T ranspupillary Thermotherapy the Management of Choroidal Melanoma

Carol L. Shields, MD, Jerry A. Shields, MD, Patrick De Potter, MD, Sundeep Kheterpal, MD

• In

Background: Several methods of treatment have been used for choroidal mela­noma. The purpose of this report is to evaluate the effectiveness of transpupillary thermo­therapy using near-infrared radiation to treat choroidal melanoma.

Methods: A prospective study was conducted to evaluate the clinical features, treat­ment results, and complications of patients with choroidal melanoma who were treated with transpupillary thermotherapy delivered over one to four sessions and followed for at least a 6-month period. All treated tumors had either documentation of growth or ophthalmoscopic risk factors for future growth and/or metastasis.

Results: There were 17 patients with choroidal melanoma treated with transpupillary thermotherapy. The mean tumor size before treatment was 6.6 mm in base and 3.0 mm in thickness. The tumor margin was a mean of 2.3 mm from the optic disc and 2.7 mm from the foveola. Seven tumors (41 %) touched the optic disc margin and three (18%) were under the fovea. The tumor responded to treatment in all patients, with a decrease in thickness and resolution of associated subretinal fluid. At a minimum of 6 months of follow-up, the mean tumor thickness was 1.7 mm, and the tumor site was a residual chorioretinal scar with partial visibility of the sclera in all patients. Despite the proximity to the optic disc and foveola, the final visual acuity was the same or improved in ten eyes (59%) and decreased in seven (41 %). The improved vision was due to resolution of subfoveal fluid, whereas the decreased vision was primarily the result of treatment in the fovea with ultimate retinal vascular occlusion or preretinal traction. Although long­term follow-up is not yet available, there were no patients with tumor recurrence or tumor metastases.

Conclusions: This preliminary study demonstrates that transpupillary thermother­apy appears to be an effective treatment for selected small choroidal melanomas and may be a particularly useful modality for treating those tumors near the foveola and optic disc. Longer follow-up is necessary to assess for local recurrence and the impact of treatment on survival. Ophthalmology 1996;103:1642-1650

The goal of conservative management of choroidal mela­noma is to preserve the eye and salvage as much vision as possible.1,2 Conservative techniques include irradiation with plaque radiotherapy or charged particles, local tumor

resection, ablative laser photocoagulation, or hyperther­mia.2,3 Hyperthermia is a new technique that generally has been used in conjunction with plaque radiotherapy. The thermotherapy delivery systems have included ultra­sound, microwaves, localized current field, ferromagnetic thermoseeds, and infrared radiation.3

-2o These devices

have been designed to increase the temperature of the melanoma to a range of 42° to 44°C as an adjuvant treat-

Originally received: March 20, 1996, Revision accepted: June 10 1996.

From the Ocular Oncology Service, Wills Eye Hospital, Thomas Jeffer­son University, Philadelphia.

Support provided by the Eye Tumor Research Foundation, Philadelphia, Pennsylvania.

The authors do not have proprietary interest in any of the devices re­ferred to in this publication.

1642

Reprint requests to Carol L. Shields, MD, Ocular Oncology Service, Wills Eye Hospital, 900 Walnut St, Philadelphia, PA 19107.

Shields et al . Transpupillary Thermotherapy in the Management Choroidal Melanoma

ment to enhance the effects of radiotherapy and to mini­mize damage to the normal ocular tissue.

Nuijs-Beems and associates l? and Journee-de Korver

and associates l8-

2o have reported on their experience us­ing hyperthermia as a primary treatment in animals. More recently, Oosterhuis and associates,3 from the same group, reported favorable results in three patients with small cho­roidal melanoma treated only with thermotherapy. The authors used near-infrared radiation delivered through the dilated pupil and termed this method transpupillary ther­motherapy. To achieve tumor necrosis, the authors found that a temperature rise in the tumor ranging from 45° to 60°C was optimal. We have used the technique described by Oosterhuis and co-workers3 and present our prelimi­nary experience with primary transpupillary thermother­apy in 17 patients with posterior uveal melanoma who have 6 months or more of follow-up.

Patients and Methods

A prospective evaluation was performed on all patients with small choroidal melanoma treated with transpupil­lary thermotherapy alone on the Ocular Oncology Service at Wills Eye Hospital between January 1995 and January 1996. The criteria for treatment eligibility was based on prior reports and included selected patients with a primary choroidal melanoma measuring 12.0 mm or less in base and 5.5 mm or less in thickness, located posterior to the equator of the eye, and with evidence of documented growth or ophthalmoscopic risk factors for growth and! or metastasis.3

•21 Excluded were those patients with reti­

nal, optic nerve, vitreous, intrascleral, or extrascleral ex­tension, subretinal hemorrhage, or media opacity prohib­iting adequate visualization of the tumor.

