Treatment of Experimental Choroidal Melanoma With an Nd:Yttrium-Lanthanum-Fluoride Laser at 1047 nm

Objective  To evaluate the effect of a new infrared laser in the destruction of pigmented choroidal melanomas.

Methods  B16F10 melanomas were implanted in the subchoroidal space of 64 rabbits(tumor height, 2.0-4.0 mm). Laser radiation from an Nd:yttrium-lanthanum-fluoride laser (1047 nm) was delivered as a focused (beam waist, 25 µm; irradiance, 100 kW/cm²) raster-scanned transpupillary beam. To investigate melanin heating, treatment with focused light was compared with collimated light (beam waist, 2 mm; irradiance, 16 W/cm²). Fine-wire thermocouples were implanted at the base of 3 tumors for in vivo temperature measurements. Untreated animals were used as controls.

Results  Of 64 animals, 27 received a single treatment with focused 1047-nm light. The rate of complete tumor eradication was 91% (10 of 11 animals) at a dosage of 125 J/cm² and 75% (9/12) at 63 J/cm² to 87 J/cm². The eradication rate dropped to 25% (1 of 4) at 38 J/cm² or less (P<.001). Continuous tumor growth was observed in all animals treated with collimated radiation and in untreated controls. Temperature measurements indicated that tissue heating at the tumor base was more rapid at 1047 nm than at 805 nm.

Conclusions  These data suggest that a single treatment with a focused, raster-scanned beam at 1047 nm may play a role in the destruction of pigmented choroidal melanoma. Focused irradiation at 1047 nm may provide more effective submillisecond heating of melanin than collimated irradiation, resulting in immediate photothermal disruption of tumor cells.

ALTHOUGH SEVERAL strategies exist for the treatment of choroidal melanoma, investigators continue to search for alternative approaches associated with less ocular morbidity. Thermotherapy with laser light at 810 nm is a new strategy under investigation for treatment of smaller choroidal melanomas.^1,2 In human studies, concern has been raised that intrascleral melanoma cells may not be destroyed, ³ and histologically intact tumor cells have actually been reported in the sclera after thermotherapy.⁴ In addition, recurrences may not become apparent until years after treatment, ⁵ and several sessions are usually required for tumor control.¹ Moreover, applicability of this method appears limited to tumors of minimal thickness.² Photodynamic therapy (PDT), on the other hand, destroys abnormal tissue by combining systemic application of a photosensitizer with light irradiation of appropriate wavelength, and PDT regimens have been tested in experimental choroidal melanomas in our laboratory with encouraging results.^6-11 Potential risks of PDT are drug toxicity and skin photosensitivity. The major concern with this modality is the extreme optical density of melanotic tissue, which limits the penetration of activating light to the base of thick tumors.

The present study investigated a treatment based on near-infrared laser light^12-14 using an Nd:yttrium-lanthanum-fluoride (Nd:YLF) laser emitting light at 1047 nm. The study was initiated after earlier murine studies with cutaneous melanoma¹³ and prior pilot studies of transpupillary therapy demonstrated feasibility of tumor eradication with the Nd:YLF laser at high fluences (1000 and 2000 J/cm²). Treatment with our Nd:YLF laser system differs from other laser-based treatments, such as PDT and thermotherapy at 810 nm, in a number of important ways. First, light at 1047 nm is highly focused, whereas conventional strategies of transpupillary thermotherapy and PDT use collimated laser irradiation. Focusing the light dramatically increases the effective irradiance of the applied light. Second, the beam of focused 1047-nm light is scanned across the tumor area, whereas the collimated beam of conventional thermotherapy and PDT is stationary. Scanning is necessary when focused laser beams are used. Third, the wavelength is 1047 nm, as opposed to a range between 600 nm and 810 nm for conventional thermotherapy and PDT. Reduced scattering and absorptivity at 1047 nm in comparison with light of shorter wavelength results in improved depth of penetration for optically dense tumor tissue and reduced collateral photodamage in nonpigmented tissue.¹⁵ Finally, no exogenous photosensitizer is used. The main goal of this study was to determine whether a single treatment with focused 1047-nm laser irradiation at different light doses is effective in the destruction of pigmented choroidal melanoma in a rabbit model.

