Photodiagnosis and Photodynamic Therapy (2011) 8, 1—6
available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/pdpdt
Drug and light delivery strategies for photodynamic antimicrobial chemotherapy (PACT) of pulmonary pathogens: A pilot study Corona M. Cassidy a, Michael M. Tunney a, Nicholas D. Magee b, J. Stuart Elborn b, Steven Bell c, Thakur Raghu Raj Singh a, Ryan F. Donnelly Bsc, PhD a,∗ a
School of Pharmacy, Queens University Belfast, Medical Biology Centre, 97 Lisburn Road, Belfast BT9 7BL, UK Centre for Infection and Immunity, 3rd Floor, Medical Biology Centre, 97 Lisburn Road, Belfast BT9 7BL, UK c School of Chemistry and Chemical Engineering, Queens University Belfast, David Keir Building, Stranmillis Road, Belfast BT9 5AG, UK b
KEYWORDS Methylene blue; Light delivery; Pulmonary drug delivery
∗
Summary Pulmonary disease is the main cause of morbidity and mortality in cystic fibrosis (CF) suffers, with multidrug-resistant Pseudomonas aeruginosa and Burkholderia cepacia complex as problematic pathogens in terms of recurrent and unremitting infections. Novel treatment of pulmonary infection is required to improve the prognosis and quality of life for chronically infected patients. Photodynamic antimicrobial chemotherapy (PACT) is a treatment combining exposure to a light reactive drug, with light of a wavelength specific for activation of the drug, in order to induce cell death of bacteria. Previous studies have demonstrated the susceptibility of CF pathogens to PACT in vitro. However, for the treatment to be of clinical use, light and photosensitizer must be able to be delivered successfully to the target tissue. This preliminary study assessed the potential for delivery of 635 nm light and methylene blue to the lung using an ex vivo and in vitro lung model. Using a fibre-optic light delivery device coupled to a helium—neon laser, up to 11% of the total light dose penetrated through full thickness pulmonary parenchymal tissue, which indicates potential for multiple lobe irradiation in vivo. The mass median aerodynamic diameter (MMAD) of particles generated via methylene blue solution nebulisation was 4.40 m, which is suitable for targeting the site of infection within the CF lung. The results of this study demonstrate the ability of light and methylene blue to be delivered to the site of infection in the CF lung. PACT remains a viable option for selective killing of CF lung pathogens. © 2010 Elsevier B.V. All rights reserved.
Corresponding author. Tel.: +44 (0) 28 90 972 251; fax: +44 (0) 28 90 247 794. E-mail address:
[email protected] (R.F. Donnelly).
1572-1000/$ — see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.pdpdt.2010.12.007
2
Introduction Cystic fibrosis (CF) is the most common autosomal recessive condition in the UK [1], characterised by increased viscosity of bodily secretions, including mucus in the lungs, which is difficult to expectorate and, thus, easily infected. Pulmonary disease, as a result of the unremitting cycle of infection and inflammation, is the main cause of morbidity and mortality in these patients [2]. Chronic infection with pathogens such as Pseudomonas aeruginosa and Burkholderia cepacia are virtually impossible to eradicate, with antibiotic treatment of an acute exacerbation aimed at improving lung function by decreasing bacterial load and inhibiting enzymes such as proteases which damage lung tissue. There is clearly a need for alternative antimicrobial methods for treatment of chronic CF pulmonary infection caused by bacteria which are resistant to conventional antibiotic treatment. Photodynamic antimicrobial chemotherapy (PACT) combines highly coloured compounds with visible light to produce reactive oxygen species that are lethal to target pathogens [3]. A wide range of organisms from the Grampositive Staphylococcus aureus [4] to the Gram-negative P. aeruginosa [5] have been shown to be susceptible to PACT with a range of photosensitizers in vitro. For PACT to be of use clinically, effective delivery methods of light and photosensitizer to the site of action are necessary. Factors such as patient acceptability, target site and photosensitizer dose to be delivered all impact on treatment success. The physical and chemical mucus compositions and the degree of parenchymal destruction and bronchiectasis in the lungs of CF patients can significantly alter drug distribution and bioavailability of drugs delivered to the lung [6]. More specifically, Donnelly et al. identified CF mucus as a potential barrier to efficient photosensitizer and light delivery [5]. Pulmonary photodynamic therapy (PDT) is currently indicated for the treatment of pre-invasive carcinoma or in the palliative treatment of tumours that obstruct bronchi, with irradiation mediated by a fibre optic cable coupled to a laser [7,8]. Photosensitizers for PDT tend to be relatively large molecules, are usually administered parenterally, and accumulate selectively in neoplastic tissue. Targeting of photosensitizers to sites of infection in this way is not possible and so the drug must be applied topically. Physicochemical properties of the photosensitizer, photosensitizer dose to be delivered, barrier properties of the target site and patient acceptability are important aspects to consider when choosing the method of drug delivery. The aim of this preliminary study was to assess, using ex vivo and in vitro lung models, whether light and a photosensitizer could be delivered to the site of infection within the CF lung.