The data collected for evaluation included patient age, race, and sex, and visual acuity at initial examination. The tumor data included largest tumor base size (millimeters measured by indirect ophthalmoscopy), greatest tumor thickness (millimeters measured by A-scan and B-scan ultrasonography), proximity of the tumor margin to the optic disc and foveola (millimeters measured by indirect ophthalmoscopy), meridional location (nasal, superona­sal, superior, superotemporal, temporal, inferotemporal, inferior, and inferonasal), pigmentation (0 = none, 1 = minimal, 2 = moderate, 3 = heavy), morphology (sessile, dome, mushroom-shaped), subretinal fluid (present, ab­sent), orange pigment (present, absent), drusen (present, absent), and photographic evidence for documented growth. At each follow-up period at 1 to 3 months, visual acuity and ultrasonographic tumor thickness were mea­sured. Fundus photography, fluorescein angiography, and indocyanine green angiography were selectively per­formed. Associated anterior segment and fundus findings were recorded. Specifically, the presence of cornea, iris, or lens damage; retinal vein obstruction; retinal artery obstruction; retinal hemorrhage; choroidal hemorrhage; vitreous hemorrhage; macular edema; surface wrinkling retinopathy; optic atrophy; neovascularization of the disc

or retina; retinal hole; or tumor break through Bruch membrane were recorded.

The treatment was delivered using a specially modified infrared diode laser at 810 nm, with an adjustable beam width of 1.2,2.0, and 3.0 mm as provided by Iris Medical Instruments (Mountain View, CA). The infrared delivery system was adapted to a Haag-Streit slit-lamp biomicro­scope (Koniz, Switzerland) and delivered through one of two contact lenses (panfunduscope, Rodenstock, Munich Germany, or Mainster standard laser lens, Ocular Instru­ments, Bellevue, W A).

Immediately before treatment, topical anesthetic 0.5% proparacaine hydrochloride was applied, and the pupil was dilated with 2.5% phenylephrine hydrochloride, 1 % tropicamide, and 1 % cyclopentolate hydrochloride. A ret­robulbar injection of 5 ml 2% carbocaine hydrochloride was performed into the muscle cone.

Treatment was initiated in one spot for 1 minute at a power setting of 300 mW. Continuous observation through the slit-lamp biomicroscope directed on the tumor was maintained at all times. The power was increased by 50-m W steps for 1 minute to achieve an endpoint of a light-gray appearance to the tumor at the completion of the treatment spot. If the color change was found early in the exposure or coagulation of the retinal vessels was observed, then the power setting was lowered. In general, a 3.0-mm spot size was used to cover large portions of the melanoma tissue. In instances where there was a need to avoid the optic disc or retinal vessels, a smaller spot size of 2.0 and 1.2 mm was applied. With diminution of the spot size (and concomitant increase in the power density), the power was decreased accordingly to achieve the endpoint. The initial spot was applied in a region of the tumor most distant from the optic disc and foveola. Thereafter, spots were delivered in overlapping conflu­ency, including 0.5 mm of clinically normal tissue around the margin of the tumor.

Follow-up examinations were performed at 1- to 3-month intervals. Clinical examination, fundus photogra­phy, and ocular ultrasonography were performed. The decision to perform additional treatment was based on the ophthalmoscopic appearance of the tumor as well as tumor thickness. If the tumor were judged completely regressed, then additional treatment was not performed. If the tumor were judged partially regressed, then further treatments were performed to reach the endpoint of a completely regressed chorioretinal scar, as has been de­scribed?

Results

Between January 1995 and January 1996, we treated 85 patients with choroidal melanoma using transpupillary thermotherapy. Ofthese 85 patients, 43 had transpupillary thermotherapy alone and 42 had transpupillary thermo­therapy as an adjuvant to plaque radiotherapy.

Seventeen of the 43 patients treated with trans pupillary thermotherapy alone had a minimum of 6 months of fol­low-up (mean, 7 months; range, 6-10 months) and were

1643

Ophthalmology Volume 103 , Number 10, October 1996

selected for this study. Of these 17 patients, the mean age at treatment was 57 years (range, 20-79 years; median, 60 years). All patients were white, and there were eight women and nine men.

Ten patients were asymptomatic, six had blurred vision and one had pain in the affected eye at the time of presen­tation. Visual acuity before treatment was 20/20 in six eyes (35%), 20/25 in four (24%), 20/30 in one (6%), 201 40 in one (6%), 20/50 in two (12%), 20/60 in none, 201 70 in none, 20/80 in one (6%), 20/100 in one (6%), and 20/200 in one (6%). The tumor meridian was nasal in two eyes (12%), superonasal in none, superior in five (29%), superotemporal in four (24%), temporal in two (12%), inferotemporal in three (18%), inferior in none, and inferonasal in one (6%). The posterior tumor margin was a mean of 2.3 mm from the optic disc margin (range, 0-6.5 mm; median, 2.0 mm) and a mean of 2.7 mm from the foveola (range, 0.1-5 .7 mm; median, 3.0 mm) (Figs 1 and 2). Seven tumors (41 %) touched the optic disc margin and three (18%) were under the fovea (Fig 2). The mean tumor base was 6.6 mm (range, 3.0-12.0 mm; median, 6.5 mm), and mean tumor thickness was 3.0 mm (range, 1.7-5.1 mm; median, 2.9 mm).

