You are on page 1of 25

Journal of Porphyrins and Phthalocyanines

J. Porphyrins Phthalocyanines 2001; 5: 105–129

REVIEW

Basic principles of photodynamic therapy

IAN J. MACDONALD* and THOMAS J. DOUGHERTY

Photodynamic Therapy Center, Roswell Park Cancer Institute, Buffalo, NY 14263, USA

Received 15 April 2000


Accepted 8 June 2000

Table of Contents: 1.6.1 Vascular Shutdown and Inflammation


1.6.2 Direct Cell Effects
1. INTRODUCTION 1.6.2.1 DNA Damage
1.1 History of Photodynamic Therapy 1.6.2.2 Lysosome Damage
1.2 Photochemistry of Photodynamic Therapy 1.6.2.3 Mitochondrial
1.3 Oxygen in PDT Damage
1.3.1 Singlet Oxygen 1.6.2.4 Membrane Damage
1.3.2 Tissue Oxygenation 1.7 Apoptosis
1.4 Photosensitizers 1.7.1 Apoptosis and Necrosis
1.4.1 Porphyrin Photosensitizers 1.7.2 Apoptosis in PDT
1.4.2 Photofrin 1.7.3 Mitochondrially Controlled Apoptosis
1.4.3 New Photosensitizers 1.7.3.1 Reactive Oxygen
1.4.4 Photosensitizer Pharmacology Species (ROS) and the
1.4.4.1 Structure–Activity and Mitochondrial Membrane
Quantitative Structure– Potential (D m) in Apoptosis
Activity Relationships 1.7.3.2 Permeability
1.4.4.2 Targeted Photosensitizers Transition (PT)
1.4.4.3 Amphiphilicity 1.7.3.3 Apoptosis Inducing Factor
1.4.4.4 How Amphiphilicity, and Cytochrome c in
Aggregation, and Apoptosis
Serum Protein Interactions 1.7.4 Summary, PDT and
Affect the Photodynamic Apoptosis
Activity of Photofrin1 1.8 Immune Effects in PDT
1.4.4.5 How Amphiphilicity, 1.8.1 Immune System Activation and
Aggregation, and Suppression
Serum Protein Interactions 1.8.2 Immune System Modulators
Affect the Photodynamic 1.9 Clinical Photodynamic Therapy
Activity of Other
Photosensitizers Acknowledgement
1.5 Light
1.5.1 PDT Dose REFERENCES AND NOTES
1.5.2 Light Penetration
1.5.3 Photobleaching
1.5.4 Light Sources and Delivery
1.6 Biological Mechanisms of Action
1. INTRODUCTION
1.1 History of Photodynamic Therapy

——————— The therapeutic use of light, as we know it, begins in 1900


*Correspondence to: I. MacDonald, Photodynamic Therapy Center, when Raab reported that the combination of acridine orange
Roswell Park Cancer Institute, Buffalo, NY 14263, USA. and light could destroy living organisms (paramecium) [1].
In the 1920s Policard [2] noted that tumor tissue was
Copyright # 2001 John Wiley & Sons, Ltd. inherently more fluorescent than healthy tissue. In the 1950s
106 I. J. MACDONALD AND T. J. DOUGHERTY

Ronchese [3] attempted to activate endogenous fluorescent


molecules in tumor tissue to delineate its boundaries more
accurately. Between 1940 and 1960 Figge et al. [4–6] and
Rasmussen-Taxdal et al. [7] administered natural porphy-
rins to patients and tumor-bearing animals in an attempt to
more accurately detect tumor tissue by fluorescence. During
the 1960s Winkelman [7–10] used synthetic porphyrins
(tetraphenylporphines) to detect tumor tissue. Throughout
the 20th century a few attempts were made to treat tumor
tissue with photosensitizing agents, mainly with non-
porphyrin photosensitizers. Tappeiner et al. and Jesionek
et al. [11, 12] used sunlight to activate eosin in tumor tissue
in vivo. Bellin et al. examined the photosensitizing activity
of many different dyes in vitro. However, until the 1960s the Fig. 1. A Jablonski diagram showing the various modes of
field of photodynamic therapy lay fallow until a chance excitation and relaxation in a chromophore: (A) excitation; (B)
connection between Richard Lipson and Samuel Schwartz fluorescence; (C) intersystem crossing; (D) phosphorescence;
occurred. Schwartz had isolated a tumor localizing impurity (E) non-radiative transfer of energy to singlet oxygen; (F)
substrate oxidation by singlet oxygen; (G) internal conversion.
from hematoporphyrin preparations that was later named
hematoporphyrin derivative (HpD). Lipson was investigat-
ing how to detect tumor tissue by observing the intratumoral
fluorescence of hematoporphyrin. Unable to obtain repro- [19]. Illuminating the chromophore with light of the
ducible results with hematoporphyrin, Lipson began appropriate wavelength excites it to excited singlet states
experimenting with Schwartz’s HpD. He used it as a tumor (A). The chromophore can relax back to the ground state by
detection agent [13], and recognized that it could be used as emitting a fluorescent photon (B) or to excited triplet states
a photosensitizer to destroy tumor tissue [14]. via intersystem crossing (ISC) (C). From triplet excited
In the 1970s Dougherty rediscovered that fluorescein states the chromophore can relax back to the ground state by
diacetate could photodynamically destroy TA-3 cells in emitting a phosphorescent photon (D) or transferring energy
vitro [15]. Dougherty then began treating tumor-bearing to another molecule via a radiationless transition (E). In the
animals with fluorescein and found that it could work as a first-order approximation, radiative, triplet to singlet transi-
photosensitizer [16]. Weishaupt et al. identified the cyto- tions are quantum mechanically ‘forbidden’ since a change
toxic product of the photochemical reaction to be singlet of electron spins is required, however these transitions occur
oxygen [17]. However fluorescein has a low singlet oxygen to a very small extent [20]. Additionally, the chromophore
quantum yield and a long wavelength absorption in the can also lose energy through internal conversion or
green portion of the electromagnetic spectrum that does not radiation-less transitions during collisions with other mol-
penetrate deeply into tissue. Porphyrin photosensitizers ecules [19] (G). In oxygenated environments the chromo-
were then examined as photosensitizers because they are phore readily transfers its energy to ground-state molecular
efficient singlet oxygen generators and have absorption oxygen (3O2) to produce singlet oxygen (1O2) which can
maxima in the red portion of the electromagnetic spectrum. form adducts with organic substrates [S(O)](F).
Eventually Dougherty rediscovered Schwartz’s HpD, which The photosensitizer and oxygen interact through the triplet
by then was known to have a high singlet oxygen quantum states because oxygen has a unique, triplet-ground state and
yield, an absorption maximum in the red, and is selectively low-lying excited states. The energy required for the triplet
retained in tumor tissue [15]. After several years spent to singlet transition in oxygen is 22 kcal molÿ1, which
isolating and identifying the active fractions of HpD, a corresponds to a wavelength of 1274 nm (infrared light)
purified version named Photofrin1, was produced. In the [21]. Thus relatively low energy is needed to produce singlet
following years Photofrin1 was approved for use in the oxygen. Photochemical reactions of this type are known as
United States against early-and late-stage lung cancers and Type-II photoreactions and are characterized by a depen-
esophageal cancers and dysplasias with other indications dence on the oxygen concentration (O2) (Fig. 2) [22].
pending [15]. At the time of this review it has been used on Although Type-II photoreactions are commonly associated
nearly 10 000 patients in the United States, Canada, with singlet oxygen production some other compounds have
Netherlands, Japan, France, and Italy (and is pending triplet-ground states and can be involved in Type-II
approval for use in eight other countries). In addition to the photoreactions, among these are nitric oxide and vitamin A
approved indications, Photodynamic therapy (PDT) has [22]. In anoxic environments excited photosensitizers (3P*)
been used against bladder cancers, brain cancers, breast can react directly with organic substrates (S) by electron
metastases, skin cancers, gynecological malignancies, exchange producing an oxidized substrate (S‡) and reduced
colorectal cancers, thoracic malignancies, and oral, head photosensitizer (Pÿ). In hypoxic environments the reduced
and neck cancers [18]. Since Photofrin has gained legal photosensitizer (Pÿ) can react with oxygen to produce
status many of the mechanisms by which PDT works have superoxide anions …Oÿ 2 † which. can then form the highly
been elucidated, many new photosensitizers have been reactive hydroxyl radical (OH ) [22]. The excited photo- .
created [18] and research and development continues. sensitizer (3P*) can also react with superoxide radicals …O2 †
ÿ
to produce superoxide anions …O2 † which can then create the
.
highly reactive hydroxyl radical (OH ) [22]. Collectively
1.2 Photochemistry of Photodynamic Therapy
these reactions are classified as Type-I photoreactions and
The photochemical reaction that generates singlet oxygen, are characterized by a dependence on the target–substrate
the putative cytotoxic agent in PDT [17], from ground-state concentration (Fig. 2) [22]. Although Type-II reactions are
oxygen is represented by the Jablonski diagram (Fig. 1) reported to dominate during PDT [23], Foote has suggested

Copyright # 2001 John Wiley & Sons, Ltd. J. Porphyrins Phthalocyanines 2001; 5: 105–129
BASIC PRINCIPLES OF PHOTODYNAMIC THERAPY 107

Fig. 2. Type-I and Type-II photoreactions, where 1P is a


photosensitizer in a singlet ground state, 3P* is a photosensi-
tizer in a triplet excited state, S is a substrate molecule, Pÿ is
reduced photosensitizer molecule, S‡ is an oxidized substrate
molecule, O2 is molecular oxygen (triplet ground state), Oÿ 2 is
the superoxide anion, O2 is the superoxide radical, P‡ is the
oxidized photosensitizer, 3O2 is triplet ground-state oxygen,
1
O2 is oxygen in a singlet excited state, and S(O) is an oxygen
adduct of a substrate. Fig. 3. Molecular orbital diagrams showing the electron
distribution in triplet and singlet oxygen (top). Lewis structure
depicting the zwitterionic character of singlet oxygen (bottom).
Type-I reactions may become more dominant under
conditions where the photosensitizers are highly concen- oxygen’s lifetime is reduced to approximately 2 ms because
trated, and especially under hypoxic conditions [22]. the energy of oxygen–hydrogen (O–H) stretching in water
molecules nearly equals the excited-state energies of singlet
oxygen. The energy is dissipated as heat by the stretching
1.3 Oxygen in PDT
and vibrational motions of water molecules. Because singlet
1.3.1 Singlet Oxygen. Singlet oxygen, the predominant oxygen reacts so rapidly, PDT-induced oxidative damaged
cytotoxic agent produced during PDT [17] is a highly is highly localized to regions no larger in diameter than the
reactive form of oxygen that is produced by inverting the thickness of a cell membrane. Photodynamic damage is
spin of one of the outermost electrons. Normally, ground- probably confined to targets near to or within hydrophobic
state oxygen has two unpaired electrons residing separately regions of the cell due to the hydrophobic character of most
in the outermost antibonding orbitals (Fig. 3). In the absence photosensitizers.
of a magnetic field the electronic configuration is indis-
tinguishable, however in a magnetic field (B) the spins of
the electrons can be revealed to be in one of three possible 1.3.2 Tissue Oxygenation. The importance of tissue
configurations, both spins aligned up, both spins aligned oxygenation in PDT was demonstrated by several research-
down, or one up and one down [24]. When the spins are both ers [25–27]. Hypoxic cells, those with less than 5%
aligned up the oxygen molecule will deflect down in a oxygenation, were found to be resistant to PDT. Henderson
magnetic field, when both spins are aligned down they will et al. have reported that oxygen consumption depends on the
deflect the molecule upwards in a magnetic field, and when fluence rate of the light [28], and that tissue destruction
the spins are anti-parallel the molecule will pass unper- might be enhanced by using lower fluence rates. Lower
turbed through the magnetic field. Because of these three fluence rates do not deplete the tissue’s oxygen supply as
possible states the ground state of oxygen is called a triplet rapidly and as a result the tissue is exposed to singlet oxygen
state. Any molecule with such a valence electron config- for a longer time during treatment [25, 26].
uration is considered to be in a triplet state [19, 24]. d‰1 O2 Š "C DPDT
The extreme reactivity of singlet oxygen arises from the ˆ ˆ  ˆ k‰3 O2 Š …1†
dt E tE
pairing of two electrons into one of the 2p antibonding
The rate of singlet oxygen production …d‰ O2 Š† is shown
1

orbitals (Fig. 3). In the ground state the outermost electrons dt


are distributed, according to Hund’s rule, in the Px and Py mathematically by equation (1) as a function of the light
antibonding orbitals [24]. Since the orbitals are degenerate fluence rate ( ), the total PDT dose (DPDT) and the
and the electrons have aligned spins, the quantum numbers concentration of ground-state oxygen. The rate of singlet
of each electron are identical and thus are forced to occupy oxygen production also varies as a function of the
separate orbitals to comply with the Pauli exclusion concentration (C), the molar extinction coefficient (e), and
principle [21]. During an interaction with the excited the singlet oxygen quantum yield of the compound (Φ).
photosensitizer the spin of one electron inverts, making its Equation (1) shows the rate of oxygen consumption
quantum numbers unique, allowing them to pair together increases linearly with the fluence rate. Clinically, high
into the antibonding orbital which destabilizes the molecule. fluence rates are usually chosen to decrease treatment times
Although singlet oxygen is often depicted as diradical, it is [18]. However it was reported that high fluence rates may
actually a highly polarized zwitterion (Fig. 3, bottom) [21]. not be appropriate for eliciting an optimal effect. Foster et
Singlet oxygen is so reactive that it has a lifetime that al. showed that a fluence rate of 50 mW cmÿ2 produced a
ranges from 10–100 ms in organic solvents. This restricts its greater therapeutic response than 200 mW cmÿ2, when
activity to a spherical volume, 10 nm in diameter, centered tumor-bearing rats were treated with 5 mg/kgÿ1Photofrin1
at its point of production. In an aqueous environment singlet and 360 J cmÿ2 of 630 nm light [29]. Similar results were

Copyright # 2001 John Wiley & Sons, Ltd. J. Porphyrins Phthalocyanines 2001; 5: 105–129
108 I. J. MACDONALD AND T. J. DOUGHERTY

also reported by Reed et al. [30], Tromberg et al. [31], and


Sitnik et al. [32]. Data from the reported experiments and
oxygen diffusion models suggest that lower fluence rates
expose greater portions of the tumor to singlet oxygen for a
longer period of time [29, 31]. Direct measurements of
tumor tissue oxygenation showed a rapid decrease in the
oxygen tension during the first seconds of treatment. The
extent of the decrease during treatment was fluence rate
dependent, as was the treatment response [33]. The degree
of damage to healthy tissue was also reduced, possibly
because healthy tissue can repair minor damage more
efficiently, avoiding a greater inflammatory response [32].
While an ample oxygen supply is necessary during PDT
treatment, eventual destruction of the tumor microvasculature
is also crucial to starve tumor cells of oxygen and nutrients
[30]. Sitnik et al. have suggested that fluence rates could be
judiciously chosen to maximize exposure to singlet oxygen
Fig. 5. Diagram depicting an electron (fuzzy ball) in a region of
during treatment, while the total fluence could be adjusted to length L bounded by infinitely high potential energy barriers.
ensure that the tumor microvasculature eventually collapses,
starving the tumor of oxygen and nutrients [33].

explained by assuming that the delocalized p-electrons in


1.4 Photosensitizers
the aromatic ring behave as particles (fuzzy ball) in a one-
1.4.1 Photophysics of Photosensitizers. Porphyrins, dimensional square-well potential, as described by Hoff-
chlorins, and bacteriochlorins (Fig. 4) are among the most man. The square-well potential consists of a region of length
useful photosensitizers for in vivo PDT, although other L bounded on each side by an infinitely high potential
classes of porphyrinoids such as phthalocyanines and energy barrier (Fig. 5). Equation (2) shows the energy
texaphyrins are also used. Porphyrins, chlorins, and levels, calculated from solutions of the Schrödinger equa-
bacteriochlorins have absorption maxima in the red portion tion. In equation (2), h is Plank’s constant, n is the energy
of the electromagnetic spectrum and are efficient singlet state of the electron (or valence electron in the case of an
oxygen generators [34]. Red absorption maxima allow ensemble of electrons), m is the mass of the electron and L is
activating light to penetrate deeper into tissue. They are the length of the confining region. The term E(n,N) is a
generally planar–aromatic molecules composed of four perturbation to the energy levels due to electron–electron
symmetrically arranged pyrrol units linked by methine interactions when there is more than one electron in the box;
bridges. Although texaphyrins have three pyrrolic units and N is the number of additional particles in the box.
phthalocyanines are linked by azone bridges, both
h2 …n ‡ 1†2
molecules are planar and aromatic and have similar En ˆ ‡ E…n; N † n ˆ 0; 1; 2 . . .
photophysical properties as porphyrins, chlorins and 8mL2
bacteriochlorins. Porphyrins, chlorins and bacteriochlorins N ˆ 1; 2; 3 . . . …2†
are all fully conjugated, but chlorins have one pyrrolic
double bond reduced and bacteriochlorins have two pyrrolic The perturbation term E(n,N) is equal to zero when there is
double bonds reduced. It is because of their extensive only one electron in the potential well but raises the energy
aromatic systems that these molecules absorb light in the of the levels as the number of electrons in the box increases.
visible spectrum [34]. Equation (2) is an great oversimplification but can be used
All porphyrin-like compounds have a strong absorption to illustrate the principle. Fig 6 shows how the energy states
band around 400 nm called the Soret band. Unfortunately accessible to a valence electron rise in the presence of other
this band is not useful for PDT since blue light does not electrons. In this example the energy required to excite an
penetrate very deeply into tissue; thus the weaker satellite electron from the ground state (n = 0) to an excited state
absorption bands (Q-bands) between 600 nm and 800 nm (n = 9) is greater when more than one electron (N > 1) is
are used for treatment. Porphyrins exhibit weak absorption confined in the potential well. Since the energy of light is
maxima around 630 nm while chlorins and bacteriochlorins inversely proportional to the wavelength via the relation
have strong absorption maxima around 650 nm and 710 nm DE = hn = hc/, where n is the frequency,  is the wave-
respectively. The different absorption spectra can be length, and c is the speed of light, the wavelength of light
needed to excite the valence electron is shorter, i.e. bluer
than the light needed to excite the single particle. Porphy-
rins, chlorins, and bacteriochlorins (Fig. 4) have 22-, 20-,
and 18-p electrons in their aromatic rings confined to
approximately the same region in space. Therefore the
absorption maxima of porphyrins are more blue-shifted than
chlorins, and chlorins more than bacteriochlorins because of
increasing numbers of p-electrons interacting in the ring.
All of these classes of compounds are highly efficient
singlet oxygen generators. The efficiency of production of
Fig. 4. The basic chemical structures of porphyrins, chlorins singlet oxygen is an empirically determined quantity called
and bacteriochlorins. the singlet-oxygen quantum yield (Φ). The singlet-oxygen

Copyright # 2001 John Wiley & Sons, Ltd. J. Porphyrins Phthalocyanines 2001; 5: 105–129
BASIC PRINCIPLES OF PHOTODYNAMIC THERAPY 109

Fig. 6. A diagram depicting how increasing numbers of p-


electrons in a ring system cause a given transition to blue shift
to higher energies.

