Alternative Dose for Choroidal Melanoma Treated With an Iodine-125 Radioactive Plaque: A Single-Institution Retrospective Study

Int. J. Radiation Oncology Biol. Phys., Vol. 78, No. 3, pp. 844–848, 2010 Copyright Ó 2010 Elsevier Inc. Printed in the USA. All rights reserved 0360-3016/$–see front matter

doi:10.1016/j.ijrobp.2009.08.059

CLINICAL INVESTIGATION

Eye

ALTERNATIVE DOSE FOR CHOROIDAL MELANOMA TREATED WITH AN IODINE125 RADIOACTIVE PLAQUE: A SINGLE-INSTITUTION RETROSPECTIVE STUDY PAUL A. SACONN, M.D.,* CHRISTOPHER J. GEE, M.D.,y CRAIG M. GREVEN, M.D.,y THOMAS P. MCCOY, M.S.,z KENNETH E. EKSTRAND, PH.D.,* AND KATHRYN M. GREVEN, M.D.* Departments of *Radiation Oncology, yOphthalmology, and zBiostatistical Sciences—Division of Public Health Sciences, Wake Forest University School of Medicine, Winston-Salem, NC Purpose: The Collaborative Ocular Melanoma Study (COMS) established iodine-125 plaque brachytherapy as an accepted standard treatment for medium-size choroidal melanoma. In the COMS, the prescription dose was 85 Gy. This is a retrospective review of our outcomes in patients treated with lower doses than those used in the COMS. Methods and Materials: From 1990 to 2004, 62 patients were treated with iodine-125 plaque brachytherapy for choroidal melanoma. COMS eye plaques were used with dose prescribed to the apex of the tumor. The median and average dose rates at the tumor apex were 63.5 cGy/h and 62.7 cGy/h, respectively. The median and average total doses were 63.0 Gy and 62.5 Gy (range, 56–69 Gy), respectively. The median and mean durations of implant were 100.0 hours and 101.1 hours (range, 71–165 hours). Results: Median follow-up time was 58.2 months. The 5-year outcomes including overall survival, disease-free survival, cause-specific survival, local failure, secondary enucleation rate, and visual acuity (VA) <20/200 were estimated using the Kaplan-Meier method. Overall, there were 7 local failures, 4 distant failures, and 10 secondary enucleations (6 due to local failure and 4 due to treatment complications). Univariate analysis was performed to identify significant prognostic factors associated with disease-free survival (baseline VA in tumor eye, tumor shape), cause-specific survival (diabetic retinopathy), local failure (none found), secondary enucleation rate (diabetic retinopathy, basal tumor dimension) and VA <20/200 (diabetic retinopathy, tumor shape, age, retinal detachment, treatment depth, and history of vision-limiting condition). Conclusions: Our survival and local control outcomes are comparable to those of the COMS. However, VA at 5 years seems to be better. Lower doses of radiation could potentially lead to better visual outcomes. Ó 2010 Elsevier Inc. Choroidal melanoma, Plaque, Brachytherapy, Dose.

sen so that the actual doses delivered before and after TG 43 would be the same (2). From 1990 through 2004, we treated all patients with medium-size choroidal melanoma at our institution with a dose-reduced COMS design. This article reports the all-cause mortality, enucleation rate, and visual outcomes in our patients by the use of dosage reduction.

INTRODUCTION Choroidal melanoma is the most common primary malignant intraocular tumor in adults. Until the 1980s, the standard treatment of choroidal melanoma was enucleation. In 1986 the Collaborative Ocular Melanoma Study (COMS) initiated a large randomized multicenter clinical trial comparing enucleation and iodine plaque brachytherapy and demonstrated equivalent survival while offering the potential for visual preservation (1). Consequently, iodine-125 plaque brachytherapy is now widely accepted as a preferred alternative to enucleation in the treatment of medium-size choroidal melanoma. From 1986 to 1996, the COMS used a prescription dose of 100 Gy delivered to the tumor apex or 5-mm point depending on tumor height. When the Task Group (TG 43) calculation protocol became the recommendation of the American Association of Physicists in Medicine, the dose of 85 Gy was cho-

METHODS AND MATERIALS A retrospective chart review of all patients with medium-size choroidal melanoma treated with iodine-125 plaque brachytherapy at Wake Forest University Baptist Medical Center from 1990 to 2004 was performed. This retrospective review was approved by the Institutional Review Board. A total of 62 patients were treated during the time period. All patients had unilateral choroidal melanoma measuring 2.0–9.6 mm in height and less than 17.0 mm in diameter. All patients were over 21 years of age and had visual acuity (VA) in the fellow eye of better than 20/200.

