Photodynamic Therapy Induced Immunosuppression in Humans Is Prevented by Reducing the Rate of Light Delivery

Transcription

1 ORIGINAL ARTICLE Photodynamic Therapy Induced Immunosuppression in Humans Is Prevented by Reducing the Rate of Light Delivery Georgia A. Frost 1, Gary M. Halliday 1 and Diona L. Damian 1 Photodynamic therapy (PDT) of non-melanoma skin cancers currently carries failure rates of 1 4%. The optimal irradiation protocol is as yet unclear. Previous studies showed profound immunosuppression after PDT, which may compromise immune-mediated clearance of these antigenic tumors. Slower irradiation prevents immunosuppression in mice, and may be at least as effective as high-fluence-rate PDT in preliminary clinical trials. The photosensitizers 5-aminolaevulinic acid and/or methyl aminolaevulinate were applied to discrete areas on the backs of healthy Mantoux-positive volunteers, followed by narrowband red light irradiation (632 nm) at varied doses and fluence rates. Delayed type hypersensitivity (Mantoux) reactions were elicited at test sites and control sites to determine immunosuppression. Human ex vivo skin received low- and highfluence-rate PDT and was stained for oxidative DNA photolesions. PDT caused significant, dose-responsive immunosuppression at high (75 mw cm 2 ) but not low (15 or 45 mw cm 2 ) fluence rates. DNA photolesions, which may be a trigger for immunosuppression, were observed after high-fluence-rate PDT but not when light was delivered more slowly. This study demonstrates that the current clinical PDT protocol (75 mw cm 2 )is highly immunosuppressive. Simply reducing the rate of irradiation, while maintaining the same light dose, prevented immunosuppression and genetic damage and may have the potential to improve skin cancer outcomes. Journal of Investigative Dermatology (211) 131, ; doi:1.138/jid ; published online 2 January 211 INTRODUCTION Photodynamic therapy (PDT) for skin cancer comprises topical application of a photosensitizer, followed by irradiation with visible light in the presence of oxygen. PDT with systemic photosensitizers has been used to treat a range of internal malignancies, including lung, brain, esophagus, bladder, and head and neck cancers (Brown et al., 24). Topical PDT, using the photosensitizing heme precursors 5-aminolaevulinic acid (ALA) or methylaminolaevulinate (MAL), is increasingly used to treat superficial basal cell carcinoma (BCC), Bowen s disease (squamous cell carcinoma in situ), and premalignant actinic keratoses (Morton et al., 22). Compared with surgery, ALA/MAL-PDT offers excellent functional and cosmetic outcomes and is a more 1 Department of Dermatology, Bosch Institute, The University of Sydney and Sydney Cancer Centre at Royal Prince Alfred Hospital, Camperdown, New South Wales, Australia Correspondence: Diona L. Damian, Department of Dermatology, Gloucester House Level 3, Royal Prince Alfred Hospital, Missenden Road, Camperdown, New South Wales 25, Australia. diona.damian@sswahs.nsw.gov.au Abbreviations: 8oxoG, 8-hydroxy-2 -deoxyguanosine; ALA, 5-aminolaevulinic acid; BCC, basal cell carcinoma; MAL, methylaminolaevulinate; PDT, photodynamic therapy; PPD, purified protein derivative; ROS, reactive oxygen species Received 3 August 21; revised 22 November 21; accepted 3 November 21; published online 2 January 211 practical option for the treatment of superficial but large diameter lesions, and for patients with large numbers of skin cancers. Although ALA/MAL-PDT clearance rates of 49% have been found for smaller superficial BCCs (Haller et al., 2), reported long-term clearance rates for BCC typically range from B5 84% depending on lesion size and number of PDT sessions (Szeimies, 27). Clearance rates of B76% at 5 years have been reported after two PDT sessions for nodular BCCs (Rhodes et al., 27) and B8% at 12 months for Bowen s disease (Morton et al., 26), whereas actinic keratoses showed initial clearance of B75% and recurrence of 2% at 12 months (Tschen et al., 26). Hence, ALA/MAL- PDT using currently recommended protocols carries relatively high failure rates. Immune suppressed transplant recipients show even higher PDT failure rates, reflecting the key role of skin immunity in the resolution of these antigenic tumors (Dragieva et al., 24). For ALA/MAL-PDT of skin cancers, a maximum fluence rate of 2 mw cm 2 is recommended (Babilas et al., 26), however, there is no defined lower limit. At high-fluence rates, more rapid oxygen consumption likely outstrips the oxygen supply from surrounding blood vessels, whereas at low-fluence rates, tissue oxygenation remains at levels adequate for completion of the photodynamic process (Foster et al., 1991). Murine studies have shown that lower-fluence rates cause more efficient photobleaching and greater tissue 962 Journal of Investigative Dermatology (211), Volume 131 & 211 The Society for Investigative Dermatology