Materials and methods Photosensitizers The photosensitizer used in this study was methylene blue (MB; Sigma—Aldrich, Dorset, UK). MB was dissolved in phosphate buffered saline pH 7.4 (PBS; Oxoid, Hampshire, UK)
C.M. Cassidy et al. to make stock photosensitizer solutions, before sequential sterile filtration through 0.45 m and 0.22 m filters. Stock solutions were then aseptically diluted in PBS to achieve the required photosensitizer concentration. All solutions were protected from light, stored at 4 ◦ C, and used within 7 days of preparation.
Light delivery through lung tissue Porcine lungs, free from disease, were sourced locally from an abattoir. On receipt, the tissue was frozen at −20 ◦ C until required for use, when it was thawed for 24 h at 4 ◦ C. Using an endotracheal tube (Smiths Medical International Ltd., Kent, UK) to intubate the lung, a fibre optic probe with a diffuser tip of 2 cm (Medlight SA, Axcon Pharma, Switzerland) was inserted into one of the lower lobes of the pulmonary tissue. For the duration of this study, a flow of nitrogen was used to maintain inflation of the pulmonary tissue. Initially, the probe was coupled to a 633 nm helium—neon laser with an output of 5 mW, with the fluence output at the tip determined prior to insertion using an Ophir Orion PD light power meter (Bfi Optilas, Milton Keynes, UK). The fluence through the tissue was measured by placing the probe of the light meter against the tissue at the nearest point to the diffuser tip, with the laser switched off, and measuring fluence. The laser was then switched on and fluence re-measured. The fibre optic probe was then withdrawn from the lung at 5 mm increments along the perpendicular plane, with the fluence measured at each point, until no light was detected with the meter. This was repeated on two occasions at 633 nm, and at 785 nm. Fig. 1 illustrates the experimental set up used in this study.
Figure 1 Experimental set-up for light delivery through porcine lung tissue. Where (a) is the light power meter, (b) is the point where the probe of the light meter is placed against the lung tissue at the nearest point to the diffuser tip of the optical probe (c), (d) is the excised porcine lung, inflated with nitrogen, (e) is the connection of the endotracheal tube to the lung, (f) is the endotracheal tube, (g) is the connecter between the endotracheal tube and the nitrogen supply and (h) is the fibre optic probe.
How about: Drug and light delivery for PACT of pulmonary pathogens
3
Fine particle characteristics of nebulised photosensitizer and delivery to the lung A Copley Multistage Liquid Impinger (MSLI, Copley Scientific Limited, Nottingham, UK) was used, in accordance with the manufacturer’s instructions, to determine the fine particle characteristics of a nebulised MB solution. Aliquots (20 ml) of PBS were added to each of the four upper stages of the MSLI, with a filter paper in the fifth stage. A Vacuubrand chemistry diaphragm pump (Model MZ 2C NT + AK + EK; Vacuubrand, Wertheim, Germany) was connected to the outlet of the apparatus and the airflow adjusted to measure 35 ± 5 l/min at the inlet to the induction port with the airflow switched off. A MB solution of 1 mg/ml was introduced into the reservoir of the PARI TurboBoy N (PARI Medical Ltd., Surrey, UK) nebuliser and the mouthpiece of the nebuliser connected to the adapter of the apparatus. The pump was switched on for 10 s before switching on the nebuliser. After 60 s, the nebuliser was switched off, and 5 s later, the airflow switched off. The filter paper was then removed from stage 5 of the apparatus, placed into a container containing an aliquot of PBS, with the drug extracted into the solvent. All parts of the apparatus, including the induction port and mouthpiece adapter, were placed in a container with a solvent aliquot in order to extract all trace of MB. MB was then extracted from the inner walls and the collection plate of each of the four upper stages by careful tilting and rotating the apparatus, observing that no liquid transfer occurred between stages. Sample analysis was carried out using a Cary UV—vis spectrophotometer (Varian, Dublin, Ireland) at an absorbance wavelength of 665 nm. The instrument was calibrated between 0.00 and 10.00 g/ml for MB: limit of detection 0.041 g/ml; limit of quantification 0.13 g/ml. Analysis of resulting drug-loaded aliquots determined the quantity of drug contained in each of the six volumes of solvent. The percentage of drug collected at each stage and the cumulative percentage were determined and calculation of cut-off diameter was performed according to Eq. (1).