Reprint requests to: Kathryn M. Greven, M.D., Department of Radiation Oncology, Wake Forest University School of Medicine, Medical Center Boulevard, Winston-Salem, NC 27157. Tel: (336) 713-3600; Fax: (336) 713-6565; E-mail: [email protected]

Conflict of interest: none. Received April 22, 2009, and in revised form Aug 20, 2009. Accepted for publication Aug 25, 2009. 844

Plaque brachytherapy dose for choroidal melanoma d P. A. SACONN et al.

In accord with the COMS measures, two poor vision outcomes were used to evaluate VA outcome at the 3- and 5-year follow-up examinations. These outcomes included (1) a VA of 20/200 or worse or (2) loss of six lines or more of VA from baseline (3). During the follow-up interval, visual outcomes were documented at regular intervals. The time at which a patient met the definition for a poor vision outcome was recorded. All patients who underwent enucleation were recorded as having a poor vision outcome if they had not already met the definition for a poor vision outcome before enucleation.

Treatment protocol In accord with the COMS design, the size of the plaque was chosen to cover the tumor base plus a tumor-free margin of 2–3 mm on all borders except in instances in which the optic nerve limited the margin. For tumors 5.0 mm or more in apical height, the radiation dose was prescribed to the tumor apex, and for tumors 2.5–4.9 mm in apical height, the prescription point was 5 mm from the interior surface of the sclera (1). The dosimetry recommended by TG 43 of the American Association of Physicists in Medicine Radiation Therapy Committee was followed. Originally our doses were determined in accordance with the pre–TG 43 factors. However, the doses reported here have all been corrected to be in accordance with the TG 43 recommendations (2). Our median prescription dose was 63 Gy. In accord with the COMS design, the number of seeds was dependent on plaque size, as follows: 12 mm (8 seeds), 14 mm (13 seeds), 16 mm (13 seeds), 18 mm (21 seeds), and 20 mm (24 seeds). The seed strength varied depending on dose rate, plaque size, and tumor thickness and ranged from 1.1 mCi (1.4 U) to 4.95 mCi (6.2 U). Additionally, like the COMS dosimetric calculations, our dose calculations neither accounted for the increased photon attenuation within the silastic seed carrier nor reflected the decrease in backscattering from the gold plaque.

Statistical methods Patient demographics and clinical characteristics were summarized using descriptive statistics. The Kaplan-Meier method was used to estimate the event times of outcomes (4). Inasmuch Six patients already had a VA of 20/200 or less at baseline and were excluded from analyses for the visual outcome. Baseline demographic, ophthalmic, and tumor and treatment characteristics were analyzed. These characteristics were tested individually for association with outcome using bivariate Cox proportional hazards regression. Characteristics significant with p < 0.20 were included in multivariable Cox proportional hazards modeling. A backwards elimination procedure was then used to select a final multivariable model, retaining characteristics significantly associated with outcome with p < 0.05. All analyses were performed using SAS v9.1.3 (SAS Institute, Cary, NC) and Stata v10 (StataCorp, College Station, TX).

RESULTS Patient characteristics Table 1 gives the distribution of characteristics for the 62 patients compared with the patients in the COMS (1). Of those patients, 56% were female, all were Caucasian, and the average age was 63.5 years. Associated presenting factors included diabetic retinopathy (5%), glaucoma (2%), cataract (27%), and macular degeneration (5%). At baseline, the average visual acuities in the tumor and fellow eye were 0.64 

845

Table 1. Patient demographics and ophthalmic characteristics (N = 62) Patient characteristics Demographics Sex F M Race White Other Age at plaque insertion (y) <50 50–69 $70 Median age (y) Ophthalmic characteristics Baseline visual acuity in study eye $20/20 20/40 to <20/20 20/160 to <20/40 <20/160 Median baseline visual acuity History of diabetic retinopathy No Yes

CCCWFU n

%

COMS %

35 27

56 44

50 50

62 0

100

98 2

11 55 34

25 50 25

17 24 15 6 20/30

27 39 24 10

33 38 19 10

59 3

95 5

7 34 21 64.1

Abbreviations: CCCWFU = Comprehensive Cancer Center, Wake Forest University; COMS = Collaborative Ocular Melanoma Study.