2 damage by topical ALA-PDT (Robinson et al., 1998), and yield better tumor kill after systemic PDT (Sitnik and Henderson, 1998; Iinuma et al., 1999). Preliminary human studies also suggest that ALA/MAL-PDT with lower-fluence rates may be as or more effective than current protocols (Langmack et al., 21; Ericson et al., 24; Cottrell et al., 28). PDT is highly immunosuppressive in mice (Elmets and Bowen, 1986), with suggestion that this may be prevented or even reversed by lowering fluence rates (Sitnik and Henderson, 1998; Henderson et al., 24). The immune effects of varying PDT fluence rates in humans are as yet unknown. PDT is generally considered to have little carcinogenic risk (Morton et al., 28). There are, however, reports of malignancies developing in ALA/MAL-PDT-treated fields (Wolf et al., 1997; Varma et al., 2). The oxidative photoproduct 8-hydroxy-2 -deoxyguanosine (8oxoG) is a marker of DNA damage that is observed after UVA and UVB irradiation (Javeri et al., 28). PDT induces 8oxoG in mice (Takahashi et al., 25) and DNA strand breaks in human fibroblasts (Haylett et al., 23), however, there are no reported studies in human skin. The Mantoux reaction, a T-cell-mediated delayed type hypersensitivity response to intradermally injected tuberculin purified protein derivative (PPD), offers a practical and ethical model of in vivo skin immunity (Damian and Halliday, 22). Using this model, we recently reported that topical ALA/MAL-PDT, delivered at 78 mw cm 2, is highly immunosuppressive in humans (Matthews and Damian, 21). We now report that ALA/MAL-PDT at high but not low-fluence rates is immunosuppressive and causes generation of 8oxoG lesions in humans. RESULTS High-fluence-rate MAL-PDT and ALA-PDT are immune suppressive Fifteen volunteers of Fitzpatrick s skin types II and III (Fitzpatrick, 1998; 8 men, 7 women, mean age 4 years, range 23 65) had mirrored templates on each side of their lower backs, which were treated with MAL, ALA, vehicle, or no lotion. One side of the back was irradiated with high-fluence-rate red light (fluence rate 75 mw cm 2, total dose 37 J cm 2 ), whereas the other side remained light protected. MAL-PDT significantly suppressed Mantoux induration (mean±sem Ddiameter 3.1±.7 mm, P ¼.3, n ¼ 15; Figure 1a) corresponding to a 33% mean reduction in diameter. MAL-PDT also suppressed erythema (Derythema index (EI) 34.7±9.8 erythema units, P ¼.3, n ¼ 15; Figure 1b) equating to a 37% mean reduction in EI. Likewise, ALA-PDT significantly suppressed Mantoux diameter (Ddiameter 3.6±.5 mm, Po.1, Figure 1a) corresponding to a 4% mean reduction in diameter. ALA-PDT EI was suppressed (DEI 53.6±9.6 erythema units, Po.1, Figure 1b) by 67% compared with control. Neither photosensitizer caused immune suppression in the absence of irradiation. Mantoux reactions at unirradiated, MAL-treated sites did not differ from control Mantoux reactions (Ddiameter.4±.1 mm, P ¼.8, Figure 1a; DEI a b Immunosuppression (ΔEI) Immunosuppression (Δmm) P=.3 P=.8 P=.7 P=.1 P=.2 2 P =1. P =1. P =1. P =1. P =1. 2 P=.3 P<.1 VEH MAL ALA Light only VEH MAL ALA Unirradiated Irradiated P=.3 P<.1 Light VEH ALA MAL only VEH MAL ALA Unirradiated Irradiated Figure 1. Photodynamic therapy (PDT; fluence rate 75 mw cm 2 and total dose 37 J cm 2 ) with either methyl aminolaevulinate (MAL) or 5-aminolaevulinic acid (ALA) is highly immune suppressive in humans. Red light alone or with vehicle, or photosensitizer alone (no light) did not cause immunosuppression. Mantoux diameter and erythema at each treatment site was subtracted from that at the unirradiated control site to determine immunosuppression (a, Ddiameter; b, DEI; Student s paired t test, n ¼ 15). EI, erythemia index; VEH, vehicle. 5.4±5.3 erythema units, P ¼ 1., Figure 1b). Similarly, ALA alone was not immunosuppressive (Ddiameter.4±.1 mm, P ¼.7, Figure 1a; DEI 8.±5.4 erythema units, P ¼ 1., Figure 1b). Sites receiving red light only did not show suppression of Mantoux diameter or erythema compared with unirradiated control sites (Figure 1). This was further investigated in three men and seven women (skin types II and III, mean age 4, range 26 74). At separate sites on the back, these volunteers received red light only (75 mw cm 2 ) with total doses of 37, 111, and 222 J cm 2 (up to six times the standard therapeutic light dose) and MAL-PDT as a positive control (37 J cm 2, 75mWcm 2 ). Two additional sites served as unirradiated controls. Red light did not suppress Mantoux induration nor erythema at any dose (Ddiameter.4±.2 mm, P ¼.4;.5±.2 mm, P ¼.3;.6±.5 mm, P ¼ 1.; and DEI 1.7±6.8, P ¼ 1.; 8.8±7.6, P ¼ 1.;.7±7.8 P ¼ 1. erythema units for 37, 111, and 222 J cm 2, respectively). Again, MAL-PDT caused immunosuppression of B2% in diameter (Ddiameter 963

