Journal of Photochemistry and Photobiology B: Biology 57 (2000) 142–148www.elsevier.nl / locate / jphotobiol

The phototoxicity of photodynamic therapy may be suppressed orenhanced by modulation of the cutaneous vasculature

*´Salvador Gonzalez, Chitralada Vibhagool, Margaret Sherwood, Thomas J. Flotte, Nik KolliasWellman Laboratories of Photomedicine, Department of Dermatology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA

Received 9 June 2000; accepted 19 July 2000

Abstract

In photodynamic therapy, the threshold for light induced toxicity depends on the drug concentration and the light dose. This study wasaimed to show for vascular photosensitizers that the toxicity threshold on normal tissue may be predictably modified by modulation of thecutaneous vasculature. Albino rabbits were injected with 1.0 mg/kg of a vascular photosensitizer, benzoporphyrin derivative monoacidring-A. The threshold light dose for toxicity to normal skin was determined at an absorption maximum of the drug (694 nm), 1 h afterdrug injection. The cutaneous vasculature was dilated by prior skin exposure to ultraviolet radiation or was constricted by iontophoreticapplication of epinephrine. Threshold toxicity was determined clinically and by assessing the effective concentration of hemoglobin in theskin by diffuse reflectance spectroscopy (DRS). Tissue samples that received threshold doses were investigated with light and electronmicroscopy. The toxicity threshold increased by 3.260.9 (mean6S.D.) following vasoconstriction and decreased by 3.660.8 followingvasodilation, compared to control sites. Light and electron microscopy showed similar findings at threshold for both vasodilated andvasoconstricted sites. Therefore vascular modulation may be used to predictably enhance or suppress the level of phototoxicity of normalskin. 2000 Elsevier Science S.A. All rights reserved.

Keywords: Vascular modulation; Benzoporphyrin derivative; Skin phototoxicity; Erythema; UVB radiation; Epinephrine-iontophoresis

1. Introduction acid ring A, Verteporfin ) is a second generation vascularphotosensitizing drug which distributes to both, vascula-

Photodynamic therapy (PDT) involves the activation of ture and tissues, and has been found to produce tissuea photosensitizer with light of the appropriate wavelength necrosis following damage to the skin vasculature. Ato produce reactive oxygen species, including singlet concentration gradient appears to be established betweenoxygen [1] which produces focal damage to tissue [2,3]. the vessels and the surrounding tissue that decreases withPDT has been approved in several countries for treating time after drug administration. In the skin of the New

tumors with Photofrin , while a second generation of PDT Zealand albino rabbit (NZW), the threshold dose todrugs has been introduced and are going through clinical produce a minimal purpura reaction that leads to skintrials [4,5]. PDT drugs may be classified by the compart- necrosis and scarring has been found to depend strongly onment they concentrate in as their main target, for example, the plasma drug level and only minimally on the tissuein the skin this may be vascular or tissue specific. A drug concentration at times 1–4 h after drug injection [6].vascular drug upon exposure to light of the right wave- In dermatology, BPD-MA has been used in clinical trialslength would produce damage to the endothelial cells or for skin tumors and for psoriasis [5]. It shows skincause clotting and blockage of the vessels. A tissue drug photosensitivity that usually resolves in a few days de-that localizes in the epidermis or the dermis when exposed pending on the infused drug concentration, there is noto light would produce necrosis of the epidermis or the prolonged drug induced photosensitivity as with photofrindermis [6]. BPD-MA (benzoporphyrin derivative mono- [7–9]. While the results from the skin tumor studies [10]

have been encouraging in terms of tumor eradiacation, theuninvolved perilesional skin is frequently injured because

*Corresponding author. Johnson & Johnson, Consumer Productsof the presence of drug during light exposure. AnyWorlwide, 199 Grandview Road, Skillman, NJ 08558-9418, USA. Tel.:successful therapy has to spare the perilesional normal skin11-908-8741-343; fax: 11-908-8741-317.

E-mail address: nkollia@cpcus.jnj.com (N. Kollias). while treating the involved lesion.