Fig. 7. The synthesis of Photofrin1.


quantum yield describes the number of singlet oxygen
molecules that are formed per photon of energy absorbed.

1.4.2 Photofrin. The most commonly used and studied longer than 630 nm, citing that Photofrin’s relatively weak
photosensitizer to date is Photofrin1, the only commercially absorption band at 630 nm does not allow optimal light
available photosensitizer [18]. Photofrin is easily synthe- penetration. After identifying a useful chromophore it is
sized by reacting hematoporphyrin in a mixture of sulfuric tested in vitro and in vivo for PDT activity. The
and acetic acid to form hematoporphyrin-IX (di- and mono-) pharmacological properties of the new compounds are also
acetates. Hydrolytic treatment with base yields a crude evaluated. Many porphyrin-like compounds allow a large
mixture commonly referred to as hematoporphyrin deriva- number of alterations to be made at the peripheral positions
tive (HpD) [35]. Commercial Photofrin1 is made by of the chromophore [34]. A quantitative structure–activity
removing low-molecular weight components (Fig. 7). The relationship (QSAR) study can be performed to identify
purified fraction was found to be optimally active against optimally active compounds as was done by Henderson et
tumor tissue, and is composed of monomers, dimers, al. with the pyropheophorbide-a derivatives [43]. Ideally, a
trimers, and larger oligomers [35, 36] up to eight or nine photosensitizer should be non-toxic, selectively retained in
porphyrin units [37]. Pandey et al. demonstrated that ether- tumor tissue in high concentrations, water soluble, and
linked hematoporphyrin dimers and trimers, synthesized to cleared in a reasonable time from the body and rapidly from
mimic Photofrin1 components, were more efficacious in the skin to avoid photosensitivity reactions. Although the
mice than carbon–carbon and ester-linked dimers [35, literature is rife with reports of ‘novel’ photosensitizers for
38, 39]. The active HpD fraction that comprises Photofrin1 PDT only Photofrin1 has been legalized by health agencies
still contains many porphyrins oligomers of varying lengths for use around the world [18].
[18, 36, 40], linkages [38, 39, 41], and stereochemistries Several benzoporphyrin derivatives (BPDs) were synthe-
[35, 39]. sized from protoporphyrin [44, 45]. The BPD chromophores
Photofrin’s longest wavelength absorption maximum is a have absorption maxima around 690 nm. A comparative
relatively weak (e630  3000 Mÿ1 cmÿ1) Q-band at 630 nm. study of several BPD analogues showed that the monoacid
Light at 630 nm can adequately activate Photofrin1 to a derivative with the benzo ring attached to the A-ring (BPD-
depth of about 5 mm in tissue. MA) (Fig. 8A) was more photodynamically active than the
Clinically Photofrin1 has exhibited no systemic toxicity diacid or dimethyl ester versions [45]. The pharmacological
and does not appear to be carcinogenic or mutagenic at the properties of BPD-MA are unfavorable for PDT treatment
doses used [42]. Photofrin1 has been used successfully of some cancers because it clears too rapidly from serum
against a wide variety of cancers, does not appear to and tissues including cancerous lesions [46]. Three hours
succumb to multidrug resistance, and Photofrin1-PDT has after injection BPD-MA reaches its maximum tumor con-
no known cumulative dose ceiling as do radiation and centration and quickly decreases thereafter [46]. Although
chemotherapy. Thus patients can be treated repeatedly as an adequate PDT response can be achieved by treating three
needed. Photofrin1 does accumulate in the skin so that a hours after injection, tumor selectivity is low because of the
possible complication can be a severe sun burn or photo- large circulating concentration in all tissues. While BPD-
reaction. These are obviated by avoiding sunlight and high- MA may not be entirely suitable for treating cancer it has
intensity light, and wearing protective clothes and sun- been proven useful for treating the wet form of age-related
glasses for approximately 6 weeks after treatment. macular degeneration (AMD) [47, 48]. Since most of the
BPD-MA remains in circulation at short times after injec-
1.4.3 New Photosensitizers. Many new compounds have tion, laser light can be focused to destroy the neovasculature
been synthesized in an attempt to create a better photo- that grows near the macula and causes blindness in AMD
sensitizer than Photofrin1. Often, researchers examine new [47, 48]. The FDA has recently recommended that BPD-
chromophores with absorption maxima at wavelengths MA, given the tradename Visudyne1 (verteporfin for

Copyright # 2001 John Wiley & Sons, Ltd. J. Porphyrins Phthalocyanines 2001; 5: 105–129
110 I. J. MACDONALD AND T. J. DOUGHERTY

Fig. 8. Photosensitizer candidates for photodynamic therapy.

injection), be approved for use in Visudyne1 therapy, which treatment can be administered within four hours of injection
is essentially PDT to destroy the neovasculature on the [44, 53]. At low doses (<1.65 mg kgÿ1 ‡ 100 J cmÿ2)
retina [49]. tumor regression was short lived while at high doses
Tin etiopurpurin (SnEt2), tradenamed Purlytin1, is a (>2.5 mg kgÿ1 ‡ 100 J cmÿ2) regression was prolonged but
chlorin photosensitizer being investigated for use in PDT tissue selectivity was sacrificed [57]. Skin photosensitivity
(Fig. 8B) [44]. Purlytin1 has a tin atom chelated into its was noted to be ‘temporary’ [57]. This compound is also
center and has an absorption maximum near 650 nm. It has being tested to destroy neovasculature on the retina, and
been reported that Purlytin1 localizes into the skin and can initial results suggests that it can effectively destroy
produce a photoreaction for 7 to 14 days after administration choroidal vessels with minimal damage to the neurosensory
[50, 51]. Purlytin1 has been evaluated for use against retina [58].
recurrent metastatic breast cancer [50] and is currently Lutetium texaphyrin, tradenamed Lutex (Fig. 8F), is a
being evaluated for use in treating corneal neovasculariza- non-porphyrin photosensitizer currently in Phase-II clinical
tion [51]. trials for the treatment of cancer [59]. Lutex has been used
Another chlorin, meta-tetrahydroxyphenyl chlorin (m- to treat malignant melanoma [60] because it is highly
THPC), tradenamed Foscan1 (Fig. 8C), was the most active selective for tumor tissue and it can be activated with
photosensitizer in a series of tetrahydroxyphenyl analogues 732 nm light that penetrates more deeply into pigmented
[44, 52, 53]. Although Foscan1 has a singlet-oxygen lesions than the light needed to excite Photofrin1,
quantum yield comparable to other chlorins, only low drug Verteporfin1, Foscan1, or Npe6. Lutex is also being used
(typically 0.1 mg kgÿ1) and light doses (<20 J cmÿ2) are to prevent re-stenosis of vessels after cardiac angioplasty
needed to achieve photodynamic responses equivalent to [61] by photo-inactivating foam cells that accumulate
Photofrin [18]. Foscan1 is currently under investigation for within arteriolar plaques. For this indication Lutex has been
use in treating esophageal, lung, laryngeal [54], thoracic given the tradename Antrin1.
[55], and skin cancers [56]. Foscan1 can induce photo- A prodrug being used clinically to treat superficial
sensitivity for up 20 days after administration [56]. Foscan1 cancers is D5-aminolevulinic acid (ALA, Fig 8E) ALA is
is so photodynamically active that efforts are underway to a precursor in the biosynthesis of heme [62]. Overloading
treat with less penetrating green light and lower drug doses cells with ALA forces them to produce protoporphyrin-IX
to reduce this compounds phototoxicity [18]. (the photosensitizer) faster than iron ferrochelatase can
Monoaspartyl chlorin e6 (Npe6) is another synthetic convert protoporphyrin-IX to heme, a non-photosensitizer.
chlorin developed to utilize the long wavelength absorption Because the ALA used for treatment is endogenous the
maximum of chlorins (Fig. 8D). Npe6 has been investigated feedback mechanisms are unable to prevent the over-
for use against adenocarcinoma of the breast, basal cell production of protoporphyrin-IX, which accumulates in
carcinoma, and squamous cell carcinoma [57]. Like BPD- high concentrations in the cells making them photosensitive
MA, Npe6 clears rapidly from tissues, therefore light [62]. ALA-PDT destroys superficial disease well but is not

Copyright # 2001 John Wiley & Sons, Ltd. J. Porphyrins Phthalocyanines 2001; 5: 105–129
BASIC PRINCIPLES OF PHOTODYNAMIC THERAPY 111

well suited for destroying tumor tissue greater than one Some attempts have been made to create photosensitizers
millimeter in depth. Aminolevulinic acid has recently been that are specific to tumor cell targets. In vitro, cationic
used successfully in the treatment of actinic keratosis and photosensitizers localize selectively in high concentrations
may have a role in aiding detection of certain bladder within mitochondria which have been shown to be
tumors [62]. Furthermore, esterified ALA derivatives may susceptible PDT targets (see Section 1.6.2.3). Nile blue
have improved pharmacological properties improving derivatives (NB) [68], rhodamine-123 (R-123), chalcogen-
retention in tumor tissue, and thus use as a detection agent, pyrylium dyes, and cationic porphyrins all localize to the
and possibly as a prodrug for PDT treatment [63]. mitochondria, however some NB derivatives, R-123, and
some chalcogenpyrylium dyes have very low singlet-
oxygen quantum yields (<10%). The NB derivatives also
1.4.4 Photosensitizer Pharmacology have low pKa values and are protonated only below pH 5. At
1.4.4.1 Structure–Activity and Quantitative Structure– physiologic pH (7.4) NB derivatives are neutral and have
Activity Relationships. Quantitative structure–activity re- unfavorable absorption maxima around 500 nm [64]. The
lationships (QSARs) and structure–activity relationships largest drawback that most cationic photosensitizers have is
(SARs) have been performed with several series of that they clear very rapidly from tissues [69, 70].
porphyrin and porphyrinoid compounds to identify opti- Highly anionic photosensitizers also appear to be less
mally active derivatives within a series of structurally effective photosensitizers. Generally photodynamic activity
related analogues [43, 45, 64–66]. Structure–activity rela- decreases with increasing numbers of negative charges on
tionship studies are performed to identify structural charac- the chromophore. Mono-acid benzoporphyrin derivatives
teristics that affect the biological activity (other endpoints were 10 times more photodynamically active then di-acid
can also be studied) of a compound [67]. Quantitative benzoporphyrin derivatives in vitro [45]. However, in vivo
structure–activity relationship studies are performed to mono-acid BPD clears so rapidly that treatment must be
quantify the contributions from physicochemical charac- administered three hours after injection (see Section 1.4.3)
teristics that effect the compound’s activity; in this case a [46, 64]. Di and tri-sulfonated phthalimidomethyl deriva-
photosensitizer’s photodynamic activity in biological sys- tives were 200 times more active then the tetra-sulfonated
tems [67]. In QSAR studies it is assumed that the biological derivative, despite the fact that cells accumulated higher
activity of a compound is determined by its ability to bind to levels of the tetra-sulfonated compound [66]. Tetra-
a target receptor. Theoretically, the affinity for the receptor sulfonated zinc phthalocyanine and tetra-sulfonated phtha-
is related to the steric, hydrophobic, and charge interactions limidomethyl aluminum phthalocyanine both required high
between the receptor and the compound [67]. The general drug concentrations to achieve a photodynamic response in
regression equation describing a QSAR takes the form of vivo [65, 66]. Highly anionic compounds appear to be less
equation (3) where BA is the biological activity of the photodynamically active because they are cleared rapidly in
compound which is related to the sum of the contributions vivo [71]. The exception is Photofrin which is a di-acid that
from steric, hydrophobic and charge interactions [67]. is retained to a high degree in tumor tissue and is
X X X
BA ˆ steric ‡ hydrophobic ‡ charge …3† photodynamically active. Unlike the other compounds
described Photofrin1 consists of high-molecular weight
Most of the photosensitizers described in Section 1.4.3 components that are highly aggregated, possibly because of
have been optimized through SAR studies [45, 64, 66] while amphiphilicity as well as the hydrophobicity of the
few have been put through rigorous QSAR studies [43, 65]. components.
Some compounds, such as sulfonated phthalocyanines [65]
and pyropheophorbide-a ether derivatives [43], exhibit 1.4.4.2 Targeted Photosensitizers. Porphyrins have been
relationships between the hydrophobicities and photody- determined to be ligands for the mitochondrial peripheral
namic activities of the compounds. The QSAR exhibited by benzodiazepine receptor (PBR) [72] and a correlation
the sulfonated phthalocyanines [65] shows a dependence on between the photodynamic activities of several porphyrin
the log P values, possibly the result of optimal binding to the photosensitizers and their binding affinities for the PBR was
outer membrane of the cell. The pyropheophorbide-a shown by Verma et al. [73]. The benzodiazepine binding
derivatives are thought to exhibit a parabolic QSAR because site (Site-II) on human serum albumin (HSA) has also been
the optimal members bind most selectively to a protein found to be a binding site for several porphyrins, and
target within the cell [43]. chlorins. A correlation between the binding of certain
A photosensitizer’s structure, charge, and hydrophobicity putative Photofrin1 components to Site-II and photo-
determine how it interacts with itself and its surroundings, dynamic efficacy of the components was also demonstrated
these factors influence how it is accumulated by cells, to by Tsuchida et al. [74]. Tsuchida et al. suggested that HSA
what site, etc. In vivo there are many serum proteins and is not a photodynamic target but that Site-II binds photo-
blood cells with which the compounds can interact. Within sensitizers with a similar affinity as the PBR because Site-II
tumor tissue, pH gradients, irregular leaky vasculature, and and the PBR both bind benzodiazepines [74]. Therefore
abnormal blood flow contribute to where the photosensitizer HSA binding could be used to screen for active photo-
will localize and to what degree. A photosensitizer that is sensitizers.
considered useful has to bypass these pharmacological Some attempts have been made to direct photosensitizers
hurdles to force lethal concentrations into susceptible to known cellular targets by creating a photosensitizer
regions of the tumor tissue. Theoretically photosensitizers conjugate, where the other molecule is a ligand that is
can be rationally designed, but more realistically they are specific for the target. To improve localization to cell
discovered during blind probing. Usually a structure with membranes cholesterol conjugates were synthesized [75].
reasonable photophysical characteristics is found; then the Antibody-conjugates have also been synthesized to direct
pharmacology is tweaked by chemically modifying the photosensitizers to specific tumor antigens [76–78] and
structure. insulin–porphyrin conjugates have been created to target the

Copyright # 2001 John Wiley & Sons, Ltd. J. Porphyrins Phthalocyanines 2001; 5: 105–129
112 I. J. MACDONALD AND T. J. DOUGHERTY

chromophore into cells in high concentrations through a targets within a tumor cell. This will be discussed in detail in
known uptake mechanism [79]. Chemotherapeutic agents the following sections.
have been attached to porphyrin chromophores to increase
the lethality of the PDT treatments [80], and protein- and 1.4.4.4 How Amphiphilicity, Aggregation, and Serum
microsphere-conjugates were made to improve the pharma- Protein Interactions Affect the Photodynamic Activity of
cology of the compounds [81]. These strategies seldom Photofrin1. It has been shown that aggregation changes the
work well because the pharmacological properties of both photophysical properties of porphyrin, chlorin, and phthalo-
compounds are drastically altered. cyanine photosensitizers [85–96]. Aggregation shortens the
triplet-state lifetime and decreases the singlet-oxygen
quantum yield by dissipating the energy through internal
1.4.4.3 Amphiphilicity. Amphiphilicity describes mol- conversion [92–97]. Therefore highly aggregated photo-
ecules that are partially hydrophobic and hydrophilic at sensitizers in vivo are generally considered to be less
different positions so that each region interacts with the effective photosensitizers [87, 98]. Aggregation can some-
solvent independently. No method has been established to times be avoided by using emulsifying detergents such as
quantify amphiphilicity of small molecules, however polysorbates, Cremophore EL1, polyethylene glycol, or
amphiphilicity is analogous to zwitterionicity. Margaron lipid emulsions. Solvents such as DMSO and ethanol can
et al. [65] and others [53] have reported that amphiphilic also be used to dissolve the compounds. Emulsifying agents
photosensitizers are generally more photodynamically increase the solubility of some compounds allowing them to
active than symmetrically hydrophobic or hydrophilic be administered easily [99, 100], however, not much is
molecules. They clearly demonstrated the importance of known about the behavior of these agents in vivo [101–104].
amphiphilicity by comparing the in vitro photodynamic Amphiphilic molecules are thought to bind most
activity of sulfonated tetraphenylporphine derivatives efficiently to proteins at the interface between the hydro-
(TPPS), each derivative having a different amphiphilicity philic surface and hydrophobic interior, thus the interaction
depending on the number of sulfonic acid moieties [65]. The between serum proteins and photosensitizers has been
amphiphilicity of TPPS derivatives (Fig. 9) is readily examined in detail. Photofrin1 has the largest body of
apparent because the phenyl sulfonic acid moieties are fixed pharmacology data related to photosensitizers. While it is
in space (to the meso position of the porphine). It is harder to clear that serum proteins interact with Photofrin1 the exact
determine the amphiphilicity of more complicated struc- nature of the interactions is still not fully understood
tures. Several researchers have suggested that amphiphilic [105, 106].
photosensitizers are more photodynamically active because Since albumin is the most abundant serum protein and is
they can localize to the hydrophobic–hydrophilic interfaces known to associate with many different compounds [107],
in membranes and at the surface of proteins [45, 53, including porphyrins [98, 108, 109], the interaction between
65, 66, 82–84]. Amphiphilicity is also important because it Photofrin and albumin has been examined in detail
can affect the degree to which a compound aggregates [98, 108–115]. In vitro, HpD, Photofrin, hematoporphyrin,
which, in turn, can affect the photophysical properties of a and deuteroporphyin all bound to albumin in solution
photosensitizer. In this thesis amphiphilicity has been [98, 108, 109]. Hepatocytes incubated in vitro with HpD in
identified as a contributing factor in the photodynamic phosphate-buffered saline (PBS) containing one percent
activity of the pyropheophorbide-a derivatives. It appears bovine serum albumin accumulated lower levels of HpD
that amphiphilicity modulates aggregation which in turn than hepatocytes incubated with HpD in PBS with no
determines how the compounds are accumulated to various albumin [110]. The experiments indicate directly and
indirectly that albumin binds these photosensitizers [110].
To determine if binding to albumin is important to achieve
an adequate PDT response Hamblin et al. conjugated
hematoporphyrin to albumin and tested the conjugate in
vivo [106]. Non-phagocytic cells (e.g. tumor cells) accu-
mulated small amounts of the hematoporphyrin–albumin
conjugate on their outer membranes but did not accumulate
any intracellularly [106]. However, the hematoporphyrin-
albumin conjugate was very efficiently accumulated by
macrophages [106].
Lipoproteins are the next most abundant serum proteins
after albumin. Kessel reported that in vitro HpD initially
binds to albumin but rapidly redistributes to lipoproteins
long before light treatment would occur in vivo [112]. Jori et
al. also showed that HpD was undetectable in the albumin
fraction of the serum 48 hours after administration in vivo
[111]. Other researchers have used electrophoresis, ultra-
centrifugation and spectroscopic analysis to confirm that
lipoproteins are the main reservoir for circulating Photo-
frin1 [98, 116].
The apparent reason that lipoproteins have a high affinity
for photosensitizers (especially amphiphilic photosensiti-
zers) is that they are composed of a core of 1500
cholesteryl esters encapsulated in a monolayer made of
Fig. 9. The structures of sulfonated tetraphenylporphines. 800 phospholipid and 500 cholesterol molecules