All animals were treated in accordance with the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research, and the experiments were approved by the animal care committee of the Massachusetts Eye and Ear Infirmary, Boston.

B16F10 melanoma cells were passaged in vitro using RPMI medium (Gibco, Grand Island, NY) containing 15% fetal calf serum (Gibco). Ten to 14 days before subchoroidal implantation, 5 × 10⁶ melanoma cells in phosphate-buffered saline were inoculated subcutaneously into the flank region of C57BL/6 mice.

Sixty-four female New Zealand albino rabbits each weighing between 2.9 and 3.5 kg were immunosuppressed with daily intramuscular injections of cyclosporine A (Sandimmune; Novartis Pharmaceuticals Corp, East Hannover, NJ). A dosage of 20 mg/kg was initiated 3 days before tumor implantation and maintained for a total of 10 days; this dosage was reduced to 15 mg/kg daily for the remainder of the experiment until the animal was killed. Tumors were implanted as described previously.^16,17 In brief, the tumor-bearing C57BL/6 mouse was killed immediately before implantation. The tumor was dissected free, minced, and stored in cold 0.9% isotonic sodium chloride solution. The rabbits were anesthetized with an intramuscular injection of 50 mg/kg ketamine hydrochloride (Ketalar; Parke-Davis, Morris Plains, NJ) and 5 mg/kg xylazine hydrochloride (Rompun; Bayer, Shawnee Mission, Kan). Two 7-0 polyglactin traction sutures were placed at the limbus at the 3 and 9-o'clock positions to rotate the globe superiorly. An inferior conjunctival peritomy was performed, and a 2-mm sclerotomy using a No. 64 Beaver blade(Beaver Surgical Products, Waltham, Mass) was made 5 mm posterior from the limbus to expose the subchoroidal space. A curved 22-gauge spinal needle was used to deliver a 0.5-mm³ tumor fragment into the subchoroidal space. Tumor placement was confirmed with indirect ophthalmoscopy. In a standardized manner, all tumors were implanted approximately 2 disc diameters inferior to the optic nerve to facilitate visualization. The traction sutures were removed.

The tumors were evaluated by indirect ophthalmoscopy starting 5 days after implantation. Ophthalmoscopy, fluorescein angiography, red-free photography, and color fundus photography were performed immediately before as well as 24 hours, 72 hours, 1 week, 2 weeks, and 4 weeks after treatment. Additional follow-up examinations were performed whenever regrowth of the tumor was suspected. B-scan ultrasonography (Quantum 200; Siemens, Issaquah, Wash) was used to determine the tumor height. Tumors between 2.0-mm and 4.0-mm thick were eligible for the study, corresponding to 2- to 4-week follow-up after tumor implantation. The color fundus photographs were obtained with a fundus camera (TRC 50 VT; Topcon, Tokyo, Japan). For fluorescein angiography, 10% sodium fluorescein(Ak-Fluor; Akorn Inc, Abita Springs, La) was injected at 0.1 mg/kg into the marginal ear vein. Anesthesia was the same as for the surgical procedure. Eyes with multiple elevated tumors and tumors located in the fundus periphery were excluded from the study.

