Clinical, histopathological and immunohistochemical assessment of human skin field cancerization before and after photodynamic therapy

Authors


  • Funding sources
    None.

  • Conflicts of interest
    R.M.S. has served as a consultant for, and has received speakers’ honoraria and financial support to perform clinical trials from Almirall, Biofrontera, Energist, Galderma, Intendis, Novartis, Photonamic and Wyeth. The remaining authors declare no conflicts of interest.

  • R.M.S. and L.T. contributed equally to this work.

Rolf-Markus Szeimies.
E-mail: rolf-markus.szeimies@klinikum-vest.de

Abstract

Summary Background  The field cancerization concept in photodamaged patients suggests that the entire sun-exposed surface of the skin has an increased risk for the development of (pre)-malignant lesions, mainly epithelial tumours. Topical photodynamic therapy (PDT) is a noninvasive therapeutic method for multiple actinic keratosis (AK) with excellent outcome.

Objectives  To evaluate the clinical, histological and immunohistochemical changes in human skin with field cancerization after multiple sessions of PDT with methylaminolaevulinate (MAL).

Methods  Twenty-six patients with photodamaged skin and multiple AK on the face received three consecutive sessions of MAL-PDT with red light (37 J cm−2), 1 month apart. Biopsies before and 3 months after the last treatment session were taken from normal-appearing skin on the field-cancerized area. Immunohistochemical stainings were performed for TP-53, procollagen-I, metalloproteinase-1 (MMP-1) and tenascin-C (Tn-C).

Results  All 26 patients completed the study. The global score for photodamage improved considerably in all patients (< 0·001). The AK clearance rate was 89·5% at the end of the study. Two treatment sessions were as effective as three MAL-PDT sessions. A significant decrease in atypia grade and extent of keratinocyte atypia was observed histologically (< 0·001). Also, a significant increase in collagen deposition (= 0·001) and improvement of solar elastosis (= 0·002) were noticed after PDT. However, immunohistochemistry showed only a trend for decreased TP-53 expression (not significant), increased procollagen-I and MMP-1 expressions (not significant) and an increased expression of Tn-C (= 0·024).

Conclusions  Clinical and histological improvement in field cancerization after multiple sessions of MAL-PDT is proven. The decrease in severity and extent of keratinocyte atypia associated with a decreased expression of TP-53 suggest a reduced carcinogenic potential of the sun-damaged area. The significant increase of new collagen deposition and the reduction of solar elastosis explain the clinical improvement of photodamaged skin.

The deleterious effects of skin exposure to ultraviolet (UV) irradiation are well established. One of the most relevant characteristics of photoaged skin is the appearance of actinic keratosis (AK) which is the most common carcinoma in situ. Interestingly, most of these patients have multiple lesions, often aggregated on severely photodamaged, sun-exposed areas. First introduced by Slaughter et al.1 in 1953, the concept of field cancerization can easily be adopted for dermatological purposes to define a chronically photodamaged skin with multiple premalignant lesions (AK) and skin cancers. As AK has the potential to transform into squamous cell carcinoma (SCC), it is recommended to treat these lesions. The reported annual transformation rate ranges from 0·025% to 16%.2 Also, it is necessary to consider that clinically visible AK arises on an altered chronically sun-exposed field (field cancerization) and therefore many other clusters of mutated cells are present in the same field.3 These clusters transform into small patches and may, in turn, transform into AK or consecutively SCC.3,4 We believe that it is important to treat not only the visible AK (lesion-directed therapy), but the entire photodamaged surface (field-directed therapy), thus reducing the potential risk of invasive carcinoma.

Chronically sun-exposed skin also develops several changes such as roughness, sallowness, dyschromia, wrinkles and fine lines, erythema, telangiectasias and sebaceous gland hypertrophy. These changes are referred to as the photoageing process.5,6 Reactive oxygen species (ROS) generated after UV irradiation give rise to increased transcription of matrix metalloproteinases (MMPs) through complex signalling pathways, decreased expression of the procollagen-I and procollagen-III genes, and culminating in reduced dermal matrix generation. UV radiation also induces an accumulation of degraded collagen in the dermis, which antagonizes neocollagenesis.5–7 Sun-related changes in the skin involve the appearance of elastosis in association with degeneration and decrease of collagen, clinically apparent as yellow discoloration and coarse wrinkles. Histologically accumulated abnormal elastic fibres in the papillary dermis can be detected. As a result of UV-induced hyperplasia of melanocytes or increased melanogenesis, pigmentary alterations like ephelides, lentigines and a diffuse irreversible hyperpigmentation are apparent. Also, alterations in cutaneous microvasculature such as regression of small blood vessels and neoangiogenesis, resulting in telangiectasias, are seen on chronically light-exposed skin.5,8

