Radiation Maculopathy after Proton Beam Irradiation for Choroidal Melanoma David R. Guyer, MD, Shizuo Mukai, MD, Kathleen M. Egan, MPH, Johanna M. Seddon, MD, Susan M. Walsh, BS, Evangelos S. Gragoudas, MD Purpose: Radiation maculopathy is a microangiopathy of the retina, which is often observed after irradiation of the eye. To quantitatively determine the frequency of anatomic and functional features of this condition, the authors reviewed a large series of patients with choroidal melanomas who were treated by proton beam irradiation. Methods: Color photographs and/or fluorescein angiograms of 218 patients with paramacular choroidal melanomas were graded by two independent masked readers to determine the frequency of the various lesions of radiation maculopathy. Results: Overall, 194 (89%) of the 218 patients in this study developed some degree of radiation maculopathy. The earliest and most common finding was macular edema, which was observed in 87% of patients, overall, by the end of year 3. Microvascular changes (microaneurysms and/or telangiectasia), intraretinal hemorrhages, and capillary nonperfusion were noted in 76%, 70%, and 64% of eyes, respectively, by the end of postirradiation year 3. Visual acuity was retained at 20/200 or better in 90% of eyes 1 year after irradiation and in 67% of eyes 3 years after treatment. Conclusion: Radiation maculopathy is common after proton beam irradiation of paramacular choroidal melanomas. However, ambulatory vision is preserved in many eyes. Ophthalmology 1992;99:1278-1285
Radiation retinopathy is a delayed vaso-occlusive microangiopathy of the retinal vasculature. Its diverse manifestations include intraretinal hemorrhages, cotton-wool spots, microaneurysms, telangiectasia, vascular sheathing, capillary nonperfusion, macular edema, lipid exudation, retinal neovascularization, retinal pigment epithelium changes, and vitreous hemorrhage. Subretinal neovascularization, J central retinal vein occlusion, 2 and central retinal artery occlusion 3 are less common findings. In Originally received: October 13, 1991. Revision accepted: February 17, 1992. From the Retina Service, Massachusetts Eye and Ear Infirmary, Harvard Medical School, Boston. Presented at the American Academy of Ophthalmology Annual Meeting, Anaheim, October 1991, and the Association for Research in Vision and Ophthalmology Meeting, Sarasota, May 1991 . Supported in part by a Heed Foundation Fellowship to Dr. Guyer and in part by the Macula Foundation, New York, New York. Reprint requests to Evangeios Gragoudas, MD, Retina Service, Massachusetts Eye and Ear Infirmary, 243 Charles St, Boston, MA 02114.
1278
1933, Stallard4 first described radiation retinopathy in cases of retinal capillary hemangioma and retinoblastoma treated with radon-seed irradiation. Since then, many investigators have reported radiation retinopathy occurring after local or external beam irradiation. 5- 27 To quantitate the frequency of anatomic and functional features of radiation maculopathy, we studied a large series of patients with paramacular choroidal melanoma who were treated with proton beam irradiation and who were at high risk of radiation maculopathy. Intrinsic tumor factors and damage of the tumor vessels may playa role in post-treatment morbidity and may confuse the picture ofradiation injury to normal retinal vessels. To minimize this, we included only eyes with tumors of small and moderate size.
Patients and Methods Eyes with choroidal melanomas that were less than 15 mm in diameter, less than 5 mm in height, within 4 disc
Guyer et al . Radiation Maculopathy after Proton Beam Irradiation diameters of the center of the fovea (paramacular tumor), and treated with proton beam irradiation28 - 3o before January 1988, were eligible for inclusion in the study. Of 337 eligible eyes, the 218 eyes in which at least 1 postirradiation color photograph and/or fluorescein angiogram was performed at the Massachusetts Eye and Ear Infirmary (MEEI) were included in the current study. Patients included in the study were similar to patients who were eligible but lacked MEEI follow-up in age, gender, eye involved, percent with poor preirradiation visual acuity (worse than 20/200), percent with macular detachment, location of the posterior tumor margin in relation to the macula, radiation dose received by the macula, and tumor diameter (Table 1). However, those excluded had slightly taller tumors (mean, 3.4 versus 3.1 mm; P = 0.03). The median number of postirradiation photographic or fluorescein angiographic evaluations was 3 (range, 1 to 8). The mean time between irradiation and the first postirradiation examination was 11.8 months (range, 1 month to 5.9 years). The mean time between irradiation and the most recent evaluation was 39.7 months (range, 5.2 months to 14.7 years). Color photographs and/or fluorescein angiograms were graded by two independent masked readers (ORG and SM). The presence or absence within 1500 J.tm of the foveal avascular zone of each of the following was evaluated: macular edema, intraretinal hemorrhages, cotton-wool spots, microvascular abnormalities (telangiectasia and/or microaneurysms), vascular sheathing, intraretinal lipid, subretinal fluid and/or lipid, and capillary nonperfusion. Cases that were graded differently by the two readers were re-evaluated by the readers jointly, and an arbitrated grade was assigned. A randomly selected sample of 50
color photographs and 38 fluorescein angiograms were blindly re-reviewed in the above manner to assess intrareader reliability. Agreement between the readings ranged from 70% for macular edema to 94% for intraretinallipid. The median kappa value was 0.57 (range, 0.39 for macular edema to 0.84 for intraretinallipid). Annual and cumulative rates of each retinal lesion were calculated by the Kaplan-Meier method. 3 ! Rates were estimated for all eyes combined and separately for eyes stratified by proximity of the tumor to the macula (within 2 disc diameters [~O] of the macula versus 2 to 4 DO from the macula). The log-rank test 32 was used to evaluate the equality of rates in groups stratified by distance to the macula. The timing of an event was estimated as the interval between irradiation and the midpoint between the date of first occurrence of the lesion and the preceding evaluation. The proportion of eyes with resolution after diagnosis of the individual retinal lesions also was evaluated. Rates of visual loss to worse than 20/200 were evaluated among patients with 20/200 or better initial visual acuity using the Kaplan-Meier approach.
