause a number of adverse effects, the most dangerous of which is symptomatic and potentially fatal hypoglycemia. ^{24, 25} Current recommendations therefore advise caution in systemic use of propranolol for this novel indication in an otherwise normotensive pediatric population until clinical indications and appropriate treatment protocols are more clearly defined by multicenter trials. ⁵

We are addressing the above mentioned barriers in IH treatment in this study by using NP as a delivery vehicle for PDT and potentially for pharmaceutical therapy. NP is made of porphyrin that is naturally found in the human body (e.g. hemoglobin); therefore, it is significantly less toxic to normal cells, and more biocompatible than the other inorganic nanoparticles such as Gold and Silica. ¹³ As shown by PET imaging NP selectively accumulated in HTs at high levels via EPR effect. This finding confirms that HT vasculature is leaky as we hypothesized. The high levels of NP in HTs enabled a localized, high-efficacy treatment. NP was activated only in HTs with laser light illumination, while remaining inactive in other organs limiting the systemic side effects of the treatment. ²⁶ Additionally, NIRL that we used in this study has an improved depth of penetration than conventional laser light (5-10 mm vs. 1 mm). ^{27, 28} Increased depth of penetration combined with the superficial location of HTs further increased the effectiveness of the treatment.

A limited number of clinical studies exist evaluating the use of PDT as potential treatment for IHs and other vascular malformations. PDT was used for the treatment of port wine stains with favorable results, and secondary scar formation was noted only in 3 out of 238 patients demonstrating the safety of the procedure. ²⁹ Additionally, a prospective evaluation of the outcomes following interstitial PDT for patients with vascular anomalies, including IHs, provided evidence that PDT is a successful modality in the management of these pathologies that are resistant to conventional modalities, with minimal side effects. ¹² However, both of these studies were clinical studies and none of them has the systematic, basic science approach focusing on the drug pharmacokinetics and biodistribution like we are presenting here. Moreover, considering the favorable chemical properties of NP and deeper penetration of NIRL, we believe in that the treatment we used in this study will be more effective in the clinical treatment of IHs.

Limitations of the current study

The HT model in nude mice does not reflect exactly the same biological behavior of human IH. For example, mouse hemangioendothelioma cell line originated HTs did not show the spontaneous regression that is frequently seen in IH in humans. On the other hand, histological characteristics (most importantly highly vascular nature, and hyperproliferation of the vessels) were similar to IHs. These histological characteristics enabled the selective accumulation of NP in HTs, and they are critical for the success of the proposed treatment and NP. Since IHs are also highly vascular tumors, there is a strong possibility that the biodistribution of NP will be similar in IHs.

Conclusion

Systemic injection of NP carrier leads to selective accumulation in our vascular cutaneous tumor model, predisposing them to successful ablation by PDT. This data strongly suggest that NP mediated PDT can be an effective treatment for IHs and will lay the ground for future studies using more clinically relevant animal models.

Supplementary Material

Supplemental Figure 1

Supplemental Figure 1. The figure shows the ulceration on an untreated HT suggesting the superficial location of the tumor and skin involvement.