- Email: [email protected]
Plaque Brachytherapy for Uveal Melanoma: A Vision Prognostication Model
International Journal of
Radiation Oncology biology
physics
www.redjournal.org
Clinical Investigation: Central Nervous System Tumor
Plaque Brachytherapy for Uveal Melanoma: A Vision Prognostication Model Niloufer Khan, BA,* Mohammad K. Khan, MD, PhD,x James Bena, MS,y Roger Macklis, MD,* and Arun D. Singh, MDz *Department of Radiation Oncology, Taussig Cancer Center, yDepartment of Quantitative Health Sciences, zDepartment of Ophthalmic Oncology, Cole Eye Institute, Cleveland Clinic, Cleveland, Ohio; and xDepartment of Radiation Oncology, Emory University School of Medicine, Atlanta, Georgia Received Sep 3, 2011, and in revised form Mar 16, 2012. Accepted for publication Apr 5, 2012
Summary A visual prognostication model for predicting visual acuity after treatment with I-125 or Ru-106 plaque brachytherapy for uveal melanoma was generated, based on the experience of treating 189 patients. The resulting nomogram provides a means to predict vision loss at 3 years after reatment, which may have an impact on patient selection and treatment counseling.
Purpose: To generate a vision prognostication model after plaque brachytherapy for uveal melanoma. Methods and Materials: All patients with primary single ciliary body or choroidal melanoma treated with iodine-125 or ruthenium-106 plaque brachytherapy between January 1, 2005, and June 30, 2010, were included. The primary endpoint was loss of visual acuity. Only patients with initial visual acuity better than or equal to 20/50 were used to evaluate visual acuity worse than 20/50 at the end of the study, and only patients with initial visual acuity better than or equal to 20/200 were used to evaluate visual acuity worse than 20/200 at the end of the study. Factors analyzed were sex, age, cataracts, diabetes, tumor size (basal dimension and apical height), tumor location, and radiation dose to the tumor apex, fovea, and optic disc. Univariate and multivariable Cox proportional hazards were used to determine the influence of baseline patient factors on vision loss. Kaplan-Meier curves (log rank analysis) were used to estimate freedom from vision loss. Results: Of 189 patients, 92% (174) were alive as of February 1, 2011. At presentation, visual acuity was better than or equal to 20/50 and better than or equal to 20/200 in 108 and 173 patients, respectively. Of these patients, 44.4% (48) had post-treatment visual acuity of worse than 20/50 and 25.4% (44) had post-treatment visual acuity worse than 20/200. By multivariable analysis, increased age (hazard ratio [HR] of 1.01 [1.00-1.03], PZ.05), increase in tumor height (HR of 1.35 [1.22-1.48], P<.001), and a greater total dose to the fovea (HR of 1.01 [1.00-1.01], P<.001) were predictive of vision loss. This information was used to develop a nomogram predictive of vision loss. Conclusions: By providing a means to predict vision loss at 3 years after treatment, our vision prognostication model can be an important tool for patient selection and treatment counseling. Ó 2012 Elsevier Inc.
Reprint requests to: Arun D. Singh, MD, Department of Ophthalmic Oncology, Cole Eye Institute (i3-129), Cleveland Clinic Foundation, 9500
Int J Radiation Oncol Biol Phys, Vol. 84, No. 3, pp. e285ee290, 2012 0360-3016/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.ijrobp.2012.04.005
Euclid Avenue, Cleveland, OH 44195. Tel: (216) 445-9479; Fax: (216) 445-2226; E-mail: [email protected] Conflict of interest: none.
e286 Khan et al.
International Journal of Radiation Oncology Biology Physics
Introduction
only 43% of patients would have visual acuity better than 20/200 (2). Many subsequent studies have also reported loss of visual acuity after treatment with plaque brachytherapy (Table 1) (2, 9-18). The decline in visual acuity can be explained by several factors. Some are consequences of inherent tumor characteristics, such as proximity to the fovea. Therefore, preoperative visual acuity must be taken into account in predictions of postoperative visual acuity. Other mechanisms known to be responsible for a loss of visual acuity include radiation effects on normal ocular structures adjacent to the tumor such as radiation-induced cataract (9), radiation retinopathy (ischemic, exudative, neovascular) (9), radiation choroidopathy (9), and radiation optic neuropathy (9). Radiation effects on the tumor such as retinal or vitreous hemorrhage secondary to tumor necrosis or intratumoral choroidal vasculopathy (toxic tumor syndrome) may also contribute to loss of visual acuity. Although previous studies have explored risk factors associated with the development of radiation retinopathy or radiation optic neuropathy, these radiation sequelae may not occur in isolation. Moreover, specific etiopathogenetic mechanisms for vision loss may be difficult to ascertain or classify in an individual patient. Therefore, determining the factors that influence visual acuity after plaque brachytherapy may be a more reliable way of determining the patient’s prognosis. A vision prognostication model that can predict the visual outcome for an individual patient, based on baseline tumor characteristics and a specific
In a landmark randomized multicenter study initiated in the mid1980s, the Collaborative Ocular Melanoma Study (COMS) established plaque brachytherapy as a viable alternative to enucleation for the treatment of medium-sized choroidal melanomas (defined as 2.5-10.0 mm in apical height and 5-16 mm in largest basal diameter) (1, 2). The study demonstrated a low risk (10.3%) of local treatment failure in patients undergoing brachytherapy (3) and a 5-year tumor-related mortality rate of 11% after enucleation and 9% after brachytherapy (4). These equivalent rates of local control and survival between the 2 treatments firmly established plaque brachytherapy as an appropriate treatment for medium-sized uveal melanomas. Although ruthenium-106 (5-7), a beta-emitter, has been widely used in European countries, the COMS study used iodine-125. The study initially prescribed a dose of 100 Gy to the tumor apex, which was later revised to 85 Gy based on recommendations of the American Association of Physicists in Medicine (1). Proton therapy has also been reported to provide effective local control rates (8). Along with proton therapy, plaque brachytherapy has now become the favored treatment alternative to enucleation for medium-sized melanomas because it is preferred by patients and can potentially preserve some degree of visual acuity in the treated eye. In practice, however, plaque brachytherapy leads to a significant loss of visual acuity. The COMS study showed that at 3 years,
Table 1
Loss of visual acuity in patients treated with I-125 brachytherapy for treatment of uveal melanoma* No. of patients
Median follow-up time (mo)
VA worse than 20/200 (% patients)
Factors associated with decreased VA
Sia et al, 2000 (10) Melia et al, 2001 (2)
50 623
39.5 -
37% 45%
Jones et al, 2002 (11)
63
No significant predictors found Tumor-associated retinal detachment Tumor not dome-shaped Diabetes mellitus Tumor proximity to macula and optic disc Higher doses to macula and optic disc Macula dose rate 111 11.1 cGy/h Greater tumor height Increased age -
Study
Bechrakis et al, 2002 (12)
152
Puusaari et al, 2004 (14)
96
36
27.1 (mean) 42
-
95.4% 18% of patients with VA above 20/400 49% 56%
Stack et al, 2005 (15) Jensen et al, 2005 (16)
92 156
31 74.4
Krohn et al, 2008 (17)
111
61
60%
Saconn et al, 2010 (18)
62
58.2
22%
189
-
25.4%
Khan et al, 2012
Dose rate to optic disc >60 cGy/h Dose rate to lens <15 cGy/h Atherosclerosis/coronary artery disease Tumor height >8 mm Large tumory Diabetic retinopathyz Poor initial VA in affected eye Greater total dose to fovea Greater tumor height Increased age
Abbreviation: VA Z visual acuity. * The table includes English-language studies published between 1995 and 2010, evaluating visual acuity after I-125 plaque brachytherapy for uveal melanoma. Studies were found by conducting a PubMed search for the terms “uveal melanoma,” “choroidal melanoma,” “ocular melanoma,” “visual acuity,”“I-125,” “I125,” and “brachytherapy.” Studies evaluating the efficacy of treatment for radiation retinopathy (not the occurrence of radiation retinopathy) were excluded. Case reports were also excluded. y According to criteria of the Collaborative Ocular Melanoma Study. z Factors associated with shorter time to visual acuity worse than 20/200.
