Plaque radiotherapy is the most common treatment modality for uveal melanoma.1,2 In a study of 638 patients undergoing plaque radiotherapy, the Collaborative Ocular Melanoma Study (COMS) demonstrated Kaplan-Meier estimates of tumor control in 90% and globe salvage in 88% at 5 years.3 The COMS group also studied long-term mortality outcomes after plaque radiotherapy and showed a 12-year rate of death from uveal melanoma metastasis at 21%.4 However, radiotherapy of uveal melanoma exposes the surrounding normal ocular tissues to potential radiation-related damage, often resulting in decreased vision.2–6
Several studies have previously examined visual acuity outcomes after plaque radiotherapy.7–9 In a comprehensive study of 1106 patients with plaque-irradiated uveal melanoma, Shields et al found poor visual acuity (defined as 20/200 or worse) in 34% of patients at 5 years and 68% of patients at 10 years follow-up.7 In a separate study of 1780 patients with small choroidal melanoma, measuring 3 mm in thickness or less, patients treated with plaque radiotherapy had a 10-year risk of experiencing ≥3 lines of Snellen visual acuity loss in 49%.10
More recently, anti-vascular endothelial growth factors (VEGF) have been employed to reduce tumor-related and radiation-related ischemic retinal findings. Shah et al explored the visual benefits of additional prophylactic anti-VEGF (intravitreal bevacizumab) in 292 patients with plaque-irradiated uveal melanoma compared with controls who were initially observed and found (bevacizumab group vs control group) a reduction in the 2 year rate of optical coherence tomography (OCT)-evident macular edema (26% vs 40%, P = 0.004), clinically evident radiation maculopathy (16% vs 31%, P = 0.001), moderate vision loss (33% vs 57%, P < 0.001), and poor visual acuity (Snellen <5/200) (15% vs 28%, P = 0.004),11 suggesting the benefit of prophylactic intravitreal bevacizumab injections for prevention of radiation side effects and maintenance of visual acuity. Herein, we study a larger cohort of 1131 patients over a longer period of time to further evaluate prophylactic bevacizumab after plaque-irradiated uveal melanoma and specifically investigate visual outcomes, tumor response, and radiation complications based on patient age.
A retrospective chart review was performed on all patients with uveal melanoma managed with plaque radiotherapy and additional prophylactic intravitreal bevacizumab (1.25 mg in 0.05 mL) injections at the Ocular Oncology Service (Wills Eye Hospital, Philadelphia, PA) between July 7, 2000 and January 19, 2018. Patients with cystoid macular edema at presentation that required therapeutic intravitreal bevacizumab injections were excluded from analysis. The study was approved by the Wills Eye Hospital Institutional Review Board and was adherent to the tenets of the Declaration of Helsinki. Informed consent was obtained from all participants in this study. All patients had comprehensive ophthalmic examination including slit lamp biomicroscopy, indirect ophthalmoscopy, and imaging with fundus photography, ocular ultrasonography, fundus autofluorescence, and optical coherence tomography.
Patients were categorized into 3 age groups including younger (<50 years), middle aged (50–70 years, inclusive), and older (>70 years) patients. Demographic and initial clinical features included patient age, involved eye, race, sex, medical history, ocular history, best-corrected visual acuity, and presenting symptoms. Snellen visual acuity was defined as good (20/20–20/40), intermediate (20/50–20/150), and poor (20/200 and worse). LogMAR visual acuity was recorded and expressed as a Snellen visual acuity equivalent. Recorded tumor features included distance from the optic disc and foveola in millimeters (mm), largest basal diameter (millimeter), thickness (millimeter), presence of Bruch membrane rupture, episcleral sentinel vessels, extraocular extension, anteroposterior location of epicenter, quadrantic location, location of the anterior margin, and configuration. Presence of subretinal fluid was determined by clinical examination and optical coherence tomography. Presence and extent of orange pigmentation over the tumor was determined by funduscopy and autofluorescence imaging. Tumor classification was recorded based on the American Joint Committee on Cancer (AJCC) staging system, 8th edition.12
All patients were treated with customized Iodine-125 (I125) plaque radiotherapy. Plaque features included applicator shape, total treatment duration (hours), total dose [centiGray (cGy)] to tumor apex, base, optic disc, foveola, lens, sclera, and opposite retina, and dose rate (cGy/hour) to the tumor apex, base, optic disc, foveola, lens, sclera, and opposite retina.