A diode-pumped Nd:YLF-laser (DPM-1000; Microlase Optical Systems Ltd, Glasgow, Scotland) emitting laser radiation at 1047 nm was used for tumor treatment. The laser system was provided and serviced by Provectus Pharmaceuticals, Inc (Knoxville, Tenn). The specific laser parameters are given in Table 1. The laser system was coupled to a standard ophthalmic slitlamp (Zeiss SL-130; Carl Zeiss Gmb H, Oberkochen, Germany) using a 7-mirror articulated arm (Oxid Corp, Farmington Hills, Mich) and confocal dual-axis laser scanner (Provectus Pharmaceuticals, Inc). The scanner system was mounted on the tonometer mount of the slitlamp and included a pair of computer controlled galvanometers (model 6210; Cambridge Technology, Inc, Cambridge, Mass), a beam expander, a focusing lens system, and a confocal beam splitter. Adjustment of the focusing lens system allowed the delivered beam waist to be varied between 25 µm and 2 mm at the retina. The beam waist (in air) was determined with the knife-edge technique.

The animals were anesthetized as described above, and the pupils were dilated with 2.5% phenylephrine (Bausch & Lomb, Tampa, Fla) and 1% tropicamide(Bausch & Lomb). The laser beam was delivered from the scanner to the retina via an 18-mm Peyman wide-field YAG capsulotomy lens (Ocular Instruments, Inc, Bellevue, Wash). The curved lens surface oriented towards the cornea was covered with hydroxypropyl methylcellulose (Gonak; Akorn, Inc, Buffalo Grove, Ill). The treatment beam was rapidly scanned in a stepwise mode using a preprogrammed rectangular raster pattern to cover the desired treatment area. During scanning, the beam was stepped at 40 kHz (25-microsecond dwell time per step; 2-Hz raster refresh rate for a 3 × 3-mm raster pattern) with both duration and overlap of each spot controlled by predefined computer settings. Typical treatment parameters are given in Table 1. The size of the treatment spot was calculated on the basis of the specified magnification of the contact lens.¹⁸ The rectangular raster pattern was set at 3 × 3 mm with reference to the retinal level. This size was considered to be optimal in pilot studies. Larger-sized squares were avoided to ensure that the laser beam was delivered through the center of the fundus contact lens without impinging on the pupil. Overlapping treatment areas were added sequentially until the whole area of the tumor was covered. To achieve thorough coverage of the whole tumor surface, an overlap of 0.5 mm was allowed between adjacent treatment areas. A 1-mm margin was treated to target microscopic growth of melanoma tissue beyond the ophthalmoscopically visible borders.

For the dose-finding work, tumor-bearing animals were randomized to either laser treatment or no treatment. Animals assigned to the treatment group received a single treatment with focused 1047-nm radiation at fluences of 125 J/cm², 87 J/cm², 63 J/cm², 38 J/cm², 25 J/cm², or 13 J/cm².

In addition, to elucidate the role of heating of melanin, treatment with focused light was compared with a comparable dose of collimated light. For collimated light treatment, the dual-axis scanner system was used to deliver a collimated beam (waist, 2 mm; area, 3.1 mm²) with the rest of the parameters identical to that used for focused laser treatment. Overlapping treatment areas were added sequentially until the whole area of the tumor was covered. To produce collimated 805-nm irradiation, a diode laser (F-79-1000-100; Coherent Semiconductor Group, Santa Clara, Calif) was coupled into the same slitlamp and confocal laser scanner previously used with the Nd:YLF laser(Table 1).

All animals were followed for at least 4 weeks after laser treatment, except animals with documented massive tumor growth or regrowth. Additional animals were used for histopathological and temperature measurement analysis.

A precision fine-wire thermocouple of 25-µm diameter (CHAL-001; Omega Engineering, Inc, Stamford, Conn) was connected to a high-speed amplifier(SR 570, sensitivity 1 µV/A at 10 kΩ input impedance; Stanford Research Systems, Sunnyvale, Calif) and digitizing oscilloscope (TDS-210; Tektronix, Beaverton, Ore) for data capture and analysis. The temporal resolution of this thermal measurement system was approximately 1 millisecond, permitting a comparison of tissue heating at 1047 nm and 805 nm (the latter wavelength was used in several animals to provide reference to conventional transpupillary thermotherapy). Treatment at 1047 nm (focused and collimated) and 805 nm (collimated) was performed as described at an extended treatment interval of 90 seconds corresponding to a light dose of 500 J/cm². All other parameters, including instantaneous irradiance, were the same as for the other animals(Table 1). Thermal response of the measurement system was calibrated at 5 points using water baths of different temperatures (41°C-86°C).