Volume 84 Number 3 2012
Visual prognostication model for plaque brachytherapy e287 performed with plaque simulator treatment planning software (Plaque Simulator v 5.3.4; Bebig GmbH, Berlin, Germany) (19, 20) designed to give a total prescription of 85 Gy to the tumor apex. Isotopes used were iodine-125 (Model MED3631-A/M; North American Scientific Inc, Chatsworth, CA) and ruthenium106 (Models CCA, CCD, and CCB, with diameters of 15.3 mm, 17.9 mm, and 20.2 mm, respectively; Bebig GmbH, Berlin, Germany). As part of the initial workup, all patients underwent a detailed ophthalmoscopic examination, photography, and ultrasonographic A and B scan. Tumor size (basal dimensions and height) was assessed by ophthalmoscopic examination and by ultrasonography, respectively. For the placement of the radioactive plaque, tumor localization was performed by use of transillumination or indirect ophthalmoscopy, based on tumor location. An intraoperative ultrasound was performed to ensure that the tumor margins were covered by the plaque. The patients were evaluated for follow-up 1 week after plaque brachytherapy, then every 3 months throughout the first year, every 6 months throughout the second through fifth years, and then annually thereafter. The best corrected visual acuity was measured at each visit. Exploratory frequency tables and descriptive statistics were created for the outcomes of visual acuity worse than 20/50 and visual acuity worse than 20/200. Categoric factors were described by using frequencies and percentages, and continuous measures were summarized by using median and interquartile range. Kaplan-Meier estimates of survival (freedom from vision loss) and univariable Cox proportional hazards models were performed to identify potential predictors of visual acuity loss. Multivariable Cox proportional hazards models were fit by using variables that showed minimal association in univariable analysis (P<.20). Collinearity diagnostics were performed before model fitting, and predictors were chosen from this set. Among the size measurements, largest basal diameter, circumference, surface area, and percentage of retinal surface were very highly correlated, so only the largest basal diameter was used for multivariable modeling. Based on the findings from the multivariable models, a nomogram was fit for each visual acuity model. Bootstrap estimates of the concordance index were fit for each model, and the calibration curve showing actual vs predicted probabilities of vision loss were fit. Predicted risk was calculated at 1 year and 3 years. These analyses were performed with SAS software (version 9.2; Cary, NC) and the Design package within R software (version 2.8; Vienna, Austria).
Table 2 Summary demographic patient characteristics: visual acuity (VA) worse than 20/50 and worse than 20/200 Features Age, y (median) Sex F M Medical history Diabetes Hypertension Heart disease Ocular comorbidities Cataracts Other conditions*
VA 20/50 (nZ116)
VA <20/200 (nZ29)
65.6
64.0
62 (53.5%) 54 (46.6%)
27 (46.6%) 31 (53.5%)
15 (12.9%) 43 (37.1%) 18 (15.5%)
7 (12.1%) 18 (31.0%) 7 (12.1%)
48 (41.4%) 33 (28.5%)
23 (39.7%) 19 (32.85)
* Other ocular comorbidities were glaucoma, macular degeneration, strabismus, amblyopia, retinal disease, history of injury, and surgery.
prescribed radiation dose, would be a useful management tool. To our knowledge, such a vision prognostication model has not been reported. We present our initial attempt at generating a vision prognostication tool that can predict the chances of retaining visual acuity at 3 years after treatment.
