Near infrared photoimmunotherapy prevents lung cancer metastases in a murine model.
Near infrared photoimmunotherapy (NIR-PIT) is a new cancer treatment that combines the specificity of intravenously injected antibodies with the acute toxicity induced by photosensitizers after exposure to NIR-light. Herein, we evaluate the efficacy of NIR-PIT in preventing lung metastases in a mouse model. Lung is one of the most common sites for developing metastases, but it also has the deepest tissue light penetration. Thus, lung is the ideal site for treating early metastases by using a light-based strategy. In vitro NIR-PIT cytotoxicity was assessed with dead cell staining, luciferase activity, and a decrease in cytoplasmic GFP fluorescence in 3T3/HER2-luc-GFP cells incubated with an anti-HER2 antibody photosensitizer conjugate. Cell-specific killing was demonstrated in mixed 2D/3D cell cultures of 3T3/HER2-luc-GFP (target) and 3T3-RFP (non-target) cells. In vivo NIR-PIT was performed in the left lung in a mouse model of lung metastases, and the number of metastasis nodules, tumor fluorescence, and luciferase activity were all evaluated. All three evaluations demonstrated that the NIR-PIT-treated lung had significant reductions in metastatic disease (*p < 0.0001, Mann-Whitney U-test) and that NIR-PIT did not damage non-target tumors or normal lung tissue. Thus, NIR-PIT can specifically prevent early metastases and is a promising anti-metastatic therapy.
Photoimmunotherapy Inhibits Tumor Recurrence After Surgical Resection on a Pancreatic Cancer Patient-Derived Orthotopic Xenograft (PDOX) Nude Mouse Model.
Photoimmunotherapy (PIT) uses a target-specific photosensitizer based on a near-infrared (NIR) phthalocyanine dye, IR700, to induce tumor necrosis after irradiation with NIR light to kill cancer cells, such as those that remain after surgery. The purpose of the present study was to sterilize the surgical bed after pancreatic cancer resection with PIT in carcinoembryonic antigen (CEA)-expressing, patient-derived, orthotopic xenograft (PDOX) nude mouse models.
After confirmation of tumor engraftment, mice were randomized to two groups: bright light surgery (BLS)-only and BLS + PIT. Each treatment arm consisted of seven tumor-bearing mice. BLS was performed under standard bright-field with an MVX10 long-working distance, high-magnification microscope on all mice. For BLS + PIT, anti-CEA antibody conjugated with IR700 (anti-CEA-IR700) (50 µg) was injected intravenously in all mice 24 h before surgery. After the surgery, the resection bed was then irradiated with a red-light-emitting diode at 690 ± 5 nm with a power density of 150 mW/cm2.
Anti-CEA-IR700 labelled and illuminated the pancreatic cancer PDOX. Minimal residual cancer of the PDOX was detected by fluorescence after BLS. The local recurrence rate was 85.7 % for BLS-only and 28.6 % for BLS + PIT-treated mice (p = 0.05). The average recurrent tumor weight was 1149.0 ± 794.6 mg for BLS-only and 210.8 ± 336.9 mg for BLS + PIT-treated mice (p = 0.015).
Anti-CEA-IR700 was able to label and illuminate a pancreatic cancer PDOX nude mouse model sufficiently for PIT. PIT reduced recurrence by eliminating remaining residual cancer cells after BLS.
Near Infra-Red Photoimmunotherapy with Anti-CEA-IR700 Results in Extensive Tumor Lysis and a Significant Decrease in Tumor Burden in Orthotopic Mouse Models of Pancreatic Cancer
Photoimmunotherapy (PIT) uses tumor specific monoclonal antibodies that are conjugated to the photosensitizer phthalocyanine dye, IR700, which is cytotoxic upon irradiation with near-infrared (NIR) light [1–3].
Several monoclonal antibodies (mAbs) have been used with PIT in mouse models of breast cancer including trastuzumab, a monoclonal antibody (mAb) directed against human epidermal growth factor receptor-2 (HER-2), and panitumumab, a monoclonal antibody directed against human epidermal growth factor receptor-1 (HER-1) [4,5]. Cell death was induced immediately after irradiating mAb-IR700–bound target cells with NIR light. In vivo tumor shrinkage after irradiation with NIR light was demonstrated in target cells expressing the epidermal growth factor receptor. The mAb-IR700 conjugates were effective when bound to the cell membrane and produced no phototoxicity when not bound, suggesting a different mechanism for PIT as compared to conventional photodynamic therapies .
Pancreatic cancer is a lethal tumor with high rates of local and distant recurrence [6,7]. In the present study, we used a chimeric monoclonal antibody against the carcinoembryonic antigen (CEA) for PIT, which is often overexpressed in pancreatic cancer and has been previously utilized by our laboratory for fluorescence-guided surgery and fluorescence laparoscopy [8–17]. The anti-CEA antibody was conjugated to IR700 and used for PIT treatment of human pancreatic cancer in orthotopic mouse models as well as pancreatic cancer cells in vitro.
Materials and Methods
The human pancreatic cancer cell line BxPC-3 was stably transduced to express green fluorescent protein (GFP) as previously described [18,19]. Cells were maintained in RPMI 1640 medium supplemented with 10% fetal bovine serum (Hyclone, Logan, UT), penicillin/streptomycin (Gibco-BRL, Carlsbad, CA), sodium pyruvate (Gibco-BRL), sodium bicarbonate (Cellgro, Manassas, VA), l-glutamine (Gibco-BRL), and minimal essential medium nonessential amino acids (Gibco-BRL). All cells were cultured at 37°C in a 5% CO2 incubator.
Determination of CEA antigen expression level
BxPC-3-GFP cells from a 75 cm2 flask were harvested with enzyme-free cell dissociation buffer, washed once with incubation buffer (PBS + 0.5% FBS + 0.1% sodium azide), and recovered in incubation buffer at 2 x 106 cells/ml and kept at 4°C. The cells (5 x 105) were incubated with chimeric anti-CEA antibody (Genara Biosciences LLC, Morgan Hill, CA) (10 μg/ml) in 300 μl incubation buffer for 1 hour at 4°C, washed three times with PBS and stained with an Alexa488-conjugated donkey anti human IgG (H+L) antibody (Jackson Immunoresearch, West Grove, PA) for 45 minutes, followed by three washes with PBS. A control without anti-CEA was prepared in parallel. Flow cytometry profiles from the anti-CEA antibody-treated cells and the untreated cells were established on a Guava EasyCyte Plus flow cytometer (EMD Millipore, Billerica, MA). The antibody binding capacity for anti-CEA was determined by using reference beads (Bangs Laboratories, Inc., Fishers, IN) and geometric means according to the manufacturer’s protocol.
Athymic nu/nu nude mice (AntiCancer Inc., San Diego, CA), 4–6 weeks old, were used in this study. Mice were kept in a barrier facility under HEPA filtration. Mice were fed with an autoclaved laboratory rodent diet. All mouse surgical procedures and imaging were performed with the animals anesthetized by intramuscular injection of 50% ketamine, 38% xylazine, and 12% acepromazine maleate (0.02 ml). Animals received buprenorphine (0.10 mg/kg ip) immediately prior to surgery and once a day over the next 3 days to ameliorate pain. CO2 inhalation was used for euthanasia of all animals at 5 weeks after surgery. To ensure death following CO2 asphyxiation, cervical dislocation was performed. All animal studies were conducted with an AntiCancer, Inc. Institutional Animal Care and Use Committee (IACUC)-protocol specifically approved for this study and in accordance with the principals and procedures outlined in the National Institute of Health Guide for the Care and Use of Animals under Assurance Number A3873–1.
A water-soluble silicon-phthalocyanine derivative, IRDye 700DX NHS ester was obtained from LI-COR Bioscience (Lincoln, NE). of Chimeric anti-CEA antibody (Genara Biosciences LLC) (2 mg [~ 14 nmol]) at a concentration of 2 mg/ml in 0.1 M Na2HPO4 (pH = 8.6) was incubated for 2 hours at room temperature with IR700dye NHS ester (135 ug, 70 nmol) prepared in anhydrous DMSO at 5 mmol/L. After the incubation period, the IR700-conjugate was buffer exchanged and purified with phosphate buffer saline (PBS, pH = 7.1) using Amicon Ultra Centrifugal Filter Units (EMD Millipore Corporation, Billerica, MA). The IR700-mAb conjugate was repeatedly diluted with 10 ml volumes of PBS and then concentrated using the filter units until less than 2% of the unconjugated IR700 dye species remained, as determined by size exclusion HPLC (SE-HPLC). Analysis of the conjugates by SE-HPLC was performed using an Agilent 1100 HPLC system fitted with a TSKgel G2000SWxl column (Tosoh Biosciences, Tokyo, Japan). The SE-HPLC elution buffer was 1X PBS (pH = 7.1) with a flow rate of 1 ml/min. UV/Vis detection at 280 nm and 690 nm was used to determine the average dye-to-antibody ratio (DAR) for each conjugates. With this sample, a purity of 97.6% with 0.5% free dye and a DAR of 4.1 was achieved.
After confluence, BxPC-3-GFP human pancreatic cancer cells (1 x 106) were injected subcutaneously into the flanks of nude mice and allowed to engraft and grow over a period of 4–6 weeks. Tumors were then harvested and tumor fragments (1 mm3) from subcutaneous tumors were sutured to the tail of the pancreas using 8–0 nylon surgical sutures (Ethilon; Ethicon Inc., Somerville, NJ). On completion, the tail of the pancreas was returned to the abdomen, and the incision was closed in one layer using 6–0 nylon surgical sutures (Ethilon) [20,21]. The tumor fragments were allowed to grow over a period of 2 weeks.
BxPC-3 cells were seeded in white-wall 96-well plates (4,000/well) and allowed to attach overnight. Cells were incubated with the antibody conjugate, anti-CEA-IR700, (dye-antibody ratio of 5.1) at 10 mg/ml for 2 hours at 37°C. Four wells at a time were subjected to treatment with 690 nm light from an LED (Marubeni Corporation, Tokyo, Japan) at a power density of 50 mW/cm2 that was calibrated with a power meter equipped with a photodiode power sensor (Thorlabs Inc., Newton, NJ). After light treatment, the antibody solution was replaced with complete RPMI 1640 medium, containing CytoTox Green (Promega, Madison, WI) to monitor cell killing.
Photoimmunotherapy in vivo
Anti-CEA antibody (100μg) (Genara Biosciences LLC, Morgan Hill, CA) conjugated to IR700DX reconstituted to 100 μl was injected via tail vein in the treatment group 24 hours prior to intervention, while the control group had 100 μl of PBS similarly injected 24 hours prior to injection. Each group consisted of 10 mice with orthotopic BxPC-3-GFP tumors.
After 24 hours, the pancreatic tumors in all 10 mice in the treatment group were exposed via a left lateral incision and imaged to detect both the GFP signal and the 700 nm signal. All the mice were subsequently subjected to photoimmunotherapy by exposing the tumor to a 690 nm laser at 150 mW/cm2 for 30 minutes for a total of 270 J/cm2. The surrounding normal tissues were protected with aluminum foil during PIT. Mice were imaged at the time of therapy and weekly thereafter with tumor exposed to evaluate response to therapy. After 5 weeks the mice were sacrificed, at which point they were imaged and had their tumors resected and weighed.
Mice were imaged weekly using the Olympus OV100 small animal imaging system (Olympus Corp. Tokyo, Japan), containing an MT-20 light source (Olympus Biosystems Planegg, Germany) and DP70 CCD camera (Olympus Corp. Tokyo, Japan) . All images were analyzed using Image-J (National Institute of Health Bethesda, MD) and were processed with the use of Photoshop elements-11 (Adobe Systems Inc. San Jose, CA).
Tumor size determination
The mice in both groups had weekly laparotomy to expose the pancreatic tumors via a left lateral incision. Tumors were imaged with the OV-100 by GFP expression. Tumor size was assessed using Image-J software (National Institutes of Health, Bethesda, Maryland).
All statistical analysis was done using SPSS software version 21 (IBM, Armonk, NY). For pairwise comparisons, quantitative variables were calculated using the paired-samples Student’s t-test and confirmed with the Wilcoxon rank-sum test. A p-value ≤0.05 was considered significant. 95% confidence intervals obtained on analysis of the data were configured into the error bars of the appropriate figures and graphs.
Results and Discussion
Anti-CEA-IR700 binds to CEA-expressing pancreatic cancer cells in vitro and causes extensive cell death after light activation compared to control
The anti-CEA antibody binding capacity to BxPC3 human pancreatic cancer cells was 2,227,000 binding sites per cell by FACS analysis (Fig. 1). Cells were incubated with anti-CEA-IR700 and treated with 690 nm light. At the end of the incubation, there was 100% cell death in the anti-CEA IR700 + 690 nm light group compared to a negligible amount of cell death in the no-690 nm light group (Fig. 2). Death of 690 nm light-treated cells in absence of Anti-CEA-IR700 was negligible (data not shown).
Two weeks after orthotopic implantation of BxPC-3-GFP tumors, engraftment was ensured and mice were divided into 2 groups with the treatment group receiving anti-CEA-IR700 conjugate (100 μg) and the control group receiving PBS (Fig. 3). Tumor size was assessed on a weekly basis to evaluate response to therapy and overall differences in progression. In the control group there was an initial exponential increase in tumor size that began to plateau at week 4 achieving a maximum average of 390.7 mm2 (95% CI [347.7, 433.7]). In contrast, in the treatment group there was an initial decrease in tumor size from baseline with a maximal response seen at week 1 (Figs. ?(Figs.44 and ?and5)5) with an average size of 6.65mm2 (95% CI [1.75, 11.5]). Over the course of the experiment, the tumor began to regrow, reaching a maximum average of 29.5 mm2 (95% CI [16.5, 42.5]) at 5 weeks post-treatment. The difference in tumor size between the control and the treatment groups was significant (p<0.001).
The average body weights of the mice after one week of treatment were 26.3 grams (95% CI [25.1, 27.4]) for the PIT group and 25.1 grams for the control group (95% CI [24, 26.2]) which was not statistically different (p = 0.23). The average body weights of the mice 5 weeks after treatment were 29.2 grams (95% CI [28, 30.5]) for the PIT group and 28.7 grams (95% CI [27, 30.3]) for the control group which was also not statistically different (p = 0.64), indicating that PIT was well tolerated by the mice.
Despite the anti-tumor effects of PIT, there was however a 100% recurrence rate. Previous studies have shown that the rate and amount of tumor cell destruction is dependent on both the conjugate dose and the light dose, the product of which results in the same cytotoxic effect regardless of the method of light delivery (continuous or intermittent) [2,23]. In this regard, further investigation is needed to assess how dosing of the anti-CEA-IR700 complex and varying the mode and amount of energy delivery could increase efficacy.
A single round of treatment was employed in the present report as proof of principle in an orthotopic model of pancreatic cancer. Multiple rounds of PIT will be performed in future experiments. Repeated rounds of therapy have been shown to increase the efficacy of PIT . It is expected that repeated rounds of PIT would reduce the recurrence rate. Repeated rounds of PIT should be feasible due to the low toxicity observed. PIT should be able to add the efficacy of surgery and radiation therapy when used in combination with these therapies. This will be tested in orthotopic models, including immunocompetent mice, and in experimental high metastatic models, as well as in patient-derived orthotopic xenograft (PDOX) models in future experiments.