Recently, many different lasers and light sources have been used in dermatology to reduce the signs of photoageing as well as to improve the clinical appearance of the skin.8–12 Also, topical photodynamic therapy (PDT) has been examined as a treatment for nonmelanoma skin cancer (NMSC) and the improvement of photoageing.8,11–14 PDT typically involves the application of a topical photosensitizer such as 5-aminolaevulinic acid (ALA) or its methyl ester (MAL), which is activated by exposure to a visible light source. As a consequence of the combination of light, photosensitizer and tissue oxygen, cytotoxic ROS are formed in diseased tissues, inducing necrosis and apoptosis of the malignant and premalignant cells.15 Although numerous clinical studies show that PDT improves the signs of photoageing, there remain very few studies which provide relevant histological and molecular data in this field.15

In the present study, we evaluated the clinical effect of three sessions of topical PDT with MAL on the face of patients with severe photodamaged skin and AK. Additionally, histopathology and immunohistochemical analyses of the treated field were performed to gain deeper insight into the epidermal and dermal changes before and after the procedure.

Patients and methods

Subjects

Twenty-six patients (18 women and eight men; mean age 59·4 years, range 35–90) were enrolled in the study. All patients were of skin types I–IV (Fitzpatrick classification), and had severely photodamaged skin. Inclusion criteria were at least three AKs per face, with signs of photoageing-related solar lentigines, coarse wrinkles, fine lines, diffuse mottled pigmentation and past history of sun exposure. Patients were recruited from June 2008 to July 2009. The study was approved by the local ethical committees at the Hospital das Clinicas (Universidade de São Paulo, Brazil) and the University Hospital Regensburg (Germany), and the study was registered at clinicaltrials.gov (NCT00843323). Written informed consent was obtained from all subjects prior to entry into the study. The exclusion criteria were: history of photosensitivity-related disorders, history of keloids, active infectious disease, immunosuppression, pregnancy, lactation, history of oral retinoid use within 1 year of the study entry, laser or any cosmetic treatment in the previous 6 months, other topical agents in the treatment area such as retinoids, 5-fluorouracil, imiquimod or diclofenac sodium in the previous 3 months, allergy to MAL or excipients of the cream, allergy to lidocaine, and/or poor patient compliance.

Treatment protocol

Before the procedure, a gentle bloodless curettage was performed in all patients in order to remove the scales and crusts of AK. Consecutively, a 1 mm thick layer of MAL cream 160 mg g−1 (Metvix®; Galderma, Paris, France) was applied over the AK lesions (licensed procedure) and a thinner layer over the other areas of the face without AK lesions. The entire face was then covered with an occlusive and opaque dressing. An average of 1·5 tubes of MAL (one tube contains 2 g of cream) was used for each patient and each treatment session. After 3 h of incubation, the dressing was removed and the surface was cleansed with 0·9% saline solution. The whole face was then illuminated with a light-emitting diode (LED) device with red light with a peak at 630 nm (Aktilite; PhotoCure, Oslo, Norway) with a total light dose of 37 J cm−2. The LED device was correctly positioned at 5 cm from the surface of the face and the illumination was performed on both hemifaces separately, ensuring that the whole face was thoroughly illuminated. Two additional sessions were performed monthly in all patients with the same protocol, except that the incubation time was reduced to 1·5 h. During illumination, an air-forced cooling device was used to reduce pain discomfort (Zimmer Elektromedizin, Ulm, Germany). Patients were instructed to prevent exposure of the treated skin to direct light for 48 h after each treatment and also advised to apply sunblocks [sun-protection factor (SPF) 50+] three times a day.

Clinical evaluation

During light exposures all patients were asked to evaluate the intensity of pain using a visual analogue scale (VAS), considering 0 as absence of pain and 10 as the most severe pain.16 Local and systemic adverse events were recorded during treatment and after follow-up visits.

