In vivo fluorescence kinetics and photodynamic therapy efficacy of d-aminolevulinic acid-induced porphyrins in basal cell carcinomas and actinic keratoses; implications for optimization of photodynamic therapy

ARTICLE

In recent years several innovative therapies of cancer have been under investigation. Photodynamic therapy (PDT) is a new, promising treatment modality for malignancies, with potentially high effectiveness and low morbidity.

Photodynamic therapy refers to light activation of a tumor-localizing photosensitizer to generate highly reactive oxygen intermediates, causing tissue injury and necrosis by oxidizing essential cellular components. Topical PDT using delta-aminolevulinic acid (ALA) involves photosensitation with endogenous porphyrins and activation with visible light. ALA is a small molecule which easily penetrates the abnormal stratum corneum overlaying many skin tumors, while it hardly penetrates the normal skin. In situ conversion of ALA to protoporphyrin IX (PpIX), an extremely potent photosensitizer and fluorescence emitter, is accomplished in normal and neoplastic keratinocytes to a different degree, by enzymes in the heme pathway, resulting in selective accumulation in neoplastic tissue and tissue-specific phototoxic effects [1, 2].

The emission of light following absorption of incident photons by chromophores is termed fluorescence. PpIX accumulation is characterized by a bright orange-red fluorescence and can be visualized in the skin when illuminated with light of appropriate wavelengths. Since the PpIX accumulation is a gradual process, fluorescence kinetics could be used to determine the ideal time point post application to initiate the irradiation of the tumor.

In a previous study, we showed that the erythema inspection and quantification analysis, as a result of tissue-specific phototoxic effect with topical ALA-PDT, provides a reliable predictor of the therapeutic outcome [3]. The purpose of this study was to assess the clues for optimization of topical ALA-PDT conditions, to treat epithelial tumors such as basal cell carcinomas (BCC) and precancerous lesions such as actinic keratoses (AK), by determination of the accumulation efficiency of the photosensitizer versus time and the localization efficiency of the photosensitizer within the tumor versus time.

Materials and methods

Patients

After informed consent, 29 patients (21 men and 17 women, median age 72 years) with 37 histologically proven malignant and precancerous skin lesions (17 BCC and 20 AK), median diameter 1 cm, were included in this study. Of the 17 BCC (2 superficial, 7 nodular, 1 nodular-micronodular, 1 micronodular, 1 nodular-cystic, 1 pigmented and 4 metatypical- basal squamous), median diameter 0.9 cm, 13 were located on the face, 2 on the neck, 1 on the ear lobe and 1 on the trunk. Of the 20 AK (9 atrophic, 5 acantholytic, 3 hypertrophic and 3 bowenoid), 18 were on the face and scalp and 2 on the extremities. No lesion had been previously treated.

ALA-PDT procedure

After cleansing the whole area with sterile saline solution, a freshly prepared w/o cream containing 20% ALA (Sigma Co, St Louis) was applied with a margin of about 1cm beyond the skin lesion and kept under occlusion impervious to light for 3-6 hrs, followed by irradiation, without anesthesia or sedation. The irradiation was carried out with a xenon lamp 300W, fitted with bandpass filters selecting visible light. The exact moment of starting the irradiation was deduced according to the findings of peak fluorescence detection. In 13 BCC and 16 AK a session of PDT with light dose 30 J/cm² was performed. In a second group of patients, tumors with clinical and histological factors of poor responsiveness (large size and/or thickness, pigmentation, marked cellular atypia) were included; four BCC and three AK received 3 sessions of 60 J/cm^2, one monthly (total light dose 180 J/cm²), while 1 AK of an extremity received a total light dose of 240 J/cm².

Studies of fluorescence

The equipment used to image, measure and map the fluorescence intensity of PpIX in the skin, as well as the color change evaluation by imaging the light diffuse reflection, consisted of a thermoelectrically cooled charge coupled device (CCD) color camera, a light source, a system of optical filters, a frame grabber and a personal computer. The camera had a wide responsivity, ranging from UV to NIR (320-1,150 nm). The selection of imaging spectral band was performed with 6 optical filters adapted on the camera and controlled by specially developed software, which was also employed for the image calibration processing and analysis. The video signal obtained from the output of the camera was digitized by a video grabber, which has been installed on a Pentium personal computer 200 MHz, with 16 MB RAM and graphics accelerator with 4MB video RAM. The monitor' s display resolution capacity was 1,024 x 768 pixels, with color depth 24 bit (16.7 million colors). The tissue irradiation was carried out with a 300 W xenon lamp fitted with bandpass filters to excite PpIX fluorescence and white light respectively. The light was delivered to tissue via a ring fiber-optic bundle placed in the front of the camera and surrounding a zoom macro-lens.