A secondary retinal detachment was present in 14 eyes (82%), orange pigment clumps on the tumor surface in 15 (88%), and drusen on the tumor surface in 4 (24%). The tumor was dome-shaped in 16 patients (94%) and mushroom-shaped in 1 (6%). Growth had been photo­graphically documented in seven patients (41 %) before treatment.

The total number of treatment sessions per patient was 2 in 11 patients, 3 in 4 patients, and 4 in 2 patients. The mean number of sessions per patient was 2.5. Treatment session 1 was received by all patients and included 3-and 2-mm spots of thermotherapy, each at I-minute dura­tion. The 3-mm spots were at a mean power of 595 mW (range, 300-781 mW; median, 616 mW) and averaged a total of eight spots per tumor (range, 1-16 spots; median, 7 spots). The 2-mm spots were at a mean power of 600 mW (range, 450-700 mW; median, 625 mW) and were used in only four patients, with an average of four spots per tumor (range, 2-8 spots). There were no patients with I-mm treatment spots. Immediately after treatment session 1, 15 tumors had a white appearance and 2 were gray. There was no instance of retinal vascular obstruc­tion, despite treatment directly through the retinal vessels. In two patients, a single 200-J.Lm retinal hemorrhage at the treatment site was noted.

Treatment session 2 was received by all patients and included 3- and 2-mm spots of thermotherapy, each at 1-minute duration. The 3-mm spots were received in 15 patients at a mean power of 518 m W (range, 175 - 745 mW; median, 483 mW) and averaged a total of 11 spots per tumor (range, 3-11 spots; median, 7 spots). The 2-mm spots were at a mean power of 572 m W (range, 317-700 m W; median, 700 m W) and were used in only three patients, with an average of eight spots per tumor (range, 3-11 spots). There were no patients with I-mm treatment spots. Immediately after treatment session 2, three tumors had a white appearance, ten had a gray appearance, and

1644

four had a barely gray color. In one patient, a single 400-J.Lm retinal hemorrhage at the treatment site was noted.

Treatment session 3 was received by six patients and included 3-, 2-, and I-mm spots of thermotherapy, each at I-minute duration. The 3-mm spots were received in five patients at a mean power of 621 mW (range, 400-700 mW; median, 669 mW) and averaged a total of five spots per tumor (range, 2-10 spots; median, 6 spots). The 2-mm spots were at a mean power of 520 m W (range, 240-800 mW; median, 520 mW) and were used in only two patients, with an average of four spots per tumor (range, 3-5 spots). There were two patients with I-mm treatment spots, a mean power of 450 mW (range, 400-500 mW; median, 450 mW), and a mean of six spots per tumor (range, 5-7 spots). Immediately after treatment session 3, no tumors had a white appearance, three had a gray appearance, and three had a barely gray color. At treatment, a twig retinal vein occlusion developed in one patient at the treatment site.

Treatment session 4 was received by two patients and included only 3-mm spots of thermotherapy, each at 1-minute duration. The 3-mm spots were at a mean power of 620 mW (range, 600-640 mW; median, 620 mW) and averaged a total of three spots per tumor (range, 1-5 spots; median, 3 spots). Immediately after treatment ses­sion 4, no tumors had a white appearance, none had a gray appearance, and two had a barely gray color.

The mean tumor thickness measured ultrasonographi­cally before treatment was 3.0 mm. The mean tumor thickness gradually decreased to 2.3 mm at month 1,2.1 mm at month 2, 2.0 mm at month 4, 1.7 mm at month 6 after initial treatment, and stabilized thereafter (Figs 1-4). Choroidal atrophy with partial or complete visibil­ity of the sclera at the site of treatment was present in all patients. When comparing the percent reduction of tumor thickness as a function of tumor pigmentation, we found that at the initial follow-up visit (mean, 1.7 months), tu­mor reduction was 15% for amelanotic melanoma and 34% for heavily pigmented melanoma (overall mean re­duction thickness, 27%). At the 6-month follow-up, the percent reduction of the tumor thickness was 21 % for amelanotic melanoma and 50% for heavily pigmented tumors (overall mean tumor reduction, 43%). All tumors showed some degree of regression to the thermotherapy. None of the patients had local tumor recurrence.