Cut-off diameter for stage n = fn
60 Q
1/2 (1)
where Q is the flow rate (l/min) between 30 and 100 l/min and for each stage, f2 = 6.8, f3 = 3.1 and f4 = 1.7. From this data, the mass median aerodynamic diameter (MMAD) and the geometric standard deviation (GSD) were calculated, via a plot of the cumulative percentage versus cut-off diameter on log probability paper. The MMAD correlates to the cut-off diameter corresponding to 50% on the cumulative percentage scale, with the GSD calculated from Eq. (2) if a straight line drawn between the data points is a good fit:
0.5 GSD =
x y
(2)
where x and y are the particle sizes corresponding to, for x, 84.13% of cumulative percentage and for y, 15.87%.
Figure 2 A graph showing percentage transmission of fluence detected through porcine lung tissue, with progressive retraction of light diffusing tip through inflated lung. Where wavelength = 633 nm, output at source = 55mW, output at diffuser tip = 2.4 W. Where wavelength = 785 nm, output at source = 125mW, output at diffuser tip = 77 W.
Statistical analysis Where appropriate, the Kruskal—Wallis test for independent samples was used to determine significance of difference in results between treatments, with the Mann—Whitney’s U test used to determine significance between individual treatments. In all cases, p < 0.05 denoted significance.
Results Light delivery through lung tissue Fig. 2 shows the relative decrease in fluence when the fibre optic light delivery device was removed in 5 mm increments from an excised inflated porcine lung. In all cases, a decrease in fluence was noted with progressive removal of the light diffusing cable from the lung. A percentage decrease in fluence of approximately 89% was noted in the first lobe tested when the diffuser tip was removed 10 mm from the initial position. Therefore, approximately 11% of the total light dose penetrated through the entire thickness of the inflated pulmonary tissue. In the second lobe tested, this was reduced to approximately 4%. When the emitting wavelength was changed from 663 nm to 785 nm, the light dose delivered through porcine pulmonary tissue was increased to approximately 23%. Fig. 3 shows the penetration of light from the diffuser tip through the porcine lung in ambient light and darkness.
Fine particle characteristics of nebulised photosensitizer and delivery to the lung The results presented in Table 1 show that an average of 87% of particles generated by the Pari Turboboy N at a tidal pressure of 35 l/min were less than 11.21 m in diameter, with 58% and 26% less than 5.11 m and 2.80 m in diameter respectively. The average MMAD was 4.40 m and the GSD
4
C.M. Cassidy et al. Table 1 Average percentage of total mass collected from stages in multi-stage liquid impinger, cumulative percentage and calculated cut-off diameter of aerosolised particles (±SD, n = 5). Stage
Stage collection (%)
Cumulative (%)
Cut-off diameter (m)
5 (filter paper) 4 3
26.39 (±4.35) 31.96 (±4.24) 32.17 (±5.43)
26.39 (±4.35) 58.35 (±5.92) 87.81 (±1.39)
2.80 5.11 11.21
Stage 2, pre-separator. Stage 1, mouthpiece adaptor and induction port.
was 2.30. The MMAD may be used as a guide for particle deposition within the lung.
Discussion For PACT to be a successful treatment option for multidrugresistant pathogens infecting the CF lung, sufficient light and photosensitizer must be able to reach the target site of infection within the lung. Donnelly et al. have previously demonstrated that 635 nm light can be delivered through artificial CF mucus in vitro, at fluence levels suitable for treatment of CF pathogens [5].
Figure 3 Light penetration through inflated pig lung in ambient light (a) and in darkness (b).