The fluorescence images were recorded in a dark room and between the fluorescence recordings the area of interest was kept under occlusion. The power density of the light source was very low (approximately 0.5 mW/cm²) and the exposure time was very short, in order to avoid possible photobleaching of the photosensitizer. For the study of fluorescence kinetics, serial in vivo fluorescence images were captured in 9 patients (5BCC, 4AK), every 15 min for 7 hrs and every 2 hrs afterwards until 24 hrs. For the remaining patients, fluorescence was detected between 2-7 hrs after ALA application. For any image, the medians of the fluorescence intensity of the pixels which correspond to the lesion area were calculated, after correcting for background autofluorescence and the sequential integrated fluorescence signals were then plotted versus time. This correction was obtained by subtracting the pre-ALA administration medians of autofluorescence intensity. The maximal fluorescence intensity, as compared with the zero time baseline intensity, is defined as photosensitizer accumulation efficiency and the correlation of fluorescence spatial distribution with the area of the lesion assessed with visual light, is defined as photosensitizer localization efficiency. In order to better show the spatial and temporal variation of the PpIX fluorescence (accumulation and localization efficiency) we used pseudo-colors. The recorded alterations from the baseline are expressed on a color scale, increasing from black (0%) to white (30%). The detection of tumor tissue by imaging PpIX fluorescence with pseudo-colors is based on the fact that each image consists of multiple pixels with specific coordinates in the RGB color space [3].

The prerequisite for reproducible color measurements is the standardization of the imaging conditions, such as illumination and capturing conditions. In the case of diffuse reflection, calibration was performed against a white perfect diffuser (BaSO₄) and for fluorescence emission imaging against a white fluorescence standard (Spectralon). In both cases calibration against black was performed with the camera aperture closed. When the calibration procedure ended, the acquisition was performed with the stored calibration parameters, thus, all the real time displayed images were calibrated in the spectral and the intensity domain. The compensation for any spatial variation of the image intensity caused by the non-uniform transfer function of the optics or by the non uniform illumination of the subject was performed with shading correction.

Results

ALA-PDT

Seventeen BCC were treated. The overall cure response rate, without recurrence was 70.6%. The follow up ranged from 20-35 months (mean 24 months).

Twenty AK were treated with ALA-PDT. The overall cure rate was 85%, without recurrence and the follow up period was 26-48 months (mean 36 months) (Table I).

Fluorescence kinetics and spatial distribution

In vivo fluorescence kinetics evaluation over time showed that most of the skin lesions developed maximum fluorescence emission intensity (photosensitizer accumulation efficiency) between 4-6.5 hrs after ALA application. There were considerable variations in both the rate of increase and the maximum intensity of the fluorescence achieved, expressed in arbitrary units, between tumors of the same type. We, therefore, varied the beginning of the irradiation, adapting PDT planning to the patient lesion individualities.

Protoporphyrin IX fluorescence kinetics data of BCC and AK are presented in Figure 1. In principle, by fluorescence detection, BCC were characterized by high photosensitizer accumulation efficiency, while AK had much lower photosensitizer accumulation efficiency. The best localization efficiency of BCC was noted 3.5-5 hrs after ALA application and thereafter a progressive decrease of fluorescence intensity became evident on the tumor, as well as diffusion of fluorescence outside the margins of BCC (Figs. 2 and 3). Despite the persistence of satisfactory fluorescence intensity values until 14 hrs after ALA application (Fig. 1), irradiation starting was chosen to be early, 3.5-5 hrs, when high photosensitizer accumulation efficiency and maximal localization efficiency on the BCC was noted, ensuring the selectivity of the procedure (Figs. 1, 2 and 3).

In the case of AK, the fluorescence intensity after achieving the maximal value (median 5 hrs) declined gradually within the margins of AK (high photosensitizer localization efficiency) (Fig. 4). Irradiation, therefore was started based on the data of accumulation efficiency, (approximately 5 hrs after ALA application (Figs. 1 and 4).

High spatial variability of fluorescence intensity was recorded on the tumor surface, possibly attributed to the unequal distribution of atypical cells within the tumor or to existing local variations of the tissue optical properties. Twenty-four hours after topical ALA application hardly any fluorescence was detectable in tumors, nor in surrounding normal skin.

The fluorescence kinetics study demonstrated that the findings were reproducible.

Discussion

In the present study a high cure rate was proven in superficial skin tumors such as BCC and precancerous lesions, such as AK. Some previous clinical studies yielded high efficacy rates [1, 4, 5]. The perceived advantages of ALA-PDT, such as localized and short-term photosensitivity and rapid photo-degradation by light illumination have led this modality to be very useful for the treatment of superficial non-melanoma skin tumors. The success of treatment requires an optimal interplay among a number of several parameters, such as drug and light doses, treatment sessions, criteria of tumors and patient selection. Limited depth of light and/or ALA penetration significantly limits PDT for hypertrophic AK and thick nodular BCC and may offer remarkable risk of tumor recurrence in case of inadequate photosensitizer concentration. A variety of strategies have been used to improve the desired PpIX production such as iron chelators [6] and to enhance the effectiveness of PDT using different wavelengths and light doses [7, 8].