To implant the thermocouple at the tumor base, the rabbits were anesthetized as for the operation. The eye was rotated to expose the conjunctiva and sclera overlying the choroidal melanoma. The conjunctiva was removed from the sclera in the region of the melanoma. The extension of the tumor was determined by transillumination and marked with a pen on the scleral surface. The eye was penetrated at a flat angle next to the tumor with a sharp 27-gauge needle. The needle tip was then carefully moved through the base of the tumor parallel to the underlying retina and out of the eye by creating another penetration. The wire thermocouple was threaded into the needle, and the temperature-sensitive point of the wire was moved to the tip of the needle. The distance of the 2 penetrations was measured, and the needle was moved backward by half this distance, with the thermocouple held in a constant position relative to the needle. Then, the needle was completely removed, with the thermocouple left in an unchanged position relative to the eye. This procedure allowed placement of the thermocouple in the center of the tumor base. No intratumoral or choroidal hemorrhage was detected by ophthalmoscopy. The temperature was measured before, during, and after laser treatment. Subsequent treatments at different wavelengths were performed in the same tumor to eliminate variability of thermocouple positions in different tumors. The measurements were repeated with reversed order of the wavelengths. The probe was placed at the base of the tumor because this area (distant from the laser and close to the choroid) is critical for tumor regrowth.¹⁹

After the animals were killed, the eyes were enucleated, fixed in 10% buffered formaldehyde solution (Formalde-Fresh; Fisher Chemical, Fairlawn, NJ), and processed for methacrylate sectioning. The sections were stained with hematoxylin-eosin and evaluated by light microscopy (BHS System; Olympus Optical Co, Tokyo).

Continuous variables are expressed as mean ± SD. Discrete data are expressed as frequencies. The χ² test was used for discrete variables. P<.05 was considered statistically significant.

A total of 64 animals, each with a melanoma lesion, were included in the study. Of these, 27 animals were treated with 1047 nm of focused irradiation at different light doses (Table 2). The mean tumor thickness was 3.0 ± 0.6 mm at the time of tumor treatment. The rate of tumor eradication following a single treatment was 91% (10/11) for animals treated at 125 J/cm², 75% (3/4) for animals treated at 87 J/cm², and 75% (6/8) for animals treated at 63 J/cm². The eradication rate dropped to 25% in average (1/4) for animals treated at a 38 J/cm² or lower dose. This decrease was statistically significant when compared with the overall eradication rate of animals treated at a 63-J/cm² or higher dose (P<.001). In addition, 4 animals were treated with collimated 1047-nm radiation at 125 J/cm². The mean tumor thickness in this group was 3.1 ± 0.5 mm at the time of treatment. In marked contrast to treatment with focused irradiation at 125 J/cm², no tumor arrest was achieved in this group (P = .002). Another 20 of 64 animals did not receive any laser treatment. All untreated tumors showed continuous growth beyond 12 mm in height. No lung or liver metastases were found in any of the treatment groups, although lung metastases were detected in 1 untreated animal.