This work was supported by grants from the National Cancer Institute CA142669 and CA132971 (to M.B. and AntiCancer, Inc). AntiCancer, Inc. and Aspyrian Therapeutics provided support in the form of salaries for authors YZ, RH, LM, and MG-G but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Near Infrared Photoimmunotherapy in the Treatment of Pleural Disseminated NSCLC: Preclinical Experience
Pleural metastases are common in patients with advanced thoracic cancers and are a cause of considerable morbidity and mortality yet is difficult to treat. Near Infrared Photoimmunotherapy (NIR-PIT) is a cancer treatment that combines the specificity of intravenously injected antibodies for targeting tumors with the toxicity induced by photosensitizers after exposure to NIR-light. Herein, we evaluate the efficacy of NIR-PIT in a mouse model of pleural disseminated non-small cell lung carcinoma (NSCLC). In vitro and in vivoexperiments were conducted with a HER2, luciferase and GFP expressing NSCLC cell line (Calu3-luc-GFP). An antibody-photosensitizer conjugate (APC) consisting of trastuzumab and a phthalocyanine dye, IRDye-700DX, was synthesized. In vitro NIR-PIT cytotoxicity was assessed with dead staining, luciferase activity, and GFP fluorescence intensity. In vivo NIR-PIT was performed in mice with tumors implanted intrathoracic cavity or in the flank, and assessed by tumor volume and/or bioluminescence and fluorescence thoracoscopy. In vitro NIR-PIT-induced cytotoxicity was light dose dependent. In vivo NIR-PIT led significant reductions in both tumor volume (p = 0.002 vs. APC) and luciferase activity (p = 0.0004 vs. APC) in a flank model, and prolonged survival (p < 0.0001). Bioluminescence indicated that NIR-PIT lead to significant reduction in pleural dissemination (1 day after PIT; p = 0.0180). Fluorescence thoracoscopy confirmed the NIR-PIT effect on disseminated pleural disease. In conclusion, NIR-PIT has the ability to effectively treat pleural metastases caused by NSCLC in mice. Thus, NIR-PIT is a promising therapy for pleural disseminated tumors.
Lung cancer is the most common cause of cancer-related deaths worldwide. In the USA in 2014, 224,210 people were diagnosed with lung cancer and 159,260 died 1. Lung cancer is an aggressive disease with a very low 5-year survival. About 80% of lung cancers are histologically classified as non-small cell lung carcinoma (NSCLC). During the course of lung cancer, pleural spread of NSCLC, which is a lethal complication, frequently occurs in advanced patients 2. Although early stage and locally advanced NSCLC can be treated with a combination of surgery, chemotherapy, and radiation therapy, palliative chemotherapy is the only practical treatment for NSCLC with pleural metastases, resulting in only 6-9 month median survival 3. In recognition of the poor prognosis associated with pleural metastasis, such disease has recently been reclassified from T4 to M1a 4. Therefore, therapies that could treat pleural metastases without excessive collateral damage to the lungs might be predicted to prolong survival.
Intrapleural conventional photodynamic therapy (PDT) has been previously tested in patients after surgical debulking of pleural disease 5. However, this treatment (using porfimer sodium as the PDT agent) produced some toxicities due to the poor selectivity of the agent. PDT for malignant pleural mesothelioma was also performed after surgical debulking and immunochemotherapy, this phase III study for malignant pleural mesothelioma failed to show a difference in overall survival or progression free survival for the group with additional intraoperative PDT 6. More recently, a phase II trial of pleural PDT after surgery for NSCLC with pleural spread demonstrated that surgery and conventional PDT could be performed safely resulting in good local control and prolonged median survival 7 Thus, conventional PDT results in equivocal benefits for patients with metastases to the pleural. One clear problem with conventional PDT is that produces considerable damage to adjacent tissues thus, negating any potential benefit from the treatment itself.
The concept of using targeted light therapy is over three decades old 8,9. However, the original PDT agents were highly hydrophobic and therefore the pharmacokinetics of antibody conjugated PDT agents were difficult to target to tumors alone. Previous studies have attempted to target conventional PDT agents by conjugating them to antibodies. Unfortunately, these conjugates were usually trapped in the liver and could only be used in isolated body cavities such as the peritoneum 10,11. A study using a more hydrophilic phthalocyanine-based photosensitizer (Aluminum (III) Phthalocyanine Tetrasulfonate) has been published, however, no in vivo treatment response data was reported 12. The recognition that substituting a water soluble phthalocyanine-based photosensitizer (IR700) in the conjugation with an antibody and applying near infrared light has led to much higher selectivity. NIR-PIT differs from these prior PDT not only in the water-solubility of the photosensitizer, but also in its reliance on NIR light that has better tissue penetration than the lower wavelengths used for exciting PDT agents. This antibody-photosensitizer conjugates (APC) demonstrates similar intravenous pharmacokinetics to naked antibodies, resulting in highly targeted tumor accumulation with minimal non-target binding. When bound to targeted cells, APCs induce rapid, selective cytotoxicity after exposure to NIR light. In vitro studies have demonstrated that NIR-PIT is highly target cell-specific and leads to rapid and irreversible cell death due to membrane damage 13–16.
One obvious limitation of NIR-PIT is that it would seem limited to tumors located relatively shallow from the surface that can be easily exposed to NIR light. However, light can be administered endoscopically and among the organs, the lungs have the best ability to transmit light because they are mostly filled with air. Thus, NIR light administered to the thoracic cavity could easily penetrate within pleural disease. In this study, we investigate the efficacy of NIR-PIT for treating pleural disease in a NSCLC mouse model.
Materials and methods
Water soluble, silicon-phthalocyanine derivative, IRDye 700DX NHS ester and IRDye 800CW NHS ester were obtained from LI-COR Bioscience (Lincoln, NE, USA). Panitumumab, a fully humanized IgG2 mAb directed against EGFR, was purchased from Amgen (Thousand Oaks, CA, USA). Trastuzumab, 95% humanized IgG1 mAb directed against HER2, was purchased from Genentech (South San Francisco, CA, USA). All other chemicals were of reagent grade.
Synthesis of Antibody-dye conjugates
Conjugation of dyes with mAbs was performed according to previous reports 13,14. In brief, panitumumab or trastuzumab (1 mg, 6.8 nmol) was incubated with IR700 NHS ester (60.2 µg, 30.8 nmol) or IR800CW NHS ester (35.9 µg, 30.8 nmol) in 0.1 mol/L Na2HPO4 (pH 8.6) at room temperature for 1 hr. The mixture was purified with a Sephadex G25 column (PD-10; GE Healthcare, Piscataway, NJ, USA). The protein concentration was determined with Coomassie Plus protein assay kit (Thermo Fisher Scientific Inc, Rockford, IL, USA) by measuring the absorption at 595 nm with spectroscopy (8453 Value System; Agilent Technologies, Santa Clara, CA, USA). The concentration of IR700 or IR800 was measured by absorption at 689 nm or 774 nm respectively to confirm the number of fluorophore molecules conjugated to each mAb. The synthesis was controlled so that an average of four IR700 molecules or two IR800 molecules were bound to a single antibody. We performed SDS-PAGE as a quality control for each conjugate as previously reported 13. We abbreviate IR700 conjugated to trastuzumab as tra-IR700, to panitumumab as pan-IR700 and IR800 conjugated to trastuzumab as tra-IR800.
HER2 and luciferase/GFP-expressing Calu3-luc-GFP cells were established with a transfection of RediFect Red-FLuc-GFP (PerkinElmer, Waltham, MA, USA). High GFP and luciferase expression was confirmed with 10 passages of the cells. Balb/3T3 cells were transfected with RFP (EF1a)-Puro lentiviral particles (AMSBIO, Cambridge, MA, USA). High, stable RFP expression was confirmed after 10 passages in the absence of a selection agent. To evaluate specific cell killing by NIR-PIT, 3T3 cells stably expressing RFP (3T3-RFP) were used as negative controls. Cells were grown in RPMI 1640 (Life Technologies, Gaithersburg, MD, USA) supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin (Life Technologies) in tissue culture flasks in a humidified incubator at 37°C at an atmosphere of 95% air and 5% carbon dioxide.
Fluorescence from cells after incubation with pan-IR700 or tra-IR700 was measured with a flow cytometer (FACS Calibur, BD BioSciences, San Jose, CA, USA) and CellQuest software (BD BioSciences). Calu3-luc-GFP cells (1×105) were incubated with each APC for 6 hr at 37°C. To validate the specific binding of the conjugated antibody, excess antibody (50 µg) was used to block 0.5 µg of antibody-dye conjugates 16.
To detect the antigen specific localization of antibody-dye conjugates, fluorescence microscopy was performed (IX61 or IX81; Olympus America, Melville, NY, USA). Ten thousand cells were seeded on cover-glass-bottomed dishes and incubated for 24 hr. Tra-IR700 was then added to the culture medium at 10 µg/mL and incubated at 37°C for 6 hr. The cells were then washed with PBS; Propidium Iodide (PI)(1:2000)(Life Technologies) and Cytox Blue (1:500)(Life Technologies), were used to detect dead cells. They were added into the media 30 min before the observation. The cells were then exposed to NIR light (2 J/cm2) and serial images were obtained. The filter was set to detect IR700 fluorescence with a 590-650 nm excitation filter, and a 665-740 nm band pass emission filter.
In vitro NIR-PIT
One hundred thousand cells were seeded into 24 well plates or ten million cells were seeded onto a 10 cm dish and incubated for 24 hr. Medium was replaced with fresh culture medium containing 10 µg/mL of tra-IR700 which was incubated for 6 hr at 37°C. After washing with PBS, phenol red free culture medium was added. Then, cells were irradiated with a NIR laser, which emits light at 685 to 695 nm wavelength (BWF5-690-8-600-0.37; B&W TEK INC., Newark, DE, USA). The actual power density of mW/cm2 was measured with an optical power meter (PM 100, Thorlabs, Newton, NJ, USA).
Cytotoxicity/ Phototoxicity assay
The cytotoxic effects of NIR-PIT with tra-IR700 were determined by the luciferase activity and flow cytometric PI staining. For luciferase activity, 150 μg/mL of D-luciferin-containing media (Gold Biotechnology, St Louis, MO, USA) was administered to PBS-washed cells 1 hr after NIR-PIT, and analyzed on a bioluminescence imaging (BLI) system (Photon Imager; Biospace Lab, Paris, France). For the flow cytometric assay, cells were trypsinized 1 hr after treatment and washed with PBS. PI was added to the cell suspension (final 2 μg/mL) and incubated at room temperature for 30 min, prior to flow cytometry.
To investigate the specificity of tra-IR700, excess trastuzumab 1,000 µg/mL added to the medium for 1 hr, and 10 µg/mL of tra-IR700 was added to the media for 6 hr. Without washing with PBS, NIR light was administered and 1 hr later PI staining was performed as above.
Estimation of GFP fluorescence intensity in vitro
Two hundred thousand cells were seeded on cover-glass-bottomed dishes and incubated for 12 hr. Tra-IR700 was then added to the culture medium (phenol red free) at 10 µg/mL and incubated at 37°C for 6 hr, followed by NIR-PIT. Cells were trypsinized 1 hr after treatment and washed with PBS, then analyzed by flow cytometry.
Animal and tumor models
All in vivo procedures were conducted in compliance with the Guide for the Care and Use of Laboratory Animal Resources (1996), US National Research Council, and approved by the local Animal Care and Use Committee. Six- to eight-week-old female homozygote athymic nude mice were purchased from Charles River (NCI-Frederick). During procedures, the mice were anesthetized with inhaled isoflurane.
Six million Calu3-luc-GFP cells were injected subcutaneously in the right dorsum of the mice. The greatest longitudinal diameter (length) and the greatest transverse diameter (width) were measured with an external caliper. Tumor volumes based on caliper measurements were calculated by the following formula; tumor volume = length × width2 × 0.5. Tumors reaching approximately 100 mm3 in volume were selected for further experiments. Body weight was checked on the scale.
For BLI, D-luciferin (15 mg/mL, 200 μL) was injected intraperitoneally and the mice were analyzed with a Photon Imager for luciferase activity at day 11. Mice were selected for further study based on tumor size and bioluminescence.
In order to create a pleural disseminated NSCLC model, six million Calu3-luc-GFP NSCLC cells in PBS (total 200 μL) were injected into the thoracic cavity through a right intercostal space using a 30G needle. To avoid lung injury, the needle could only be inserted 5 mm (a foam styrol stopper prevented deeper insertion). Twenty days later, bioluminescence was performed after D-luciferin (15 mg/mL, 200 μL) was injected intraperitoneally and the mice were imaged with the Photon Imager; mice with sufficient activity were selected for further study.
In vivo fluorescence imaging
In vivo fluorescence images were obtained with a Pearl Imager (LI-COR Bioscience) for detecting IR700/ IR800 fluorescence, and a Maestro Imager (CRi, Woburn, MA, USA) for GFP. For GFP, a band-pass filter from 445 to 490 nm (excitation) and a long-pass blue filter over 515 nm (emission) were used. The tunable emission filter was automatically stepped in 10 nm increments from 500 to 600 nm for the green filter sets at a constant exposure (1000 msec). The spectral fluorescence images consist of autofluorescence spectra and the spectra from GFP (Calu3-luc-GFP tumor), which were then unmixed, based on the characteristic spectral pattern of GFP, using Maestro software (CRi).
A model BF XP-60 bronchoscope system was inserted by a trained bronchoscopist/thoracosciopist (KS) via an intercostal space after the animal was euthanized, and the intrathoracic cavity was observed with white light and fluorescence imaging using multi-band excitation filters. Thoracoscopic images were obtained via a dichroic splitter, in which both the excitation light images were displayed using the image processor (OTV-S7; Olympus Co., Tokyo, Japan; not commercially available), and the fluorescence images, which were filtered by in-house designed multicolor emission filters (516 to 556 nm band-pass for GFP and 680 to 710 nm band-pass for IR700) were detected with an (EM)-CCD camera (Texas Instruments, Dallas, TX, USA). Both images were displayed side by side on the PC monitor with DualView 2 software (RGB Spectrum). Real-time images of both white light and fluorescence images were recorded. Camera gain, exposure time, and binning for the fluorescence images were held constant in each fluorescent protein throughout the study. Analysis of the images was performed with ImageJ software (http://rsb.info.nih.gov/ij/).
Characterization of the pleural disseminated mouse model
Both the disseminated pleural model and the subcutaneous bilateral flank models received 100 μg of tra-IR700 or tra-IR800 intravenously (tra-IR800 was used to avoid auto-fluorescence). One day after injection, serial images were performed with a fluorescence imager (Pearl Imager) for detecting IR700/ IR800 fluorescence, with the Photon Imager for BLI, and the Olympus BF XP-60 thoracoscopy. Images of the mice were obtained with an iphone5 (Apple Inc., Cupertino, CA, USA).
In vivo NIR-PIT
Calu3-luc-GFP right dorsum tumor xenografts were randomized into 4 groups of at least 7 animals per group undergoing one of the following treatments: (repeated PIT)18: (1) no treatment (control); (2) only NIR light exposure at 50 J/cm2 on day 1 and 100 J/cm2 on day 2; (3) 100 μg of tra-IR700 i.v., no NIR light exposure; (4) 100 μg of tra-IR700 i.v., NIR light was administered at 50 J/cm2 on day 1 after injection and 100 J/cm2on day 2 after injection. These therapies were performed only once at day 14 after cell implantation. Mice were monitored daily, and tumor volumes and body weight were measured three times a week until the tumor diameter reached 2cm, whereupon the mouse was euthanized with carbon dioxide.
In vivo imaging was acquired with a fluorescence imager (Pearl Imager) for detecting IR700 fluorescence, and the Photon Imager for BLI. For analyzing BLI, ROI of similar size were placed over the entire tumor.