Clinical evaluation was assessed by digital photographs before treatment (T0), before second treatment (T1), before third treatment (T2) and 3 months after last treatment session (T3). Photoageing was also evaluated at the same time points and quantified using a five point scale (0–4) based on the studies of Dover et al. and Zane et al.8,11,12,17 by two dermatologists not related to the study. All patients were scored for global photoageing, mottled pigmentation, fine lines, sallowness, roughness, facial erythema, telangiectasias and coarse wrinkles (Table 1). The number of AKs was also assessed before and after each treatment session.

Table 1.   Five-point scale for photodamage (adapted from Dover et al.8 and Zane et al.17)
Global score for photoageing
 0 Facial skin smooth to the touch, without fine lines or unevenness in pigmentation in any skin areas (cheeks, forehead, perioral)
 1 One area of significant roughness, dyspigmentation (hyper- or hypo-) or fine lines
 2 Two areas of significant roughness, dyspigmentation or fine lines or one area of roughness, dyspigmentation and fine lines
 3 Three areas of significant roughness, dyspigmentation or fine lines or two areas of roughness, dyspigmentation and fine lines
 4 Four areas of significant roughness, dyspigmentation or fine lines or three areas of roughness, dyspigmentation and fine lines
Mottled pigmentation
 0 Evenly pigmented skin
 1 Small areas of light hypo- or hyperpigmentation
 2 Small areas of moderate hypo- or hyperpigmentation
 3 Moderate areas of moderate hypo- or hyper-, large areas of light hypo- or hyper-, or small areas of heavy hypo- or hyperpigmentation
Facial erythema
 0 No evidence
 1 Small areas of light erythema
 2 Small areas of moderate erythema or moderate areas of light erythema
 3 Moderate areas of erythema, large areas of light erythema or small areas of erythema
Telangiectasia
 0 No evidence
 1 Rare and spaced
 2 Several, discrete
 3 Moderate and in close proximity
 4 Many and densely packed
Fine surface lines
 0 No evidence
 1 Rare
 2 Several, discrete
 3 Moderate and in close proximity
 4 Many and densely packed
Coarse wrinkling
 0 No evidence
 1 Superficial on one area (forehead, glabella, chin, nasolabial and periorbital)
 2 Superficial on more than one area or moderate in one
 3 Moderate on more than one area or deep in one
 4 Deep on more than one area
Sallowness
 0 Pink skin
 1 Slight suggestion of yellowness
 2 Pale with moderate suggestion of yellowness
 3 Pale with a distinct suggestion of yellowness
Tactile roughness
 0 Skin is smooth
 1 Smooth with occasional rough areas
 2 Mild roughness
 3 Moderate roughness
 4 Severe roughness

Histological and immunohistochemical assessments

Skin biopsies were taken before the first and 3 months after the last treatment. A 3-mm punch biopsy specimen was obtained under sterile conditions after local anaesthesia with lidocaine 1%. We selected the ‘normal-appearing’ skin on the field-cancerized area as the site for skin biopsies, with the aim of avoiding any AK lesion. In all patients, the selected area was preauricular. A second biopsy was performed 3 months after the last PDT at 0·5 cm distant to the previous biopsy, to eliminate the possibility of histological alterations from wound healing and scar formation. The skin samples were fixed (10% formalin), embedded in paraffin, sectioned (3 μm thickness) and stained by haematoxylin and eosin. Elastic fibres were also evaluated in this study. The specimens were submitted to Weigert counterstained by van Gieson. This last combination enhances the contrast between the elastic fibres and the background, allowing more precise digital morphometry which was quantified by image analysis and compared before and after the procedure (Image-Pro Plus version 5.1; Media Cybernetics, Bethesda, MD, U.S.A.).

Histological variables were studied and compared before and after treatments by two independent dermatopathologists. Paraffin-embedded skin samples were stained simultaneously under the same conditions with antibodies as follows: anti-TP53 (dilution 1 : 100; Dako, Glostrup, Denmark), antiprocollagen-I (dilution 1 : 100; Abcam, Cambridge, U.K), anti-MMP-1 (dilution 1 : 25; Imgenex, San Diego, CA, U.S.A.) and antitenascin-C (dilution 1 : 400; Tn-C; Biohit, Helsinki, Finland). Digital morphometry was used to analyse the images before and after PDT with the Image-Pro Plus version 5.1. Images were evaluated and the selected pixels expressed as percentage of the total area.