1011-1344/00/$ – see front matter 2000 Elsevier Science S.A. All rights reserved.PI I : S1011-1344( 00 )00089-0

´S. Gonzalez et al. / Journal of Photochemistry and Photobiology B: Biology 57 (2000) 142 –148 143

BPD-MA is primarily a vascular drug and PDT with battery powered controllable direct current source (Phor-BPD-MA results in vascular occlusion and damage to the eser II system, IOMED, Salt Lake City, UT, USA). Thevascular endothelial cells [9]. Vascular modulation as in electrodes were placed at least 10 cm apart on the back ofvasoconstriction and vasodilation may result in suppression the animals. A 3-ml aliquot of the prepared epinephrine

2or enhancement of PDT induced toxicity with BPD-MA or solution was placed under the anode (232 cm ) andother vascular drugs. The goal of the present study was to distilled water was placed under the cathode. The applieddetermine the extent to which the phototoxicity induced by current was 0.2 mA for 5 min which corresponded to 60PDT with BPD-MA on normal skin could be predictably mV. Skin sites were then carefully and thoroughly dried inadjusted by modulation of the cutaneous vasculature. order to remove residual epinephrine solution. The vas-Elucidation of this dependence can shed light on the PDT oconstriction lasted approximately 45 min, that is 15–20action of vascular photosensitizers and a means to provide min beyond the laser exposure.protection for normal skin during PDT of skin diseases.

2.2. Photosensitizing agent

2. Materials and methods Liposomal Benzoporphyrin Derivative Mono Acid RingA (BPD-MA) was supplied by Quadra Logic Technologies

2.1. Animals and modification of cutaneous vasculature Inc. (Vancouver, BC, Canada) and was stored in the dark at48C until use. Liposomal BPD-MA was reconstituted in

The New Zealand albino rabbit (Millbrook Breeding sterile water (Abbott Laboratories, North Chicago, IL,Farms, Amherst, MA, USA) was chosen as the animal USA) at 1.47 mg BPD/cc and was used within 72 h aftermodel due to the availability of many skin sites for reconstitution. The drug was intravenously administered, 1treatment, easy i.v. access for drug administration, and mg/kg body weight, via the peripheral ear vein using aprevious experience with this animal model. The animals 24-gauge angiocatheter (Becton Dickinson Vascular Ac-weighed between 3 and 4 kg and were prepared as cess, Sandy, UT, USA) followed by a flush with 2.5 ml ofpreviously described [11]. Briefly, fur on the back was 5% dextrose in water (Baxter Healthcare Corp., Deerfield,clipped and thereafter carefully shaved in order to obtain a IL, USA).skin surface with minimum hair interference in lightdosimetry. In order to assure equivalent photodynamic 2.3. PDT light exposureresponses, skin sites were randomly selected, avoidingobviously non-uniform skin, such as cervical, vertebral, Visible radiation at a wavelength of 694 nm and an

2and extremely lateral areas as well as areas of dense hair irradiance of 100 mW/cm (CR 599 Dye Laser; Coherent,growth. Four animals were pretreated with UVB radiation Palo Alto, CA, USA) was delivered to the skin through aand ten animals were treated with epinephrine before microlens-tipped quartz fiber (Quadra Logic Technologicsreceiving PDT light exposure. Inc., Vancouver BC, Canada). The irradiance was measured

Ultraviolet B radiation (290–320 nm) was used to before and after every exposure with a power meterinduce cutaneous vasodilation [12]. The irradiation source (Model 210, Coherent). The laser beam, 3.5 cm in

2was a bank of 12 UVB-HO-908 fluorescent tubes (Elder diameter (10 cm ) was adjusted perpendicular to thePharmaceuticals Inc., Cleveland, OH). The spectral ir- surface of the skin.radiance of these lamps was measured with a spec- For PDT treatment, the animals were always treatedtroradiometer (model 742, Optronics Inc., Orlando, FL, with light 1 h after i.v. administration of BPD-MA. FourUSA). The UVB intensity of the light was quantified and skin sites 1 cm in diameter were simultaneously exposed inmonitored with a IL-1700 radiometer equipped with a SED order to test four different fluences at increments of 40%.240 UVB detector (International Light Inc., Newburyport, The ranges of light doses, determined based upon previous

