Original Contributions Am J Otolaryngol 12-5.

1991

Cutaneous Photoprotection Using a Hydroxyl Radical Scavenger in Photodynamic Therapy NORMAND. HOGIKYAN,MD, RICHARDE. HAYDEN,MD, AND PATRICKW. MCLEAR, MD Photodynamic therapy (PDT) is emerging as an effective therapy for a variety of malignant diseases, including head and neck cancer. Prolonged cutaneous photosensitivity following therapy, however, remains the most significant side effect. The biochemical mechanism of this sensitivity, and indeed of the tumoricidal effect of PDT, is uncertain, but is believed to involve formation of singlet oxygen and possibly other oxygen-derived free radicals. This laboratory recently reported that a singlet oxygen scavenger, diphenylisobenzofuran (DPIBF), afforded cutaneous photoprotection to 67% of animals treated with PDT. Those results, the first from an in vivo study, supported the idea that singlet oxygen plays a significant role in PDT and its associated toxicity. They also, however, suggested that it is not the sole intermediate. The current study looks at the photoprotective effects of the hydroxyl radical scavenger dimethyl thiourea, alone and in conjunction with DPIBF. Our results strongly support a role for the hydroxyl radical in producing the cutaneous phototoxicity associated with PDT. AM J OTOLARYNGOL12:1-5. Copyright 0 1991 by W.B. Saunders Company

Photodynamic therapy (PDT) is a treatment modality that holds great promise in the management of a variety of malignant diseases. It involves the systemic administration of a photoactive drug that is preferentially retained by tumor. This is followed by exposure to light of the appropriate wavelength, which activates the drug in tissues with high concentrations of retained drug, that is, the target tumor. The most thoroughly investigated photosensitizer is dihematoporhyrin ether (DHE), the more purified form of hematoporphyrin derivative. Human trials of DHE have involved a broad spectrum of malignancies, including head and neck cancer,’ with phase III clinical trials under way in cancers of the bladder, esophagus, and lung. Overall results have been encouraging’; however, cutaneous photosensitivity following treatment remains the most significant side effect. This is due to retention of the photosensitizer

Received September 2, 1990, from the Department of Otolaryngology-Head and Neck Surgery, Washington University School of Medicine, St Louis, MO. Accepted for publication October 29, 1990. Presented at the Research Forum of the AAO-HNS annual meeting, September 1990. Supported in part by NIH grant no. T32-NS07278-05. Address correspondence and reprint requests to Norman D. Hogikyan, MD, Department of dtolaryngology, Washington Universitv Medical School, 517 S Euclid, Box 8115, St Louis, MO 6311b. Copyright 0 1991 by W.B. Saunders Company 0196-0709/91/1201-0008$5.00/O

in skin where resultant photoactivation of the drug, even by sunlight, results in skin damage. Photosensitivity lasts approximately 4 to 8 weeks following therapy. The mechanism of PDT, and presumably of the related cutaneous photosensitivity, has generally been accepted to involve the formation of singlet oxygen and possibly other oxygen-derived free radicals. These reactants are formed via two separate photooxidation pathways, labeled type I and type II3 (Fig 1). The role of singlet oxygen has been supported by a variety of in vitro studies.*r5 More recently, this laboratory provided further support by demonstrating, in a rat model, that cutaneous photosensitivity was prevented in 67% of experimental animals by administering the singlet oxygen scavenger 1,3 diphenylisobenzofuran (DPIBF) prior to light exposure.” These results, however, also imply that some other mechanisms played a role in the other 33% and that additional targeted scavenging might achieve 100% protection. Whether oxygen-free radicals formed via type I reactions play a significant role in PDT remains uncertain; however, in vitro evidence from studies using DHE and hematoporphyrin derivative supports the formation of the potent hydroxyl radical during PDT.5,7 These findings, together with the data concerning DPIBF, indicated the need for in vivo investigation of the role of the hydroxyl radical in PDT. Using a cutaneous phototoxicity

CUTANEOUS PROTECTION IN PHOTODYNAMIC THERAPY sensitizer

1 hv

TYPE radicals 4

I

I substrate

TYPE

7

sensitizer’

J

II 07

*‘0

t2

02

oxygenated products

Figure 1. Types and II photooxidation

oxygenated products

reactions. (Repro-

duced with permission.6)

model, this study was designed to look not only at the role of the hydroxyl radical in PDT, but also at the use of hydroxyl radical scavengers, alone and in conjunction with DPIBF, in the prevention of cutaneous phototoxicity. MATERIALS AND METHODS Forty-eight Sprague Dawley rats weighing 266 to 466 g were given a 10 mg/kg intraperitoneal (IP) injection of DHE (Photofrin II, Quadralogic Technologies). They were then randomly divided into four groups (A, B, C, and D) and returned to their cages in a dimly lit room. All animals used in this study were treated humanely in compliance with the guidelines formulated by the National Society for Medical Research and the National Academy of Sciences. Approximately 24 hours later, all animals were light treated. Rats in group A (control group) received a 1.5 mL/166 g body weight IP injection of a l:3 emulsion of Cremophor El (Sigma Chemical, St Louis, MO) and normal saline 4 to 6 hours prior to light exposure, and a 1-mL intravenous (IV) injection of normal saline 20 minutes prior to light exposure. Group B animals were given a 15 mg/kg body weight IP injection of DPIBF (Sigma Chemical) 4 to 6 hours prior to light treatment. This injectate was prepared by slowly dissolving 6 mg

Figure 2. Grading scale for gross skin changes: A, grade 0; B, grade 1 [treated area in arrows); C, grade 2; and D, grade 3.