3 a Immunosuppression (Δmm) b Immunosuppression (ΔEI) P =1. 1 P=1. P=1. P=1. P=1. P =.66 P=.24 P=.6 P= mw cm 2 75 mw cm 2 37 J cm 2 P =.15 P=.4 P = mw cm 2 75 mw cm 2 37 J cm 2 Figure 2. High-fluence-rate PDT (75 mw cm 2 ) with methyl aminolaevulinate (MAL) caused significant, dose responsive suppression of Mantoux reactions. In contrast, low-fluence-rate MAL-PDT (15 mw cm 2 ) was not immune suppressive at any dose (a, Ddiameter; b, DEI; Student s paired t-test, n ¼ 15). EI, erythemia index. a Immunosuppression (Δmm) b Immunosuppression (ΔEI) P=.6 P =.4 P = mw cm 2 P=1. P=.1 P = P=1. P=.7 P= J cm 2 75 mw cm 2 P=.11 P=.5 P = J cm 2 45 mw cm 2 75 mw cm 2 Figure 3. High-fluence-rate PDT (75 mw cm 2 ) with methyl aminolaevulinate (MAL) significantly suppressed both Mantoux diameter and erythema with a total light dose of 37 J cm 2 but not 1 or 2 J cm 2. Low-fluence rate MAL-PDT (45 mw cm 2 ) was not immunosuppressive at any dose (a, Ddiameter; b, DEI; Student s paired t-test, n ¼ 15). EI, erythemia index. 1.9±.5, P ¼.2) and 17% in erythema (DEI 17.7±5.6, P ¼.48), in these 1 volunteers. MAL-PDT is immune suppressive at high but not low fluence rates Fifteen volunteers (skin types II and III, 7 men, 8 women, mean age 39, range 26 74) had MAL-PDT administered to each side of the lower back. On one side total doses of 1, 2, and 37 J cm 2 were delivered at a high-fluence rate of 75 mw cm 2. These three doses were also delivered to the contralateral back at a low-fluence rate of 15 mw cm 2. Mantoux reactions at each site were compared with unirradiated controls (no light, no MAL). High-fluence-rate MAL-PDT significantly suppressed Mantoux reactions at total light doses of 37 J cm 2 (Ddiameter 2.9±.6 mm, P ¼.2; DEI 34.5±7.5 erythema units, P ¼.2) and 2 J cm 2 (Ddiameter 1.9±.5 mm, P ¼.6; DEI 35.±6.4 erythema units, P ¼ o.1) but not 1 J cm 2 (Ddiameter 1.4±.5 mm, P ¼.24; DEI 23.3±9.7 erythema units, P ¼.15; Figure 2). MAL-PDT delivered at a low-fluence rate (15 mw cm 2 ) did not, however, suppress Mantoux reactions at any dose (37 J cm 2 Ddiameter.2±.4 mm, P ¼ 1.; DEI 13.7±8.5 erythema units, P ¼.7; 2 J cm 2 Ddiameter.2±.4 mm, P ¼ 1., DEI 1.5±7.1 erythema units, P ¼ 1.; 1 J cm 2 Ddiameter.1±.4 mm, P ¼ 1., DEI 3.4±6.6 erythema units, P ¼ 1.; Figure 2). Delivery of 37 J cm 2 at a fluence rate of 75 mw cm 2 takes B8 minutes, hence, delivery of this dose at 15 mw cm 2 takes B4 minutes, which is less feasible in clinical practice. Fifteen additional volunteers (skin types II and III, seven men, eight women, mean age 4, range 24 65) were thus recruited to compare the immune effects of an intermediate fluence rate (45 mw cm 2 ; irradiance time 13 minutes) to irradiance at 75 mw cm 2. High-fluence (75 mw cm 2 ) MAL-PDT was immunosuppressive in this group of volunteers at a dose of 37 J cm 2 (Ddiameter 1.8±.5 mm, P ¼.2; DEI 24.7±7. erythema units, P ¼.2), but not at lower doses (2 J cm 2 Ddiameter 1.1±.4 mm, P ¼.7, DEI 3.5±1.3 erythema units, P ¼.5; 1 J cm 2 ; Ddiameter.4±.3 mm, P ¼.1, DEI 16.9±6.5 erythema units, P ¼.11, Figure 3). MAL-PDT delivered with a fluence rate of 45 mw cm 2 did not suppress Mantoux reactions at any total light dose (37 J cm 2 Ddiameter.6±.5 mm, P ¼ 1., DEI 13.±1.1 erythema units, P ¼ 1.; 2 J cm 2 Ddiameter 964 Journal of Investigative Dermatology (211), Volume 131

4 8-oxo Staining (% of control) PDT: 75 mw cm 2 PDT: 15 mw cm 2 MAL only P= Time (min) 9 12 Figure 4. Human skin samples (duplicates from three volunteers) were treated with methylaminolaevulinate (MAL) only, or low-fluence (15 mw cm 2 ) or high-fluence (75 mw cm 2 ) -MAL-photodynamic therapy. Staining for oxidative DNA photolesions (8-hydroxy-2 -deoxyguanosine) was performed at various time points up to 2 hours to assess DNA repair as well as damage. The percentage of epidermal area positively staining for 8-oxo at each treatment time point was compared with untreated control skin (unpaired two-tailed Student s t-test, Bonferroni correction). Figure 5. Human skin (bar ¼ 1 lm) was stained for 8-hydroxy-2 - deoxyguanosine. Control (untreated) skin showed low-level baseline staining consistent with oxidative damage resulting from normal cellular metabolism (a). Methylaminolaevulinate (MAL)-photodynamic therapy (PDT)-treated skin using a high-fluence rate of 75 mw cm 2, total light dose 37 J cm 2, t ¼ 6 minutes following irradiation showed significantly higher levels of oxidative DNA damage (b) compared with MAL-PDT-treated skin using a low-fluence rate (c, 15mWcm 2 )..6±.3 mm, P ¼.4, DEI 16.3±6.3 erythema units, P ¼ 1.; 1 J cm 2 Ddiameter.6±.3 mm, P ¼.6, DEI 8.