For evaluation of NIR-PIT effects in the pleural disseminated NSCLC mouse model, mice were randomized into 4 groups of 7 animals per group including: (1) no treatment (control); (2) only NIR light exposure at 50 J/cm2 on day 1 and 100 J/cm2 on day 2; (3) 100 μg of tra-IR700 i.v., no NIR light exposure; (4) 100 μg of tra-IR700 i.v., NIR light was administered at 50 J/cm2 on day 1 after injection and 100 J/cm2 on day 2 after injection. NIR light was applied transcutaneously followed by serial fluorescence imaging and BLI.
To evaluate histological changes of lung at 1 day after PIT, microscopy was performed (BX51, Olympus America). Lungs with tumors were harvested and placed in 10% formalin. Serial 10-μm slice sections were fixed on glass slide for H-E staining.
Data are expressed as means ± s.e.m. from a minimum of four experiments, unless otherwise indicated. Statistical analyses were carried out using a statistics program (GraphPad Prism; GraphPad Software, La Jolla, CA, USA). For multiple comparisons, a one-way analysis of variance (ANOVA) with Tukey's test was used. The cumulative probability of survival, determined herein as the tumor diameter failing to reach 2 cm, was estimated in each group with the use of the Kaplan-Meier survival curve analysis, and the results were compared with the log-rank test and Wilcoxon test. p < 0.05 was considered to indicate a statistically significant difference.
Characterization of the cell line and the NIR-PIT effect
To monitor optically the effect of NIR-PIT, NSCLC cell line Calu3 was genetically modified to express GFP and luciferase (Calu3-luc-GFP)(Fig. ?(Calu3-luc-GFP)(Fig.1A).1A). The fluorescence signals obtained with pan-IR700 and tra-IR700 using Calu3-luc-GFP cells were evaluated by FACS. After 6 hr incubation with either pan-IR700 or tra-IR700, Calu3-luc-GFP cells showed higher brightness with tra-IR700 than with pan-IR700 consistent with the expression profile (Fig. ?(Fig.1B).1B). These signals were completely blocked by the addition of excess trastuzumab, suggesting specific binding and validating that the addition of the luciferase/ GFP gene had not altered the cell expression profile. Serial fluorescence microscopy of Calu3-luc-GFP cells performed before and after NIR-PIT (2 J/cm2) demonstrated rapidly appearing cellular swelling, bleb formation and rupture of the lysosome (Fig. ?(Fig.1C).1C). Time-lapse imaging showed acute morphologic changes in the cell membrane within 25 minutes and fluorescence of PI indicating cell death (Additional File 2: Video S1). No significant changes were observed in HER2-negative 3T3 cells after exposure to NIR light, suggesting NIR-PIT induced no damage in non-target cells (Additional File 1: Fig. S1). Based on the incorporation of PI, the cell death percentage increased in a light dose dependent manner. No significant cytotoxicity was observed with NIR light exposure alone or with tra-IR700 alone (Fig. ?(Fig.1D).1D). NIR-PIT was blocked with excess trastuzumab even in tra-IR700 containing media (Additional File 1: Fig. S2). Bioluminescence showed significant decreases of relative light units (RLU) in NIR-PIT treated cells (Fig. ?(Fig.1E).1E). BLI also showed a decrease of luciferase activity in a light dose dependent manner (Fig. ?(Fig.1F).1F). GFP fluorescence intensity was greatly reduced in dead cells (stained positive with PI), while GFP fluorescence was preserved in surviving cells (Fig. ?(Fig.1G).1G). GFP fluorescence was likely reduced after NIR-PIT because the GFP was extruded from the cytoplasm after membrane rupture leading to dilution and/or denaturation. The GFP fluorescence ratio on FACS showed decreases in a light dose dependent manner, while no decrease was detected with NIR light exposure or Pan-IR700 alone (Fig. ?(Fig.1H).1H). Collectively, these data suggested that the effects of NIR-PIT on Calu3-luc-GFP could be monitored with GFP fluorescence and bioluminescence.
In vivo NIR-PIT reduced tumor volume and luciferase activity in a flank xenograft model
In vivo NIR-PIT experiments were first conducted on flank xenografts of Calu3-luc-GFP. The NIR-PIT regimen and imaging protocol are depicted in Fig. ?Fig.2A.2A. Both BLI and fluorescence decreased after NIR-PIT (Fig. ?(Fig.2B2B and Additional File 1: Fig. S3A). RLU of tumor in other groups showed a gradual increase due to tumor growth. In contrast, luciferase activity decreased 1 day after repeated NIR-PIT (*p = 0.002 < 0.01, PIT vs. APC, Tukey's test with ANOVA)(Fig. ?ANOVA)(Fig.2C).2C). The body weight (BW) ratio showed no remarkable acute toxicity (Fig. ?(Fig.2D).2D). Significant decreases (**p = 0.0004 < 0.001, PIT vs. APC, Tukey's test with ANOVA) in tumor volume were confirmed, which was consistent with luciferase activities (Fig. ?(Fig.2E).2E). Survival was prolonged significantly in the PIT group (***P < 0.0001, Long-rank test and Wilcoxon test)(Fig. ?test)(Fig.2F).2F). Since bioluminescence is more sensitive to tumor killing as it is based on live cells, the physical tumor volume took longer to show the effect of NIR-PIT. Collectively, these data suggest that NIR-PIT caused significant tumor reduction and prolonged survival in the in vivo flank tumor model.
Characterization of the pleural disseminated NSCLC mouse model
Prior to therapy, implanted thoracic tumors were evaluated with serial fluorescence imaging, BLI and fluorescence thoracoscopy. The implanted thoracic disseminated tumors demonstrated high activity with fluorescence imaging based on IR700, IR800 and GFP, but also high activity on bioluminescence, which co-localized with each other (Fig. ?(Fig.3).3). Fluorescence thoracoscopy indicated that disseminated tumor establishment and the good contrast of IR700 between tumors and intrathoracic organs (Fig. ?(Fig.33 and Additional File 3: video S2), which confirmed pleural metastases that fluoresced preferentially with tra-IR700. These data suggest that pleural disseminated NSCLC cancer mouse model with Calu3-luc-GFP cells was successfully established; intravenously injection of agent could reach the disseminated tumors.
In vivo NIR-PIT effect in pleural disseminated cancer mouse model
After treatment with NIR-PIT pleural disseminated tumors decreased in bioluminescence and fluorescence (Fig. ?(Fig.4A4A and ?and4B4B and Additional File 1: Fig. S3B). While the RLU decreased in the NIR-PIT treated tumors, RLU of tumor in other groups showed a gradual increase due to tumor growth. In contrast, luciferase activity decreased 1 day after repeated NIR-PIT (*p = 0.0180 < 0.05, PIT vs. APC, Tukey's test with ANOVA)(Fig.?ANOVA)(Fig.4C).4C). The BW ratio showed no change (ns, PIT vs. APC, light, control, Tukey's test with ANOVA) (Fig. ?(Fig.4D).4D). Taken together, these data suggest that NIR-PIT caused significant tumor reduction in vivo pleural disseminated model.
In vivo NIR-PIT effect assessed with GFP fluorescence imaging
Finally, to assess the effect of repeated NIR-PIT on Calu3-luc-GFP tumor in vivo, GFP fluorescence imaging was performed in both the flank model and pleural disseminated model (Fig. ?(Fig.5A).5A). With the flank model, both GFP/ IR700 fluorescence disappeared at 1 day after NIR-PIT, which was confirmed by ex vivo tumor imaging (Fig. ?(Fig.5B).5B). Using fluorescence thoracoscopy, GFP and IR700 fluorescence disappeared (Fig. ?(Fig.5C5C and videos S3 and S4 in Additional Files 4-5). A small effusion was observed with thoracoscopy (Fig. ?(Fig.5C5C arrow). Moreover, there was no apparent damage to the normal lung by NIR-PIT as observed with histological analysis (Fig. ?(Fig.66).
In this study, we demonstrate that an APC can be delivered to both flank and intrathoracic tumor after intravenous injection and that subsequent NIR-PIT can be successfully performed transcutaneously to the mouse thorax with acceptable morbidity. Among the imaging tools used to document tumor regression, which included BLI, fluorescence imaging and fluorescence thoracoscopy, the latter two could be used in clinical practice using the IR700 dye 19,20. BLI using firefly luciferase, although less suitable for clinical translation, was useful as a primary outcome measure as it requires both oxygen and ATP to actively transport the substrate luciferin and subsequently catalyze the photochemical reaction 21,22. Since NIR-PIT-induced necrotic cell death releases ATP, BLI is an appropriate and sensitive biomarker for NIR-PIT 16,23.In vivo GFP fluorescence imaging enables the full process of tumorigenesis, treatment, regression, metastasis, or recurrence, to be detected although this also is not translatable 24. The high level of GFP tumor fluorescence in this model permitted imaging with quantification of tumor growth and dissemination without the need for additional contrast agents. By employing cytoplasmic GFP expressing cells, antitumor effects induced by NIR-PIT could be clearly monitored as extrusion of GFP from treated cells resulted in a diminution of signal 15. The development of a mini-endoscope mimicking thoracoscopy permits intrathoracic fluorescence imaging and is the most likely method by which NIR-PIT would be administered and monitored in humans. By changing the filter sets, multicolor endoscopic imaging becomes possible to simultaneously monitor tumor regression with GFP fluorescence and accumulation of APC with IR700 fluorescence as shown in videos S3 and S4 25 (Additional Files 4-5). Both real-time color capability and direct access to the disseminated tumors, resulted in much higher resolution imaging. From the photophysical point of view, the endoscope can minimize light scattering and absorbance that is caused by overlapping tissue, resulting in more precise depiction of the lesion.
In this study, we use a pleural disseminated tumor model by simple tumor cell injection in the thoracic cavity. A variety of animal models could be used including percutaneous orthotopic injection (POI), surgical orthotopic injection (SOI), and transpleural orthotopic injection (TOI)26. Various advantages and disadvantages exist among these models. For example, SOI requires high skills and is invasive, resulting in high pre-procedure mortality, however SOI is thought to be more physiologic than others 26,27. The POI model has the advantages of simplicity and less invasiveness with a very low pre-procedure mortality rate 26. A recent study reported that implantation rates were similar among these models 26. With these considerations, we chose the POI approach. This approach had a high implantation rate (around 85%) of NSCLC pleural dissemination, which was confirmed by non-invasive BLI.
The survival of patients with NSCLC patients with pleural disseminations is only 6 to 9 months even with systemic chemotherapy 3. Surgery is not currently performed because of its morbidity and limited benefit 28. While not likely to be curative, NIR-PIT could offer the benefit of local control with minimal invasiveness. Moreover, NIR-PIT could be readily used as an adjunct to conventional surgery at the time of initial diagnosis.
There are several limitations to this study. First, not all lung cancers overexpress HER2, and therefore this particular target may not be ideal in other lung cancers. Fortunately, NIR-PIT has proven effective with almost all APCs with which it has been attempted and therefore, it is likely that the proper APC or combination of APCs could be found to treat a specific phenotype of lung cancer cell membrane expression13–15,29,30. We were also unable to determine the long-term side effects of NIR-PIT in this limited model. Short-term studies of the mice demonstrated no apparent adverse events after NIR-PIT. It is possible that sudden widespread cell necrosis could cause either acute or delayed toxicity but none was observed in this model. Only small reactive pleural effusions were observed by thoracoscopy. Additionally, it is clear that NIR-PIT alone will not be sufficient to cure thoracic metastases, although the use of NIR light to activate IR700 will produce deeper tissue penetration within larger masses than the shorter wavelengths of light used in conventional PDT photoactivation or other light therapies such as those using UV light 31. Therefore, we foresee NIR-PIT as an adjuvant to surgery with an initial debulking procedure followed by NIR-PIT to “mop up” residual disease. Furthermore, it is interesting to consider the possibility that systemic chemotherapy may be more effective after NIR-PIT. Previous studies have shown that NIR-PIT causes treated tumors to exhibit increased permeability to nano-sized drugs. Therefore, current or future chemotherapies for lung cancer may benefit from prior treatment with NIR-PIT 32. Finally, in this study we irradiated transcutaneously which is difficult to translate to the clinic, however, it would be feasible to deliver light via thoracoscopy, bronchoscopy or even during open-surgery. Thus, although this particular animal model is not directly translatable, the principle of treating thoracic malignancies with light therapy is feasible.
In conclusion, this study demonstrates that NIR-PIT effectively treated pleural metastases in a mouse model of NSCLC. NIR-PIT could be a promising adjuvant for treating pleural carcinomatosis replacing or adding to existing therapies such as surgery and chemotherapy.
This research was supported by the Intramural Research Program of the National Institutes of Health, National Cancer Institute, Center for Cancer Research. K.S. is supported with JSPS Research Fellowship for Japanese Biomedical and Behavioral Researchers at NIH.
K.S. mainly conducted experiments, performed analysis and wrote the manuscript; T.N. conducted thoracoscopy with K.S. P.L.C. wrote the manuscript and supervised the project; and H.K. planned and initiated the project, designed and conducted experiments, wrote the manuscript, and supervised the entire project.
Metastatic Recurrence in a Pancreatic Cancer Patient Derived Orthotopic Xenograft (PDOX) Nude Mouse Model Is Inhibited by Neoadjuvant Chemotherapy in Combination with Fluorescence-Guided Surgery with an Anti-CA 19-9-Conjugated Fluorophore
The aim of this study is to determine the efficacy of neoadjuvant chemotherapy (NAC) with gemcitabine (GEM) in combination with fluorescence-guided surgery (FGS) on a pancreatic cancer patient derived orthotopic xenograft (PDOX) model. A PDOX model was established from a CA19-9-positive, CEA-negative tumor from a patient who had undergone a pancreaticoduodenectomy for pancreatic adenocarcinoma. Mice were randomized to 4 groups: bright light surgery (BLS) only; BLS+NAC; FGS only; and FGS+NAC. An anti-CA19-9 or anti-CEA antibody conjugated to DyLight 650 was administered intravenously via the tail vein of mice with the pancreatic cancer PDOX 24 hours before surgery. The PDOX was brightly labeled with fluorophore-conjugated anti-CA19-9, but not with a fluorophore-conjugated anti-CEA antibody. FGS was performed using the fluorophore-conjugated anti-CA19-9 antibody. FGS had no benefit over BLS to prevent metastatic recurrence. NAC in combination with BLS did not convey an advantage over BLS to prevent metastatic recurrence. However, FGS+NAC significantly reduced the metastatic recurrence frequency to one of 8 mice, compared to FGS only after which metastasis recurred in 6 out of 8 mice, and BLS+NAC with metastatic recurrence in 7 out of 8 mice (p?=?0.041). Thus NAC in combination with FGS can reduce or even eliminate metastatic recurrence of pancreatic cancer sensitive to NAC. The present study further emphasizes the power of the PDOX model which enables metastasis to occur and thereby identify the efficacy of NAC in combination with FGS on metastatic recurrence.
Complete tumor resection improves overall survival of pancreatic cancer patients, which is presently 5% at five years . Metastatic relapse often occurs following attempted curative resection of the primary tumor as a result of invisible microscopic tumor deposits left behind. Making tumors fluoresce offers great advantages for tumor detection during surgery in order to achieve complete resection , . We have previously shown that fluorescence-guided surgery (FGS) for pancreatic cancer decreased the residual tumor burden and improved overall and disease-free survival in mouse models using fluorescently-labeled human pancreatic cancer cell lines –.
Patient-derived orthotopic xenografts (PDOX) recapitulate the biological characteristics of the disease of origin, including metastases – and are a clinically-relevant model for fluorescence-guided surgery ,–.