Statistical analysis

Clinical and histological scores were compared with the Wilcoxon nonparametric statistical test. For elastic fibre measurement and immunohistochemical results, the parametric Student’s t-test was used. The nonparametric Spearman and Pearson correlation tests were employed to determine the level of association of histological and immunohistochemical findings. < 0·05 was considered to represent statistical significance. All statistics were performed with PASW Statistics 18 for Windows (SPSS, Chicago, IL, U.S.A.).

Results

Clinical findings

All 26 patients completed the study. Under the air-forced cooling device they tolerated the illumination well: VAS mean score was 4·4 (range 3–7). Interruption of LED exposure was not necessary in any patient. As expected, the most common side-effects were erythema and oedema. The mean time for resolution of these side-effects was 5 days. Crusts and erosions were observed in the site of pre-existing AK and resolved within 1 week (Fig. 1a–c). No bacterial or viral infection or other systemic adverse events occurred during the study. No pigmentary changes or scarring were noticed.

Figure 1.

 Field cancerization of the skin before (a), immediately after illumination (b), 5 days after methylaminolaevulinate–photodynamic therapy (MAL-PDT) with crusts and scaling (c) and 3 months after three MAL-PDT sessions (d).

The Wilcoxon nonparametric test showed a statistical difference between scores of the clinical variables (Table 1) at baseline and T3 and scores at T1 and T2 for almost all variables studied. Interestingly, no statistical difference was noted between scores at T2 and T3 for the mentioned variables, except for facial erythema, which improved between T2 and T3 (= 0·046) (Fig. 2). No statistically significant difference was observed for coarse wrinkling, although a slight improvement was noted (mean ± SD score at baseline 3·35 ± 0·17, at T1 3·35 ± 0·17, at T2 3·35 ± 0·17 and at T3 3·31 ± 0·17).

Figure 2.

 Statistical values between treatments and the score improvements for clinical variables (global score, fine surface lines, mottled pigmentation, tactile roughness, coarse wrinkling, facial erythema, telangiectasia, sallowness). T0, before treatment; T1, before second treatment; T2, before third treatment; T3, 3 months after last treatment session.

The total number of AKs (lesion base) at baseline was 268. After the first treatment (T1) the total remaining number was 67, with 75% clearance rate. At T2, 39 AKs were observed, with a clearance rate of 85·4% and at T3, 28 AKs with a clearance rate of 89·5%. Mean ± SD numbers of AKs per patient were 10·31 ± 3·11 at T0, 2·58 ± 1·92 at T1, 1·50 ± 1·48 at T2 and 1·08 ± 1·47 at T3.

Histological findings

Patients showed significant changes on histology before and after treatment with MAL-PDT (Fig. 3). The mean ± SD score for atypia grade improved from 1·46 ± 0·58 at baseline to 0·69 ± 0·47 at 3 months after the last treatment session (< 0·001). The mean ± SD score for atypia extent was 0·49 ± 0·19 at baseline and 0·26 ± 0·20 at 3 months after the last PDT session (< 0·001). The mean ± SD score for epidermal thickness did not differ significantly (0·08 ± 0·02 at baseline and 0·07 ± 0·02 at 3 months) (= 0·214). The mean ± SD score for degree of elastosis was 1·77 ± 0·65 at baseline and 1·38 ± 0·64 at 3 months (= 0·002). The subepidermal collagen layer was 0·04 ± 0·05 mm thick (mean ± SD) at baseline and 0·17 ± 0·43 mm at 3 months (= 0·001). The Wilcoxon nonparametric test showed a statistical difference between baseline and after treatment for all histological variables, except for epidermal thickness. Based on the Spearman correlation coefficient, we observed a positive correlation between the variables atypia grade and atypia extent. Baseline atypia grade and extent scores showed a positive correlation (= 0·028). Notably, the correlation was even more pronounced when post-treatment scores were compared (< 0·001).

Figure 3.

 Photomicrograph of cutaneous field cancerization before (a, c) and after photodynamic therapy (b, d) showing decrease in the amount and extent of keratinocyte atypia, collagen deposition and decrease in telangiectatic blood vessels. Haematoxylin and eosin; original magnification: (a, b) × 200; (c, d) × 400.

