- Email: [email protected]
Proton beam radiotherapy for uveal melanoma: Results of Curie Institut–Orsay Proton Therapy Center (ICPO)
Int. J. Radiation Oncology Biol. Phys., Vol. 65, No. 3, pp. 780 –787, 2006 Copyright © 2006 Elsevier Inc. Printed in the USA. All rights reserved 0360-3016/06/$–see front matter
doi:10.1016/j.ijrobp.2006.01.020
CLINICAL INVESTIGATION
Eye
PROTON BEAM RADIOTHERAPY FOR UVEAL MELANOMA: RESULTS OF CURIE INSTITUT–ORSAY PROTON THERAPY CENTER (ICPO) RÉMI DENDALE, M.D.,*‡ LIVIA LUMBROSO-LE ROUIC, M.D.,† GEORGES NOEL, M.D.,‡ LOÏC FEUVRET, M.D.,‡ CHRISTINE LEVY, M.D.,† SABINE DELACROIX, PH.D.,‡ ANNE MEYER, M.D.,§ CATHERINE NAURAYE, PH.D.,‡ ALEJANDRO MAZAL, PH.D.,*‡ HAMID MAMMAR, M.D.,* PAUL GARCIA, PH.D.,‡ FRANÇOIS D’HERMIES, M.D.,§ ERIC FRAU, M.D.,储 CORINE PLANCHER, D.M.,¶ BERNARD ASSELAIN, M.D.,¶ PIERRE SCHLIENGER, M.D.,*‡ JEAN JACQUES MAZERON, M.D.,‡ †‡ AND LAURENCE DESJARDINS, M.D. Departments of *Radiation Oncology, †Ophthalmology, and ¶Biostatistics, Curie Institut, Paris, France; ‡Curie Institut–Orsay Protonthérapie Center, Orsay, France; §Department of Ophthalmology, Hotel Dieu Hospital, Paris, France; 储Department of Ophthalmology, Le Kremlin Bicêtre Hospital, Le Kremlin Bicêtre, France Purpose: This study reports the results of proton beam radiotherapy based on a retrospective series of patients treated for uveal melanoma at the Orsay Center. Methods and Materials: Between September 1991 and September 2001, 1,406 patients with uveal melanoma were treated by proton beam radiotherapy. A total dose of 60 cobalt Gray equivalent (CGE) was delivered in 4 fractions on 4 days. Survival rates were determined using Kaplan–Meier estimates. Prognostic factors were determined by multivariate analysis using the Cox model. Results: The median follow-up was 73 months (range, 24 –142 months). The 5-year overall survival and metastasis-free survival rates were 79% and 80.6%, respectively. The 5-year local control rate was 96%. The 5-year enucleation for complications rate was 7.7%. Independent prognostic factors for overall survival were age (p < 0.0001), gender (p < 0.0003), tumor site (p < 0.0001), tumor thickness (p ⴝ 0.02), tumor diameter (p < 0.0001), and retinal area receiving at least 30 CGE (p ⴝ 0.003). Independent prognostic factors for metastasis-free survival were age (p ⴝ 0.0042), retinal detachment (p ⴝ 0.01), tumor site (p < 0.0001), tumor volume (p < 0.0001), local recurrence (p < 0.0001), and retinal area receiving at least 30 CGE (p ⴝ 0.002). Independent prognostic factors for local control were tumor diameter (p ⴝ 0.003) and macular area receiving at least 30 CGE (p ⴝ 0.01). Independent prognostic factors for enucleation for complications were tumor thickness (p < 0.0001) and lens volume receiving at least 30 CGE (p ⴝ 0.0002). Conclusion: This retrospective study confirms that proton beam radiotherapy ensures an excellent local control rate. Further clinical studies are required to decrease the incidence of postirradiation ocular complications. © 2006 Elsevier Inc. Uveal melanoma, Proton beam radiotherapy, Local control, Survival, Enucleation, Ocular complication.
Uveal melanoma is the most frequent ocular tumor but remains rare with a reported incidence of 6 to 7 cases per 100,000 inhabitants (1, 2). Recently, surgical enucleation was the standard treatment, but various eye-preserving treatment modalities have subsequently been developed, such as local resection (3– 6), plaque brachytherapy (7, 8), external beam radiotherapy with photons using a linear accelerator (9), helium ions (10), and proton beam therapy (11–15). Proton beam radiotherapy is a conservative alternative to enucleation for the management of uveal melanoma, because a comparison between enucleation and radiotherapy
does not reveal any significant difference in terms of overall survival and metastasis-free survival (16). In 1954, at the initiative of Irène Joliot-Curie, a synchrocyclotron particle accelerator was constructed by Philips at the Orsay University campus. It was initially devoted to nuclear physics research, but a phase of experimentation to develop medical applications was initiated in 1987 (17). Proton beam radiotherapy at Orsay, essentially for uveal melanoma, was performed for the first time in 1991 (18 –20). In this study, we will report the results of proton beam radiotherapy for uveal melanoma on a retrospective consecutive series of patients with a minimum follow-up of 2 years.
Reprint requests to: Rémi Dendale, M.D., Department of Radiotherapy, Curie Institut, 26 rue d’Ulm, 75248, Paris Cedex 05, France. Tel: (⫹33) 1-4432-4632; Fax: (⫹33) 1-4432-4616;
E-mail: [email protected] Received April 26, 2005, and in revised form Jan 17, 2006. Accepted for publication Jan 18, 2006.