Final visual acuity was 20/20 in four eyes (24%), 201 25 in three (18%), 20/30 in two (12%), 20/40 in three (18%),20/50 in none, 20/60 in two (12%), 20/70 in none, 20/80 in none, 201100 in one (6%), 20/200 in none, and 20/400 in two (12%) (Fig 5). Visual acuity was the same before and after treatment in four eyes (24%), improved after treatment in six eyes (35%), and decreased in seven eyes (41 %). The reason for improved vision was resolu­tion of subretinal fluid in all patients. The reasons for decreased vision were retinal vascular obstruction to fo­veal vessels in four patients and preretinal fibrosis in the fovea in three. All patients were alive with no evidence of melanoma metastasis.

The complications of transpupillary thermotherapy were generally limited to the site of treatment. There were

Shields et al . Transpupillary Thermotherapy in the Management Choroidal Melanoma

Figure 1. Choroidal melanoma in a 68-year-old man. A, wide-angle fun­dus photograph shows choroidal melanoma superior to superotemporal arcade. B, close-up 45° fundus photograph of the choroidal melanoma shows subtle orange pigment and subretinal fluid. C, B-scan ultrasound demonstrates acoustic hollowness; tumor thickness measured 4.1 mm be­fore treatment. D, 6 months after transpupil\ary thermotherapy to the entire tumor plus tumor-free margins, the choroidal melanoma has re­gressed, leaving minimally elevated pigment, retinal pigment epithelial alterations, and retinal traction, dragging the superotemporal arcade into the treatment site. E, 6 months after transpupillary thermotherapy the B­scan ultrasound demonstrated marked flattening of the tumor and a thick­ness measurement of 1. 7 mm.

1645

Ophthalmology Volume 103, Number 10, October 1996

no occurrences of direct heat-related alterations in the cornea, lens, iris, vitreous, or optic nerve. Glaucoma, tu­mor extension through Bruch membrane, choroidal hem­orrhage, and retinal hole formation did not develop in any of the patients. The complications of treatment included

1646

Figure 2. Juxtapapillary choroidal melanoma in a 45-year-old man. A, wide­angle fundus photograph shows the choroidal melanoma in the inferonasal juxtapapillary region. B, close-up 45° fundus photograph demonstrates the choroidal melanoma touching the nasal margin of the optic disc. Overlying subretinal fluid and orange pigment are visible. C, B-scan ultrasound demon­strates acoustic hollowness; the tumor thickness measured 3.4 mm. D, 5 months after treatment, the tumor has regressed, leaving an excavated scar of flat pigment with partial visibility of the sclera, retinal pigment epithelium alterations, and obstruction of the overlying retinal vessels. Visual acuity remained 20/20. E, B-scan ultrasound after treatment shows marked flat­tening of the tumor; the thickness measured 1.6 mm.

overlying branch retinal artery or vein occlusion in eight patients, less than a 500-j.lm retinal hemorrhage in four, retinal traction in four, macular edema in two, choroidal hemorrhage in one, and focal neovascularization of the retina in two (presumably secondary to a branch retinal

Shields et al . Transpupillary Thermotherapy in the Management Choroidal Melanoma

5.5

4.5

1.5

0.5 +---4---4---4---4---4---4---~--~---r--~

10

Months after treatment

Figure 3. Transpupillary thermotherapy for choroidal melanoma in 17 patients: individual tumor-thickness response. (Each tumor is indicated by a line and a unique symbol t o differentiate overlapping lines.)

vein obstruction). Management of these problems was observation in all patients except for the patients with retinal neovascularization, who were treated with sector scatter retinal photocoagulation. In two patients, delayed ocular pain with stable clinical findings developed nearly 1 week after treatment. The pain was attributed to heat­related tumor necrosis, inflammation, or neuralgia of the ciliary nerves, and was treated with topical corticoste­roids.

Discussion

Hyperthermia is the elevation of temperature above the normal body temperature of 37°C. It is well recognized that hyperthermia can be used as a treatment for a variety of cancers?2-25 Hyperthermia can be used as a primary treatment by devitalizing cells through a direct cytotoxic effect or it can be used as an adjunctive treatment by sensitizing cells to radiotherapy or chemotherapy. The cytotoxic effect is general1y achieved with more intense treatment at a higher temperature, whereas the sensitizing effect is found with a lower temperature?·17.26.27 Hyper­thermia is believed to preferentially affect neoplastic tis­sue over normal tissue because the neoplastic tissue typi­cally has inefficient vascular channels and thus an im­paired cooling system, allowing a heat sink to develop and be maintained. The heat sink may be cytotoxic in itself or it may act to sensitize the tumor by inhibiting

DNA repair mechanisms within the cell so that radiation or chemotherapy is more effective.26