Digital morphometry of elastic fibres was quantified by image analysis (Image-Pro Plus version 5.1). The mean ± SD baseline score of 12·93 ± 2·40 decreased to 7·35 ± 1·44, showing a decrease in the amount of elastotic material deposited in the dermis (= 0·013, Student’s t-test).

Immunohistochemical findings

Immunohistochemical studies were performed with TP-53, procollagen-I, MMP-1 and Tn-C. The mean ± SD expression of TP-53 at baseline was 5·64 ± 3·63, decreasing to 5·44 ± 3·14 after PDT. No statistical difference was noted for TP-53 (= 0·5804) although 14 of 26 patients showed a decrease of TP-53 expression. Interestingly, a positive correlation of the histological parameter ‘extent of atypia’ and TP-53 before treatment was noted based on the Pearson correlation test (= 0·002), which means that higher TP-53 expression was observed with higher amounts of cellular atypia. No correlation was found between these two parameters after PDT.

The mean ± SD expression of MMP-1 was 1·56 ± 0·97 at baseline, increasing to 1·97 ± 1·09 (= 0·08). Although no statistical difference was noted, 12 of the 26 patients showed elevation of MMP-1 after PDT (Fig. 4). The Spearman correlation test also showed a negative correlation between the histological parameter ‘degree of elastosis’ and MMP-1 after PDT, indicating that higher expression of MMP-1 was correlated with decreased elastosis (= 0·033). Likewise, for procollagen-I expression no statistical difference was noted, although 14 of 26 patients showed elevated levels. The mean ± SD baseline score was 3·97 ± 4·88 compared with 4·05 ± 5·06 at final biopsy (= 0·4769).

Figure 4.

 Immunohistochemical expressions before and after methylaminolaevulinate–photodynamic therapy of TP-53 (a, b), matrix metalloproteinase-1 (c, d), procollagen-I (e, f) and tenascin-C (g, h). Original magnifications: (a, b, e–h) × 200; (c, d) × 100.

The level of Tn-C expression increased (Fig. 4) and was statistically significant (= 0·024). At baseline the mean ± SD expression was 1·99 ± 1·51, increasing to 3·46 ± 3·31 after PDT. Tn-C is a glycoprotein that may be expressed by mesenchymal tissue, AK and SCC and also during tissue repair processes. We emphasize that the biopsy was taken from the ‘clinically normal skin’ within the field of cancerization and not in a suspected AK lesion.

Discussion

Topical PDT with either ALA or MAL has been extensively studied for skin ageing and NMSC, although most of these studies provide clinical rather than histological and immunohistochemical data.

The aim of the present study was to investigate the effect of multiple topical MAL-PDT sessions in patients with clinical signs of cutaneous field cancerization. The first treatment session was performed according to the recommended protocol for the treatment of AK with MAL-PDT (3 h incubation, 37 J cm−2 illumination). As we achieved a 75% AK clearance after the first session, we decreased the incubation time to 1·5 h, so that fewer side-effects with similar final outcome would be expected at the end of the study. As demonstrated by Braathen et al.18 a shorter incubation time of 1 h with MAL proved to be as effective as the recommended protocol for thin AK of the face. After the first session, only thin and nonhyperkeratotic lesions remained in the field of our patients, justifying the reduced incubation time.

A high clearance rate of AK with excellent cosmetic outcome was achieved after multiple MAL-PDT sessions in our patients (Fig. 1d). The treatment was well tolerated with the use of the air-forced cooling device during illumination with the LED. Also, already after two treatments efficacy results were very similar to the results after completing all three sessions of PDT (AK clearance rate 85·4% and 89·5%, respectively). All clinical parameters improved and were statistically significant, except coarse wrinkles. Interestingly, we also observed a statistically significant improvement of facial erythema and telangiectasias. These results are partially comparable with previous studies. Zane et al.17 showed no improvement for facial erythema and telangiectasias. However, they performed only two treatment sessions compared with three sessions in the present paper. We postulate that probably the collagen deposition in the upper dermis might compress the telangiectatic vessels towards the deep dermis, thus improving the erythema and telangiectasias on the surface, as a selective photothermolysis effect is not expected with PDT performed with continuous LED illumination in contrast to the use of intense pulsed light for photoactivation.

