Topical ALA-Photodynamic Therapy for the Treatment of
`Acne Vulgaris
`
`Wichai Hongcharu, Charles R. Taylor, Yuchiao Chang,* David Aghassi, Kittisak Suthamjariya, and
`R. Rox Anderson
`Massachusetts General Hospital, Wellman Laboratories of Photomedicine, Departments of Dermatology and *Medicine, Harvard Medical School,
`Boston, Massachusetts, U.S.A.
`
`Topical aminolevulinic acid is converted into a
`potent photosensitizer, protoporphyrin,
`in human
`hair follicles and sebaceous glands. Photodynamic
`therapy with topical aminolevulinic acid was tested
`for the treatment of acne vulgaris, in an open-label
`prospective human study. Each of 22 subjects with
`acne on the back was treated in four sites with ami-
`nolevulinic acid plus red light, aminolevulinic acid
`alone, light alone, and untreated control. Half of the
`subjects were treated once; half were treated four
`times. Twenty percent topical aminolevulinic acid
`was applied with 3 h occlusion, and 150 J per cm2
`broad-band light (550±700 nm) was given. Sebum
`excretion rate and auto-¯uorescence from follicular
`bacteria were measured before, and 2, 3, 10, and
`20 wk after, treatment. Histologic changes and pro-
`toporphyrin synthesis in pilosebaceous units were
`observed from skin biopsies. Aminolevulinic acid
`plus red light caused a transient acne-like folliculitis.
`Sebum excretion was eliminated for several weeks,
`and decreased for 20 wk after photodynamic therapy;
`
`A cne vulgaris is a skin disease affecting more than 80%
`
`of young people. Propionibacterium acnes and sebum
`secretion play major roles in the pathogenesis of acne.
`Topical and systemic antibiotics are mainstays for
`treatment of acne, but the success rate varies in part
`due to the gradual resistance to antibiotics. Sun exposure has a well-
`known bene®cial effect on acne, which is not
`the case for
`ultraviolet exposure (Sigurdsson et al, 1997). Studies show that the
`bacteria produce porphyrins as a by-product of their metabolism.
`Visible light
`is known to activate the porphyrins,
`inducing a
`photodynamic reaction that
`subsequently kills
`the pathogenic
`bacteria (Kjeldstad, 1984). Furthermore, photodynamic reactions
`can kill all strains of bacteria (Soukos et al, 1998).
`Photodynamic therapy (PDT) with topical aminolevulinic acid
`(ALA) has been used to treat nonmelanoma skin cancer, actinic
`keratoses, and psoriasis (Szeimies et al, 1996). Topically applied
`
`Manuscript received December 15, 2000; revised April 18, 2000;
`accepted for publication May 8, 2000.
`Reprint requests to: Dr. R. Rox Anderson, Wellman Laboratories of
`Photomedicine, Massachusetts General Hospital, BHX 630, 50 Blossom
`Street, Boston, MA 02114. Email: andersonr@helix.mgh.harvard.edu
`Abbreviations: ALA, aminolevulinic acid; PDT, photodynamic therapy;
`PpIX, protoporphyrin IX; SO, sebum output.
`
`multiple treatments caused greater suppression of
`sebum. Bacterial porphyrin ¯uorescence was also
`suppressed by photodynamic therapy. On histology,
`sebaceous glands showed acute damage and were
`smaller 20 wk after photodynamic therapy. There
`was clinical and statistically signi®cant clearance of
`in¯ammatory acne by aminolevulinic acid plus red
`light, for at least 20 wk after multiple treatments and
`10 wk after a single treatment. Transient hyperpig-
`mentation, super®cial exfoliation, and crusting were
`observed, which cleared without scarring. Topical
`aminolevulinic acid plus red light
`is an effective
`treatment of acne vulgaris, associated with signi®cant
`side-effects. Aminolevulinic acid plus red light causes
`phototoxicity to sebaceous follicles, prolonged sup-
`pression of sebaceous gland function, and apparent
`decrease in follicular bacteria after photodynamic
`therapy. Potentially, aminolevulinic acid plus red
`light may be useful for some patients with acne. J.