±7.5 erythema units, P ¼ 1., Figure 3). PDT does not cause systemic immunosuppression Systemic immunosuppression was determined by comparing initial Mantoux reactions (measured 3 days before PDT) to unirradiated control Mantoux reactions following PDT. There was no significant difference in Mantoux-induced induration nor erythema before and after PDT in any group of volunteers. MAL-PDT induces 8oxoG at high but not low fluence rates Three women (skin types II, III, and IV, mean age 37 years, range 29 45) donated skin from elective surgical procedures, which was then treated with MAL only, high (75 mw cm 2 ) or low-fluence-rate (15 mw cm 2 ) MAL-PDT and subsequently stained for epidermal 8oxoG (Figures 4 and 5). MAL only and low-fluence-rate PDT were not significantly different from control skin in area of epidermis staining for 8oxoG at any time point studied. However, high-fluencerate PDT caused a significant increase in epidermal 8oxoG staining compared with control skin at t ¼ 6 minutes (% of control 215±45, P ¼.4, n ¼ 6; Figure 4). We also stained a separate MAL-PDT-treated skin sample for cyclobutane pyrimidine dimers, but found no increase in these direct DNA photolesions compared with unirradiated skin. Solar-simulated UV-irradiated skin, showing large numbers of cyclobutane pyrimidine dimers, served as a positive control (data not shown). DISCUSSION Consistent with our previous findings (Matthews and Damian, 21), ALA-PDT was profoundly immunosuppressive in humans, measured by both Mantoux erythema and induration, using a high-fluence-rate protocol. ALA-PDT caused a higher level of immunosuppression (B5%) than MAL-PDT (B35%). ALA and MAL produce the same intracellular photosensitizer (PpIX), penetrate human skin to similar depths and produce a similar spatial distribution of PpIX in murine skin (de Bruijn et al., 28). However, ALA produces relatively greater amounts of PpIX (Juzeniene et al., 26) and higher levels of fluorescence with a slower time course than MAL in human skin (Lesar et al., 29). This, together with greater local edema reported after ALA compared with MAL- PDT in mice (de Bruijn et al., 28) and the possibility of different microscopic PpIX distributions following ALA and MAL application to human skin, could explain the greater immunosuppression observed after ALA-PDT in our study. We examined normal skin, but intralesional variations in PpIX distribution within tumors could also influence the immune effect of PDT (de Haas et al., 28). ALA/MAL-PDT using our protocol impaired local immunity to existing antigens (recall immunity), and could thus impair local antitumor immunity and tumor eradication. The effects of PDT on the development of immunity to new tumor antigens (primary immunity) are unknown. In our studies, immunosuppression was evident 6 days after PDT; however, it is unknown how long this effect persists in humans. In clinical practice, skin cancers receive ALA/MAL-PDT on two occasions 1 week apart (Morton et al., 28), suggesting 965

5 immune suppression for at least 2 weeks, but possibly longer given that PDT immunosuppression persists for at least 28 days in mice (Gollnick et al., 21). We found significant, dose responsive immune suppression with MAL-PDT delivered at 75 mw cm 2, which was completely prevented by simply lowering the fluence rate to deliver an identical total dose five times more slowly. An intermediate fluence rate of 45 mw cm 2, a more practical option requiring irradiation of only an additional 5 minutes beyond current clinical recommendations, also did not cause immunosuppression. The putative central mechanism of PDT is generation of reactive oxygen species (ROS) by irradiation of a photosensitizer. PDT-induced ROS can then react with a range of cellular targets, including cell membrane lipids and mitochondria as well as DNA. Irradiation with lower light doses can alter cell signalling and cytokine and receptor expression in the absence of cell death, suggesting that there are different dose responses for these ROS effects (Calzavara-Pinton et al., 27). In human fibroblasts treated with ALA-PDT, 7 95% of cell kill occurred without detectable DNA damage, suggesting that mitochondrial damage, which we did not directly examine, is also a key mechanism of cell death (Haylett et al., 23). In normal human skin, we found less oxidative DNA damage after low-fluence-rate MAL-PDT, although the effect of reduced fluence rate on DNA damage in tumor cells is unknown. ALA-based photosensitizers are more readily taken up by tumors than by normal skin, with a 1-fold increase in PpIX production in dysplastic or malignant lesions (Fritsch et al., 2). It is thus possible that low-fluence-rate PDT might maintain efficacy of tumor kill by more selectively damaging tumor DNA, causing other ROS