Methods and Materials An internal review board approval was obtained. All 189 patients treated at the Cleveland Clinic for a primary single ciliary body or choroid melanoma with iodine-125 or ruthenium-106 plaque brachytherapy between January 1, 2005, and June 30, 2010, were included. Patients with 2 primary uveal melanomas (1 in the ipsilateral eye and 1 in the contralateral eye) were excluded, as were any patients with nonmelanoma tumors of the ciliary body or choroid. Any patient who received adjuvant ocular therapy such as transpupillary thermotherapy or had received more than 1 plaque insertion for the same tumor were also excluded. Patient characteristics and demographics were extracted from the Cleveland Clinic electronic medical records system (EpicCare EMR, Madison, WI). Information regarding patients’ pretreatment and post-treatment ocular histories was extracted from paper charts at the Department of Ophthalmic Oncology, Cole Eye Institute, Cleveland Clinic. Dosimetry for the plaques was Table 3
Summary tumor characteristics and radiation dose: visual acuity (VA) worse than 20/50 and worse than 20/200 Features
Largest basal diameter (mm) Tumor height (mm) Distance to fovea (mm) Distance to disc (mm) Total dose to disc (Gy) Total dose to macula (Gy) Total dose to lens (Gy) Isotope I-125 Ru-106
VA <20/50 (nZ116) Median (25th, 75th) 12.0 4.0 4.5 4.8 29.8 50.5 9.9
(9.0, 14.0) (2.5, 6.0) (1.5, 9.5) (3.0, 9.5) (18.6, 42.0) (18.6, 97.3) (4.8, 23.7)
92 (79.3%) 24 (20.7%)
VA <20/200 (nZ58) Median (25th, 75th) 12.0 4.9 3.0 4.5 34.4 61.5 13.6
(9.0, 14.5) (3.0, 7.5) (1.5, 7.5) (3.0, 7.5) (23.7, 45.3) (25.7, 110.9) (7.9, 28.6)
51 (87.9%) 7 (12.1%)
International Journal of Radiation Oncology Biology Physics
e288 Khan et al. Table 4 Kaplan-Meier estimates of visual acuity preservation at 12, 24, and 36 months for 20/50 and 20/200 vision Time
No VA loss
20/50 (nZ108) 12 41.1% 24 51.2% 36 56.6% 20/200 (nZ179) 12 16.2% 24 27.0% 36 41.4%
95% CI
No. failed
No. left
(31.1, 51.1) (40.0, 62.3) (44.4, 68.8)
40 46 48
38 18 8
(10.3, 22.0) (18.9, 35.1) (30.2, 52.7)
25 35 43
87 48 22
Abbreviations: CI Z confidence interval; VA Z visual acuity. The analysis is based on 108 patients with VA better than 20/50 at presentation and the 179 patients with VA better than 20/200 at presentation.
Results For the time period of January 1, 2005, to June 30, 2010, 224 cases of uveal melanoma were treated at the Department of Ophthalmic Oncology, Cole Eye Institute, Cleveland Clinic Foundation. Of these, 9 cases of iris melanoma were excluded. Twenty-six additional cases were excluded: 22 in which adjuvant treatment with transpupillary thermotherapy or laser was given and 4 with 2 consecutive plaques. This left 189 cases of ciliary body and choroid melanoma for analysis. Five of these patients did not return for follow-up. These patients were included in the analysis but treated as censored immediately (time 0) for all outcomes. Of the 189 patients, 174 (92%) were alive as of February 1, 2011. Of the 15 deaths, 4 (2%) were attributable to metastatic uveal melanoma, 5 (3%) were unrelated, and 6 (4%) were due to unknown causes. Seven patients (3.2%) in the study cohort required enucleation (tumor recurrence 6; neovascular glaucoma 1). Visual acuity was evaluated, excluding all patients with vision worse than the specified endpoints at presentation. Therefore, 108 patients with visual acuity better than or equal to 20/50 at presentation were included in the analysis for factors associated with visual acuity worse than 20/50, and 173 patients with vision better than or equal to 20/200 were included in the analysis for factors associated with visual acuity worse than 20/200. The demographics of the patients are given in Table 2, and radiation dose distribution is described in Table 3. Kaplan-Meier estimates of visual acuity loss are presented in Table 4. Significant factors contributing to loss of visual acuity on multivariable analysis with reduced Cox proportional hazards Table 5 Hazard ratios, confidence intervals, and P values from reduced Cox proportional hazards model for vision loss to worse than 20/50 Parameter
Hazard ratio
Lower CL
Upper CL
P value
Diabetes mellitus Heart disease Other ocular problems Tumor height Total dose to lens
0.44 0.46 1.73 1.41 0.98
0.16 0.14 0.91 1.13 0.95
1.23 1.50 3.29 1.76 1.01
.12 .20 .097 .002 .14
Abbreviations: Lower CL Z lower limit of confidence interval; Upper CL Z upper limit of confidence interval.
Table 6 Hazard ratios, confidence intervals, and P values from reduced Cox proportional hazards model for vision loss to worse than 20/200 Parameter
Hazard ratio
Lower CL
Upper CL
P value
Female sex Tumor height Total dose (macula)
1.85 1.42 1.01
0.99 1.28 1.01
3.46 1.58 1.02
.055 <.001 <.001
Abbreviations: Lower CL Z lower limit of confidence interval; Upper CL Z upper limit of confidence interval.
models are presented in Tables 5and 6. On multivariable analysis, with regard to visual acuity worse than 20/50, no factors were found to be protective, whereas increased tumor height (hazard ratio [HR] of 1.41 [1.13-1.76], PZ.002) and a history of other ocular comorbidities (HR of 1.73 [0.91-3.32], PZ.097) were found to be detrimental. These other ocular comorbidities included both historical and currently diagnosed systemic conditions such as hypertension, diabetes, and heart disease and ocular conditions such as cataracts, glaucoma, macular degeneration, strabismus, amblyopia, and other retinal diseases, Patients did not undergo specific correction of these comorbidities before radiation therapy. With regard to visual acuity worse than 20/200, female sex (HR of 1.85 [0.99-3.46], PZ.055), increased tumor height (HR of 1.42 [1.28-1.58], P<.001), and a greater total dose to the macula (HR of 1.01 [1.01-1.02], P<.001) were found to be detrimental (Table 4). Nomogram fit was attempted based on the findings from the multivariable models for patients with visual acuity worse than 20/50 at 1 and 3 years and visual acuity worse than 20/200 at 1 and 3 years. A nomogram could not be generated for visual acuity worse than 20/50 because it was not possible to fit a statistically significant model with adequate discrimination or calibration. The nomogram generated for visual acuity worse than 20/200 assigns a certain number of points to sex, tumor height in millimeters (apex), and total dose to the macula in Gray because these were the significant factors associated with this level of vision loss on multivariable analysis (Fig. 1). The total number of points can be added up and correlated with a predicted probability of visual acuity worse than or equal to 20/200 at 1 year and at 3 years. For example, a man with a tumor height of 2 mm and a dose to the macula of 40 Gy would have 15 þ 15 þ 30 Z 60 total points. This number of total points (line 5) correlates with a 5% to 10% chance of visual acuity worse than 20/200 at 1 year and a 30% chance of visual loss below 20/200 at 3 years. A concordance index (0.81) for this nomogram between actual and predicted loss of vision at 1 year indicates a fairly strong fit for this model (Fig. 2).