Tumors with complete regression after laser treatment showed a typical course. Upon application of focused 1047-nm irradiation, bleaching within 3 seconds after the start of laser treatment was noted. Occasionally, small retinal hemorrhages were observed during or within minutes after treatment. No initial changes were detected in the surrounding choroid. Maximum surrounding subretinal exudation was found several hours after treatment. Exudation decreased within 3 days after treatment, and all subretinal fluid was completely absorbed within 7 days after treatment. As the subretinal fluid resolved gradually, reductions of the tumor dimensions and vascular occlusion of the adjacent choroid were evident. Later, fibrotic pigmented chorioretinal scars evolved, surrounded by choroidal atrophy. No retinal traction or optic nerve atrophy was noted. A typical time course is given in Figure 1 and Figure 2.On fluorescein angiography, untreated tumors showed patchy hyperfluorescence increasing with time as a result of leaking intratumoral vessels. Fluorescein angiography performed 3 days after treatment indicated hypofluorescence of the tumor with some leakage at the margins (Figure 1 and Figure 2).Further follow-up revealed disappearance of the marginal leakage. For eyes with complete tumor regression, no difference between different fluences was noted on ophthalmoscopy and fluorescein angiography. No posttreatment vitreous haze as a sign of intraocular inflammation was noted.

Continuous growth of tumors was evident by pigmentation in the treated area followed by persistent pigmentation through the treated margins. As the subretinal fluid overlying the tumor was absorbed, the increase in thickness was detectable by indirect ophthalmoscopy. On fluorescein angiography, areas of growth showed presence of tumor vessels associated with diffuse leakage.

In contrast to focused radiation, tumors treated with collimated 1047-nm laser radiation did not show regression by ophthalmoscopy or fluorescein angiography. No bleaching of these tumors was observed during treatment, whereas all tumors treated with focused irradiation showed substantial intense whitening during treatment.

Another 3 of 64 animals were used for in vivo temperature measurements. Two tumors (height, 0.5 mm and 6.0 mm) were irradiated with collimated 1047-nm radiation, and, after a brief recovery interval, they were treated with focused 1047-nm light. Both for focused and collimated light, the temperature increases at the tumor base were 25°C to 28°C for the thin tumor and 10°C for the thick tumor. In each case, a maximum stable temperature was reached within 10 to 20 seconds after the start of laser irradiation. However, the thermal data for focused irradiation also exhibited 2 transient patterns superimposed on this gradual mean increase: a sawtooth-shaped repetitive increase of approximately 10°C with a periodicity of approximately 500 milliseconds and a similarly repetitive burst of rapid spikes of 5°C to 15°C (each with a duration on the order of 1 millisecond). None of these patterns were observed with collimated irradiation.

A third tumor (height, 2.0 mm) was first irradiated with 1047-nm collimated light. Measurements at the base of the tumor revealed a maximum temperature increase of 9°C. No bleaching of the tumor was observed. Subsequent measurements of the same tumor using collimated irradiation at 805 nm (with otherwise-unchanged experimental settings) revealed a similar temperature increase of 9°C. However, the maximum temperature was reached at different time intervals after the initiation of treatment. With collimated 1047-nm irradiation, the time interval was 15 seconds, whereas at 805 nm it was 60 seconds. The same results were observed for each tumor when the experiments were repeated with reversed order of laser treatments.

Ten of 64 animals were used for histopathological analysis. Histopathological specimens were obtained immediately and later after laser treatment at different light doses as well as for controls and animals treated with collimated light(Figure 3). Light microscopic evaluation of melanomas before treatment and of areas of regrowth of treated melanomas showed histologically intact tumor tissue (Figure 3A). The cells were polygonal with vesicular cell nuclei containing prominent nucleoli. Numerous mitotic figures were noted in the microscopic field. The tumors were vascular. Immediately after laser treatment, the intratumoral blood vessels were dilated and densely packed with swollen erythrocytes throughout the full thickness of the tumor (Figure 3B). Twenty-four hours and 3 days later, homogenous necrosis was seen throughout the treated lesion (Figure 3C). The tumor cells presented hyperchromatic nuclei and loss of cytoplasmic features. The dilated vascular structures showed disintegration. Two weeks after treatment and later, multiple islands of necrosis infiltrated with pigment-laden macrophages were observed (Figure 3D). No intact tumor cells or mitotic figures were found. No histopathological differences were detected between melanomas successfully treated at different light doses. Also, no histopathological differences were found between untreated tumors and those treated with collimated 1047-nm radiation.