INTRODUCTION
780
Protontherapy of uveal melanoma
METHODS AND MATERIALS Patients Patients were referred to the Curie Institut–Orsay Proton Therapy Center by three hospitals (Curie Institut, Hôtel Dieu Hospital, Kremlin Bicêtre Hospital). All patient and tumor data were prospectively recorded in a database. From this database, we retrospectively selected patients with uveal melanoma treated with proton beam radiotherapy and followed during a minimum of 2 years. Not included in this study were patients with a previous local treatment, metastases, extrascleral tumor invasion, or iris melanoma at diagnosis. A total of 1,406 case files were selected for patients treated between September 1991 and September 2001. The clinical diagnosis was established 3 or 4 weeks before radiotherapy by the referring hospital on the basis of clinical ocular fundus examination, A and B mode ultrasound, and fluorescein angiography. Patients with little and anterior uveal melanoma were treated by 125I brachytherapy plaque and were not analyzed in this study.
Treatment procedure The procedure after diagnosis was precisely defined. One to 2 weeks before radiotherapy, patients underwent a surgical procedure for precise demarcation of tumor margins by an ophthalmologist. This surgery consisted of suturing 4 or 5 tantalum clips (diameter: 2.5 mm, thickness: 0.5 mm) onto the sclera around the tumor base. The limits of the tumor base were determined by transillumination. Proton beam radiotherapy was planned and performed at Curie Institut–Orsay Proton Therapy Center 1 to 2 weeks after surgery. The treatment technique has been described previously (18 –21). Briefly, head immobilization during radiotherapy was ensured by a custom-made thermoplastic mask associated with a bite block. EYEPLAN software (22–24) was used for dosimetry based on three-dimensional (3D) reconstruction of the eye, including the tumor. Tumor margins were drawn by the referring center ophthalmologist and were verified by the radiation oncologist. The optimal dosimetric eye position was determined according to the following criteria: total irradiation of the tumor surrounded by a safety margin of 2.5 mm; and reduction of the dose delivered to the lacrimal gland, macula, optic disc, and lens in decreasing order of priority. Beam characteristics were selected to place the 90% isodose 2.5 mm behind the tumor. The lateral collimator border defined the site of the isodose 50%. During treatment, patients were seated on a robot chair facing the beam and were asked to fix a small target light placed at a known angle adopted during the treatment planning process. As the beam remained horizontal and immobile, the patient was moved into the treatment position. The robot chair can be moved in all directions and angles with a precision of about 0.1 mm and 0.1°. A systematic attempt was made to keep the eyelids out of the beam’s field. Before each irradiation session, clip positions were verified by 2 orthogonal Polaroid X-rays, and the chair was then moved to the correct position. The fixed gaze was monitored by a magnifying camera (⫻10). Irradiation was systematically suspended whenever the eye shifted from the tumor position and was resumed only after the eye had returned to the optimal position. A total dose of 60 cobalt Gray equivalent (CGE) (using a relative biologic effectiveness of 1.1) was delivered in 4 fractions on 4 consecutive days. An ophthalmologic examination was systematically performed by an ophthal-
● R. DENDALE et al.
781
mologist during the week of irradiation. No transpupillary thermotherapy and no other local treatments were done on these patients.
Follow-up Patients were regularly followed by their ophthalmologist from the referring hospital 1 month after irradiation, every 6 months for the first 2 years, and annually thereafter. A complete ocular functional assessment, including determination of visual acuity, lid damage, keratitis, cataract, glaucoma, optic neuropathy, maculopathy, retinopathy, and ocular inflammation, was performed 1, 2, 3, 5, and 10 years after irradiation. A liver ultrasound for detection of metastases was performed every 6 months for 10 years and annually thereafter. Enucleation was performed for local recurrence and for ocular complications such as severe glaucoma, major retinal detachment, retinal hemorrhage, disabling eye pain, or for ocular complications preventing good clinical and B-mode ultrasound follow-up.
Statistical methods Survival rates were calculated from the day of irradiation to the date of the event of interest or the date of last follow-up. Five-year local recurrence, metastasis-free survival, overall survival, and enucleation for complication (EC) rates were determined by the Kaplan–Meier method (25) and were expressed with a 95% confidence interval (CI). The log–rank test (26) was used to compare survival distributions between the various putative prognostic parameters, based on univariate analysis of age, gender, eye involved, tumor site in relation to the equator, tumor diameter, tumor thickness, retinal detachment, tumor volume, tumor–macula distance, tumor– optic disc distance, eye volume receiving ⱖ30 CGE (corresponding to 50% of the total dose), lens volume receiving ⱖ30 CGE, ciliary body volume receiving ⱖ30 CGE, retinal area receiving ⱖ30 CGE, and macular area receiving ⱖ30 CGE. For some parameters, such as optic disc area receiving ⱖ30 CGE and length of optic nerve receiving ⱖ30 CGE, the median values were equal to zero. We therefore analyzed the optic disc area irradiated and the length of optic nerve irradiated at the dose ⱖ12 CGE (corresponding to 20% of the total dose) with a median value different from zero. Only parameters found to be significant on univariate analysis were introduced into the Cox model. For multivariate analysis of metastasis-free survival, the “local recurrence” variable was introduced into the Cox model to evaluate the impact of this event on the incidence of metastasis. A p value ⬍0.05 was considered to be significant. The tumor site was defined in relation to the equator line. If the melanoma reached the equator line, it was called “at the equator.” If the melanoma was behind or in front of the equator line without reaching it, it was called “posterior” or “anterior,” respectively.