Recently, many studies reported positive outcomes with neoadjuvant chemotherapy (NAC) of pancreatic cancer –. NAC allows for the identification of those patients with rapidly progressive metastatic disease at the time of preoperative restaging, and can increase the R0 resection rate and reduce the risk of local tumor recurrence . However, a significant number of patients still develop recurrent disease immediately after NAC treatment and subsequent surgical resection –. Therefore, new strategies in addition to NAC are needed to reduce the recurrence of pancreatic cancer. In this study, we determined the efficacy of CA19-9 conjugated with a fluorescent dye to illuminate pancreatic cancer PDOXs for FGS in combination with NAC.
Materials and Methods
Athymic nu/nu nude mice (AntiCancer Inc., San Diego, CA), 4–6 weeks old, were used in this study. Mice were kept in a barrier facility under HEPA filtration. Mice were fed with an autoclaved laboratory rodent diet. All mouse surgical procedures and imaging were performed with the animals anesthetized by intramuscular injection of 50% ketamine, 38% xylazine, and 12% acepromazine maleate (0.02 ml). Animals received buprenorphine (0.10 mg/kg ip) immediately prior to surgery and once a day over the next 3 days to ameliorate pain. CO2 inhalation was used for euthanasia of all animals at 90 days after surgery. To ensure death following CO2 asphyxiation, cervical dislocation was performed. All animal studies were conducted with an AntiCancer, Inc. Institutional Animal Care and Use Committee (IACUC)-protocol specifically approved for this study and in accordance with the principals and procedures outlined in the National Institute of Health Guide for the Care and Use of Animals under Assurance Number A3873-1.
Establishment of patient derived orthotopic xenograft (PDOX) of pancreatic cancer
Pancreatic cancer patient tumor tissues were obtained at surgery and cut into fragments (3-mm3) and transplanted subcutaneously in nude mice , . The subcutaneous tumors were then passaged in nude mice both orthotopically and subcutaneously. All patients provided written informed consent and samples were procured and initially transplanted in NOD/SCID under the approval of the Institutional Review Board of MD Anderson Cancer Center.
Orthotopic tumor implantation
A small 6- to 10-mm transverse incision was made on the left flank of the mouse through the skin and peritoneum. The tail of the pancreas was exposed through this incision, and a single 3-mm3 tumor fragment from subcutaneous tumors was sutured to the tail of the pancreas using 8-0 nylon surgical sutures (Ethilon; Ethicon Inc., NJ, USA). On completion, the tail of the pancreas was returned to the abdomen, and the incision was closed in one layer using 6-0 nylon surgical sutures (Ethilon) , .
Antibody conjugation and tumor labeling
Monoclonal antibodies specific for carbohydrate antigen 19-9 (CA19-9) and carcinoembryonic antigen (CEA) were obtained from Abcam Inc. (Cambridge, MA, USA) and (Aragen Bioscience, Inc. (Morgan Hill, CA, USA), respectively. The antibodies were labeled with the DyLight 650 Protein Labeling Kit (Thermofisher Scientific, Waltham, MA, USA) according to the manufacturer’s instructions , , . To determine if the anti-CA19-9 antibody, conjugated with DyLight 650 (anti-CA19-9-650), and the anti-CEA antibody, conjugated with DyLight650 (anti-CEA-650), could label the pancreatic tumor in vivo, 50 µg of anti-CA19-9-650 or anti-CEA-650 were injected into the tail vein of the mice with subcutaneous pancreatic tumors. Twenty-four hours later, whole body images were obtained with the OV100 Small Animal Variable Magnification Imaging System (Olympus, Tokyo, Japan).
After confirmation of tumor engraftment, 32 mice were randomized to 4 groups; BLS only; BLS+NAC; FGS only; and FGS+NAC. Each treatment arm involved 8 tumor-bearing mice. The mice randomized to NAC-treatment were administered 80 mg/kg gemcitabine (GEM) (Eli Lilly and Company, Indianapolis, IN, USA). GEM was injected i.p. on day 8, 15 and 22. No significant effects on body weight, morbidity, or severe toxicity were observed in NAC-treated mice.
For fluorescence-guided surgery (FGS), a 15-mm transverse incision was made on the left flank of the mouse through the skin and peritoneum which was kept open with a retractor. The tail of the pancreas was exposed through this incision. Fifty µg of anti-CA19-9 antibody, conjugated to DyLight 650, was injected via the tail vein in the mice in the FGS group 24 hours before surgery. A MINI MAGLITE LED PRO flashlight (MAG INSTRUMENT, Ontario, CA, USA) coupled to an excitation filter (ET 640/30X, Chroma Technology Corporation, Bellows Falls, VT, USA) was used as the excitation light source. A Canon EOS 60D digital camera with an EF–S18–55 IS lens (Canon, Tokyo, Japan) coupled with an emission filter (HQ700/75M-HCAR, Chroma Technology Corporation) was used as the real-time image capturing device for FGS. BLS was performed under standard bright-field using an MVX10 microscope (Olympus, Tokyo, Japan). After completion of surgery, the incision was closed in one layer using 6-0 nylon surgical sutures, and the mice were allowed to recover in their cages.
Tumor samples were removed with surrounding normal tissues at the time of resection. Fresh tissue samples were fixed in 10% formalin and embedded in paraffin before sectioning and staining. Tissue sections (3 µm) were deparaffinized in xylene and rehydrated in an ethanol series. Hematoxylin and eosin (H & E) staining was performed according to standard protocols. For immunohistochemistry, the sections were then treated for 30 min with 0.3% hydrogen peroxide to block endogenous peroxidase activity. The sections were subsequently washed with PBS and unmasked in citrate antigen unmasking solution (Mitsubishi Kagaku Iatron, Inc., Tokyo, Japan) in a water bath for 40 min at 98°C. After incubation with 10% normal goat serum, the sections were incubated with anti-CA19-9 antibody (1?100) and anti-CEA antibody (1?100) at 4°C overnight. The bound primary antibodies were detected by binding with an anti-mouse secondary antibody and an avidin/biotin/horseradish peroxidase complex (DAKO Cytomation, Kyoto, Japan) for 30 min at room temperature. The labeled antigens were visualized with a DAB kit (DAKO Cytomation). The sections were counterstained with hematoxylin and observed with a BH-2 microscope (Olympus, Tokyo, Japan) equipped with an INFINITY1 2.0 megapixel CMOS digital camera (Lumenera Corporation, Ottawa, Canada). All images were acquired using INFINITY ANALYZE software (Lumenera Corporation) without post-acquisition processing.
Evaluation of histopathological response to NAC
Histopathological response to chemotherapy drugs was defined according to Evans’s grading scheme: Grade I, little (<10%) or no tumor cell destruction is evident; Grade IIa, destruction of 10%–50% of tumor cells; Grade IIb, destruction of 51%–90% of tumor cells; Grade III, few (<10%) viable-appearing tumor cells are present; Grade IV, no viable tumor cells are present .
Evaluation of tumor recurrence and progression
To assess for recurrence postoperatively, animals underwent laparotomy 12 weeks after surgery, and the tumors were imaged with the Canon EOS 60D digital camera with an EF–S18–55 IS lens (Canon), excised, harvested and weighed for analysis.
PASWStatistics 18.0 (SPSS, Inc.) was used for statistical analyses. Tumor weight was expressed as mean ± SD. The two-tailed Student’s t-test was used to compare continuous variables between 2 groups. Comparisons between categorical variables were analyzed with Fisher’s exact test. A p value <0.05 was considered statistically significant for all comparisons.
The pancreatic PDOX tumor was diagnosed as moderately differentiated adenocarcinoma with H&E staining (Figure 1A). Based on immunohistochemistry, the PDOX tumor was found to be CA19-9-positive and CEA-negative (Figures 1B and 1C). The PDOX was brightly labeled with anti-CA19-9-650 (Figure 1D), but the fluorescence signal with anti-CEA-650 was very weak (Figure 1E). The fluorescence results were consistent with the immunohistochemical results, and based on them, it was decided to use anti-CA19-9-650 to label the PDOX for FGS. Anti-CA19-9-650 was injected in the tail vein of the mice with PDOX tumors 24 hours before FGS.
Sensitivity of PDOX to NAC
The PDOX mice were randomized to 4 groups; BLS only; BLS+NAC; FGS only; FGS+NAC. Each treatment arm involved 8 PDOX mice. The mice randomized to the NAC group were treated with GEM on days 8, 15 and 22 (Figure 2). All animals underwent surgery on day 29 (Figures 3). The average excised PDOX tumor weight was 188.5±53.1 mg for BLS-only; 84.5±51.6 mg for BLS+NAC; 299.0±86.3 mg for FGS-only; and 141.8±48.9 mg for FGS+NAC. The average excised tumor weight in the BLS+NAC mice was significantly less than in the BLS-only mice (p?=?0.001). The average excised tumor weight in the FGS+NAC mice was also significantly less than FGS-only mice (p<0.001). Upon histological examination, over 50% of cancer cells were dead and replaced by necrotic tissue or stromal cells in the PDOX tumor treated with FGS+NAC and was judged as Evan’s grade IIb – III (Figure 4).
Effect of NAC on tumor recurrence with BLS or FGS
With regard to the recurrent tumor weight, the average local recurrent tumor weight was 389.2±356.6 mg in BLS-only treated mice; 369.1±251.9 mg in BLS+NAC-treated mice; 73.0±77.2 mg in FGS-only treated mice; and 78.4±90.8 mg in FGS+NAC-treated mice. The average local recurrent tumor weight in FGS-only treated mice was significantly less than in BLS-only treated mice (p?=?0.041). The average metastatic recurrent tumor weight of the pancreatic cancer PDOX was 170.7±184.2 mg for BLS-only treated mice; 40.0±19.7 mg for BLS+NAC-treated mice; 31.3±37.6 mg for FGS-only mice; and 1.3±3.7 mg for FGS+NAC-treated mice. The average metastatic recurrent tumor weight in FGS+NAC was significantly less than BLS+NAC (p?=?0.001). The metastatic recurrent weight in the FGS+NAC group compared to the FGS only group was marginally significant (0.059). The average total recurrent tumor weight in FGS only was significantly less than BLS only (p?=?0.037), and that in FGS+NAC was also significantly less than BLS+NAC (p?=?0.004) (Figures 5 and ?and6).6). The recurrence rate of FGS+NAC was also significantly less than BLS+NAC (p?=?0.008). FGS+NAC significantly reduced the metastatic recurrence frequency to one of 8 mice compared to FGS only where metastasis recurred in 6 out of 8 mice and BLS+NAC where it occurred in 7 out of 8 mice (p?=?0.041) (Table 1).
In a previous study, we conjugated a monoclonal antibody specific for the tumor-associated antigen CA19-9 with the AlexaFluor 488 green fluorophore. We were able to demonstrate in vivo binding of the antibody fluorophore conjugate to the tumor tissue in an orthotopic mouse model of human pancreatic cancer . This fluorescence facilitated differentiation between normal and tumor tissue within the pancreas and also revealed microscopic foci or tumor implants within the spleen, liver, and peritoneum which were not visible under standard light microscopy. This study offered a novel technique to facilitate the intraoperative identification of both primary tumor and small metastatic lesions that may be missed at the time of surgery in those patients whose tumors express the tumor-associated antigen CA19-9.
In another study, we compared a hand-held imaging system with larger imaging systems previously used for FGS . In a PDOX model labeled with Alexa Fluor 488-conjugated anti-CA 19-9 antibody, only the portable hand-held device could distinguish the residual tumor from the background, and complete resection of the residual tumor was achieved under fluorescence navigation, suggesting this system can be applied to the clinic in the near future to enable widespread application of FGS.
There are several novel aspects to the present study that should be emphasized. To the best of our knowledge, this is the first study that has utilized the combination of NAC and a CA 19-9 antibody conjugated fluorophore for FGS of pancreatic cancer. Furthermore, the present study took advantage of a longer wavelength dye, DyLight 650, which we have previous shown has better tissue penetration compared to AlexaFlour 488 . In addition, the PDOX model developed in our laboratory, and used in the present study, allows for individualized therapy that is not available with pancreatic cancer cell line models –,. PDOX models can be helpful to determine if an individual’s tumor is sensitive to various NAC regimens. The most novel and unexpected finding was that FGS+NAC eliminated pancreatic cancer metastases in seven out of eight mice.
For bright light surgery, tumors were removed with grossly negative margins under standard bright-field using an MVX10 microscope. For fluorescence-guided surgery, tumor resection was guided by labeling the tumors with an anti-CA 19-9 antibody labeled with a 650 nm fluorophore. The pancreatic cancer PDOX used in this study had a very aggressive behavior. At FGS, we detected some tiny tumors spreading around the primary tumors, which could not be detected under normal macroscopic inspection. At the first surgery, the surgical margin was exterior to the tumor border which was recognized macroscopically. However, a larger margin provided by FGS appears insufficient to lower or prevent metastatic recurrence which required NAC in addition to FGS (Table 1). However, the larger margins afforded by FGS are necessary to lower or prevent metastatic recurrence, as the combination of BLS and NAC are ineffective to lower or prevent metastatic recurrence (Table 1).
All mice in this study were euthanized 90 days after BLS or FGS and therefore, we were not able to compare survival differences between the groups. However, as seen in Table 1, the metastatic recurrence rate in FGS+NAC was significantly less than FGS only (p?=?0.041), suggesting that FGS+NAC improves the survival of pancreatic cancer patients compared to FGS only.
In summary, we have determined the efficacy of NAC with GEM in combination with FGS on a pancreatic cancer PDOX model. The results from this study indicate that NAC in combination with FGS can reduce or even eliminate metastatic recurrence of pancreatic cancer sensitive to NAC. This is an important result for the future more effective treatment of pancreatic cancer. The present study further emphasizes the power of the PDOX model which enables metastasis to occur and thereby identify the efficacy of NAC on metastatic recurrence.
This study was supported in part by National Cancer Institute grants CA132971 and 142669 (to MB and AntiCancer, Inc.); and Japanese Ministry of Education, Culture, Sports, Science and Technology for Fundamental Research Grant Numbers 26830081 to YH; and 26462070 to IE and 24592009 to KT. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Laser vaccine adjuvants
Immunologic adjuvants are essential for current vaccines to maximize their efficacy. Unfortunately, few have been found to be sufficiently effective and safe for regulatory authorities to permit their use in vaccines for humans and none have been approved for use with intradermal vaccines. The development of new adjuvants with the potential to be both efficacious and safe constitutes a significant need in modern vaccine practice. The use of non-damaging laser light represents a markedly different approach to enhancing immune responses to a vaccine antigen, particularly with intradermal vaccination. This approach, which was initially explored in Russia and further developed in the US, appears to significantly improve responses to both prophylactic and therapeutic vaccines administered to the laser-exposed tissue, particularly the skin. Although different types of lasers have been used for this purpose and the precise molecular mechanism(s) of action remain unknown, several approaches appear to modulate dendritic cell trafficking and/or activation at the irradiation site via the release of specific signaling molecules from epithelial cells. The most recent study, performed by the authors of this review, utilized a continuous wave near-infrared laser that may open the path for the development of a safe, effective, low-cost, simple-to-use laser vaccine adjuvant that could be used in lieu of conventional adjuvants, particularly with intradermal vaccines. In this review, we summarize the initial Russian studies that have given rise to this approach and comment upon recent advances in the use of non-tissue damaging lasers as novel physical adjuvants for vaccines.