It becomes, however, clear that there is a need to monitor parameters other than delivered light dose and irradiance and to correlate these with ALA-PDT outcome. The determination of the time course of photosensitizer fluorescence in skin tumors is crucial for effective photodiagnosis and PDT.

Our imaging procedure is noninvasive, simple, reliable and reproducible and sets no limits on the size or the shape of the area under examination. Serial, in vivo and real-time measurements of fluorescence intensity are of particular advantage when an endogenous photosensitizer is used, which might be synthesized in various amounts over time in different tumors of the same type, depending on the metabolic activity of the target cell. This novel system facilitates further study of skin photobiology and ALA pharmacokinetics.

Some authors, based on the principle that the intensity of the emitted fluorescence is a function of the amount of the fluorochrome present, claimed that the fluorescence of skin lesions after ALA application, using a Wood's lamp or laser-induced fluorescence, provided an estimate for prediction of PDT [1, 4, 8]. However, no direct correlation was found between fluorescence intensity and clinical response by other authors [6, 9]. On the other hand, the optimal time for PDT could be deduced from the time-dependent concentration of PpIX, since fluorescence kinetics is a temporal process. In order to maximize PDT effectiveness after application of ALA, irradiation of the treatment area should occur at the time of sufficiently increased concentration of the endogenous photosensitizer within the lesion.

In our study, fluorescence emission evaluation was shown to be an accurate diagnostic tool for BCC and AK, since a strong fluorescence was observed in all lesions under examination. Wennberg et al. found a good correlation between the fluorescence imaging in vivo and histological mapping in 50% of BCC and a partial one in other 23%, four hours after ALA application [10]. Fritsch et al. compared the macroscopical fluorescence intensity after topical ALA-application by means of Wood's light, with the time-dependent tissue formation of porphyrins, in epithelial skin tumors and normal skin, and recommended irradiation 1-6 hrs after ALA-application [11]. However, awareness of fluorescence time course kinetics is considered critical to the interpretation of photosensitizer accumulation in skin tumors. Unfortunately, despite the encouraging results of ALA-induced fluorescence kinetic studies in animal models [9, 12, 13], little is known concerning the in vivo porphyrin synthesis and elimination in tumor and normal tissue after topical application of ALA, in humans. Differences in photosensitizer accumulation and localization efficiency and transient selectivity may be most likely attributed to a faster uptake or increased porphyrin production in tumor versus normal keratinocytes, due to a higher demand for heme biosynthesis, rather than a reduced activity of the enzyme ferrochelatase [14].

In vivo quantitative determination of the time course of the porphyrin fluorescence is an essential pre-condition for an effective PDT, offering valuable information about the optimal time delay for efficacious and selective PDT. Tope et al. studied fluorescence kinetics in BCC after oral administration of ALA and showed a full thickness PpIX accumulation in all BCC histological subtypes and maximal tumor: normal skin fluorescence ratios 1-3 hrs after ALA ingestion [15].

In our study, excitation of the photosensitizer was performed by light of 425 ± 10 nm, which penetrates well in depth of about 0.1-0.15 mm, including the normal epidermis and superficial dermis [16]. Therefore, an estimation of the fluorescence lateral extent in superficial skin lesions could be achieved only by means of the in vivo kinetic study, whereas deep situated tumors might not be accurately delineated.

Our finding (Figs. 2 and 3) that in BCC, contrary to AK, fluorescence diffuses outside the margins of the visible lesion as the time lapses might indicate a deeper tumor volume [15], or the presence of subclinical deeper seated tumor nests, or finally be due to neovascularization in and out of the tumor mass, to support the blood supply required for the increased metabolic demand of a skin cancer. In conclusion, our results suggest that the optimum irradiation time for BCC is approximately 3.5-5 hrs after ALA application and for AK 5 hrs, taking into account both the parameters of photosensitizer accumulation and localization efficiency (Figs. 1 to 4).

Established methods of treatment for BCC (excisional surgery and cryosurgery) and for AK (cryotherapy and chemotherapy) are known to yield very high cure response rates [17-19]. However, PDT is not invasive, has excellent cosmetic results and is suitable for the treatment of multiple and large lesions. It is particularly recommended for older patients or with poor compliance, as well as patients with many other health problems, where invasive methods are contraindicated.