Copyright # 2001 John Wiley & Sons, Ltd. J. Porphyrins Phthalocyanines 2001; 5: 105–129
BASIC PRINCIPLES OF PHOTODYNAMIC THERAPY 113

[117, 118]. The apoprotein is approximately 50 kD and Early in the distribution phase the most-soluble compo-
wraps around the ‘equator’ of a hydrophobic core. With a nents disaggregate and bind with albumin, possibly because
total molecular weight of 150 kD, lipoproteins are less it is the most-available binding site. Simultaneously, the
than 50% protein by weight and are more like hydrophobic concentration of soluble components decreases because of
reservoirs. clearance mechanisms. Since clearance is an irreversible
The interaction between Photofrin1 and lipoproteins and sink the system continuously shifts towards unbound
the effect this interaction has on the photodynamic proper- molecules to preserve equilibrium. As a result no Photo-
ties of Photofrin1 was examined by Korbelik who exposed frin1 can be detected in the albumin fraction 10 hours after
cultured cells to Photofrin1 and high-density lipoprotein administration in vivo [111]. The less-soluble oligomers
(HDL), low-density lipoprotein (LDL), or very low-density disaggregate more slowly and can bind to albumin
lipoprotein (VLDL) [113]. Summarizing Korbelik’s results, (especially if albumin is the only species present) [98,
the cells accumulated less Photofrin1 in the presence of the 114], however it is more likely that they will partition into
lipoproteins than in the presence of albumin [115]. After lipoproteins because the final free energy is lower than if
incubating the cells for 24 hours with Photofrin, and they were bound to albumin [112]. Although lipoproteins
lipoproteins or albumin, efflux of Photofrin1 was most make up a smaller portion of serum than albumin they are
enhanced by medium-containing albumin [115]. Interest- considerably larger, more hydrophobic, and have a large
ingly, in both experiments the globulin fractions of the capacity for hydrophobic molecules [112, 117, 118]. At the
serum bound the Photofrin1 more tightly than the time of light treatment, 24–48 h after injection, mainly
lipoproteins; however globulins comprise a much smaller lipoprotein-bound oligomers and some large aggregates
fraction of the serum. remain in circulation [112, 121, 122]. Interestingly, macro-
Tumor cells display higher numbers of LDL receptors on phages have a nearly unsaturable capacity for LDL (and the
their surfaces, presumably because they require more photosensitizer bound to it) thus it has been speculated that
cholesterol than normal cells to build membranes [119]. macrophages accumulate high concentrations of Photofrin
Hamblin et al. created hematoporphyrin–lipoprotein con- and then localize in the skin causing skin photoreactions
jugates to direct the photosensitizer into cells through the [105, 123].
LDL receptor-mediated pathway. The VLDL and HDL The pathways by which Photofrin1 is accumulated by
conjugates were not accumulated by cells and the LDL cells is not well understood, however Maziere et al. and
conjugate was accumulated to a lesser extent than hemato- Hamblin et al. demonstrated that a small amount of
porphyrin alone [106]. Significantly, Hamblin et al. reported material, bound to LDL, is internalized by the LDL-
that minor alterations to the LDL complex greatly increased receptor-mediated pathway [105, 123]. However the LDL-
the probability that it would bind to scavenger receptors on receptor-mediated pathway does not appear to be the main
macrophages and become phagocytosed [105, 106]. Co- route of internalization. This was demonstrated by Moan et
valently binding hematoporphyrin to transferrin, a serum al. who reported that cultured cells accumulate Photofrin1
protein with a known uptake pathway also did not increase components with a degree of selectivity that does not
the cellular uptake of hematoporphyrin [106]. Korbelik’s correlate with their hydrophobicity or their affinity for LDL
and Hamblin’s results demonstrate that lipoprotein binding [121, 124]. Therefore the main routes by which cells
is important. However their results also show that Photo- accumulate Photofrin1 components may be by passive
frin1 is not significantly accumulated by the LDL receptor- diffusion through the outer membrane or some other
mediated pathway. receptor-mediated transport process that is highly dependent
Summarizing the reports cited, a description of the on the aggregation state of the molecules. This will be
pharmacokinetics and pharmacodynamics of Photofrin1 discussed in more detail in the context of the current study.
can be constructed. Photofrin1 consists of many compo-
nents, each having different hydrophobicities and amphi- 1.4.4.5 How Amphiphilicity, Aggregation, and Serum
philicities. Therefore, it is reasonable to assume that each Protein Interactions Affect the Photodynamic Activity of
component behaves independently from the others. Bellnier Other Photosensitizers. Allison et al. demonstrated that in
et al. fitted a triphasic exponential curve to the pharmaco- the presence of serum proteins BPD-MA disaggregates and
kinetic profile of Photofrin1, however it is likely that the binds to lipoproteins [125]. They also showed that BPD-MA
curve is actually an n-phasic curve where n is the number of was more photodynamically active if it is non-covalently
components [120]. Each component has two factors that bound to LDL or HDL before use in vivo [126, 127]. When
control its behavior in solution, the rates of aggregation and BPD-MA was bound to LDL it was accumulated to higher
disaggregation (related to the free energy [DG‡] barrier to levels by cultured tumor cells and living tumor tissue than
form the transition state) and the total free energy released when it was administered alone [127]. Allison et al. bound
upon binding (DG). This is shown graphically in Fig. 9 BPD-MA to acetylated LDL, which cannot bind to the LDL
where f is a reaction coordinate. In Fig. 9 four possible receptor, and demonstrated that the amount of drug
situations are shown: (A) kinetically favorable, thermo- accumulated by the cells was markedly reduced compared
dynamically favorable; (B) kinetically unfavorable, thermo- to the normal LDL-associated BPD-MA [127]. Unfortu-
dynamically favorable; (C) kinetically unfavorable, nately, no direct comparison was made between the
thermodynamically unfavorable; and (D) kinetically favor- intracellular levels of BPD-MA and the corresponding
able, thermodynamically unfavorable. The Photofrin1 photodynamic activity to assure that greater photodynamic
components are probably described by these four examples activity resulted from higher intracellular levels of BPD-
and many other possible combinations. Therefore two MA [127].
components may have identical log P values [related to The interaction between sulfonated tetraphenylporphines
DG by DG = ÿ RT ln (P)] but have different pharmaco- (TPPS) has also been examined in some detail. Although
logical properties because they have different amphiphili- these compounds cannot be used to treat human cancer
cities [related to DG‡ by]. patients because they are neurotoxic, they are interesting

Copyright # 2001 John Wiley & Sons, Ltd. J. Porphyrins Phthalocyanines 2001; 5: 105–129
114 I. J. MACDONALD AND T. J. DOUGHERTY

because they have well-defined amphiphilicities and are a especially lipoproteins. Glanzmann et al. and others have
useful model to study how amphiphilicity and hydrophobi- reported that m-THPC exhibits an unusual pharmacokinetic
city determine photodynamic activity (Fig. 10) [128]. The profile in vivo [132, 133]. Initially, half of the m-THPC
TPPS1 derivative has one sulfonic acid group while TPPS2a administered disappears from circulation gradually re-
has two at opposing meso positions and TPPS2o has two at appearing in the serum after about 10 h and decreases
adjacent meso positions. The TPPS3 and TPPS4 derivatives thereafter [132, 133]. The shape and extinction coefficient
have three and four sulfonic acid groups respectively. The of the absorption maxima and the fluorescence signal of
most interesting compounds in this series are the TPPS2 m-THPC, by comparison with disaggregated m-THPC, did
derivatives because they have identical log P values and not change during the time that the drug disappeared
very different tumor localizing and photosensitizing proper- indicating that the disappearance is not an artifact of
ties in vitro [128]. The TPPS2a derivative is the most aggregation or binding to serum proteins [133]. Glanzmann
photodynamically active member of this series followed by et al. and others have suggested that m-THPC is initially
the 2o-, 1-, 3- and 4- derivatives, respectively. The 2a- sequestered in the liver and then released over time. Others
derivative is accumulated the slowest, and in the highest [134] have suggested that m-THPC metabolites are released
concentrations by cells in vitro. It also has the highest from the liver, although Glanzmann et al. have suggested
affinity for lipoproteins and a low affinity for albumin. In that this is not the case and have shown that the reappearing
contrast the 4-derivative binds mainly to albumin and does fraction is as photodynamically active as m-THPC [133].
not bind to lipoproteins. It is accumulated to the highest Bellnier et al. have shown that when the hexyl-ether
levels by tumor cells and is considered to be the most tumor derivative of pyropheophorbide-a is administered in
selective compound in the series; however it is also the least ethanol, as opposed to 1.0% (v/v) Tween-80, the same
photodynamically active [129]. pharmacokinetic profile as that of m-THPC was obtained in
Phthalocyanines and naphthalocyanines are classes of mice [134]. Bellnier et al. suggested that ethanol stimulates
photosensitizers that also have clearly defined amphiphili- the liver to rapidly accumulate the compound which is then
cities. They can be used as models, like the TPPS released back into circulation over time [134]. Since m-
derivatives, because they can have up to four sulfonic acid THPC is formulated for clinical use in 20% (v/v) ethanol,
groups, one at each pyrrol position. Di-sulfonated phthalo- 30% (v/v) polyethylene glycol, and 50% (v/v) water, it may
cyanines also have two configurations either on opposite be possible that ethanol is responsible for the unusual
sides of the ring or at adjacent pyrrolic positions. Berg et al., pharmacokinetic profile.
Brasseur et al., and Margaron et al. have demonstrated that In the previous section the pharmacological properties of
the 2a-sulfonated phthalocyanine is the most efficient Photofrin1 were described. They consisted of independent
photosensitizer in vitro followed by the 1-, 3-, 2o-, and actions of many different species, having various combina-
4-derivatives [65, 130, 131]. This is the same trend tions of hydrophobicity and amphiphilicity. The informa-
exhibited by the TPPS series. Other phthalocyanine and tion presented in this section shows that amphiphilic
naphthalocyanine derivatives exhibit the same general photosensitizers, even among varied classes of compounds,
trend; the more amphiphilic they are the more photo- are more photodynamically active than symmetrically
dynamically active they are. hydrophobic or hydrophilic ones suggesting that the most
Other new photosensitizers, m-THPC and Npe6 [113, active fractions of Photofrin1 may be those that are most
115, 132] were also found to interact with serum proteins amphiphilic.

1.5 Light
1.5.1 PDT Dose. Selective tumor destruction can be
achieved if the amount of photosensitizer accumulated in
tumor tissue is greater than that accumulated in the
surrounding healthy tissues. Selectivity can also be
enhanced by selectively delivering the light to the tumor
tissue and excluding it as fully as possible from the healthy
tissues nearby. Star et al. and Svaasand et al. have stated the
PDT dose mathematically (equation (4)) as the product of
the light fluence rate [ (,q,t)], the molar-extinction
coefficient of the photosensitizer ["()], and the photo-
sensitizer concentration in the region of treatment [C(q,t)]
[135, 136]. In equation (4), t is the illumination time, q is a
generalized spatial coordinate, and  is the wavelength of
the activating light.
Z Z
DPDT ˆ "…† C…q; t† …; q; t†dtdq …4†
0 q

The PDT dose is a measure of the energy absorbed by the


Fig. 10. The diagrams depict four possible interactions
between photosensitizer molecules and serum proteins: photosensitizer in a volume of tissue. Theoretically the PDT
(A) thermodynamically favorable, kinetically favorable; dose could be calculated exactly if the fluence rate and drug
(B) thermodynamically favorable, kinetically unfavorable; concentration and the extinction coefficient are known at
(C) thermodynamically unfavorable, kinetically unfavorable; every point and at all times in the treatment volume
(D) thermodynamically unfavorable, kinetically favorable. [135, 136]. Equation (4) is general so that the target

Copyright # 2001 John Wiley & Sons, Ltd. J. Porphyrins Phthalocyanines 2001; 5: 105–129
BASIC PRINCIPLES OF PHOTODYNAMIC THERAPY 115

concentration C(q,t) could pertain to the amount of Equation (7) shows why there is a window around 800 nm
photosensitizer in a volume of tissue, or to smaller regions where light penetrates most deeply into tissue. At
such as a cell, organelles within the cell, or even specific wavelengths longer than 800 nm water molecules begin to
proteins. Equation (4) shows that if a photosensitizer absorb the light and convert its energy to heat. Also highly
localizes to a target with a concentration (Ct), localizes to pigmented tissues can dramatically reduce the depth that
regions around the target with a concentration (C), the PDT light can penetrate into tissue. This was demonstrated by
dose to the target will be scaled by a factor equal to the ratio treating separate groups of black and white guinea pigs with
of the two concentrations (Ct/C). In reality photosensitizer graded doses of dihematoporphyrin ether (DHE)-PDT until
concentrations usually differ only slightly between tumor they achieved equal responses. White guinea pigs required
and healthy tissue; therefore the light must be administered 26 J cmÿ2 of green light to produce eschars while black ones
as specifically as possible with fiber optics and specialized needed 58 J cmÿ2 because of the high absorptivity of
lenses. melanin [143]. Dougherty et al. also reported that the
photosensitizer can decrease the attenuation depth if it has a
1.5.2 Light Penetration. Svaasand [137], Wilson et al. high extinction coefficient and is present in high concentra-
[138], Bolin et al. [139] and others have demonstrated that tions in tumor tissue [141]. In equation (7) the reduced
light with a wavelength near 800 nm passes most freely absorptivity has been expanded (equation (6)) to include the
through tissues. This is illustrated nicely by placing a absorptivity of endogenous chromophores and the concen-
flashlight to one’s cheek and observing the red glow on the tration of the photosensitizer (C), the distance from the light
inside surface (personal observation). We see the red source (r), and the molar extinction coefficient (e).
because red-light penetrates most deeply into tissue, and  ˆ f3…ma ‡ "Cr†‰m0s ‡ …ma ‡ "Cr†Šgÿ1=2 …7†
not due to the color of blood as is sometimes thought. The
depth that light penetrates into tissue is dependent upon the Increases in any one of these values without an equivalent
optical properties of the tissue and the wavelength of the decrease in the others will result in a decrease in the
light [137–140]. When photons enter tissue a portion is attenuation depth [141].
reflected by the surface and the rest scatter around in the
tissue until they either escape or are absorbed by chromo-
phores (either endogenous tissue chromophores or exogen- 1.5.3 Photobleaching. Photobleaching is destruction of
ous molecules such as the photosensitizer) [135, 137, 138, the photosensitizer by light-mediated processes. It was
140, 141]. Theoretically a few photons may pass completely observed that light degrades photosensitizers in solution
through the tissue volume although this number is very according to equation (8), where b is the experimentally
small [140]. Wavelengths less than 800 nm are scattered determined decay constant that is specific for each
with increasing efficiency by macromolecules because they compound, and J is the total fluence (J = t) [144].
are equal to or less than the size of the particles [140]. C ˆ C0 eÿ J …8†
The attenuation coefficient [a (cmÿ1)] describes the
exponential decrease of a specific wavelength of light as it Generally photobleaching is measured as the decrease in the
penetrates tissue and is different for each tissue [137–140]. fluorescence over time since fluorescence intensity and
The fluence rate of light at a depth (r) from a point light concentration are proportional [144]. Measuring photo-
source within a volume of scattering medium with an bleaching in vivo is more difficult and less reliable because
attenuation coefficient a is given by equation (5) and was aggregation, clearance mechanisms, and changes in the
derived from diffusion theory and describes experimental optical properties of the tissue can also decrease the
data well [135, 136]. observed fluorescence [64, 145]. Photobleaching has been
shown to be mediated by singlet oxygen [146, 147],
eÿ r independent of oxygen [64, 148], or indifferent to either
ˆ 0 …5†
r condition [147]. In aqueous solutions Photofrin photo-
Fluence rates decrease exponentially with depth. Thus small bleaches very slowly, presumably because it is so highly
changes in the attenuation coefficient lead to large changes aggregated. Since in the aggregated state Photofrin1 has a
in the fluence rate with depth. Therefore it is advantageous low singlet-oxygen quantum yield it was assumed that
to have photosensitizers that absorb near 800 nm to Photofrin1 was bleached by an oxygen-dependent mech-
maximize the treatment depth in many different tissues. anism [64]. However, Photofrin1 bleached rapidly in de-
The inverse of the attenuation coefficient is the attenuation oxygenated solutions suggesting that a Type-I photoreaction
depth [ (cm)] which describes the depth that light entering may mediate photodestruction. In the first experiment,
tissue is reduced to 1/e (approximately 37%) of its initial aggregation probably inhibits photobleaching by inhibiting
intensity (equation (6)). Typical values of the attenuation Type-I photoreactions. Dougherty suggested that in bio-
depth range from 1 to 3 mm for non-pigmented tissues logical systems proteins and other cellular components
[137–140, 142]. Dougherty has reported that the destructive disaggregate Photofrin1 so that photobleaching may be an
effects of PDT extend to two attenuation depths or about 5 important factor in the PDT response [64]. Experimentally it
to 6 mm, indicating that as little as 10 percent of the entering was determined that Photofrin was nearly 60% photo-
light is required for a photodynamic effect [64]. Although bleached after a typical PDT treatment in vivo.
the attenuation depth is dominated by scattering at the While oxygen-mediated photodestruction of photo-
wavelength used for PDT, the attenuation depth actually sensitizers has been theoretically and experimentally
depends on the reduced scattering coefficient (m's) and the demonstrated, most experiments suggest that Type-I photo-
reduced absorptivity coefficient (ms) (equation (6)) [64, degradation dominates [146].
135, 138, 140–142]. Although photobleaching decay constants for several
porphyrins have been measured in solution, in vitro, and in
 ˆ ÿ1 ˆ ‰3ma …m0s ‡ ma †Šÿ1=2 …6† vivo, there is no agreement as to whether photobleaching has