Potential Role of Lasers as Vaccine Adjuvants
The challenge of adjuvant development for intradermal vaccines
Immunological adjuvants, (from the latin adjuvare, meaning to help), effect qualitative and quantitative changes in immune responses to a simultaneously administered vaccine antigen that result in sufficient immunological memory and protection against pathogens.1 Contemporary clinical vaccines generally use highly targeted recombinant molecules as antigens that are often poorly immunogenic in their own right and therefore require enhancement with immunologic adjuvants.2–6 Development of safe and potent immunologic adjuvants therefore represents an important element of current and future vaccine development.2,3 The immune potentiating effects of adjuvants are often accompanied by an increased risk of local reactogenicity or systemic toxicity. As a result, in spite of a steady proliferation of potential adjuvant candidates, the number of adjuvants that have actually been approved for use with human vaccines by regulatory agencies such as the FDA is surprisingly limited. Until the recent approval of AS04 (alum with monophosphoryl lipid A) for use in Cervarix® (GlaxoSmithKline) and AS03 (squalene-based oil-in-water emulsion) for Q-pan H5N1 influenza vaccine (GlaxoSmithKline), the FDA had only approved particulate aluminum-salt adjuvants for use with vaccines, mainly due to concerns over side effects. The European Medicines Agency (EMEA) has approved only 5.7–12
The tradeoff between efficacy and safety is evident in recent global experiences with influenza vaccines. Comparisons of adjuvanted to unadjuvanted vaccines in different populations consistently show that rates of seroconversion with adjuvanted vaccines is higher than with unadjuvanted vaccines, but that rates of injection site reactogenicity are also higher with the former.13–15 After the release of vaccines for H1N1 in 2009, the AS03-adjuvanted influenza vaccine Pandemrix® (GlaxoSmithKline) was linked to hundreds of cases of narcolepsy in the EU.16–19 In the US, where only unadjuvanted H1N1 vaccines were approved, no cases were reported.20 Since a potential link has been made between H1N1 epitopes and autoimmune narcolepsy,21 the AS03 adjuvant may have contributed to this adverse effect. The development of new vaccine adjuvants with improved safety profiles is a highly desirable factor in vaccine development.
The challenge of adjuvant development is increased when it comes to intradermal vaccines. There has been a growing interest in targeting vaccines to the skin due to the potential for the skin-based immune system to yield protective immune responses with smaller amounts of vaccine antigen.22–24 Vaccination at this epithelial surface can effectively prime the body to respond to pathogens22,23,25–28 and induce a robust recall immune response.29–31 The intradermal route of vaccination appears to be more effective than the conventional intramuscular route for many vaccines including influenza vaccines22,23,26,28 and hepatitis B vaccines.22–24 An intradermal influenza vaccine Fluzone Intradermal® (Sanofi Pasteur) was approved by the FDA in 2011 and the company has licensed such a vaccine in more than 40 countries.32
In spite of its promise, several distinct challenges have kept intradermal delivery from becoming a standard of practice in vaccination. One of these is the paucity of appropriate adjuvants.33,34 Most adjuvants used in licensed intramuscular vaccines like aluminum salt and oil-in-water adjuvants are simply too reactogenic when delivered intradermally.35,36 While some clinical trials have suggested that alum-adjuvanted vaccines are tolerable when given intradermally,37,38 reports of injection-site reactions with intradermally-delivered adjuvanted vaccines are much higher compared with non-adjuvanted formulations.39 Finally, current adjuvants may not be compatible with intradermal delivery or formulation requirements, especially with newer intradermal vaccination technologies.33,34 Not surprisingly, there is no adjuvanted intradermal vaccine licensed to date.
The use of non-destructive lasers to adjuvant vaccines
A number of studies conducted over the last 2 decades suggest that non-tissue damaging lasers can be used to modify local skin immune responses in a way that enhances systemic immune response to a vaccine introduced into the treated skin. Treating the skin with non-harmful laser light represents an adjuvanting approach that is potentially compatible with intradermal vaccination. This review focuses on the historical development, current status, and future prospect of lasers that do not breach or destroy skin tissue for this purpose. It is well established that skin injury such as scarification or burn can enhance immune responses40and lasers can be used to induce such injuries to the skin.39,41–43 A number of investigators have explored the immune-stimulating effects of thermally-destructive lasers in treating cancer44,45 and a recent review has been published covering this approach.46 In addition, fractional laser devices used to enhance delivery of drugs and vaccines have been examined for their ability to alter immunologic responses in the skin to vaccines and immunotherapies.47 These lasers breach the tissue barrier by ablating tissue in small cylindrical volumes, creating a field of skin micropores whose diameter, depth, and density can be highly controlled. The tissue damage that results from the process of microporation can activate the immune system through release of damage-associated molecular patterns (DAMPs) from coagulated tissues, leading to expression of pro-inflammatory cytokines and activation of antigen presenting cells.47 This effect is enhanced by the cytokine cascades induced by the loss of cutaneous barrier integrity.48,49 The use of fractional lasers for vaccination has also been recently reviewed.50
The use of non-destructive lasers to alter tissue immune responses in a manner that can enhance systemic vaccine responses, (laser vaccine adjuvants or LVAs), is a novel approach that has just begun to be explored. LVA treatment of the skin is characterized by a combination of relatively low power densities or irradiances (0.7 to 6.0 W/cm2) combined with fairly high total fluences (energy dose supplied per unit area)—typically in the hundreds of Joules/cm2—a combination that is unique compared with most other clinical applications of lasers to the skin35,51–53 (summarized in Fig. 1). LVA exposures generate moderate but non-damaging thermal responses in the tissue that are quite distinct from “athermal” low-level laser therapy (LLLT), where both laser irradiances and fluences are typically 1–2 logs lower.54,55 The fluences applied with LVAs are well beyond the range where most simulative biological responses to LLLT have been identified. Such lasers also operate at irradiances about one log greater than high fluence, low power laser treatments used to induce apoptotic effects in a variety of cancer cell lines both in vitro and in vivo.56–59 With recently reported progress on a new type of LVAs, the combination of LVAs with skin-based vaccination now has the potential to yield more effective vaccine responses in a safe and cost-effective manner.
History of Laser Vaccine Adjuvants
Initial Russian studies
The concept of using non-destructive lasers as vaccine adjuvants evolved from several decades of laser research in Russia, some of which was transferred to US laboratories over the decade. The origin of this work goes back to Russian investigations of photobiomodulation that began soon after the first descriptions of the effects of low-power laser on biological processes by Endre Mester in Hungary in 1967.60 One area of clinical application that attracted early medical interest was the promotion of wound healing in various tissues.61,62 Many of these early wound healing studies utilized low-power helium-neon lasers.63–66 In 1978, the copper vapor laser was clinically introduced in Russia. Unlike the continuous laser light emitted by helium-neon lasers, copper vapor (CV) lasers release very short duration pulses (10–25 ns) at very high repetition rates (5–20 kHz). CV laser light is emitted at 2 wavelengths—about 510 and 580 nm—representing the yellow-green part of the visible light spectrum. The distinctive laser emission characteristics of copper vapor lasers were of interest to Russian physicians to explore a variety of clinical applications, including wound healing.67 These explorations also gained impetus from medical research funded by the Soviet military during the Afghan conflict in the early 1980s.
While early Russian studies of CV lasers for wound healing featured either low-energy, athermal treatments at doses typical of LLLT, or high-power treatments to induce tissue coagulation,68,69 a small subset of studies in the 1980s and 1990s, many led by Dr Anatoly I Soldatov of the St. Petersburg Academy of Postgraduate Medical Education, utilized much higher irradiances and doses that induced significant photothermal and photoacoustic responses in the irradiated tissue but did not cause tissue damage typical of high power lasers. Some of these studies showed that this new type of treatment effectively promoted wound healing for specific chronic medical conditions such as gastric and duodenal ulcers, bronchitis, and bronchial hyperplasia.70–72
A group of St. Petersburg investigators, led by Dr Sergei Onikienko of the St. Petersburg Military Medical Academy, hypothesized that this higher-power but non-destructive type of laser treatment might also affect immune responses in healthy tissues. These scientists began to explore the potential of CV laser to improve vaccine responses and conducted a number of studies in mice and humans in the late 1990s and early 2000s that used CV lasers to enhance responses to both prophylactic and therapeutic vaccines. These studies, the majority of which were published in Russian, non-peer reviewed journals, showed that CV laser treatment of the skin improved responses to intradermal delivery of commercial influenza and hepatitis B vaccines in documented vaccine non-responders, and also potentiated the effects of experimental therapeutic vaccines for chronic hepatitis B and cancer.
Recent US studies
In 2004, a Massachusetts General Hospital team conducting scientific assessments as part of the BioIndustry Initiative of the US State Department—a program within the cooperative nonproliferation efforts between the US and Russia73—began to meet with these investigators, explore their approach and introduce them to US investigators for potential collaborations. In 2008, a US biotechnology company initiated an effort to replicate these earlier Russian studies using a more structured approach. The first of these studies, published by the Wellman Center for Photomedicine in 2010 provided support for the Russian preclinical results by showing that a 532 nm nanosecond pulsed Nd:YAG laser could enhance antibody titer responses to a model ovalbumin vaccine and a split-virion influenza vaccine.53 Chen et al. subsequently showed that this type of laser could enhance immune responses to a nicotine vaccine35,53 and a dendritic cell vaccine.74
In parallel studies conducted at the laboratories of the Vaccine and Immunotherapy Center (VIC) at Massachusetts General Hospital, Kashiwagi et al. showed that a Nd:YVO4 laser emitting either nanosecond pulsed light at 532 nm or continuous wave, near-infrared light at 1064 nm could enhance immune responses to a model vaccine (ovalbumin, OVA) and to a live attenuated influenza vaccine. Surprisingly, the 1064 nm laser provided superior efficacy to the 532 nm laser in a lethal challenge study.51 The responses to the 532 nm laser were anticipated based on the earlier work done in Russia and at Wellman, but the responses to the 1064 nm continuous wave laser were quite unanticipated and represent a promising avenue of exploration for this approach. Taken together, the research conducted by Russian and US scientists suggests that non-destructive lasers have the potential to enhance vaccine responses and are worthy of further exploration. In addition, the recent discovery by Kashiwagi et al. supports the view that potent vaccine responses can be induced by relatively simple, low power laser systems. This finding enhances the potential and clinical applicability for the commercial development of LVAs.
Immunologic Effects of Laser Vaccine Adjuvants
Systemic effects on vaccine responses
Three different research groups to date have shown systemic vaccine enhancement in response to LVA treatment. Table 1 summarizes the laser types and treatment parameters used by these different groups. The initial Russian studies on LVAs, summarized in Tables 2 and ?and3,3, provide a rationale to further pursue this approach but for the most part lack adequate descriptions of methods and are published largely in non-peer review publications. In these studies, Onikienko and his colleagues used a copper vapor (CV) laser (D.V. Efremov Institute of Electrophysical Apparatus) emitting light at both 510 and 578 nm (mix of 10% 510 nm and 90% 578 nm) with a pulse width of 10–12 ns and a pulse frequency of 10–20 kHz. The laser beam had a flat top profile. Irradiances were typically between 1–6 W/cm2 and skin exposures were typically 5 mm spots.
Vaccine response studies were performed in both mice and humans. A vaccination study in white mongrel mice involved a single 1- or 2-min treatment of the ear skin at an irradiance of 1–3 W/cm2 followed by an intradermal injection of 50 μL of a split inactivated influenza vaccine (Vaxigrip, Sanofi Pasteur). CV laser treatment of the skin resulted in a 54% to 86% increase in anti-influenza antibody titer at 4 wk compared with vaccine alone.75,77 Induction of cell-mediated immune response was also shown using the leukocyte migration inhibition test (LMIT).78 In a follow up study, protective immunity was examined by exposure of 2 groups of 20 of the same breed of mice to a lethal dose of H3N2 influenza (?/Aichi/2/68) via an inhalation route 14 d after vaccination. 70% of the mice receiving vaccination and laser treatment survived compared with 35% of mice receiving vaccine only.75,77
Onikienko’s team extended the vaccination approach into clinical applications with prophylactic vaccines. In one study, 42 people with documented low antibody titer responses to influenza vaccination were intradermally vaccinated with a 15 μg dose of Vaxigrip via jet injector; 22 of these received skin site exposure to CV laser at an irradiance of 1.0 W/cm2 for 2 min (120 J/cm2 total dose) right before vaccination.77 Blood was drawn from each vaccine at 4 wk and a number of immune end points, including anti-influenza antibody titration, LMIT, lymphocyte cytotoxic activity, monocyte cytokine secretory activity, and the increase in activity of the lymphocyte enzymes, were examined and compared with 22 healthy controls who were also vaccinated. Based on assessment of these assays, it was determined that 14 of the 22 laser treated non-responder subjects showed a statistically significant increase in vaccine responses compared with only 5 of 20 non-responder subjects in the vaccine-only group. In the healthy control group, significant responses were measured in 20 out of 24 subjects.
A similar study was performed in 17 people documented as hepatitis B vaccine non-responders (failure to maintain HBsAb antibody titer of greater than 10 mIU/mL 6 mo after completing the course of vaccination).77 All subjects received a course of 3 intradermal injections (0, 1, and 3 mo) of 20 µg of a recombinant hepatitis B vaccine (Combiotech) using the Mantoux technique. Nine of the subjects received a 3 min treatment of a 5 mm spot on the shoulder skin with a CV laser at 1 W/cm2 average power (180 J/cm2total dose) prior to each vaccination. Eight other subjects received hepatitis B vaccination with concomitant IL-2 injections (2 500 000 IE via subcutaneous injection) in a manner similar to Jungers et al.79 In the laser treated group, 7 out of 9 subjects reached the international standard for protection at 6 mo after the end of vaccination (10 IU/mL), while none of the IL-2 treated subjects did so.77
The St. Petersburg group also applied the laser approach to therapeutic vaccination. In one study, the laser was used to enhance responses to an investigational vaccine (recombinant HBsAg without alum, Combiotech) combined with laser treatment of the injection site. Subjects diagnosed with chronic hepatitis B for at least 2 y received either a series of 12 weekly intradermal vaccinations with or without CV laser skin pretreatment (1.0–1.5 W/cm2 on a 5 mm skin spot for 1–3 min). A control group received Lamivudine 100 mg daily for 12 wk. The effect of the treatment was evaluated by clinical indicators of disease including liver function tests, circulating HBV DNA by PCR and serum HBsAg, and immune responses using LMIT with HBsAg to measure cell-mediated immune response. At 12 wk, 5 of 9 of the laser-pretreated vaccines showed normalized ALT, HBV DNA copies below 300, and positive LMIT compared with 4 out of 11 of the Lamivudine-treated group.
The Wellman group’s initial studies at MGH similarly used a nanosecond pulsed laser operating in the green spectrum. This was a Q-switched neodymium-doped yttrium aluminum garnet (Q-Nd:YAG) laser emitting light at 532 nm with 6–7 ns pulse widths and repetition rate of 10 Hz. Their initial studies used 2 min exposures to 4 separate 6 mm spots of skin on BALB/c mice at an irradiance of 0.78 W/cm2 followed by intradermal injection of ovalbumin or inactivated influenza vaccines in each irradiated spot. The illumination of skin with the laser increased the motility of antigen presenting cells (APCs), leading to enhanced antigen uptake by APCs and helper T cell priming in the draining lymph nodes. This 2-min laser exposure increased humoral immune responses to a model vaccine (OVA) by 300 to 500% and a split-virion influenza vaccine (Fluvirin) by 400% in primary vaccination and 900% in booster vaccination compared with a non-adjuvanted group.53
In the most recently published LVA studies, the VIC group of Kashiwagi et al. at MGH used a neodymium-doped yttrium orthovanadate laser emitting light either at 532 nm in high frequency Q-switched mode (Q-Nd:YVO4) with an irradiance of 1.0 W/cm2, a pulse duration of 10 ns and a pulse repetition rate of 10 kHz, or in a continuous wave mode at the near-infrared spectrum at 1064 nm (CW-NIR) at an irradiance of 5 W/cm2. The study showed that the CW-NIR laser adjuvant induces the transient expression of a limited set of cytokines and chemokines in skin resulting in recruitment and activation of dendritic cells in skin draining lymph nodes (dLNs). Furthermore, a 1-min application of the CW-NIR laser augmented antibody response most efficiently to OVA and an influenza vaccine (whole inactivated PR8 virus) with a TH1-TH2 balanced T cell response, and conferred protection in a murine influenza lethal challenge model, whereas the 532 nm Q-Nd:YVO4 induced a TH1-skewed response with little impact on protection.51 Importantly, the protective immune responses induced by the CW-NIR were comparable to those induced by a licensed adjuvant and support the view that LVA might have utility in augmenting responses to intradermal vaccines.
