During photodynamic treatment, in addition to the reaction with biological substrate, self-photosensitization occurs and reactive oxygen intermediates interact with the photosensitiser, leading to its transformation and/or destruction. This phenomenon is called photobleaching. The first relevant observation of photobleaching in the photodynamic therapy field was made in 1986 by Moan (Moan, 1986).
The main reactions leading to photobleaching are presented in fig. 4.1. The photosensitiser undergoes Type I and/or Type II mechanisms upon light irradiation, leading to the production of oxygen radical species. These oxygen radical species react with the neighbouring molecules, including the photosensitisers, leading to their destruction. Thus photobleaching can occurs via two pathways, the Type I way involving reactive oxygen species and Type II way involving singlet oxygen. Photobleaching quantum yield of different photosensitisers varies significantly and can be attributed to oxidation potential, lipophilicity, presence of a metallic ion, kind of reactions involved (Type I or II).
Kinetic parameters of photobleaching are mainly derived from spectroscopic measurements assessed by UV-Vis or fluorescence spectroscopy. Several important mechanistic issues of photobleaching were obtained from the detailed analysis of spectroscopic modifications. In the earlier studies on photobleaching of photosensizer molecules, the kinetic decay of photosensitiser was considered as a mechanism depending only on the light dose delivered to the tissue, materialized by the mono-exponential decay e-αD, where α stands for the photobleaching constant and D stands for the fluence of irradiation (J/cm-2). It became clear later on that the photobleaching is a complex phenomenon, which cannot be described by a single exponential decrease (Moan et al., 2000; Sørensen et al., 1998). For some photosensitisers the decay rates have been shown to be practically independent of the concentration of the dye during illumination (Mang et al., 1987; Moan, 1986; Sørensen et al., 1998) and thus exhibit a first order decay. However, for the majority of dyes, the photobleaching decay is highly dependent on the initial photosensitiser concentration (Moan et al., 1988), meaning that the photoproducts from the chromophore can cause the decay of a neighbouring chromophore (Moan et al., 1997).
Parameters affecting photobleaching. Aggregation state, pH, ionic strength and oxygen concentration
Bezdetnaya et al. (Bezdetnaya et al., 1996) demonstrated that for HpD and PpIX, quantum yield of photobleaching obtained by matching fluorescence were higher than those obtained by matching absorbance (10 and 11 times for HpD and PpIX respectively). The authors concluded that this difference reflected the preferential photobleaching of photolabile monomeric forms compared to aggregates. Another study confirmed the preferential photobleaching of monomeric species of m-THPC (Belitchenko et al., 1998).
Several studies of Rotomskis and co-workers demonstrated that photobleaching efficiency of haematoporphyrin-like sensitisers seemed to be consistent with their aggregation state and the presence of covalently linked structures. Both dimethoxyhaematoporphyrin (DMHp) and Hp are present in an equilibrium of monomeric and aggregated forms in aqueous solutions (Streckyte and Rotomskis, 1993). Their absorption bleaching rate constants are two to four times higher than that of HpD, a sensitiser containing mostly linear structures of porphyrins linked by ether, ester and/or carbon-carbon bonds (Dougherty et al., 1984). It is also 10 to 20 times higher than that of Photofrin® (PF), which contains covalently linked ” sandwich” type structure (Streckyte and Rotomskis, 1993b). In HpD, some of the side chains are involved in ether and ester linkages and therefore this compound is more photostable than DMHp and Hp. In PF and Photosan-3 (PS) (highly aggregated “sandwich” type structure (Strec kyte and Rotomskis, 1993)), almost all side chains are involved in covalently linked structures, probably accounting for the high photostability of these sensitisers. The presence of a certain amount of protoporphyrin in PS is probably responsible for its lower photostability compared to PF.
Lowering the pH value of a photosensitiser solution results in a shift of both the absorption and the fluorescence spectra as well as in a decrease of the fluorescence intensity, indicating an aggregation at low pH values (pH < 5) (Cunderlikova et al., 1999). Reddi et al. (Reddi and Jori, 1988) also demonstrated an aggregation of hematoporphyrin and Photofrin® when decreasing the pH from 7.4 to 5.0 and they also demonstrated the decrease of the photobleaching quantum yield to 70 % for hematoporphyrin and 30 % for Photofrin® , thus suggesting a resistance toward photobleaching of aggregated species.
Changing the ionic strength by varying the buffer concentration can affect the aggregation state of a sensitiser. An increase of the buffer concentration of a TPPS4 solution increases the aggregation of the sensitiser and reduces the photobleaching quantum yield by 50 % (Davila and Harriman, 1990). Thus, it follows from all these studies that the quantum yield of photobleaching is inversely proportional to the aggregation state of the photosensitisers.