`Invest Dermatol 115:183±192, 2000
`
`taken up by epithelial cells and metabolized via the
`ALA is
`porphyrin pathway to protoporphyrin IX (PpIX), the precursor of
`heme (Kappa et al, 1989;
`Iinuma et al, 1994). PpIX is a
`photosensitizer that accumulates not only in the epidermal cells
`but also in the pilosebaceous units (Divaris et al, 1990; Kennedy and
`Pottier, 1992). When intense visible light is delivered on the ALA-
`treated skin, PpIX is excited into a triplet state, which reacts with
`oxygen to produce singlet oxygen, causing membrane damage and
`cell destruction. Topical ALA may directly enter hair follicles,
`where sebaceous glands actively synthesize and retain PpIX. We
`conducted this pilot study to test the hypothesis that photodynamic
`destruction of P. acnes, sebaceous glands, or both would occur in
`human skin, improving acne vulgaris.
`
`MATERIALS AND METHODS
`
`Subject selection Twenty-two subjects of both sexes with mild to
`moderate acne vulgaris (grades 1±4) (Burke and Cunliffe, 1984) on their
`backs were enrolled between October 1998 and March 1999. People were
`excluded if they had used any topical acne treatment, systemic antibiotics in
`the past 2 wk, or systemic retinoids in the past year. People were also
`excluded who were using medication that may exacerbate or alleviate acne,
`who were planning to have excessive sunlight exposure, who had a history
`of keloid or photosensitivity disorder, or who had Fitzpatrick's
`skin
`phototype V-VI; pregnant and lactating women were also excluded.
`
`0022-202X/00/$15.00 ´ Copyright # 2000 by The Society for Investigative Dermatology, Inc.
`
`183
`
`

`

`184 HONGCHARU ET AL
`
`THE JOURNAL OF INVESTIGATIVE DERMATOLOGY
`
`Table I. Subject characteristics in both groups
`
`Single-treatment group
`
`Multiple-treatment group
`
`Age (y, mean + SD)
`Gender (M/F)
`Skin phototype
`
`Disease history (y, mean + SD)
`Previous systemic antibiotic treatment (number of subjects)
`Previous topical antibiotic treatment (number of subjects)
`Previous systemic isotretinoin treatment (number of subjects)
`Number of baseline comedones (median, range)
`Number of baseline in¯ammatory comedones (median, range)
`Number of baseline papules (median, range)
`Number of baseline pustules (median, range)
`Number of baseline nodules (median, range)
`Number of baseline cysts (median, range)
`
`30 + 8.74
`9/2
`Type I 9.1%, Type II 27.3%
`Type III 54.5%, Type IV 9.1%
`11.45 + 8.38
`3 (27%)
`3 (27%)
`3 (27%)
`3.0, 30
`3.0, 41
`4.5, 22
`0.0, 3
`0.0, 3
`0.0, 0
`
`27 + 4.56
`8/3
`Type I 18.2%, Type II 36.4%
`Type III 27.3%, Type IV 18.2%
`11.27 + 4.24
`4 (36%)
`4 (36%)
`2 (18%)
`3.5, 34
`2.5, 17
`6.5, 33
`0.0, 2
`0.5, 13
`0.0, 0
`
`In the multiple-treatment group, subjects were treated once a week for four
`consecutive weeks. In this group, if severe exfoliation, erosions, or purpura
`occurred, treatment was postponed to the following week. In both groups,
`subjects returned 1 wk after treatment for clinical evaluation and at weeks
`2, 3, 10, and 20 for clinical, ¯uorescence, and SO evaluations.
`
`Clinical evaluations Each subject's acne was visually assessed using an
`in¯ammatory acne score modi®ed from that previously described
`(Michaelsson et al, 1977). The modi®cation we used in this
`study
`accounted for both number and size of acne lesions. The numbers of
`comedones, in¯ammatory comedones, papules, pustules, nodules, and cysts
`in each test area were recorded. Each type of lesion was given a severity
`index as follows: 0.5 for comedo (<1 mm), 0.75 for in¯ammatory comedo,
`1 for papule (1-5 mm), 2 for pustule, 3 for nodule (>5 mm), and 4 for
`in¯ammatory cyst.
`Clinical
`improvement was globally assessed by three dermatologists
`unaware of the status of treatment, who blindly graded changes in acne
`from ®xed-magni®cation clinical photographs, after being shown a small set
`of standardized series of training slides not used in the data evaluation. The
`grading scale was de®ned as ±3 for >50% exacerbation, ±2 for 25%+-50%
`exacerbation, ±1 for 1%+-25% exacerbation, 0 if unchanged, 1 for 1%+-
`25% improvement, 2 for 25%+-50% improvement, 3 for 50%+-75%
`improvement, 4 for 75%+-99% improvement, and 5 for 100% improve-
`ment, compared with the baseline.