The present study investigated a new strategy for treatment of experimental pigmented choroidal melanoma based on near-infrared laser light. AnNd:YLF laser emitting focused 1047-nm light was used. Our results show that a single treatment with a highly focused, raster-scanned laser beam at 1047 nm can obtain a 75% tumor eradication rate with total light doses of 63 J/cm² and 87 J/cm². Enhanced efficacy of 91% was observed for total light doses of 125 J/cm².

In marked contrast, no tumor arrest was achieved in any of the 4 tumors treated with collimated 1047-nm light at otherwise unchanged experimental conditions (ie, same total light dose). This result suggests that the increased instantaneous irradiance, associated with a tightly focused laser beam, may play a crucial role in tumor destruction. The instantaneous irradiance is defined as the total irradiation energy that falls per unit time on a unit area. In our experimental settings, this measure characterizes the amount of laser light that is incident upon a particular area of tissue during the period of time the treatment beam impinges on the area. Focusing the laser beam decreases the beam area from 3.1 mm² to 4.9 × 10⁴ mm² (a factor of 10⁴; Table 1), thereby increasing the instantaneous irradiance by the same factor (from 1.6 × 10 W/cm² to 1.0 × 10⁵ W/cm²; Table 1). The high instantaneous irradiance of the focused beam results in a transient, highly elevated tissue temperature in a given volume of tumor tissue while the rapidly moving focus is directed into that volume. Because the focused beam is only present in a particular focal zone for a brief interval (well below 1 millisecond), the high instantaneous irradiance produces a localized thermal spike in this zone.¹⁴ This spike in temperature may contribute to cellular photodisruption. In contrast, conventional transpupillary thermotherapy uses a stationary beam of lower instantaneous irradiance to achieve hyperthermia of tumor tissue.

Although the focus of the treatment beam was stepped at 40 kHz (ie, 25-microsecond dwell time per step) in our study, the temporal resolution of our thermocouple was limited to about 1 millisecond, making it impossible to detect any transient temperature spikes that occurred when the intense, tightly focused laser beam scanned across a particular volume of tumor tissue. However, these spikes caused distinct time-averaged thermal patterns that were superimposed on the general elevation of temperature and were observable with the thermocouple. Transient spikes (ie, 5°C-15°C, having a duration on the order of 1 millisecond) were indicative of heating of tissue in proximity to the thermocouple during each pass of the beam (ie, 1 spike per scan line) and occurred in a brief burst during each scanning period. In addition, a repetitive sawtooth-shaped pattern (approximately 10°C increase, with a periodicity of approximately 500 milliseconds) represented cumulative localized heating of tissue proximal to the thermocouple as the rastered treatment beam repeatedly passed through the immediate vicinity of the thermocouple during each 2-Hz scanning period; the peak of this longer-term phenomenon corresponded with the burst of transient spikes. These patterns were observed during treatment with focused light but were absent during treatment with collimated light. Otherwise, focused and collimated light at equivalent experimental conditions(ie, mean intensity and total light dose) yielded comparable gradual mean thermal increases of 25°C and 10°C at the base of tumors, 0.5 mm and 6 mm in thickness, respectively. For both focused and collimated light, equilibration of the thermal input with diffusion into the surrounding tissue explains why the gradual mean thermal pattern was similar.

In our study, tumor response to treatment with focused light was observed promptly after treatment for all tumors, including those treated with low light doses. Bleaching of tumors appeared nearly immediately upon initiation of treatment (within 3 seconds) with light doses as low as 13 J/cm².On fluorescein angiograms obtained immediately and 24 hours after treatment, hypofluorescence indicated tumor vascular closure. Tumor shrinkage was noted within 2 to 3 days after treatment.

In contrast, no angiographic or histopathological changes were observed upon treatment with collimated irradiation, and no tumor arrest was achieved in this group. In addition, no tumor bleaching was observed. These observations are consistent with data reported for similar treatment of murine cutaneous melanoma.¹³ Photodisruption occurring upon rapid localized heating¹⁵ may explain the immediately visible effects following focused irradiation.