Localized effects on irradiated tissues
In general, LVAs appear to work by modifying the immunologic environment within the tissue that receives the vaccine, resulting in enhancement of the vaccine response. The specific modifications in the local immune environment appear to differ depending on the type of laser used and are likely related to significant differences in the laser wavelength, pulse duration, pulse energy, and pulse frequency.
Effect on heat shock protein expression and release
A fundamental principle of vaccine adjuvant development, based on Matzinger’s danger theory of immune response80 is to trigger a danger signal to the immune system that can promote more vigorous and long-lasting responses to a vaccine antigen. These danger signals are often in the form of either DAMPs or pathogen-associated molecular patterns (PAMPs) that can trigger cytokine and chemokine cascades in the tissues, usually through Toll-like receptors (TLRs).81 Significant development work was put into developing adjuvants in the form of DAMPs or PAMPs that can trigger these TLR-mediated pathways to promote vaccine responses.7
Onikienko et al. identified an important role for heat shock protein 70 (HSP70) in mediating vaccine responses to CV laser treatment. HSPs are a family of ubiquitous intracellular molecules that function as molecular chaperones as part of numerous intracellular processes (e.g., protein folding and transport). Under conditions of stress, some of these play important roles in refolding or disposing of misfolded and denatured proteins,82 stabilizing cellular membranes and enhancing cell signaling,83 and inhibiting specific apoptotic pathways.84 The intracellular expression of many of these HSPs are significantly induced under stress conditions such as fever, radiation, infections, and neoplasia.85 Some HSPs, such as HSP70, play additional roles if and when released outside the cell. In these circumstances they can act as potent DAMP-like inducers of immunity and have been harnessed as adjuvants in experimental vaccines targeted to cancers and infections.86 HSPs expressed on the surface of stressed and damaged cells or released from necrotic cells can serve as a kind of danger signal87 and are recognized by APCs through specific receptors, such as TLRs, scavenger receptors (LOX-1), CD91, and CD14 resulting in increased antigen display by MHC class I and II molecules and priming T cells.86,88
Onikienko’s group showed that the CV laser treatment (irradiance of 1–3 W/cm2 with a pulse width of 10–12 ns and pulse repetition frequency of 10–20 kHz) to a 5 mm skin spot on a mouse ear for 3 min induced rapid, dose-related increases in extracellular HSP70 as determined by a whole-mount in vivo immunostaining of epidermal sheet of the mouse ear.52 Western blot analysis on epidermal tissue further showed an increased expression of HSP70 in the ear skin that persisted for 7–14 d.52 Since adjuvant effects in these mice similar to those caused by the CV laser could be induced by injection of exogenous by itself, Onnikienko et al. concluded that high-frequency pulsed laser treatment enhances immune responses via release and sustained expression of HSP70 by fibroblast and/or keratinocytes in the laser irradiated skin.
Cells in the skin harbor a high baseline level of intracellular HSP70 that could potentially be released under conditions of stress89 with the highest levels found in keratinocytes.90 While overexpression of heat shock proteins is a normal cellular response to stress, a number of investigators have shown that under conditions of stress a portion of constituent HSP70 is mobilized to the cell membrane and can be released from the cell through a variety of mechanisms.91–93 Subsequent LVA studies performed in the US and published to date have not linked LVA skin treatment to the release or overexpression of HSP70.51,53 This difference in expression and release profiles for HSP70 may be related to the differences in the laser parameters used by the different laboratories.
Effect on immune cell migration
The use of chemical adjuvants in skin-based vaccination studies directly activate and induce migration of APCs from the skin to the proximal dLN.94,95 LVAs similarly induce APC migration to the skin and dLNs, increasing the concentration of APCs in volume of treated tissue and enhancing their ability to activate, pick up antigen, and migrate to dLNs. The means by which they accomplish these effects may be different depending on the laser type.
Effects on inflammatory and chemokine signaling
With chemical and biological adjuvants, the activation and mobilization of APCs in the skin is a result of both autocrine and paracrine signaling through cytokines and chemokines.96 DCs are the most versatile APCs; licensed and experimental adjuvants activate DC-mediated innate immune responses that result in robust adaptive immune responses.7–9,97,98 Intradermal administration of adjuvants typically induces inflammatory responses including cytokine release and leukocyte infiltration.99These adjuvants are effective also because the inflammatory responses mediated by chemical adjuvant lasts for several weeks.99 Unfortunately, this persistence of inflammatory signaling may also play a role in diminishing the safety of these adjuvants.12,100
LVAs appear to function quite distinctly from chemical adjuvants in that they result in tissue signaling and the activation of APCs, but do not appear to trigger significant inflammatory responses. Chen et al. reported that the activation and migration of APCs in the skin following 532 nm Q-Nd:YAG laser treatment was not accompanied by a significant increase in inflammatory cytokines including TNF-α, IL-1β, IL-6, and CCL2.36,53 The activation picture for APCs following exposure to the Q-Nd:YAG laser-treated skin was mixed. While the overall number of DCs migrating to skin-draining LNs was significantly increased by 24 h and activation markers including CD80 and MHC class I were upregulated in skin-dLNs, critical activation markers such as CD40 and MHC class II were not upregulated.74
Kashiwagi et al. showed that the CW 1064 nm laser treatment did not result in the significant expression of proinflammatory cytokine genes such as Il1b, Il6, and Tnf and that, without introduction of a vaccine, cytokine responses and expression return to basal levels by 24 h after laser illumination.51 Nevertheless, CW-NIR laser treatment results in CD11c+ dendritic cell recruitment into the irradiated skin, increases the expression of MHC class II and co-stimulatory molecules including CD40 and CD86 on these cells, and increases the number of activated DCs in skin-dLNs 24 h after the laser irradiation.51 Since the magnitude of the CD4+ T cell responses is proportional to the number and quality of DCs that reach the dLNs,101 it is not surprising that the CW-NIR laser adjuvant enhances CD4+ T cell responses. These qualitative and quantitative DC responses correlate to the transient expression (measured 6 h after the CW-NIR laser treatment) of a set of chemokine genes including Ccl2, Ccl6, Ccl11, Ccl17, Ccl20, and Ccr7 that mediate DC migration96,102–105 (Fig. 2). The migration of mature DCs to dLNs through afferent lymphatic vessels is regulated by multiple cytokines and chemokines.101,106 The transient tissue response to the CW-NIR laser illumination results in expression of chemokines sufficient to initiate DC migration and maturation in situ but may not be sufficient to provide DCs with additional guidance cues including CCL19/21 expression in lymphatic endothelial cells. Further investigations of the effects of LVA on DC migration and/or activation are needed.
Mechanisms of Action for Laser Vaccine Adjuvants
Photobiological basis for laser adjuvant effects
A fundamental principle of laser medicine is that emitted photons must be absorbed in order to have a biological effect. The absorbers of laser light, called chromophores, have specificity for and sensitivity to particular wavelengths of light due to how specific wavelength photons interact with electrons within the molecular structure of the chromophore. As a result, tissues preferentially absorb some wavelengths of light over others, showing greater absorption efficiency at different wavelengths.54,55 Three key chromophores in the skin are melanin, hemoglobin and water. In the ultraviolet (UV) and visible spectrum, absorption by melanin and hemoglobin dominate. The effective absorption of melanin drops off quickly beyond 700 nm and ends at around 1100 nm. Hemoglobin has a high coefficient of absorption in the visible light range with a peak at 578 nm and also falls quickly beyond 700 nm; it plays a relatively insignificant role as a chromophore beyond about 1000 nm. Water has a much lower absorption coefficient of light compared with melanin and hemoglobin until about 1000 nm, but as it makes up almost 70% of the composition of the skin, it presents a large absorption target. Beyond 1000 nm, water becomes the dominant dermal chromophore. The relatively weak absorption of the 3 main skin chromophores between 700 and 1000 nm provides an “optical window” in the skin that permits laser light to penetrate much deeper (Fig. 3). This means that the 510/578 nm CV laser, 532 nm Q-Nd:YAG and 532 nm Q-Nd:YVO4 lasers will have a much shallower effective penetration depth (depth at which the intensity of the laser energy falls to 1/e or about 37% of the incident intensity) about 1 mm, as compared to CW-NIR laser, which will be about 4 mm.107 These differences in absorption efficiency in the skin account for most of the difference in emitted irradiance between visible light and NIR systems (i.e., 1 W/cm2 in the Q-Nd:YVO4 system at 532 nm and 5 W/cm2 in the CW Nd:YVO4 system at 1064 nm).
Photothermal effects of adjuvanting lasers
All lasers that have been used as LVAs to date induce moderate thermal effects in the treated tissues, but irradiance and fluence are specifically calibrated to remain below the level that causes skin damage. Increases in skin temperature induced by these minute-range laser exposures do not reach the pain threshold (42–43 °C)51,53 much less than the temperatures known to induce skin damage within such exposure times.112–114The non-destructive nature of the LVA exposures used in the recent US mouse studies were validated by both visible skin inspection and independent analyses of biopsy for histopathological evidence of tissue damage.51,53 In addition, a clinical safety study, performed with a 1064 nm nanosecond pulse laser using irradiances and fluences equivalent to those used in the murine laser vaccine enhancement experiments, was well tolerated in humans with no subject reporting uncomfortable skin sensations or pain and no significant skin damage or changes in skin appearance noted during or as a result of any laser exposure.51
While overall increases in temperature in the laser exposed skin is modest, lasers featuring high power, nanosecond duration laser pulses are nevertheless likely to cause significant thermal stresses in the skin. This is due to a phenomenon called thermal containment. When the time over which a laser pulse is absorbed into a volume of the skin around a chromophore (tp) is significantly shorter than the time the resulting heat can be dissipated from that volume to the surrounding tissue (tr), the result is a significant increase in the temperature within that volume relative to the surrounding tissue.108 This condition is known as thermal containment and is considered to be met when tp ≈ 0.25tr. As the duration of a laser pulse decreases, the volume in which the thermal energy is contained also decreases. Thermal confinement is the basis of selective photothermolysis, a phenomenon first described by Anderson and Parrish,115–117 and is the key mechanism of many medical laser applications.
In the nanosecond range, thermal containment is expected to occur at the organelle scale (e.g., 0.5 to 1.0 μm in diameter). In the visible light range where melanin is an important absorber, such targets are typically melanosomes in the basal epidermis.118,119 Given the same pulse energy, shorter pulse durations will result in larger temperature rises within more highly localized volumes around the chromophore. Eventually, sufficiently large pulse energies or sufficiently short pulse durations will result in temperature rises large enough to induce transient protein unfolding or permanent denaturation.120,121 Once these localized temperature spikes significantly exceed 100 °C, the result will be microcavitation or explosive vaporization of the water in and around the target chromophore.122–124 This phase transition can cause significant damage to the tissue. Laser pulse durations in the nanosecond range with pulse powers in the kW or MW range like the CV, Q-Nd:YAG, and Q-Nd:YVO4 lasers, contain sufficient energy to induce significant temperature perturbations at the subcellular level (microhyperthermia), resulting in cellular stress from heat shock.125,126The propagation of high power (kW or MW) nanosecond-range pulses, repeated tens or thousands of times per second over a matter of minutes, likely triggers a number of stress responses in the skin that leads to enhanced immune processing of introduced vaccine antigens. As long as the combination of peak power, pulse duration and overall fluence are limited, these temperature spikes will not lead to significant irreversible damage to the tissue.
Aside from the tissue stress effects induced by highly localized generation of heat by nanosecond duration laser pulses, the modest overall increase of heat within the tissue does not appear to contribute significantly to the impact of the laser on the immune system. Chen et al. reported that skin heating did not result in a significant enhancement of immune responses.53 Kashiwagi et al. reported no correlation between measured maximal skin temperature and antibody titer in a model vaccine experiment.51
Photoacoustic effects of adjuvanting lasers
The photothermal effects generated by high energy, nanosecond pulsed lasers are accompanied by photoacoustic effects. The significant temperature discontinuities that high power, nanosecond duration laser pulses create between the absorbing target and the surrounding tissue results in different rates of thermal expansion and thus pressure differences.127,128 These pressure differences can generate an acoustic wave that propagates at a much slower rate than that of heat dissipation. When the laser pulse duration (tp) is shorter than the time required for these stress waves to propagate (tσ), a condition of acoustic containment is reached. At pulse durations below 100 ns, significant acoustic waves are generated in the tissue.129 The combination of microhyperthermia and shock wave generation from nanosecond pulse lasers can induce significant stress within cells and tissues, even when no significant damage is apparent.130–132 Not surprisingly, Chen et al. reported that laser treatment of mouse skin with 6–7 ns pulses with a peak power of about 5 MW at a frequency of 10 Hz over 2 min resulted in the disruption of the dense protein network in the skin tissue, resulting in disconnected tissue with collagen fibers, as determined by electron microscopy, even while the overall tissue appearance remained normal.74 Photoacoustic effects would be expected to be reduced in the pulse lasers with much lower peak powers, which may explain why such tissue changes were not reported by Kashiwagi et al.51 Finally, these effects are not expected with the CW-NIR laser. In this case, photochemical effects are more likely to be the mechanism.
One of the key photochemical effects of laser light is the generation of reactive oxygen and reactive nitrogen species (ROS and RNS).133 Generation of oxygen and nitrogen radicals from lasers have been demonstrated at a wide range of the spectrum including UV,134 blue,135 visible,136 and near-infrared.54,55,137–140Generation of these radical species form the basis for a wide variety of biological effects. Since ROS and RNS (particularly nitric oxide, NO) have been shown to stimulate cytokine production in epithelial cells via activation of MAPKs (p38, ERK), JNKs, NF-κB, AP-1, soluble guanylate cyclase (sGC)/protein kinase G (PKG)141–145 and recently NLRP3 inflammasome pathways,146 it is possible that laser adjuvants can mediate activation (phosphorylation) of these pathways via ROS generation. Endogenous or exogenous ROS and NO have been shown to modulate the function of skin cells including keratinocytes147–150 and mast cells.151–153
The results of published LVA studies to date clearly demonstrate that these lasers result in migrational and functional changes of DCs in the skin, which may be a common pathway for laser immune enhancement. To this end, different types of lasers appear to engage distinct molecular mechanisms. Photothermal and photoacoustic stress may play a more important role in immune signaling induced by nanosecond pulsed lasers, while photobiomodulation may play a larger role in the CW-NIR laser (Fig. 4). Elucidation of the exact photoreceptors and signaling pathways that mediate these effects is warranted.