`
`Fluorescence photography A Nikon E2N digital camera body with a
`Nikon 105 mm macro lens was used. A ®lter (Corion LL-550S) was placed
`on the lens to block light below 550 nm. The excitation light source was
`composed of
`two synchronized photo¯ashes with Norman 400 W s
`lampheads (FT400/FT6), mounted on a stationary tower with angles of
`incidence of 60° bilaterally. Two 400 nm bandpass ®lters with 5 nm
`bandwidth (Corion S40±400S) were placed on the ¯ashes. By this method,
`the punctate orange-red ¯uorescence of hair follicles populated with P.
`acnes was seen (Lucchina et al, 1996). Fluorescence emission has been
`attributed to bacterial coproporphyrin III
`and protoporphyrin IX
`(Cornelius and Lugwig, 1967; Lucchina et al, 1996), and intensity of
`¯uorescence is related to the P. acnes population (Cornelius and Lugwig,
`1967; Lucchina et al, 1996). Fluorescence photography was performed at
`weeks 0, 2, 3, 10, and 20 in all sites. The number of punctute red
`¯uorescent dots was counted blindly for each test area.
`
`Sebum-absorbent tape (Sebutapes, CuDerm, Dallas,
`SO measurement
`TX) is a noninvasive, easy, and reproducible method to evaluate human SO
`(Pagnoni et al, 1994a). The subject's skin was shaved and then cleansed for
`15 s with cotton pads
`soaked in 70% ethanol. When the skin was
`completely dry, a strip of Sebutape was adhered to each test site for an hour.
`After removal from the skin, the white tape was placed on a black card for
`image analysis. Small transparent spots due to sebum excretion from follicles
`were visualized as a black spot on the white background. A CCD camera
`and digital frame grabber were used to capture images of the Sebutape,
`which were then examined using a computer-assisted image analysis (IP-
`LAB) system. The percentage of Sebutape area covered by sebum spots
`(black) was calculated. We considered the percentage area covered by the
`spot as the relevant measure of SO (Pagnoni et al, 1994a). The percentage
`
`Figure 1. Transient acneiform eruption caused by a single PDT
`treatment. (a) Baseline; (b) one week post-treatment.
`
`Study design Subjects were randomly divided into single-treatment and
`multiple-treatment groups. Each patient's back was equally divided into
`four 7.5 3 10 cm areas for ALA plus red light (ALA-PDT), ALA alone,
`light alone, and untreated control. Sites were marked with templates to
`precisely relocate each test area. At baseline, clinical evaluations, natural
`bacterial porphyrin ¯uorescence photography, and sebum output (SO)
`evaluation were performed. Before application of ALA, the skin was
`cleaned with 70% isopropyl alcohol. Then, 20% topical ALA in a
`hydroalcoholic vehicle (Levulan, DUSA Pharmaceuticals) was applied for
`3 h under occlusion with plastic ®lm (Saran wrap), and 150 J per cm2 broad-
`band light (550±700 nm) was given to the ALA-PDT and light alone areas.
`
`

`

`VOL. 115, NO. 2 AUGUST 2000
`
`PHOTODYNAMIC THERAPY OF ACNE
`
`185
`
`area correlates directly with the SO (Pierard, 1987). Sebutape assays of SO
`were done this way, at weeks 0, 2, 3, 10, and 20, in all sites.
`
`Adverse effects Adverse effects were scored by clinical evaluation of
`erythema, edema,
`loss of epidermis, hyperpigmentation, hemorrhage,
`vesiculation, and exfoliation on a visual analog scale from 0 to 3 (0 = absent,
`1 = mild, 2 = moderate, 3 = severe) for each ®nding. Subjective sensation of
`pain, burning, and itching was generally maximum about 10 min into light
`exposure, and was ranked at that time and at the end of treatment (1 h) by
`subjects on a scale from 0 to 3 similar to above.