Streckyte et al. (Streckyte and Rotomskis, 1993) showed that in micellar media (Triton X-100), which leads to the monomerisation of the photosensitisers, several dyes such as DMHp and HP had a different photostability and different photoproducts formation compared to aqueous media. Spikes (Spikes, 1992) reported that adding CTAB (cetyltrimethylammonium bromide) to PBS solution increases the quantum yield of PF photobleaching by 90%. The photobleaching of uroporphyrin I and hematoporphyrin in the same conditions was unchanged and the bleaching of TPPS4 decreased by 25%. Spectroscopic studies demonstrated that there was significant monomerisation of hematoporphyrin, TPPS4 and PF in CTAB, however the reasons for this opposite effect between TPPS4 and PF were not clear. The authors proposed that TPPS4 penetrates into the CTAB micelles (Reddi and Jori, 1988) and that it localizes in a low dielectric constant region and that under these conditions photobleaching would probably be slower.
Spikes (Spikes, 1992) investigated the quantum yield of photobleaching of several porphyrins in phosphate buffer solution, and found that the bleaching was reduced by nitrogen bubbling. Also, Streckyte and co-workers demonstrated that the photobleaching process of ALA-induced PpIX in cells was slowed down by bubbling nitrogen through the sample (Streckyte et al., 1994). König et al. also made the same observation for endogenously formed porphyrins in bacteria during argon flushing (König et al., 1993). An observation of the involvement of oxygen in vivo has been realised by Robinson and co-workers (Robinson et al., 1998). During a photobleaching experiment with ALA-induced PpIX, the mice died and they observed a slowdown of the photobleaching. They correlated this bleaching decrease to the oxygen decline in the skin, due to the death of the animal.
Several studies from the laboratory of TH. Foster documented oxygen depletion during PDT. Oxygen consumption model was refined by Georgakoudi and co-workers (Georgakoudi and Foster, 1998; Georgakoudi et al., 1997) by taking into account the parameter of photobleaching of Photofrin in EMT6 spheroids. This improvement considerably changed the kinetic profile of the oxygen aspects of Photofrin-PDT. The authors observed a rapid decrease in oxygen concentration during irradiation followed by a progressive return to the values measured before the irradiation. The first phase is due to the photochemical oxygen consumption which is faster than the diffusion of the oxygen through the spheroid. The second phase, corresponding to the comeback of oxygen to the initial value, is due to a slowdown of the photochemical consumption of the oxygen explained by a decrease in photosensitiser concentration (photobleaching), together with the diffusion of oxygen. This was in total agreement with the data from the mathematical model that they had developed, assuming that the photobleaching was based on a reaction between singlet oxygen and photosensitiser at the ground state. The validity of the developed model was confirmed by applying it to the experimental results on photobleaching in NHIK 3025 cells loaded with Photofrin from the study of Moan (Moan, 1986).
In their further studies Foster and co-workers investigated the impact of irradiance on photobleaching (Finlay et al., 2001; Finlay et al., 2002). In a study reporting the photobleaching of ALA-induced Protoporphyrin IX (Pp IX) in normal rat skin (Finlay et al., 2001) it was demonstrated that the photobleaching kinetics were different with different irradiances. High irradiance led to rapid oxygen consumption and a slow down of the photobleaching. In addition, the photoproducts of PpIX also exhibited an irradiance dependant photobleaching. In a second study, Finlay et al. (Finlay et al., 2002) showed that photobleaching kinetics of m-THPC on normal rat skin exhibited two distinct phases. The first phase was shown to be irradiance independent, whereas the second phase revealed an irradiance dependency consistent with an oxygen-dependant reaction process.
Presently, two mechanisms of photobleaching are acknowledged (Bonnett and Martínez, 2001). The first one, true photobleaching, corresponds to the photodegradation of the porphyrin macrocycle with the formation of photoproducts, which do not absorb in the visible light region. The second mechanism is called photomodification, where the chromophore is retained in a modified form with the formation of new visible spectral bands. For the majority of photosensitisers the photoproducts arise from both photodegradation pathways.
Photomodification is featured by the loss of absorbance or fluorescence at some wavelength and the appearance of new spectral bands, this being in agreement with the photoformation of new compounds. For macrocyclic compounds, photomodification appears when the rupture of the macrocycle doesn’t occur. While true photobleaching leads to the destruction of the tetrapyrrolic cycle and results in the formation of small products that do not absorb visible light. It appears that, where photomodification occurs, true photobleaching often occurs concomitantly and one also should notice that photomodification can be mistaken for photorelocalisation.