2MA, USA). The minimum perceptible erythema dose experience [11], were (i) in control sites, 8.4–27 J /cm ;2 2(MED) was determined to be 60 mJ/cm . One site (535 (ii) in epinephrine-treated sites, 8.4–56.0 J /cm ; and (iii)

2 2cm ) on each animal was irradiated with 120 mJ/cm of in skin sites previously exposed to UVB radiation, 3.9–2UVB (2 MED) in order to obtain a well expressed 12.0 J /cm .

vasodilatory response. This dose of UVB did not producean intense edema. PDT was delivered 24 h after UVB 2.4. Evaluation of erythema response by diffuseexposure. reflectance spectroscopy

On the other hand, diluted epinephrine (2%, 1:5000)was delivered into the skin of the rabbits by iontophoresis Diffuse reflectance spectroscopy (DRS) was used be-immediately before light exposure in order to obtain a well cause it is a very sensitive and reproducible method forcontrolled and reproductible vasoconstrictive response. assessing apparent oxyhemoglobin (HbO ) concentration2

Iontophoresis was carried out following injection with the as a measure of skin erythema [13,14]. The attenuation ofphotosensitizer (15 min before light exposure) using a light incident on skin allows us to obtain information about

´144 S. Gonzalez et al. / Journal of Photochemistry and Photobiology B: Biology 57 (2000) 142 –148

blood content within the upper dermis [13]. Reflected light 2.7. Statistical treatmentcontaining such information was analyzed by comparingspectra obtained before and after each intervention and Mean and standard deviation from the mean for MPuDthen calculated the changes in the apparent concentration and apparent HbO concentration in each experimental2

of HbO . The DRS spectrophotometer was a Hewlett- situation were obtained. A one-way analysis of the vari-2

Packard 8452A DAS fitted with optical fiber bundles [13]. ance (ANOVA) and Turkey Kramer as multiple com-The joined end of a bifurcated fiber bundle was placed parison post-tests were made.against the skin, one leg brought light to the skin while theother delivered the remitted light from the skin to thespectrograph. DRS spectra from the skin were recorded 3. Resultsevery 2 nm, at wavelengths ranging from 400 to 820 nm.DRS measurements were carried out (i) before and 24 h 3.1. Clinical responsesafter cutaneous exposure to UVB radiation; (ii) before and15 min after iontophoresis-delivered epinephrine; and (iii) The minimum threshold light dose to produce purpura 2immediately before and 1 h after PDT treatment (laser days after exposure has been determined to correlateexposure). strongly with skin necrosis resulting in scar formation 2

weeks after PDT [11]. It was determined by clinicalexamination that PDT treatment of UVB-irradiated skin

2.5. Clinical scoring sites and epinephrine-treated skin sites had a similarclinical follow-up as PDT-treatment of control skin sites. It

Clinical responses were carefully monitored for 5 days was found that 694 nm laser light or BPD-MA alone didfollowing PDT treatment. Cutaneous responses such as not produce skin phototoxicity on either UVB or epineph-erythema, blanching, edema, purpura, scab, and scar were rine treated skin sites (data not shown). MPuD in the

2individually assessed on a five point scale. BPD-MA control skin sites was 13.866.7 (mean6S.D.) J /cm ,induced phototoxicity has been characterized in terms of which was significantly greater than in the skin site

2the threshold fluence to produce purpura (MPuD) [11]. previously exposed to UVB radiation (3.961.3 J /cm )MPuD corresponds to the minimum threshold light dose (P,0.05), and significantly lower than the MPuD in

2required for purpura 2 days after exposure and correlates epinephrine-treated skin (40.7612.4 J /cm ) (P,0.05).strongly with skin necrosis resulting in scar formation 2 The MPuD in the epinephrine-treated sites increased by aweeks after PDT. Cutaneous responses were graded as factor of 3.260.9 (mean6S.D.; range: 0.2–0.5). Thefollows: (–) no response; (6) just perceptible response, MPuD in the UVB-treated sites decreased by a factor ofdiameter smaller than 1 cm; (11) response with clear 3.660.8.margins of identical size to the exposed site; (21) wellpronounced response extending beyond the limits of 3.2. Changes in apparent oxyhemoglobin concentrationexposure; (31) marked response extending more than 0.5cm beyond the borders of the exposed skin site. All clinical From the DRS measurements (spectra) we determinedassessments were made and recorded independently by two the difference between all PDT treated sites and those ofinvestigators. adjacent control sites. The difference in apparent HbO2

concentration between treated and control sites was calcu-lated [14] (Table 1). A significant increase (P,0.0001) of