Results Characteristics of the patients included in the series are presented in Table 1. The series consisted of 218 patients, of whom 50% were male. The mean age was 56 years (range, 16 to 91 years). The median preoperative visual acuity was 20/40 (range, 20/15 to hand motions). Ten percent of patients had a preoperative visual acuity poorer than 20/200. Almost all eyes were treated with 70 Gy de-
Table 1. Characteristics of Patients With Paramacular Intraocular Melanomas WithMEEI Follow-up* No. of patients Mean age (range) No. male (%) No. left eye (%) No. pretreated vision worse than 20/200 (%) No. macular detachment (%) Mean diameter (range) Mean height (range) Distance from macula (no./%) Involved :o::2DD >2 -4DD Dose to macula (no./%) 100% 50-99% 10-49% 2-4 DD
29 0.30 33 0.19
34 0.16 26 0.16 37 0.11 35
0.14 39 0.05
Combo
::;2 DD
75 0.61 0.61
13
116 0.40 0.40
35
140 0.27 0.27
48
121 0.28 0.28
49
155 0.20 0.20
69
151 0.21 0.21
64
172
84
0.10 0.10
0.90
0.70
0.62
0.56
0.42
0.46
0.32
>2-4 DD
13 0.56 17 0.43
21 0.33 16 0.40 23
0.27 20 0.34 29 0.08
• Estimated by the Kaplan-Meier method; annual rates presented for the combined group only. + Number
1280
Three
Two
One
followed to the end of the specified interval without having developed the retinal lesion.
Combo
26
0.44 0.83
::;2 DD
5 0.95
52 0.40 0.64
15
69 0.39 0.56
21
65 0.34 0.53
22
92 0.25 0.39
35
84 0.28 0.43
43
113 0.18 0.27
53
0.83
0.77
0.68
0.59
0.52
0.42
>2-4 DD
11 0.56 12 0.50
14 0.43 9 0.48
14 0.38 12 0.34 18 0.15
Combo
16 0.25 0.87 27 0.33 0.76
35 0.33 0.70 31 0.24 0.64 49 0.25 0.55
55
0.08 0.48
71
0.13 0.36
Guyer et al . Radiation Maculopathy after Proton Beam Irradiation MACULAR EDEMA
PERCENT
100 + 90 80 70 60 50 40 30 20 10
PERCENT
100 90 80 70 60 50 40 30 20 10 2 YEAR
3
4
5
POST - IRRADIATION
Figure 1. Top left, macular edema after proton beam irradiation for choroidal melanoma. Top right, this finding was present in 87% of patients cumulatively by postirradiation year 3. Figure 2. Microvascular changes including microaneurysms and telangiectasia may occur following proton beam irradiation for choroidal melanoma (second row left and right). These microvascular changes are even more apparent in this fluorescein angiogram (third row left) of the patient shown in the second row right, which also shows capillary nonperfusion. Seventy-six percent of the patients in this series developed such changes cumulatively by the end of postirradiation year 3 (third row right).
5) were diagnosed in 64%, 55%, and 48% of eyes, respectively, by the end of postirradiation year 3. For intraretinal hemorrhage, capillary nonperfusion, and intraretinallipid, the risk was constant year to year, whereas NFL infarcts were unlikely to develop after year 2. With the exception
of NFL infarcts, these lesions were significantly more common when the tumor was in close proximity to the macula. A less prevalent retinal lesion was vascular sheathing (Table 2B)_ Annual rates were low (10% to 18%), and
1281
Ophthalmology
Volume 99, Number 8, August 1992 INTRARETINAL HEMORRHAGES
100 90
PERCENT
80 70 60 50 40 30 20 10
Figure 3. Intraretinal hemorrhages (top left) are commonly observed in patients with choroidal melanoma after proton beam irradiation (top right). Figure 4. Pretreatment (center left and right) and posttreatment (bottom left) photographs illustrate the marked capillary nonperfusion that may occur after proton beam irradiation of choroidal melanomas.
approximately one third of patients had evidence of vascular sheathing by the end of year 3. Sheathing was uncommon when the tumor was more than 2 DD from the macula (P = 0.0002). Neovascularization (NVE) was encountered in only 13 eyes (6%) in this series.
Resolution of Retinal Lesions The proportion of eyes whose retinal lesions spontaneously regressed are presented in Table 3. Of the 175 patients who developed macular edema, 138 (79%) had follow-up subsequent to its detection. Only 7 (5%) of these
1282
patients showed spontaneous resolution of the edema. Microvascular changes and capillary nonperfusion also were unlikely to have resolved by the last available evaluation. In contrast, NFL infarcts occurred transiently with more than two thirds of affected eyes free of the lesion by the last available follow-up.
Visual Outcome Eyes with tumors near the macula are at higher risk of visual loss after proton beam irradiation. 33 ,34 Visual loss to worse than 20/200 was evaluated among the 196 pa-
Guyer et al . Radiation Maculopathy after Proton Beam Irradiation COTTON - WOOL SPOTS
100 90
PERCENT
80 70 60
50
40 30 20 10
2 YEAR
3
4
5
POST· IRRADIATION
Figure 5. Cotton-wool spots were observed in 48% of patients cumulatively by postirradiation year 3.
tients with initial visual acuity of 20/200 or better. In this series, consisting exclusively of eyes with paramacular tumors of small or moderate size, visual acuity was retained at 20/200 or better in 90% of eyes 1 year after irradiation and in 67% of eyes 3 years after treatment. Results were progressively better in eyes with tumors at greater distances from the macula; the 3-year rate of loss was 47% in eyes with tumors less than 1 DD from the macula, 29% when the tumor was 1 to 2 DD from the macula, and 15% when the tumor was 2 to 4 DD from the macula. The most recently measured visual acuity in this group was 20/50 or better in 69 eyes (35%), 20/60 to 20/200 in 41 eyes (21 %), 20/300 to counting fingers in 57 eyes (29%), and hand motions or worse in the remaining 29 eyes (15%).
