Comparison of Physical Pretreatment Regimens to Enhance Protoporphyrin IX Uptake in Photodynamic Therapy: A Randomized Clinical Trial

Key Points

Question  What is the relative potential of curettage, microdermabrasion, microneedling, and fractional laser pretreatment to enhance methyl aminolevulinate hydrochloride–induced protoporphyrin IX formation in photodynamic therapy?

Findings  This randomized clinical trial of 12 healthy participants demonstrates that pretreatment with an ablative fractional laser significantly intensifies protoporphyrin IX fluorescence to a larger extent than curettage, microdermabrasion, microneedling, and nonablative fractional laser. Moreover, increasing laser densities (2%-6%) and the number of pretreatment passes (1-3) did not further enhance protoporphyrin IX fluorescence.

Meaning  Ablative fractional laser treatment has a higher potential than curettage, microdermabrasion, microneedling, and nonablative fractional laser to enhance the PDT response in normal skin.

Importance  Skin pretreatment is recommended for adequate penetration of topical photosensitizing agents and subsequent protoporphyrin IX (PPIX) accumulation in photodynamic therapy (PDT).

Objective  To compare the relative potential of different physical pretreatments to enhance PPIX fluorescence in normal skin.

Design, Setting, and Participants  This intraindividual, randomized clinical trial was performed from November 28 to December 20, 2014, at Bispebjerg Hospital, Copenhagen, Denmark, among 12 healthy volunteers 18 years or older. Analysis was based on intention to treat. All participants completed the study protocol.

Interventions  Participants were each exposed to standardized skin preparation with curettage, microdermabrasion with abrasive pads, microneedling with dermarollers, ablative fractional laser (AFXL), non-AFXL, and no pretreatment, followed by 3 hours of methyl aminolevulinate hydrochloride incubation and subsequent red light illumination.

Main Outcomes and Measures  The primary outcome measure was methyl aminolevulinate–induced PPIX fluorescence accumulation. Secondary outcome measures were PPIX photobleaching and clinical local skin reactions, supported by noninvasive reflectance measurements of percentage of skin redness, transepidermal water loss, and participant-assessed pain.

Results  Among the 12 healthy study participants (8 men; 4 women; mean [SD] age, 33 [15] years), histologic findings confirmed standardization of interventions with partial removal of the stratum corneum after curettage and microdermabrasion and similar vertical penetration depths for microneedling, AFXL, and non-AFXL (median, 125 μm). PPIX fluorescence reached highest intensities in skin pretreated with AFXL (median, 8661 arbitrary units [AU]) compared with microdermabrasion (median, 6731 AU), microneedling (median, 5609 AU), and curettage (median, 4765 AU) (P < .001), among which similar enhancement was shown. Comparatively lower fluorescence levels were demonstrated for skin pretreated with non-AFXL (median, 2898 AU), methyl aminolevulinate–treated controls (median, 2254 AU), and untreated controls (median, 239 AU) (P < .03). Increasing laser densities (2% vs 4% vs 6%) and the number of pretreatment passes (1, 2, and 3 passes) did not enhance PPIX fluorescence. Local skin reactions were most intensified in AFXL-pretreated skin and correlated with PPIX fluorescence and degree of PPIX photobleaching.

Conclusions and Relevance  Under standardized conditions, PPIX accumulation was most enhanced after AFXL pretreatment, followed by microdermabrasion, microneedling, and curettage. Increasing the number of pretreatment passes and laser densities did not further augment PPIX accumulation. These results may indicate relatively enhanced PDT response by AFXL pretreatment in diseased skin.

