Skin and mucosal fluorescence diagnosis with different light sources

ARTICLE

Auteur(s) : Jurgita Liutkeviciute-Navickiene, Aleksandras Mordas, Laimute Rutkovskiene, Laima Bloznelyte-Plesniene

Laboratory of Laser and Photodynamic Treatment, Institute of Oncology, Vilnius University, Santariškių 1, Vilnius LT-08660, Lithuania

accepté le 11 Novembre 2008

The growing incidence of cutaneous and mucosal malignancies each year necessitates the development of new and more effective methods for the diagnosis of cancerous lesions, while assuring better treatment results and improving patient satisfaction [1]. Unlike those used formerly, which were only systemically-applicable haematoporphyrin derivates (HpD), the recently developed topical photosensitizers, 5-aminolevulinic acid (ALA) or its methyl ester (methyl aminolevulinate – MAL), induce photosensitizing porphyrins [2]. The targeted photosensitization by ALA or MAL-induced porphyrins or HpD of mucosal and skin cancers, particularly superficial and extensive lesions, including superficial basal cell carcinoma and Bowen’s disease, leads to a selective red fluorescence which can be demonstrated by Wood’s lamp or other appropriate light sources [3]. This technique may be useful to define lesion margins better and/or to detect multifocal recurrences earlier. Performed early, fluorescence diagnosis has been shown to be highly efficient for superficial non-melanoma skin cancer, despite the low level of invasiveness.

Porphyrin-enriched tumour tissue irradiation with a fluorescence excitation system leads to the emission of a pink-red fluorescence. This principle is used as a diagnostic procedure and is called fluorescence diagnosis (FD), also known as photodynamic diagnosis (PDD) [4]. The term PDD is not actually correct, since reactive oxygen species are not involved in fluorescence diagnosis techniques [5]. The aim of this study was to investigate the possibilities of FD, using different wavelengths of light sources in skin and mucosal lesion diagnostics.

All photosensitizers are excited in ultraviolet and visible parts of the spectrum. The illuminated field must have distinct boundaries with the use of light in a well-defined spectrum. Conventional light sources are divided into four subcategories (all of which have rather wide spectral ranges, with the first three covering the entire visible spectrum): incandescent lamps; high-pressure arc lamps; low-pressure arc lamps and light diodes. Incandescent lamps are essentially conventional light bulbs with a range from 400 nm to infrared, and a peak wavelength of emission, lambda max (nm) equal to 2897×2/T (k). Arc lamps contain a gas that conducts electricity at high temperatures: high-pressure arc lamps contain mercury or xenon, while low-pressure arc lamps contain fluorescent material, and we are familiar with their use in ordinary room lighting. Finally, light diodes are small semi-conductors with a narrow wavelength band of 20-50 nm, with no infrared emission. Their main disadvantage, according to Brancaleon et al., is the difficulty of focusing the light on the target tissue, because they produce a large spread of light [1]. In our opinion, that is true for photodynamic therapy, but for fluorescence diagnosis, light diodes serve very well.

In FD, porphyrin fluorescence is detected under irradiation with different lights (under blue or near – UV excitation in range 300-450 nm) [6]: a Wood light (370-405 nm) [5], xenon lamp (375-400 nm) [7], filtered xenon lamp (380-430 nm) [8], filtered mercury arc lamp (400 nm) [9], a krypton ion laser with a 405 nm wavelength [9], an incoherent light source that provided blue light in addition to white light [10], a 150 W mercury lamp, with a pass-band filter selecting the fluorescence excitation band 405 ± 15 nm [11], 370 nm [12], 410 nm [12, 13], a Blu-U light system that uses visible blue light [9], blue light at 380-440 nm [14], the violet excitation wavelength at 405 nm is used as it matches the PpIX fluorescence excitation peak (the Soret band) well [13, 15, 16], light diodes (401 nm) and others. We used light sources composed of seven light diodes in order to determine which wavelength suits best for diagnosing tumours with particular histological structures. Our light source produced light in the range of 378 to 426 nm with seven different waves.