DPIBF in 2 mL Cremophor El overnight, and emulsifying with 6 mL normal saline on the day of treatment. Animals in group C received a 506 mg/kg IV injection of the selective hydroxyl radical scavenger dimethyl thiourea (DMTU)’ in 1 mL normal saline 20 minutes prior to light therapy. Group D animals received both DPIBF and DMTU as given in groups B and C. All animals had abdominal skin shaved and treated with depilatory cream, and were anesthetized with 56 mg/kg IP injection of sodium pentobarbital for light treatment. A fiberoptic light source [Fiber Lite, Dolan-Jenner Industries) equipped with a 156-W, 21-V Tungsten Halogen Bulb (Sylvania), heat absorbing glass, and a 618-nm sharp-cut filter (#CS 2-73; Kopp Glass) was used in this study, and emitted a broad band of light greater than 618 nm in wavelength. To avoid injury to abdominal viscera, animals were positioned in the left lateral position and the loosely adherent abdominal skin was pulled away from the body and fixed in position using sterile 26-gauge needles. The fiberoptic bundle was then positioned over this skin and light was directed onto an area 2 cm in diameter. Power output measured using a Coherent power meter (#210; Coherent) was 390 mW, giving a power density of 124 mW/cm’. Light treatment was given for 45 minutes for a total dose of 335 J/cm*. Treated skin was cooled throughout therapy using humidified air. Skin temperature was monitored with a cutaneous probe and tele-thermometer (#42SC; Yellow Springs Instruments) and kept at or below 37°C. Rats were observed daily for 72 hours following phototreatment, and the greatest degree of tissue change was graded grossly by an unbiased observer on the following scale: 0, normal skin; 1, edema or erythema; 2, contraction and blanching; and 3, ulceration. This scale, which was determined by observing progressive skin changes in a pilot study using the same treatment conditions, is demonstrated in Fig 2. Skin changes were photographically documented, and animals were euthanized at 72 hours.

HOGIKYAN,

HAYDEN,

AND McLEAR

RESULTS Results are summarized in Fig 3. By 72 hours, all group A (control) animals exhibited moderate to severe (grade 2 or 3) skin changes. Animals in group B (DPIBF) showed similar changes in 92% of cases, while no animals in group C (DMTU) showed more than a minimal (grade 1) change. Rats in group D (DPIBF + DMTU) demonstrated minimal change in 66.7% of cases. Figures 4 through 7 illustrate representative gross findings in each treatment group. Figure 8 shows a group C animal beside a control animal. Data were analyzed using Fisher’s exact test for equal proportions. Groups C and D showed significant photoprotection compared with group A, with two-tailed probability values of .OOOl and .0013, respectively. Proportions in group B did not differ significantly from group A, nor was there a significant difference between groups C and D. Recognizing that the categorization of skin changes was difficult in a few cases, further statistical analyses were performed after collapsing grades 0 or 1 and 2 or 3 into single categories. These analyses yielded results identical to those derived using all four categories. DISCUSSION Understanding the physiologic basis of a therapeutic modality is fundamental to any subsequent modification of that therapy. This is true for attempts to augment its positive effects as well as to minimize side effects. The physiologic mechanism of photodynamic therapy is not entirely understood. The results of this study make significant progress toward that end, while raising new questions for future investigation, Dimethyl thiourea, a selective scavenger of hydroxyl radicals in vitro,’ clearly demonstrated a photoprotective effect on the skin of treated animals. This is the first study showing such an effect N

u

Figure

4.

Group

A (control)

animal with grade 3 change.

in vivo. These results strongly implicate hydroxyl radicals in the cutaneous phototoxicity associated with PDT. Furthermore, if it is assumed that the tumoricidal and phototoxic effects of PDT are caused by the same mechanism this further implicates hydroxyl radical as a factor in PDT. We recognize, however, that protection with radical scavengers constitutes only indirect evidence of the radical’s presence; care must be taken in extrapolating radical scavenging specificity in vitro to the in vivo situation. Results with DPIBF alone are somewhat more difficult to explain. Only one in 12 animals treated with this singlet oxygen scavenger alone

12

M

E

IO,

R

0 F A N

8

1

6

!

4

0

1

GRADE I

Figure 3. animals.

A ,CONTROL, c (DMTU) Summary

2

3

OF SKIN CHANGE 7.----B (DPIBF) D (DPIBF.DMTU,

of skin change grading for each group of Figure

5.