Recently, Kleinpenning et al.19 demonstrated that two sessions of MAL-PDT, 3 months apart, induced complete clearance of AK for patients with a moderate degree of actinic damage, while for patients with severe photodamage only partial clearance was observed, with an overall response of 61·9%. Also, none of the patients reaching clearance showed relapsing or new AK at the 3-month follow-up. The authors propose that maybe more PDT sessions are necessary to achieve even better response for patients with multiple AK and a severe degree of photodamage.19 Apalla et al.20 evaluating ALA-PDT for immunocompetent patients with clinical and histological signs of field cancerization, observed a significant delay in the mean time of appearance of new AK lesions in the fields treated with ALA-PDT, compared with those under placebo, during a 12-month follow-up period. In our study, we did not aim to provide extended follow-up data to prove if MAL-PDT could prevent the development of new lesions on the field, although several studies have shown the ability of PDT to reduce NMSC in humans and in animal models.21–24

Histological variables in our study showed relevant changes (Fig. 3). The amount and degree of keratinocyte atypia improved considerably after PDT in the biopsy performed 3 months after the last treatment session in all patients. Histologically, the epidermis became more stratified, with a normal appearance, and the keratinocytes tended to a normal polarization after MAL-PDT. This demonstrates that a decrease in the overall keratinocyte dysplasia in the field could be seen. These results may have profound implications as the biopsy was taken from clinically normal skin, rather than AK lesions. The improvement of these histological parameters in the field may suggest that treating the entire field could decrease the potential appearance of new lesions, as shown in previous studies.23–25

Bagazgoitia et al.26 demonstrated that a single session of MAL-PDT not only improved the level of keratinocyte dysplasia in the field but also significantly decreased TP-53 expression. Nevertheless, the authors observed that TP-53 expression still remained stable in 50% of the patients and that 45% of them still presented dysplasia after PDT. In our study, we observed a decrease in TP-53 expression in 14 of 26 patients, but the difference was not statistically significant. Importantly, in the study of Bagazgoitia et al. the biopsy was taken from a clinically visible AK lesion, so that it was more likely to obtain considerable differences in TP-53 expression after a therapeutic procedure such as topical PDT. In our study, the biopsy was obtained from clinically normal skin and we observed a trend towards a decrease in TP-53 expression. Orringer et al.27 demonstrated a 99% decrease in the overall TP-53-positive keratinocytes in field cancerization after one single session of ablative CO2 laser 21 days after treatment. After 6 months, the average staining density remained decreased at 98%. However, both mutated and wild-type TP-53 may be present in the field, and the authors utilized a nonspecific staining for TP-53. The removal of the clustered and mutated TP-53 keratinocytes is clearly suggestive of some prophylactic benefit, but the impact of reducing the number of cells overexpressing the wild-type TP-53 is still unclear. Unfortunately, we did not perform a specific immunohistochemical study for mutated TP-53. Acute exposure to UV irradiation may transiently increase expression of positive-staining wild-type TP-53, but not mutated TP-53.28 Although we asked all subjects to avoid sun exposure and advised all of them to apply SPF 50+ sunscreen three times per day, we cannot affirm that they avoided sunlight before the final biopsy.

In our study, we observed an increase in Tn-C expression in the papillary dermis after multiple MAL-PDT sessions in the field cancerization area. Tn-C is expressed in the papillary dermis and epidermis of AK lesions and is widely distributed in papillary and reticular dermis and epidermis of SCC.29 As we performed the biopsy in the normal-appearing skin, we postulate that the increase in Tn-C expression might have been due to the tissue-repair process secondary to the PDT effect on the skin. It is believed that Tn-C is particularly prominent in the basal tumour cells at the SCC invasion front and induces MMP-9 synthesis.29 The role of Tn-C in field cancerization is unclear, as we expected a decrease in the expression after multiple sessions of PDT. Latijnhouwers et al.30,31 demonstrated that elevated cytokines such as tumour necrosis factor (TNF)-α, interleukin (IL)-4 and interferon-γ may contribute to the overexpression of Tn-C in the tissue. As many cytokines are induced by PDT, it is possible that the increase in Tn-C expression observed in our study might be secondary to the healing process induced by the inflammatory effect of PDT.