Discussion This study represents the largest series to date of patients with choroidal melanomas evaluated for radiation maculopathy after radiation therapy. In contrast to previous studies, we used analytic methods that accounted for variable follow-up among the patients in the series. The color photographs and fluorescein angiograms were independently read by two reviewers who were masked to the clinical information. Limitations of the study were that it was retrospective in nature, and the rate estimates may be less precise than those that would be obtained in a prospective study, and that the series consisted only of those patients who returned to MEEI for follow-up. However, the groups with and without follow-up were broadly similar, and the rates reported should apply to all patients meeting the eligibility criteria for the study. Eighty-nine percent of the patients in our series developed radiation maculopathy. The most common finding was macular edema, which was present in 87% of patients cumulatively at 3 years after irradiation. Only 5% of the macular edema cases showed resolution at follow-up examination. Laser photocoagulation may be
beneficial to patients with macular edema because it does not appear to resolve spontaneously. Other common findings were microvascular changes (microaneurysms and/or telangiectasia), intraretinal hemorrhages, and capillary nonperfusion, which were present in 76%, 70%, and 64% of patients cumulatively by 3 years after irradiation. Retinal neovascularization was rare in our series. In the report by Brown and coworkers, s the most common findings in patients irradiated with cobalt plaques were hard exudates (85%), microaneurysms (75%), and intraretinal hemorrhages (65%). In their patients receiving external beam irradiation, the most common findings were intraretinal hemorrhages (88%) and microaneurysms (81 %). Hard exudates were observed in only 38% of these patients. Macular edema was detected in 65% of the cobalt plaque-irradiated patients and in 58% of the external beam-irradiated group. Haye and co-workers35 have reported that 18 (23%) of 79 of their patients with choroidal melanomas treated with cobalt irradiation developed a severe maculopathy with visual acuity of 20/200 or worse. Capillary obliteration, exudation, and cystoid macular edema were the most common findings reported in their study. Our fluorescein angiographic findings are consistent with earlier studies. 5 ,18,19 The fluorescein angiographic hallmarks of radiation retinopathy include capillary nonperfusion, telangiectasia, microaneurysms, and macular edema. Irvine and Wood 36 irradiated monkeys with 3000 rad and described an animal model of radiation retinopathy. Histopathologic examination showed focal loss of capillary endothelial cells and pericyctes. This lesion was followed by confluent capillary cell loss, nerve fiber layer infarcts, and capillary nonperfusion. The deeper and smaller retinal vessels were involved earlier. Intraretinal neovascularization was observed, but neovascularization of the disc and NVE were not present. Interestingly, one histopathologic case report described a patient with subretinal neovascularization secondary to telangiectatic retinal vessels. 1 In our series, NVE was encountered in only 13 (6%) eyes. Several reports describe the histopathologic findings of choroidal and/or ciliary body melanomas treated with
Table 3. Resolution Of Signs Of Radiation Maculopathy Lesion Mascular edema Microvascular changes Hemorrhage Capillary nonperfusion Lipid NFL infarcts Sheathing
No. with No. Diagnosed Follow-up (%)
No. Resolved (%)
175
138 (79)
7 (5)
147 134
103 (70) 95 (71)
8 (8) 38 (40)
106 104 94 79
69 (65) 71 (68) 68 (72) 47 (59)
5 (7) 28 (39) 49 (68) 12 (25)
1283
Ophthalmology
Volume 99, Number 8, August 1992
proton beam irradiationY-41 These findings include endothelial cell swelling, decreased size of the blood vessel lumen, basement membrane thickening, sinusoidal vessel collapse, vessel thrombosis, perivascular fibrin deposits, lipid droplets in tumor cells, and lymphocytic infiltration. In several of these cases, it could not be determined if the retinal findings were radiation-induced or secondary to chronic retinal detachment. Ferry and co-workers38 observed edema and vascular changes only in areas adjacent to the tumor. Kincaid and associates 39 also observed localized changes. They found total hyalinization of the retina and choroid in the irradiated area in three of five cases. In one case, these authors observed diffuse retinopathy with severe atrophy, vascular occlusion, fibrosis, and cholesterol plaque formation. Saomil and co-workers41 recently reported on the histopathologic results of a large series of patients with uveal melanomas treated with proton beam irradiation. The most common retinal lesions observed by these authors were neovascularization, intraretinal hemorrhages, protein exudate, lipid deposition, cystoid bodies, and blood vessel damage (thickening and thrombosis). Retinal lesions were usually noted only overlying or in close proximity to the tumor. These findings may reflect in part the fact that most eyes had been enucleated because of complications, mainly neovascular glaucoma. Our clinical findings in patients with radiation retinopathy are very similar to those seen with diabetes mellitus. In both radiation maculopathy and diabetic retinopathy, early changes include microaneurysms, intraretinal hemorrhages, exudates, cotton-wool spots, capillary nonperfusion, and macular edema. However, proliferative neovascularization occurs much more frequently in diabetes mellitus. In addition, papillopathy is often seen with radiation maculopathy, but not with diabetic retinopathy. This study demonstrates that radiation maculopathy is common after proton beam irradiation of paramacular choroidal melanomas. Previous studies have shown that radiation maculopathy is a leading cause of visual loss in this group of patients. However, despite the high prevalence of radiation maculopathy in this series, two thirds of patients retained visual acuity of 20/200 or better by the end of postirradiation year 3. Our findings suggest that although radiation maculopathy occurs frequently in eyes with paramacular melanomas treated with proton beam irradiation, ambulatory vision is preserved in many eyes.