Copyright # 2001 John Wiley & Sons, Ltd. J. Porphyrins Phthalocyanines 2001; 5: 105–129
116 I. J. MACDONALD AND T. J. DOUGHERTY

any discernible effect on PDT activity [144, 145]. It is derived growth factor are vasodilators and inhibit platelet
thought that photobleaching may rapidly destroy large aggregation. They are produced simultaneously with
amounts of the drug that is localized in the healthy tissue endothelin-1, an agent that sensitizes endothelial cells to
surrounding the tumor. If this happens fast enough the vasoconstricting stimuli [158]. When some event exposes
normal tissue should be spared from some of the photo- the basement membrane to the blood serum, in this case
dynamic damage thus increasing tumor selectivity. There is PDT [159], cascades of eicosinoids tip the system towards
also some anecdotal evidence that suggests that low level vasoconstriction. In the region of injury vessel constrictions
illumination or controlled illumination of patients could form. Platelets and neutrophils adhere to the vessel wall, roll
effectively photobleach Photofrin out of the skin and that toward the constriction and aggregate, at which point they
this is the mechanism by which Photofrin is finally removed migrate into the surrounding tissues following chemokine
from the skin [149, 150]. There is no clear proof that gradients [160].
photobleaching affects PDT treatments [136]. McMahon et al. showed that thromboxane and other
vasoactive leukotrienes were produced in large quantities in
1.5.4 Light Sources and Delivery. The activating light is vivo after Photofrin1-PDT. McMahon et al. showed that the
most often generated by lasers, or in some cases by arc inflammatory response induced by Photofrin1-PDT is not
lamps or fluorescent light sources. Lasers are most general since other photosensitizers do not necessarily
frequently used because they produce highly coherent produce the same degree or type of vascular response [161].
monochromatic light that can be efficiently channeled into Although NPe6-PDT produced blood stasis it was mainly
quartz fibers used as light delivery devices. Since the lasers, due to platelets aggregated on the artery walls. Throm-
their power sources and cooling apparatus are often noisy, boxane was not detected and arteriolar constriction was not
bulky, and labor intensive they usually cannot be used in the observed after NPe6-PDT. Although not confirmed directly,
operating room or clinic. Therefore light is usually delivered McMahon et al. suggested that the vascular response
through fiber-optic cables to the treatment site, often observed after Npe6-PDT resulted from the lack of
through an endoscope. The fibers can also be tipped with thromboxane production. They suggested that NPe6-PDT
a variety of lenses and diffusers that allow optimal stimulates the excretion of prostacyclin which inhibits
illumination of the tissues to be treated. For example thromboxane production [161]. Although the mechanisms
hollow, spherical organs (e.g. bladder) can be illuminated are unknown, SnEt2 produces an inflammatory response but
from within with spherical diffuser bulbs, while tubular does not cause vessel constriction or platelet aggregation
regions (e.g. esophagus) can be illuminated with cylindrical [161].
diffuser fibers of varying lengths [151, 152]. Graded index Many eicosinoids are derived from arachidonic acid,
(GRIN) lenses can be used to illuminate flat surfaces [28]. which is produced by phospholipases that hydrolyze
Conveniently the thin fiber-optic cables can also be fed phospholipids stored in cells [160, 162–164]. Arachidonic
through the existing channels in bronchoscopes and acid is converted to prostaglandins by prostaglandin
endoscopes eliminating the need for more complicated endoperoxide synthase, which catalyzes the addition of
surgical procedures. In the future light patches may be two molecules of oxygen to arachidonic acid [165].
available to illuminate surfaces even more effectively. Agarwal et al. demonstrated that phospholipases C and A2
Inexpensive diode lasers are currently in use in Europe and were active in RIF-1 cells in vitro following Photofrin1-
may soon be available in the United States, eliminating the PDT [166]. Inhibiting these phospholipases reduced the
labor intensive argon-pumped dye lasers. tendency for these cells to undergo apoptosis (See Section
1.6). Fingar et al. also showed that Indomethacin, which
inhibits the cyclooxygenase activity of prostaglandin
1.6 Biological Mechanisms of Action
endoperoxide synthase, decreases the amount of throm-
1.6.1 Vascular Shutdown and Inflammation. Although it boxane produced after Photofrin1-PDT, in turn reducing
was demonstrated that HpD-PDT killed tumor cells in vitro, the PDT response by inhibiting vessel constriction [154].
Henderson et al. reported vascular destruction occurred to a Although vascular destruction is generally considered to
greater extent in vivo [153]. Blood stasis and hemorrhaging be one of the major contributing effects, direct cytotoxic
were found to starve tumor cells of oxygen and nutrients. effects are not entirely unimportant and are described more
This was confirmed by Fingar et al. who treated tumor- fully in the next section [155].
bearing animals within minutes of injecting Photofrin1, to
confine photodynamic damage to the vasculature, and 1.6.2 Direct Cell Effects. Originally PDT was thought to
observed pockets of necrotic-tumor cells remote from the destroy tissue by damaging cells which had accumulated
damaged vessels [154]. Henderson et al. also demonstrated enough photosensitizer and received enough light to
that tumor bed effects (destruction of the vasculature) are produce lethal amounts of singlet oxygen. This had been
crucial to reducing the survivability of EMT-6 and RIF cells demonstrated effectively in vitro by Christensen et al. [167],
in vivo [155]. Henderson et al. also showed directly, by Berns et al. [168], Salvatore et al. [169] and others.
measuring oxygen concentrations in tumor tissue, that Henderson and others showed (see previous section) that a
oxygen deprivation kills tumor cells after PDT [156]. greater percentage of tumor cells in vivo die after PDT
Korbelik et al. reported that vascular destruction, treatment because vascular shutdown starves them of
observed after PDT is similar to the inflammatory response oxygen and nutrients [153–156]. However, Henderson
observed after tissue injury or infection [157]. Thus acknowledged that direct cell cytotoxicity is not entirely
eicosinoids and other inflammatory agents have been unimportant [155, 156]. Fingar et al. confirmed this by
examined as mediators of the PDT response. Endothelial injecting DBA mice bearing SMT-F tumors with 5 mg kgÿ1
cells normally produce a balance of vasoactive and of Photofrin1 24 h before light treatment or 0.1 mg mlÿ1
vasoconstrictive substances that maintain a healthy vascular Photofrin1 5 min before treatment. Both doses achieved
tone [158]. For example, prostacyclin and endothelium- equivalent circulating concentrations of Photofrin1

Copyright # 2001 John Wiley & Sons, Ltd. J. Porphyrins Phthalocyanines 2001; 5: 105–129
BASIC PRINCIPLES OF PHOTODYNAMIC THERAPY 117

(1.2 mg mlÿ1), except that the 24 h delay allowed time for carcinogenic potential of Photofrin1 and aluminum
the photosensitizer to accumulate in tumor cells while the phthalocyanine-mediated PDT to X-ray irradiation in
five-minute delay confined most of the drug in the B-lymphoblast cell lines in vitro [42]. Both cell lines are
vasculature. The five-minute treatment produced no tumor p53 deficient but one was transfected with a functional p53
responses while the 24-hour treatment elicited curative gene. The cells with functional p53 genes did not accu-
responses, proving that some degree of cellular damage was mulate the numbers of mutations as did the p53 mutants. In
important to obtain complete tumor destruction [170, 171]. either case the mutation potential of PDT was determined to
Tumor responses could be obtained with the short time be insignificant compared to that of X-ray irradiation [42].
treatment regimen by using doses as high as 10 mg kgÿ1 of Recently McNair et al. examined the ability of HpD,
Photofrin, however selectivity was poor and damage to the m-THPC, and methylene blue to damage DNA in vitro. No
surrounding tissues was extensive [170, 171]. DNA damage was detected after m-THPC-PDT, but minor
Wilson et al. [172], Sharkey et al. [173], and Adams et al. DNA damage was observed after HpD-and methylene blue-
[174], have used Photofrin1-PDT-resistant RIF cells to PDT. The damage from methylene blue and HpD was
demonstrate the importance of direct cytotoxicity. Although transient and was resolved 4 h after treatment [180]. Thus
these cells were created in vitro, they were also found to be the published reports suggest that DNA damage is not
resistant to PDT in vivo although the vascular damage initially a major contributor to the PDT response (DNA
produced by the treatments was similar [174]. alterations will be a major factor during the apoptotic
response).
1.6.2.1 DNA Damage. Initially it was thought that the
DNA could be susceptible to photodynamic modification. 1.6.2.2 Lysosomal Damage. Many porphyrinoid photo-
Gutter et al. and Fiel et al. showed DNA damage could be sensitizers can be observed intracellularly with a fluores-
observed after HpD- or cationic-porphyrin-mediated PDT in cence microscope. Most photosensitizers localize to
vitro [175, 176]. Modification of the guanine moiety was cytosolic targets such as the Golgi, endoplasmic reticulum,
most commonly observed [175]. Indirect evidence of PDT- mitochondria, lysosomes, and membranes [172, 181–185].
induced DNA damage was demonstrated by Gomer et al. The localization sites, as reported in the literature, of the
who noted decreased cell survivability in cells with retarded most commonly studied photosensitizers are shown in
DNA-damage repair mechanisms in vitro [177]. However, Table 1.
when Gomer compared the PDT-DNA damage to X-ray- The data in Table 1 show that many photosensitizers
induced DNA damage, after equitoxic doses, there was less localize to the lysosomes, thus it was suggested by Berg et
DNA damage from PDT treatment than from X-ray al. that the lysosomes were highly susceptible PDT targets
treatment. Moan et al. and Evensen et al. also showed that [194]. Okada et al. demonstrated that lysosomal disruption
X-ray treatment produced 80% more strand breaks, 5% was not directly cytotoxic [195] but Geze et al. [183] and
more sister-chromatid exchanges, and more chromatid others [196–198] suggested that hydrolases and acid that
aberrations than PDT treatment. Furthermore PDT-induced leak out of the damaged lysosomes may degrade cellular
strand breaks were repaired more efficiently than those components. However Berg et al. showed that TPPS-
caused by X-ray irradiation [178, 179] and the number of mediated PDT disrupted lysosomes but did not significantly
chromatid aberrations did not correlate with the photo- decrease cell viability [194]. The activities of the lysosomal
toxicity of HpD [179]. Evans et al. compared the enzymes b-N-acetyl-D-glucosaminidase (b-AGA) and ca-

Table 1. Intracellular localization sites of various photosensitizers

Photosensitizer, dose, (incubation time) Cell line Localization target Ref.

Npe6, 15 mg mlÿ1 (24 h) Rat Heart, CHO Lysosomes [186]


TPPS2o, S2a, S4, 30 (40 h), 3.0 (18 h), 75 mg mlÿ1 (18 h) NHIK3025 Lysosomes [128]
TPPS1, 5 mg mlÿ1 (18 h) NHIK3025 Perinuclear [128]
ÿ1
TPPS1, 1 mg ml (18 h) CT26 murine colon Perinuclear [187]
carcinoma
AlPcS4, 75 mg mlÿ1 (18 h) NHIK3025 Lysosomes [128]
AlPcS4, 10 mM (4 h) RIF Lysosomes [188]
ÿ1
AlPcS4, 50 mg ml (18 h) V79 Lysosomes [189]
AlPcS2, S1 40 mg mlÿ1 (18 h) NHIK3025 Cytoplasm/lysosomes [128]
ALPcS2, 25 mg mlÿ1 (5 h) V79 Lysosomes [189]
SnEt2, 2 mM (15 min) P388 Lysosomes [190]
ÿ1
BPD-MA, 1 mg ml (4 h) NHIK3025 ER/Golgi [191]
m-THPC, 0.5 mg mlÿ1 V79 Diffuse cytoplasm [52]
ÿ1
Lu-Tex, 20 mg ml (4 h) EMT-6 Lysosomes [192]
1
Photofrin Outer membrane then mitochondria [193]

Copyright # 2001 John Wiley & Sons, Ltd. J. Porphyrins Phthalocyanines 2001; 5: 105–129
118 I. J. MACDONALD AND T. J. DOUGHERTY

thepsins L and B were also measured in the cytosol of unaffected by PDT, probably because the active portion of
cultured cells after photodynamic treatment with TPPS the protein juts out into the cytosol. It was further
derivatives. The TPPS4 derivative localizes mainly to the demonstrated that the susceptibility of the protein to PDT
lysosomes (Table 1), however after PDT treatment no extra- correlates with the degree to which it is embedded into a
lysosomal enzyme activity was detected. Some cytosolic membrane. Cytochrome c oxidase resides within the inner
enzyme activity could be detected after treating the cells mitochondrial membrane, succinate dehydrogenase is
with TPPS-1 or -2a, but only after non-lethal doses were anchored in the inner mitochondrial membrane but
used. Apparently cytosolic hydrolase inhibitors and the pH protrudes out into the mitochondrial matrix, and malate
change (7.3 compared to 4.2 in the lysosomes) block the dehydrogenase floats free in the mitochondrial matrix.
enzyme activity. Lastly, Berg suggested that the enzymes Cytochrome c oxidase was inhibited to a greatest extent
detected might have been actually released during the followed by succinate dehydrogenase which was inhibited
extraction procedure rather than during PDT [194]. It was to a greater extent than malate dehydrogenase by HpD-PDT
concluded that TPPS-mediated PDT does not kill cells by [82]. Hilf suggested that cytochrome c oxidase was exposed
damaging lysosomes or facilitating the release of hydrolytic to the highest porphyrin concentrations because it is
enzymes or acid [194]. embedded in a hydrophobic environment where active
To reconcile the inconsistency between lysosomal HpD components tend to localize.
localization and cellular insensitivity to lysosomal damage Gibson et al. compared the photosensitizing ability of
Moan et al. suggested that lysosomes are a primary HpD, hematoporphyrin (Hp), hydroxyvinyldeuteroporphyin
localization site but not the photodynamic target [189]. (HVDP), and protoporphyrin-IX (PP-IX) to the degree to
Moan explained that upon illumination the lysosomes may which they inhibited cytochrome c oxidase activity in vitro
rupture and the photosensitizer could relocalize to more [200]. Reduced mitochondrial functioning and cytochrome
susceptible organelles such as the membranes of the c oxidase activity were observed after HpD, HP and HVDP-
mitochondria or nuclei. Moan supported this theory by mediated PDT, but not after PP-IX-PDT. These results
demonstrating that di- and tetra-sulfonated aluminum further supported Hilf’s hypothesis that cytochrome c
phthalocyanines (AlPcS2 and AlPcS4) exhibit large in- oxidase is a susceptible target since the degree of inhibition
creases in their fluorescence intensities within cells after the matched the in vivo efficacies of these compounds [200].
cells were treated with low doses of light. Presumably the A protein complex, termed the permeability transition
compounds are disaggregated, thus more fluorescent, after pore (PT pore) [72, 201, 202], that consists of hexokinase,
being released from the lysosomes. Relocalized into new the peripheral-benzodiazepine receptor (PBR), the voltage-
cellular targets the compounds are more photodynamically dependent anion channel (VDAC or porin), creatine kinase,
active because they are disaggregated and have higher the adenine nucleotide translocator (ANT), and cyclophyllin
singlet oxygen quantum yields. However, it has not been D [202] has been proposed to be a susceptible PDT target
determined whether the lysosomes rupture after photody- [28, 203]. Miccoli et al. reported that photodynamically
namic damage and release the compounds or whether some damaging hexokinase may inhibit the enzyme’s ability to
other cellular damage, caused by unobservable fractions of phosphorylate glucose for entry into glycolysis, thereby
the photosensitizer elsewhere in the cell, causes the crippling the cell’s metabolic pathway [203]. Interestingly,
lysosomes to rupture and subsequently release the photo- hexokinase was damaged by hypericin-PDT only after
sensitizer. hypericin was bound to hexokinase while it was bound to
the PT pore complex [203]. Dubbelman et al. considered the
1.6.2.3 Mitochondrial Damage. Since ATP, required to ANT to be a susceptible-PDT target since its active site
power cellular functions, is produced by mitochondria Hilf contains several reactive thiol groups [28], and since Atlante
et al. considered this organelle to be susceptible to et al. showed that the functions of other anion translocators
photodynamic damage. In support of this, Hilf et al. decreased before ATP production declined [204]. The PBR
demonstrated that HpD-PDT inhibits several mitochondrial is also thought to be a susceptible target because it
proteins in vivo. Malate dehydrogenase, succinate dehy- transports cholesterol into the mitochondrial matrix for
drogenase, and cytochrome c oxidase were inhibited by steroidogenesis [205] and because hematoporphyrin and
15%, 20–25%, 50–60% respectively by HpD-PDT [82]. protoporphyrin readily bind to it [72]. Verma et al. reported
These three proteins are required to maintain the electro- a positive correlation between the doses needed to kill 63%
chemical gradient across the inner-mitochondrial mem- of the cells in culture and the PBR inhibition constants of a
brane. Murant et al. also examined the functioning of the series of porphyrin photosensitizers. Verma also demon-
mitochondrial proteins adenylate kinase (AK) and mono- strated that cells expressing higher levels of PBR were more
amine oxidase (MAO) after in vivo HpD-PDT [199]. susceptible to Photofrin-PDT than cells expressing normal
Adenylate kinase and MAO were inhibited by 30% and levels of PBR [73].
80% respectively, but regained 100% activity within two Perlin et al. demonstrated that in vitro HpD-PDT reduced
hours of treatment possibly because of the presence of the activity of the adenosine triphosphatase (ATPase),
antioxidants in the mitochondria. Hilf et al. also examined which extracts energy from the proton gradient across the
the functioning of glucose phosphate isomerase, pyruvate mitochondrial membrane to form adenosine triphosphate
kinase, and lactate dehydrogenase, all cytosolic proteins. (ATP) [206]. Perlin demonstrated that PDT-induced cell
None of the these proteins were inhibited by HpD-mediated death occurs after the mitochondrial membrane potential
PDT [82]. They suggested that cytosolic proteins may not be decreases. Leakage of markers from the mitochondria was
as susceptible to photochemical oxidation as membrane- observed to occur only after oxidative phosphorylation
bound proteins because the singlet-oxygen lifetime is stopped due to photo-oxidation of mitochondrial proteins
drastically reduced in aqueous environment (see Section such as cytochrome c oxidase [206]. Atlante et al. came to a
1.3.1). This was supported by the fact that pryruvate kinase similar conclusion by showing that mitochondrial swelling
is anchored to the inside of the outer membrane yet was still occurs only after disruption of the mitochondrial membrane