RESULTS Ninety-one percent of the 1,406 patients were referred by the Institut Curie ophthalmology department. There were 700 (49.8%) males, and 703 (50%) left eyes were involved. Sixty-one (4.4%) tumors were anterior, 588 (41.8%) were at the equator, and 757 (53.8%) were posterior. Extensive retinal detachment was presented in 371 (26.4%) eyes, but for 4 eyes (0.3%), retinal detachment status was unknown. Other patient characteristics, tumor characteristics, and dosimetric parameters are reported in Table 1. The median follow-up of
782
I. J. Radiation Oncology
● Biology ● Physics
Volume 65, Number 3, 2006
Table 1. Patient, tumor, and dosimetric characteristics (n ⫽ 1406)
Age Tumor diameter Tumor thickness Tumor volume Tumor–optic disc distance Tumor–macula distance Retina area receiving ⱖ30 CGE Eyeball volume receiving ⱖ30 CGE Lens volume receiving ⱖ30 CGE Ciliary body volume receiving ⱖ30 CGE Macular area receiving ⱖ30 CGE Optic disc area receiving ⱖ12 CGE Length of optic nerve receiving ⱖ12 CGE
Mean
Median
Range
61 years 13.3 mm 5.4 mm 0.6 cc 4.3 mm 3.8 mm 31% 2.9 cc 58 cc 41% 48% 45% 2.6 mm
58 years 13 mm 4.8 mm 0.4 cc 3.5 mm 2.4 mm 31% 2.8 cc 58.8 cc 37% 41% 33% 0.8 mm
15–90 years 2.5–24.4 mm 1.2–14 mm 0–3.6 cc 0–21.6 mm 0–23.3 mm 0–74% 0.2–6.6 cc 0–100% 0–100% 0–100% 0–100% 0–12.3 mm
Abbreviation: CGE ⫽ cobalt Gray equivalent.
surviving patients was 73 months (range, 24 –142 months). Nine hundred sixty-six patients (69.4%) were alive with no evidence of disease. Forty-six patients (3.3%) were alive with metastatic disease. Eleven patients (0.8%) were alive with a second cancer. Thirty-one patients (2.2%) were lost to follow-up (25 patients with no evidence of disease and 6 patients with metastatic disease at last follow-up). Local control The 5-year local control rate was 96% (range, 95.4 – 96.6%). Fifty-two local recurrences were diagnosed during follow-up: 36.5% during the first 18 months and 61.5% during the first 3 years after irradiation. The median interval between irradiation and local recurrence was 21 months. The univariate analysis results are reported in Table 2. Two independent unfavorable prognostic factors were identified on multivariate analysis: large tumor diameter (ⱖ13 mm) with a relative risk (RR) of 2.18 (95% CI, 1.21–3.91; p ⫽ 0.003) and small macular area (⬍41%) receiving ⱖ30 CGE with RR of 2 (95% CI, 1.12–3.57; p ⫽ 0.01). The therapy for local recurrence consisted of enucleation for 33 patients (63%), another course of proton beam radiotherapy for 3 patients, 125I plaque brachytherapy for 3 patients, and thermotherapy for 1 patient. Two patients did not receive local therapy because of concomitant metastatic disease. Twenty-six (50%) of the 52 patients with local recurrence developed metastases, and 30 (58%) died. Overall survival Three hundred forty-two patients (24.3%) died during follow-up. Causes of death were melanoma for 242 patients (70.8%), intercurrent disease for 37 patients (10.8%), second cancer for 16 patients (4.7%), and unknown for 47 patients (13.7%). The 5-year overall survival rate was 79% (range, 77.8 – 80.2%). The univariate analysis results are reported in Table 2. Six independent prognostic factors were identified on multivariate analysis: advanced age (ⱖ61 years) with an RR of 2.40 (95% CI, 1.91–3.01; p ⬍ 0.0001), male gender with
an RR of 1.44 (95% CI, 1.17–1.81; p ⫽ 0.0003), large tumor diameter (ⱖ13 mm) with an RR 1.55 (95% CI, 1.17–2.07; p ⬍ 0.0001), median and anterior tumor sites with an RR 1.63 (95% CI, 1.26 –2.08; p ⬍ 0.0001), large tumor thickness (ⱖ4.8 mm) with an RR 1.34 (95% CI, 1.03–1.75; p ⫽ 0.02), and large retinal area (ⱖ31%) receiving ⱖ30 CGE with an RR 1.40 (95% CI, 1.08 –1.8; p ⫽ 0.003). Metastasis-free survival The 5-year metastasis-free survival rate was 80.6% (range, 79.4 – 81.7%). The liver was the first metastatic site in 81.4% of cases. The univariate analysis results are reported in Table 2. Six independent prognostic factors were identified on multivariate analysis: older age (ⱖ61 years) with RR 1.42 (95% CI, 1.12–1.79; p ⫽ 0.004), retinal detachment with RR 1.38 (95% CI, 1.04 –1.73; p ⫽ 0.01), large tumor volume (ⱖ0.41 cc) with RR 1.74 (95% CI, 1.25–2.41; p ⬍ 0.0001), median and anterior tumor sites with RR 1.85 (95% CI, 1.4 –2.5; p ⬍ 0.0001), large retinal area (ⱖ31%) receiving ⱖ30 CGE with RR 1.38 (95% CI, 1.04 –1.83; p ⫽ 0.002), and local recurrence with RR 4.17 (95% CI, 2.71– 6.41; p ⬍ 0.0001). Ocular complications Five-year ocular complication rates are reported in Table 3 for the following complications: maculopathy, papillopathy, keratitis, vitreous hemorrhage, and intraocular inflammation. Some of these ocular complications were not associated with any visual symptoms and were exclusively diagnosed on systematic follow-up examinations. Enucleation for complications (EC) Enucleation performed for local recurrence was excluded from this analysis. Ninety-nine patients (7%) underwent enucleation for complications: neovascular glaucoma with severe pain in 83 patients, loss of vision in 5 patients, and for an unknown cause in 1 patient. Enucleation for complications was performed during the first 2 years after irradi-
Protontherapy of uveal melanoma
● R. DENDALE et al.