`
`Histologic examination Punch biopsy specimens (4 mm) were taken
`immediately after PDT, a few weeks after PDT, and at 20 wk, from both
`the untreated control and ALA-PDT areas. Specimens were sectioned in
`either vertical or horizontal fashion and stained with hematoxylin and
`eosin,
`Fontana-Masson,
`and Masson-trichrome
`stains. Histologic
`examination was performed. Cross-sectioned areas of sebaceous glands,
`representative sebocytes, and the sebocyte nuclear area were measured from
`planimetric analysis of serial sectioned specimens of the skin using a
`computer-assisted planimetry system (Weissmann et al, 1984). To minimize
`the variation of sebaceous gland and sebocyte areas due to huge differences
`in cross-section and to provide a representative estimate of sebaceous gland
`and sebocyte areas, the largest sebaceous area of each follicle and largest
`sebocytes near the center of follicles from each serial sectioned specimen
`were measured. The area of sebaceous gland and the cytoplasm/nuclear
`area ratio in sebocytes were calculated and compared between control and
`PDT areas at each follow-up. To determine the level of PpIX converted
`from ALA in the pilosebaceous units, punch biopsy specimens were also
`taken from ALA-treated areas after 3 h occlusion as described above. A
`series of horizontal cross-sections of fresh-frozen specimens was obtained,
`and localization of PpIX production was noted by ¯uorescence microscopy.
`Histologic examinations were performed to get a qualitative picture of
`reactions to PDT. A total of 15 specimens were obtained. Eight biopsies of
`PDT-treated areas were taken with accompanying specimens from the
`nontreatment area: four from multiple PDT-treated areas at follow-up 5,
`one from a multiple PDT-treated area at follow-up 3, one from a single
`PDT-treated area immediately after PDT, one from a single PDT-treated
`area at follow-up 3, and one from a single PDT-treated area at follow-up 5.
`Seven biopsies were obtained without an accompanying specimen from
`control areas, and were analyzed for morphologic changes due to PDT:
`two from single PDT-treated areas immediately after PDT, one from an
`acneiform lesion appearing 3 d after PDT, one from a single PDT-treated
`area at follow-up 2, one from a single PDT-treated area at follow-up 5, and
`one from a multiple PDT-treated area at follow-up 3.
`
`Statistical analysis Treatment effects were determined based on the
`following analyses: (1) comparing the scores from each follow-up visit to
`the baseline scores using paired t tests; (2) comparing the change from
`baseline among the four treatment sites using paired t tests; (3) comparing
`the change from baseline between the single-treatment and multiple-
`treatment groups using two-sample t tests; and (4) comparing the change
`from baseline between the single-treatment and multiple-treatment groups
`using a repeated measures analysis to combine data from all follow-up visits.
`Statistical signi®cance was de®ned as a p-value of less than 0.05.
`
`RESULTS
`
`the 23 subjects enrolled, 22 (17 males and ®ve females)
`Of
`completed the study. One was dropped from the study because his
`asthma necessitated systemic steroid treatment, which is one of the
`exclusion criteria. The age of patients completing the study ranged
`from 18 to 44 y. Characteristics of the subjects in both groups are
`shown in Table I.
`An impressive, acute eruption of in¯ammatory acneiform lesions
`was observed in the ALA-PDT sites only, in all patients (100%) in
`both groups, starting approximately 3±4 d post-treatment (Fig 1).
`The induced lesions were papules, pustules, and nodules that lasted
`for 4 d to 3 wk in the single-treatment group. In the multiple-
`treatment group, subsequent treatments induced progressively less
`in¯ammatory acne, such that almost no new acneiform lesions were
`observed after treatment 4.
`
`In¯ammatory acne score (Figs 2a and 3)
`
`Single-treatment group Only the area treated with PDT showed
`improvement in acne, which was statistically signi®cant starting
`
`Figure 2. The mean improvement (+ SEM) by treatment sites,
`treatment groups, and follow-up visits. (a) Reduction in in¯ammatory
`acne score;
`(b) global clinical-improvement grading;
`(c) reduction in
`auto¯uorescence of follicles, related to P. acnes; (d) reduction in sebum
`excretion rate. - - -, single treatment group; ÐÐ, multiple treatment
`group; red, PDT; blue, untreated; deep blue, light alone; green, ALA alone.
`
`3 wk after treatment. The other three areas (ALA alone, light alone,
`untreated) showed slightly worse acne not signi®cantly different
`from baseline, for all visits. When comparing the change from
`baseline between the area treated with single PDT and the other
`three areas, the differences were statistically signi®cant at 3, 10, and
`20 wk.