effects such as cytokine triggering and mitochondrial damage (Hilf, 27) and inducing tumor cell apoptosis (Singh et al., 21). We examined only total DNA and it remains possible that our low-fluence rate may not have spared oxidative damage to mitochondria. Despite the possibility that reduced oxidative DNA damage, as we observed here, might impair efficacy, studies using low-fluence-rate protocols have so far demonstrated comparable tumor clearance rates when compared with high fluence-rate regimens in actinic keratoses treated with ALA and nm or nm light sources (Ericson et al., 24) and in superficial BCCs treated with ALA and narrowband 633 nm light (Cottrell et al., 28). This implies that high-fluence rates might not be required for efficacy. In BCCs, photodynamic efficiency (loss of light absorbance as photosenstizer is used in the photodynamic process) decreased with increased fluence rate; 2 mw cm 2 of narrowband red light was most efficient (Cottrell et al., 28). A 635 nm red light source was as effective in the treatment of superficial BCCs at 12.6 J cm 2, delivered at 7mWcm 2 over 3 minutes, as standard treatment (37 J cm 2 over 8 minutes; i.e., 75 mw cm 2 ; Langmack et al., 21). There is a suggestion that low-irradiance ambulatory ALA- PDT (12 mw cm 2 ) may also be effective (Moseley et al., 26). Similarly, high clearance rates for actinic keratoses were reported with lower fluence rates (3 mw cm 2 ; Ericson et al., 24) or even irradiation using low-intensity Copenhagen sunlight (Wiegell et al., 28). Higher numbers of neutrophils, greater inflammation, and greater expansion of tumor-specific T cells are seen after low-irradiance, low-dose systemic PDT compared with a high-irradiance, high-dose regimen in mice (Kousis et al., 27). Fractionation of irradiation has also been suggested to improve the efficacy of topical ALA-PDT for superficial BCC (de Haas et al., 26), with suggestion that neutrophils induced by systemic PDT may be key mediators of anti-tumor immunity (Kousis et al., 27) and anti-tumor efficacy (de Bruijn et al., 26) in this setting. The role and fluence rate-dependence of neutrophil infiltration in topical ALA/MAL-PDT in humans is unknown. Ultraviolet radiation induces 8oxoG photolesions that are products of oxidative damage, and are highly mutagenic if left unrepaired (Kozmin et al., 25). We found that high but not low-fluence-rate MAL-PDT induced these photolesions. UV-induced cyclobutane pyrimidine dimers are a key trigger of UV-induced immunosuppression (Kuchel et al., 25) and it has been suggested that 8oxoG may have similar effects. Inhibition of ROS-mediated damage, such as 8oxoG, prevents immunosuppression (Yao et al., 29). More recently it has been shown that UV-induced platelet activating factor and cis-urocanic acid both induce 8oxoG formation and cause immunosuppression, whereas inhibition of these mediators reduced 8oxoG and immunosuppression (Sreevidyaet al., 21). It is not known yet whether enhanced repair of 8oxoG helps to prevent immunosuppression, but topical application of the 8oxoG repair enzyme 8 oxoguanosine DNA glycosylase reduced tumor size and progression in photoinduced tumor models (Wulff et al., 28). Hence, ROS-mediated damage, such as 8oxoG, is a key molecular mechanism causing immunosuppression (Halliday, 21). In our study, the immunosuppression following high-fluencerate ALA/MAL-PDT may reflect oxidative DNA damage, which at lower fluence rates is no greater than the expected baseline level of oxidative damage that results from normal cellular metabolism (Javeri et al., 28). Additional studies are now indicated to better evaluate the cellular and cytochemical changes involved in PDT-induced immunomodulation, to improve PDT efficacy in skin cancer, but possibly also to use immune suppressive effects in the treatment of immune-mediated disorders such as psoriasis (Morton et al., 28). Topical PDT at high but not low-fluence rates causes both immunosuppression and induction of oxidative DNA photolesions in human skin. As both immunosuppression and DNA damage are recognized tumor promoters, this suggests that simply irradiating more slowly may have the potential to reduce tumor recurrence and boost cure rates. Further studies are now urgently needed to assess the role of PDTimmunosuppression in PDT treatment failures and the mechanisms of cell toxicity following low-fluence-rate PDT. Using randomized controlled trials, we then need to determine the comparative efficacy of low irradiance rate PDT, so that we can establish and recommend to clinicians the optimal PDT protocol. 966 Journal of Investigative Dermatology (211), Volume 131