0.31 and 0.76  0.28, respectively. The average intraocular pressure in the tumor eye was 14.7  3.3. The proportion of patients with a relative afferent papillary defect, VFD, and retinal detachment were 6%, 25%, and 13%, respectively. Most patients had a dome-shaped tumor (84%), 10% had a button-shaped tumor, and 6% had tumors of other shapes.

Tumor characteristics Table 2 illustrates the tumor characteristics compared with the COMS (1). The average tumor heights and basal diameters were 4.6  1.9 mm and 11.4  2.8 mm, respectively. The average tumor distances to the optic nerve and fovea were 2.8  1.3 mm and 3.1  1.4 mm, respectively.

Treatment characteristics Table 3 illustrates the treatment characteristics. The average plaque size was 15 mm. The median and average dose rates at the tumor apex were 63.5 cGy/h and 62.7 cGy/h, respectively. The median and average total doses were 63.0 Gy and 62.5 Gy (range, 56–69 Gy), respectively. The median and mean durations of implant were 100.0 hours and 101.1 hours (range, 71–165 hours). The average treatment depth was 6.0  1.3 mm. The total dose distributions were as follows: 56–60 Gy (32%), 61–63 Gy (31%), 64 Gy (13%), and 64–69 Gy (24%).

I. J. Radiation Oncology d Biology d Physics

846

Table 2. Tumor characteristics Tumor characteristics Largest basal dimension (mm) 4.5–8.0 >8.0–11.0 >11.0–14.0 >14.0–16.0 >16.0 Median diameter (mm) Average diameter (mm) Apical tumor height (mm) 2.5–5.0 5.1–10.0 Distance to fovea (mm) 0.0 0.1–2.0 2.1–5.0 5.1–8.0 >8.0 Average distance Distance to optic disc (mm) 0.0–2.0 2.1–4.0 4.1–6.0 6.1–8.0 >8.1 Average distance

CCCWFU n

Volume 78, Number 3, 2010

Table 4. Disease outcome measures %

COMS %

9 24 19 9 1 11.0 11.4

15 39 31 15 2

13 34 37 16

45 17

73 27

63 37

8 13 17 8 14 3.1

13 21 27 13 23

13 25 29 17 15

10 22 9 11 9 2.8

16 35 15 18 15

15 31 24 15 15

Abbreviations: CCCWFU = Comprehensive Cancer Center, Wake Forest University; COMS = Collaborative Ocular Melanoma Study.

Disease outcomes Table 4 illustrates the 5-year outcome measures compared with the COMS (1, 5). Median follow-up time for the 62 patients was 4.8 years (58.2 months). Twenty-five patients were still at risk for outcomes analysis 5 years after plaque therapy. The 5-year local control rate was 91.3% (Fig. 1). The 5-year enucleation rate was 12.0%, with 10 patients requiring secondary enucleations (6 secondary to local failure and 4 secondary to treatment complications). The median survival was 12.8 years, and the 5-year overall survival was 78.6% (Fig. 2). The 5-year disease-free survival and cause-specific

Disease outcome measures

CCCWFU

COMS

Median follow-up (y) Median survival (y) 5-year disease outcome measures (%) Overall survival Disease-free survival Cause-specific survival Local failure rate 5-year enucleation rate

4.8 12.8

5.6

79 90 96 9 12

Abbreviations: CCCWFU = Comprehensive Cancer Center, Wake Forest University; COMS = Collaborative Ocular Melanoma Study.

survival were 90.4% and 95.8%, respectively (Fig. 3). There were 19 deaths over the course of study follow-up (with 4 deaths from disease). There were 7 local failures and 4 distant failures (3 with liver metastasis and 1 with multiple site metastases). Visual outcomes Table 5 illustrates the visual outcomes compared with the COMS) (3). A decline in VA to 20/200 or worse was observed in only 16% and 22% of eyes at 3 and 5 years, respectively (Fig. 4). A loss of 6 or more Snellen lines from baseline was observed in 10% and 16% of eyes at 3 and 5 years, respectively. Prognostic factors On univariate analysis, diabetic retinopathy was the only statistically significant adverse factor related to disease-free survival (p = 0.034) and cause-specific survival (p = 0.014). The only significant predictor of local failure was blurred vision and VFD at presentation (p = 0.005). In multivariable Cox regression modeling, the variables associated with poorer overall survival included cataract (hazard ratio [HR] = 2.60, p = 0.048), lower baseline VA in tumor eye (HR = 0.04, p = 0.003), and distant failure (HR = 24.9, p < 0.001). Diabetic retinopathy (HR = 7.32,