All patients were counseled on the risks and benefits of intravitral injections and then offered prophylactic intravitreal bevacizumab injections at the time of plaque removal followed by at 4 month intervals for a duration of 2 years. Only patients who consented to prophylactic injections were included in this study to control for this variable between comparison groups. Prophylactic injections were continued at 4-month intervals as long as the ophthalmic examination was unremarkable for clinical features of radiation damage [eg, radiation maculopathy, radiation retinopathy, or cystoid macular edema (CME)]. Once features of radiation damage were detected, patients were offered therapeutic injections. Prophylactic injections were discontinued if the patient refused additional injections or required injection of anti-vascular endothelial growth factor (anti-VEGF) agents for therapeutic treatment of post-radiation complications. The total number of prophylactic injections received by each patient was recorded. Injections given to treat radiation complications (eg, cystoid macular edema) were not considered prophylactic.
Clinical outcomes included final best-corrected visual acuity, total lines of visual acuity loss, final tumor features (basal diameter and thickness), tumor recurrence, interval to recurrence, radiation complications (radiation maculopathy, nonproliferative retinopathy, proliferative retinopathy, papillopathy, cystoid macular edema, retinal vascular occlusion, iris neovascularization, neovascular glaucoma, cataract, and scleral necrosis), and interval to radiation complications. Radiation maculopathy was defined by the presence of retinal microaneurysms, exudation, hemorrhages, and/or nerve fiber layer infarction in the macular region. Radiation retinopathy was defined as similar changes found outside the macular region. Radiation papillopathy was defined by the presence of optic disc congestion, edema, surrounding retinal nerve fiber infarction, or atrophy.
Data were analyzed using SPSS statistics software (version 18 for Windows; SPSS Inc., Chicago, IL). Continuous variables are expressed as mean (median, range). Demographic, clinical, and imaging features were compared by age group (<50 vs 50–70 vs >70 years) using Chi-square test or Fischer exact test for categorical variables and 1-way analysis of variance for continuous variables. Post hoc analysis of continuous variables with Bonferroni test was used to make pairwise comparisons between the age groups. Logistic regression analysis was used to adjust the role of potential confounders on categorical variables. Kaplan-Meier graphs depict the rate of complications after plaque therapy by age group. Hazard ratio (HR) was calculated using Cox regression analysis. A P value <0.05 was considered statistically significant.
There were 1131 eyes in 1131 patients with uveal melanoma. Of these, there were 231 (20%) younger, 657 (58%) middle aged, and 243 (21%) older patients. Demographic features are described in Table 1. A comparison by age category (<50 vs 50–70 vs >70 years) revealed mean age of 40 versus 61 versus 77 years (P < 0.001). Fewer younger patients presented with diabetes (5% vs 14% vs 21%, P < 0.001), systemic hypertension (16% vs 45% vs 60%, P < 0.001), and ocular melanocytosis (2% vs 2% vs 6%, P < 0.03). There was no difference in patient race, sex, or other medical and ocular history (Table 1).
Clinical features at presentation are described in Table 2 . Comparison by age category revealed that a smaller proportion of younger patients presented with moderate visual acuity (20/50–20/150) (22% vs 22% vs 31%, P = 0.04). Presenting mean LogMAR visual acuity (expressed as Snellen visual acuity) was different among the age groups (20/40 vs 20/40 vs 20/50, P = 0.02). Post hoc analysis showed better presenting LogMAR visual acuity in the younger compared with the older age group (P = 0.04) and the middle aged compared with the older age group (P = 0.05). Younger patients presented with more symptoms of decreased visual acuity (40% vs 33% vs 30%, P < 0.001) and visual field defect (13% vs 9% vs 5%, P < 0.001) with fewer younger patients presenting asymptomatically (26% vs 33% vs 45%, P < 0.001). Younger patients were more likely to present with subretinal fluid (83% vs 78% vs 71%, P = 0.006) with greater extension beyond the tumor margin (32% vs 22% vs 17%, P = 0.04), more likely to have subretinal fluid involvement of 3 or more quadrants (18%, 8%, 4%, P < 0.001), and subretinal fluid extension under the foveola (35% vs 29% vs 19%, P = 0.001). There was no difference by age group in tumor laterality, distance to optic disc or foveola, tumor thickness, presence of Bruch membrane rupture, sentinel vessels, or extraocular extension, anteroposterior and quadrantic location of tumor epicenter, tumor configuration, orange pigment presence and extent, and AJCC 8th edition classification.