We also emphasize the importance of further investigation in the field of ALA-induced fluorescence kinetics, in order to optimize treatment protocols.

Article accepted on 27/3/00

CONCLUSION

REFERENCES

1. Kennedy JC, Pottier RH, Pross DC. Photodynamic therapy with endogenous protoporphyrin IX: basic principles and present clinical experience. J Photochem Photobiol B Biol 1990; 6: 143-8.

2. Peng Q, Berg K, Moan J, Kongshaug M, Nesland JM. 5-Aminolevulinic acid-based photodynamic therapy: principles and experimental research. Photochem Photobiol 1997; 65(2): 235-51.

3. Tosca AD, Balas CJ, Stefanidou MP, Katsantonis JC, Georgiou SK, Tzardi MN. Photodynamic treatment of skin malignancies with aminolevulinic acid: emphasis in anatomical observations and in vivo erythema visual assessment. Dermatol Surg 1996; 22: 929-34.

4. Svanberg K, Andersson T, Killander D, Wang I, Stenram U, Anderson-Engels S, et al. Photodynamic therapy of non-melanoma malignant tumors of the skin using topical delta-aminolevulinic acid sensitization and laser irradiation. Br J Dermatol 1994; 130: 743-51.

5. Cairnduff E, Stringer MR, Hudson EJ, Ash DV, Brown SB. Superficial photodynamic therapy with topical 5-aminolevulinic acid for superficial primary and secondary skin cancer. Br J Cancer 1994; 69: 605-8.

6. Fijan S, Hönigsmann H, Ortel B. Photodynamic therapy of epithelial skin tumours using delta-aminolevulinic acid and desferrioxamine. Br J Dermatol 1995; 133: 282-8.

7. Szeimies RM, Abels C, Fritsch C, Karrer S, Steinbach P, Baumler W, et al. Wavelength dependency of photodynamic effects after sensitization with 5-aminolevulinic acid in vitro and in vivo. J Invest Dermatol 1995; 105: 672-7.

8. Jeffes EW, McCullough JL, Weinstein GD, Fergin PE, Nelson JS, Shull TF, et al. Photodynamic therapy of actinic keratosis with topical 5-aminolevulinic acid. A pilot dose-ranging study. Arch Dermatol 1997; 133: 727-32.

9. Der Veen van N, Leengoed van HLLM, Star WM. In vivo fluorescence kinetics and photodynamic therapy using 5-aminolevulinic acid-induced porphyrin: increased damage after multiple irradiations. Br J Cancer 1994; 70: 867-72.

10. Wennberg AM, Gudmundson F, Stenquist B, Ternesten A, Mölne L, Rosén A, Larkö O. In vivo detection of basal cell carcinoma using imaging spectroscopy. Acta Derm Venereol(Stockh) 1999; 79: 54-61.

11. Fritsch C, Lehmann P, Stahl W, Schulte KW, Blohm E, Lang K, Sies H, Ruzicka T. Optimum porphyrin accumulation in epithelial skin tumours and psoriatic lesions after topical application of delta-aminolaevulinic acid. Br J Cancer 1999; 79(9/10): 1603-8.

12. Abels C, Heil P, Dellian M, Kuhnle GEH, Baumgartner R, Goetz AE. In vivo kinetics spectra of 5-aminolevulinic acid-induced fluorescence in a amelanotic melanoma of the hamster. Br J Cancer 1994; 70: 826-33.

13. Der Veen van N, de Bruijn HS, Berg RJW, Star WM. Kinetics and localisation of PpIX fluorescence after topical and systemic ALA application, observed in skin tumours of UVB-treated mice. Br J Cancer 1996; 73: 925-30.

14. Langer S, Abels C, Botzlar A, Pahernik S, Rick K, Szeimies RM, Goetz AE. Active and higher intacellular uptake of 5-aminolevulinic acid in tumors may be inhibited by glycine. J Invest Dermatol 1999; 112: 723-8.

15. Tope WD, Ross EV, Kollias N, Martin A, Gillies R, Anderson RR. Protoporphyrin IX fluorescence induced in basal cell carcinoma by oral delta-aminolevulinic acid. Photochem Photobiol 1998; 67(2): 249-55.

16. Anderson RR, Parrish JA. The optics of human skin. J Invest Dermatol 1981; 77: 13-9.

17. Telfer NR, Golver GB, Bowers PW. Guidelines for the management of basal cell carcinoma. Br J Dermatol 1999; 141: 415-23.

18. Pearlman DL. Weekly pulse dosing: effective and confortable 5-fluorouracil treatment of multiple facial actinic keratoses. J Am Acad Dermatol 1991; 25: 665-7.

19. Lubritz RR, Smolewski SA. Cryosurgery rate of actinic keratoses. J Am Acad Dermatol 1982; 7: 631-2.