Discussion Our data set demonstrated decreased visual acuity with increased tumor height and increased total dose to the macula, similar to the findings in many previous studies (12, 17, 21). However, other factors that had been associated with severe vision loss in previous studies, such as proximity to optic disc and fovea (11, 21), total dose to optic disc and fovea, and systemic diseases such as diabetes mellitus (18, 21), were not significant in this data set. Inasmuch as our study was based on a limited number of cases, it is possible that some of the prognostic factors may have been excluded because they did not achieve adequate statistical significance to be included in the model.
Volume 84 Number 3 2012
Fig. 1.
Visual prognostication model for plaque brachytherapy e289
The nomogram for visual acuity loss worse than 20/200. In this nomogram, tumor height is the most significant predictor of risk.
Our data also reveal that the main modifiable factor associated with severe visual loss is the total radiation dose to the macula (Table 6). Therefore, if the radiation dose to the macula could be minimized, visual outcome after brachytherapy could be improved (22). However, a larger data set is required to identify the threshold doses necessary to avoid radiation maculopathy. This is an important consideration for designing new classes of ophthalmic brachytherapy implants or identifying new isotopes for ophthalmic applications. The novel aspect of this study was the generation of a nomogram to predict visual loss, based on patient characteristics, tumor characteristics, and dosimetry, which are available before treatment. Furthermore, although the calibration plot indicates a fairly strong correlation between actual and predicted visual loss, the inclusion of more data points will make it a more robust predictive
Fig. 2. The calibration plot for visual acuity 20/200 or worse. The concordance index for the model of vision 20/200 or worse is 0.81. The actual and predicted risk measures are very good across the range of predicted and actual risk levels.
model. Additional studies based on analysis of published data are under way.
Conclusions The vision prognostic tool is a first step in providing clinicians and patients with a way to estimate visual acuity in a specific patient even before the plaque brachytherapy treatment is delivered. By providing a practical means to predict vision loss at 3 years after treatment, our vision prognostication model can be an important tool for patient selection and treatment counseling.
References 1. Krintz AL, Hanson WF, Ibbott GS, et al. A reanalysis of the Collaborative Ocular Melanoma Study Medium Tumor Trial eye plaque dosimetry. Int J Radiat Oncol Biol Phys 2003;56: 889-898. 2. Melia BM, Abramson DH, Albert DM, et al. Collaborative ocular melanoma study (COMS) randomized trial of I-125 brachytherapy for medium choroidal melanoma: I. Visual acuity after 3 years COMS report no. 16. Ophthalmology 2001;108:348-366. 3. Jampol LM, Moy CS, Murray TG, et al. The 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. Ophthalmology 2002;109: 2197-2206. 4. Diener-West M, Earle JD, Fine SL, et al. The COMS randomized trial of iodine 125 brachytherapy for choroidal melanoma: III. initial mortality findings. COMS Report No. 18. Arch Ophthalmol 2001;119: 969-982. 5. Rouberol F, Roy P, Kodjikian L, et al. Survival, anatomic, and functional long-term results in choroidal and ciliary body melanoma after ruthenium brachytherapy (15 years’ experience with beta-rays). Am J Ophthalmol 2004;137:893-900.
e290 Khan et al. 6. Damato B, Patel I, Campbell IR, et al. Visual acuity after Ruthenium(106) brachytherapy of choroidal melanomas. Int J Radiat Oncol Biol Phys 2005;63:392-400. 7. Summanen P, Immonen I, Kivela T, et al. Radiation related complications after ruthenium plaque radiotherapy of uveal melanoma. Br J Ophthalmol 1996;80:732-739. 8. Gragoudas ES. Proton beam irradiation of uveal melanomas: the first 30 years. The Weisenfeld Lecture. Invest Ophthalmol Vis Sci 2006;47: 4666-4673. 9. Wen JC, Oliver SC, McCannel TA. Ocular complications following I125 brachytherapy for choroidal melanoma. Eye (Lond) 2009;23: 1254-1268. 10. Sia S, Harper C, McAllister I, et al. Iodine-I25 episcleral plaque therapy in uveal melanoma. Clin Experiment Ophthalmol 2000;28:409-413. 11. Jones R, Gore E, Mieler W, et al. Posttreatment visual acuity in patients treated with episcleral plaque therapy for choroidal melanomas: dose and dose rate effects. Int J Radiat Oncol Biol Phys 2002; 52:989-995. 12. Bechrakis NE, Bornfeld N, Zoller I, et al. Iodine 125 plaque brachytherapy versus transscleral tumor resection in the treatment of large uveal melanomas. Ophthalmology 2002;109:1855-1861. 13. Shields CL, Naseripour M, Cater J, et al. Plaque radiotherapy for large posterior uveal melanomas (> or Z 8-mm thick) in 354 consecutive patients. Ophthalmology 2002;109:1838-1849. 14. Puusaari I, Heikkonen J, Kivela T. Effect of radiation dose on ocular complications after iodine brachytherapy for large uveal melanoma:
International Journal of Radiation Oncology Biology Physics
15.
16.
17.
18.
19.
20. 21.
22.
empirical data and simulation of collimating plaques. Invest Ophthalmol Vis Sci 2004;45:3425-3434. Stack R, Elder M, Abdelaal A, et al. New Zealand experience of I125 brachytherapy for choroidal melanoma. Clin Experiment Ophthalmol 2005;33:490-494. 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 2005;63:101-108. Krohn J, Monge OR, Skorpen TN, et al. Posterior uveal melanoma treated with I-125 brachytherapy or primary enucleation. Eye (Lond) 2008;22:1398-1403. Saconn PA, Gee CJ, Greven CM, et al. Alternative dose for choroidal melanoma treated with an iodine-125 radioactive plaque: a singleinstitution retrospective study. Int J Radiat Oncol Biol Phys 2010; 78:844-848. Knutsen S, Hafslund R, Monge OR, et al. Dosimetric verification of a dedicated 3D treatment planning system for episcleral plaque therapy. Int J Radiat Oncol Biol Phys 2001;51:1159-1166. Astrahan MA. A patch source model for treatment planning of ruthenium ophthalmic applicators. Med Phys 2003;30:1219-1228. Finger PT, Chin KJ, Yu GP. Risk factors for radiation maculopathy after ophthalmic plaque radiation for choroidal melanoma. Am J Ophthalmol 2010;149:608-615. Lumbroso-Le Rouic L, Charif Chefchaouni M, Levy C, et al. 125I plaque brachytherapy for anterior uveal melanomas. Eye (Lond) 2004; 18:911-916.