`
`

`

`186 HONGCHARU ET AL
`
`THE JOURNAL OF INVESTIGATIVE DERMATOLOGY
`
`Figure 3. In¯ammatory acne improved by a single PDT treatment. (a) Baseline. (b) Ten weeks post-PDT. (c) Acne starts to resume 20 wk after PDT.
`Long-term remission of acne after multiple PDT treatments. (d) Baseline. (e) Two weeks post-PDT (an irritation reaction to Sebutape is seen on the right-
`hand side in this subject). (f) Twenty weeks post-PDT.
`
`Figure 4. Fluorescence of porphyrin from
`bacteria in follicles (red dots) decrease after a
`single PDT. Photographs were taken as described
`at baseline (A), week 3 (B), and week 20 (C)
`post-PDT.
`
`and statistically
`group There was obvious
`Multiple-treatment
`signi®cant improvement in acne at all follow-up visits after multiple
`PDT treatment. There was no improvement in ALA-alone, light
`alone, or untreated sites. Change from baseline was signi®cantly
`greater at sites of multiple PDT compared with the other three sites
`for all visits (p < 0.05). At visit 2 only (week 2), there was a barely
`signi®cant
`improvement
`in the area treated with ALA alone
`compared with the untreated area (p = 0.046).
`
`Comparison between single- and multiple-treatment groups The multi-
`ple PDT treatment group showed signi®cantly more improvement
`than the single PDT treatment group at the ®rst three follow-up
`visits. This difference diminished after week 3. No signi®cant
`differences between the multiple- and single-treatment groups
`were observed in the non-PDT sites with respect to each individual
`visit. When data from all follow-up visits were combined, the
`
`multiple PDT and multiple ALA alone treatment sites showed
`more improvement than the single-treatment group (p < 0.001 and
`p = 0.007, respectively).
`
`Global clinical-improvement score (Fig 2b)
`
`Single-treatment group The PDT site showed signi®cant global
`improvement starting week 3 and extending through week 20. The
`area without treatment, and the area treated with light alone, also
`showed improvement reaching statistical signi®cance at weeks 3
`and 20 (p = 0.017 and 0.018, respectively). The difference between
`PDT and the other three treatment sites was statistically signi®cant
`at weeks 3 and 10.
`
`for the PDT
`improvement
`Signi®cant
`Multiple-treatment group
`treated area was observed starting visit 1 (week 1) and this
`
`

`

`VOL. 115, NO. 2 AUGUST 2000
`
`PHOTODYNAMIC THERAPY OF ACNE
`
`187
`
`Figure 5. Sebum excretion is suppressed by a
`single PDT, then gradually recovers. (A) At
`baseline,
`(B) week 2 post-PDT,
`(C) week 10
`post-PDT, (D) week 20 post-PDT.
`
`sup-
`remains
`Figure 6. Sebum excretion
`pressed after multiple PDT treatments, for at
`least 20 wk. (A)±(D) as in Fig 5 above.
`
`improvement persisted throughout all the four follow-up visits (up
`to 20 wk at least). The area treated with ALA alone at visit 2 and the
`area treated with light alone or ALA alone at visit 5 also showed
`improvement reaching statistical signi®cance. There was signi®-
`cantly more improvement, however, in the PDT treated site than
`the other three sites, at all follow-up visits.
`
`Comparison between single- and multiple-treatment groups The multi-
`ple PDT treatment group showed signi®cantly more improvement
`than the single PDT treatment group when evaluated at the ®rst
`two follow-up visits (weeks 1 and 2). The single-treatment group
`did not have signi®cantly more acne improvement than multiple
`treatment, at any time. When data from all follow-up visits were
`combined, the comparison between single PDT and multiple PDT
`reached statistical signi®cance (p = 0.008).
`
`Fluorescence photography evaluation (Figs 2c and 4)
`
`showed
`group Only the PDT treated sites
`Single-treatment
`signi®cant loss of ¯uorescence related to P. acnes, which lasted for
`all four follow-up visits. The differences between PDT and the
`other three test sites were also statistically signi®cant for all visits.
`
`showed
`group Again, only the PDT sites
`Multiple-treatment
`signi®cant loss of P. acnes ¯uorescence, starting at follow-up visit
`2. The sites treated with ALA alone or untreated had signi®cantly
`greater ¯uorescence than baseline, at weeks 10 and 20. The
`differences between the PDT area and the other three test sites
`were statistically signi®cant for all visits.