Fluorescence detection is the basic principle of FD [17]. The radiation or the emission from the molecule can easily be demonstrated if the molecule concentration in stimulated conditions is very high, or if the velocity of the radiation-less deactivation is low in relation to the velocity of the radiation. The emission is termed fluorescence if the emission fades quickly (10^-9-10^-3 sec.) after light absorption. In patients with advanced or small recurrent skin and mucosal cancer, FD can improve the efficacy of the treatment [18]. These lesions may represent therapeutic and diagnostic challenges because of special subtypes, location, previous therapy or accompanying diseases. Fluorescence can help in effective detecting and delineating of neoplastic areas. Recently, the use of fluorescence diagnosis and photodynamic therapy has been proposed for the management of cancer [19-22]. In view of the easy accessibility of the mechanism to adequately screen and detect pre-malignant changes and early lesions in the upper aerodigestive tract and skin, FD is being tried as a diagnosis modality with the potential to bridge the gap between clinical examination and invasive biopsies [12]. In order to further enhance the tumour demarcation, exogenous sensitising agents can be administered. Several groups are carrying out research to develop FD methods for early detection of premalignant lesions in the upper aerodigestive tract, most of them using ALA-induced protoporphyrin IX (PpIX) as a photosensitizer. Aminolevulinic acid has been shown to be the drug with the most experimental and clinical use [4, 19]. No generalized photosensitivity has been reported following topical ALA application, and ALA-induced PpIX appears to be almost completely washed from the body within 24 h from its induction [15]. Topical ALA application does not provide prolonged generalized photosensitivity. The most used agent (in the past) was hematoporphyrin derivative (Photofrin), originally developed for photodynamic therapy (PDT) by intravenous administration. In therapeutic doses it exhibits an unwanted side effect of transient skin sensitisation, which lasts for at least 1-2 weeks. Low-dose injection can be used for tumour diagnostic purposes, provided that sensitive detection equipment is employed [13]. Hematoporphyrin derivative was the first photosensitizer applied in diagnostics and treatment, though its use was limited due to systemic administration and consequent prolonged generalized photosensitivity [23, 24]. Currently it is not used for only diagnostic purposes. Topically active agents are preferable for PDT and FD, and to date most experience come from using ALA [6].

In our study we compared FD for tumours with different histological structures, using seven different wavelength light sources and both systemic and topical sensitizing agents.

Materials and methods

For the patients with advanced malignant disease, 2.5-5 mg/kg HpD (Photohem, Moscow, Russia) was injected i.v. and within 12 to 24 hours after the injection malignant lesions were illuminated with blue light for cancerous tissue detection. HpD in our clinic is used for the treatment both of advanced malignant and metastatic tumours. So it is permissible for treatment purposes. In parallel, for those patients, the skin and mucous malignant and benign lesion FD was applied, because such tissues are saturated with porphyrins. FD for patients with T1-2 was carried out within 2 to 8 hours (mostly 4 hours) after topical ALA - 20% cream of 5-aminolevulinc acid (MEDAC GmbH Hamburg, Germany) on Exsipiale basement application or within 3 hours after topical Metvix -16% cream of methyl aminolevulinate (Photocure ASA, Oslo, Norway) application. The cream was applied topically with 1 cm margins surrounding the lesion, and an occlusive dressing covering the cream. The sensitizer PpIX is synthesized to highly elevated levels by the haem cycle in the cells, due to the amounts of ALA or ALA-ME applied [25]. The lesions were not specifically prepared before performing fluorescent diagnostics. Before administering photodynamic therapy we usually remove scales and crusts by curettage or laser destruction. For the diagnostic procedure the lesions were not particularly prepared, in order to assess and differentiate the glowing of the lesions as if they were seen the first time.

As a fluorescence excitation system, we used the light system based on blue light emitting diodes which allows easy switching from conventional white-light mode to an ALA/MAL or HpD-induced violet-blue light (378-426 nm) mode (table 1).

The instrument was designed to provide multi-spectral (seven band) light. Its functional feature is its ability to switch between white light imaging and fluorescence imaging with multiple emission wavelength bands, using all of them together or separately. Diagnostic illumination usually lasted for 5 to 30 seconds. Two or three doctors separately assessed the fluorescence that was seen. Results were summarised and the average was calculated. At the beginning of the study, spectroscopic analysis was performed for ten lesions. That enabled us to compose and to use a visual scale reasonably. Finally, a digital camera was used for taking pictures of the fluorescing area. Different wavelengths were delivered in ascending order. That is, in the beginning, the lesion was illuminated with the light of the shortest wavelength (378 nm). When fluorescence was inspected and a picture was taken with a digital camera, that illumination was turned off and illumination with the light of longer wavelength (i.e. 389 nm) was started and so on. Such procedures lasted only a few seconds, so no decrease in fluorescence was seen. Of course, if illuminating with one of abovementioned lights for several minutes, fluorescence substantially decreases and disappears totally.