Trial Registration  clinicaltrials.gov Identifier: NCT02372370

Physical pretreatment of the skin facilitates local uptake of photosensitizing agents and is recommended to obtain optimal outcomes for treatment of actinic keratoses (AKs) in photodynamic therapy (PDT).¹ Removal of scales, crusts, and hyperkeratotic tissue is, therefore, part of the approved PDT procedure in Europe.^1,2 The most common pretreatment technique is curettage.^3-6 However, even after surface debridement, moderate and thick AKs still respond to PDT with lower cure rates than nonhyperkeratotic AKs,^3-6 which may be owing to limited penetration of photosensitizing agents and suboptimal protoporphyrin IX (PPIX) accumulation.^2,7,8 In recent years, new physical methods, such as the ablative fractional laser (AFXL), non-AFXL, microdermabrasion, and microneedling, have been introduced to impair the stratum corneum (SC) barrier and to increase the uptake of photosensitizer.^8-19 However, further studies are needed to examine and directly compare the ability of physical pretreatments to prime the skin for PDT. Each technique affects the skin barrier differently; curettage and microdermabrasion cause varying degrees of SC removal in an operator-dependent manner,^1,11,17 whereas microneedling creates a uniform pattern of microchannels that penetrate the SC and enter the underlying tissue.^13,18,19 The AFXL forms microscopic vertical holes of ablated tissue, each surrounded by a thin layer of coagulated tissue,¹⁵ and non-AFXL induces vertical microscopic treatment zones of coagulated tissue with minor epidermal effect.^20,21 However, the relative effect of AFXL, curettage, microdermabrasion, microneedling, and non-AFXL pretreatment in inducing PPIX production in human skin is unknown. In this study, we aimed to assess the relative potential of skin preparation with curettage, microdermabrasion with abrasive pads, microneedling, AFXL, and non-AFXL to enhance methyl aminolevulinate hydrochloride–induced PPIX fluorescence and to assess PDT reactions in normal skin.

Twelve healthy volunteers underwent assessment for eligibility and subsequent treatment at the Department of Dermatology, Bispebjerg Hospital, Copenhagen, Denmark (Figure 1). Inclusion criteria consisted of 18 years or older, Fitzpatrick skin types I to III, and normal skin in the study areas. Exclusion criteria consisted of lactation or pregnancy among women, allergy to any constituents of the methyl aminolevulinate cream, porphyria, topical treatment, recent sun burn, tattoos or moles within the test areas, history of keloid formation, and conditions associated with a risk for poor adherence. The study protocol (available in Supplement 1) was approved by the Danish Health and Medicines Authority and the Ethics Committee of Region Hovedstaden. The study was conducted in accordance with the guidelines for good clinical practice and the Declaration of Helsinki.²² Written and oral informed consent was obtained from all participants.

The study was designed as an intraindividual, randomized clinical trial. In each participant, 17 test areas of 2 × 4 cm were mapped on the upper back and exposed to a panel of standardized physical interventions (Table). Thus, test areas represented 5 columns with 3 rows in each, as well as untreated and methyl aminolevulinate–treated control areas. Based on computer-generated random sequences, we randomized (1) type of pretreatment to be applied in each column (curettage, microdermabrasion, microneedling, AFXL, and non-AFXL) and (2) the intensity of the individual pretreatments in each row (1, 2, and 3 passes and 2%, 4%, and 6% laser density). Treatment allocations were selected from opaque, sequentially numbered, sealed envelopes. Follow-up visits were performed at days 1, 3, and 7 after PDT.

Curettage was performed with disposable 7-mm ring curettes (Acu-Dispo-Curette; Acuderm Inc); microdermabrasion was performed with abrasive pads (grain diameter, 58.5 μm) (2121M P240; Ambu A/S). A 0.2-mm dermaroller system was used for microneedling procedures (DRS20 [0.02 mm]; Derma Roller System, Ltd). We applied AFXL with a 10 600-nm fractional carbon dioxide laser system using 30 mJ/microbeam and a 1-millisecond pulse duration at densities of 2%, 4%, and 6% (DotScan; GME German Medical Engineering GmbH). We applied the non-AFXL with a 1540-nm erbium:glass laser delivering 26 mJ/microbeam and a pulse duration of 15 milliseconds (15-mm XF Microlens; Icon Aesthetic System, Palomar Medical Technologies). Each pretreatment preparation took approximately 30 seconds to complete.

After pretreatment interventions, 125 μL of methyl aminolevulinate hydrochloride cream, 16% (160 mg/g), (Metvix;Galderma Nordic) was applied on pretreated and nonpretreated test areas and occluded for 3 hours. Subsequently, all test areas were illuminated with red light-emitting diode (LED) light (632 nm, 68 mW/cm²) (Aktilite CL128; Galderma) at a total light dose of 37 J/cm². The PPIX photographs were obtained before and after LED illumination and after 1 and 2 hours of occlusion, during which methyl aminolevulinate cream was wiped off and followed by reapplication of 50 μL of methyl aminolevulinate cream per treatment area. We quantified PPIX photobleaching as the light-induced reduction in PPIX fluorescence intensity.