Radiation Oncology biology
physics
www.redjournal.org
Clinical Investigation: Central Nervous System Tumor
Plaque Brachytherapy for Uveal Melanoma: A Vision Prognostication Model Niloufer Khan, BA,* Mohammad K. Khan, MD, PhD,x James Bena, MS,y Roger Macklis, MD,* and Arun D. Singh, MDz *Department of Radiation Oncology, Taussig Cancer Center, yDepartment of Quantitative Health Sciences, zDepartment of Ophthalmic Oncology, Cole Eye Institute, Cleveland Clinic, Cleveland, Ohio; and xDepartment of Radiation Oncology, Emory University School of Medicine, Atlanta, Georgia Received Sep 3, 2011, and in revised form Mar 16, 2012. Accepted for publication Apr 5, 2012
Summary A visual prognostication model for predicting visual acuity after treatment with I-125 or Ru-106 plaque brachytherapy for uveal melanoma was generated, based on the experience of treating 189 patients. The resulting nomogram provides a means to predict vision loss at 3 years after reatment, which may have an impact on patient selection and treatment counseling.
Purpose: To generate a vision prognostication model after plaque brachytherapy for uveal melanoma. Methods and Materials: All patients with primary single ciliary body or choroidal melanoma treated with iodine-125 or ruthenium-106 plaque brachytherapy between January 1, 2005, and June 30, 2010, were included. The primary endpoint was loss of visual acuity. Only patients with initial visual acuity better than or equal to 20/50 were used to evaluate visual acuity worse than 20/50 at the end of the study, and only patients with initial visual acuity better than or equal to 20/200 were used to evaluate visual acuity worse than 20/200 at the end of the study. Factors analyzed were sex, age, cataracts, diabetes, tumor size (basal dimension and apical height), tumor location, and radiation dose to the tumor apex, fovea, and optic disc. Univariate and multivariable Cox proportional hazards were used to determine the influence of baseline patient factors on vision loss. Kaplan-Meier curves (log rank analysis) were used to estimate freedom from vision loss. Results: Of 189 patients, 92% (174) were alive as of February 1, 2011. At presentation, visual acuity was better than or equal to 20/50 and better than or equal to 20/200 in 108 and 173 patients, respectively. Of these patients, 44.4% (48) had post-treatment visual acuity of worse than 20/50 and 25.4% (44) had post-treatment visual acuity worse than 20/200. By multivariable analysis, increased age (hazard ratio [HR] of 1.01 [1.00-1.03], PZ.05), increase in tumor height (HR of 1.35 [1.22-1.48], P<.001), and a greater total dose to the fovea (HR of 1.01 [1.00-1.01], P<.001) were predictive of vision loss. This information was used to develop a nomogram predictive of vision loss. Conclusions: By providing a means to predict vision loss at 3 years after treatment, our vision prognostication model can be an important tool for patient selection and treatment counseling. Ó 2012 Elsevier Inc.
Reprint requests to: Arun D. Singh, MD, Department of Ophthalmic Oncology, Cole Eye Institute (i3-129), Cleveland Clinic Foundation, 9500
Int J Radiation Oncol Biol Phys, Vol. 84, No. 3, pp. e285ee290, 2012 0360-3016/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.ijrobp.2012.04.005
Euclid Avenue, Cleveland, OH 44195. Tel: (216) 445-9479; Fax: (216) 445-2226; E-mail: [email protected] Conflict of interest: none.
e286 Khan et al.
International Journal of Radiation Oncology Biology Physics
Introduction
only 43% of patients would have visual acuity better than 20/200 (2). Many subsequent studies have also reported loss of visual acuity after treatment with plaque brachytherapy (Table 1) (2, 9-18). The decline in visual acuity can be explained by several factors. Some are consequences of inherent tumor characteristics, such as proximity to the fovea. Therefore, preoperative visual acuity must be taken into account in predictions of postoperative visual acuity. Other mechanisms known to be responsible for a loss of visual acuity include radiation effects on normal ocular structures adjacent to the tumor such as radiation-induced cataract (9), radiation retinopathy (ischemic, exudative, neovascular) (9), radiation choroidopathy (9), and radiation optic neuropathy (9). Radiation effects on the tumor such as retinal or vitreous hemorrhage secondary to tumor necrosis or intratumoral choroidal vasculopathy (toxic tumor syndrome) may also contribute to loss of visual acuity. Although previous studies have explored risk factors associated with the development of radiation retinopathy or radiation optic neuropathy, these radiation sequelae may not occur in isolation. Moreover, specific etiopathogenetic mechanisms for vision loss may be difficult to ascertain or classify in an individual patient. Therefore, determining the factors that influence visual acuity after plaque brachytherapy may be a more reliable way of determining the patient’s prognosis. A vision prognostication model that can predict the visual outcome for an individual patient, based on baseline tumor characteristics and a specific
In a landmark randomized multicenter study initiated in the mid1980s, the Collaborative Ocular Melanoma Study (COMS) established plaque brachytherapy as a viable alternative to enucleation for the treatment of medium-sized choroidal melanomas (defined as 2.5-10.0 mm in apical height and 5-16 mm in largest basal diameter) (1, 2). The study demonstrated a low risk (10.3%) of local treatment failure in patients undergoing brachytherapy (3) and a 5-year tumor-related mortality rate of 11% after enucleation and 9% after brachytherapy (4). These equivalent rates of local control and survival between the 2 treatments firmly established plaque brachytherapy as an appropriate treatment for medium-sized uveal melanomas. Although ruthenium-106 (5-7), a beta-emitter, has been widely used in European countries, the COMS study used iodine-125. The study initially prescribed a dose of 100 Gy to the tumor apex, which was later revised to 85 Gy based on recommendations of the American Association of Physicists in Medicine (1). Proton therapy has also been reported to provide effective local control rates (8). Along with proton therapy, plaque brachytherapy has now become the favored treatment alternative to enucleation for medium-sized melanomas because it is preferred by patients and can potentially preserve some degree of visual acuity in the treated eye. In practice, however, plaque brachytherapy leads to a significant loss of visual acuity. The COMS study showed that at 3 years,
Table 1
Loss of visual acuity in patients treated with I-125 brachytherapy for treatment of uveal melanoma* No. of patients
Median follow-up time (mo)
VA worse than 20/200 (% patients)
Factors associated with decreased VA
Sia et al, 2000 (10) Melia et al, 2001 (2)
50 623
39.5 -
37% 45%
Jones et al, 2002 (11)
63
No significant predictors found Tumor-associated retinal detachment Tumor not dome-shaped Diabetes mellitus Tumor proximity to macula and optic disc Higher doses to macula and optic disc Macula dose rate 111 11.1 cGy/h Greater tumor height Increased age -
Study
Bechrakis et al, 2002 (12)
152
Puusaari et al, 2004 (14)
96
36
27.1 (mean) 42
-
95.4% 18% of patients with VA above 20/400 49% 56%
Stack et al, 2005 (15) Jensen et al, 2005 (16)
92 156
31 74.4
Krohn et al, 2008 (17)
111
61
60%
Saconn et al, 2010 (18)
62
58.2
22%
189
-
25.4%
Khan et al, 2012
Dose rate to optic disc >60 cGy/h Dose rate to lens <15 cGy/h Atherosclerosis/coronary artery disease Tumor height >8 mm Large tumory Diabetic retinopathyz Poor initial VA in affected eye Greater total dose to fovea Greater tumor height Increased age
Abbreviation: VA Z visual acuity. * The table includes English-language studies published between 1995 and 2010, evaluating visual acuity after I-125 plaque brachytherapy for uveal melanoma. Studies were found by conducting a PubMed search for the terms “uveal melanoma,” “choroidal melanoma,” “ocular melanoma,” “visual acuity,”“I-125,” “I125,” and “brachytherapy.” Studies evaluating the efficacy of treatment for radiation retinopathy (not the occurrence of radiation retinopathy) were excluded. Case reports were also excluded. y According to criteria of the Collaborative Ocular Melanoma Study. z Factors associated with shorter time to visual acuity worse than 20/200.