Copyright # 2001 John Wiley & Sons, Ltd. J. Porphyrins Phthalocyanines 2001; 5: 105–129
BASIC PRINCIPLES OF PHOTODYNAMIC THERAPY 119

potential [204]. Atlante et al. treated mitochondria with Apoptosis is characterized by a series of cellular changes
ammonium acetate to induce colloidosmotic swelling, and that can be seen with a microscope; they include chromatin
compared the rates of swelling with Photofrin1-PDT- condensation and the formation of apoptotic bodies.
treated mitochondria. They reported that Photofrin1-treated Apoptosis can be also detected by examining the electro-
mitochondria swelled faster than ammonium acetate-treated phoretic pattern of DNA extracted from cells. The DNA
mitochondria indicating that PDT was damaging the from cells that die by apoptosis forms a ‘ladder’ pattern
mitochondria’s ability to regulate the membrane potential because the DNA is cleaved into multiples of 180 kDa
rather than permeablizing the membrane, as does ammo- fragments. This is the classical sign of apoptotic cell death,
nium acetate. Atlante et al. also spectrophotometrically although many new techniques have been devised, such as
monitored the transport of substrates by mitochondria the TUNNEL assay and flow cytometric techniques, to
during Photofrin-PDT. Before colloidosmotic swelling detect apoptosis more easily and quntitatively.
occurred tricarboxylate (i.e. citrate, oxalosuccinate) and
oxodicarboxylate (i.e. malate and oxaolacetate) carriers 1.7.2 Apoptosis in PDT. Argarwal et al. showed that
were observed to be inhibited by 50% and 65% respectively murine L5178Y lymphoma cells progressed to apoptosis in
[204]. Thus it appears photodynamic damage to the as little as 30 min after PDT treatment in vitro [208].
mitochondrial proteins constitutes the primary event in a Generally apoptotic responses require several hours to days
chain leading to breakdown of the electron transport chain, to complete. Argarwal suggested that PDT-induced cellular
disruption of the mitochondrial-membrane potential and damage bypasses the normal apoptotic mechanisms and
mitochondrial swelling [204]. initiates late-stage processes directly [208]. Zaidi et al.
observed apoptosis in vivo (apoptotic bodies and DNA
1.6.2.4 Membrane Damage. Photo-oxidizing cellular laddering) in RIF-1 tumors treated with Photofrin-PDT
membranes was initially thought to be an efficient way to [209]. However Zaidi reported that the apoptotic bodies and
kill tumor cells. Furthermore Photofrin1 was found to digested DNA may have come from host cells infiltrating
localize to the cellular membranes at early incubation times the tumor rather than the tumor cells [209]. He et al.
[193]. Dubbelman et al. reported that PDT can cause demonstrated that PDT does not induce apoptosis in all cell
lysosomal rupture, potassium ions and small molecules leak lines in vitro [210]. Despite having similar sensitivities to
from cells, and increase the flip-flop rate of membranes lipid Photofrin1-PDT, an LD50 dose of Photofrin1-PDT induced
but that these events usually occur after protein damage, apoptosis in rat MTF7 cells while an LD85 dose was
especially mitochondrial [28]. In general, cultured cells required to induce apoptosis in PC3 human prostatic
seem to be unaffected by direct membrane oxidation, except adenocarcinoma cells, and apoptosis could not be induced
at very high levels. The increased flip-flop rates of inserted in H322a non-small cell carcinoma cells. He et al. suggested
lipid probes and the leakage of small molecules described that there are many different pathways through which PDT
above appears to be the result of damage to the membrane can induce apoptosis [210].
proteins band 3 and the 2-aminoisobutyric acid transport Kessel et al. [190] and Luo et al. [211] showed that
protein (2-AATP) [28]. different photosensitizers induce apoptosis to varying
Although direct photo-oxidation of membrane lipids may degrees. They treated cultured cells with SnET2 or tin
not be especially lethal to a cell, membranes may still have octaethylpurpurin amidine (SnOPA, a cationic variant of
an important role in PDT. Oxidation of lipids within the SnEt2) -PDT. Tin ethyl etiopurpurin induced apoptosis
mitochondrial membranes may not be directly cytotoxic but within 60 min of treatment while SnOPA required 24 h or
these oxidized lipids may serve as potent signals that signal longer [190, 211]. Tin octaethylpurpurin amidine localized
a cell to die. Phospholipids are natural substrates for uniformly throughout the cell and its membranes while
phospholipases which, as reported in Section 1.6.1, can be SnET2 localized to the lysosomes and mitochondria, thus
converted into arachidonic acid by phospholipases. Luo suggested that SnOPA damaged cellular membranes,
Membranes, especially mitochondrial membranes, may retarding the cell’s ability to regulate apoptosis [190].
also act as localization sites for photosensitizers. Photo- Argarwal et al. demonstrated that phospholipases C and
dynamically active photosensitizers may have no inherent A2, which are involved in apoptosis, are activated after
affinity for specific protein targets but may partition into the Photofrin1-PDT [166]. When the actions of phospholipases
surrounding lipid layers. Therefore the photosensitizer’s C and A2 were blocked with inhibitors, the transient
affinity for particular membranes determines the protein increases in intracellular calcium levels and DNA fragmen-
target, i.e. those that reside in that membrane. Wilson et al. tation commonly observed during apoptosis were elimi-
have demonstrated that Photofrin1 competes with, 10- nated [166].
nonyl acridine orange (which binds selectively to cardio- Kick et al. reported that the proto-oncogenes c-jun and c-
lipin), for the cardiolipin-rich membranes of the inner fos, which are commonly associated with apoptosis, were
mitochondrial membrane [172]. activated after treating cultured cells with Photofrin1-PDT
[212]. C-jun and c-fos may be turned on to transcribe
1.7 Apoptosis apoptosis mediating factors. Fisher et al. showed that
cultured HL-60 cells expressing higher levels of p53 were
1.7.1 Apoptosis and Necrosis. Two ways in which less sensitive to Photofrin1-PDT than cells expressing
eukaryotic cells can die have been identified. Necrosis is normal levels or no p53 [213]. The tumor suppressor gene
characterized by non-ATP-driven processes such as colloid- p53 is known to arrest damaged cells in the S-phase of the
osmotic swelling while apoptosis is mediated by ATP- cell cycle and force them into apoptosis if they cannot repair
driven processes such as caspase activation [207]. Sub- the damage [214].
lethal genetic damage signals the cell to die apoptotically In the last ten years there has been a shift away from
while increasingly severe damage will destroy the cell’s thinking that the nucleus is the only organelle to control
ability to produce ATP and it will die by necrosis. apoptosis. Guido Kroemer and other (next sections) have

Copyright # 2001 John Wiley & Sons, Ltd. J. Porphyrins Phthalocyanines 2001; 5: 105–129
120 I. J. MACDONALD AND T. J. DOUGHERTY

established that the mitochondria can induce and control would undergo apoptosis. If the mitochondria were treated
nuclear apoptosis. In light of this discovery and the with ATT, then with BA, and then mixed with isolated
information presented in Section 1.6.2.3, Kessel et al. nuclei the nuclei did not undergo apoptosis. Thus it was
hypothesized that PDT may kill cells by mitochondrially clearly demonstrated that PT can be a early apoptotic event
controlled apoptosis [215, 216]. that precedes the loss of D m and ROS generation. It was
thought that many apoptosis inducers may function by
1.7.3 Mitochondrially Controlled Apoptosis triggering PT [219].
1.7.3.1 Reactive Oxygen Species (ROS) and the Mito-
chondrial Membrane Potential (D m) in Apoptosis. 1.7.3.3 Apoptosis Inducing Factor and Cytochrome c in
Zamzami et al. showed that non-nuclear elements could Apoptosis. Zamzami et al. and Petit et al. reported that
direct apoptosis by inducing apoptosis in enucleated cells cytochrome c and a 50 kDa protein, named apoptosis
[217]. Zamzami also showed that splenocytes decreased inducing factor (AIF), were released after mitochondrial
their uptake of dihexyloxacarbocyanine (DiOC6), a mito- swelling [219, 220]. Cytochrome c and AIF both induce
chondrially localizing dye, after forcing them to become apoptosis but by different mechanisms. Apoptosis inducing
apoptotic with dexamethasone. The amount of DiOC6 factor directly initiates apoptosis in the nucleus while
accumulated by the mitochondria can be correlated with cytochrome c activates caspases. In the presence of cyto-
the mitochondrial membrane potential (D m) which con- chrome c and dATP the Apaf-1/procaspase-9 complex
trols its accumulation. In apoptotic splenocytes the DiOC6 (Apaf-1 is the human homologue of the Caenorhabditis
signal decreased long before classical apoptotic signs could elegans cell-death gene that forms a complex with
be observed. While D m decreased, mitochondrial gluta- procaspase-9) can activate caspase-9 which in turn activates
thione levels decreased and reactive oxygen species (ROS) other death signals such as caspase-3, 6, and 7 [221, 222].
production increased. All three signals (decreased D m, The Apaf-1/procaspase-9 complex has also been shown to
decreased glutathione levels, and increased production of be an essential component of p53 mediated apoptosis [222].
ROS) were proposed to be very early, mitochondrially
controlled events that eventually lead to nuclear apoptosis 1.7.4 Summary, PDT and Apoptosis. There are many
[217]. similarities between the mechanisms by which mitochon-
Zamzami et al. also reported a D m threshold value dria regulate apoptosis and mechanisms by which PDT kills
below which the cell was irreversibly committed to die by cells. Kessel et al. have proposed that mitochondrially
apoptosis [218]. This was demonstrated by treating controlled apoptosis, similar to the mechanisms described
lymphocytes with various apoptosis-inducing agents (cyclo- above, is occurring after PDT treatment [215, 216]. Active
sporin A, rotenone, m-ClCCP, A23187, N-t-butyl-a-phenyl- photosensitizers may localize to mitochondrial targets such
nitrone, trolox, and L-ascorbate), each which induces as the membranes surrounding the PT pore or specific
apoptosis by different mechanisms. The lymphocytes were proteins in the PT pore. Hilf et al. and others have shown
then stained with DiOC6 as described above and separated other important mitochondrial proteins may be involved as
into two populations, those with high D m values and those well. During PDT singlet oxygen attacks mitochondrial
with low D m values. Neither population initially exhibited proteins and possibly alters their amino acid structures,
classical apoptotic signs (DNA laddering, apoptotic bodies) thereby disrupting the protein’s folding conformation. This
but within 18 h the low-D m lymphocytes progressed to may force them into other conformations that disrupt their
apoptosis while the high-D m lymphocytes did not, clearly normal functioning. Ruck et al. recently demonstrated that
indicating that the D m collapse is an early apoptotic signal ANT, reconstituted into liposomes, could act normally as an
[218]. ADP–ATP exchange carrier or it could form nonspecific
channels with conductance properties similar to the PT-pore
1.7.3.2 Permeability Transition (PT). Zamzami reported [223]. The PT pores allow water molecules to flow into ions
that mega-channels in the inner mitochondrial membrane and antioxidants flow out as the system settles towards
can open in response to certain compounds. When these equilibrium. The membrane potential decreases and the
channels open (termed permeability transition [PT]) small electron transport chain uncouples spewing out detrimental
molecules (1.5 kDa) can flow in and out of the ROS. The remaining antioxidants may scavenge some ROS
mitochondrial matrix. When PT occurs the mitochondria but if PT is not reversed damage will accumulate and the
cannot maintain the proton gradient, D m collapses then the electron transport chain will collapse. Petit et al. reported
electron transport chain uncouples, ROS production in- that bcl-2 decreases the likelihood of apoptosis by inhibiting
creases (produced by an inefficiently functioning electron PT-pore formation [224] and He et al. have shown that cells
transport chain) and consumes antioxidants. Eventually overexpressing bcl-2 were more resistant to in vitro
antioxidants are depleted and ROS damage accumulates and Photofrin1-PDT than parental cells [225]. Kim et al. also
more PT-pores open, the mitochondria swell and burst, demonstrated that bcl-2 mediates PDT-induced cell death,
spewing out detectable molecules such as glutathione and that aluminum phthalocyanine-PDT selectively de-
(GSH) and calcium ions, and the cell proceeds into stroys bcl-2, enhancing the apoptotic response after PDT
apoptosis [219]. [226]. Both results imply that PT-pore formation is a
The PT-pore complex consists of a conglomerate component of PDT-mediated cell death [225, 226]. Below a
hexokinase, VDAC, PBR, ANT, creatine kinase, and certain D m threshold level the mitochondria will cease to
cyclophyllin D and probably others that have not been function and swell, bursting off the outer membrane because
identified [202]. Zamzami et al. used attractyloside (ATT) it has less surface area than the inner membrane. This
to force open PT-pores and bongkrekic acid (BA) to force releases cytochrome c and AIF, which reside between the
them closed on isolated mitochondria. If the isolated inner and outer membranes, and at this point apoptosis is
mitochondria were exposed to ATT to induce PT and then irreversible. Cytochrome c activates caspases and AIF
the mitochondria were mixed with isolated nuclei, the nuclei initiates nuclear apoptosis. Varnes et al., Kessel et al. and

Copyright # 2001 John Wiley & Sons, Ltd. J. Porphyrins Phthalocyanines 2001; 5: 105–129
BASIC PRINCIPLES OF PHOTODYNAMIC THERAPY 121

Kim et al. have demonstrated that cytochrome c is released the immune system immediately after treatment while it
during PDT and that it enhances the apoptotic response to required 72 h for immune suppression to develop in animals
PDT [215, 216, 226, 227]. A diagram showing the pathways treated with Photofrin and m-THPC-PDT. Musser et al.
described and the putative PDT targets is shown in Fig. 11. suggested that different photosensitizers behave differently
The star bursts represent the putative PDT targets which because they have different hydrophobicities and localize to
when damaged initiate this apoptotic pathway. different subcellular targets which in turn affects the
immune response [228]. Alternatively, Lynch et al.
suggested macrophages modulate immune suppression
1.8 Immune Effects in PDT
because they accumulate large amounts of highly aggre-
1.8.1 Immune System Activation and Suppression. gated compounds (i.e. Photofrin1) and then secrete
Although it is not completely understood, researchers have suppressive agents during PDT [229]. Lynch et al.
reported that PDT activates and suppresses the immune supported this theory by adoptively transferring macro-
system, depending upon certain variables. Using PDT to phages from PDT-treated animals to di-nitrofluorobenezene
generate tumor antigens to induce a specific immune (DNFB) contact-sensitized animals. They reported a 60%
response against tumor tissue could be useful to treat a reduction in the immune response after a second DNFB
primary tumor and its metastases. However, it has been challenge [229]. Furthermore Lynch et al. showed that
reported that PDT also suppresses immune system function, PDT-treated macrophages produce and secrete cytokines,
limiting the effect of specific immune responses. Photo- such as interleukin-1 (IL-1). This seemed logical since IL-1
dynamic treatment seems to alter the balance of agents that can facilitate production of acute phase proteins (i.e. serum
regulate the immune system and appears to be able to tip it amyloid A, serum amyloid P, and C-reactive protein) which
towards activation or suppression by the release of specific modulate immune suppression. They reported that serum
cytokines. This is probably controlled by many complicated amyloid A, serum amyloid P, and C-reactive protein were
factors including, but not limited to, the particular detectable after PDT [229].
photosensitizer, light dose, and dose rate. Korbelik et al. have demonstrated that PDT can induce
The mechanism by which PDT induces specific immune long-term specific immunity [230]. Korbelik used vitamin
responses has been suggested by Korbelik [157]. Summariz- D3 binding protein-derived macrophage activating factor
ing, PDT-treated tissue releases large quantities of cell (DBPMAF), which hyper-stimulates macrophage prolifera-
debris, and inflammatory signals, cytokines and chemotactic tion, and may be important for retrieving antigens, to
agents. Within minutes of light treatment large numbers of overcome immune suppression. They reported increased
neutrophils can be observed invading the treated tissue. selective tumor destruction after DBPMAF and PDT [231].
They secrete lysosomal enzymes and ROS that destroy
endothelial and tumor cells and further amplify the 1.8.2 Immune System Modulators. The immune system
inflammatory response. When the neutrophils die they produces cytokines, many of which appear to have
release their cellular contents which then act as chemotactic numerous and subtle roles depending on the situation, to
signals for future waves of inflammatory cells. As the regulate itself. Cytokine signals have been studied during
neutrophil invasion wanes mast cells flood into the damaged and following PDT to determine how the immune system
tissue and release granules containing vasoactive agents and functions at a particular time.
cytokines. Monoctyes and macrophages invade, proliferate, Nseyo et al. detected interleukin-1b (IL-1b), IL-2, and
collect cell debris, and destroy pockets of surviving tumor tumor necrosis factor-a (TNF-a) in the urine of patients who
cells. At the end of the inflammatory response the macro- received whole bladder PDT, demonstrating that the
phages and monocytes secrete immunosuppressive factors immune system was active in the tissues of treated patients
which down-regulate the response and may hinder any [232]. It was not reported whether these cytokines were
future, specific immune response [157]. produced during an inflammatory or a specific immune
Musser et al. showed that PDT-induced immune sup- response.
pression was a function of the photosensitizer used [228]. Herman et al. showed that protoporphyrin-IX (PP-IX)
Musser demonstrated that TPPS4 and HpD-PDT suppress and Photofrin-PDT induced secretion of IL-2, IL-3, TNF-a,

Fig. 11. A summary of the pathways and feedback loops involved in mitochondrially controlled
apoptosis. Star bursts represent possible points at which PDT initiates mitochondrially controlled
apoptosis.