783
Table 2. Univariate analysis Factors Age ⬍61 years ⱖ61 years Gender F M Laterality Left Right Tumor location from equator Posterior At equator Anterior Tumor diameter ⬍13 mm ⱖ13 mm Median tumor thickness ⬍4.8 mm ⱖ4.8 mm Tumor volume ⬍0.41 cc ⱖ0.41 cc Retinal detachment No Yes Distance tumor-papilla ⬍3.5 mm ⱖ3.5 mm Distance tumor-macula ⬍2.4 mm ⱖ2.4 mm Surface-irradiated retina 30 CGE ⬍31% ⱖ31% Volume of irradiated eyeball 30 CGE ⬍2.8 cc ⱖ2.8 cc Volume of irradiated lens 30 CGE ⬍43% ⱖ43% Volume of irradiated ciliary body 30 CGE ⬍37% ⱖ37% Irradiated optic nerve length 12 CGE ⬍0.8 mm ⱖ0.8 mm Irradiated papilla surface 12 CGE ⬍33% ⱖ33% Irradiated macula surface 30 CGE ⬍41% ⱖ41%
Local recurrence RR (p) 1 1.04 ns
Metastatic disease RR (p) 1 1.38 (0.004)
Overall survival RR (p)
Secondary enucleation RR (p)
1 2.28 (0.0001)
1 1 ns
1 1.18 ns
1 1.06 ns
1 1.40 (0.0013)
1 1.21 ns
1 1.17 ns
1 1.17 ns
1 1.2 ns
1 1.12 ns
1 1.51 5.75 (⬍0.0001)
1 3 2.05 (⬍0.0001)
1 2.41 2.03 (⬍0.0001)
1 2.25 2.23 (0.0004)
1 2.30 (0.003)
1 2.81 (⬍0.0001)
1 2.60 (⬍0.0001)
1 2.43 (⬍0.0001)
1 1.91 (0.02)
1 2.83 (⬍0.0001)
1 2.24 (⬍0.0001)
1 6.27 (⬍0.0001)
1 1.94 (0.02)
1 3.11 (⬍0.0001)
1 1.63 (⬍0.0001)
1 5.02 (⬍0.0001)
1 2 (0.01)
1 2.27 (⬍0.0001)
1 1.77 (⬍0.0001)
1 2.57 (⬍0.0001)
1 1.15 ns
1 1.25 ns
1 1.2 ns
1 1.27 (0.038)
1 2.06 (0.01)
1 1.77 (⬍0.001)
1 1.40 (0.0015)
1 1.47 (0.05)
1 1.73 (⬍0.0001)
1 1.61 (⬍0.0001)
1 2.15 (⬍0.0001)
1 2.61 (⬍0.0001)
1 1.48 ns
1 1.91 (⬍0.0001)
1 1.76 (⬍0.0001)
1 3.73 (⬍0.0001)
1 2.21 (0.006)
1 1.82 (⬍0.0001)
1 1.82 (⬍0.0001)
1 4.46 (⬍0.0001)
1 2.22 (0.005)
1 2.15 (⬍0.0001)
1 2.11 (⬍0.0001)
1 4.91 (⬍0.0001)
1 1.25 ns
1 1.11 ns
1 1.09 ns
1 1.08 ns
1 1 ns
1 1.05 ns
1 1.04 ns
1 1.1 ns
1 0.81 ns
1 0.94 ns
1 0.54 (0.032)
1 0.69 (0.0013)
Abbreviations: RR ⫽ relative risk; ns ⫽ not significant; CGE ⫽ cobalt Gray equivalent.