Volume 84 Number 3 2012
Visual prognostication model for plaque brachytherapy e287 performed with plaque simulator treatment planning software (Plaque Simulator v 5.3.4; Bebig GmbH, Berlin, Germany) (19, 20) designed to give a total prescription of 85 Gy to the tumor apex. Isotopes used were iodine-125 (Model MED3631-A/M; North American Scientific Inc, Chatsworth, CA) and ruthenium106 (Models CCA, CCD, and CCB, with diameters of 15.3 mm, 17.9 mm, and 20.2 mm, respectively; Bebig GmbH, Berlin, Germany). As part of the initial workup, all patients underwent a detailed ophthalmoscopic examination, photography, and ultrasonographic A and B scan. Tumor size (basal dimensions and height) was assessed by ophthalmoscopic examination and by ultrasonography, respectively. For the placement of the radioactive plaque, tumor localization was performed by use of transillumination or indirect ophthalmoscopy, based on tumor location. An intraoperative ultrasound was performed to ensure that the tumor margins were covered by the plaque. The patients were evaluated for follow-up 1 week after plaque brachytherapy, then every 3 months throughout the first year, every 6 months throughout the second through fifth years, and then annually thereafter. The best corrected visual acuity was measured at each visit. Exploratory frequency tables and descriptive statistics were created for the outcomes of visual acuity worse than 20/50 and visual acuity worse than 20/200. Categoric factors were described by using frequencies and percentages, and continuous measures were summarized by using median and interquartile range. Kaplan-Meier estimates of survival (freedom from vision loss) and univariable Cox proportional hazards models were performed to identify potential predictors of visual acuity loss. Multivariable Cox proportional hazards models were fit by using variables that showed minimal association in univariable analysis (P<.20). Collinearity diagnostics were performed before model fitting, and predictors were chosen from this set. Among the size measurements, largest basal diameter, circumference, surface area, and percentage of retinal surface were very highly correlated, so only the largest basal diameter was used for multivariable modeling. Based on the findings from the multivariable models, a nomogram was fit for each visual acuity model. Bootstrap estimates of the concordance index were fit for each model, and the calibration curve showing actual vs predicted probabilities of vision loss were fit. Predicted risk was calculated at 1 year and 3 years. These analyses were performed with SAS software (version 9.2; Cary, NC) and the Design package within R software (version 2.8; Vienna, Austria).
Table 2 Summary demographic patient characteristics: visual acuity (VA) worse than 20/50 and worse than 20/200 Features Age, y (median) Sex F M Medical history Diabetes Hypertension Heart disease Ocular comorbidities Cataracts Other conditions*
VA 20/50 (nZ116)
VA <20/200 (nZ29)
65.6
64.0
62 (53.5%) 54 (46.6%)
27 (46.6%) 31 (53.5%)
15 (12.9%) 43 (37.1%) 18 (15.5%)
7 (12.1%) 18 (31.0%) 7 (12.1%)
48 (41.4%) 33 (28.5%)
23 (39.7%) 19 (32.85)
* Other ocular comorbidities were glaucoma, macular degeneration, strabismus, amblyopia, retinal disease, history of injury, and surgery.
prescribed radiation dose, would be a useful management tool. To our knowledge, such a vision prognostication model has not been reported. We present our initial attempt at generating a vision prognostication tool that can predict the chances of retaining visual acuity at 3 years after treatment.