Standardization was performed histologically by adjusting fractional depths using a validated ex vivo pig skin model²³ and by relating it to the number of channels created per square centimeter. Ablative fractional laser at densities of 2% delivered 25 channels/cm² compared with 54 channels/cm² per single microneedling pass and 115 channels/cm² from a single non-AFXL exposure (Table). Pretreatments were characterized from vertically sectioned histologic slides as presented in Figure 2 and in eFigure 1 in Supplement 2. After each intervention, punch biopsy specimens were collected, cut into 10-µm frozen horizontal sections, and stained with hematoxylin-eosin. Skin samples exposed to 1 treatment pass and AFXL of 15% density were analyzed under a calibrated microscope from 9 images for each intervention. Pretreatments were performed by the same investigator (C.B.) and were standardized by applying 1 pass in a horizontal direction, 2 passes in a vertical direction, and 3 passes in a diagonal direction on the skin.

The primary outcome measure was PPIX fluorescence accumulation. Secondary outcome measures were PPIX photobleaching and clinical local skin reactions (LSRs) supported by noninvasive reflectance measurements of the percentage of skin redness, transepidermal water loss (TEWL), and participant assessment of pain.

Protoporphyrin IX fluorescence and PPIX photobleaching (in arbitrary units [AU]) were quantified by fluorescence photography (Medeikonos PDD/PDT, Medeikonos AB) as previously described.²⁴ All photographs were adjusted for autofluorescence by subtracting the baseline value from each specific PPIX fluorescence measure and calibrated with a fluorescence standard (F322; Bioscience).

Local skin reactions were assessed on-site by the same investigator (C.B.) and scored in all participants on a clinical scale from 0 to 6 in which 0 indicated no skin reactions; 1, erythema with poorly defined demarcated borders; 2, sharply demarcated erythema; 3, well-defined erythema with (partial) palpable edema; 4, sharply demarcated erythema with obvious palpable edema; 5, sharply demarcated strong erythema with edema above skin level; and 6, sharply demarcated strong erythema with edema extending beyond the treatment area. Clinical assessments were unblinded. To ensure consistency in evaluations, a blinded, trained dermatologist (K.T.-B.) performed random, on-site quality control assessments of the unblinded assessor. Clinical photographs were obtained for photodocumentation.

Reflectance spectroscopy objectively assessed the percentage of redness with a reflectance meter (UV Optimize Scientific 558; Chromo-Light). Measurements of water loss through the skin were used as a measure for skin barrier integrity and subsequent healing after PDT, whereby the more intact the skin barrier, the lower the TEWL. We expressed TEWL in grams per hour per square meter and quantified it with a TEWL apparatus (Cortex Technology ApS). Participants assessed pain intensity during each pretreatment intervention and overall pain sensation during LED illumination using a numerical rating scale from 0 to 10 in which 0 indicates no pain and 10, the worst imaginable pain.

Analysis was based on intention to treat. Nonparametric analyses were used because the Kolmogorov-Smirnov test showed nonnormally distributed outcome measures. Descriptive data were presented as medians with minimum and maximum ranges. A univariate general linear model was performed testing intervention effect (type of pretreatment and number of passes) vs outcome measures (PPIX fluorescence, photobleaching, LSRs, redness, TEWL, and pain). Posttest Bonferroni correction was applied to avoid bias from multiple testing. We used the Spearman rank correlation to assess correlations between individual outcome measures. P < .05 was considered significant. All statistics were performed using SPSS software (version 22.0; SPSS, Inc).

A total of 12 healthy participants (8 men; 4 women; mean [SD] age, 33 [15] years) were included. All participants completed the study protocol and underwent analysis for outcome measures.