After the blue light inspection, biopsy samples from tumour foci were taken. The evaluated fluorescence data were compared with the cytological and/or histopathological tissue investigation.
Table 1 Characteristics of light diodes composing light source

Results

Red or red-pink fluorescence was observed in 92 malignant epithelial tumours; 58 of them fluorescent sharp, 32 – not so intensive, 2 malignant tumours – nasopharyngeal area chondrosarcomas – had no fluorescence (table 2).

The most intensive red fluorescence was detected in thin superficial malignant lesions (figure 1). In 9% cases, tumour foci were identified with the blue light in an area that initially appeared normal when examined with conventional light. All the tumour foci were carcinomas confirmed histologically (by biopsy).

There were multiple isles of actinic keratosis seen in the lesions on photodamaged skin, but healthy skin around them was not glowing and there was no difference between fluorescence of such skin and skin surrounding a solitary lesion.

From 268 benign lesions, very slight fluorescence was detected in a few haemangiomas and paratracheal papillomas, two foci of Darier diseases, one fragment of herpes zoster and some superficial open wounds with very intensive capillarity. Fluorescence of these benign lesions was different from malignant lesions – not so red, bluer and less intensive. Darier disease produced more intense fluorescence than other benign lesions, but these lesions were seen as very minute ill-defined scattered red fluorescenting dots, different from malignancies where the fluorescence is all over malignant tissues (figure 2).

We noticed fluorescence of the tongue mucosa in one in 5-6 patients after systemic application of photosensitizer. This occurred because of saprofitic bacterium. Chewing gum or lollipops induced very intensive, bright raspberry fluorescence of the centre part of the tongue as well. Naevi, papillomas, seborrhoeic keratoses and scars had no fluorescence. Patients did not note any subjective symptoms such as pain, itching, burning sensations or other events during the diagnostic procedure. These observations confirm a good efficacy and tolerance of FD in the cohort of cancer patients.

Performing FD with different wavelength light sources, it was found that for assessing malignant skin and mucosal cancer, the optimal wavelength is 401 nm. During illumination with light at such a wavelength, the brightest fluorescence of malignancies (a bright raspberry colour) is seen. Usage of that particular wavelength delineates the border between lesions and healthy tissue best. Intact skin and mucosa do not fluoresce and remain blue coloured. Light having a wavelength of 389 nm serves better for detecting benign lesions (table 3).

While illuminating with 426 nm light, dark green to grey, non-intense fluorescence was seen in the case of four patients. In areas producing fluorescence of such a colour, a focus of infiltrative carcinoma under the healthy skin was confirmed, rather than a superficial lesion.

We investigated the usefulness of ALA/MAL-induced porphyrin fluorescence in the pre-operative demarcation of ill-defined clinical tumour margins and as a control before and after PDT (figure 3). There was a strong correlation between clinical extension and the fluorescence pattern of the tumours. In addition, all fluorescent areas proved to be neoplastic by histopathologic examination.
Table 2 Fluorescence intensity according malignant tumour morphology

Discussion

In clinical practice, the first FD was made by Lipson in 1961. His team went on to study the potential localization of HpD in tumours in patients undergoing bronchoscopy or esophagoscopy for suspected malignant disease. Light of the appropriate wavelength to activate HpD (400 nm) was produced by a filtered mercury arc lamp and transmitted via a fiber optic cable to the endoscope. Tumour fluorescence was observed through a filter, which excluded reflected light from the mercury arc lamp. This study was the first to demonstrate that tumour localization and fluorescence had a potential clinical use in the detection of human tumours [9]. ALA also appears to have potential for clinical use as a FD agent in conditions that require histological surveillance but are not readily visible [9]. For that reason, all popular agents (HpD, ALA and MAL) were used in our study. The fluorescence produced by them does not differ.

Attention should be paid to lesions showing only moderate fluorescence in the mucosa, as this might already indicate the onset of carcinoma in situ [7]. The additional information we received from this method may be of great significance. Recent technological advances in light sources, high-sensitivity imaging detectors, and high-performance spectrographs, together with advances in digital imaging/processing, are enabling a wide variety of medical applications [11]. In our study we paid more attention to light sources used for fluorescence diagnosis, because in many FD studies, mainly photosensitizers are analysed and compared, paying less attention to light sources. It is known that porphyrins show five strong absorption peaks. Their absorption spectrum exhibits a maximum in the so-called Soret band, ranging from 360 to 400 nm. Maximum absorbance is around at 405 nm. This maximum is followed by 4 additional peaks with decreasing intensity between 500 and 635 nm (Q-Bands). For FD, light around 405 nm is used [16, 17]. For this reason the aim of our study was to analyse the fluorescence produced separately by different light sources having a wavelength with near-maximum or maximum absorbance. The fluorescence of skin and mucosal tumours with different histological structures was examined.