Copyright # 2001 John Wiley & Sons, Ltd. J. Porphyrins Phthalocyanines 2001; 5: 105–129
122 I. J. MACDONALD AND T. J. DOUGHERTY

and interferon-g (IFN-g) from human and mouse monocytes uses are in various stages of clinical trials demonstrating the
[233]. They also reported that PP-IX sensitized the immune broad applicability of PDT to human disease.
system without light and suggested that the PDT-induced
immune response was enhanced because of pre-sensitiza-
tion before light treatment [233]. Caution must be used Acknowledgements
while interpreting these results since the doses of PP-IX in The authors thank the National Institutes of Health (CA
the experiments used were excessively high so that even 55791) and the Oncologic Foundation of Buffalo for
small amounts of light may have induced PDT. financial support.
Bellnier demonstrated that a combination of low-dose
recombinant human TNF-a, administered adjuvant to
Photofrin-PDT doubled the phototoxicity of Photofrin and
improved the selectivity of the treatment. Since TNF
stimulates macrophage growth and differentiation, Bellnier
REFERENCES
suggested that exogenous TNF may enhance the inflamma-
tory response at the treatment site [234]. Evans et al. and 1. Raab O. Ueber die wirkung fluoreszierender stoffe auf
Schumaker et al. also detected TNF-a in the supernatants infusorien. Z. Biol. 1900; 39: 524–546.
from mouse macrophages, and human cells treated in vitro 2. Policard A. Etudes sur les aspects offerts par des tumeurs
with PDT [235, 236]. experimentales examinee a la lumiere de Woods. Compt.
Kick et al. detected IL-6 mRNA and IL-6 protein in Rend. Soc. Biol. 1924; 91: 1423–1424.
epithelial-HeLa cells treated with Photofrin-PDT [237]. 3. Ronchese F. The fluoreseence of cancer under the Wood
Since IL-6 production from PDT peaked four hours earlier, light. Oral. Surg. Oral. Med. Oral. Path. 1954; 7: 967–
and in higher quantities than from cells treated with UV 971.
irradiation Kick et al. suggested that that PDT killed cells by 4. Figge FHJ. The relationship of pyrrol compounds to
a different pathway than radiation. They reported that PDT- carcinogenesis. In AAAS Research Conf. on Cancer.
stimulated IL-6 production was more similar to that from Moulton FR. (ed). 1945; 117–128.
cells treated with 12-O-tetradecanoylphorbol-13-acetate 5. Figge FHJ. Editorial, on the etiology of cancer carcino-
(TPA), an apoptosis inducer. However, IL-6 production genic agents. Ann. Int. Med. 1947; 24: 143–146.
peaked nearly two hours earlier in the PDT-treated cells 6. Figge FHJ, Weiland GS, Manganiello LOJ. Cancer
than in the TPA-treated cells, suggesting that PDT induces detection and therapy. Affinity of neoplastic, embryonic
immune responses by unique pathways [237]. and traumatized regenerating tissues for porphyrins and
Henderson et al. examined the changes in levels of IL-1, metalloporphyrins. Proc. Soc. Exp. Biol. Med. 1948; 68:
IL-3, IL-6, IL-10, IFN-g, granulocyte/macrophage-colony- 143–146.
stimulating factor (GMCSF), TNF-a, and transforming 7. Rasmussen-Taxdal DS, Ward DE, Figge FHJ. Fluores-
growth factor b (TGFb) after in vivo and in vitro PDT cence of human lymphatie and cancer tissue following
[238]. After in vivo Photofrin1-PDT Henderson reported high doses of intravenous hematoporphyrin. Cancer 1955;
that only IL-6 and IL-10 levels changed significantly. 8: 78–81.
Interleukin-6 mRNA and protein levels in the skin, spleen, 8. Winkelman J. Intracellular localization of ‘hematopor-
and tumor cells peaked three hours post-treatment and phyrin’ in transplanted tumor. J. Natl. Cancer Inst. 1961;
decreased thereafter. At high light doses (>100 J cmÿ2), 27: 1369–1377.
IL-10 mRNA levels rapidly increased then decreased to 9. Winkelman J. The distribution of tetraphenylporphine-
undetectable levels within six hours. However, in the skin sulfonate in the tumor-bearing rat. Cancer Res. 1962; 22:
IL-10 levels increased up to 120 h after treatment (at that 589–596.
point observation was stopped). Flow cytometry experi- 10. Winkelman J. Metabolic studies on the accumulation of
ments revealed that granulocyte and macrophage popula- tetraphenylporphine sulfonate in tumors. Experimentia
tions increased 89% and 40% respectively and tumor cells 1967; 23: 949–950.
decreased by 60% after treatment. These data demonstrate 11. Tappeiner H, Jesionek A. Therapeutische versuche mit
that macrophage proliferation and activity was destroying fluoreszier enden stoffe. Muench. Med. Wochschr. 1904;
tumor cells long after irradiation had stopped [238]. 1: 2042–2044.
12. Jesionek A, Tappeiner H. Zur behandlung der haufcarci-
nome mit fluorescierenden stoffen. Deut. Arch. Klin. Med.
1.9 Clinical Photodynamic Therapy 1904; 82: 223–227.
While the clinical aspects of PDT are not the main focus of 13. Henderson BW, Dougherty TJ, Schwartz S et al.
this review it is noteworthy to point out that Photofrin-PDT Historical perspective. In Photodynamic Therapy: Basic
has been approved for use in the United States against Principles and Clinical Applications. Henderson BW,
advanced stage esophageal, advanced non-small-cell lung Dougherty TJ. (eds). Marcel Decker: New York, 1992; 1–
cancer and early-stage lung cancer. Photofrin-PDT has also 15.
been approved for use in five European countries, Canada 14. Lipson RL, Baldes EJ, Olsen AM. The use of a derivative
(including bladder cancer), and Japan. Other trials that are of hematoporphyrin in tumor detection. J. Natl. Cancer
pending approval use Photofrin-PDT to treat early-stage Inst. 1961; 26: 1–11.
esophageal cancer. Photofrin1-PDT is also being investi- 15. Dougherty TJ. A brief history of clinical photodynamic
gated as an adjuvant therapy for treatment of the surgical therapy at Roswell Park Cancer Institute. J. Clin. Laser
bed after resection of malignant glioma, malignant astro- Med. Surg. 1996; 14: 219–221.
cytoma, malignant mesothelioma, head and neck cancers, 16. Dougherty TJ. Activated dyes as antitumor agents. J. Natl.
and some intraperitoneal cancers [18]. In addition numerous Cancer Inst. 1974; 52: 1333–1336.
new photosensitizers for both oncologic and non-oncologic 17. Weishaupt KR, Gomer CJ, Dougherty TJ. Identification of

Copyright # 2001 John Wiley & Sons, Ltd. J. Porphyrins Phthalocyanines 2001; 5: 105–129
BASIC PRINCIPLES OF PHOTODYNAMIC THERAPY 123

singlet oxygen as the cytotoxic agen in photo-inactivation 37. Siegel MM, Tabei K, Tsao R et al. Comparative mass
of a murine tumor. Cancer Res. 1976; 36: 2326–2329. spectrometric analyses of Photofrin oligomers by fast
18. (a) Dougherty TJ, Gomer CJ, Henderson BW et al. atom bombardment mass spectrometry, UV or IR matrix-
Photodynamic Therapy. J. Natl. Cancer Inst. 1998; 90: assisted laser desorption/ionization mass spectrometry,
889–905. (b) Pandey RK, Zheng G. Porphyrins as electrospray ionization mass spectrometry and laser
Photosensitizers in Photodynamic Therapy, vol. 6. desorption/jet-cooling photoionization mass spectrome-
Kadish, Smith, Guilard. (eds). Academic Press: San try. J. Mass. Spectrom. 1999; 34: 661–669.
Diego, 2000. 38. Pandey RK, Shiau F, Dougherty TJ, Smith KM. Regio-
19. Gilbert A, Baggott J. Molecular photochemistry. Essen- selective synthesis of ether-linked porphyrin dimers and
tials of Molecular Photochemistry: CRC Press: Boca trimers related to Photofrin II. Tetrahedron 1991; 47:
Raton, 1991; 1–10. 9571–9584.
20. Gilbert A, Baggott J. Molecular Photophysics. Essentials 39. Pandey RK, Smith KM, Dougherty TJ. Porphyrin dimers
of Molecular Photochemistry. CRC press: Boca Raton, as photosensitizers in photodynamic therapy. J. Med.
1991; 91–144. Chem. 1990; 33: 2032–2038.
21. Turro NJ. Singlet oxygen and chemilluminescent organic 40. Bonnet R, Berenbaum MC. HPD-A study of its compo-
reactions. Modern Molecular Photochemistry. University nents and their properties. Adv. Exp. Biol. Med. 1983; 160:
Science Books: California, 1991; 583–593. 241–250.
22. Foote CS. Mechanisms of photo-oxygenation. In Porphy- 41. Kessel D. Proposed structure of the tumor localizing
rin Localization and Treatment of Tumors. Doiron DR, fraction of HpD (hematoporphyrin derivative). Photo-
Gomer CJ. (eds). Alan R. Liss: New York, 1984; 3–18. chem. Photobiol. 1986; 44: 193–196.
23. Henderson BW, Dougherty TJ. How does photodynamic 42. Evans HH, Horng M, Ricanati M et al. Mutagenicity of
therapy work? Photochem. Photobiol. 1992; 55: 147–157. photodynamic therapy as compared to UVC and ionizing
24. Gilbert A, Baggott J. Photo-oxygenation reactions. radiation in human and murine lymphoblast cell lines.
Essentials of Molecular Photochemistry. CRC Press: Photochem. Photobiol. 1997; 66: 690–696.
Boca Raton, 1991; 501–525. 43. Henderson BW, Bellnier DA, Greco WR et al. An in vivo
25. Pass HI. Photodynamic therapy in oncology. J. Natl. quantitative structure–activity relationship for a conge-
Cancer Inst. 1993; 86: 443–456. neric series of pyropheophorbide-a derivatives as photo-
26. See KL, Forbes IJ, Betts WH. Oxygen dependence of sensitizers for photodynamic therapy. Cancer Res. 1998;
phototoxicity with haematoporphyrin derivative. Photo- 57: 4000–4007.
chem. Photobiol. 1984; 39: 631–634. 44. Morgan AR. 1992. Reduced porphyrins as photosensiti-
27. Henderson BW, Fingar VH. Relationship of tumor zers: Synthesis and biological effects. In Photodynamic
hypoxia and response to photodynamic therapy in an Therapy, Basic Principles and Clinical Applications.
experimental mouse tumor. Cancer Res. 1987; 47: 3110– Henderson BW, Dougherty TJ, (eds). Marcell Decker:
3114. New York, 1992; 157–172.
28. Dubbelman TMAR, Prinsze C, Penning LC, Van 45. Richter AM, Waterfield E, Jain AK et al. In vitro
Steveninck J. In Photodynamic Therapy: Basic Principles evaluation of phototoxic properties of four structurally
and Clinical Applications. Henderson BW, Dougherty TJ. related benzoporphyrin derivatives. Photochem. Photo-
(eds). Marcel Decker: New York, 1992; 37–46. biol. 1990; 52: 495–500.
29. Foster TH, Murant RS, Bryant RG et al. Oxygen 46. Richter AM, Cerruti-Sola S, Sternberg ED et al.
consumption and diffusion effects in photodynamic Biodistribution of tritiated benzoporphyrin derivative
therapy. Radiat Res. 1991; 126: 296–303. (3H-BPD-MA), a potent photosensitizer, in normal and
30. Reed MWR, Mullins AP, Anderson GL et al. The effects tumor bearing mice. J. Photochem. Photobiol. B: Biol.
of photodynamic therapy on tumor oxygenation. Surgery 1990; 5: 231–244.
1989; 106: 94–99. 47. Lin SC, Lin CP, Feld JR et al. The photodynamic
31. Tromberg BJ, Orenstein A, Kimel SJ et al. In vivo tumor occlusion of choroidal vessels using benzoporphyrin
oxygen tension measurements for the evaluation of the derivative. Curr. Eye Res. 1994; 13: 513–522.
efficiency of photodynamic therapy. Photochem. Photo- 48. Schmidt-Erfurth U, Miller J, Sickenberg M et al.
biol. 1990; 52: 375–385. Photodynamic therapy of subfoveal choroidal neovascu-
32. Sitnik TM, Henderson BW. The effect of fluence rate on larization:clinical and angiographic examples. Graefes
tumor and normal tissue responses to photodynamic Arch. Clin. Exp. Opthalmol. 1998; 230: 365–374.
therapy. Photochem. Photobiol. 1998; 67: 462–466. 49. Visudyne2, FDA approval. CIBA Vision. 11-17-1999.
33. Sitnik TM, Hampton JA, Henderson BW. Reduction of Ref Type: Internet Communication.
tumour oxygenation during and after photodynamic 50. Mang TS, Allison R, Hewston G et al. A phase II/III
therapy in vivo: effects of fluence rate. Br. J. Cancer clinical study of tin ethyl etiopurpurin (Purlytin)-induced
1998; 77: 1386–1394. photodynamic therapy for the treatment of recurrent
34. Spikes JD. Photobiology of porphyrins. In Porphyrin cutaneous metastatic breast cancer. Cancer J. Sci. Am.
Localization and Treatment of Tumors. Doiron DR, 1998; 4: 378–384.
Gomer CJ, (eds). Alan R. Liss: New York, 1984; 19–39. 51. Primbs GB, Casey R, Wamser K et al. Photodynamic
35. Pandey RK, Majchrzycki JF, Smith KM, Dougherty TJ. therapy for corneal neovascularization. Ophthalmic Surg.
Chemistry of Photofrin II and some new photosensitizers. Lasers 1998; 29: 832–832.
Proc SPIE 1989; 1065: 164–174. 52. Ma L, Moan J, Berg K. Evaluation of a new photo-
36. Pandey RK, Marshall SM, Tsao M et al. Fast atom sensitizer, meso-tetra-hydroxyphenyl-chlorin, for use in
bombardment mass spectral analysis of Photofrin II and photodynamic therapy: A comparison of its photobiolo-
its synthetic components. Biomed. Environ. Mass Spec. gical properties with those of two other photosensitizers.
1990; 19: 404–415. Int. J. Cancer 1994; 87: 883–888.

Copyright # 2001 John Wiley & Sons, Ltd. J. Porphyrins Phthalocyanines 2001; 5: 105–129
124 I. J. MACDONALD AND T. J. DOUGHERTY