Physical interventions exerted different effects on the skin. Curettage resulted in almost full-thickness removal of SC with localized areas of intact SC, whereas microdermabrasion disrupted SC more superficially (Figure 2 and eFigure 1 in Supplement 2). Fractional penetration depths were similar after microneedling, AFXL, and non-AFXL, reaching the superficial dermis (median, 125 μm; P = .16) (Table). Microneedling generated narrow channels without surrounding coagulation zones. The AFXL created vertical, open channels consisting of a central, ablated zone of 50 to 75 μm in width, surrounded by a 22- to 31-μm cuff of coagulated tissue. Laser-tissue interactions from non-AFXL exposure showed well-defined columns of dermal coagulation, 32 to 64 μm wide, with overlying subepidermal clefting, epidermal edema, and structurally intact SC.

Skin integrity was assessed by TEWL, which increased by AFXL interventions (P < .001). Increasing TEWL values were seen with AFXL densities of 6% (72.2 g/h/m²) vs 4% (62.4 g/h/m²; P < .006) and 4% vs 2% (40.1 g/h/m²; P < .001). Curettage, microdermabrasion, and microneedling did not alter TEWL rates (Table). Transepidermal water loss values immediately after physical interventions correlated with the severity of LSRs on days 1 to 7 (range, r = 0.486 to r = 0.615; P < .001).

Qualitative PPIX accumulation was evaluated on fluorescence photographs, revealing a homogeneous and intensified PPIX fluorescence in AFXL-pretreated skin, which contrasted with a nonuniform, speckled fluorescence formation after curettage, microdermabrasion, microneedling, and non-AFXL interventions. Methyl aminolevulinate–treated control areas had the lowest fluorescence intensities, as shown in Figure 3 and in eFigure 2A in Supplement 2. Quantitative PPIX fluorescence intensities increased during the 3 hours of incubation as shown in Figure 4A and were independent of number of the pretreatment passes and laser densities (P = .11) (Figure 4B and eTable 1 in Supplement 2). Significantly higher fluorescence intensities occurred in skin pretreated with AFXL (median, 8661 AU) compared with microdermabrasion (median, 6731 AU), microneedling (median, 5609 AU), and curettage (median, 4765 AU) (P < .001), among which similar fluorescence enhancement was shown compared with methyl aminolevulinate–treated controls (median, 2254 AU; P < .03) and untreated controls (median, 239 AU; P < .001). Non-AFXL (median, 2898 AU) did not alter fluorescence in comparison with methyl aminolevulinate–treated controls (median, 2254 AU; P > .99).

Furthermore, higher PPIX photobleaching was found after AFXL (median, 7838 AU; P < .05), followed by microdermabrasion (median, 5821 AU), microneedling (median, 5105 AU), and curettage (median, 4654 AU) (Figure 4C). Non-AFXL did not influence photobleaching (median, 2460 AU) compared with methyl aminolevulinate–treated controls (median, 2039 AU; P > .99) (Figure 4C). The degree of photobleaching was independent of the number of pretreatment passes and laser densities (eTable 2 in Supplement 2) (P = .72). Formation of PPIX fluorescence and photobleaching correlated with the immediate TEWL values after skin pretreatments (range, r = 0.469 to r = 0.489; P < .001).

Physical interventions induced mild to moderate erythema and, furthermore, edema was seen after laser interventions. No oozing or bleeding occurred from 1, 2, or 3 passes with curettage, microdermabrasion, microneedling, AFXL, or non-AFXL. Pretreatment-induced LSRs were unaffected by increasing the number of passes and the laser densities (eFigure 2 in Supplement 2) (P = .07). Immediate skin reactions appeared most intensely after non-AFXL (median, 4; range, 1-5) owing to formation of edema, followed by AFXL pretreatment (median, 2; range, 1-3; P < .004). Similar skin reactions occurred after curettage (median, 2; range, 1-2), microneedling (median, 1; range, 0-2), and microdermabrasion (median, 1; range, 0-2) (P > .99). Reflectance measurements showed the highest median percentages of redness after AFXL (44.7%), non-AFXL (40.5%), and curettage (40.1%) compared with microneedling (36.4%) and microdermabrasion (36.0%) (P < .03) and with untreated control skin (27.6%) (P < .001).