After analysing different wavelength light sources used for performing FD, it was found that the optimal wavelength is 401 nm. It is interesting to notice that authors who studied photosensitizers refer to 405 nm as the maximum absorbance peak, whereas the brightest fluorescence of skin and mucosal malignancies was seen under light with a wavelength of 401 nm in our study. Light having a wavelength of 405 nm produced fluorescence too, but not as bright as that of 401 nm. It is possible that in organisms photosensitizers have peak absorbance at 401 nm in skin and mucosal malignancies, but it needs more detailed investigation.

We had an interesting experience – green fluorescence was seen when infiltrative carcinoma was illuminated with light having a wavelength of 426 nm. As reviewed by Yang et al., green fluorescence has been seen, but in other tissues: connective tissue rich in collagen and elastin fibers, such as bone and dura, appeared green in the composite fluorescence image, as would be expected from the known fluorescence spectra of these molecules [11].

The use of FD allowed delineation of clinically ill-defined tumours and the detection of tumour relapses or new tumours that were not clinically detectable [5, 27]. FD is a simple diagnostic method, without doubt this technique has a practical role in everyday clinical practice. Despite some previously expressed thoughts from several authors [28] that it will not be a clinically useful method, there is clear evidence that sometimes it is impossible to manage without FD (e.g., to determine radicality of treatment after surgical operations and especially after PDT), also large cutaneous malignancies can present a diagnostic challenge. Fluorescence imaging is an attractive diagnostic technique for skin and mucosal tumour demarcation, with the potential to come more and more into clinical use.

The observations in this group of patients suggest that red fluorescence may correlate with the precise location of minimal (millimetre) residual malignant skin and mucosal tissue, the same findings were found by Yang et al. [11] studying brain tumours. Hence, fluorescence could be used to guide surgical resection or PDT and promote better tumour tissue removal in silent areas.

In vivo fluorescence can provide new tools for clinical oncology: screening and diagnosis of early-stage malignancy, defining tumour extent, and optimising localized treatment of solid tumours [11].

Conclusion

The most appropriate wavelength for FD is 401 nm in order to achieve complete visualization of malignant lesions after the topical or systemic application of a tumour selective photosensitizer. Malignant tumours were found to be fluorescent, whereas no fluorescence was observed in normal skin and mucosa. In the blue light mode, there is background blue fluorescence in normal tissue and red fluorescence in malignant areas. This method is applicable for detecting early superficial tumors, delineating their margins and managing follow-up after therapy.

It has been shown to be high effective in superficial malignant skin and mucosal lesion diagnostics. In a few cases very slight fluorescence was observed in benign lesions. FD may be used to optimise the detection of lesions in post-PDT patients, and also to guide tumour therapy.

Acknowledgements

References

2 Babilas P, Landthaler M, Szeimies RM. Photodynamic therapy in dermatology. Eur J Dermatol 2006; 16: 340-8.

3 Vereecken P, Da Costa CM, Steels E, De Lathouwer O, Heenen M, De Mey A. Fluorescence diagnosis of face-located basal cell carcinomas: a new dermatological procedure which may help the surgeon. Acta Chir Belg 2007; 107: 205-7.

4 Fritsch C, Ruzicka T. Fluorescence diagnosis and photodynamic therapy in dermatology from experimental state to clinic standard methods. J Environ Pathol Toxicol Oncol 2006; 25: 425-40.

5 Fritsch C, Goerz G, Ruzicka T. Photodynamic therapy in dermatology. Arch Dermatol 1998; 134: 207-14.

6 Scott MA, Hopper C, Sahota A, Springett R. Fluorescence photodiagnostics and photobleaching studies of cancerous lesions using ratio imaging and spectroscopic techniques. Laser Med Sci 2000; 15: 63-72.

7 Csanady M, Kiss JG, Kiss GJ, Laszlo I, Jori J. ALA (5-aminolevulinic acid) - induced protoporphyrin IX fluorescence in the endoscopic diagnostic and control of pharyngo-laryngeal cancer. Eur Arch Otorhinolaryngol 2004; 261: 262-6.