53. Boyle RW, Dolphin D. Structure and biodistribution meso-tetra(4-N-methylpyridyl)porphine. Cancer Lett.
relationships of photodynamic sensitizers. Photochem. 1993; 73: 59–64.
Photobiol. 1996; 63: 469–485. 71. Woodburn K, Stylli S, Hill JS et al. evaluation of tumour
54. Murrer LHP, Marijnissen JPA, Star WM. Short- and long- and tissue distribution of porphyrins for use in photo-
term normal tissue damage with photodynamic therapy in dynamic therapy. Br. J. Cancer 1999; 65: 321–328.
pig trachea: a fluence-response pilot study comparing 72. Verma A, Snyder SH. Characterization of porphyrin
Photofrin and mTHPC. Br. J. Cancer 1999; 80: 744–755. interactions with peripheral type benzodiazepine recep-
55. Ris HB, Altermatt HJ, Inderbitzi R et al. Photodynamic tors. Mol. Pharm 1988; 34: 800–805.
therapy with chlorins for diffuse malignant mesothelioma: 73. Verma A, Facchina SL, Hirsch DJ et al. Photodynamic
initial clinical results. Br. J. Cancer 1991; 64: 1116–1120. tumor therapy: mitochondrial benzodiazepine receptors as
56. Kuebler A, Haase T, Staff C et al. Photodynamic therapy a therapeutic target. Mol. Med. 1998; 4: 40–45.
of primary nonmelanomatous skin tumours of the head 74. Tsuchida T, Zheng G, Pandey RK et al. Correlation
and neck. Lasers Surg. Med. 1999; 25: 60–68. between site-II specific human serum albumin (HSA)
57. Taber SW, Fingar VH, Coots CT, Wieman TJ. Photo- binding affinity and murine in vivo photosensitizing effi-
dynamic therapy using mono-L-aspartyl chlorin e6 cacy of some Photofrin components. Photochem. Photo-
(Npe6) for the treatment of cutaneous disease: a phase I biol. 1997; 66: 224–228.
clinical study. Clin. Cancer Res. 1998; 4: 2741–2746. 75. Hombrecher HK, Schell C, Thiem J. Synthesis and
58. Mori K, Yoneya S, Ohta M et al. Angiographic and investigation of a galactopyranosyl-cholesteryloxy sub-
histologic effects of fundus photodynamic therapy with a stituted porphyrin. Bioorg. Med. Chem. Lett. 1999; 6:
hydrophilic sensitizer (mono-L-aspartyl chlorin e6). 1199–1202.
Ophthalomology 1999; 106: 1384–1391. 76. Donald PJ, Cardiff RD, He D, Kendall K. Monoclonal
59. Young SW, Woodburn KW, Wright M et al. Lutetium antibody–porphyrin conjugate for head and neck cancer:
texaphyrin (PCI-0123): A near-infrared, water-soluble the possible magic bullet. Otolaryng Head Neck Surg.
photosensitizer. Photochem. Photobiol. 1996; 63: 892– 1991; 105: 781–787.
897. 77. Hemming AW, Davis NL, Dubois B et al. Photodynamic
60. Woodburn KW, Fan Q, Kessel D et al. Photodynamic therapy of squamous cell carcinoma. An evaluation of a
therapy of B16F10 murine melanoma with lutetium new photosensitizing agent, benzoporphyrin derivative
texaphyrin. J. Invest, Dermatol. 1998; 110: 746–751. and new photoimmunoconjugate. Surg. Oncol. 1993; 2:
61. Woodburn KW, Qing F, Kessel D, Young SW. Photo- 187–196.
irradiation and imaging of atheromatous plaque with 78. Vrouenraets MB, Visser GWM, Stewart FA et al.
texaphyrins. Proceedings of SPIE-the International So- Development of meta-tetrahydroxyphenylchlorin-mono-
ciety for Optical Engineering, 1997; 44: 50. clonal antibody conjugates for photoimmunotherapy.
62. Kennedy JC, Marcus SL, Pottier RH. Photodynamic Cancer Res. 1999; 59: 1505–1513.
therapy (PDT) and photodiagnosis (PD) using endogenous 79. Akhlynina TV, Tamara V, Rosenkranz AA et al. Insulin-
photosensitization induced by 5-aminoluvulinic acid mediated intracellular targeting enhances the photo-
(ALA): Mechanisms and clinical results. J. Clin. Laser dynamic activity of chlorin e6. Cancer Res. 1995; 55:
Med. Surg. 1996; 14: 289–304. 1014–1019.
63. Lange M, Jichlinski P, Zellweger M et al. Photodetection 80. Karagianis G, Reiss JA, Marchesini R et al. Biophysical
of early human bladder cancer based on the fluorescence and biological evaluation of porphyrin–bisacridine con-
of 5-aminolevulinic acid hexylester-induced protoporphy- jugates. Anti-Cancer Drug Design 1996; 11: 205–220.
rin IX: a pilot study. Br. J. Cancer 1999; 80: 185–193. 81. Bachor RS, Shea CR, Gillies R, Hasan T. Photosensitized
64. Dougherty TJ. Photochemistry in the treatment of cancer. destruction of human bladder carcinoma cells treated with
Adv. Photochem. 1992; 17: 275–311. chlorin e6-conjugated microspheres. Proc. Natl. Acad.
65. Margaron P, Gregoire M, Scasnar V et al. Structure- Sci. USA 1991; 88: 1580–1584.
photodynamic activity relationships for a series of 4- 82. Hilf R, Warne NW, Smail DB, Gibson SL. Photodynamic
substituted phthalocyanines. Photochem. Photobiol. 1996; inactivation of selected intracellular enzymes by hemato-
63: 217–223. porphyrin derivative and their relationship to tumor cell
66. Ali H, Langlois LJ, Wagner JR et al. Biological activities viability in vitro. Cancer Lett. 1984; 24: 165–172.
of phthalocyanines-X. Synthesis of sulfonated phthalo- 83. Singh G, Jeeves WP, Wilson BC, Jang D. Mitochondrial
cyanines. Photochem. Photobiol. 1988; 47: 713–717. photosensitization by Photofrin II. Photochem. Photobiol.
67. Hansch C, Leo A. Electronic effects on organic reactions. 1987; 46: 645–659.
Exploring QSAR, Fundamentals and Applications in 84. Chatterjee SR, Srivastava TS, Kamat JP, Devasagayam
Chemistry and Biology. American Chemical Society: TPA. Lipid peroxidation induced by novel porphyrin plus
Washington D.C., 1995; 1–22. light in isolated mitochondria: possible implications in
68. Cincotta L, Foley JW, Cincotta AH. Novel red absorbing photodynamic therapy. Mol. Cell. Biochem. 1997; 166:
benzo[a]phenoxazinium and benzo[a]phenothiazinium 25–33.
photosensitizers: in vitro evaluation. Photochem. Photo- 85. Margalit R, Rottenberg M. Thermodynamics of porphyrin
biol. 1987; 46: 751–758. dimerization in aqueous solutions. Biochem. J 1983; 219:
69. Bellnier DA, Young DN, Detty MR et al. pH-dependent 445–450.
chalcogenopyrylium dyes as potential sensitizers for 86. Margalit R, Shaklai N, Cohen S. Fluorometric studies on
photodynamic therapy: selective retention in tumors by the dimerization equilibrium of protoporphyrin IX and its
exploiting pH differences between tumor and normal haemato derivative. Biochem. J 1983; 209: 457–552.
tissue. Photochem. Photobiol. 1999; 70: 630–636. 87. Aveline BM, Hassan T, Redmond RW. The effects of
70. Villanueva A, Jori G. Pharmacokinetic and tumour- aggregation, protein binding and cellular incorporation on
photosensitizing properties of the cationic porphyrin the photophysical properties of benzoporphyrin derivative

Copyright # 2001 John Wiley & Sons, Ltd. J. Porphyrins Phthalocyanines 2001; 5: 105–129
BASIC PRINCIPLES OF PHOTODYNAMIC THERAPY 125

monoacid ring A (BPDMA). J. Photochem. Photobiol. B: 106. Hamblin MR, Newman EL. Photosensitizer targeting in
Biol. 1995; 30: 161–169. photodynamic therapy I. Conjugates of hematoporphyrin
88. Kessel D, Byrne CJ, Ward AD. Configuration of tripor- with albumin and transferrin. J. Photochem. Photobiol. B:
phyrin ethers probed by fluorescence measurements. J. Biol. 1994; 26: 45–56.
Photochem. Photobiol. B: Biol. 1992; 13: 153–160. 107. Peters Jr. T. Ligand binding by albumin. All About
89. Schell C, Hombrecher HK. Synthesis, self-assembling Albumin. Academic Press: San Diego, 1995; 76–132.
properties and incorporation of carbohydrate-substituted 108. Rottenberg M, Margalit R. Deuteroporphyrin-albumin
porphyrins into cell membrane models. Chem. Eur. J. binding equilibrium. Biochem. J. 1985; 229: 197–203.
1999; 5: 587–594. 109. Grossweiner LI, Goyal GC. Binding of hematoporphyrin
90. Csik G, Balog E, Voszka I et al. Glycosylated derivatives derivative to human serum albumin. Photochem. Photo-
of tetraphenyl porphyrin: photophysical characterization, biol. 1984; 40: 1–4.
self aggregation and membrane binding. J. Photochem. 110. Jori G, Cozzani I, Reddi E et al. in vitro and in vivo studies
Photobiol. B: Biol. 1998; 44: 216–224. on the interaction of hematoporphyrin and its dimethyl-
91. Akins DL, Ozcelik S, Zhu H, Guo C. Fluorescence decay ester with normal and malignant cells. In Porphyrin
kinetics and structure of aggregated tetrakis(p-sulfonato- Localization and Treatment of Tumors. Doiron DR,
phenyl)porphyrin. J. Phys. Chem. 1996; 100: 14390– Gomer CJ. (eds). Alan R. Liss: New York, 1984; 471–482.
14396. 111. Jori G, Beltramini M, Reddi E et al. Evidence for a major
92. Smith GJ. The effects of aggregation on the fluorescence role of plasma lipoproteins as hematoporphyrin carriers in
and triplet state yield of hematoporphyrin. Photochem. vivo. Cancer Lett. 1984 24: 291–297.
Photobiol. 1985; 41: 123–126. 112. Kessel D. Porphyrin-lipoprotein association as a factor in
93. Bennet LE, Chiggino KP, Henderson RW. Singlet oxygen porphyrin localization. Cancer Lett. 1986; 33: 183–188.
formation in monomeric and aggregated porphyrin c. J. 113. Korbelik M, Hung J. Cellular delivery and retention of
Photochem. Photobiol. B: Biol. 1988; 3: 81–89. Photofrin II: The effects of interaction with human plasma
94. Zhang X, Xu H. Influence of halogenation and aggrega- proteins. Photochem. Photobiol. 1991; 53: 501–510.
tion on photosensitizing properties of zinc phthalocyanine 114. Korbelik M. Cellular delivery and retention of Photofrin
(ZnPc). J. Chem. Soc., Faraday Trans. 1993; 89: 3347– II. The effects of human versus mouse and bovine serum.
3351. Photochem. Photobiol. 1992; 56: 391–397.
95. Fernandez DA, Awruch J, Dicello LE. Photophysical and 115. Korbelik M. Cellular delivery and retention of Photofrin
aggregation studies of t-butyl-substituted phthalo- III. Role of plasma proteins in photosensitizer clearance
cyanines. Photochem. Photobiol. 1996; 63: 784–792. from cells. Photochem. Photobiol. 1993; 57: 846–850.
96. MacDonald IJ, Morgan J, Bellnier DA et al. Subcellular 116. Beltramini M, Fiery PA, Fernanda R et al. Steady-state
localization pattens and their relationship to photody- and time-resolved spectroscopic studies on the hemato-
namic activity of pyropheophorbide-a derivatives. Photo- porphyrin-lipoprotein complex. Biochemistry 1987; 26:
chem. Photobiol. 1999; 70: 789–797. 6852–6858.
97. Howe L, Zhang JZ. The effects of biological substrates on 117. Gutierrez MM, Tsai S, Phillips ML et al. Studying low-
the ultrafast excited-state dynamics of zinc phthalocya- density lipoprotein–monoclonal antibody complexes
nine tetrasulfonate in solution. Photochem. Photobiol. using dynamic laser light scattering and analytical
1998; 67: 90–96. ultracentrifugation. Biochemistry 1999; 38: 1284–1292.
98. Moan J, Rimington C, Evensen JF, Western A. Binding of 118. Chatterton JE, Phillips ML, Curtiss LK et al. Immuno-
porphyrins to serum proteins. Adv. Exp. Biol. Med. 1985; electron microscopy of low density lipoproteins yields a
193: 193–220. ribbon and bow model for the conformation of apolipo-
99. Wan LSC, Lee PFS. CMC of polysorbates. J. Pharm. Sci. protein B on the lipoprotein surface. J. Lipid Res. 1995;
1974; 63: 136–137. 36: 2027–2037.
100. Farhadieh B. Determination of CMC and partial specific 119. Mosely ST, Goldstein JL, Brown MS et al. targeted killing
volume of polysorbates 20, 60, and 80 from densities of of cultured cells by receptor-dependent photosensitiza-
their aqueous solutions. J. Pharm. Sci. 1973; 62: 1685– tion. Proc. Natl. Acad. Sci. USA 1981; 79: 5717–5721.
1688. 120. Bellnier DA, Ho Y, Pandey RK et al. Distribution and
101. Kongshaung M, Cheng LS, Moan J, Rimington C. elimination of Photofrin II in mice. Photochem. Photo-
Interaction of Cremophor EL with human plasma. Int. J. biol. 1989; 50: 221–228.
Biochem. 1991; 4: 473–478. 121. Moan J, Sommer S. Uptake of the components of
102. McKean DL, Pesce AJ, Koo W. Analysis of polysorbate hematoporphyrin derivative by cells and tumors. Cancer
and its polyethoxylated metabolite. Anal. Biochem. 1987; Lett. 1983; 21: 167–174.
161: 348–351. 122. Kongshaug M. Distribution of tetrapyrrolic photosensiti-
103. Mayhew E, Vaughan LA, Panus A et al. Lipid-associated zers among human plasma proteins. Int. J. Biochem. 1992;
methylpheophorbide-a(hexyl-ether) as a photodynamic 24: 1239–1265.
agent in tumor-bearing mice. Photochem. Photobiol. 123. Maziere JC, Morliere P, Santus R. The role of the low
1993; 58: 845–851. density lipoprotein receptor pathway in the delivery of
104. Woodburn K, Chang CK, Sangan L et al. Biodistribution lipophilic photosensitizers in the photodynamic therapy of
and PDT efficacy of a ketochlorin photosensitizer as a tumors. J. Photochem. Photobiol. B: Biol. 1991; 8: 351–
function of the delivery vehicle. Photochem. Photobiol. 360.
1994; 60: 154–159. 124. Hamblin MR, Newman EL. on the mechanism of the
105. Hamblin MR, Newman EL. Photosensitizer targeting in tumor-localizing effect in photodynamic therapy. J.
photodynamic therapy II. Conjugates of hematoporphyrin Photochem. Photobiol. B: Biol. 1991; 23: 3–8.
with serum lipoproteins. J. Photochem. Photobiol. B: 125. Allison BA, Pritchard PH, Richter AM, Levy JG. The
Biol. 1994; 26: 147–157. plasma distribution of benzoporphyrin derivative and the

Copyright # 2001 John Wiley & Sons, Ltd. J. Porphyrins Phthalocyanines 2001; 5: 105–129
126 I. J. MACDONALD AND T. J. DOUGHERTY

effects of plasma lipoproteins on its biodistribution. of Tumors. Doiron DR, Gomer CJ. (eds.) Alan R, Liss:
Photochem. Photobiol. 1990; 52: 501–507. New York, 1984; 149–161.
126. Allison BA, Waterfield E, Richter AM, Levy JG. The 141. Dougherty TJ, Potter WR. Of what value is a highly
effects of plasma lipoproteins on in vitro tumor cell killing absorbing photosensitizer? J. Photochem. Photobiol. B:
and in vivo tumor photosensitization with benzoporphyrin Biol. 1991; 8: 233–234.
derivative. Photochem. Photobiol. 1991; 54: 709–715. 142. Potter WR, Mang TS. Photofrin II levels by in vivo
127. Allison BA, Pritchard PH, Levy JG. Evidence for low- fluorescence photometry. In Porphyrin Localization and
density lipoprotein receptor-mediated uptake of benzopor- Treatment of Tumors. Doiron DR, Gomer CJ. (eds.) Alan
phyrin derivative. Br. J. Cancer 1994; 69: 833–839. R. Liss: New York, 1984; 177–186.
128. Kessel D, Thompson P, Saatio K, Nantwi KD. Tumor 143. Bernstein EF, Thomas GF, Smith PD et al. Response of
localization and photosensitization by sulfonated deriva- black and white guinea pig skin to photodynamic
tives of tetraphenylporphine. Photochem. Photobiol. treatment using 514-nm light and dihematoporphyrin
1987; 45: 787–790. ether. Arch. Derm. 1990; 126: 1303–1307.
129. Kongshaug M, Moan J, Brown SB. The distribution of 144. Mang TS, Dougherty TJ, Potter WR et al. Photobleaching
porphyrins with different tumor localizing ability among of porphyrins used in photodynamic therapy and implica-
human plasma proteins. Br. J. Cancer 1989; 59: 184–188. tions for therapy. Photochem. Photobiol. 1987; 45: 501–
130. Berg K, Bommer JC, Moan J. Evaluation of sulfonated 506.
aluminum phthalocyanines for use in photochemotherapy. 145. Belitchenko I, Melnikova V, Bezdetnaya L et al.
A study on the relative efficiencies of photoinactivation. Characterization of photodegradation of meta-tetra (hy-
Photochem. Photobiol. 1989; 49: 587–594. droxyphenyl)chlorin (mTHPC) in solution: Biological
131. Brasseur N, Ali H, Langlois LJ, Van Lier J. Biological consequences in human tumor cells. Photochem. Photo-
activities of phthalocyanines-IX. Photosensitization of V- biol. 1998; 67: 584–590.
79 Chinese hamster cells and EMT-6 mouse mammary 146. Geirgakoudi I, Foster TH. Singlet oxygen versus non-
tumor by selectively sulfonated zinc phthalocyanines. singlet oxygen-mediated mechanisms of sensitizer photo-
Photochem. Photobiol. 1988; 47: 705–711. bleaching and their effects on photodynamic dosimetry.
132. Jori G, Reddi E. The role of lipoproteins in the delivery of Photochem. Photobiol. 1998; 67: 612–625.
tumor-targeting photosensitizers. Int. J. Biochem. 1993; 147. Spikes JD, Bommer JC. Photobleaching of mono-L-
25: 1369–1375. aspartyl chlorin e6 (NPe6): A condidate sensitizer for the
photodynamic treatment of tumors. Photochem. Photobiol.
133. Glanzmann T, Hadjur C, Zellweger M et al. Pharmaco-
1993; 58: 346–350.
kinetics of tetra(m-hydroxyphenyl)chlorin in human
148. Spikes JD. Quantum yields and kinetics of the photo-
plasma and individualized light dosimetry in photo-
bleaching of hematoporphyrin, Photofrin II, tetra(4-
dynamic therapy. Photochem. Photobiol. 1998; 67: 596–
sulfonatophenyl)-porphine and uroporphyrin. Photochem.
602.
Photobiol. 1992; 55: 797–808.
134. Bellnier DA, Henderson BW, Pandey RK et al. Murine
149. Potter WR, Mang TS, Dougherty TJ. Theory of photo-
pharmacokinetics and anti-tumor efficacy of the photo-
dynamic dosimetry: consequences of photo-destruction of
dynamic sensitizer 2-[1-hexyloxyethyl]-2-devinyl pyro-
sensitizer. Photochem. Photobiol. 1987; 46: 97–101.
pheophorbide-a. J. Photochem. Photobiol. B: Biol. 1999; 150. Boyle DG, Potter WR. Photobleaching Photofrin II as a
20: 55–61. means of eliminating skin photosensitivity. Photochem.
135. Star WM, Wilson BC, Patterson MS. Light delivery and Photobiol. 1987; 46: 997–1001.
optical dosimetry in photodynamic therapy of solid 151. McCaughan JS. Photodynamic therapy of endobronchial
tumors. In Photodynamic Therapy: Basic Principles and and esophageal tumors: an overview. J. Clin. Laser Med.
Clinical Applications. Henderson BW, Dougherty TJ. Surg. 1997; 14: 223–233.
(eds.) Marcel Dekker: New York, 1992; 335–368. 152. Overholt BF, Panjehpour M. Photodynamic therapy in
136. Svaasand LO, Potter WR. The implications of photo- Barret’s esophagus. J. Clin. Laser Med. Surg. 1996; 14:
bleaching for photodynamic therapy. In Photodynamic 245–249.
Therapy: Basic Principles and Clinical Applications. 153. Henderson BW, Dougherty TJ, Malone PB. Studies on the
Henderson BW, Dougherty TJ. (eds.) Marcel Dekker: mechanism of tumor destruction by photoirradiation
New York, 1992; 369–385. therapy. In Porphyrin Localization and Treatment of
137. Svaasand LO. Optical dosimetry for direct and interstitial Tumors. Doiron DR, Gomer CJ. (eds.) Alan R. Liss: New
photoirradiation therapy of malignant tumors. In Porphy- York, 1984; 601–612.
rin Localization and Treatment of Tumors. Doiron DR, 154. Fingar VH, Wieman TJ, Wiehle SA, Cerrito PB. The role
Gomer CJ. (eds.) Alan R. Liss: New York, 1984; 91–114. of microvascular damage in photodynamic therapy: the
138. Wilson BC, Jeeves WP, Lowe DM, Adam G. Light effects of treatment on vessel constriction, permeability,
propagation in animal tissues in the wavelength range and leukocyte adhesion. Cancer Res. 1992; 52: 4914–
375–825 nanometers. In Porphyrin Localization and the 4921.
Treatment of Tumors. Doiron DR, Gomer CJ. (eds.) Alan 155. Henderson BW, Waldow SM, Mang TS et al. Tumor
R, Liss: New York, 1984; 115–132. destruction and kinetics of tumor cell death in two
139. Bolin FP, Preuss LE, Cain BW. A comparison of spectral experimental mouse tumors following photodynamic
transmittance for several mammalian tissues: effects at therapy. Cancer Res. 1999; 45: 572–576.
PRT frequencies. In Porphyrin Localization and Treat- 156. Henderson BW, Fingar VH. Relationship between tumor
ment of Tumors. Doiron DR, Gomer CJ. (eds.) Alan R hypoxia and response to photodynamic therapy treatment
Liss: New York, 1984; 211–225. in an experimental mouse model. Cancer Res. 1987; 47:
140. Wilkisch PA, Jacka F. Studies of light propagation 3110–3114.
through tissue. In Porphyrin Localization and Treatment 157. Korbelik M. Induction of tumor immunity by photo-