Thermotherapy combined with radiation therapy has been found to be beneficial for superficial malignant tu­mors, including cutaneous malignant melanoma, head and neck tumors, and breast cancer.26-29 Engin et af9 found that the addition of thermotherapy to the radiation proto­col for superficial malignant tumors provided a greater chance for complete response of the tumor, especially if it were within the limits of heat penetration (within 3 cm deep to the skin surface). In addition, other investigators found that those patients with head and neck cutaneous melanoma treated with radiotherapy had both a more im­mediate response and an enhanced response durati --.n if thermotherapy were added to the regimen.27

With regard to the eye, thermotherapy has been investi­gated for its effects on intraocular tumors, particularly uveal malignant melanoma. Several thermotherapy sources have been used, including ultrasound,4-s micro­waves,9-13 localized current field,16 ferromagnetic ther­moseeds,14.15 and infrared radiation.3.17- 2o In most patients, the thermotherapy has been delivered as an adjuvant to radiotherapy, usually plaque brachytherapy. Although most of the studies have been preliminary, it is believed that the addition of thermotherapy to brachytherapy is safe, enhances tumor necrosis, and may possibly allow for a decrease in the radiation dose to ultimately minimize radiation complications.13

The main drawback of the currently used thermother­apy units of microwaves, localized current field , and fer­romagnetic thermoseeds is that the heat is delivered in a transscleral fashion, similar to brachytherapy. For tumors abutting the optic nerve, localization of the tumor using indirect methods can be very tedious and possibly inaccu­rate because of the far posterior location of the mass and obstruction by the meninges surrounding the optic nerve. In the juxtapapillary region, the thermotherapy measure­ments would be altered using the above methods due to displacement of the delivery system as it impinges on the meninges and optic nerve, We have encountered this problem over the years with radioactive plaque placement and have used a specially designed shallow or deep­notched plaque to deliver adequate radiation to the por­tions of the tumor within 3 mm of the optic nerve that are possibly skewed posteriorly off the plaque edge. I

,2,3o

This potential malposition should be considered in the thermal dose calculations when applying trans scleral ther­motherapy to tumors in the juxtapapillary region.

One way to avoid the theoretical aberrations of trans­scleral thermotherapy treatment for juxtapapillary tumors is to deliver the thermotherapy in a direct transpupillary fashion, as described by Oosterhuis and co-workers.3 This method uses near-infrared radiation delivered through the dilated pupil and is termed transpupillary thermotherapy. Using this method, posterior tumors in the juxtapapillary and macular region can be adequately heated to induce tumor necrosis with or without the addition of radiation. 3

Investigation into the effects of infrared thermotherapy on uveal malignant melanoma in animals has been con­ducted, primarily in The Netherlands. 17-2o Journee-de

1647

Ophthalmology Volume 103) Number 10) October 1996

Korver and co-workers ' 8-2o studied infrared hyperthermia

on pigmented and nonpigmented melanoma in hamsters and found that the optimal hyperthermia for tumor necro­sis was delivered at a temperature of 45° to 60°C at 1-minute duration, a level that they termed intermediate hyperthermia. They emphasized that low-level hyperther­mia at 42° to 45°C did not induce tumor necrosis, but was perhaps most useful as an adjuvant treatment for enhancement of the effects of radiotherapy. The interme­diate level hyperthermia at 45° to 60°C caused a direct destructive effect on melanoma, and deep penetration of the heat effect was documented. High-level hyperthermia at 60°C or more induced coagulation that impaired heat penetration into the tumors, thus allowing for only super­ficial treatment. 18

-20 Using ideal parameters for transpu­

pillary thermotherapy of uveal melanoma (3.0-4.5-mm beam; 45°-60°C at 1 minute per spot), the authors found that the cytotoxic effects of the heat generally penetrated 3 to 4 mm into the tumor.20 They concluded that transpu­pillary thermotherapy by infrared radiation was suitable for tumor therapy, because this radiation allowed for high tumor penetration and low absorption by the ocular media. Others have confirmed their findings in human choroidal melanoma that penetration of infrared thermotherapy is optimized at a I-minute exposure time and a broad beam of 3.0 mm with heat-induced necrosis reaching 4.7 mm depth (unpublished data; Langmann G, Keinert R, Faul­born J. Presented at the Austrian Society of Ophthalmol­ogy, Eisenstadt, Austria, May 1995).

-Oosterhuis and co-workers,3 in 1995, reported their experience on the safety and efficacy of transpupillary thermotherapy for human choroidal melanoma as a pri­mary treatment in three patients and as adjuvant treatment to plaque radiotherapy in nine patients. All tumors showed a reduction of tumor height, and side effects were minimal. In those three patients treated with primary transpupillary thermotherapy alone, the tumor was a mean thickness of 2.8 mm before treatment and completely flat at O-mm thickness after approximately 9 months of fol­low-up.