ation in 54% of cases and during the first 3 years in 71% of cases. Five-year EC rates were 7.7% (range, 6.7– 8.7%). The univariate analysis results are reported in Table 2. Two independent unfavorable prognostic factors were identified on multivariate analysis: large tumor thickness (ⱖ4.8 mm) with RR 5.42 (95% CI, 2.87– 40.23; p ⬍ 0.0001) and large
lens volume (ⱖ43%) receiving ⱖ30 CGE with RR 2.47 (95% CI, 1.48 – 4.12; p ⫽ 0.0002). Visual acuity Baseline visual acuity and visual acuity at 5 years of follow-up are reported in Table 4. At 5 years, visual acuity
784
I. J. Radiation Oncology
● Biology ● Physics
Table 3. Actuarial ocular complication rates
Maculopathy Papillopathy Glaucoma Cataract Keratitis Vitreous hemorrhage Intraocular inflammation
5-year (%)
Confidence interval
66.5 23.4 28.6 61.8 11.5 13.9 27.5
(63.3–69.7) (20.5–26.3) (26–31.2) (59–64.7) (9.6–13.4) (11.9–16) (24.9–30.1)
remained stable for 38% of patients, decreased for 56% of patients, and improved for 6% of patients. DISCUSSION This series comprises 1,406 patients with a median follow-up of 73 months and a minimum follow-up of 2 years. This is a very homogeneous series in terms of both patient and tumor characteristics and treatment modalities; clip surgery, treatment planning, and treatment delivery were performed according to predefined procedures. These procedures have remained unchanged since 1991, apart from successive versions of EYEPLAN software. This series was comprised of only a few small anterior tumors as a result of the initial patient selection, because small anterior tumors are usually treated by 125I plaque brachytherapy, whereas small posterior tumors are usually treated by proton beam radiotherapy. Surgical access to the posterior segment for placement of episcleral plaques is fairly difficult, and coverage of the tumor base by a plaque may be difficult or incomplete for tumors invading the optic disc, because of the emergence of the optic nerve. Distribution of the intraocular dose delivered by brachytherapy is also much more heterogeneous than that delivered by proton beam radiotherapy. This dose heterogeneity is even more marked for larger tumor volumes. The maximum intraocular dose can exceed 200 Gy for some applications, whereas it is always equal to the prescribed dose of 60 CGE for proton beam radiotherapy. Some published studies have compared the results obtained by plaque brachytherapy and heavy particle (protons or Helium) beam irradiation for tumors with similar characteristics and seem to confirm a higher ocular toxicity after brachytherapy (27, 28). Proton beam radiotherapy is therefore probably a more appropriate treatment option for large posterior tumors, very close to the optic nerve, whereas small anterior tumors can be effectively treated by plaque brachytherapy. On the basis of these selection criteria (small anterior melanomas treated by plaque brachytherapy and other melanomas treated by proton beam radiotherapy), we have not found any difference in terms of toxicity or local control between these two treatment modalities (29). Local control The 5-year local control rate of 96% observed in this series was very similar to the rates reported by others (21,
Volume 65, Number 3, 2006
27, 30 –33), ranging from 90.5% to 96%. The median time to local recurrence was 21 months, and two-thirds of local recurrences were during the first 3 years after irradiation. Munzenrider et al. (31) reported a median interval of 14 months between irradiation and local recurrence. Like others (27), we found that tumor diameter was an independent prognostic factor for local recurrence. The second independent prognostic factor identified by multivariate analysis was a greater macula area (ⱖ41%) receiving ⱖ30 CGE. This means that a better local control rate was observed for eyes receiving a greater macular dose compared to eyes receiving a lower dose. This suggests better local control for tumors situated close to the macula. Three possible explanations that are probably intimately related can be proposed. First, this result could reflect the recruitment bias of our series, which included only a small number of small anterior tumors, treated mostly by plaque brachytherapy. The majority of anterior tumors treated by proton therapy were therefore larger tumors with a higher risk of recurrence. The second explanation could be a more precise definition of tumor margins for tumors situated close to the macula. Tumor margins were determined by two different techniques: The posterior limit of the melanoma was determined by ocular fundus photographs or angiography, allowing precise definition of its relation to the macula, whereas the limits of more anterior tumors are determined by clips sutured onto the sclera over the tumor margins as determined by transillumination. However, the anterior limit of an achromic tumor can be difficult to determine by transillumination. The third explanation is related to the earlier diagnosis of these posterior tumors, which would therefore be associated with less necrosis and possibly more radiosensitivity. Tumors situated close to the macula are more rapidly symptomatic, leading to earlier ophthalmologic examination and diagnosis, whereas a more anterior tumor can remain asymptomatic for a long time. Overall survival and metastasis-free survival The 5-year overall survival rate in our study was 79%, corresponding to the published rates (14, 30, 32, 34), ranging from 70.3% to 78%. The 5-year metastasis-free survival rate was 80.6% and was comparable to the rates reported in other published series, ranging from 74.3% to 91%. This indicated that in 20% of our patients, local treatment alone Table 4. Comparison of VA between initial examination and 5-year follow-up: percentage of patients VA at 5 years Baseline VA
⬍20/200
20/200 to ⬍20/40
ⱖ20/40
Total
⬍20/200 20/200 to ⬍20/40 ⱖ20/40 Total
9.1% 26.4% 18.6% 54.1%
1% 9.6% 10.9% 21.5%
0.6% 4.2% 19.6% 24.4%
10.5% 40.3% 49.2% 100%
Abbreviation: VA ⫽ visual acuity.