Copyright # 2001 John Wiley & Sons, Ltd. J. Porphyrins Phthalocyanines 2001; 5: 105–129
BASIC PRINCIPLES OF PHOTODYNAMIC THERAPY 127

dynamic therapy. J. Clin. Laser Med. Surg. 1996; 14: modification of DNA by hematoporphyrin derivative.
329–334. Biochim. Biophys. Acta 1985; 475: 307–314.
158. Merhi Y, Guidoin R, Provost P et al. Increase in neutrophil 176. Fiel RJ, Datta-Gupt N, Mark E, Howard JC. Induction of
adhesion and vasoconstriction with platelet deposition DNA damage by porphyrin photosensitizers. Cancer Res.
after deep arterial injury by angioplasty. Am. Heart J. 1981; 41: 3543–3545.
1995; 129: 445–451. 177. Gomer CJ, Rucker N, Ferrario A, Murphree AL.
159. Chaudhuri K, Keck RW, Selman SH. Morphological Expression of potentially lethal damage in Chinese
changes of tumor microvasculature following hematopor- hamster cells exposed to hematoporphyrin derivative
phyrin derivative sensitized photodynamic therapy. photodynamic therapy. Cancer Res. 1986; 46: 3348–
Photochem. Photobiol. 1999; 46: 823–827. 3352.
160. Steele PM, Cheseboro JH, Stanson AW et al. Balloon 178. Moan J, Waksvik H, Christensen T. DNA single-stranded
angioplasty: natural history of the pathophysiological breaks and sister chromatid exchanges induced by treat-
response to injury in a pig model. Circ. Res. 1985; 57: ment with hematoporphyrin and light or by X-rays in
105–112. human NHIK 3025 cells. Cancer Res. 1981; 40: 2915–
161. McMahon KS, Wieman TJ, Moore PH, Fingar VH. 2918.
Effects of photodynamic therapy using mono-L-aspartyl 179. Evensen JF, Moan J. Photodynamic action and chromo-
chlorin e6 on vessel constriction, vessel leakage, and somal damage: a comparison of haematoporphyrin deriva-
tumor response. Cancer Res. 1994; 54: 5374–5379. tive (HpD) and light with X-irradiation. Br. J. Cancer
162. Steele PM, Cheseboro JH, Stanson AW et al. Natural 1982; 45: 456–465.
history of the pathophysiological response to injury in a 180. McNair FL, Marples B, West CML, Moore JV. A comet
pig model. Circ. Res. 1985; 57: 105–112. assay of DNA damage and repair in K562 cells after
163. Merhi Y, Lacoste L, Lam JYT. Neutrophil-platelet inter- photodynamic therapy using hematoporphyrin derivative,
actions and vasoconstriction. Circulation 1994; 90: 997– methylene blue and meso-tetrahydroxyphenylchlorin. Br.
1002. J. Cancer 1997; 75: 1721–1729.
164. Fuster V, Badimon L, Badimon J, Cheseboro JH. The 181. Gaullier J, Geze M, Santus R et al. Subcellular localiza-
pathogenesis of coronary artery disease and the acute tion and photosensitization by protoporphyrin IX in
coronary syndromes. N. Engl. J. Med. 1992; 5: 310–318. human keratinocytes and fibroblasts cultivated with 5-
165. Voet D, Voet JG. Arachidonate metabolism: prostaglan- aminolevulinic acid. Photochem. Photobiol. 1995; 62:
dins, prostacyclins, thromboxanes, and leukotrienes. Bio-
114–122.
chemistry. Wiley, New York: 1990; 658–665.
182. Kessel D, Woodburn K. Biodistribution of photosensitiz-
166. Agarwal ML, Larkin HE, Zaidi SIA et al. Phospholipase
ing agents. Int. J. Biochem. 1993; 10: 1377–1383.
activation triggers apoptosis in photosensitizer mouse
183. Geze M, Morliere P, Maziere JC et al. Lysosomes, a key
lymphoma. Cancer Res. 1993; 53: 5897–5902.
target of hydrophobic photosensitizers proposed for
167. Christensen T, Feren K, Moan J, Pettersen E. Photo-
chemotherapeutic applications. J. Photochem. Photobiol.
dynamic effects of hematoporphyrin derivative on syn-
B: Biol. 1993; 20: 23–25.
chronized and asynchronous cells of different origin. Br.
184. Leach MW, Higgins RJ, Autry SA et al. In vitro photo-
J. Cancer 1981; 44: 717.
dynamic effects of lysyl chlorin p6: cell suvival, localiza-
168. Berns MW, Dahlman A, Johnson FM et al. In vitro
cellular effects of hematoporphyrin derivative. Cancer tion and ultrastructural changes. Photochem. Photobiol.
Res. 1982; 42: 2325. 1993; 58: 653–660.
169. Salvatore G, Ambesi-Impiombato SS, Tromantano D et 185. Woodburn K, Vardaxis NJ, Hill JS et al. Subcellular
al. Effects of laser irradiation on hematoporphyrin-treated localization of porphyrins using confocal laser scanning
normal and transformed thyroid cells culture. Proc. 13 Int. microscopy. Photochem. Photobiol. 1991; 54: 725–732.
Cancer. Cong. 1982; 237. 186. Roberts WG, Berns MW. In vitro photosensitization I.
170. Fingar VH. Drug, light, and oxygen dependence of photo- Cellular uptake and subcellular localization of mono-L-
dynamic therapy of murine tumors. Roswell Park Cancer aspartyl chlorin e6, chloro-aluminum phthalocyanine, and
Institute, 1988. Photofrin II. Lasers Surg. Med. 1989; 9: 90–101.
171. Fingar VH, Potter WR, Henderson BW. Drug and light 187. Malik Z, Amit I, Rothmann C. Subcellular localization of
dose dependence of photodynamic therapy: A study of sulfonated tetraphenyl porphines in colon carcinoma cells
tumor cell clonogenicity and histologic changes. Photo- by spectrally resolved imaging. Photochem. Photobiol.
chem. Photobiol. 1987; 45: 643–650. 1997; 65: 389–396.
172. Wilson BC, Olivo M, Singh G. Subcellular localization of 188. Wood S, Holroyd JA, Brown SB. The subcellular localiza-
Photofrin and aminolevulinic acid and photodynamic tion of Zn(II) phthalocyanines and their redistribution on
cross-resistance in vitro in radiation-induced fibrosarcoma exposure to light. Photochem. Photobiol. 1997; 65: 397–
cells sensitive or resistant to Photofrin-mediated photo- 402.
dynamic therapy. Photochem. Photobiol. 1997; 65: 166– 189. Moan J, Berg K, Anholt A, Madslien K. Sulfonated
176. aluminum phthalocyanines as sensitizers for photochemo-
173. Sharkey SH, Wilson BC, Moorehead R, Singh G. therapy. Effects of small doses on localization, dye
Mitochondrial alterations in photodynamic therapy-resis- fluorescence and photosensitivity in V79 cells. Int. J.
tant cells. Cancer Res. 1993; 53: 4994–4999. Cancer 1994; 58: 865–870.
174. Adams KE, Espiritu M, Wilson BC, and Singh G. 190. Kessel D, Lou Y, Deng Y, Chang CK. The role of
Photodynamic therapy-resistant cells maintain their re- subcellular localization in the initiation of apoptosis by
sistance in vivo. Photochem. Photobiol. 63[S], 98S–98S. photodynamic therapy. Photochem. Photobiol. 1997; 65:
1996. Ref Type: Abstract. 422–426.
175. Gutter B, Speck WT, Rosenkranz HS. The photodynamic 191. Berg K, Moan J. Lysosomes and microtubules as targets

Copyright # 2001 John Wiley & Sons, Ltd. J. Porphyrins Phthalocyanines 2001; 5: 105–129
128 I. J. MACDONALD AND T. J. DOUGHERTY

for photochemotherapy of cancer. Photochem. Photobiol. tumors in C3H mice: electron microscopic and histo-
1997; 65: 403–409. pathalogic and biochemical evidence. Photochem. Photo-
192. Woodburn K, Fan Q, Miles DR et al. Localization and biol. 1993; 58: 771–776.
efficacy analysis of the phototherapeutic lutetium texa- 210. He X, Sikes RA, Thomsen S et al. Photodynamic therapy
phyrin (PCI-0123) in the EMT6 sarcoma model. Photo- with Photofrin II induces programmed cell death in
chem. Photobiol. 1997; 65: 410–415. carcinoma cell lines. Photochem. Photobiol. 1994; 59:
193. Kessel D. Chemical and biochemical determinants of 468–473.
porphyrin localization. In Porphyrin Localization and 211. Lou Y, Kessel D. Initiation of apoptosis versus necrosis by
Treatment Tumors. Doiron DR, Gomer CJ. (eds.) Alan R. photodynamic therapy with chloroaluminum phthalocya-
Liss: New York, 1984; 405–418. nine. Photochem. Photobiol. 1997; 66: 479–483.
194. Berg K, Moan J. Lysosomes as photochemical targets. Int. 212. Kick G, Messer G, Plewig G et al. Strong and prolonged
J. Biochem. 1994; 59: 814–822. induction of c-jun and c-fos proto-oncogenes by photo-
195. Okada JG, Rechsteiner M. Introduction of macromol- dynamic therapy. Br. J. Cancer 1996; 74: 30–36.
ecules into cultured mammalian cells by osmotic lysis of 213. Fisher AMR, Danenberg K, Nanerjee D et al. Increased
pinocytotic vesicles. Cell 1982; 29: 33–41. photosensitivity in HL60 cells expressing wild-type p53.
196. Wilson PD, Firestone RA, Lenard J. The role of lysosomal Photochem. Photobiol. 1997; 66: 265–270.
enzymes in killing of mammalian cells by the lysoso- 214. Apoptosis 1996, publication of Oncogene Research
motrpoic detergent N-dodecylimidizole. J. Cell Biol. Products, 1996. Calbiochem. Ref Type: Pamphlet.
1987; 104: 1223–1229. 215. Kessel D, Luo Y. Photodynamic therapy: A mitochondrial
197. Miller DK, Griffiths E, Lenard J, Firestone RA. Cell inducer of apoptosis. Cell Death Diff. 1999; 6: 28–35.
killing by lysosomotropic detergents. J. Cell Biol. 1983; 216. Kessel D, Luo Y. Mitochondrial photodamage and PDT-
97: 1841–1851. induced apoptosis. J. Photochem. Photobiol. B: Biol.
198. De Duve C, De Barsy T, Poole B et al. Lysosomotropic 1998; 42: 89–95.
agents. Biochem. Pharm. 1974; 23: 2495–2531. 217. Zamzami N, Marchetti P, Castedo M et al. Sequential
199. Murant RS, Gibson SL, Hilf R. Photosensitizing effects of reduction of mitochondrial transmembrane potential and
Photofrin II on the site-selected mitochondrial enzymes generation of reactive oxygen species in early apoptosis.
adenylate kinase and monoamine oxidase. Cancer Res. J. Exp. Med. 1995; 182: 367–377.
1987; 47: 4323–4328. 218. Zamzami N, Marchetti P, Castedo M et al. Reduction in
200. Gibson SL, Hilf R. Photosensitization of mitochondrial mitochondrial potential constitutes an early irreversible
cytochrome c oxidase by hematoporphyrin derivative and step of programmed cell death in vivo. J. Exp. Med. 1995;
related porphyrins. Cancer Res. 1983; 43: 4191–4197. 181: 1661–1672.
201. Kroemer G, Zamzami N, Susin SA. Mitochondrial control 219. Zamzami N, Susin SA, Marchetti P et al., Mitochondrial
of apoptosis. Immunol Today 1997; 18: 44–51. control of nuclear apoptosis. J. Exp. Med. 1996; 183:
202. Hirsch T, Decaudin D, Susin SA et al. PK11195, a ligand 1533–1544.
of the mitochndrial benzodiazepine receptor, facilitates 220. Petit PX, Goubern M, Diolez P et al. Disruption of outer
the induction of apoptosis and reverses Bcl-2-mediated mitochondrial membrane as a result of large amplitude
cytoprotection. Exp. Cell Res. 1998; 241: 426–434. swelling: the impact of irreversible permeability transi-
203. Miccoli L, Beurdeley-Thomas A, De Pinieux G et al. tion. FEBS Lett. 1998; 426: 111–116.
Light induced photoactivation of hypericin affects the 221. Marzo I, Susin SA, Petit PX et al. Caspases disrupt
energy metabolism of human glioma cells by inhibiting mitochondrial membrane barrier function. FEBS Lett.
hexokinase bound to mitochondria. Cancer Res. 1998; 58: 1998; 427: 198–202.
5777–5786. 222. Zhou P, Chou J, Olea RS et al. Solution structure of Apaf-
204. Atlante A, Moore G, Passarella S, Salet C. Hematopor- 1 CARD and its interaction with caspase-9 CARD: A
phyrin derivative (Photofrin II): impairment of anion structural basis for specific adaptor/caspase interaction.
translocation. Biochem. Biophys. Res. Commun. 1986; Proc. Natl. Acad. Sci. USA 1999; 96: 11265–11270.
141: 584–590. 223. Ruck A, Dolder M, Wallimann T, Brdiczka D. Recon-
205. Popadopoulos V, Brown S. Role of the peripheral-type stituted adenine nucleotide translocase forms a channel
benzodiazepine receptor and the polypeptide diazepam for small molecules comparable to the permeability
binding inhibitor in steroidogenesis. J. Steroid. Biochem. transition pore. FEBS Lett. 1998; 426: 97–101.
Mol. Biol. 1995; 53: 103–110. 224. Petit PX, Susin SA, Zamzami N et al. Mitochondria and
206. Perlin DS, Murant RS, Gibson SL, Hilf R. Effects of programmed cell death: back to the future. FEBS Lett.
photosensitization by hematoporphyrin derivative on 1996; 396: 7–13.
mitochondrial adenosine triphosphatase-mediated proton 225. He J, Agarwal ML, Larkin HE et al. The induction of
transport and membrane integrity of R3230AC mammary partial resistance to photodynamic therapy by the proto-
adenocarcinoma. Cancer Res. 1995; 45: 653–658. oncogene bcl-2. Photochem. Photobiol. 1996; 64: 845–
207. Lemasters JJ. Mechanisms of hepatic toxicity, V. 852.
necraptosis and the mitochondrial permeability transition: 226. Kim HR, Luo Y, Li G, Kessel D. Enhanced apoptotic
shared pathways to necrosis and apoptosis. Am. J. Physiol. response to photodynamic therapy after bcl-2 transfection.
1999; 276: G1–G6. Cancer Res. 1999; 59: 3429–3432.
208. Agarwal ML, Clay ME, Harvey HE et al. Photodynamic 227. Varnes ME, Chiu S, Xue L, Oleinick NL. Photodynamic
therapy induces rapid cell death by apoptosis in L5178Y therapy-induced apoptosis in lymphoma cells: transloca-
mouse lymphoma cells. Cancer Res. 1991; 51: 5993– tion of cytochrome c causes inhibition of respiration as
5996. well as caspase activation. Biochem. Biophys. Res.
209. Zaidi SIA, Oleinick NL, Zaim MT, Mukhtar H. Apoptosis Commun. 1999; 255: 673–679.
during photodynamic-therapy induced ablation of RIF-1 228. Musser DA, Fiel RJ. Cutaneous photosensitizing and

Copyright # 2001 John Wiley & Sons, Ltd. J. Porphyrins Phthalocyanines 2001; 5: 105–129
BASIC PRINCIPLES OF PHOTODYNAMIC THERAPY 129

immunosuppressive effects of a series of tumor localizing 234. Bellnier DA. Potentiation of photodynamic therapy in
porphyrins. Photochem. Photobiol. 1991; 53: 119–123. mice treated with recombinant human tumor necrosis
229. Lynch DH, Haddad S, King VJ et al. Systemic immuno- factor. J. Photochem. Photobiol. B: Biol. 1991; 8: 203–
suppression induced by photodynamic therapy (PDT) is 210.
adoptively transferred by macrophages. Photochem. 235. Evans S, Mathews W, Perry R et al. The effect of photo-
Photobiol. 1989; 49: 453–458. dynamic therapy on tumor necrosis factor production by
230. Korbelik M, Dougherty GJ. Photodynamic therapy- murine macrophages. J. Natl. Cancer Inst. 1990; 82: 34–
mediated immune response against subcutaneous mouse 39.
tumors. Cancer Res. 1999; 59: 1941–1946. 236. Schumaker BP, Hetzel FW. Clinical laser photodynamic
231. Korbelik M, Naraparaju VR, Yamamoto N. Macrophage- therapy in the treatment of bladder cancer. Photochem.
directed immunotherapy as adjuvant to photodynamic Photobiol. 1987; 46: 899–901.
therapy of cancer. Br. J. Cancer 1997; 75: 202–207. 237. Kick G, Messer G, Goetz AE et al. Photodynamic therapy
232. Nseyo UO, Whalen RK, Duncan MR et al. Urinary induces expression of interleukin-6 by activation of AP-1
cytokines following photodynamic therapy for bladder but not NF-kB DNA binding. Cancer Res. 1995; 57:
cancer. Urology 1990; 36: 167–171. 2373–2379.
233. Herman S, Kalechman Y, Gafter U et al. Photofrin II 238. Gollnick SO, Liu X, Owczarczak B et al. Altered
induces cytokine secretion by mouse spleen cells and expression of interleukin-6 and interleukin-10 as a result
human peripheral mononuclear cells. Immunopharm. of photodynamic therapy in vivo. Cancer Res. 1997; 57:
1996; 31: 195–204. 3904–3909.

Copyright # 2001 John Wiley & Sons, Ltd. J. Porphyrins Phthalocyanines 2001; 5: 105–129

You might also like