6 MATERIALS AND METHODS Mantoux testing We recruited healthy Mantoux positive hospital staff and students, who had previously been vaccinated with Bacille Calmette-Guerin. Volunteers were not taking immunosuppressive or photosensitizing medications and were asked to avoid sun exposure for 4 weeks before and throughout the study, as even small UV exposures can suppress skin immunity (Damian and Halliday, 22). Approval for these studies was provided by the Sydney South West Area Health Service and University of Sydney Ethics Committees and all participants provided written informed consent. All experiments were performed in accordance with the precepts established by the Helsinki Declaration. Volunteers were initially Mantoux tested with three graded doses of tuberculin PPD (Tubersol; Sanofi-Pasteur, Toronto, ON, Canada) to determine both Mantoux positivity and the appropriate PPD dose to elicit a Mantoux diameter of B8 mm. PPD was diluted with normal saline to a volume of.1 ml and injected intradermally on the lateral lower back with 31-gauge needles. Mantoux reactions were assessed 72 hours later using the pen method, which measures Mantoux diameter of induration in two perpendicular planes (Bouros et al., 1991). The EI at each Mantoux site and at adjacent skin was measured in triplicate using a portable reflectance meter (Dia-stron, Andover, UK). The EI of adjacent skin was then subtracted from the EI overlying Mantoux reactions to determine Mantoux-induced erythema, and to adjust for any PDT-induced erythema (Damian and Halliday, 22). Photodynamic therapy Templates were created on each side of the lower back using Comfeel Plus Ulcer Dressing (Coloplast A/S, 35 Humlebaek, Denmark) to create discrete 3 2 cm areas separated by 1 cm. Treatment allocation at each square was randomized based on order of study entry, with a mirror image of treatment site allocation on the contralateral back. In all, 4 mg cm 2 of the photosensitizers 2% ALA in white soft paraffin (Biosynth AG, Staad, Switzerland) and MAL 16% cream (Metvix: Galderma, Belrose, New South Wales, Australia) or vehicle (emulsifying ointment) were applied to the various sites, which were then occluded for 3 hours under Tegaderm dressings (3M Healthcare, St Paul, MN) and light protected with aluminum foil. The creams were then removed and test sites exposed to varied fluence rates and total light doses from an Aktilite CL128 LED array (PhotoCure, ASA, Oslo, Norway). The light source emitted narrowband red light with a peak of 632 nm as measured at 1 nm intervals with an Optronics OL754 spectroradiometer (Optronics, Orlando, FL) calibrated against standard lamps. Irradiance was monitored daily with an IL17 broadband radiometer (International Light Technologies, Peabody, MA), calibrated against the spectroradiometer. After irradiation, all treatment sites remained light protected with aluminum foil for a further 24 hours. Measurement of immunosuppression and data analysis At 72 hours after irradiation each of the treatment and control sites were injected with PPD, and the resulting Mantoux reactions quantitated by diameter and erythema measurements 72 hours thereafter. Immunosuppression was calculated as the difference between control site (no photosensistizer, no light) and treatment site Mantoux reactions. Results are shown as mean±sem; significance was determined by Bonferroni corrected paired Student s t-tests with Po.5 considered significant. Measurement of PDT-induced DNA photolesions Three healthy adults undergoing abdominoplasty (n ¼ 2) or breast reduction (n ¼ 1) consented for their excised skin to be used in the study. Skin was collected in normal saline, trimmed of excess fat, washed in chlorhexidine, rinsed with phosphate-buffered saline (Invitrogen, Calsbad, CA), and cut into 3 cm squares, which received MAL only (no light), low-fluence-rate MAL-PDT (15 mw cm 2 ), high-fluence-rate MAL-PDT (75 mw cm 2 ), or no treatment (control). MAL (16%, 4 mg cm 2 ) was applied and occluded with Tegaderm. The skin was then placed into separate petri dishes containing RPMI-164 medium (Sigma-Aldrich, St Louis, MO) supplemented with 1% fetal bovine serum and 1 U ml 1 penicillin,.1 mg ml 1 streptomycin, and 25 ng ml 1 amphotericin B (Sigma-Aldrich). Petri dishes were light protected with aluminum foil and incubated for 3 hours at 37 1C in5%co 2 in air. MAL was removed using phosphate-buffered saline and specimens were rinsed twice in phosphate-buffered saline and cut into 5 mm squares. The two PDT groups were then exposed to 37 J cm 2 of red light (Aktilite) with a fluence rate of 15 or 75 mw cm 2. Duplicate skin pieces were collected immediately or after various times of incubation in the media described above, snap frozen in OCT (Tissue-Tek, Sakura, Zoeterwoude, the Netherlands) and stored at 71C. Frozen sections (6-mm thick) were stained by immunohistochemistry