References 1. Boozalis GT, Schachat AP, Green WR. Subretinal neovascularization from the retina in radiation retinopathy. Retina 1987;7:156-61. 2. Cogan DG. Lesions of the eye from radiant energy. JAMA 1950; 142: 145-51. 3. Shukovsky D, Fletcher GH. Retinal and optic nerve complications in a high dose irradiation technique of ethmoid sinus and nasal cavity. Radiology 1972; 104:629-34.
1284
4. Stallard HB. Radiant energy as (a) a pathogenic and (b) a therapeutic agent in ophthalmic disorders. Gifford Edmonds prize essay for 1932. Br J Ophthalmol 1933;Monograph SuppI6:1-126. 5. Brown GC, Shields JA, Sanborn G, et al. Radiation retinopathy. Ophthalmology 1982;89: 1494-150 I. 6. Brown GC, Shields JA, Sanborn G, et al. Radiation retinopathy. Trans PA Acad Ophthalmol Otolaryngol1981 ;34: 144-51. 7. Tomsak RL, Smith JL. Radiation retinopathy in a patient with lung carcinoma metastatic to brain. Ann Ophthalmol 1980; 12:619-22. 8. Wara WM, Irvine AR, Neger RE, et al. Radiation retinopathy. Int J Radiat Oncol Bioi Phys 1979;5:81-3. 9. Bagan SM, Hollenhorst RW. Radiation retinopathy after irradiation of intracranial lesions. Am J Ophthalmol 1979;88:694-7. 10. Egbert PR, Donaldson SS, Moazed K, Rosenthal AR. Visual results and ocular complications following radiotherapy for retinoblastoma. Arch Ophthalmol 1978;96: 1826-30. II. Rotman M, Long RS, Packer S, et al. Radiation therapy of choroidal melanoma. Trans Ophthalmol Soc UK 1977;97: 431-5. 12. Char DH, Lonn LI, Margolis LW. Complications of cobalt plaque therapy of choroidal melanomas. Am J Ophthalmol 1977;84:536-41. 13. Harris JR, Levene MB. Visual complications following irradiation for pituitary adenomas and craniopharyngiomas. Radiology 1976;120:167-71. 14. Chan RC, Shukovsky D. Effects of irradiation on the eye. Radiology 1976;120:673-5. 15. Lommatzsch P. Treatment of choroidal melanomas with I06RujI06Rh beta-ray applicators. Surv OphthalmoI1974;19: 85-100. 16. MacFaul PA, Bedford MA. Ocular complications after therapeutic irradiation. Br J Ophthalmol 1970;54:237-47. 17. Bedford MA, Bedotto C, MacFaul PA. Radiation retinopathy after the application of a cobalt plaque. Report ofthree cases. Br J Ophthalmol 1970;54:505-9. 18. Hayreh SS. Post-radiation retinopathy. A fluorescence fundus angiographic study. Br J Ophthalmol 1970;54:705-14. 19. Gass JDM. A fluorescein angiographic study of macular dysfunction secondary to retinal vascular disease. VI. xray irradiation, carotid artery occlusion, collagen vascular disease, and vitritis. Arch Ophthalmol 1968;80:606-17. 20. Chee PHY. Radiation retinopathy. Am J Ophthalmol 1968;66:860-5. 21. Stallard HB. Radiotherapy for malignant melanoma of the choroid. Br J Ophthalmol 1966;50:147-55. 22. Perrers-Taylor M, Brinkley D, Reynolds T. Choroido-retinal damage as a complication of radiotherapy. Acta Radiol Ther Phys Bioi 1965;3:431-40. 23. Rosengren B. Two cases of atrophy of the optic nerve after previous roentgen treatment of the chiasmal region and the optic nerves. Acta Ophthalmol 1958;36:874-7. 24. Cibis PA, Noell WK, Eichel B. Ocular effects produced by high-intensity x-radiation. Arch Ophthalmol 1955;53:65163. 25. Flick 11. Ocular lesions following the atomic bombing of Hiroshima and Nagasaki. Am J Ophthalmol 1948;31:13754. 26. Stallard HB. Radiotherapy of malignant intra-ocular neoplasms. Br J Ophthalmol 1948;32:618-39. 27. Moore RF. Presidential address. Trans Ophthalmol Soc UK 1935;55:3-26. 28. Gragoudas ES, Goitein M, Koehler AM, et al. Proton ir-
Guyer et al . Radiation Maculopathy after Proton Beam Irradiation
29. 30. 31. 32. 33. 34. 35.
radiation of choroidal melanomas. Preliminary results. Arch Ophthalmol 1978;96: 1583-91. Gragoudas ES, Goitein M, Verhey L, et al. Proton beam irradiation: an alternative to enucleation for intraocular melanomas. Ophthalmology 1980;87:571-8l. Goitein M, Miller T. Planning proton therapy of the eye. Med Phys 1983;10:275-83. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457-8l. Peto R, Peto J. Asymptotically efficient rank invariant test procedures. J Roy Stat Soc 1972;135:185-207. Seddon JM, Gragoudas ES, Polivogianis L, et al. Visual outcome after proton beam irradiation of uveal melanoma. Ophthalmology 1986;93:666-74. Seddon JM, Gragoudas ES, Egan KM, et al. Uveal melanomas near the optic disc or fovea. Visual results after proton beam irradiation. Ophthalmology 1987;94:354-6l. Haye C, Desjardins L, Bouder P, et al. Maculopathies par radiations chez les patients traites pour melanomes choroidiens. Ophthalmologie 1990;4:229-31.