Protontherapy of uveal melanoma
was insufficient. As reported by other authors, patients who developed local recurrence had a poorer prognosis than those who did not, because 50% with local recurrence developed metastatic disease. Our multivariate analysis demonstrated local recurrence to be an independent unfavorable prognostic factor for metastasis-free survival, but it is unclear whether metastases are directly related to local recurrence or whether they reflect a particularly aggressive phenotype of the primary tumor. The prognostic value of a local recurrence event on metastasis incidence has already been described by others (34, 35). Prescher et al. (36) showed that monosomy 3 had a negative prognostic impact on the metastasis incidence and survival rate. Further studies are necessary to identify such tumors, but they would require performing systematic biopsy in the primary tumors. Eye complications and enucleation Proton beam radiotherapy can induce ocular complications, particularly glaucoma (31, 37). According to Char et al. (32), 34.9% of treated eyes developed neovascular glaucoma. In our series, the enucleation rate related to ocular complications was 7%, corresponding to the published EC rates (14, 30 –34, 37, 38), which range from 2.2% to 18% and are related to neovascular glaucoma in 71% to 100% of cases. In our series, 64% of these enucleations were performed during the first 3 years. Egan et al. (34) and Fuss et al. (30) reported EC rates during the first 3 years of 64% and 100%, respectively. Wilson and Hungerford (38) reported a mean interval between irradiation and EC of 28 months. Some authors have reported retinal detachment (37) and tumor– optic disc distance (14, 34) to be independent prognostic factors for EC. In our series, only tumor thickness and lens volume irradiated were found to be significant independent prognostic factors on multivariate analysis. The risk of secondary enucleation is therefore related to a tumor characteristic (tumor thickness) and a dosimetric parameter (dose to lens). Lens protection was not checking in our dosimetric planning. However, we always tried to reduce the dose delivered to the corneal limbus, but dosimetric planning for large tumors, anterior tumors, and tumors invading the ciliary body often required eye positions leading to irradiation of the anterior chamber and therefore the lens. Complications leading to enucleation could be related to irradiation of the anterior chamber, despite the avoidance of the lens itself. There was a high cataract rate in our series, but cataract surgery did not seem to have any impact on the rate of glaucoma and EC (data not shown). Unfortunately, data corresponding to the dose received by the anterior chamber were not recorded in this study. Tumor thickness is a reflection of tumor size, which determines the ocular volume irradiated. The tumor thickness and the size of the collimator are indeed closely related. With a large collimator, it becomes difficult to avoid the irradiation of the anterior chamber, and the irradiated ocular volume is always large. Various methods of reducing treatment toxicity are being investigated. They include total dose reduction, reduction of
● R. DENDALE et al.
785
safety margins, increased fractionation, and concomitant treatment with radioprotectors, anticoagulants, or thermotherapy.
Total dose reduction The initial protocols for ocular proton beam radiotherapy were established with a small number of high-dose fractions, because the clinical data available at that time indicated favorable results for this type of irradiation on skin melanomas (12, 39). Plaque brachytherapy, the first radiotherapy modality for uveal melanoma, used empiric doses between 70 and 140 Gy. These high doses were possible, because no substantial normal-tissue reaction was observed (12). Proton scatter in the tissues is much less marked than during conventional external beam radiotherapy (photons or electrons). The sudden dose drop posterior to the Bragg peak and the use of collimators manufactured according to the tumor profile allow the use of very small safety margins around the tumor, resulting in a much smaller intraocular volume irradiated. The equivalent dose delivered by current proton beam radiotherapy regimens is much higher than that delivered by conventional radiotherapy with 2 Gy fractions. Bentzen et al. (40) reported that the estimated dose to control 80% of tumors measuring 10 mm and 20 mm in diameter was 22 Gy and 28 Gy, respectively, delivered in 9 fractions. The dose of 60 CGE delivered in our series corresponds to 2 to 2.5 times these values. Some authors (41, 42) have recently challenged the concept of the radioresistance of melanoma. In a dose reduction study, Gragoudas et al. (43) reported the 5-year results of a randomized trial on 188 patients comparing doses of 50 CGE and 70 CGE delivered in 5 fractions over 5 days and found no significant difference between the two dose regimens in terms of ocular toxicity or local control.
Reduction of the tumor safety margin A reduction of the safety margin around the tumor would decrease the ocular irradiation volume. However, based on a series of patients with a follow-up of 15 years, Egger et al. (21) reported that localized reduction of the tumor safety margins to avoid irradiation of the optic nerve and optic disc induced a decrease of the 5-year local control rate from 96.3% to 71.3%. Other teams (44) have reported that reduction of tumor safety margins is possible in particular cases by using 3D CT reconstruction and by introducing anatomic modifications in their EYEPLAN dosimetric planning to obtain a better representation of the region of interest concerned by the tumor safety margin reduction. We recently reported (45) the preliminary results of a study combining very localized reduction of the safety margin close to the macula and localized transpupillary thermotherapy and obtained good preservation of visual acuity with no reduction of the local control rate.
786
I. J. Radiation Oncology
● Biology ● Physics
Increased fractionation Late complications are directly correlated with the dose per fraction. During proton beam radiotherapy, 14 to 15 CGE are delivered in each fraction. An increased number of fractions (4 to 8 fractions) with the same total dose of 60 CGE could possibly decrease the ocular complication rate without compromising local control. However, doubling of the number of fractions compared to the current regimen would raise a number of practical and cost problems. This type of fractionation study would have to be conducted very cautiously with strict tumor selection and rigorous discontinuation criteria. Concomitant treatments WR2721 (amifostine) or WR-77913 (aminophosphorothioate) have a known radioprotective effect (46 – 48). Several teams (46, 47, 49, 50) have tested these molecules in patients treated with irradiation for various tumor locations. They have reported an encouraging radioprotective effect on normal tissues. These products must be administered by i.v. injection before each treatment session. Hypofractionated proton beam radiotherapy seems to be particularly suitable for radioprotective therapy. Anticoagulants or platelet aggregation inhibitors can decrease tissue hypoxia by reducing the number of micro-
Volume 65, Number 3, 2006
thrombi (51) and could therefore reduce the toxic effect of irradiation. In fact, tumor and vascular necrosis, vascular thrombosis, and intraocular inflammation after proton beam radiotherapy can be responsible for severe hypoxia of healthy intraocular tissues (52). This hypoxia predisposes to angiogenesis, which can be responsible for the appearance of neovascular glaucoma. A combination of transpupillary thermotherapy and protons could reduce post proton therapy ocular toxicity. Char et al. (53) reported an exudative retinal detachment reduction by thermotherapy after a proton radiation. CONCLUSION We have reported the results of a homogeneous series of 1,406 patients treated for uveal melanoma by proton beam radiotherapy in a single center using the same treatment procedure. The local control rate was good and similar to previously published rates, but ocular complication rates and the secondary enucleation rate without complication during the first year remain significant. The most frequent ocular toxicity was neovascular glaucoma, but enucleation was avoided in the majority of these patients. Future research should try to reduce the ocular complication rate while maintaining the current good local control rates.