`
`Comparison between single- and multiple-treatment groups The group
`treated four times with PDT had more, but not signi®cantly more,
`
`

`

`188 HONGCHARU ET AL
`
`THE JOURNAL OF INVESTIGATIVE DERMATOLOGY
`
`SO in the area treated with PDT was also signi®cantly lower than
`any other test sites, at each follow-up visit.
`
`Multiple-treatment group The PDT sites showed signi®cant de-
`crease in SO, at all follow-up visits. The differences between the
`area treated with PDT and the other three areas were also
`statistically signi®cant at all four visits.
`
`Comparison between single- and multiple-treatment groups Multiple
`PDT suppressed SO more than single PDT; however,
`the
`difference was signi®cant only at
`the longest
`follow-up time,
`20 wk. When data from all
`follow-up visits were combined,
`multiple PDT caused far lower SO than single PDT (p = 0.001).
`
`in pilosebaceous
`Histology Marked focal histologic changes
`glands were observed in all samples treated by PDT. In the control
`(untreated) biopsy specimens, all
`subjects had well-developed
`sebaceous gland with typical round or oval lobules. Immediately
`after PDT, there was a mixed, neutrophil-predominant in®ltrate
`along pilosebaceous units and perivascular area, and retiform
`degeneration of sebocytes (Fig 7). There was an apparent reduction
`of sebaceous gland size, with a mean decrease of 40% immediately
`after PDT. Epidermal changes were also observed with epidermal
`necrosis, vacuolization of keratinocytes
`from the mid stratum
`spinulosum to the stratum granulosum, and neutrophilic exocytosis.
`At 3 d after PDT, the acneiform lesions induced by PDT showed
`large intraepidermal pustules, disruption of hair follicles, and frank
`sebaceous gland destruction replaced by a mixed, neutrophil-
`predominant dermal in®ltrate (Fig 8).
`Reduction of sebaceous gland size compared with the untreated
`control area was observed 3 wk after both single and multiple PDT
`(30% vs 55%). Focal vacuolization of sebocytes (Fig 9), granulo-
`matous reaction, and perifollicular ®brosis were also observed,
`although some of the sebaceous glands had regained a normal
`morphology with smaller size relative to control. Cytoplasm/
`nuclear cross-section area ratio in sebocytes was reduced, relative to
`nontreatment, by 38% and 56% in single and multiple PDT,
`respectively.
`In the multiple-treatment group, improvement continued longer
`into the follow-up period, such that by the end of the study (20 wk
`after the last treatment), reduction of sebaceous gland size and
`sebocyte cytoplasm/nuclear area were 45% (range from 15% to
`80%) and 46% (range from 39% to 53%), respectively. At 20 wk
`after multiple PDT treatments, there was complete destruction or
`marked atrophy of sebaceous gland lobules, with comparatively few
`sebocytes present (Fig 10a). Frequently, a granulomatous reaction
`composed of multinucleated giant cells and histiocytic in®ltrates
`was seen at the remnant of destroyed sebaceous glands (Fig 10b).
`Perifollicular ®brosis (Fig 10c), in¯ammation, and spongiosis were
`seen occasionally, but
`these ®ndings were not constant. The
`epidermis appeared completely normal.
`Twenty weeks after a single PDT treatment, there was only a
`slight reduction of sebaceous gland size (a mean decrease of 17%)
`without other apparent morphologic changes or in®ltrates. The
`sebocyte cytoplasm/nuclear area ratio was not reduced.
`Masson-trichrome
`stained specimens
`showed perifollicular
`®brosis caused by single or multiple PDT, and mild disarray of
`collagen bundles in the mid-reticular dermis by multiple PDT.
`Fontana-Masson stain showed higher epidermal pigmentation after
`PDT, and slightly more dermal melanophages (pigment incon-
`tinence) compared with untreated sites.
`Fluorescence microscopy of
`fresh-frozen sections after ALA
`application showed bright porphyrin ¯uorescence in epidermis and
`pilosebaceous units, compared with untreated skin. There was
`brighter PpIX ¯uorescence in sebocytes
`than in the adjacent
`follicular epithelial cells (Fig 11).