36. Irvine AR, Wood IS. Radiation retinopathy as an experimental model for ischemic proliferative retinopathy and rubeosis iridis. Am J Ophthalmol 1987;103:790-7. 37. Seddon JM, Gragoudas ES, Albert DM. Ciliary body and choroidal melanomas treated by proton beam irradiation. Histopathologic study of eyes. Arch OphthalmoI1983;101: 1402-18. 38. Ferry AP, Blair CJ, Gragoudas CS, Volk Sc. Pathologic examination of ciliary body melanoma treated with proton beam irradiation. Arch Ophthalmol 1985; 103: 1849-53. 39. Kincaid Me, Folberg R, Torczynski E, et al. Complications after proton beam therapy for uveal malignant melanoma. A clinical and histopathologic study of five cases. Ophthalmology 1988;95:982-91. 40. Zinn KM, Stein/Pokorny K, Jakobiec FA, et al. Protonbeam irradiated epithelioid cell melanoma of the ciliary body. Ophthalmology 1981;88:1315-2l. 41. Saornil Alvarez MA, Egan KM, Gragoudas ES, et al. Histopathology of proton beam irradiated versus enucleated uveal melanomas. Arch Ophthalmol 1992; [in press].
1285
Radiation retinopathy is a delayed vaso-occlusive microangiopathy of the retinal vasculature. Its diverse manifestations include intraretinal hemorrhages, cotton-wool spots, microaneurysms, telangiectasia, vascular sheathing, capillary nonperfusion, macular edema, lipid exudation, retinal neovascularization, retinal pigment epithelium changes, and vitreous hemorrhage. Subretinal neovascularization, J central retinal vein occlusion, 2 and central retinal artery occlusion 3 are less common findings. In Originally received: October 13, 1991. Revision accepted: February 17, 1992. From the Retina Service, Massachusetts Eye and Ear Infirmary, Harvard Medical School, Boston. Presented at the American Academy of Ophthalmology Annual Meeting, Anaheim, October 1991, and the Association for Research in Vision and Ophthalmology Meeting, Sarasota, May 1991 . Supported in part by a Heed Foundation Fellowship to Dr. Guyer and in part by the Macula Foundation, New York, New York. Reprint requests to Evangeios Gragoudas, MD, Retina Service, Massachusetts Eye and Ear Infirmary, 243 Charles St, Boston, MA 02114.
1278
1933, Stallard4 first described radiation retinopathy in cases of retinal capillary hemangioma and retinoblastoma treated with radon-seed irradiation. Since then, many investigators have reported radiation retinopathy occurring after local or external beam irradiation. 5- 27 To quantitate the frequency of anatomic and functional features of radiation maculopathy, we studied a large series of patients with paramacular choroidal melanoma who were treated with proton beam irradiation and who were at high risk of radiation maculopathy. Intrinsic tumor factors and damage of the tumor vessels may playa role in post-treatment morbidity and may confuse the picture ofradiation injury to normal retinal vessels. To minimize this, we included only eyes with tumors of small and moderate size.
Patients and Methods Eyes with choroidal melanomas that were less than 15 mm in diameter, less than 5 mm in height, within 4 disc
Guyer et al . Radiation Maculopathy after Proton Beam Irradiation diameters of the center of the fovea (paramacular tumor), and treated with proton beam irradiation28 - 3o before January 1988, were eligible for inclusion in the study. Of 337 eligible eyes, the 218 eyes in which at least 1 postirradiation color photograph and/or fluorescein angiogram was performed at the Massachusetts Eye and Ear Infirmary (MEEI) were included in the current study. Patients included in the study were similar to patients who were eligible but lacked MEEI follow-up in age, gender, eye involved, percent with poor preirradiation visual acuity (worse than 20/200), percent with macular detachment, location of the posterior tumor margin in relation to the macula, radiation dose received by the macula, and tumor diameter (Table 1). However, those excluded had slightly taller tumors (mean, 3.4 versus 3.1 mm; P = 0.03). The median number of postirradiation photographic or fluorescein angiographic evaluations was 3 (range, 1 to 8). The mean time between irradiation and the first postirradiation examination was 11.8 months (range, 1 month to 5.9 years). The mean time between irradiation and the most recent evaluation was 39.7 months (range, 5.2 months to 14.7 years). Color photographs and/or fluorescein angiograms were graded by two independent masked readers (ORG and SM). The presence or absence within 1500 J.tm of the foveal avascular zone of each of the following was evaluated: macular edema, intraretinal hemorrhages, cotton-wool spots, microvascular abnormalities (telangiectasia and/or microaneurysms), vascular sheathing, intraretinal lipid, subretinal fluid and/or lipid, and capillary nonperfusion. Cases that were graded differently by the two readers were re-evaluated by the readers jointly, and an arbitrated grade was assigned. A randomly selected sample of 50
color photographs and 38 fluorescein angiograms were blindly re-reviewed in the above manner to assess intrareader reliability. Agreement between the readings ranged from 70% for macular edema to 94% for intraretinallipid. The median kappa value was 0.57 (range, 0.39 for macular edema to 0.84 for intraretinallipid). Annual and cumulative rates of each retinal lesion were calculated by the Kaplan-Meier method. 3 ! Rates were estimated for all eyes combined and separately for eyes stratified by proximity of the tumor to the macula (within 2 disc diameters [~O] of the macula versus 2 to 4 DO from the macula). The log-rank test 32 was used to evaluate the equality of rates in groups stratified by distance to the macula. The timing of an event was estimated as the interval between irradiation and the midpoint between the date of first occurrence of the lesion and the preceding evaluation. The proportion of eyes with resolution after diagnosis of the individual retinal lesions also was evaluated. Rates of visual loss to worse than 20/200 were evaluated among patients with 20/200 or better initial visual acuity using the Kaplan-Meier approach.