Advantages and Limitations of Laser Vaccine Adjuvants
Advantages of LVA over conventional adjuvants
As an adjuvanting approach, lasers have several inherent advantages over chemical and biological adjuvants. (1) Laser light does not persist in the tissue or excessively perpetuate immune signaling, so it reduces the potential for toxic adjuvant effects. (2) It does not appear to depend on conventional inflammatory pathways, similarly reducing the potential for adverse events. (3) LVAs do not cause reactogenicity at the immunization site, a common factor with nearly all current vaccine adjuvants. (4) It cannot induce anti-adjuvant antibodies since it is not a chemical or biological substance. (5) There is little risk of allergic responses from lasers. Millions of people have been treated with visible light and NIR lasers for tattoo removal, hair removal, skin tightening and regeneration. While there are several published reports of allergic response after Nd:YAG laser treatment in tattoo removal, these were essentially delayed hypersensitivity reactions against the tattoo ink in the skin.154,155 (6) LVAs are separate from the vaccine antigen and do not require formulation with the vaccine. Some adjuvants are difficult to combine with the vaccine antigen because co-formulation may cause instability in the vaccine formulation. (7) The laser has no special storage requirements such as cold-chain requirements. Finally, as work by Chen et al. has shown, the laser can be used with several other conventional adjuvants to enhance their effect.35,36
Distinctions of the continuous wave, near-infrared laser for LVA over other laser parameters
When all LVA approaches to date are considered, the CW-NIR system has some compelling advantages over the nanosecond pulsed visible light lasers used in other studies. First, the CW-NIR system is much less sensitive to differences in skin pigmentation. Laser light in the green-yellow spectrum is significantly absorbed by melanin, resulting in highly variable absorption of the same laser dose across different skin phototypes.156,157 In addition, the melanin distribution in different types of skin is not uniform.114 This means that the laser dosing in dark skin may be different than in lighter skin, requiring a more careful recalibration of dose for different skin phototypes. Under these conditions, it may prove more challenging to tolerably treat a dark-skinned recipient due to the high efficiency of light absorption. While Russian investigators conducted initial studies of CV lasers in humans (Table 3), most of their subjects were fairly light skinned. Compared with the yellow-green spectrum, at 1064 nm in the NIR spectrum the coefficient of absorption of melanin is nearly 10-fold less.156,157 Differences in thermal responses to 1064 nm laser treatment between very light skinned and very dark skinned recipients appear to be modest and these differences appear to be shaped by absorption of laser light by blood.158,159 Initial pilot studies in humans using volunteers with skin phototypes V and VI have already established the probability that CW-NIR doses used in the mouse models will be tolerable in humans of all skin tones.33,51 The reduced differences in absorbance across skin phototypes makes the CW-NIR laser ideal for clinical use.
Second, the use of a device emitting continuous, low power laser light represents an engineering advantage for the development of LVA devices. Generation of nanosecond, high peak power laser pulses require technical specifications that may limit the compactness, simplicity and cost of devices to be deployed in the clinic or field. The generation of high frequency, high power pulses can be variable over the lifetime of the laser and the functional lifespan of a Q-switch diode laser is around 10 000 h. In contrast, CW 1064 nm lasers can be produced by very small and economic diode systems and have lifespans that exceed 100 000 h. Based on current technology capabilities, a handheld device for clinical use at under $1000 is feasible.
Limitations of LVA compared with conventional adjuvants
One key limitation in the use of LVA for potentiating vaccines is that their effects occur at a relatively shallow depth within exposed skin and require vaccination to occur intradermally in order to potentiate vaccine responses. This limitation may prove less troublesome as new technologies for intradermal and transcutaneous vaccines further develop.33,160
Another potential limitation of LVA pertains to the inherent limits of the immune response potential in treated tissue. LVA treatment alters DC responses to vaccine antigens based on transient light adjuvant exposure and without induction of significant inflammation. It remains to be seen whether this provides clinically meaningful increases in systemic immunity. Reported results of Russian experiments with the CV laser in human vaccine recipients are certainly provocative (Table 3) and Kashiwagi et al. have shown that the CW NIR laser outperforms alum in intradermal vaccination with model ovalbumin and influenza virus in the mouse model.51 Longer term, well-controlled studies are obviously needed. Further investigation will also be required to understand the optimal dose configurations for each type of system and to conduct comparative studies with other approved adjuvants.
Finally, the lack of appropriate laser devices for inducing vaccine responses is a key limitation in advancing the field. To date, laser devices used in studies of LVA have been either research prototypes or laser systems typically applied to cosmetic dermatology or manufacturing. While the treatment parameters needed to enhance vaccine responses can be produced by some of the existing clinical dermatologic laser systems if they are modified to generate significantly lower power levels and irradiances, such devices are relatively expensive, with the simplest devices typically costing more than $25 000, while more sophisticated devices can top $100 000. Lasers that are expensive and need extensive technical knowledge to operate and maintain reduce the opportunity for investigators working on vaccines, who typically do not have expertise in lasers, to test the potential of such an approach in many different vaccines. The development of a simple, low-cost, portable device would be essential for more rapid preclinical development. Ultimately, the use of LVA for mass vaccination will require simple, small, and economic systems that allow for repeated usage with little need for training or extensive maintenance.
Future Applications of Laser Vaccine Adjuvants
Research performed to date clearly underlines the promise of laser vaccine adjuvants. The technology is aligned with the long-term objectives of governmental and non-governmental organizations to reduce or eliminate chemical adjuvants whenever possible and promote the development of needleless vaccination approaches.34,161 LVA opens a new pathway toward those 2 important goals.
Use of LVA for prophylactic vaccines
In the near term, the most likely application for LVAs is to enhance intradermal vaccination. Clinical trials have been conducted on several different intradermal vaccines including influenza, HAV, HBV, polio, measles, and rabies.33,34 Some traditional vaccines such as the BCG (Bacille de Calmette et Guérin) and the smallpox vaccine are routinely given intradermally. BCG has an established track record as a safe vaccine and has been given to over 3 billion individuals, making it the most widely administered vaccine to date.162Intradermal influenza vaccines are now available for clinical use.32 As a physically separate adjuvant, LVA represents a broadly-applicable technology for this approach to vaccine delivery. LVA could be easily combined with approved vaccines and experimental candidates, and could be developed into single devices for delivery of adjuvant and vaccine. Since, the efficacy of inactivated influenza vaccine in children and elderly has been suboptimal without adjuvant163 and BCG produces highly variable levels of protection against MTB especially in preventing adult pulmonary TB.164–166 These vaccines would be attractive and feasible candidates for addition of LVA. In addition, LVA could be also used to augment established vaccines when they fail to achieve clinical significance in certain populations. For example, HBV vaccination is suboptimal in immunocompromised populations on dialysis and immunosuppressive therapy. LVA appears to be effective in enhancing intradermal influenza and HBV vaccines in non-responders to previous vaccinations,77 which justifies exploration of the use of LVA this case. Further determination of the efficacy of LVA for currently approved vaccines is an important research goal.
Use of LVA for therapeutic vaccines
The versatility of dendritic cells (DC) and their key role in regulating adaptive immunity167,168 has led to extensive investigation of DC-based vaccines for cancer and chronic infectious diseases.168,169 Activation of cognate T cells in the dLN is also necessary.170 For such vaccines, the intradermal delivery route offers a way to generate superior T-cell induction.171 Some preliminary work has been done on the use of LVAs with DC vaccines. Onikienko et al. explored the use of DC vaccines loaded with laser-treated (photomodulated) antigens of hepatitis B virus (HBV) or tumor autoantigens for therapeutic HBV and cancer vaccinations.75 A recent study in the US showed that treatment of skin with 532 nm Q-Nd:YAG laser followed by a intradermal DC vaccine injection increased the efficacy of the cancer vaccine.74 In this study, the laser induced more efficient migration of intradermally injected DC vaccine into the skin-dLNs resulting in an increase of anti-tumor IFN-γ+CD8+ cytotoxic T cell responses and conferred a survival benefit in 4T1 breast tumor and B16F10 melanoma models in mice.74 Together, these data support the view that laser adjuvant may be able to augment and induce sustained anti-tumor immune response in the context of intradermal injection of autoantigen- or DC-based therapeutic cancer vaccines.
The use of a physical adjuvant for cancer vaccines has some distinct advantages over chemical or biological adjuvants. Unlike many cancer vaccine adjuvants, LVA provides a transient immune stimulation and does not produce a depot effect.51 Depot formation and the associated slow release of antigen is related to persistence of adjuvant in exposed tissues and induce a prolonged inflammatory cytokine response.172 This is a tried and true approach in prophylactic vaccines, but may not be optimal for cancer vaccines. A recent study shows that persisting vaccine depots could induce a prolonged inflammation at vaccination sites and reduce vaccine efficacy.173 In this study, persistent vaccine depots induced specific T cell sequestration, dysfunction and deletion at vaccination sites, while a non-persisting vaccine formulation shifted T cell localization toward tumors, inducing superior antitumor effects. The ability of LVAs to induce effective cancer vaccine responses without depot effect merits further exploration.
The clinical utility of vaccine adjuvants requires both safety and efficacy, with an overriding emphasis on safety for prophylactic vaccines.174 The scientific literature contains reports of several hundred adjuvants with efficacy profiles that could be better than most adjuvants that currently approved for use in licensed vaccines, but none of these are clinically relevant because of their unsatisfactory safety profiles.12,100 From this perspective, the potential of LVA for use in human vaccines is supported by evidence from the preclinical and clinical studies and a compelling safety profile. In the most recent published study, LVA doses equivalent to those shown to be immunogenic in mice were tolerable in humans and did not appear to cause any skin damage.51 As with any other adjuvant, the safety and efficacy of LVA will have to be demonstrated through further rigorous preclinical safety and toxicology studies, and ultimately through a series of clinical trials of the combination of a clinically approved vaccine with the laser adjuvant, but current evidence in mice and humans suggests efficacy without tissue damage or significant inflammatory changes. The use of LVAs would further enable the use of intradermal vaccines, which could also be delivered needle-free, as with patches or micro-needle, a tremendous advantage in resource-limited and epidemic situations. Additionally, the discovery of NIR laser adjuvant effects enables the development of low-cost, low-power, hand-held LVAs that could be used in such resource-limited situations.
Finally, the adjuvant effects of LVAs would permit additional antigen sparing, avoid chemical interactions with adjuvant, and could even potentially be applied to mucous membranes, where inflammatory adjuvants are unworkable. Finally, the ability of LVAs to activate dendritic cells and augment cell-mediated immunity, without reactogenicity or long pharmaceutical developmental time-lines, makes them ideal for inclusion in both prophylactic and therapeutic vaccines being developed for infectious agents ranging from influenza to dengue to HPV, as well as therapeutic vaccines targeting cancer, cell- or epitope-based.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
We thank Dr R Rox Anderson and Dr Apostolos Doukas (Masachusetts General Hospital), Dr Holger Schlüter (HIGHYAG Lasertechnologie GmbH), and Dr Gregory Altschuler (Dental Photonics) for their critical input into our understanding of lasers and their impact in medicine; we thank Dr Michael Hamblin (Masachusetts General Hospital) and Mr Eugene L.Q. Lee (Imperial College London) for their contributions to figures. This material is based upon work supported by Center for Integration of Medicine and Innovative Technology (CIMIT), Defense Advanced Research Projects Agency under Contract No. Space and Naval Warfare Systems Center Pacific Award N66001-10-1-2132, Bill and Melinda Gates Foundation (Grand Challenges Explorations OPP1046276), The Friends of VIC and National Institute of Allergy and Infectious Diseases grant (R01AI105131). Any opinions, findings and conclusions, or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the Defense Advanced Research Projects Agency. We also would like to acknowledge the additional support of the Edmund C. Lynch Jr. Cancer Fund.
|AS03 (4)||adjuvant system 03 (04)|
|APC||antigen presenting cell|
|BCG||Bacille de Calmette et Guérin|
|CTL||cytotoxic T lymphocyte|
|DAMP||damage-associated molecular pattern|
|dLN||draining lymph node|
|EMEA||European Medicines Agency|
|HA(B)V||hepatitis A (B) virus|
|HBsAg||hepatitis B surface antigen|
|HSP70||heat shock protein 70|
|LLLT||low-level laser therapy|
|LMIT||leukocyte migration inhibition test|
|LVA||laser vaccine adjuvants|
|NSCLC||non-small cell lung carcinoma|
|Nd:YVO4||neodymium-doped yttrium orthovanadate|
|Nd:YAG||neodymium-doped yttrium aluminum garnet|
|PAMP||pathogen-associated molecular pattern|
|RCC||renal cell carcinoma|
|RNS||reactive nitrogen species|
|ROS||reactive oxygen species|
Laser vaccine adjuvant for cutaneous immunization
The skin is an immunologically active tissue with an abundance of resident APCs, in marked contrast to the muscular tissue. Approximately 40% of the body’s APCs are located in the skin including epidermal Langerhans cells and dermal dendritic cells (DCs). There is also a thick network of capillary lymphatic vessels in the skin that direct the passage of antigens and antigen-captured DCs from the skin to the draining lymph nodes. The first vaccine, smallpox, was delivered by skin scarification nearly two centuries ago . A body of evidence has consistently shown that id. vaccination is superior to im. in the clinic. For instance, id. delivery of one fifth of the influenza vaccine dosage could induce the same level of protection as im. vaccination in humans [7,8]. id. administration of unadjuvanted influenza vaccine stimulated hemagglutination inhibition antibody titers at a level comparable to those induced by MF59-adjuvanted influenza vaccine injected intramuscularly in the elderly . In addition to influenza vaccines, Mikszta et al. showed that one dose of id. immunization of naked recombinant protective antigen (rPA) of Bacillus anthracis gave rise to a 60% seroconversion rate, whereas only 20% of mice generated a detectable antibody response after im. vaccination of alhydrogel-aduvanted rPA [10,11]. Likewise, in comparison with im. delivery, id. delivery provided a tenfold dose-sparing benefit for rabies or HBV vaccine in normal patients [12–14]. Furthermore, id. but not im. vaccination induced significant anti-HBV antibodies in nonresponsive hemodialysis patients .
Despite being an effective route of vaccination, current id. delivery techniques, such as the Mantoux method and bifurcated needles, are inconvenient and require specially trained personnel. Thus it is not practical for immunization of a large population in a short period of time by this route . This situation may soon be changed, thanks to the development of a novel microinjection system that is now in clinic trials for convenient id. immunization of influenza vaccines . Similar technologies such as jet injectors, microneedle and microprojection array patches have also been developed for convenient cutaneous immunization and to improve patient compliance [16–19].
Adjuvant for cutaneous vaccination
At present, the majority of vaccine adjuvants are developed and evaluated in im. immunizations. Different from the muscular tissue, the skin tends to develop severe local reactogenicity after id. immunization even in the absence of adjuvants, presumably attributable to the presence of a large number of resident immune cells and a high density of blood and lymphatic vessel networks in the skin [1–3,7–9,20]. The local reactions are also more readily visible in the skin than in the muscle. Importantly, the skin is the organ serving as a sensorial physical barrier between our body and the environment, and its integrity is crucial in fulfilling this task. Thus, adjuvants for cutaneous immunization must have a higher level of safety and induce less inflammation. Many adjuvants used in im. vaccinations cannot be used for cutaneous immunization.
The only US FDA-approved, widely used adjuvant over the past 80 years in the clinic has been aluminum salt-based adjuvant, referred to generically as ‘Alum’ . It is currently included in several child vaccines in the USA, such as hepatitis A/B, diphtheria–tetanus–pertussis, and Haemophilus influenza type b. In 2009, an AS04-adjuvanted recombinant human papillomavirus vaccine was approved by the US FDA to prevent cervical cancer in girls and young women . AS04 is a combinatorial adjuvant containing Alum and monophosphoryl lipid A (MPL), a low-toxicity derivative of lipopolysaccharide that activates Toll-like receptor (TLR)-4. MF59-adjuvanted seasonal influenza vaccine has been used for more than a decade in the elderly in Europe, but not in children, because a major component of MF59 adjuvant is a self substance named squalene that has the potential to induce autoantibodies after repeated use [23–25]. A dozen other adjuvants are approved by foreign authorities to be included or tested in human vaccines. These include the oil-in-water emulsion AS03, water-in-oil emulsions montanide ISA 51 and ISA 720, the TLR-7 agonist imiquimod (R837), the TLR-9 agonist unmethylated CpG oligonucleotides (CpG), saponin QS21 mixed with MPL in liposomes (AS01) or in squalene emulsion (AS02), immunostimulatory complexes, and particle formulation adjuvants, such as liposomes, virosomes and microspheres . All of these adjuvants are tested or used in im. immunizations; however, their safety for id. vaccination remains a concern.