2.6. Skin biopsies the apparent HbO concentration was found in all PDT-2

treated and UVB-exposed sites. Epinephrine-treated skinSkin specimens from sites receiving MPuD immediately sites showed a significant reduction (P,0.001) of HbO2

after laser exposure were obtained for transmission elec- concentration compared to PDT-treated control skin sitestron microscopy (TEM) and routine light microscopy. For and to those PDT-treated sites that were previously treatedTEM, 2-mm punch biopsies were further dissected lon- with epinephrine or UVB radiation. After PDT treatment,gitudinally and fixed in 4% glutaraldehyde in 0.1 M all the epinephrine-treated skin sites showed a significantcacodylate buffer (pH 7.4) overnight at 48C. They were increase of HbO concentration making the exposed sites2

then post-fixed in 2% osmium tetroxide in cacodylate indistinguishable from the surrounding unexposed skin,buffer for 2 h, dehydrated in a graded ethanol series and i.e., abrogating completely the vasoconstrictor effect of theembedded in Epon 812 epoxy resin (Tousimis Research epinephrine.Corporation, Rockville, MA, USA). Ultrathin sectionswere cut on an ultramicrotome (Reichert-Jung Ultracut, 3.3. Histological and ultrastructural findingsVienna, Austria), stained with uranyl acetate and leadcitrate, and examined with a transmission electron micro- The UVB and epinephrine-treated animals showedscope (model CM10, Philips, Eindhoven, Netherlands). similar changes, however, the relative quantities of the

´S. Gonzalez et al. / Journal of Photochemistry and Photobiology B: Biology 57 (2000) 142 –148 145

Table 1Cutaneous changes in the apparent HbO concentrations before and after2

PDT following vascular modulation by either epinephrine or exposure toaUVB radiation

Skin site HbO (mean6S.D.)2

b,cNon-PDT, but epinephrine treated site 20.3360.08bNon-PDT, but UVB-irradiated site 0.5060.50

PDT response of control untreated site 0.2260.20bPDT response of epinephrine treated site 20.0560.20bPDT response of UVB-irradiated site 0.6460.03

a The values of apparent oxyhemoglobin concentration correspond tothose values obtained after normalizing from internal control baselinevalues.

b Compared to PDT response of control untreated site (site nonpreviously treated neither with epinephrine nor exposed to UVB radia-tion) (P,0.0001).

c Compared to PDT response of control site, and of sites previouslytreated with epinephrine or UVB exposure (P,0.0001).

damage varied. The overall histological evaluation demon-strated a prominent papillary dermal edema, stasis of bloodin blood vessels, slight perivascular mixed inflammatorycell infiltrate composed of lymphocytes and neutrophils, orfocal extravasation of red blood cells. No significantepidermal alterations were found. The epinephrine-treatedanimals showed more prominent papillary dermal edema(Fig. 1a). The UVB-treated animals showed more promi-nent vascular stasis and extravasation of red blood cells(Fig. 1b). Unequivocal damage to the vessel walls couldnot be identified.

Transmission electron microscopy of control skin speci-2mens exposed at MPuD (8.4–12 J /cm ) showed damage

to endothelial cells and pericytes immediately after laserexposure. Endothelial cells showed condensation andblebbing of the nuclear chromatin, disruption of organellesand cell membranes and increased vacuolization. Epi-

2nephrine-treated skin sites exposed to MPuD (56 J /cm )2and UVB-treated sites exposed to MPuD (3.9 J /cm ) both

showed similar damage to endothelial cells and isolatedfibroblasts. Fibroblasts, in both, exhibited swollen mito-chondria, disrupted cristae, dilated endoplasmic reticulumand phagocytic inclusions. Endothelial cells showed swol-len mitochondria with loss of cristae, perinuclear spacesand peripheral condensation of chromatin with somevacuolization (Fig. 2a and b). Intravascular congestion of