Results Characteristics of the patients included in the series are presented in Table 1. The series consisted of 218 patients, of whom 50% were male. The mean age was 56 years (range, 16 to 91 years). The median preoperative visual acuity was 20/40 (range, 20/15 to hand motions). Ten percent of patients had a preoperative visual acuity poorer than 20/200. Almost all eyes were treated with 70 Gy de-
Table 1. Characteristics of Patients With Paramacular Intraocular Melanomas WithMEEI Follow-up* No. of patients Mean age (range) No. male (%) No. left eye (%) No. pretreated vision worse than 20/200 (%) No. macular detachment (%) Mean diameter (range) Mean height (range) Distance from macula (no./%) Involved :o::2DD >2 -4DD Dose to macula (no./%) 100% 50-99% 10-49% 2-4 DD
29 0.30 33 0.19
34 0.16 26 0.16 37 0.11 35
0.14 39 0.05
Combo
::;2 DD
75 0.61 0.61
13
116 0.40 0.40
35
140 0.27 0.27
48
121 0.28 0.28
49
155 0.20 0.20
69
151 0.21 0.21
64
172
84
0.10 0.10
0.90
0.70
0.62
0.56
0.42
0.46
0.32
>2-4 DD
13 0.56 17 0.43
21 0.33 16 0.40 23
0.27 20 0.34 29 0.08
• Estimated by the Kaplan-Meier method; annual rates presented for the combined group only. + Number
1280
Three
Two
One
followed to the end of the specified interval without having developed the retinal lesion.
Combo
26
0.44 0.83
::;2 DD
5 0.95
52 0.40 0.64
15
69 0.39 0.56
21
65 0.34 0.53
22
92 0.25 0.39
35
84 0.28 0.43
43
113 0.18 0.27
53
0.83
0.77
0.68
0.59
0.52
0.42
>2-4 DD
11 0.56 12 0.50
14 0.43 9 0.48
14 0.38 12 0.34 18 0.15
Combo
16 0.25 0.87 27 0.33 0.76
35 0.33 0.70 31 0.24 0.64 49 0.25 0.55
55
0.08 0.48
71
0.13 0.36
Guyer et al . Radiation Maculopathy after Proton Beam Irradiation MACULAR EDEMA
PERCENT
100 + 90 80 70 60 50 40 30 20 10
PERCENT
100 90 80 70 60 50 40 30 20 10 2 YEAR
3
4
5
POST - IRRADIATION
Figure 1. Top left, macular edema after proton beam irradiation for choroidal melanoma. Top right, this finding was present in 87% of patients cumulatively by postirradiation year 3. Figure 2. Microvascular changes including microaneurysms and telangiectasia may occur following proton beam irradiation for choroidal melanoma (second row left and right). These microvascular changes are even more apparent in this fluorescein angiogram (third row left) of the patient shown in the second row right, which also shows capillary nonperfusion. Seventy-six percent of the patients in this series developed such changes cumulatively by the end of postirradiation year 3 (third row right).
5) were diagnosed in 64%, 55%, and 48% of eyes, respectively, by the end of postirradiation year 3. For intraretinal hemorrhage, capillary nonperfusion, and intraretinallipid, the risk was constant year to year, whereas NFL infarcts were unlikely to develop after year 2. With the exception
of NFL infarcts, these lesions were significantly more common when the tumor was in close proximity to the macula. A less prevalent retinal lesion was vascular sheathing (Table 2B)_ Annual rates were low (10% to 18%), and
1281
Ophthalmology
Volume 99, Number 8, August 1992 INTRARETINAL HEMORRHAGES
100 90
PERCENT
80 70 60 50 40 30 20 10
Figure 3. Intraretinal hemorrhages (top left) are commonly observed in patients with choroidal melanoma after proton beam irradiation (top right). Figure 4. Pretreatment (center left and right) and posttreatment (bottom left) photographs illustrate the marked capillary nonperfusion that may occur after proton beam irradiation of choroidal melanomas.
approximately one third of patients had evidence of vascular sheathing by the end of year 3. Sheathing was uncommon when the tumor was more than 2 DD from the macula (P = 0.0002). Neovascularization (NVE) was encountered in only 13 eyes (6%) in this series.
Resolution of Retinal Lesions The proportion of eyes whose retinal lesions spontaneously regressed are presented in Table 3. Of the 175 patients who developed macular edema, 138 (79%) had follow-up subsequent to its detection. Only 7 (5%) of these
1282
patients showed spontaneous resolution of the edema. Microvascular changes and capillary nonperfusion also were unlikely to have resolved by the last available evaluation. In contrast, NFL infarcts occurred transiently with more than two thirds of affected eyes free of the lesion by the last available follow-up.
Visual Outcome Eyes with tumors near the macula are at higher risk of visual loss after proton beam irradiation. 33 ,34 Visual loss to worse than 20/200 was evaluated among the 196 pa-
Guyer et al . Radiation Maculopathy after Proton Beam Irradiation COTTON - WOOL SPOTS
100 90
PERCENT
80 70 60
50
40 30 20 10
2 YEAR
3
4
5
POST· IRRADIATION
Figure 5. Cotton-wool spots were observed in 48% of patients cumulatively by postirradiation year 3.
tients with initial visual acuity of 20/200 or better. In this series, consisting exclusively of eyes with paramacular tumors of small or moderate size, visual acuity was retained at 20/200 or better in 90% of eyes 1 year after irradiation and in 67% of eyes 3 years after treatment. Results were progressively better in eyes with tumors at greater distances from the macula; the 3-year rate of loss was 47% in eyes with tumors less than 1 DD from the macula, 29% when the tumor was 1 to 2 DD from the macula, and 15% when the tumor was 2 to 4 DD from the macula. The most recently measured visual acuity in this group was 20/50 or better in 69 eyes (35%), 20/60 to 20/200 in 41 eyes (21 %), 20/300 to counting fingers in 57 eyes (29%), and hand motions or worse in the remaining 29 eyes (15%).