We evaluated the local reaction of some of these adjuvants following id. injection [27–29]. As shown in Figure 1, R837, MPL/CpG, MPL/R837 and MPL/Alum induced the most severe local reactions with a lesion size of 4–6 mm in diameter, concurrent with skin ulceration (upper panels), which was persistent for weeks. Histological examination revealed a heavy infiltration of inflammatory cells into adjuvant-treated skin (lower panel). The strong and persistent reactions in the skin exclude them to be used as cutaneous vaccine adjuvants owing to potential breaching of the skin that would provide the opportunity for local and systemic infections with various microorganisms. Alum and montanide ISA 720 induced a similar lesion as described previously, with infiltration of inflammatory cells evoked by Alum greater than for montanide ISA 720, but both caused little skin ulceration. Deposition of Alum, montanide ISA 720 and MPL/Alum was readily visible at the injection site, appearing as an abscess owing to the white/silver color of the adjuvant and did not disappear for months. Although only two emulsion adjuvants are being tested, it is likely that other emulsion adjuvants (e.g., AS02, MF59 and AS03) may have similar depositions in the skin in light of their similar physiochemical properties. The unpleasant ‘abscess’ and persistent inflammation in the skin question the use of any antigen-depot adjuvants for cutaneous vaccination. In contrast to the adjuvants described earlier, MPL or CpG induced mild local reactions with a lesion size of only about 2 mm in diameter that was completely resolved within 2 weeks. MPL and CpG may thus be safe adjuvants for cutaneous vaccination.
In addition to the persistent inflammation, inclusion of emulsion-based adjuvants to a vaccine faces another challenge due to a small injection volume allowed for id. injection, typically 100 µl per site, in contrast to 0.5–1 ml for im. injection. The sticky solution is also probably difficult to inject through the novel microinjection system. Besides, microneedle array patches and needle-free vaccine patches made of dried vaccines are being developed in order to eliminate cold-chain storages. Addition of emulsion-based adjuvant to the dried vaccine may result in a loss of its adjuvanticity, since a specific physical status of the adjuvant, in particular, antigen-depot adjuvants, is crucial for the immune-enhancing effect of the adjuvant. Accordingly, an ideal adjuvant for cutaneous vaccination should be safe, easily injectable, with little inflammatory response.
Laser vaccine adjuvant
We developed a novel physical type of laser-based vaccine adjuvant capable of enhancing vaccine-induced immune responses without direct contact with the antigen . In brief, a small area (<1 cm2) of mouse skin had hair removed and was exposed to a Q-switched 532-nm Nd:YAG laser (Spectra-Physics Inc., CA, USA) with a pulse width of 5–7 ns, beam diameter of 7 mm and frequency of 10 Hz at 0.3 W for 2 min, a corresponding dose of 90 J/cm2. The illumination did not raise the skin temperature higher than 41°C as measured by an infrared camera  or cause alteration in the skin visibly or histologically (Figure 1; LVA). Following the illumination, immunogens were intradermally administered into the site of laser illumination, whereas control mice received the immunogens similarly in the absence of laser illumination. The brief laser illumination was able to enhance ovalbumin (OVA)-specific antibody production by 300–500% over OVA alone (p < 0.001) . Similar immune enhancement was observed with other immunogens like 2009–2010 seasonal influenza vaccine , nicotine vaccine and malarial liver stage antigen 1 (data not shown).
Furthermore, when LVA was combined with MPL or CpG, two adjuvants with less reactogenicity either alone or after combination with LVA (Figure 1), a synergistic immune boosting was obtained. Thus, id. administration of a mixture of OVA and MPL into the site of laser illumination augmented OVA-specific antibody production by twofold over nonlaser-treated control, or 22-fold over antigen alone (data not shown). Likewise, when a nicotine vaccine was mixed with MPL at a 1:1 ratio and intradermally injected, nicotine-specific antibody level was increased by 13-fold over nicotine-vaccine alone and incorporation of LVA further increased the antibody level by 33-fold. Importantly, similar synergistic effects were attained with OVA-induced cell-mediated immune responses, as reflected by a significant increase in the number of CD4+ and CD8+ cells secreting IL-4 or IFN-γ in the presence, as compared with the absence, of LVA. LVA also synergistically boosted OVA-induced antibody production and cell-mediated responses when combined with CpG. Taken together, id. vaccination in combination with LVA, LVA/MPL, or LVA/CpG augmented immune responses by approximately 16-, 72- and 40-times, respectively, over im. vaccination in the absence of adjuvants. This physical adjuvant is thus ideal for cutaneous vaccination either alone or in combination with other vaccine adjuvants that are known to activate DCs such as MPL and CpG [30,31].
Mechanisms of LVA function
The mechanism underlying laser-mediated immune enhancement is not well understood at present. It does not appear to be solely caused by photothermal effects, because when the skin was maintained at 42°C by a 10 × 10 × 100 mm steel bar for 2 min followed by immunization at the warm skin with OVA as previously, OVA-specific antibody production was increased by only 20% . Furthermore, immunization conducted 2 or 4 h after laser illumination, with skin temperature having returned to normal, gave rise to a similar boost as immunization did immediately after laser illumination. Russian scientists reported that illumination of the skin with a laser of a higher power and density (0.6 W and 3 W/cm2) enhanced humoral immunity against influenza vaccine id. delivered in mice, by induction of extracellular heat-shock protein 70 production and inflammatory responses . In contrast, laser used in our study did not induce heat-shock protein or a significant inflammatory response in laser-exposed skin . We also observed little alteration in the surface expression of costimulatory molecules CD86 and CD83 or MHC class II molecule on skin DCs. We therefore consider LVA as a noninflammatory vaccine adjuvant.
However, injection of the antigen into the site of laser illumination appeared to be crucial since if the antigen was injected into a distal site, for instance, 1 cm away from the laser-illuminated site, the immune-enhancing potential decreased substantially . In accordance with this, we found that the laser illumination greatly accelerated the motility of APCs only at the areas of laser illumination. As shown in Figure 2, dermal GFP+cells, mostly DCs and macrophages, were constantly shifting, albeit slowly, and extending pseudopods, but most of them remained at original locations during a 20-min period of recording in the control (Figure 2). On the contrary, the cells in the laser-treated mice showed a high migratory ability: cells leaving their original locations and a gap appearing between arrows (original locations) and the individual cells over time. OVA injection also increased migration of APCs, albeit to a lesser extent. Strikingly, a synergistic effect was observed on APC motility when OVA was administrated into the site of laser illumination (Figure 2; laser + OVA). An increase in the motility of APCs is likely to promote them to survey a greater area and facilitate their antigen sampling as recognized using dendrite surveillance extension and retraction cycling habitude (dSEARCH) . The increased motility may also lead to sufficient transportation of antigen-captured DCs to the draining lymph nodes.
We postulate that brief laser illumination can transiently alter the interstitial microarchitecture, increase the tissue permeability, and permit relatively free movement of APCs in the interstitium. Indeed, upon laser illumination, dermal collagen fibers were dissociated and the interaction between DCs and surrounding tissue scaffolds was disrupted in the site of laser illumination, in sharp contrast to the well-organized microarchitecture in the control dermal connective tissue, as revealed by transmission electron microscopy (Figure 3). Dissociation of APCs with the matrix proteins is expected to free their movement . The laser illumination may also enlarge pre-formed channels in the peri-lymphatic basement membrane to assist entry of APCs into the lymphatic vessel . In support of altered interstitial resistance resulting in enhanced migration of DCs by laser illumination, we found that id. injection of DCs into the site of laser illumination increased the number of DCs migrating into the draining lymph nodes by approximately 300%, when compared with injection of DCs into the control skin, irrespective of DC maturation status. We also found that modification of pulse width, frequency and peak power at 532 nm did not significantly influence laser adjuvant effects, arguing for a physical-based mechanism. Another potential mechanism for laser-mediated immune enhancement may be the ability of laser to transiently permeabilize cellular membranes by a shock wave, which augments antigen uptake by DCs . Laser treatment has been shown to increase uptake of antisense oligonucleotide by three- or 30-fold due to a laser pulse-generated high pressure . Zeira et al. also showed that femtosecond laser sufficiently enhanced DNA delivery into cells and induced immune responses to the encoded antigen . In addition, acceleration of interstitial flow by laser illumination can greatly assist a flow of soluble antigens from the skin to the draining lymph nodes where the antigens are presented to resident DCs. In the skin, the initial lymphatic vessels are blind-end structures with wide lamina and thin walls. These initial lymphatic vessels drain excess fluid and solutes from the interstitial space and pass them to lymph nodes via lymphatic ducts. The draining process is extremely slow under normal physiological conditions but it can be increased as many as ten times by inflammation or fever-range hyperthermia [39,40]. Other mechanisms may be also involved in laser-mediated immune enhancement, including enhanced mitochondrial activity of APCs, chemical releases, altered tissue pressure, and so on. Although most of these mechanisms, such as laser-induced shock wave, accelerated interstitial flow and altered mitochondrial activity, remain largely speculative and more studies are needed to confirm, it is clear that laser augments vaccination by novel and distinct mechanisms over traditional vaccine adjuvants.
Advantages of LVA over traditional vaccine adjuvants
Vaccine adjuvants have been traditionally defined as molecules, compounds or macromolecular complexes that ideally boost the potency and longevity of specific immune responses to antigens, but cause minimal toxicity or long-lasting immunity on their own . However, LVA is not a chemical or compound, and it thus has the following advantages over conventional vaccine adjuvants. First, it eliminates the complex formulation process of mixing adjuvant and antigens. It also circumvents the problems associated with the maintenance of a stable mixture of the resultant vaccine during the cold storage. Optimal formulation of a safe, stable mixture between vaccine and adjuvant is a challenge for some vaccines such as rPA anthrax vaccine . Alum adjuvant is a noncrystalline gel and antigen must be adsorbed onto highly charged aluminum particles for the adjuvant to be potent. At least two serious issues result from the use of Alum. First, freezing, lyophilization, or cold storage would result in separation of antigen from the aluminum particles and cause a loss of the adjuvant potency [42–44]. Second, the biophysical structure and stability of the resultant product are difficult to assay as an Alum complex. Importance of formulation has also been illustrated by the development of the malaria RTS,S vaccine. When the malaria vaccine was mixed with Alum plus MPL (AS04), it failed to protect immunized subjects against a Plasmodium falciparum challenge, whereas the same antigen mixed with QS21 plus MPL in an oil-in-water emulsion (AS02) or in liposome (AS01) induced protection [45,46]. Second, LVA can be used immediately and repeatedly at any time, which offers a great advantage when facing vaccine shortages in the event of influenza pandemic, an outbreak of a new viral strain or a bioterrorist attack. Conceivably, the vaccine dose-sparing effect of LVA can greatly enhance the bioavailability of a given stockpile vaccine, which can potentially save millions of lives during the early phase of an influenza pandemic. Third, due to the way that the laser stimulates the immune system without direct interaction with vaccine itself, the laser-based vaccine adjuvant platform may work as a type of universal and standalone adjuvant, which is especially significant for the new US National Biodefense Strategy, stressing one more flexible, broad-spectrum approach for protection against multiple diseases. Conceivably, with a valid laser device, it can be conveniently and readily applied to any vaccine whenever it is needed. Fourth, LVA can be readily combined with the newly developed id. or transcutaneous delivery strategies, such as microinjection systems and microneedle array patches. Finally, LVA does not involve administration of any foreign or self substances into the body apart from the immunogen itself and thus would not induce self-destructive immune cross-reactions, also termed ‘molecular mimicry’, which can potentially cause long-term side effects [47,48]. By contrast, other adjuvants, regardless of whether they are foreign or self to the body, have the potential to cause long-term adverse reactions after repeated uses. Recently, the Swedish and Finnish authorities suspended further vaccination with Pandemrix™ (made by GlaxoSmithKline, UK), an AS03-adjuvanted 2009 pandemic H1N1 influenza vaccine, and began to investigate the causative link of Pandemrix vaccination to the rising cases of narcolepsy, a chronic neurological disease with disturbed sleep–wake cycles, in children and adolescents. This concern is raised because narcolepsy did not rise in the USA where a nonadjuvanted H1N1 influenza vaccine was used in the same period of time. In fact, whether or not adjuvanted influenza vaccines are safe is an area of hot debate among scientific communities, health authorities and the general public. Whether adjuvant exposure causes macrophagic myofacititis, Gulf War syndrome, and other rare mental and chronic autoimmune diseases remains an overall public concern.
Recent development of sufficient and convenient id. and transcutaneous vaccination technologies raises an urgent need for safe and potent vaccine adjuvants for augmentation of cutaneous vaccination. The majority of traditional vaccine adjuvants are not suitable for use in skin immunization because many of them cause unacceptable local reactogenicity owing to high sensitivity of the skin to inflammation. The use of lasers to stimulate skin immune cells is safe, simple, unique and ideal for cutaneous vaccination. It may be possible to adjuvantate both existing and future vaccines. This less inflammatory adjuvant can be potentially used to boost either Th1 or Th2 immune responses dependent on the nature of a given immunogen or the presence of other adjuvants. In this regard, our ongoing study showed that LVA in combination with MPL or CpG greatly boosted both humoral and Th1 immunity, which may be of particular significance for increasing vaccine immunogenicity in the elderly or immunocompromised populations. Furthermore, LVA can be readily incorporated into newly developed id. or transcutaneous vaccine delivery systems, bringing about at least a tenfold antigen-dose-sparing benefit, without any adjuvant injection, as compared with im. vaccination. Conceivably, a hand-held laser device can illuminate the skin for 2 min followed by microinjection or topical application of a vaccine-coated microneedle patch on the site of illumination. These simple vaccine delivery and immune-enhancing strategies will have a great impact on vaccination of a large population in the USA and worldwide, in particular, during an unpredictable vaccine shortage, for instance, during an influenza pandemic, an outbreak of a new viral strain, a bioterrorist attack, or a major natural disaster in fear of cholera outbreaks.
Within the next 5 years, a hand-held, safe prototype laser device will be fabricated and tested first in swine and then in the clinic. The laser vaccine adjuvant is expected to gain US FDA approval and advance to clinical trials because lasers with a much higher energy have been widely used for decades in patients and healthy individuals for a variety of cosmetic and therapeutic purposes. The laser vaccine adjuvant will also be combined with newly developed microinjection or microneedle delivery systems, which will represent saltational improvement in vaccine delivery for both dosage-sparing and a mass vaccination campaign. However, changing the route of delivery and formulation of existing vaccines for vaccination will require more research to determine which vaccines work more sufficiently with this technology and what population the technology should be applied to along with a careful economic assessment.
This work is supported in part by the NIH grants AI070785 and RC1 DA028378, Sponsored Research agreement grant #2008A25652 from Boston Biocom LLC., and Grand Challenges Explorations grant #53273 from the Bill & Melinda Gates Foundation (to Mei X Wu).
Financial & competing interest disclosure
The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.
No writing assistance was utilized in the production of this manuscript.
Papers of special note have been highlighted as:
• of interest
•• of considerable interest