using mouse monoclonal anti-8oxog antibody (Trevigen, Gaithersburg, MD), as we have described previously (Javeri et al., 28). Separate pieces of skin were treated with MAL-PDT (37 J cm 2, 75mWcm 2 ) or solarsimulated UV (4 J cm 2 ), and stained for cyclobutane pyrimidine dimers (Kamiya Biomedical, Seattle, WA). The area of positively stained epidermis was determined by image analysis (ImageJ, Wayne Rasband, National Institutes of Health). The average percentage staining for each treatment group at the five time points was then calculated and compared with the average of the control by Student s unpaired t-tests (Boneferroni correction; Po.5 considered significant). CONFLICTS OF INTEREST The authors state no conflict of interest. ACKNOWLEDGMENTS Funding for these studies was provided by a Cancer Institute New South Wales Innovation Grant and by Epiderm. Dr Frost was supported by a Dermatology Research Foundation postgraduate scholarship. REFERENCES Babilas P, Landthaler M, Szeimies RM (26) Photodynamic therapy in dermatology. Eur J Dermatol 16:34 8 Bouros D, Zeros G, Panaretos C et al. (1991) Palpation vs pen method for the measurement of skin tuberculin reaction (Mantoux test). Chest 99:416 9 Brown SB, Brown EA, Walker I (24) The present and future role of photodynamic therapy in cancer treatment. Lancet Oncol 5: Calzavara-Pinton PG, Venturini M, Sala R (27) Photodynamic therapy:update 26. Part 1: Photochemistry and photobiology. J Eur Acad Dermatol Venereol 21: Cottrell WJ, Paquette AD, Keymel KR et al. (28) Irradiance-dependent photobleaching and pain in delta-aminolevulinic acid-photodynamic therapy of superficial basal cell carcinomas. Clin Cancer Res 14:

7 Damian DL, Halliday GM (22) Measurement of ultraviolet radiationinduced suppression of recall contact and delayed-type hypersensitivity in humans. Methods 28:34 45 de Bruijn HS, Meijers C, van der Ploeg-van den Heuvel A et al. (28) Microscopic localisation of protoporphyrin IX in normal mouse skin after topical application of 5-aminolevulinic acid or methyl 5-aminolevulinate. J Photochem Photobiol B 92:91 7 de Bruijn HS, Sluiter W, van der Ploeg-van den Heuvel A et al. (26) Evidence for a bystander role of neutrophils in the response to systemic 5-aminolevulinic acid-based photodynamic therapy. Photodermatol Photoimmunol Photomed 22: de Haas ER, de Bruijn HS, Sterenborg HJ et al. (28) Microscopic distribution of protoporphyrin (PpIX) fluorescence in superficial basal cell carcinoma during light-fractionated aminolaevulinic acid photodynamic therapy. Acta Derm Venereol 88: de Haas ER, Kruijt B, Sterenborg HJCM et al. (26) Fractionated illumination significantly improves the response of superficial basal cell carcinoma to aminolevulinic acid photodynamic therapy. J Invest Dermatol 126: Dragieva G, Hafner J, Dummer R et al. (24) Topical photodynamic therapy in the treatment of actinic keratoses and Bowen s disease in transplant recipients. Transplantation 77: Elmets CA, Bowen KD (1986) Immunological suppression in mice treated with hematoporphyrin derivative photoradiation. Cancer Res 46: Ericson MB, Sandberg C, Stenquist B et al. (24) Photodynamic therapy of actinic keratosis at verying fluence rates: assessment of photobleaching, pain and primary clinical outcome. Br J Dermatol 151: Fitzpatrick TB (1998) The validity and practicality of sun-reactive skin types I through VI. Arch Dermatol 124: Foster TH, Murant RS, Bryant RG et al. (1991) Oxygen consumption and diffusion effects in photodynamic therapy. Radiat Res 126: Fritsch C, Homey B, Stahl W et al. (2) Preferential relative porphyrin enrichment in solar keratoses upon topical application of deltaaminolevulinic acid methylester. Photochem Photobiol 68: Gollnick SO, Musser DA, Oseroff AR et al. (21) IL-1 does not play a role in cutaneous Photofrin photodynamic therapy-induced suppression of the contact hypersensitivity response. Photochem Photobiol 74:811 6 Haller JC, Cairnduff F, Slack G et al. (2) Routine double treatments of superficial basal cell carcinomas using aminolaevulinic acid-based photodynamic therapy. Br J Dermatol 143:127 5 Halliday GM (21) Common links among the pathways leading to UV-induced immunosuppression. J Invest Dermatol 13: Haylett AK, Ward TH, Moore JV (23) DNA damage and repair in Gorlin syndrome and normal fibroblasts after aminolevulinic acid photodynamic therapy: a comet assay study. Photochem Photobiol 78: Henderson BW, Gollnick SO, Snyder JW et al. (24) Choice of oxygenconserving treatment regimen determines the inflammatory response and outcome of photodynamic therapy of tumors. Cancer Research 64:212 6 Hilf R (27) Mitochondria are targets of photodynamic therapy. J Bioenerg Biomembr 39:85 9 Iinuma S, Schomaker KT, Wagnieres G et al. (1999) In