Discussion This study represents the largest series to date of patients with choroidal melanomas evaluated for radiation maculopathy after radiation therapy. In contrast to previous studies, we used analytic methods that accounted for variable follow-up among the patients in the series. The color photographs and fluorescein angiograms were independently read by two reviewers who were masked to the clinical information. Limitations of the study were that it was retrospective in nature, and the rate estimates may be less precise than those that would be obtained in a prospective study, and that the series consisted only of those patients who returned to MEEI for follow-up. However, the groups with and without follow-up were broadly similar, and the rates reported should apply to all patients meeting the eligibility criteria for the study. Eighty-nine percent of the patients in our series developed radiation maculopathy. The most common finding was macular edema, which was present in 87% of patients cumulatively at 3 years after irradiation. Only 5% of the macular edema cases showed resolution at follow-up examination. Laser photocoagulation may be
beneficial to patients with macular edema because it does not appear to resolve spontaneously. Other common findings were microvascular changes (microaneurysms and/or telangiectasia), intraretinal hemorrhages, and capillary nonperfusion, which were present in 76%, 70%, and 64% of patients cumulatively by 3 years after irradiation. Retinal neovascularization was rare in our series. In the report by Brown and coworkers, s the most common findings in patients irradiated with cobalt plaques were hard exudates (85%), microaneurysms (75%), and intraretinal hemorrhages (65%). In their patients receiving external beam irradiation, the most common findings were intraretinal hemorrhages (88%) and microaneurysms (81 %). Hard exudates were observed in only 38% of these patients. Macular edema was detected in 65% of the cobalt plaque-irradiated patients and in 58% of the external beam-irradiated group. Haye and co-workers35 have reported that 18 (23%) of 79 of their patients with choroidal melanomas treated with cobalt irradiation developed a severe maculopathy with visual acuity of 20/200 or worse. Capillary obliteration, exudation, and cystoid macular edema were the most common findings reported in their study. Our fluorescein angiographic findings are consistent with earlier studies. 5 ,18,19 The fluorescein angiographic hallmarks of radiation retinopathy include capillary nonperfusion, telangiectasia, microaneurysms, and macular edema. Irvine and Wood 36 irradiated monkeys with 3000 rad and described an animal model of radiation retinopathy. Histopathologic examination showed focal loss of capillary endothelial cells and pericyctes. This lesion was followed by confluent capillary cell loss, nerve fiber layer infarcts, and capillary nonperfusion. The deeper and smaller retinal vessels were involved earlier. Intraretinal neovascularization was observed, but neovascularization of the disc and NVE were not present. Interestingly, one histopathologic case report described a patient with subretinal neovascularization secondary to telangiectatic retinal vessels. 1 In our series, NVE was encountered in only 13 (6%) eyes. Several reports describe the histopathologic findings of choroidal and/or ciliary body melanomas treated with
Table 3. Resolution Of Signs Of Radiation Maculopathy Lesion Mascular edema Microvascular changes Hemorrhage Capillary nonperfusion Lipid NFL infarcts Sheathing
No. with No. Diagnosed Follow-up (%)
No. Resolved (%)
175
138 (79)
7 (5)
147 134
103 (70) 95 (71)
8 (8) 38 (40)
106 104 94 79
69 (65) 71 (68) 68 (72) 47 (59)
5 (7) 28 (39) 49 (68) 12 (25)
1283
Ophthalmology
Volume 99, Number 8, August 1992
proton beam irradiationY-41 These findings include endothelial cell swelling, decreased size of the blood vessel lumen, basement membrane thickening, sinusoidal vessel collapse, vessel thrombosis, perivascular fibrin deposits, lipid droplets in tumor cells, and lymphocytic infiltration. In several of these cases, it could not be determined if the retinal findings were radiation-induced or secondary to chronic retinal detachment. Ferry and co-workers38 observed edema and vascular changes only in areas adjacent to the tumor. Kincaid and associates 39 also observed localized changes. They found total hyalinization of the retina and choroid in the irradiated area in three of five cases. In one case, these authors observed diffuse retinopathy with severe atrophy, vascular occlusion, fibrosis, and cholesterol plaque formation. Saomil and co-workers41 recently reported on the histopathologic results of a large series of patients with uveal melanomas treated with proton beam irradiation. The most common retinal lesions observed by these authors were neovascularization, intraretinal hemorrhages, protein exudate, lipid deposition, cystoid bodies, and blood vessel damage (thickening and thrombosis). Retinal lesions were usually noted only overlying or in close proximity to the tumor. These findings may reflect in part the fact that most eyes had been enucleated because of complications, mainly neovascular glaucoma. Our clinical findings in patients with radiation retinopathy are very similar to those seen with diabetes mellitus. In both radiation maculopathy and diabetic retinopathy, early changes include microaneurysms, intraretinal hemorrhages, exudates, cotton-wool spots, capillary nonperfusion, and macular edema. However, proliferative neovascularization occurs much more frequently in diabetes mellitus. In addition, papillopathy is often seen with radiation maculopathy, but not with diabetic retinopathy. This study demonstrates that radiation maculopathy is common after proton beam irradiation of paramacular choroidal melanomas. Previous studies have shown that radiation maculopathy is a leading cause of visual loss in this group of patients. However, despite the high prevalence of radiation maculopathy in this series, two thirds of patients retained visual acuity of 20/200 or better by the end of postirradiation year 3. Our findings suggest that although radiation maculopathy occurs frequently in eyes with paramacular melanomas treated with proton beam irradiation, ambulatory vision is preserved in many eyes.
References 1. Boozalis GT, Schachat AP, Green WR. Subretinal neovascularization from the retina in radiation retinopathy. Retina 1987;7:156-61. 2. Cogan DG. Lesions of the eye from radiant energy. JAMA 1950; 142: 145-51. 3. Shukovsky D, Fletcher GH. Retinal and optic nerve complications in a high dose irradiation technique of ethmoid sinus and nasal cavity. Radiology 1972; 104:629-34.