Our results are similar to those reported by Oosterhuis and associates. The mean pretreatment tumor thickness in the 17 patients who we treated with primary transpupillary thermotherapy was 3.0 mm, and the tumor decreased to a mean of 2.3 mm after 1 month and to 1.7 mm after 6 months (Fig 4). Clinically, most tumors appeared as flat or excavated with visible sclera partially obscured by dis­persed pigment, fibrosis, and retinal edema. The measure­ment by ultrasound included the overlying retinal changes, thereby adding some thickness to the measure­ment of the scar. Despite the measured minimal elevation with ultrasound after treatment, the tumors were ophthal­moscopically flat or even excavated.

The treatment that we delivered was focused conflu­ently over the entire surface of the tumor and particularly at the margins of the tumor to eradicate subclinical foci with a margin of error. It is known that the normal choroi­dal circulation may dissipate some of the heat at the edge of the tumor, so overlap of normal choroid was important, even if it involved giving more heat to the fovea. Treat-

1648

ment was repeated to achieve our endpoint of complete tumor regression. With repeated treatment sessions, the clinically visible change in tumor appearance from the heat as viewed funduscopically was less, perhaps related to tissue atrophy at the site.

We realize that it is subjective to quantify the degree of pigmentation within a tumor, but those tumors with heavier pigmentation responded with a greater percent reduction of thickness, presumably due to greater absorp­tion of the heat. The amelanotic tumors usually required a higher power level to achieve visible heat effects. At the initial follow-up (mean, 1.7 months), the overall tumor thickness reduction was 27% as measured by ultrasound, and specifically with regard to intrinsic pigmentation of the tumor, the tumor reduction was 15% for amelanotic melanoma and 34% for heavily pigmented melanoma. At last follow-up (mean, 6 months), the overall tumor thickness reduction was 43%, and, more specifically, was 21 % for amelanotic melanoma and 50% for heavily pig­mented tumors. These findings are expected because mel­anin is the main intraocular chromophore in the near­infrared spectrum of radiation. These variations in absorp­tion should be considered when applying transpupillary thermotherapy to choroidal melanoma.

The treatment, delivered through the retina into the choroidal tumor, caused little immediate change to the overlying retinal vasculature, but the overlying retina as­sumed a light gray color. Occasionally, tiny retinal hemor­rhages could be seen. The later effects on the retina were more profound and generally occurred in the immediate vicinity of treatment. Overlying retinal destruction, mild

3.5

3.0

E g 2.5

j .>l oS

~ B :; 2.0

~

1.5

1.0 -t-----t---t---+---t---t-----i o 2 4 6

Time after initial treatment (months)

Figure 4. Transpupillary thermotherapy for choroidal melanoma in 17 patients: mean tumor thickness response over a 6-month period.

Shields et al . T ranspupillary Thermotherapy in the Management Choroidal Melanoma

traction of the retinal scar on the normal retina, adjacent macular edema, and vascular sclerosis became evident 2 to 3 months after treatment. Some patients maintained thinned but patent retinal vessels. Branch retinal vein ob­struction leading to intraretinal hemorrhages and focal retinal neovascularization occurred in two of our patients. Those patients in whom macular edema or macular retinal traction developed had tumors treated within 3 mm of the foveola. Despite treatment of seven tumors immediately touching the optic nerve, thus far there have been no cases of optic nerve inflammation. However, mild sectoral optic atrophy due to adjacent retinal atrophy could be visual­ized. We observed no anterior segment complications re­lated to the heat in our series. Others have reported focal iris atrophy, cataract, and corneal scarring? Longer fol­low-up is necessary to assess the long-term effects of transpupillary thermotherapy, especially the effects on ad­jacent structures such as the optic nerve and foveola.

In two patients, a deep, aching pain developed in the treated eye as a delayed effect approximately 1 week after treatment. There were no conclusive clinical findings to explain the symptoms, and both patients were treated em­pirically with topical corticosteroids with resolution of the pain over 48 hours. We speculate that the pain may have been related to necrosis or inflammation within the treated tumor. In the past, we have observed pain in those patients with spontaneous necrosis of uveal melanoma. Subsequent to the closure of data collection for this report, pain has developed in other similarly treated patients 1 to 3 weeks after treatment.

Visual acuity in the 17 treated eyes was the same or improved in ten patients (59%), primarily due to resolu­tion of subfoveal fluid. In seven eyes (41 %), vision de­creased after treatment, which was related to the proxim­ity of the tumor to the foveola and optic disc as well as the location of the retinal vessels at the treatment site (Fig 5). In these patients, the tumor was a mean of 2.5 mm to the foveola and 2.4 mm to the optic disc. The primary reasons for decreased vision included macular retinal vas­cular obstruction in four eyes and surface wrinkling reti­nopathy in three. When considering the proximity of the treated tumor to the foveola and optic disc, we speculate that the visual outcome may ultimately be similar to or better than plaque radiotherapy for comparable tumors, because the transpupillary thermotherapy induces focal damage, as shown histopathologically,3 and the global visual loss associated with diffuse radiation effects on the optic nerve and retina may be avoided. If feasible, a randomized study comparing the two treatment modalities would possibly provide a more precise comparison of visual and systemic outcome of the two treatments.