Treatment features are described in Table 3. A comparison by age category revealed younger patients more likely to receive recommended bevacizumab injections (4.0 vs 4.2 vs 3.8, P = 0.03), with post hoc analysis revealing a greater number of injections received by the middle aged compared with the older age group (P = 0.04). Compared with the older age group, the younger age group had a greater total dosage of radiation directed to the tumor base (196 Gy vs 184 Gy vs 173 Gy, P = 0.02), optic disc (40 Gy vs 38 Gy vs 34 Gy, P = 0.04), foveola (62 Gy vs 58 Gy vs 49 Gy, P = 0.02), and sclera (196 Gy vs 184 Gy vs 173 Gy, P = 0.02). Younger patients also had higher dose rate to the tumor base (195 cGy/h vs 186 cGy/h vs 173 cGy/h, P = 0.01), foveola (62 cGy/h vs 59 cGy/h vs 51 cGy/h, P = 0.04), and sclera (195 cGy/hr vs 185 cGy/hr vs 173 cGy/h, P = 0.01).
Clinical outcomes are described in Table 4. A comparison by age category revealed a greater follow-up time in both the younger and middle age groups (40 vs 42 vs 32 months, P < 0.001). Comparison showed fewer younger and middle age patients with moderate final Snellen visual acuity (20/50–20–150) (22% vs 26% vs 34%, P = 0.02). A greater proportion of the younger age group developed radiation complications including radiation maculopathy (48% vs 39% vs 28%, P < 0.001), nonproliferative extramacular radiation retinopathy (30% vs 25% vs 16%, P = 0.002), radiation papillopathy (21% vs 18% vs 12%, P = 0.03), cystoid macular edema (39% vs 38% vs 29%, P = 0.03), branch retinal vein occlusion (31% vs 29% vs 19%, P = 0.03), and cataract (25% vs 30% vs 20%, P = 0.004) (Figs. 1 and 2). After adjusting for potential confounding variables including distance to optic disc and foveola, tumor basal diameter and thickness at presentation, radiation dose to tumor apex, base, and foveola, extent of subretinal fluid, and number of prophylactic bevacizumab injections using logistic regression analysis, the association between age and radiation maculopathy (P < 0.001), nonproliferative extramacular radiation retinopathy (P = 0.001), and cystoid macular edema (P = 0.04) remained significant, but the associations between age and radiation papillopathy (P = 0.06) and cataract (P = 0.87) were nonsignificant. The younger patients had the shortest time to onset of radiation maculopathy (25 vs 30 vs 28 months, P = 0.03) and longest time to cataract development (32 vs 21 vs 17 months, P = 0.03). There was no difference by age group in need for enucleation, lines of visual acuity lost at last follow-up, and presence of (and interval to) tumor recurrence, proliferative radiation retinopathy, iris neovascularization, neovascular glaucoma, and scleral necrosis.
Figure 3 depicts the 48-month Kaplan-Meier estimates of radiation maculopathy, radiation retinopathy, radiation papillopathy, cystoid macular edema, and poor visual acuity. Comparing younger versus middle aged versus older age groups, the younger age group showed a higher 48-month probability of radiation maculopathy (62% vs 46% vs 47%, P = 0.001) and extra-macular radiation retinopathy (36% vs 34% vs 29%, P = 0.03). After adjusting for tumor base at presentation, location of tumor anterior margin, distance of plaque from macula, number of prophylactic intravitreal injections, and length of follow-up, Cox regression analysis showed that the 48-month HR for the development of radiation maculopathy between younger versus middle-aged was 1.5 (P = 0.001), younger versus older was 1.6, (P = 0.005), and middle-aged versus older was 1.2 (P = 0.31). After adjusting for tumor base at presentation and number of prophylactic injections, the 48-month HR for radiation retinopathy between younger versus middle-aged patients was 1.2 (P = 0.17), younger versus older patients was 1.5 (P = 0.04), and middle-aged versus older patients was 1.3 (P = 0.15). There were no differences between age groups across the 48-month follow-up period regarding the estimated risk of radiation papillopathy, cystoid macular edema, and poor visual acuity.