Mammalian lungs are composed of two distinct areas: conducting airways and pulmonary parenchyma [9]. The parenchymal tissue, composed of alveoli and alveolar ducts, forms the large surface area of the lung required for gas transfer which is folded within the respiratory tract [9]. The average thickness of alveolar septa is approximately 10 m [10]. Thus, the lung is a highly complex organ, composed of folds of tissue of micron thickness. This is the first study to assess the penetration of light through such tissue, with a view to illuminate multiple adjacent lobes for the treatment of pulmonary infection via PACT. It is commonly perceived that light cannot penetrate lung tissue more than a few millimeters; however this is in the context of photodynamic therapy for lung tumours that are solid and block light effectively [7]. Light penetration of lung parenchyma has not previously been described. In PACT, the tissue is not the target for light delivery, as is the case for lung cancer PDT. Worlitzsch et al. demonstrated the affiliation of P. aeruginosa for mucus within the CF lung, as opposed to pulmonary epithelial cells [11]. Therefore, light need only be delivered to the site of infection to exert its effect on the photosensitizer. However, this study was performed to determine if light could penetrate through tissue to reach the target site, as opposed to delivery via a fibre-optic tube to all of the smaller airways. It is technically difficult to measure light diffusion through an inflated lung because it is an air-tight system and any measurements that disrupt the seal will cause collapse of the lung. In this study, we attempted to measure light diffusion through an inflated lung whilst maintaining this air-tight system. Consequently it is difficult to ascertain the exact location of the probe tip within the lung, as the airways run at an oblique angle to the surface of the lung, where the measurements were taken. This may have skewed results in a negative manner. A further limitation of the study is that the lung tissue used was from a non-infected pig, whose lung histology would be likely to differ from that of a CF-patient. Tiddens et al. compared the larger airway dimensions of lung tissue from CF and chronic obstructive pulmonary disease patients and determined that the inner wall and smooth muscle areas of peripheral CF airways were increased by 3.3- and 4.3fold respectively compared to those of COPD airways [12]. However, it is known that progression of cystic fibrosis pulmonary disease is associated with parenchymal destruction [13], which could result in increased transmission of light through the parenchymal tissue in CF patients in comparison with non-CF patients. Despite these limitations, we have shown that light may adequately diffuse through a lobe of a lung to enable the
How about: Drug and light delivery for PACT of pulmonary pathogens activation of a nebulised photosensitizer. There is therefore potential to illuminate each of the 3 lobes of the lung individually via bronchoscopic placement of a light diffuser and combine this with a photosensitizer to treat difficult infections more effectively. During this bronchoscopy, it may be reasonable to illuminate each of the 6 lobes individually from both lungs with a light exposure of 5 min, as 30 min is a reasonable duration for a bronchoscopy. Alternatively, a light diffuser could be bronchoscopically placed in order to illuminate a whole lung, secured in place for a more prolonged period of time such as 12 h. This would, of course, necessitate a second bronchoscopy at the end of the 12 h to reposition the diffuser in the other lung. The results of this pilot study indicate that the fibre optic cable currently employed in PDT of lung carcinoma could be used to irradiate parenchymal tissue and that light can penetrate through lung tissue to a limited extent, which may allow for diffuse delivery of light to the irradiated lobe in situ. This would potentially reduce the illumination time required in PACT of pulmonary infection. Further ex vivo and in vivo studies are required before this therapy could be used clinically. It has been shown that the photosensitizers toluidine blue (TBO) and meso-tetra(N-methyl-4-pyridyl)porphine tetratosylate (TMP) are capable of diffusing through artificial CF pulmonary mucus in concentrations sufficient to produce over 99% kill upon irradiation with 100 J/cm2 of light from the Paterson lamp [5]. TBO, like MB, is a phenothiazinium compound, is structurally similar to MB, and so would be expected to have a similar mucus penetration profile. In this study, the diameter of MB particles impacting at the various stages of the MSLI was determined by the air flow within the system, as generated by the vacuum pump. We used a pump with a tidal pressure of 35 l/min because it was the most powerful vacuum readily available, and a tidal pressure of between 30 and 100 l/min was required to calculate the particle cut-off diameter. The results show that the Pari TurboBoy N, in conjunction with an inspiratory flow rate of 35 (±5) l/min can be used to form and deliver aerosolised particles of MB to targeted regions of the lung. These findings are in keeping with Laube et al., who determined that particle size and inspiratory rate could be manipulated to target pulmonary delivery of aerosols [14]. Nebulised antibiotics for CF lung infection are sized between 3 and 5 m to allow targeted delivery to the site of infection [15]. The MB particles generated in this study are of similar diameter, with a MMAD of 4.10 m, and would therefore be ideal for targeting infection in the CF lung. Limitations of the MSLI as a method of particle characterisation include difficulty in removing the total quantity of drug from the chambers and filter paper in the final chamber, which could lead to anomalous results. The variation was limited by repeating the experiment on five separate occasions, and standardizing the method of drug removal beyond the scope of the instruction manual. Additional measures included rinsing the sintered glass within each chamber five times with 2 ml aliquots of the collecting fraction before removal and vortexing the filter paper on two occasions for 30 s, with a 30 min interval for drug to partition from the filter, before filtering with a 0.22 m filter. Bonam et al. systematically reviewed the literature on variability issues regarding cas-
5
cade impacters, and determined that, although most studies focused on the limitations of the devices, other factors, such as operator-derived influences, the contribution of drug assay methodology and product-related causes should be considered [16]. The results attained showed some variability, likely as a result of a combination of the factors listed above, but the results indicate that apparatus used can easily target delivery of MB to the site of infection within the CF lung.