REFERENCES 1. Vidal JL, Bacin F, Albuisson E, et al. [“Melanoma 92”. Epidemiological study of uveal melanoma in France]. J Fr Ophtalmol 1995;18:520 –528. 2. Scotto J, Fraumeni JFJ, Lee JA. Melanomas of the eye and other noncutaneous sites: epidemiologic aspects. J Natl Cancer Inst 1976;56:489 – 491. 3. Stallard HB. Partial choroidectomy. Br J Ophthalmol 1966; 50:660 – 662. 4. Damato B. Adjunctive plaque radiotherapy after local resection of uveal melanoma. Front Radiat Ther Oncol 1997;30: 123–132. 5. Foulds WS. The local excision of choroidal melanomata. Trans Ophthalmol Soc U K 1973;93:343–346. 6. Kivela T, Puusaari I, Damato B. Transscleral resection versus iodine brachytherapy for choroidal malignant melanomas 6 millimeters or more in thickness: A matched case-control study. Ophthalmology 2003;110:2235–2244. 7. Lommatzsch PK. Results after beta-irradiation (106Ru/106Rh) of choroidal melanomas. Twenty years’ experience. Am J Clin Oncol 1987;10:146 –151. 8. Stallard HB. Radiotherapy for malignant melanoma of the choroid. Br J Ophthalmol 1966;50:147–155. 9. Dieckmann K, Georg D, Zehetmayer M, et al. LINAC based stereotactic radiotherapy of uveal melanoma: 4 years clinical experience. Radiother Oncol 2003;67:199 –206. 10. Char DH, Castro JR, Quivey JM, et al. Helium ion charged particle therapy for choroidal melanoma. Ophthalmology 1980;87:565–570. 11. Lumbroso L, Levy C, Plancher C, et al. [Results of proton beam irradiation for treatment of choroidal melanoma]. J Fr Ophtalmol 2002;25:290 –297. 12. Gradoudas ES, Goitein M, Koehler A, et al. Proton irradiation of choroidal melanomas. Preliminary results. Arch Ophthalmol 1978;96:1583–1591.
13. Gragoudas ES, Goitein M, Koehler AM, et al. Proton irradiation of small choroidal malignant melanomas. Am J Ophthalmol 1977;83:665– 673. 14. Egger E, Zografos L, Munkel G, et al. Results of proton radiotherapy for uveal melanomas. Front Radiat Ther Oncol 1997;30:111–122. 15. Brovkina AF, Zarubei GD. Ciliochoroidal melanomas treated with a narrow medical proton beam. Arch Ophthalmol 1986; 104:402– 404. 16. Seddon JM, Gragoudas ES, Albert DM, et al. Comparison of survival rates for patients with uveal melanoma after treatment with proton beam irradiation or enucleation. Am J Ophthalmol 1985;99:282–290. 17. Rosenwald JC, Bridier A, Sabattier R, et al. [Preliminary dosimetry study of the Orsay synchrocyclotron proton beam with a view to applying it therapeutically]. Ophtalmologie 1990;4:123–127. 18. Lumbroso L, Levy C, Plancher C, et al. [Results of proton beam irradiation for treatment of choroidal melanoma]. J Fr Ophtalmol 2002;25:290 –297. 19. Desjardins L, Levy C, d’Hermies F, et al. [Initial results of proton therapy in choroidal melanoma at the d’Orsay center for Proton Therapy; The first 464 cases]. Cancer Radiother 1997;1:222–226. 20. Schlienger P, Habrand JL, Schwartz L, et al. Initial results with one-year minimum follow-up of the first 146 patients with a uveal melanoma treated with protons at CPO (Orsay). Bull Cancer Radiother 1996;83:212s–214s. 21. Egger E, Schalenbourg A, Zografos L, et al. Maximizing local tumor control and survival after proton beam radiotherapy of uveal melanoma. Int J Radiat Oncol Biol Phys 2001;51:138 –147. 22. Goitein M, Miller T. Planning proton therapy of the eye. Med Phys 1983;10:275–283.