Skin reactions were accentuated after PDT, and the most intense LSR appeared in AFXL-treated skin (median, 6; range, 4-6; P < .002). Curettage (median, 5; range, 4-6), microdermabrasion (median, 5; range, 3-6), microneedling (median, 5; range, 2-6), and non-AFXL (median, 4; range, 3-5) induced similar LSRs (P > .05); all pretreatments induced enhanced LSR compared with methyl aminolevulinate–treated controls (median, 4; range, 1-4; P < .001). At 24 hours after PDT, LSRs continued to be more severe in AFXL-treated skin than after any other intervention (P < .001), persisting through day 3 (eFigure 2B in Supplement 2) and day 7 (P < .001). Severity of LSRs correlated positively with preceding degrees of PPIX fluorescence formation and PPIX photobleaching after LED exposure (immediately after PDT to day 7; range, r = 0.565 to r = 0.669; P < .001). Objectively measured percentages of redness also correlated with PPIX fluorescence and PPIX photobleaching (days 1 to 7; range, r = 0.422 to r = 0.681; P < .001).

Participants reported transient mild to moderate pain during AFXL (median, 3; range, 1-9) vs non-AFXL (median, 3; range, 0-6; P = .08) pretreatment, compared with no or slight pain during microdermabrasion (median, 0; range, 0-2), microneedling (median, 0; range, 0-2), and curettage (median, 0; range, 0-8) (P < .01) (Table). Pain scores during physical interventions were not associated with the number of interventional passes or with laser densities (P = .39), but correlated with the severity of the acute LSRs (r = 0.571) and TEWL values right after pretreatment procedures (r = 0.502; P < .001). The median overall pain score during LED illumination was 5 (range, 1-9) for all participants.

This head-to-head comparison is the first, to our knowledge, of the relative effects of curettage, microdermabrasion, microneedling, AFXL, and non-AFXL pretreatments for PDT using a standardized treatment protocol. We demonstrated that AFXL, microdermabrasion, microneedling, and curettage enhance PPIX accumulation and PDT reactions in normal skin. Pretreatment with AFXL ensured the highest and most homogeneous PPIX fluorescence, and a similar PPIX-enhancing potential was found for curettage, microdermabrasion, and microneedling. Increasing the number of pretreatment passes and laser densities did not influence PPIX fluorescence, photobleaching, PDT reactions, or pain scores.

In accordance with PDT guidelines in Europe, skin pretreatment is recommended before application of topical photosensitizers to remove hyperkeratoses, enhance photosensitizer uptake, and improve treatment efficacy.¹ Several physical pretreatment techniques are available. The most commonly used is curettage,^3-6 although the benefit of curettage to enhance treatment outcome is inconsistent in the literature.^17,25 Microdermabrasion represents an alternative method to reduce SC thickness, but no trials have so far assessed the potential of using abrasive pads. Perforating the skin in shallow, vertical fractions with microneedling has proved to intensify PPIX fluorescence,^12,13,19,26 but a split-face study in 10 patients¹⁴ found similar AK clearance from microneedling-assisted PDT (91%) compared with conventional PDT with curettage (86%). Pretreatment with the AFXL removes skin fractions in vertical ablation zones and has proved to enhance photosensitizer uptake,^15,16 significantly improving AK lesion response rates compared with conventional PDT with curettage.²⁷ Although investigated to a lesser degree, non-AFXL–induced vertical coagulation zones augmented aminolevulinic acid–induced PPIX fluorescence in normal skin, using pulse energies of 50 mJ/microbeam.¹⁰ We used the non-AFXL at 26 mJ/microbeam, which did not enhance PPIX fluorescence. Different pulse energies may explain the conflicting data from these 2 studies.