`
`Adverse effects Erythema and edema were most intense about
`10 min after the beginning of PDT and subsided to lesser intensity
`by the end of the light exposure. There was a substantial decrease in
`
`Figure 7. Retiform degeneration of sebocytes and intense mixed
`neutrophil-predominant in®ltrate. (a) Sebocytes; (b) intense mixed
`neutrophil-predominant in®ltrate. The biopsies were taken immediately
`after a single PDT treatment. Scale bars: 100 mm.
`
`improvement than that treated with single PDT. Combining data
`from all visits, the difference in ¯uorescence related to P. acnes
`between single- and multiple-treatment groups still did not reach
`statistical signi®cance (p = 0.081).
`
`SO (Figs 2d, 5, and 6)
`
`Single-treatment group Only the area treated with PDT showed a
`signi®cant decrease in SO, which was at weeks 2, 3, and 20. The
`
`

`

`VOL. 115, NO. 2 AUGUST 2000
`
`PHOTODYNAMIC THERAPY OF ACNE
`
`189
`
`Figure 8. Neutrophillic pustules are seen 3 d
`after ALA-PDT, intraepidermally and within
`pilosebaceous units, associated with the ac-
`neiform eruption caused by PDT. (a) Intraepi-
`dermally;
`(b) within pilosebaceous units. The
`sections were stained with hematoxylin and eosin.
`Scale bars: 100 mm.
`
`in both groups.
`erythema and edema by 1 h after treatment,
`Subjective reports of pain, burning, and itch were more severe at
`10 min after starting PDT than at the end of treatment. A burning
`sensation became more severe with subsequent treatments and was
`the main complaint
`(9%, 67%, 67%, and 73% of
`subjects at
`treatments 1, 2, 3, and 4, respectively). Itching was the next
`frequent subjective side-effect of subsequent treatments (73%, 73%,
`55%, and 55% of subjects at treatments 1, 2, 3, and 4). In contrast to
`multiple PDT, itching was the main discomfort in the single PDT
`group, and pain was the least.
`Erythema, hyperpigmentation, and exfoliation were commonly
`seen after PDT. Six patients in the multiple-treatment group could
`not continue the weekly treatment scheme and had to postpone
`their next treatment: two at treatments 3 and 4, two at treatment 3,
`and two at treatment 4. Erythema and hyperpigmentation faded
`away completely at 20 wk in 82% and 91%, respectively, of the
`single-treatment subjects. No subjects had exfoliation after 3 wk
`post-treatment. One subject in the single PDT group developed
`blistering in the PDT site after vigorous aerobic exercise while
`wearing a tight out®t the day after treatment. This area healed
`without scarring in 3 wk. In fact, no site in any subject had any
`scarring. Multiple PDT caused long-lasting hyperpigmentation
`with 55% of subjects still showing some degree of pigmentation at
`20 wk after treatment. In 10% of multiple-treatment
`subjects,
`super®cial but very prominent exfoliation was seen after four
`
`treatments, with transient purpura (average 1 wk) and partial loss of
`epidermis.
`
`DISCUSSION
`
`PDT has been used mainly for solid tumor treatment, but in
`dermatology its potential applications include a host of in¯amma-
`tory, proliferative, or angiogenic skin lesions
`such as actinic
`keratoses (Jeffes et al, 1997), psoriasis (Collin et al, 1997), verrucae
`(Ammann et al, 1995), and cutaneous T cell lymphoma (Wolf et al,
`1994). Beginning with Kennedy's pioneering studies (Kennedy et al,
`1992), topical ALA has been studied in skin because it is a nontoxic,
`naturally occurring, substance that targets cells expressing porphyrin
`synthesis. ALA is
`the ®rst committed precursor of
`the heme
`synthesis pathway, which is regulated by inhibition of ALA
`synthase (Kappa et al, 1989). Photosensitizing porphyrins accumu-
`late in skin after ALA is applied, presumably after depletion of
`intracellular iron stores. In normal skin, the epidermis, hair follicles,
`and sebaceous glands accumulate PpIX in high concentrations after
`systemic ALA (Divaris et al, 1990). In our study ¯uorescence
`photography shows ALA-induced PpIX ¯uorescence is greater in
`acne lesions than in surrounding tissue (Fig 12). P. acnes bacteria
`produce porphyrins to the extent that red ¯uorescence is easily
`seen, and is correlated with P. acnes colonization of sebaceous
`follicles. It has been shown that exogenous ALA can cause a
`
`

`

`190 HONGCHARU ET AL
`
`THE JOURNAL OF INVESTIGATIVE DERMATOLOGY
`
`improvement of acne after PDT is due primarily to inhibition of
`sebum, to killing of P. acnes, or to secondary host responses. We
`note, however, that suppression of sebum production appears to
`correlate best with clinical improvement.