Fig. 1. Light micrographs immediately after laser exposure. (a) Epi-small blood vessels (Fig. 2b) and occasional apoptotic 2nephrine-treated skin exposed to MPuD (56 J /cm ). Note no alterationsfigures (Fig. 2c) were also observed in UVB-treated skin of the epidermis, prominent dermal edema and only occasional congestion

2sites exposed to MPuD (3.9 J /cm ). of blood vessels. (H&E) Scale bar, 100 mm. (b) UVB-treated skin2exposed to MPuD (3.9 J /cm ). Note the dermal edema and numerous

congested blood vessels (arrows). (H&E) Scale bar, 100 mm.

4. Discussionvascular photosensitizer with the major PDT effects con-

PDT-induced skin phototoxicity with systemic photo- fined to the vasculature especially in the 4 h followingsensitizers depends on a series of factors: drug concen- drug infusion [9,11,15]. The concentration of BPD-MA intration in tissue and vasculature, light fluence and fluence the plasma is maximal immediately after injection andrate, wavelength of exposure, oxygen availability in tissue, decreases exponentially with time [6,9] while the con-etc. [15]. It has been demonstrated that BPD-MA is a centration of the sensitizer outside the vessels is one tenth

´146 S. Gonzalez et al. / Journal of Photochemistry and Photobiology B: Biology 57 (2000) 142 –148

2Fig. 2. Transmission electron micrographs immediately after laser exposure. (a) Epinephrine-treated skin exposed to MPuD (56 J /cm ). Small blood vesselshowing intravascular plasma. Note the peripheral condensation of nuclear chromatin (C) and swollen mitochondria (M) in the endothelial cell. Scale bar, 1

2mm. (b) UVB-treated skin exposed to MPuD (3.9 J /cm ). Blood vessel showing intravascular congestion. Note the swollen mictochondria with disrupted

2cristae in the endothelial cells (arrows). Scale bar, 2 mm. (c) UVB-treated skin exposed to MPuD (3.9 J /cm ). Perivascular cell demonstrating histologicmanifestations of apoptosis. Note the dense cytoplasm, nuclear condensation, and cytoplasmic vacuolization. Scale bar, 2 mm.

(0.1) that in the vessels. The goal of the present study was never complete, we do not expect the PDT response to beto give a better understanding to the action of a vascular affected by a decrease in the O supply. On the other hand,2

photosensitizer (BPD-MA) in PDT and to show that we know that cutaneous exposure to UVB radiationmodulation of the skin vasculature may affect the photo- induces vasodilation as well as expression of adhesiontoxic response of the skin to PDT as a consequence of drug molecules and mediator release such as cytokines andavailability /concentration. arachidonic acid metabolites; however, 24 h after a UV

The drug concentration in the tissue and the plasma dose of 2 MED does not produce pronounced changes inlevels are the critical paramaters in predicting skin photo- the vascular or perivascular volume tissue other than atoxicity. Modulation of the vasculature, either by vasocon- mild vasodilation. This assumption was supported by thestriction or vasodilation, produced significant changes in clinical expression of phototoxicity. The PDT-thresholdthe clinically observed phototoxic threshold reactions in reactions were identical in normal, vasodilated, and innormal skin. Since epinephrine-induced vasoconstriction is vasoconstricted skin. Thus, the vascular modulatory action