The retinal vasculature is prone to pathologic changes after hypoxic events from an underlying disease or an environmental insult.13 Similar to other ophthalmic vascular diseases, such as diabetic retinopathy, radiation toxicity can cause retinal capillary leakage, nonperfusion, and sometimes, neovascularization, predisposing the retinal vasculature to leakage and bleeding.13 The risk of developing such complications, in the setting of diabetes or after radiation therapy, has been associated with younger patient age.14–16
Radiation-induced vasculopathy manifests with retinal microaneurysms, exudation, hemorrhages, edema, and nerve fiber layer infarction after an average latent period of 14 months.15–17 Disruption of the retinal capillary endothelium leads to downstream activation of the clotting cascade, resulting in nonperfusion and ultimate retinal and optic disc ischemia.16 These complications are fairly common after treatment of uveal melanoma with plaque radiotherapy. In a series of 630 patients treated with plaque radiotherapy for macular uveal melanoma, Gündüz et al reported the 5-year risk of developing radiation maculopathy and papillopathy at 40% and 13%, respectively.18 In a separate study of 1300 patients treated with plaque radiotherapy for posterior uveal melanoma, Gündüz et al found a 46% 5-year risk of nonproliferative radiation retinopathy.19 These reports date back to the pre-anti-VEGF era when bevacizumab or similar drugs were not available for prevention or treatment of radiation maculopathy.
Age as a predictive factor for the development of radiation-induced vascular complications has been reported in both the systemic and ophthalmic literature.14–16,20 In a study of 2232 patients who received mediastinal radiotherapy for Hodgkin lymphoma, the risk of fatal acute myocardial infarction was higher in patients younger than 19 years compared with patients older than 50-years (relative risk of 45 vs 1.5, P < 0.001).20 In a cohort of 1117 patients with diabetes mellitus, younger age at presentation with systemic diabetes (<15 years) was associated with a higher long-term risk of proliferative diabetic retinopathy compared with older patients (≥15 years old).14 Based on a series of 558 patients with uveal melanoma treated with proton beam radiotherapy, evaluated by Gragoudas et al, younger age at presentation, along with presence of diabetes mellitus, hypertension, and shorter tumor distance to the foveola and/or optic disc, was identified as a predictive factor for development of radiation maculopathy and papillopathy.15 Later, Krema et al, from a cohort of 300 patients with uveal melanoma treated with I125 plaque radiotherapy, identified similar predictive factors, including young patient age, for the development of radiation retinopathy.16 However, little is known about the specific effect of patient age at presentation on the risk of development of the spectrum of radiation complications after plaque radiotherapy of uveal melanoma, especially in this era of prophylactic intravitreal bevacizumab injection.
In this study, we focused only on the impact of patient age regarding the outcomes of radiation complications after plaque radiotherapy of uveal melanoma in the anti-VEGF era in which patients were prescribed to receive a total of 7 prophylactic intravitreal injections spread uniformly over a 2-year period. We did not find differences in final visual acuity or lines of visual acuity lost between age groups. However, we found a greater proportion of younger patients developed complications including radiation maculopathy, extramacular retinopathy, papillopathy, branch retinal vein occlusion, and cystoid macular edema, even after adjusting for radiation dosages. Only cataract formation was found to be higher in older patients. By survival analysis over a 48-month period, the younger age group had a greater risk of developing radiation maculopathy compared with the middle-aged (HR = 1.5) and older (HR = 1.6) age groups and a greater risk of developing extramacular radiation retinopathy compared with the older age group (HR = 1.5).