`The acneiform eruption occurring 3±4 d after the ®rst ALA-
`PDT treatment was a constant ®nding in this
`study. The
`mechanism for this eruption is unknown. We hypothesize that
`PDT disrupts
`sebocyte and P. acnes membranes,
`activating
`complement and neutrophil migration into the perifollicular area.
`Reactive oxygen species produced by neutrophils play a signi®cant
`role in disrupting the follicular epithelium, which is responsible for
`the in¯ammatory process of acne (Akamatsu and Horio, 1998). In
`addition, P. acnes activates complement and produces C5a, a potent
`neutrophil chemotactic factor (Webster et al, 1986). The bacterial
`cell wall peptidoglycan-polysaccharide substance may also play a
`role in stimulating an immune granuloma type reaction (Vowel et al,
`1995), which was seen in this study following multiple PDT. In the
`multiple-treatment group, each subsequent PDT treatment pro-
`duced a progressively less in¯ammatory and weaker acneiform
`eruption, which is at
`least consistent with sterilization of the
`follicles by the ®rst treatment.
`ALA-PDT is not a trivial procedure, and has side-effects. We
`used an aggressive ALA-PDT treatment dose in order to test the
`possibility of effects on acne vulgaris. Fortunately, there was no
`scarring in this study, after a total of about 60 PDT sessions in 23
`subjects. Each treatment, however, takes time, is painful or pruritic,
`causes acute erythema and edema, occasionally causes blistering and
`purpura, causes an acute acneiform eruption, and usually leads to
`hyperpigmentation that fades gradually over weeks to months. To
`most people, these side-effects would be tolerable in practice only if
`PDT were able to permanently improve acne, which remains a
`distinct possibility. This situation is not unlike the use of systemic
`retinoids (Geiger, 1995), which produce both long-lasting bene®ts
`and major side-effects. We did nothing in this study to optimize or
`alleviate the side-effects of ALA-PDT, and it is rather unlikely that
`we happened to use the best conditions for PDT of acne. Dose±
`response characteristics
`for ALA-PDT treatment of acne are
`unknown. We plan future studies exploring the precise dosimetry.
`We hypothesize, for example, that killing of P. acnes requires much
`less
`aggressive treatment
`than is needed to inhibit
`sebum
`production. One might therefore ask what the least aggressive
`ALA-PDT regimen effective for improving acne is. Many other
`factors could be optimized. Uptake of ALA into sebaceous follicles
`might be enhanced or made more selective by the vehicle and
`application conditions; ALA can be delivered rapidly into skin by
`iontophoresis (Rhodes et al, 1997); using a lower light irradiance
`and fractionated exposures is known to reduce the pain of PDT;
`topical anesthesia could be used; drugs that protect patients with
`endogenous porphyrias, such as beta-carotene or anti-in¯ammatory
`agents, might decrease some of the side-effects without decreasing
`the ef®cacy of ALA-PDT; a series of less aggressive PDT treatments
`may be preferable; it may even be possible to combine ALA with
`sunscreen preparations to allow sunlight treatment over large body
`areas. These and other ideas may be worth pursuing because it is
`now clear that acne vulgaris responds to ALA-PDT.
`
`The authors acknowledge William Farinelli, Norm Michaud, Nik Kollias, and
`Kim Palli for technical assistance and clinical study coordination. Dr. Hongcharu is
`supported in part by the U.S. Department of Energy, Grant No. DOE-FG02±
`91ER61228. We thank DUSA Pharmaceuticals Inc. for donating the ALA
`supply. DUSA did not fund this research, and none of the authors has any ®nancial
`interest in DUSA or ALA-PDT for acne. The study was performed under FDA,
`IND #55249.
`
`REFERENCES
`
`Akamatsu H, Horio T: The possible role of reactive oxygen species generated by
`neutrophils in mediating acne in¯ammation. Dermatology 196:82±85, 1998
`
`Figure 9. F