Table 3. Treatment characteristics Treatment characteristics Median treatment depth (mm) Average treatment depth (mm) Median dose rate at tumor apex (cGy/h) Average dose rate at tumor apex (Gy/h) Median total dose (Gy) Average total dose (Gy)

CCCWFU 5.0 6.0 63.5 62.7 63.0 62.5

Abbreviation: CCCWFU = Comprehensive Cancer Center, Wake Forest University.

82 90 Not reported 10 13

Fig. 1. Local control (Kaplan-Meier).

Plaque brachytherapy dose for choroidal melanoma d P. A. SACONN et al.

847

Table 5. Visual outcome measures Visual outcome measures 3-year visual acuity 20/200 or worse 5-year visual acuity 20/200 or worse 3-year loss of $6 Snellen lines from baseline 5-year loss of $6 Snellen lines from baseline

CCCWFU (%) COMS (%) 16 22 10

43 Not reported 49

16

Not reported

Abbreviations: CCCWFU = Comprehensive Cancer Center, Wake Forest University; COMS = Collaborative Ocular Melanoma Study.

Before the landmark COMS, enucleation was the standard treatment for patients with uveal melanoma. The COMS, however, established iodine plaque brachytherapy as a treatment alternative with equivalent survival rates. Iodine plaque brachytherapy offers excellent tumor control with the potential for visual preservation. However, the ideal dose for achieving local control without incurring a high risk of decreased VA remains undefined. The COMS dose prescription is widely used as a standard for eye plaque treatment. That prescription dose is 85 Gy to the tumor apex, and thus the American Brachytherapy Society has recommended a prescription dose of 85 Gy, with

a dose rate of 0.60–1.05 Gy/h, using the COMS dosimetry assumptions and plaque construction techniques (6). This radiation dose, however, far exceeds the tolerance of critical surrounding structures including the retina, optic nerve, and lens. Some institutions have used lower doses without compromising disease control. For example, Quivey et al. (7) reported on 239 patients with median-size choroidal melanoma who were treated with iodine-125 plaque brachytherapy to a minimum prescription dose of 70 Gy and demonstrated a 5-year actuarial local control of 82%. Margins around the tumor were only 1 mm on each side, and there was some concern that this resulted in even lower doses to the tumor. They described a visual outcome of 20/200 or better in 58% of patients. In an attempt to reduce complications, Fontanesi et al. (8) reduced the apical dose to 75 Gy at a dose rate of 60–65 cGy/h. They described progressive tumor growth in only 4 of 144 patients and retention of vision of 20/200 or better in 59 of 102 patients. Another report by Quivey et al. (9) evaluating 150 patients with uveal melanoma with initial mean tumor size of 9.7  8.5  3.7 mm (ranging from 4.5 mm to 21.5 mm in basal dimension and 1.2–11.8 mm in height) showed that treatment parameters, including dose rate, total dose to the tumor apex, dose rate to the tumor base, and overall treatment duration were significantly correlated with local tumor control. Specifically, a dose rate less than 50 cGy/h and total dose of less than 70 Gy to the tumor apex resulted in 4.75-fold and 8-fold increases, respectively, in local failure

Fig. 3. Cause-specific survival (Kaplan-Meier).

Fig. 4. Visual acuity 20/200 or worse (Kaplan-Meier).

Fig. 2. Overall survival (Kaplan-Meier).

p = 0.019) and larger basal tumor dimension (HR = 1.35, p = 0.027) were associated with shorter time interval to secondary enucleation. Significant variables associated with shorter time to VA worse than 20/200 included diabetic retinopathy (HR = 39.1, p < 0.001), decreased initial VA in the tumor eye (HR = 0.09, p = 0.004), and relative afferent papillary defect (HR = 60.2, p = 0.004) DISCUSSION