Thermotherapy is delivered at subphotocoagulation temperatures using a specially designed delivery system. One should not confuse laser photocoagulation with infra­red thermotherapy, because the technique and pathophysi­ologic effects on the tumor differ. 19 For laser photocoagu­lation, a small beam size of 50 to 500 p,m is used, the light exposure is a fraction of a second, and the wave­length may differ, especially if argon or krypton is used?,31,32 Laser photocoagulation generally results in a

• •

• ~ 20[10 .... 1l 20~ .~ 20/50 • • t,)

'" • c;l 20/40

'" ;:;: 20130 • 20/25

• rJJ 20/20 rJJ •

~ § ~ ~ ~ ~ ~ ~ ~ ~ N N N N N N N ?:l ?:l ~

Visual acuity after treatment

Figure 5. Transpupillary thermotherapy for choroidal melanoma in 17 patients: visual acuity before and after treatment, The diagonal line indi­cates no change in vision, The points above the diagonal line represent improvement in final visual acuity after treatment and the points below the diagonal line represent decrease in final visual acuity after treatment,

depth of tumor necrosis of less than 1 mm, but with increased exposure time to 30 seconds the penetration may be somewhat increased.33 Others have reported on success in a rabbit model using indocyanine green with diode laser photocoagulation to increase penetration of the laser into the tumor, but there was concern that tumor cells persisted along large choroidal vessels.34 Infrared thermotherapy, on the other hand, is optimized by using a large beam size of 3.0 to 4.5 mm at a wavelength of 810 nm with a long exposure of 1 minute or more. As mentioned above, the depth of necrosis is generally 3.0 to 4.5 mm. With regard to the pathophysiology of tumor death, laser photocoagulation results in a rise in tempera­ture above 65°C and causes direct necrosis due to coagula­tion of tissue, whereas thermotherapy at subcoagulation temperatures of 45° to 60°C induces cell necrosis due to intracellular breakdown, primarily of mitochondria. 19

It must be stressed that we do not advocate treatment of all small melanocytic choroidal lesions. The chance of growth and metastasis from most small lesions is so low that treatment is not justified?1 However, if there is un­equivocal documentation of tumor enlargement or if there are substantial clinical risk factors for metastasis, as de­fined in a recent report,21 then treatment may be consid­ered. The complications and visual loss related to any treatment can be profound and should be realized.

In summary, preliminary observations suggest that transpupillary thermotherapy is an effective treatment modality for small choroidal melanoma, especially those in the juxtapapillary region and posterior fundus. The tumor response is relatively rapid over a few months. There are potential ocular complications related primarily to retinal vascular obstruction and retinal traction. Al­though there is usually a dramatic tumor response to treat­ment, longer follow-up is necessary to assess for local

1649

Ophthalmology Volume 103, Number 10, October 1996

recurrence of the tumor and the impact of treatment on survival.

References

1. Shields JA, Shields CL. Intraocular Tumors. A Text and Atlas. Philadelphia: WB Saunders, 1992; 171-206.

2. Shields JA, Shields CL, Donoso LA. Management of poste­rior uveal melanoma. Surv Ophthalmol 1991;36:161-95.

3. Oosterhuis JA, Journee-de Korver HG, Kakebeeke-Kemme HM, Bleeker Je. Transpupillary thermotherapy in choroidal melanomas. Arch Ophthalmol 1995;113:315-21.

4. Burgess SEP, Chang S, Svitra P, et al. Effect of hyperther­rrria on experimental choroidal melanoma. Br J Ophthalmol 1985;69:854-60.

5. Silverman RH, Coleman DJ, Lizzi FL, et al. Ultrasonic tissue characterization and histopathology in tumor xeno­grafts following ultrasonically induced hypertherrrria. Ultra­sound Med BioI 1986;12:639-45.

6. Coleman DJ, Lizzi FL, Burgess SEP, et al. Ultrasonic hy­perthermia and radiation in the management of intraocular malignant melanoma. Am J OphthalmoI1986;101 :635-42.

7. Coleman OJ, Silverman RH, Iwamoto T, et al. Histopatho­logic effects of ultrasonically induced hypertherrrria in in­traocular malignant melanoma. Ophthalmology 1988;95: 970-81.

8. Braakman R, van der Valk P, van Delft JS, et al. The effects of ultrasonically induced hypertherrrria on experimental tu­mors in the rabbit eye. Invest Ophthalmol Vis Sci 1989;30:835-44.

9. Finger PT, Packer S, Svitra PP, et al. Hypertherrrric treat­ment of intraocular tumors. Arch Ophthalmol 1984;102: 1477-81.