In our study, we confirmed by histologic findings that each physical preparation technique modulated skin integrity differently. Under standardized conditions, we surprisingly found that PPIX fluorescence intensities did not increase from applying an increasing number of passes (1-3) and increasing laser densities (2%-6%). In this normal skin model, 1 pretreatment pass was therefore sufficient to intensify PPIX fluorescence. Furthermore, we found similar PPIX-accelerating potentials of curettage and microdermabrasion despite varying extents of SC removal (eFigure 1 in Supplement 2). These results suggest that methyl aminolevulinate uptake is facilitated once the SC is disrupted. Pretreatment with the AFXL induced homogeneous and intense fluorescence in all participants but was associated with more intense PDT reactions. However, high PPIX fluorescence intensity may not necessarily translate into greater PDT efficacy. Recently, Nissen et al²⁸ suggested that PPIX accumulation must reach a certain threshold level to yield an optimal PDT effect. Exceeding this level, however, does not improve treatment efficacy but causes more adverse PDT reactions in patients with AKs. Moreover, we found that each channel per se did not increase PPIX fluorescence to the same extent. Thus, 1 pass of microneedling created approximately twice as many channels (54 channels/cm²) compared with AFXL 2% intensity (25 channels/cm²). Applying microneedling and AFXL for a similar number of channels (50 and 54 channels/cm²) showed a significantly larger PPIX-enhancing potential from AFXL, which may be explained by laser-induced loss of tissue volume, potentially serving as a reservoir for drug uptake (Figure 2). These findings indicate that channel dimensions may constitute an important determinant for PPIX accumulation. A recent study on normal human skin found similar PPIX fluorescence from 12 passes of 0.2-mm microneedling and 2.5% coverage of AFXL pretreatment (5 mJ/microbeam), suggesting that the number of channels may also be a determinant for PPIX accumulation.²⁶

Each pretreatment procedure is associated with advantages and disadvantages. Thus, curettage, microdermabrasion, and microneedling are easily accessible techniques associated with low pain scores and mild skin reactions, but are also highly operator dependent, which introduces an element of variability in the PDT regime. In contrast, AFXL and non-AFXL procedures require access to costly laser equipment, but as operator-independent techniques they have the advantage of inducing customizable and highly controllable laser-tissue interactions. Nevertheless, discomfort associated with laser pretreatment should be factored into the overall pain of the PDT procedure.

The major limitation of this study is that the interventions were performed unblinded on normal, healthy skin. The downside of increasing photosensitizer uptake in normal skin is an increased risk for adverse reactions and treatment-related pain, potentially limiting tolerance. For this reason, pretreatment regimens to increase methyl aminolevulinate uptake may be most appropriate for diseased skin with localized, thick AKs and for skin with severe photodamage and field cancerization. Therefore, which of the physical interventions has the greatest potential to induce the desired PDT reactions in specific skin diseases remains to be clarified. Finally, we used noninvasive surface fluorescence photography to measure PPIX accumulation at the skin surface, which did not allow quantification of PPIX accumulation in deeper skin layers. Still, surface evaluation of PPIX fluorescence is a recognized method to assess aminolevulinic acid– and methyl aminolevulinate–induced PPIX fluorescence. The major strengths of this study are the protocol standardization and the direct comparison of interventions.

Pretreatment with AFXL intensifies PPIX accumulation and PDT reactions in normal skin more than curettage, microdermabrasion, and microneedling. This finding raises perspectives for future improved PDT efficacy in diseased skin.

Corresponding Author: Christiane Bay, MD, Department of Dermatology D92, Bispebjerg Hospital, University of Copenhagen, Bispebjerg Bakke 23, DK-2400 Copenhagen NV, Denmark (christianebay@gmail.com).

Accepted for Publication: November 9, 2016.

Published Online: February 1, 2017. doi:10.1001/jamadermatol.2016.5268

Author Contributions: Drs Bay and Haedersdal had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Bay, Lerche, Togsverd-Bo, Haedersdal.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Bay, Haedersdal.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Bay, Philipsen.

Obtained funding: Haedersdal.

Administrative, technical, or material support: All authors.

Study supervision: Bay, Lerche, Ferrick, Togsverd-Bo, Haedersdal.

Conflict of Interest Disclosures: Drs Bay and Togsverd-Bo report receiving travel grants from Galderma. Dr Haedersdal reports receiving research funding from Galderma and LEO Pharma and loan of fractional laser equipment from Sciton and GME German Medical Engineering. No other disclosures were reported.

Funding/Support: The study was supported by a research grant from Galderma International.

Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Additional Contributions: Diana Høeg, Medical Laboratory Technician, Department of Dermatology, Bispebjerg Hospital, University of Copenhagen, Copenhagen, Denmark, performed histologic examinations. Nis Kentorp, Clinical Photographer, Department of Dermatology, Bispebjerg Hospital, University of Copenhagen, Copenhagen, Denmark, provided photographic documentation.