848

I. J. Radiation Oncology d Biology d Physics

compared with higher dose rate and dose levels. By comparison, our results with a similar population of patients demonstrate excellent local control and survival outcomes using a median total dose and dose rate to the apex of 63.0 Gy and 63.5 cGy/h, respectively. This observed discrepancy may be explained in part because the former article reported outcomes according to doses of iodine that had not been recalibrated with the current TG 43 recommended standards (2). In addition, the plaques that were being used were not COMS plaques but were custom-made designs that may not have had comparable dosimetric characteristics. The COMS demonstrated excellent local control rates and survival outcomes with the use of their dosimetry calculations and techniques and have set a standard for comparison in outcomes. At 5 years, they reported local failure, overall survival, and disease-free survival rates of 10%, 82%, and 90%, respectively. By comparison, our results demonstrated equivalent local failure and survival rates of 9%, 79%, and 90%, respectively. Additionally, the COMS demonstrated that iodine plaque brachytherapy offered excellent preservation of the globe with a 5-year enucleation rate of only 13%. By comparison, our 5-year enucleation rate was 12%. Taken together, these findings suggest that a prescription dose of 63 Gy to the tumor apex using the COMS design may be sufficient in achieving disease control and survival equivalent to that attained in the COMS. In the COMS, the two poor vision outcomes used to evaluate VA outcome at follow-up examination included (1) a VA of 20/200 or worse or (2) loss of six lines or more of VA from baseline. At 3 years after treatment, they demonstrated that 43–49% of treated eyes had substantial impairment in VA. More specifically, 49% of patients had a loss of six or more lines of VA from the pretreatment level, and 43% had a decrease in VA to 20/200 or worse during fol-

Volume 78, Number 3, 2010

low-up. By comparison, our results demonstrated that at 3 years, 10% of patients had a loss of six or more lines of VA from the pretreatment level, and 16% had a decrease in VA to 20/200 or worse during follow-up. At 5 years, our results demonstrated that 15.5% of patients had a loss of six or more lines of VA from the pretreatment level, and 22% had a decrease in VA to 20/200 or worse during follow-up. These findings suggest limited visual loss with our lower prescription dose in comparison with the COMS prescription dose while maintaining equivalent local control and survival outcomes. Direct comparisons between the COMS results and other published studies reporting VA after iodine plaque brachytherapy have been problematic. This is often due to differences in patient and tumor characteristics and in the definition of VA outcomes. Our patient demographics, ophthalmologic, characteristics and tumor characteristics seemed to be well matched with those of the patients who were included in the COMS. A subsequent report is pending that will further describe our patient and tumor characteristics with regard to visual outcomes in greater detail. Moreover, our definitions of VA outcomes are identical with those of the COMS. The obvious limitations of this study, however, include its retrospective nature, the small number of patients, and potential variation in the conditions under which VA measurements were made. In conclusion, our findings suggest that a median prescription dose of 63 Gy, using the COMS dosimetry and techniques, resulted in local disease control and survival outcomes equivalent to those of the COMS results while offering improved visual preservation. Further investigation of dosage reduction in the treatment of medium-sized ocular melanoma using iodine-125 plaque brachytherapy is warranted.

REFERENCES 1. Collaborative Ocular Melanoma Study Group. 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. 2. Nath R, Anderson LL, Luxton G, et al. Dosimetry of interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee Task Group No. 43. Med Physics 1995; 22:209–234. 3. Collaborative Ocular Melanoma Study Group. 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. 4. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457–481. 5. Collaborative Ocular Melanoma Study Group. The COMS randomized trial of iodine 125 brachytherapy for choroidal mela-

6.

7. 8. 9.

noma: Local treatment failure and enucleation in the first 5 years after brachytherapy. COMS Report No. 19. Ophthalmology 2002;198:2197–2206. Nag S, Quivey JM, Earle JD, et al. American Brachytherapy Society. The American Brachytherapy Society recommendations for brachytherapy of uveal melanomas. Int J Radiat Oncol Biol Phys 2003;56:544–555. Quivey JM, Char DH, Phillips TL, et al. High intensity 125-iodine (125I) plaque treatment of uveal melanoma. Int J Radiat Oncol Biol Phys 1993;26:613–618. Fontanesi J, Meyer D, Xu S. Treatment of choroidal melanoma with I-125 plaque. Int J Radiat Oncol Biol Phys 1993;26: 619–623. Quivey JM, Ausberger J, Snelling L, Brady LW. 1251 Plaque therapy for uveal melanoma: Analysis of the impact of time and dose factors on local control. Cancer 1996;77:2356–2362.