´S. Gonzalez et al. / Journal of Photochemistry and Photobiology B: Biology 57 (2000) 142 –148 147

is the dominant effect which is shown by the presence of may indicate some cell membrane damage. It was notprominent thrombosis without significant epidermal dam- observed in epinephrine-treated skin exposed to MPuD.age in case of PDT-treatment of UVB-irradiated skin The major problem in clinical PDT for skin tumors andobserved in the light microscopy studies (Fig. 1). In both for psoriasis is the PDT-induced necrosis of the normalcases, vasoconstriction and vasodilation, the increase and skin in the immediate vicinity of the involved tissue [7,16].decrease were of the order of three times. This implies that Simply masking the normal tissue from light duringvascular modulation may alter the threshold phototoxic exposure does not resolve the problem because subclinicallight dose by a factor of six. This dynamic range holds the disease which needs to be treated may exist beyond thepotential for providing protection to normal tissue while clinical boundaries. Secondly, PDT with systemic BPD-eradicating diseased tissue such as tumors. It has been MA in clinical trials shows a pronounced vasodilation inobserved in human studies [5,10] that the response of the normal skin surrounding the involved lesion duringnormal skin to PDT with BPD-MA involves a vasodilation light exposure leading frequently to a superthresholdthat includes a wheal and flare reaction while the lesional phototoxic dose in the normal skin. This may result inskin (tumor or psoriasis) becomes dusky due to vascular necrosis due to the fast rise of the dose–response. Addi-occlusion. Thus, the vascular reactions induced by PDT in tional studies to investigate vascular modulation of PDTnormal and diseased skin are different. phototoxicity in proliferative skin conditions are war-

The change in apparent oxyhemoglobin (HbO ) con- ranted.2

centration produced by vascular modulation may be pre-dictive of the changes in the level of phototoxicity. Wefound that the changes in apparent HbO concentration2 5. Conclusioncorresponded to changes in the MPuD. When the hemoglo-bin content of the skin increased so did the photosensitivity

Prevention of damage to normal skin during PDTof the skin, expressed by a decrease in the MPuD.treatment without altering tumor response is critical. WeIontophoretic delivery of epinephrine produced a markedhave shown, albeit in an animal model, that reduction ofconstriction of the arteriolar blood (HbO ) and resulted in2 the vascular capacity provides significant protection froman increase of the MPuD by a factor of 3.2. VasodilationPDT damage to normal skin. The modulation of the dermalinduced by UVB exposure (at the level of this research)vasculature and the changes in the local blood supplyproduced a greater than two-fold increase in hemoglobinduring PDT will affect the efficacy of this treatmentconcentration in the skin and a 3.6-fold decrease in themodality and may yield a protection factor of six. TheMPuD. The results makes sense since the phototoxicityamount of protection may be estimated by non-invasivedepends quadratically on the plasma drug concentrationmeasurement of the changes in apparent HbO concen-2[6], e.g., when the drug concentration in the plasma istration following a vascular intervention as assessed byincreased by a factor of two the light fluence to causediffuse reflectance spectroscopy.threshold toxicity is decreased by a factor of four. In this

study the drug concentration was not changed, simply thevessel diameter. And when the total drug per unit length ofthe vessel is decreased by a factor of two the fluence of 6. Abbreviationslight to produce threshold purpura is increased by almost afactor of four. Since an increase in the vessel diameter PDT Photodynamic therapyresults in more HbO as well as more BPD-MA we could2 BPD-MA Benzoporphirin derivative monoacid ring-Ause reflectance to monitor the changes in the total BPD- DRS Diffuse reflectance spectroscopyMA through the changes in the HbO concentration in the2 HbO Oxyhemoglobin2tissue. Thus HbO could prove a useful dosimeter for2 Hb Hemoglobinvascular drugs. Since the drug is primarily in the plasma it UVB Ultraviolet B radiationmakes sense that changing the total blood volume would NZW New Zealand albino rabbitresult in inversely proportional changes in the threshold MPuD Threshold fluence to produce purpuralevel of phototoxicity which turns out to be the case TEM Transmission electron microscopyclinically. It is interesting that at the microscopic level,PDT effects depend on the microscopic drug concentrationwhile the macroscopic level PDT effects may be predictedby amount of the average drug concentration in the tissue– Acknowledgementsblood vessel ensemble.

The ultrastructural findings indicate that the primary We express our special thanks to Dr Anna Richter fortarget for PDT were the endothelial cells which may result her critical reading of the manuscript and her valuablein loss of adhesion and leakage of serum. In the dilated input. This study was partially supported by Quadra Logicvessels (UVB-treated skin), the red blood cells from clots Technologies Inc. (Vancouver, BC, Canada).