vivo fluence rate and fractionation effects on tumor response and photobleaching: photodynamic therapy with two photosensitizers in an orthotopic rat tumor model. Cancer Res 59: Javeri A, Huang XX, Bernerd F et al. (28) Human 8-oxoguanine-DNA glycosylase 1 protein and gene are expressed more abundantly in the superficial than basal layer of human epidermis. DNA Repair (Amst) 7: Juzeniene A, Juzenas P, Ma ZL et al. (26) Topical application of 5- aminolaevulinic acid, methyl 5-aminolaevulinate and hexyl 5-aminolaevulinate on normal human skin. Br J Dermatol 155:791 9 Kousis PC, Henderson BW, Maier PG et al. (27) Photodynamic therapy enhancement of antitumor immunity is regulated by neutrophils. Cancer Res 67:151 1 Kozmin S, Slezak G, Reynaud-Angelin A et al. (25) UVA radiation is highly mutagenic in cells that are unable to repair 7,8-dihydro-8-oxoguanine in Saccharomyces cerevisiae. Proc Natl Acad Sci USA 12: Kuchel JM, Barnetson RS, Halliday GM (25) Cyclobutane pyrimidine dimer formation is a molecular trigger for solar-simulated ultraviolet radiationinduced suppression of memory immunity in humans. Photochem Photobiol Sci 4: Langmack K, Mehta R, Twyman P et al. (21) Topical photodynamic therapy at low fluence rates theory and practice. J Photochem Photobiol B 6:37 43 Lesar A, Ferguson J, Moseley H (29) A time course investigation of the fluorescence induced by topical application of 5-aminolevulinic acid and methyl aminolevulinate on normal human skin. Photodermatol Photoimmunol Photomed 25:191 5 Matthews YJ, Damian DL (21) Topical photodynamic therapy is immunosuppressive in humans. Br J Dermatol 162: Morton C, Horn M, Leman J et al. (26) Comparison of topical methyl aminolevulinate photodynamic therapy with cryotherapy or Fluorouracil for treatment of squamous cell carcinoma in situ: results of a multicenter randomized trial. Arch Dermatol 142: Morton CA, Brown SB, Collins S et al. (22) Guidelines for topical photodynamic therapy: report of a workshop of the British Photodermatology Group. Br J Dermatol 146: Morton CA, McKenna KE, Rhodes LE (28) Guidelines for topical photodynamic therapy: update. Br J Dermatol 159: Moseley H, Allen JW, Ibbotson S et al. (26) Ambulatory photodynamic therapy: a new concept in delivering photodynamic therapy. Br J Dermatol 154:747 5 Rhodes LE, de Rie MA, Leifsdottir R et al. (27) Five-year follow-up of a randomized, prospective trial of topical methyl aminolevulinate photodynamic therapy vs surgery for nodular basal cell carcinoma. Arch Dermatol 143: Robinson DJ, de Bruijn HS, van der Veen N et al. (1998) Fluorescence photobleaching of ALA-induced protoporphyrin IX during photodynamic therapy of normal hairless mouse skin: the effect of light dose and irradiance and the resulting biological effect. Photochem Photobiol 67:14 9 Singh G, Alqawi O, Espiritu M (21) Metronomic PDT and cell death pathways. Methods Mol Biol 635:65 78 Sitnik TM, Henderson BW (1998) The effect of fluence rate on tumor and normal tissue responses to photodynamic therapy. Photochem Photobiol 67:462 6 Sreevidya CS, Fukunaga A, Khaskhely NM et al. (21) Agents that reverse UV-Induced immune suppression and photocarcinogenesis affect DNA repair. J Invest Dermatol 13: Szeimies RM (27) Methyl aminolevulinate-photodynamic therapy for basal cell carcinoma. Dermatol Clin 25:89 94 Takahashi H, Nakajima S, Sakata I et al. (25) ATX-S1(Na)-photodynamic therapy is less carcinogenic for mouse skin compared with ultraviolet B irradiation. Br J Dermatol 153: Tschen EH, Wong DS, Pariser DM et al. (26) Photodynamic therapy using aminolaevulinic acid for patients with nonhyperkeratotic actinic keratoses of the face and scalp: phase IV multicentre clinical trial with 12-month follow up. Br J Dermatol 155: Varma S, Holt PJ, Anstey AV (2) Erythroplasia of queyrat treated by topical aminolaevulinic acid photodynamic therapy: a cautionary tale. Br J Dermatol 142:825 6 Wiegell SR, Haedersdal M, Philipsen PA et al. (28) Continuous activation of PpIX by daylight is as effective as and less painful than conventional photodynamic therapy for actinic keratoses; a randomized, controlled, single-blinded study. Br J Dermatol 158:74 6 Wolf P, Fink-Puches R, Reimann-Weber A et al. (1997) Development of malignant melanoma after repeated topical photodynamic therapy with 5-aminolevulinic acid at the exposed site. Dermatology 194:53 4 Wulff BC, Schick JS, Thomas-Ahner JM et al. (28) Topical treatment with OGG1 enzyme affects UVB-induced skin carcinogenesis. Photochem Photobiol 84: Yao Y, Wolverton JE, Zhang Q et al. (29) Ultraviolet B radiation generated platelet-activating factor receptor agonist formation involves EGF-Rmediated reactive oxygen species. J Immunol 182: Journal of Investigative Dermatology (211), Volume 131