Our results show that the spatial and temporal distribution of laser energy is essential for treatment efficacy. A relatively high instantaneous irradiance administered in a fractionated mode by repetitive, rapidly rastered, highly focused irradiation of small tissue volumes enhances tumor destruction in comparison with continuous collimated irradiation of equivalent total light dose.

Another advantage of the present method is the use of light at 1047-nm wavelength. Good tissue penetration is crucial for any light-based tumor treatment and has been questioned for conventional transpupillary thermotherapy at 810 nm. Histologically intact tumor cells were reported in the sclera of eyes in which tumor necrosis up to the inner scleral surface had been induced by transpupillary thermotherapy.⁴ Early satisfactory tumor regression after photocoagulation may be followed by recurrences that do not become apparent until years after treatment, eg, 14% recurrences on a mean of 6 years after treatment with xenon arc photocoagulation.⁵ The heat convection provided by the high blood flow in the choroid, coupled with attenuation of the treatment beam as it passes through tumor tissue, can be a limiting factor for treating the base of a tumor solely by hyperthermia.¹⁹ Our experimental results illustrate markedly faster temperature increases at the base of a tumor using 1047-nm irradiation (15 seconds) compared with 805-nm irradiation(60 seconds). These observations are in accordance with a better depth of penetration at 1047 nm. Also, better penetration of light at 1047 nm (and improved maintenance of focusing of such light within tissue) may in part explain why a single treatment with laser irradiation at 63 J/cm² or more was suitable for tumor control in 75% of tumors, with an efficacy of 91% observed for total light doses of 125 J/cm².

In conventional thermotherapy at 810 nm, several sessions are usually required for tumor control.¹ To compare the different groups, we did not perform multiple treatments in animals with tumor regrowth. Although the follow-up time is limited in our study, our past 9 years' experience with this tumor model is that the growth pattern associated with the pigmented B16F10 tumors is extremely aggressive. An inoculation of only 10⁶ melanoma cells in the subchoroidal space generates pigmented tumors within 1 week, and, thus, we expect microscopic amounts of residual viable tumor to become visible within 1 to 2 weeks of follow-up.¹⁶

In summary, a new treatment modality based on a prototype Nd:YLF laser was studied in an animal model of pigmented choroidal melanoma. Advantages include use of a 1047-nm wavelength and rapid scanning of the treatment area with a highly focused laser beam. Our experimental results illustrate markedly faster temperature increases at the base of a tumor using 1047-nm irradiation compared with 805-nm irradiation, indicating a better depth of light penetration at 1047 nm than for wavelengths of conventional transpupillary thermotherapy. Effective tumor eradication using high laser irradiance in a fractionated mode suggests that highly focused laser light at 1047 nm is superior to collimated laser light. Focused light may induce a more therapeutically effective and localized heating of melanotic tumor tissue by photodisruption.

Submitted for publication April 11, 2002; final revision received November 1, 2002; accepted November 21, 2002.

This study was supported by grant NIH EY 10975 from the National Institutes of Health, Bethesda, Md, the Massachusetts Lions Eye Research Fund, grants DFG Kr 1918/1, DFG Kr 1918/2 from the Deutsche Forschungsgemeinschaft, Bonn, Germany, and Provectus Pharmaceuticals, Inc.

We thank Eric A. Wachter, Provectus Pharmaceuticals, Inc, for designing, providing, and installing the prototype Nd:YLF laser system, including the raster-scanning device and slitlamp modifications, and for extensive consultation about laser and tissue physics.

Corresponding author and reprints: Lucy H. Y. Young, MD, PhD, Retina Service, Department of Ophthalmology, Massachusetts Eye and Ear Infirmary, Harvard Medical School, 243 Charles St, Boston, MA 02114 (e-mail: lhyoung@meei.harvard.edu).