The photobleaching of tetraphenylchlorin series sensitisers have been extensively studied because of the large clinical potential of the m-THPC and also because of their important absorption in the red region of the visible spectrum. Bonnett et al. have made a comparative study of the photobleaching of this sensitiser series by absorption measurements (Bonnett et al., 1999). The authors demonstrated that in methanol-water solution m-THPC and m-THPBC underwent only true photobleaching and photomodification mainly occurs for m-THPP. The products formed after the irradiation of m-THPP methanol-water were hydroxylated m-THPP (mono-, di-, tri- and tetra-hydroxylated m-THPP) (Bonnett and Martinez, 2002) with mono-hydroxylated m-THPP being the major photoproduct (25%). While in pure methanol small photoproducts appeared such as maleimide and methyl-3-hydroxybenzoate, the mono-hydroxylated m-THPP was still photoproduced. A recent study (Lourette et al., 2005) regarding the photobleaching of m-THPP in ethanol-water (1/99, v/v) solution revealed that using a pulsed laser as light source, m-THPP undergoes phototransformation to a hydroxylated product and several covalent oligomeric structures as dimer, trimer, tetramer and pentamer of m-THPP.
Several studies on m-THPC photobleaching demonstrated a rapid true photobleaching of m-THPC, accompanied by a photoproduct formation at λabs = 320 nm when the photosensitiser was in a PBS solution supplemented with 2% fetal calf serum (FCS) (Angotti et al., 1999, 2001; Belitchenko et al., 1998). This result was confirmed by Hadjur et al. (Hadjur et al., 1998) who showed a large formation of a 320 nm absorbing product in a 10% FCS solution. In methanol or methanol-water solution it appears that m-THPC undergoes true photobleaching (Bonnett et al., 1999), since no photoproduct at 320 nm was detected. These three observations let us propose that the photoproduct formation correlates with the FCS concentration in the incubation solutions. Mass spectrometry studies were carried out to identify the spectroscopically invisible photoproducts (Angotti et al., 1999, 2001; Jones et al., 1996; Kasselouri et al., 1999). The photobleaching was performed on m-THPC methanol solution or water-methanol solution and the products obtained are presented in the fig. 4.2 and 4.3.
The major photoproducts observed were hydroxy- and di-hydroxy-m-THPC, hydroxy-m-THPP, still the position of the hydroxyl(s) group(s) is(are) not determined, these penta or hexahydroxylated chlorin have almost the same absorption peak than m-THPC (Jones et al., 1996; Kasselouri et al., 1999). Bonnett (Bonnett and Martinez, 2002) identified several products like a chlorin and four minor products coming from true photobleaching. They are dipyrrin derivative, succinimide, and the two afore mentioned products maleimide and methyl-3-hydroxybenzoate, which were also photoproducts from m-THPP.
Liposomes are spherical self-closed structures, composed of curved lipid bilayers, which enclose part of the surrounding solvent into their interior. The size of a liposome ranges from some 20 nm up to several micrometers and they may be composed of one or several concentric membranes, each with a thickness of about 4 nm. Liposomes possess unique properties owing to the amphiphilic character of the lipids, which make them suitable for drug delivery.
Amphiphilic lipids, used for liposome preparation, consist of hydrophilic polar headgroup and hydrophobic hydrocarbon chains. This means that a polar environment, such as water solutions, promote the spontaneous aggregation of such molecules and the formation of a variety of microstructures (Tanford, 1991).
Table of contents :
I GENERAL INTRODUCTION
II ORIGINS AND CLINICAL APPLICATIONS OF PHOTODYNAMIC THERAPY
III PHOTOSENSITIZATION MECHANISMS
III.1 Pathways of molecular excitation and deactivation
III.2 Mechanism of photosensitized reactions
III.3 Type I photosensitization processes
III.4 Type II photosensitization processes
IV.1 Photobleaching mechanisms
IV.2 Parameters affecting photobleaching. Aggregation state, pH, ionic strength and oxygen concentration
IV.3 Photoproducts formation
V.1 Amphiphilic lipids
V.2 Lipid bilayers
V.3 Steric stabilization
VI LIPOSOMES FOR ANTI-CANCER DRUG DELIVERY
VI.1 Liposomes for drug delivery
VI.2 Pharmacokinetics and biodistribution of liposomes and liposomal drugs
VI.3 Accumulation of liposomal drugs in tumors
VI.4 Stability in plasma and storage
VI.5 Bioavailability of encapsulated drug
VI.6 Partitioning of lipophilic and amphiphatic drugs into liposomes
VII LIPOSOMES FOR PHOTODYNAMIC THERAPY
VII.1 Liposomal photosensitizing agents
VII.2 Photophysical properties of liposomal photosensitizing agents
VII.3 Pharmacokinetic of liposomal photosensitizing agents
VII.4 Liposomal formulations of meta(tetrahydroxyphenyl)chlorin
IX.1 Unusual photoinduced response of mTHPC liposomal formulation (Foslip)
IX.2 Study of meso-tetra(hydroxyphenil)chlorin (mTHPC) redistribution from lipid vesicles to biological substrates
IX.3 Photodynamic therapy with intratumoral administration of lipid-based mTHPC in a model of breast cancer recurrence
X GENERAL DISCUSSION
X CONCLUSIONS AND PERSPECTIVES