Relevant dermal histological changes were also observed in the field-cancerization area after MAL-PDT. The thickness of the subepidermal collagen layer increased after PDT and the difference was statistically significant. The increased deposition of new collagen in the papillary dermis is in agreement with echographic findings shown by Zane et al.17 Additionally, the degree of elastosis decreased after PDT. Histopathology showed that the thickened and amorphous elastotic materials decreased and were restored to more fibrillar elastic fibres in the dermis. Digital morphometry of elastic fibres showed a decrease in the amount of basophilic elastotic material in the dermis. These findings are all in accordance with previous studies, suggesting that PDT in fact plays a role in restoring elastic fibres and inducing collagen synthesis.13,17,32,33 In our study, the expression of procollagen-I also increased in 14 of the 26 patients, but the result was not significant. Elevated procollagen-I expression is expected after many therapeutic procedures such as deep chemical peels, dermabrasion and CO2 laser resurfacing. Also, PDT led to increased procollagen-I expression in previous studies.32–35 It is known that the accumulation of partially degraded and fragmented collagen as well as elastotic material, such as present in photoaged skin, inhibits neocollagenesis.5–7 Moreover, collagen biosynthesis, as an indirect dermal effect of PDT, is stimulated by cytokine induction.36

MMP-1 expression as a further marker of collagen metabolism also increased in 12 of the 26 patients after PDT, although the result was not significant. Almeida Issa et al.32 showed increased expression of MMP-9 and procollagen-I after 3 months of MAL-PDT, but no changes in MMP-1. Park et al.33 observed a decrease in MMP-1, MMP-3 and MMP-12, and an increase in the expression of transforming growth factor (TGF)-β, TGF-β type II receptor (TβRII), procollagen-I and procollagen-III after 1 month of ALA-PDT. In an in vivo PDT study on humans using a pulsed-dye laser, MMP-1 expression was acutely elevated and returned to baseline levels within 24 h.35 Also in vitro an induction of MMP-1 and MMP-3 could be shown after PDT of human dermal fibroblasts.37 In an animal model, Choi et al.34 observed that elevation of primary cytokines (IL-1, TGF-β1 and TNF-α) and MMPs (1, 2, 3 and 9) occurred at early points in time after one single MAL-PDT session, while procollagen-I expression increased later after treatment.

In addition to the destruction of mature collagen, UV irradiation impairs the synthesis of new collagen via a downregulation of TβRII and accordingly impaired signalling of the TGF-β/Smad pathway.7,33 After PDT, TGF-β is upregulated and not only induces collagen synthesis, but downregulates the expression of MMPs, preventing the breakdown of new collagen fibres.7

Our results regarding the expression of procollagen-I and also MMP-1 might be explained by the time point when the final biopsies were performed. Perhaps procollagen-I and MMP-1 levels would have been even higher if the biopsy had been performed earlier. MMP-1, which is responsible for the initial proteolytic clearance of fragmented collagen, was still elevated at the time point of the second biopsy, suggesting that the degradation of fragmented collagen fibres was still in progress.

In conclusion, the results of this study on 26 patients with severe photodamage provide sufficient data to support the clinical improvement of photoaged skin after multiple sessions of MAL-PDT. We also showed that a third treatment session did not achieve additional AK clearance as compared with two sessions. Histological and immunohistochemical findings showed a decrease in severity and extent of keratinocyte atypia associated with dermal collagen deposition and improvement of solar elastosis in the field-cancerization area associated with decreased expression of TP-53 and elevated levels of procollagen-I, MMP-1 and Tn-C. These results suggest that MAL-PDT could, at least, decrease the carcinogenic potential in the human skin field-cancerization area and partially reverse the signs of extrinsic and intrinsic skin ageing. The best protocol to induce a persistent clearance of TP-53-positive keratinocytes leading to a potentially prophylactic benefit of PDT remains to be elucidated.

What’s already known about this topic?

  •  Photodynamic therapy (PDT) with porphyrin precursors such as 5-aminolaevulinic acid or its methyl ester (MAL) is effective for nonmelanoma skin cancer, especially actinic keratosis (AK).
  •  A field-targeted approach not only successfully treats AK but also affects incipient lesions.

What does this study add?

  •  PDT with MAL shows positive effects (reduced keratinocyte atypia) also on severely sun-damaged facial skin.
  •  PDT with MAL demonstrates an excellent cosmetic effect in sun-damaged facial skin with a sustained induction of newly formed collagen in the dermis.

Ancillary