Vascular endothelial growth factor (VEGF) is an important inflammatory marker released in response to hypoxia, with higher levels in eyes after a vascular insult, such as radiation.13,21 Missotten et al found higher intraocular VEGF levels in eyes with uveal melanoma after radiation (median 364 pg/mL) compared with nonirradiated eyes (median of 146.5 pg/mL).21 Although an appropriate physiologic stress response to elicit angiogenesis, increased exposure to VEGF can cause pathologic changes in the retinal vasculature.13 Rivard et al described age-related variations in VEGF expression and found that the in vitro VEGF levels in vascular smooth muscle cells exposed to hypoxic conditions in old rabbits was significantly lower than young rabbits (100% vs 213%, P < 0.05).22 Greater intraocular VEGF levels in younger patients after radiotherapy could be partly responsible for the higher risk of radiation complications in this age group.
Currently, we offer intravitreal bevacizumab at 4-month intervals over a 2-year period as prophylaxis for prevention or reduction of radiation-induced side effects after plaque radiotherapy for uveal melanoma. Intravitreal bevacizumab injections have been previously studied for the treatment of radiation complications with varying results on visual acuity.6,23 When beginning treatment, injections are typically administered at 1-month intervals. Although the ideal interval for use of intravitreal bevacizumab injections as prophylaxis has not been thoroughly investigated, our group has studied the efficacy of intravitreal bevacizumab for prevention of these side effects with promising results when used at a 4-month interval.11 Of 292 patients receiving prophylactic intravitreal bevacizumab in our early series, fewer radiation complications of CME (26% vs 40%, P = 0.004) and radiation maculopathy (16% vs 31%, P = 0.001) were observed compared with controls managed without bevacizumab over a 2-year period.11 Later, we continued this study over a 4-year period and found a lower 4-year Kaplan-Meier estimated risk of developing clinically evident radiation maculopathy in patients receiving prophylactic intravitreal bevacizumab compared with a control group managed with bevacizumab (HR = 1.39, P = 0.03).24 Furthermore, over the 4-year period, the bevacizumab group demonstrated better median visual acuity compared with the control group (20/70 vs counting fingers, P < 0.001).
Herein, we have studied the impact of age on radiation complications in the setting of prophylactic anti-VEGF therapy after plaque radiotherapy. The age factor has been previously studied regarding intravitreal bevacizumab for neovascular age-related macular degeneration. Wang et al documented an association of older age with better treatment response and younger age with poorer response.25 In this analysis, we found that younger-aged patients had a greater risk of developing radiation maculopathy and extramacular retinopathy despite the use of prophylactic intravitreal anti-VEGF. However, whether there is an age-related difference with anti-VEGF response for the prevention of post-radiation complication deserves further study.
Limitations of this study include its retrospective design and lack of complete 48-month follow-up in all patients. In addition, not all patients received a total of 6 prophylactic injections for several reasons including, most commonly, patient co-management elsewhere and patient preference. We also recognize that our results could be confounded by presence of cataract or age-associated changes in visual acuity. This could explain why no difference was detected in the visual acuity outcome by age group despite differences in frequency of radiation side effects. Strengths of this study, however, were the large number of patients and long-term follow-up at a single-center with robust statistical analysis.
In summary, we have studied the impact of age on radiation complications and visual outcomes in patients with plaque-irradiated uveal melanoma and additional prophylactic intravitreal bevacizumab and have documented a greater frequency of radiation complications, including radiation maculopathy, extra-macular retinopathy, papillopathy, and cystoid macular edema, in the younger age category (<50 years). The mechanisms underlying the observed higher risk of post-radiation complications in younger patients deserve further study.
1. Shields JA, Shields CL. Management of posterior uveal melanoma: past, present, and future. The 2014 Charles L. Schepens Lecture. Ophthalmol
2. Kaliki S, Shields CL. Uveal melanoma: Relatively rare but deadly cancer. Eye (Lond)
3. Collaborative Ocular Melanoma Study Group. Collaborative Ocular Melanoma Study (COMS) randomized trial of iodine 125 brachytherapy for choroidal melanoma: IV. Local treatment failure and enucleation in the first 5 years after brachytherapy. COMS report no. 19. Ophthalmol
4. Collaborative Ocular Melanoma Study Group. Collaborative Ocular Melanoma Study (COMS) randomized trial of iodine 125 brachytherapy for choroidal melanoma: V. Twelve-year mortality rates and prognostic factors:. COMS report no. 28. Arch Ophthalmol
5. Jensen AW, Petersen IA, Kline RW, et al. Radiation complications and tumor control after 125I plaque brachytherapy for ocular melanoma. Int J Radiat Oncol Biol Phys
6. Wen JC, McCannel TA. Treatment of radiation retinopathy
following plaque brachytherapy for choroidal melanoma. Curr Opin Ophthalmol
2009; 20 (3):200–204.