8 Ackermann G, Abels C, Karrer S, Bäumler W, Landthaler M, Szeimies RM. Fluorescence-assisted biopsy of basal cell carcinomas. Hautarzt 2000; 51: 920-4.

9 Ackroyd R, Kelty C, Brown N, Reed M. The history of photodetection and photodynamic therapy. Photochem Photobiol 2001; 74: 656-69.

10 Kriegmair M, Zaak D, Stepp H, et al. Transurethral resection and surveillance of bladder cancer supported by 5-aminolevulinic acid-induced fluorescence endoscopy. Eur Urol 1999; 36: 386-92.

11 Yang XD, Muller PJ, Herman P, Wilson BC. A multispectral fluorescence imaging system: design and initial clinical tests in intra-operative Photofrin-photodynamic therapy of brain tumors. Lasers Surg Med 2003; 32: 224-32.

12 Manjunath BK, Kurein J, Rao L, Murali Krishna C, Chidananda MS. Autofluorescence of oral tissue for optical pathology in oral malignancy. J Photochem Photobiol B 2004; 73: 49-58.

13 Wang I, Clemente LP, Pratas RM. Fluorescence diagnostics and kinetic studies in the head and neck region utilizing low-dose delta-aminolevulinic acid sensitization. Cancer Lett 1999; 135: 11-9.

14 Hillemanns P, Reiff J, Stepp H, Soergel P. Lymph node metastasis detection of ovarian cancer by porphyrin fluorescence photodetection: case report. Lasers Med Sci 2007; 22: 131-5.

15 Ericson MB, Sandberg C, Gudmundson F, Rosen A. Fluorescence contrast and threshold limit: implications for photodynamic diagnosis of basal cell carcinoma. J Photochem Photobiol B 2003; 69: 121-7.

16 Rydell R, Eker C, Andersson-Engels S, Krogdahl A, Wahlberg P, Svanberg K. Fluorescence investigations to classify malignant laryngeal lesions in vivo. Head Neck 2008; 30: 419-26.

17 Fritsch C, Ruzicka T. Fluorescence diagnosis and photodynamic therapy of skin diseases. New York: Springer Wien, 2003.

18 Kriegmair M, Baumgartner R, Knuchel R, Stepp H, Hofstader F, Hofstetter A. Detection of early bladder cancer by 5-aminolevulinic acid induced porphyrin fluorescence. J Urol 1996; 155: 105-10.

19 Liutkeviciute-Navickiene J, Bloznelyte-Plesniene L, Mordas A. Topical photodynamic therapy in skin and mucosal lesions. Acta Medica Lituanica Suppl 2002; 9: 73-6.

20 Gannon MJ, Brown SB. Photodynamic therapy and its applications in gynecology. Br J Obstet Gynaecol 1999; 106: 1246-54.

21 Gahlen J, Stern J, Laubach HH, Pietschmann M, Herfart C. Improving diagnostic staging laparoscopy using intraperitoneal lavage of aminolevulinic acid (ALA) for laparoscopic fluorescence diagnosis. Surgery 1999; 126: 469-73.

22 Chatterton K, Ray E, O’Brien TS. Fluorescence diagnosis of bladder cancer. Br J Nurs 2006; 15: 595-7.

23 Morton CA, Brown SB, Collins S. Guidelines for topical photodynamic therapy: report of a workshop of British Photodermatology Group. Br J Dermatol 2002; 146: 552-67.

24 Karrer S, Abels C, Szeimies RM, et al. Topical application of a first porphycene dye for photodynamic therapy-penetration studies in human perilesional skin and basal cell carcinoma. Arch Dermatol Res 1997; 289: 132-7.

25 Pålsson S, Gustafsson L, Bendsoe N, Soto Thompson M, Andersson-Engels S, Svanberg K. Kinetics of the superficial perfusion and temperature in connection with photodynamic therapy of basal cell carcinomas using esterified and non-esterified 5-aminolaevulinic acid. Br J Dermatol 2003; 148: 1179-88.

26 Berg K, Selbo PK, Weyergang A, et al. Porphyrin-related photosensitizers for cancer imaging and therapeutic applications. J Microsc 2005; 218: 133-47.

27 Stenquist B, Ericson MB, Strandeberg C, et al. Bispectral fluorescence imaging of aggressive basal cell carcinoma combined with histopathological mapping: a preliminary study indicating a possible adjunct to Mohs micrographic surgery. Br J Dermatol 2006; 154: 305-9.

28 Profio AE, Doiron DR. Dose measurement in photodynamic therapy of cancer. Laser Surg Med 1987; 7: 71-5.