´148 S. Gonzalez et al. / Journal of Photochemistry and Photobiology B: Biology 57 (2000) 142 –148

J.G. Levy, Mouse skin photosensitization with benzoporphyrinReferencesderivative and photofrin: macroscopic and microscopic evaluation,Photochem. Photobiol. 53 (1991) 281–286.

[1] M.S. Patterson, S.J. Madsen, B.C. Wilson, Experimental tests of the [10] H. Lui, Photodynamic therapy in dermatology with porfirmerfeasibility of singlet oxygen luminescence monitoring in vivo during sodium and benzoporphyrin derivative: an update, Semin. Oncologyphotodynamic therapy, J. Photochem. Photobiol. B. Biol. 5 (1990) 21 (6) (1994) 4–10, Suppl. 15.69–84. [11] M.M. Tsoukas, G.C. Lin, M.S. Lee, R.R. Anderson, N. Kollias,

[2] T.J. Dougherty, Photosensitizers: therapy and detection of malignant Predictive dosimetry for threshold phototoxicity in photodynamictumors, Photochem. Photobiol. 45 (1987) 879–889. therapy on normal skin: red wavelengths produce more extensive

[3] T.J. Dougherty, Photodynamic therapy, Phototchem. Photobiol. 58 damage than blue at equal threshold doses, J. Invest. Dermatol. 108(1993) 895–900. (1997) 501–505.

[4] C.J. Gomer, Preclinical examination of first and second generation [12] B.A. Gilchrest, N.A. Soter, J.S. Stoff, M.C. Mihm Jr., The humanphotosensitizers used in photodynamic therapy, Photochem. Photo- suburn reaction: histologic and biochemical studies, J. Am. Acad.biol. 54 (1991) 1093–1107. Dermatol. 5 (1981) 411–422.

[5] H. Lui, L. Hruza, N. Kollias, J. Wimberly, R.R. Anderson, Photo- [13] N. Kollias, A. Baqer, K.R. Naqvi, Fiber optic spectrometer fordynamic therapy of malignant skin tumors with benzoporphyrin noninvasive transmission and diffuse reflection studies, Spectr. Lett.derivative-monoacid ring A (BPD-MA): preliminary observations, 19 (1986) 149–165.SPIE Proc. 1876 (1993) 147–151. [14] W.T. Knoeffel, N. Kollias, D.W. Rattner, N. Nishioka, A.L. War-

´[6] G.C. Lin, M.M. Tsoukas, M. Lee, S. Gonzalez, Ch. Vibhagool, R.R. shaw, Assessment of pancreatic microcirculation and oxygen deliv-Anderson, N. Kollias, Skin necrosis due to photodynamic action of ery by diffuse reflectance spectroscopy in the rat pancreas, J. Appl.benzoporphyrin depends on circulating rather than tissue drug Physiol. 80 (1996) 116–123.levels: implications for control of photodynamic therapy, Photo- ´[15] M.M. Tsoukas, S. Gonzalez, T.J. Flotte, R.R. Anderson, M.E.chem. Photobiol. 68 (1998) 575–583. Sherwood, N. Kollias, Wavelength and fluence effect on vascular

[7] A.M. Richter, B. Kelly, J. Chow, D.J. Liu, G.H.N. Towers, D. damage with photodynamic therapy on skin, J. Invest. Dermatol.Dolphin, J.G. Levy, Preliminary studies on a more effective 114 (2000) 303–308.phototoxic agent than hematoporphyrin, J. Nat. Cancer Inst. 79 [16] H. Barr, S.G. Bown, Normal tissue damage following photodynamic(1987) 1327–1332. therapy: are there biological advantages?, in: B.W. Henderson, T.J.

[8] A.M. Richter, E. Stenberg, E. Waterfield, D. Dolphin, J.G. Levy, Dougerthy (Eds.), Photodynamic Therapy, Marcel Dekker Inc, NewCharacterization of benzoporphyrin derivative, a new photosensit- York, 1992, pp. 201–216.izer, SPIE Proc. 997 (1988) 132–138.

[9] A.M. Richter, S. Yip, E. Waterfield, P.M. Logan, C.E. Slonecker,