Conclusion The results of this preliminary study demonstrate that apparatus, such as nebulisers and technology for PDT of lung carcinoma, currently available in the clinic, can be used for PACT. Planktonic and biofilm cultured CF pathogens have proven to be highly susceptible to MB-PACT, and light and MB can successfully be targeted to the site of infection in the CF lung. However, before PACT can be used clinically, factors such as total light dose to be delivered and duration of light exposure require further optimization.
Acknowledgment This work has been supported by the Department of Employment and Learning of Northern Ireland (DEL), who provided funding for the studentship of Corona Cassidy.
References [1] McCormick J, Green MW, Mehta G, Culross F, Mehta A. Demographics of the UK cystic fibrosis population: implications for neonatal screening. Eur J Hum Genet 2002;10: 583—90. [2] Boucher RC. An overview of the pathogenesis of cystic fibrosis lung disease. Adv Drug Deliv Rev 2002;54: 1359—71. [3] Donnelly RF, McCarron PA, Tunney MM. Antifungal photodynamic therapy. Microbiol Res 2008;163:1—12. [4] Griffiths MA, Wren BW, Wilson M. Killing of methicillinresistant Staphylococcus aureus in vitro using aluminium disulphonated phthalocyanine, a light-activated antimicrobial agent. J Antimicrob Chemother 1997;40:873—6. [5] Donnelly RF, McCarron PA, Cassidy CM, Elborn JS, Tunney MM. Delivery of photosensitisers and light through mucus: investigations into the potential use of photodynamic therapy for treatment of Pseudomonas aeruginosa cystic fibrosis pulmonary infection. J Control Release 2007;117: 217—26. [6] Kuhn KJ. Formulation of aerosolized therapeutics. Chest 2001;120:94S—8S. [7] Usuda J, Ichinose S, Ishizumi T, et al. Outcome of photodynamic therapy using NPe6 for bronchogenic carcinomas in central airways >1.0 cm in diameter. Clin Cancer Res 2010;16: 2198—204. [8] Usuda J, Ichinose S, Ishizumi T, et al. Management of multiple primary lung cancer in patients with centrally located early cancer lesions. J Thorac Oncol 2010;5:62—8. [9] Tsuda A, Filipovic N, Haberthur D, et al. Finite element 3D reconstruction of the pulmonary acinus imaged by synchrotron X-ray tomography. J Appl Physiol 2008;105: 964—76.
6 [10] Gehr P, Bachofen M, Weibel ER. The normal human lung: ultrastructure and morphometric estimation of diffusion capacity. Respir Physiol 1978;32:121—40. [11] Worlitzsch D, Tarran R, Ulrich M, et al. Effects of reduced mucus oxygen concentration in airway Pseudomonas infections of cystic fibrosis patients. J Clin Invest 2002;109:317—25. [12] Tiddens HA, Koopman LP, Lambert RK, et al. Cartilaginous airway wall dimensions and airway resistance in cystic fibrosis lungs. Eur Respir J 2000;15:735—42. [13] Regamey N, Jeffery PK, Alton EW, Bush A, Davies JC. Airway remodelling and its relationship to inflammation in cystic fibrosis. Thorax 2010.
C.M. Cassidy et al. [14] Laube BL, Jashnani R, Dalby RN, Zeitlin PL. Targeting aerosol deposition in patients with cystic fibrosis: effects of alterations in particle size and inspiratory flow rate. Chest 2000;118:1069—76. [15] Coates AL. Guiding aerosol deposition in the lung. N Engl J Med 2008;358:304—5. [16] Bonam M, Christopher D, Cipolla D, et al. Minimizing variability of cascade impaction measurements in inhalers and nebulizers. AAPS PharmSciTech 2008;9: 404—13.