7. Shields CL, Shields JA, Cater J, et al. Plaque radiotherapy
for uveal melanoma: long-term visual outcome in 1106 consecutive patients. Arch Ophthalmol
8. Collaborative Ocular Melanoma Study Group. Collaborative Ocular Melanoma Study (COMS) randomized trial of iodine 125 brachytherapy for medium choroidal melanoma: I. Visual acuity after 3 years. COMS report no. 16. Ophthalmol
9. Wisely CE, Hadziahmetovic M, Reem RE, et al. Long-term visual acuity outcomes in patients with uveal melanoma treated with 125I episclearal OSU-Nag Plaque Brachytherapy. Brachytherapy
10. Shields CL, Sioufi K, Srinivasan A, et al. Visual Outcome and millimeter incremental risk of metastasis in 1780 patients with small choroidal melanoma managed by plaque radiotherapy
. JAMA Ophthalmol
11. Shah SU, Shields CL, Bianciotto CG, et al. Intravitreal bevacizumab at 4-month intervals or prevention of macular edema after plaque radiotherapy
of uveal melanoma. Ophthalmol
12. Amin MB, Edge SB, Greene FL, et al. AJCC Cancer Staging Manual. Springer, 8th edNew York: 2017.
13. Miller JW, Couter JL, Strauss EC, et al. Vascular endothelial growth factor A in intraocular vascular disease. Ophthalmol
14. Hietala K, Harjutsalo V, Forsblom C, et al. Age
at onset and the risk of proliferative retinopathy in type 1 diabetes. Diabetes Care
15. Gragoudas ES, Li W, Lane AM, et al. Risk factors of radiation maculopathy and papillopathy after intraocular irradiation. Ophthalmol
16. Krema H, Xu W, Payne D, et al. Factors predictive of radiation retinopathy
post 125Iodine brachytherapy of uveal melanoma. Can J Ophthalmol
17. Brown GC, Shields JA, Sanborn G, et al. Radiation retinopathy
18. Gündüz K, Shields CL, Shields JA, et al. Radiation complications and tumor control after plaque radiotherapy
of choroidal melanoma with macular involvement. Am J Ophthalmol
19. Gündüz K, Shields CL, Shields JA, et al. Radiation retinopathy
following plaque radiotherapy
for posterior uveal melanoma. Arch Ophthalmol
20. Hancock SL, Tucker MA, Hoppe RT. Factors affecting late mortality from heart disease after treatment of Hodgkin's disease. JAMA
21. Missotten GS, Notting IC, Schlingemann RO, et al. Vascular endothelial growth factor a in eyes with uveal melanoma. Arch Ophthalmol
22. Rivard A, Berthou-Soulie L, Principe N, et al. Age
-dependent defect in vascular endothelial growth factor expression is associated with reduced hypoxia-inducible factor 1 activity. J Biol Chem
23. Finger PT, Chin KJ, Semenova EA. Intravitreal anti-VEGF therapy for macular radiation retinopathy
: a 10-year study. Eur J Ophthalmol
24. Shields CL, Dalvin LA, Chang M, et al. Visual outcome at 4 years following plaque radiotherapy
and prophylactic intravitreal bevacizumab (every 4 months for 2 years) for uveal melanoma: comparison with nonrandomized historical control individuals. JAMA Ophthalmol
2019; Epub ahead of print.
25. Wang H, Barteselli G, Freeman WR, et al. Temporal pattern of resolution/recurrence of choroidal neovascularization during bevacizumab therapy for wet age
-related macular degeneration. Int J Ophthalmol