10. Finger PT, Packer S, Svitra PP, et al. Thermoradiotherapy for intraocular tumors. Arch Ophthalmol 1985;103:1574-78.

11. Finger PT, Packer S, Paglione RW, et al. Thermoradiother­apy of choroidal melanoma. Ophthalmology 1989;96: 1384-8.

12. Swift PS, Stauffer PR, Fries PO, et al. Microwave hyper­thermia for choroidal melanoma in rabbits. Invest Ophthal­mol Vis Sci 1990;31 :1754-60.

13. Finger PT. Microwave plaque thermoradiotherapy for cho­roidal melanoma. Br J Ophthalmol 1992;76:358-64.

14. Mieler WF, Jaffe GJ, Steeves RA. Ferromagnetic hyper­thermia and iodine 125 brachytherapy in the treatment of choroidal melanoma in a rabbit model. Arch Ophthalmol 1989;107:1524-8.

15. Steeves RA, Tompkins DT, Nash RN, et al. Thermoradio­therapy of intraocular tumors in an animal model: concur­rent vs. sequential brachytherapy and ferromagnetic hyper­therrrria. Int J Rad Oncol BioI Phys 1995;33:659-62.

16. Liggett PE, Ma C, Astrahan M, et al. Combined localized current field hypertherrrria and irradiation for intraocular tumors. Ophthalmology 1991;98:1830-6.

17. Nuijs-Beems EM, Oosterhuis JA, Verburg van der Marel

1650

EH, et al. Tumor destruction by intermediate level hyper­thermia. Curr Eye Res 1990;9:771-90.

18. Journee-de Korver JG, Oosterhuis JA, Van Best JA, Fakkel 1. Xenon arc photocoagulator used for transpupillary hyper­therrrria. Doc Ophthalmol 1991;78:183-7.

19. Journee-de Korver JG, Oosterhuis JA, Kakebeeke-Kemme HM, de Wolff-Rouendaal D. Transpupillary thermotherapy (TIT) by infrared irradiation of choroidal melanoma. Doc Ophthalmol 1992;82: 185-91.

20. Journee-de Korver JG, Verburg-van der Marel EH, Ooster­huis JA, et al. Tumoricidal effect of hypertherrrria by near infrared irradiation on pigmented hamster melanoma. La­sers Light Ophthalmol 1992;4:175-80.

21. Shields CL, Shields JA, Kiratli H, et al. Risk factors for metastasis of small choroidal melanocytic lesions. Ophthal­mology 1995;102:1351-61.

22. Leeper D. Molecular and cellular mechanism of hyperther­mia alone or combined with other modalities. In: Overgaard J, ed. Hyperthermic Oncology. London, England: Taylor and Francis Ltd, 1985;2:9-40.

23. Reinhold HS. Physiological effects ofhypertherrrria. Recent Results Cancer Res 1988;107:32-43.

24. Overgaard J, Overgaard M. Hyperthermia as an adjuvant to radiotherapy in the treatment of malignant melanoma. Int J Hypertherrrria 1987;3:483-501.

25. Hall EJ, Roisin-Towie L. Biological effects of heat. Cancer Res 1984;44(suppl):4708-13.

26. Raaphorst GP, Heller DP, Bussey A, Ng CEo Thermal ra­diosensitization by 41°C hypertherrrria during low dose-rate irradiation in human normal and tumour cell lines. Int J Hypertherrrria 1994;10:263-70.

27. Arcangeli G, Benassi M, Cividalli A, et al. Radiotherapy and hyperthermia. Analysis of clinical results and identifi­cation of prognostic variables. Cancer 1987;60:950-6.

28. Shidnia H, Hornback NB, Shen RN, et al. An overview of the role of radiation therapy and hypertherrrria in treatment of malignant melanoma. Adv Exp Med Bioi 1990;267:531-45.

29. Engin K, Leeper DB, Tupchong L, et al. Thermoradiation therapy for superficial malignant tumors. Cancer 1993;72: 287-96.

30. DePotter P, Shields CL, Shields JA, et al. Impact of enucle­ation versus plaque radiotherapy in the management of jux­tapapillary choroidal melanoma on patient survival. Br J Ophthalmol 1994;78: 109-14.

31. Shields JA, Glazer LC, Mieler WF, Shields CL. Compari­son of xenon arc and argon laser photocoagulation in the treatment of choroidal melanomas. Am J Ophthalmol 1990;109:647-55.

32. Shields JA. The expanding role of laser photocoagulation for intraocular tumors. The 1993 H. Christian Zweng Me­morial Lecture. Retina 1994;14:310-22.

33. Foulds WS, Damato BE. Low energy long exposure laser therapy in the management of choroidal melanoma. Graefes Arch Clin Exp Ophthalmol 1986;224:26-31.

34. Chong LP, OzIer SA, De Queiroz 1M, Liggett PE. Indocya­nine green-enhanced diode laser treatment of melanoma in a rabbit model. Retina 1993;13:251-9.