1284
4. Stallard HB. Radiant energy as (a) a pathogenic and (b) a therapeutic agent in ophthalmic disorders. Gifford Edmonds prize essay for 1932. Br J Ophthalmol 1933;Monograph SuppI6:1-126. 5. Brown GC, Shields JA, Sanborn G, et al. Radiation retinopathy. Ophthalmology 1982;89: 1494-150 I. 6. Brown GC, Shields JA, Sanborn G, et al. Radiation retinopathy. Trans PA Acad Ophthalmol Otolaryngol1981 ;34: 144-51. 7. Tomsak RL, Smith JL. Radiation retinopathy in a patient with lung carcinoma metastatic to brain. Ann Ophthalmol 1980; 12:619-22. 8. Wara WM, Irvine AR, Neger RE, et al. Radiation retinopathy. Int J Radiat Oncol Bioi Phys 1979;5:81-3. 9. Bagan SM, Hollenhorst RW. Radiation retinopathy after irradiation of intracranial lesions. Am J Ophthalmol 1979;88:694-7. 10. Egbert PR, Donaldson SS, Moazed K, Rosenthal AR. Visual results and ocular complications following radiotherapy for retinoblastoma. Arch Ophthalmol 1978;96: 1826-30. II. Rotman M, Long RS, Packer S, et al. Radiation therapy of choroidal melanoma. Trans Ophthalmol Soc UK 1977;97: 431-5. 12. Char DH, Lonn LI, Margolis LW. Complications of cobalt plaque therapy of choroidal melanomas. Am J Ophthalmol 1977;84:536-41. 13. Harris JR, Levene MB. Visual complications following irradiation for pituitary adenomas and craniopharyngiomas. Radiology 1976;120:167-71. 14. Chan RC, Shukovsky D. Effects of irradiation on the eye. Radiology 1976;120:673-5. 15. Lommatzsch P. Treatment of choroidal melanomas with I06RujI06Rh beta-ray applicators. Surv OphthalmoI1974;19: 85-100. 16. MacFaul PA, Bedford MA. Ocular complications after therapeutic irradiation. Br J Ophthalmol 1970;54:237-47. 17. Bedford MA, Bedotto C, MacFaul PA. Radiation retinopathy after the application of a cobalt plaque. Report ofthree cases. Br J Ophthalmol 1970;54:505-9. 18. Hayreh SS. Post-radiation retinopathy. A fluorescence fundus angiographic study. Br J Ophthalmol 1970;54:705-14. 19. Gass JDM. A fluorescein angiographic study of macular dysfunction secondary to retinal vascular disease. VI. xray irradiation, carotid artery occlusion, collagen vascular disease, and vitritis. Arch Ophthalmol 1968;80:606-17. 20. Chee PHY. Radiation retinopathy. Am J Ophthalmol 1968;66:860-5. 21. Stallard HB. Radiotherapy for malignant melanoma of the choroid. Br J Ophthalmol 1966;50:147-55. 22. Perrers-Taylor M, Brinkley D, Reynolds T. Choroido-retinal damage as a complication of radiotherapy. Acta Radiol Ther Phys Bioi 1965;3:431-40. 23. Rosengren B. Two cases of atrophy of the optic nerve after previous roentgen treatment of the chiasmal region and the optic nerves. Acta Ophthalmol 1958;36:874-7. 24. Cibis PA, Noell WK, Eichel B. Ocular effects produced by high-intensity x-radiation. Arch Ophthalmol 1955;53:65163. 25. Flick 11. Ocular lesions following the atomic bombing of Hiroshima and Nagasaki. Am J Ophthalmol 1948;31:13754. 26. Stallard HB. Radiotherapy of malignant intra-ocular neoplasms. Br J Ophthalmol 1948;32:618-39. 27. Moore RF. Presidential address. Trans Ophthalmol Soc UK 1935;55:3-26. 28. Gragoudas ES, Goitein M, Koehler AM, et al. Proton ir-
Guyer et al . Radiation Maculopathy after Proton Beam Irradiation
29. 30. 31. 32. 33. 34. 35.
radiation of choroidal melanomas. Preliminary results. Arch Ophthalmol 1978;96: 1583-91. Gragoudas ES, Goitein M, Verhey L, et al. Proton beam irradiation: an alternative to enucleation for intraocular melanomas. Ophthalmology 1980;87:571-8l. Goitein M, Miller T. Planning proton therapy of the eye. Med Phys 1983;10:275-83. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457-8l. Peto R, Peto J. Asymptotically efficient rank invariant test procedures. J Roy Stat Soc 1972;135:185-207. Seddon JM, Gragoudas ES, Polivogianis L, et al. Visual outcome after proton beam irradiation of uveal melanoma. Ophthalmology 1986;93:666-74. Seddon JM, Gragoudas ES, Egan KM, et al. Uveal melanomas near the optic disc or fovea. Visual results after proton beam irradiation. Ophthalmology 1987;94:354-6l. Haye C, Desjardins L, Bouder P, et al. Maculopathies par radiations chez les patients traites pour melanomes choroidiens. Ophthalmologie 1990;4:229-31.
36. Irvine AR, Wood IS. Radiation retinopathy as an experimental model for ischemic proliferative retinopathy and rubeosis iridis. Am J Ophthalmol 1987;103:790-7. 37. Seddon JM, Gragoudas ES, Albert DM. Ciliary body and choroidal melanomas treated by proton beam irradiation. Histopathologic study of eyes. Arch OphthalmoI1983;101: 1402-18. 38. Ferry AP, Blair CJ, Gragoudas CS, Volk Sc. Pathologic examination of ciliary body melanoma treated with proton beam irradiation. Arch Ophthalmol 1985; 103: 1849-53. 39. Kincaid Me, Folberg R, Torczynski E, et al. Complications after proton beam therapy for uveal malignant melanoma. A clinical and histopathologic study of five cases. Ophthalmology 1988;95:982-91. 40. Zinn KM, Stein/Pokorny K, Jakobiec FA, et al. Protonbeam irradiated epithelioid cell melanoma of the ciliary body. Ophthalmology 1981;88:1315-2l. 41. Saornil Alvarez MA, Egan KM, Gragoudas ES, et al. Histopathology of proton beam irradiated versus enucleated uveal melanomas. Arch Ophthalmol 1992; [in press].
1285