Methods and Materials An internal review board approval was obtained. All 189 patients treated at the Cleveland Clinic for a primary single ciliary body or choroid melanoma with iodine-125 or ruthenium-106 plaque brachytherapy between January 1, 2005, and June 30, 2010, were included. Patients with 2 primary uveal melanomas (1 in the ipsilateral eye and 1 in the contralateral eye) were excluded, as were any patients with nonmelanoma tumors of the ciliary body or choroid. Any patient who received adjuvant ocular therapy such as transpupillary thermotherapy or had received more than 1 plaque insertion for the same tumor were also excluded. Patient characteristics and demographics were extracted from the Cleveland Clinic electronic medical records system (EpicCare EMR, Madison, WI). Information regarding patients’ pretreatment and post-treatment ocular histories was extracted from paper charts at the Department of Ophthalmic Oncology, Cole Eye Institute, Cleveland Clinic. Dosimetry for the plaques was Table 3
Summary tumor characteristics and radiation dose: visual acuity (VA) worse than 20/50 and worse than 20/200 Features
Largest basal diameter (mm) Tumor height (mm) Distance to fovea (mm) Distance to disc (mm) Total dose to disc (Gy) Total dose to macula (Gy) Total dose to lens (Gy) Isotope I-125 Ru-106
VA <20/50 (nZ116) Median (25th, 75th) 12.0 4.0 4.5 4.8 29.8 50.5 9.9
(9.0, 14.0) (2.5, 6.0) (1.5, 9.5) (3.0, 9.5) (18.6, 42.0) (18.6, 97.3) (4.8, 23.7)
92 (79.3%) 24 (20.7%)
VA <20/200 (nZ58) Median (25th, 75th) 12.0 4.9 3.0 4.5 34.4 61.5 13.6
(9.0, 14.5) (3.0, 7.5) (1.5, 7.5) (3.0, 7.5) (23.7, 45.3) (25.7, 110.9) (7.9, 28.6)
51 (87.9%) 7 (12.1%)
International Journal of Radiation Oncology Biology Physics
e288 Khan et al. Table 4 Kaplan-Meier estimates of visual acuity preservation at 12, 24, and 36 months for 20/50 and 20/200 vision Time
No VA loss
20/50 (nZ108) 12 41.1% 24 51.2% 36 56.6% 20/200 (nZ179) 12 16.2% 24 27.0% 36 41.4%
95% CI
No. failed
No. left
(31.1, 51.1) (40.0, 62.3) (44.4, 68.8)
40 46 48
38 18 8
(10.3, 22.0) (18.9, 35.1) (30.2, 52.7)
25 35 43
87 48 22
Abbreviations: CI Z confidence interval; VA Z visual acuity. The analysis is based on 108 patients with VA better than 20/50 at presentation and the 179 patients with VA better than 20/200 at presentation.
Results For the time period of January 1, 2005, to June 30, 2010, 224 cases of uveal melanoma were treated at the Department of Ophthalmic Oncology, Cole Eye Institute, Cleveland Clinic Foundation. Of these, 9 cases of iris melanoma were excluded. Twenty-six additional cases were excluded: 22 in which adjuvant treatment with transpupillary thermotherapy or laser was given and 4 with 2 consecutive plaques. This left 189 cases of ciliary body and choroid melanoma for analysis. Five of these patients did not return for follow-up. These patients were included in the analysis but treated as censored immediately (time 0) for all outcomes. Of the 189 patients, 174 (92%) were alive as of February 1, 2011. Of the 15 deaths, 4 (2%) were attributable to metastatic uveal melanoma, 5 (3%) were unrelated, and 6 (4%) were due to unknown causes. Seven patients (3.2%) in the study cohort required enucleation (tumor recurrence 6; neovascular glaucoma 1). Visual acuity was evaluated, excluding all patients with vision worse than the specified endpoints at presentation. Therefore, 108 patients with visual acuity better than or equal to 20/50 at presentation were included in the analysis for factors associated with visual acuity worse than 20/50, and 173 patients with vision better than or equal to 20/200 were included in the analysis for factors associated with visual acuity worse than 20/200. The demographics of the patients are given in Table 2, and radiation dose distribution is described in Table 3. Kaplan-Meier estimates of visual acuity loss are presented in Table 4. Significant factors contributing to loss of visual acuity on multivariable analysis with reduced Cox proportional hazards Table 5 Hazard ratios, confidence intervals, and P values from reduced Cox proportional hazards model for vision loss to worse than 20/50 Parameter
Hazard ratio
Lower CL
Upper CL
P value
Diabetes mellitus Heart disease Other ocular problems Tumor height Total dose to lens
0.44 0.46 1.73 1.41 0.98
0.16 0.14 0.91 1.13 0.95
1.23 1.50 3.29 1.76 1.01
.12 .20 .097 .002 .14
Abbreviations: Lower CL Z lower limit of confidence interval; Upper CL Z upper limit of confidence interval.
Table 6 Hazard ratios, confidence intervals, and P values from reduced Cox proportional hazards model for vision loss to worse than 20/200 Parameter
Hazard ratio
Lower CL
Upper CL
P value
Female sex Tumor height Total dose (macula)
1.85 1.42 1.01
0.99 1.28 1.01
3.46 1.58 1.02
.055 <.001 <.001
Abbreviations: Lower CL Z lower limit of confidence interval; Upper CL Z upper limit of confidence interval.
models are presented in Tables 5and 6. On multivariable analysis, with regard to visual acuity worse than 20/50, no factors were found to be protective, whereas increased tumor height (hazard ratio [HR] of 1.41 [1.13-1.76], PZ.002) and a history of other ocular comorbidities (HR of 1.73 [0.91-3.32], PZ.097) were found to be detrimental. These other ocular comorbidities included both historical and currently diagnosed systemic conditions such as hypertension, diabetes, and heart disease and ocular conditions such as cataracts, glaucoma, macular degeneration, strabismus, amblyopia, and other retinal diseases, Patients did not undergo specific correction of these comorbidities before radiation therapy. With regard to visual acuity worse than 20/200, female sex (HR of 1.85 [0.99-3.46], PZ.055), increased tumor height (HR of 1.42 [1.28-1.58], P<.001), and a greater total dose to the macula (HR of 1.01 [1.01-1.02], P<.001) were found to be detrimental (Table 4). Nomogram fit was attempted based on the findings from the multivariable models for patients with visual acuity worse than 20/50 at 1 and 3 years and visual acuity worse than 20/200 at 1 and 3 years. A nomogram could not be generated for visual acuity worse than 20/50 because it was not possible to fit a statistically significant model with adequate discrimination or calibration. The nomogram generated for visual acuity worse than 20/200 assigns a certain number of points to sex, tumor height in millimeters (apex), and total dose to the macula in Gray because these were the significant factors associated with this level of vision loss on multivariable analysis (Fig. 1). The total number of points can be added up and correlated with a predicted probability of visual acuity worse than or equal to 20/200 at 1 year and at 3 years. For example, a man with a tumor height of 2 mm and a dose to the macula of 40 Gy would have 15 þ 15 þ 30 Z 60 total points. This number of total points (line 5) correlates with a 5% to 10% chance of visual acuity worse than 20/200 at 1 year and a 30% chance of visual loss below 20/200 at 3 years. A concordance index (0.81) for this nomogram between actual and predicted loss of vision at 1 year indicates a fairly strong fit for this model (Fig. 2).