Group

B (DPIBF) animal

with

grade

3 change.

4

CUTANEOUS PROTECTION IN PHOTODYNAMIC THERAPY

Figure 8. Groups A and C animals illustrating toprotection with DMTU. Figure 6.

marked pho-

Group C (DMTU) animal with grade 0 change.

showed photoprotection, which is quite different from the 67% protection rate seen in the earlier study from this laboratory.6 Our attempt to explain this disparity centers on differences between the two experimental designs. Preparation of DPIBF for injection differed slightly; however, consultation with the manufac-

Figure 7. Group D (DMTU + DPIBF] animal that had demonstrated edema (grade 1 change] transiently, but now with no visible change in treated area at 72 hours.

turer’s chemists brought assurances that neither method would have altered the drug. The lot numbers were identical in both experiments as well. This would seem to rule out a problem with the drug itself. The dose of DHE was also slightly different in the two studies, with 7.5 mg/kg used in the first study and 10 mg/kg in the second. Grossweiner et al did find that higher concentrations of hematoporphyrin favored type I over type II reactions in the photosensitized lysis of phosphatidylcholine liposomes.g It is extremely unlikely, however, that their results are relevant to these experiments since their doses varied by much greater amounts and experimentation was in a controlled in vitro setting. The final difference between the two studies was in the light source used. The light source in the earlier study was a tunable dye laser driven by an argon laser that delivered monochromatic light of 630-nm wavelength at a power density of 100 mW/cm’. Treatment time in that study was 20 minutes. Obviously, this is a significantly higher dose of light at the usual wavelength for DHE activation (630 nm), as demonstrated by the frank necrosis seen in the control animals in that study. Additionally, light generated by that source did not include wavelengths longer than 630 nm that can also activate DHE, whereas the spectrum of the current source did include such wavelengths. These differences cannot be discounted entirely as possible explanations for the DPIBF results. There are recognized factors that favor type I over type II reactions, such as lower oxygen concentrations or higher reactive substrate concentrations, but light intensity has not been recognized as such a factor.3 It is possi-

HOGIKYAN, HAYDEN, AND McLEAR

ble, however, that the lower dosage and intensity of activating light used in this study favored the radical pathway versus the one involving singlet oxygen. It is also possible that the activation of DHE via its other absorption peaks in the infrared region favored type I over type II reactions in the current study. Such explanations are only speculative at this point, and remain to be investigated. The fact that DMTU and DPIBF together did not confer complete photoprotection is also worth noting. This suggests that other intermediates, possibly superoxide radical or hydrogen peroxide, are involved in this process as well. This study has demonstrated that a selective hydroxyl radical scavenger acted as a cutaneous photoprotector during PDT. The singlet oxygen scavenger DPIBF did not, under these experimental conditions, confer significant photoprotection. Additional study is warranted to explore the roles of both type I and type II photooxidation products in PDT and the prevention of cutaneous photosensitivity by targeted agents. Acknowledgment.

Statistical analyses performed by

staff of the Washington University School of Medicine Division of Biostatistics. Special thanks are due to Pro-

5

fessor K. Luszczynski for his assistance the light source for this study.

in developing

References 1. Gluckman JL, Weissler MC: Role of photodynamic therapy in the management of early cancers of the upper aerodigestive tract. Lasers Med Sci 1986; 1:217-220 2. Dougherty TJ: Photodynamic therapy-New approaches. Semin Surg Oncol 1989; 5:6-16 3. Foote CS: Mechanisms of photooxygenation. Prog Clin Biol Res 1984; 107:3-18 4. Weishaupt KR, Gomer CJ, Dougherty TJ: Identification of singlet oxygen as the cytotoxic agent in photoinactivation of a murine tumor. Cancer Res 1976; 36:2326-2329 5. Henderson BA, Miller AC: Effects of scavengers of reactive oxygen and radical species on cell survival following photodynamic treatment in vitro: Comparison to ionizing radiation. Radiat Res 1986; 108:196-205 6. McLear PW, Hayden RE: Prevention of cutaneous phototoxicity in photodynamic therapy. Am J Otolaryngol 1989; 10:92-98 7. Hariharan PV, Courtney J, Eleczko S: Production of hydroxyl radicals in cell systems exposed to haematoporphyrin and red light. Int J Radiat Biol 1980; 37:691-694 8. Rao PS, Luber JM, Milihowicz J, et al: Specificity of oxygen radical scavengers and assessment of free radical scavenger efficiency using luminol enhanced chemiluminescence. Biochem Biophys Res Commun 1988; 150:39-44 9. Grossweiner LI, Pate1 AS, Grossweiner JB: Type I and type II mechanisms in the photosensitized lysis of &osphatidylcholine liposomes by hematoporphyrin. Photochem Photobiol 1982; 36:159-167

Cutaneous photoprotection using a hydroxyl radical scavenger in photodynamic therapy.

Photodynamic therapy (PDT) is emerging as an effective therapy for a variety of malignant diseases, including head and neck cancer. Prolonged cutaneou...
3MB Sizes 0 Downloads 0 Views