Discussion Our data set demonstrated decreased visual acuity with increased tumor height and increased total dose to the macula, similar to the findings in many previous studies (12, 17, 21). However, other factors that had been associated with severe vision loss in previous studies, such as proximity to optic disc and fovea (11, 21), total dose to optic disc and fovea, and systemic diseases such as diabetes mellitus (18, 21), were not significant in this data set. Inasmuch as our study was based on a limited number of cases, it is possible that some of the prognostic factors may have been excluded because they did not achieve adequate statistical significance to be included in the model.
Volume 84 Number 3 2012
Fig. 1.
Visual prognostication model for plaque brachytherapy e289
The nomogram for visual acuity loss worse than 20/200. In this nomogram, tumor height is the most significant predictor of risk.
Our data also reveal that the main modifiable factor associated with severe visual loss is the total radiation dose to the macula (Table 6). Therefore, if the radiation dose to the macula could be minimized, visual outcome after brachytherapy could be improved (22). However, a larger data set is required to identify the threshold doses necessary to avoid radiation maculopathy. This is an important consideration for designing new classes of ophthalmic brachytherapy implants or identifying new isotopes for ophthalmic applications. The novel aspect of this study was the generation of a nomogram to predict visual loss, based on patient characteristics, tumor characteristics, and dosimetry, which are available before treatment. Furthermore, although the calibration plot indicates a fairly strong correlation between actual and predicted visual loss, the inclusion of more data points will make it a more robust predictive
Fig. 2. The calibration plot for visual acuity 20/200 or worse. The concordance index for the model of vision 20/200 or worse is 0.81. The actual and predicted risk measures are very good across the range of predicted and actual risk levels.
model. Additional studies based on analysis of published data are under way.
Conclusions The vision prognostic tool is a first step in providing clinicians and patients with a way to estimate visual acuity in a specific patient even before the plaque brachytherapy treatment is delivered. By providing a practical means to predict vision loss at 3 years after treatment, our vision prognostication model can be an important tool for patient selection and treatment counseling.
References 1. Krintz AL, Hanson WF, Ibbott GS, et al. A reanalysis of the Collaborative Ocular Melanoma Study Medium Tumor Trial eye plaque dosimetry. Int J Radiat Oncol Biol Phys 2003;56: 889-898. 2. Melia BM, Abramson DH, Albert DM, et al. Collaborative ocular melanoma study (COMS) randomized trial of I-125 brachytherapy for medium choroidal melanoma: I. Visual acuity after 3 years COMS report no. 16. Ophthalmology 2001;108:348-366. 3. Jampol LM, Moy CS, Murray TG, et al. The 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. Ophthalmology 2002;109: 2197-2206. 4. Diener-West M, Earle JD, Fine SL, et al. The COMS randomized trial of iodine 125 brachytherapy for choroidal melanoma: III. initial mortality findings. COMS Report No. 18. Arch Ophthalmol 2001;119: 969-982. 5. Rouberol F, Roy P, Kodjikian L, et al. Survival, anatomic, and functional long-term results in choroidal and ciliary body melanoma after ruthenium brachytherapy (15 years’ experience with beta-rays). Am J Ophthalmol 2004;137:893-900.
e290 Khan et al. 6. Damato B, Patel I, Campbell IR, et al. Visual acuity after Ruthenium(106) brachytherapy of choroidal melanomas. Int J Radiat Oncol Biol Phys 2005;63:392-400. 7. Summanen P, Immonen I, Kivela T, et al. Radiation related complications after ruthenium plaque radiotherapy of uveal melanoma. Br J Ophthalmol 1996;80:732-739. 8. Gragoudas ES. Proton beam irradiation of uveal melanomas: the first 30 years. The Weisenfeld Lecture. Invest Ophthalmol Vis Sci 2006;47: 4666-4673. 9. Wen JC, Oliver SC, McCannel TA. Ocular complications following I125 brachytherapy for choroidal melanoma. Eye (Lond) 2009;23: 1254-1268. 10. Sia S, Harper C, McAllister I, et al. Iodine-I25 episcleral plaque therapy in uveal melanoma. Clin Experiment Ophthalmol 2000;28:409-413. 11. Jones R, Gore E, Mieler W, et al. Posttreatment visual acuity in patients treated with episcleral plaque therapy for choroidal melanomas: dose and dose rate effects. Int J Radiat Oncol Biol Phys 2002; 52:989-995. 12. Bechrakis NE, Bornfeld N, Zoller I, et al. Iodine 125 plaque brachytherapy versus transscleral tumor resection in the treatment of large uveal melanomas. Ophthalmology 2002;109:1855-1861. 13. Shields CL, Naseripour M, Cater J, et al. Plaque radiotherapy for large posterior uveal melanomas (> or Z 8-mm thick) in 354 consecutive patients. Ophthalmology 2002;109:1838-1849. 14. Puusaari I, Heikkonen J, Kivela T. Effect of radiation dose on ocular complications after iodine brachytherapy for large uveal melanoma:
International Journal of Radiation Oncology Biology Physics
15.
16.
17.
18.
19.
20. 21.
22.
empirical data and simulation of collimating plaques. Invest Ophthalmol Vis Sci 2004;45:3425-3434. Stack R, Elder M, Abdelaal A, et al. New Zealand experience of I125 brachytherapy for choroidal melanoma. Clin Experiment Ophthalmol 2005;33:490-494. 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 2005;63:101-108. Krohn J, Monge OR, Skorpen TN, et al. Posterior uveal melanoma treated with I-125 brachytherapy or primary enucleation. Eye (Lond) 2008;22:1398-1403. Saconn PA, Gee CJ, Greven CM, et al. Alternative dose for choroidal melanoma treated with an iodine-125 radioactive plaque: a singleinstitution retrospective study. Int J Radiat Oncol Biol Phys 2010; 78:844-848. Knutsen S, Hafslund R, Monge OR, et al. Dosimetric verification of a dedicated 3D treatment planning system for episcleral plaque therapy. Int J Radiat Oncol Biol Phys 2001;51:1159-1166. Astrahan MA. A patch source model for treatment planning of ruthenium ophthalmic applicators. Med Phys 2003;30:1219-1228. Finger PT, Chin KJ, Yu GP. Risk factors for radiation maculopathy after ophthalmic plaque radiation for choroidal melanoma. Am J Ophthalmol 2010;149:608-615. Lumbroso-Le Rouic L, Charif Chefchaouni M, Levy C, et al. 125I plaque brachytherapy for anterior uveal melanomas. Eye (Lond) 2004; 18:911-916.