Protontherapy of uveal melanoma
23. Perret C. The therapy planning program EYE. Swiss Institute for Nuclear Research, Internal Report; 1987. 24. Sheen M. EYEPLAN proton therapy planning program. Bebington, Wirral, UK: Claterbridge Centre for Oncology, Internal Report; 1995. 25. Kaplan E, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1971;53:457– 481. 26. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 1966;50:163–170. 27. Wilson MW, Hungerford JL. Comparison of episcleral plaque and proton beam radiation therapy for the treatment of choroidal melanoma. Ophthalmology 1999;106:1579 –1587. 28. Char DH, Quivey JM, Castro JR, et al. Helium ions versus iodine 125 brachytherapy in the management of uveal melanoma. A prospective, randomized, dynamically balanced trial. Ophthalmology 1993;100:1547–1554. 29. Desjardins L, Lumbroso L, Levy C, et al. [Treatment of uveal melanoma with iodine 125 plaques or proton beam therapy: Indications and comparison of local recurrence rates]. J Fr Ophtalmol 2003;26:269 –276. 30. Fuss M, Loredo LN, Blacharski PA, et al. Proton radiation therapy for medium and large choroidal melanoma: Preservation of the eye and its functionality. Int J Radiat Oncol Biol Phys 2001;49:1053–1059. 31. Munzenrider JE, Verhey LJ, Gragoudas ES, et al. Conservative treatment of uveal melanoma: Local recurrence after proton beam therapy. Int J Radiat Oncol Biol Phys 1989;17:493– 498. 32. Char DH, Kroll SM, Castro J. Ten-year follow-up of helium ion therapy for uveal melanoma. Am J Ophthalmol 1998;125: 81– 89. 33. Courdi A, Caujolle JP, Grange JD, et al. Results of proton therapy of uveal melanomas treated in Nice. Int J Radiat Oncol Biol Phys 1999;45:5–11. 34. Egan KM, Ryan LM, Gragoudas ES. Survival implications of enucleation after definitive radiotherapy for choroidal melanoma: An example of regression on time-dependent. Arch Ophthalmol 1998;116:366 –370. 35. Harbour JW, Char DH, Kroll S, et al. Metastatic risk for distinct patterns of postirradiation local recurrence of posterior uveal melanoma. Ophthalmology 1997;104:1785–1792; discussion 1792–1783. 36. Prescher G, Bornfeld N, Hirche H, et al. Prognostic implications of monosomy 3 in uveal melanoma. Lancet 1996;347: 1222–1225. 37. Foss AJ, Whelehan I, Hungerford JL, et al. Predictive factors for the development of rubeosis following proton beam radiotherapy for uveal melanoma. Br J Ophthalmol 1997;81:748 –754. 38. Wilson MW, Hungerford JL. Comparison of episcleral plaque and proton beam radiation therapy for the treatment of choroidal melanoma. Ophthalmology 1999;106:1579 –1587.
● R. DENDALE et al.
787
39. Habermalz HJ, Fischer JJ. Radiation therapy of malignant melanoma: Experience with high individual treatment doses. Cancer 1976;38:2258 –2262. 40. Bentzen SM, Overgaard J, Thames HD, et al. Clinical radiobiology of malignant melanoma. Radiother Oncol 1989;16: 169 –182. 41. Fowler JF. Radioresponsiveness of malignant melanoma to fractionated irradiation. Int J Radiat Oncol Biol Phys 1984; 10:439 – 440. 42. Noel G, Simon JM, Valery CA, et al. Linac radiosurgery for brain metastasis of melanoma. Stereotact Funct Neurosurg 2002;79:245–255. 43. Gragoudas ES, Lane AM, Regan S, et al. A randomized controlled trial of varying radiation doses in the treatment of choroidal melanoma. Arch Ophthalmol 2000;118:773–778. 44. Chauvel P, Iborra N. Uveal melanoma relapses: The controversy about safety margins. PTCOG. Paris; 2004. 45. Dendale R, Lumbroso-Le Rouic R, Delacroix S, et al. Transpupillary thermotherapy and protontherapy association for posterior choroidal melanoma [Abstract]. PTCOG. Paris; 2004. 46. Bourhis J, De Crevoisier R, Abdulkarim B, et al. A randomized study of very accelerated radiotherapy with and without amifostine in head and neck squamous cell carcinoma. Int J Radiat Oncol Biol Phys 2000;46:1105–1108. 47. Antonadou D, Throuvalas N, Petridis A, et al. Effect of amifostine on toxicities associated with radiochemotherapy in patients with locally advanced non-small-cell lung cancer. Int J Radiat Oncol Biol Phys 2003;57:402– 408. 48. Belkacemi Y, Rat P, Piel G, et al. Lens epithelial cell protection by aminothiol WR-1065 and anetholedithiolethione from ionizing radiation. Int J Cancer 2001;96:15–26. 49. Athanassiou H, Antonadou D, Coliarakis N, et al. Protective effect of amifostine during fractionated radiotherapy in patients with pelvic carcinomas: Results of a randomized trial. Int J Radiat Oncol Biol Phys 2003;56:1154 –1160. 50. Reddy VN, Ikebe H, Giblin FJ, et al. Effect of radioprotective agents on X-ray cataracts. Lens Eye Toxic Res 1989;6:573– 588. 51. Salomon O, Rosenberg N, Steinberg DM, et al. Nonarteritic anterior ischemic optic neuropathy is associated with a specific platelet polymorphism located on the glycoprotein Ibalpha gene. Ophthalmology 2004;111:184 –188. 52. Lumbroso L, Desjardins L, Levy C, et al. Intraocular inflammation after proton beam irradiation for uveal melanoma. Br J Ophthalmol 2001;85:1305–1308. 53. Char DH, Bove R, Phillips TL. Laser and proton radiation to reduce uveal melanoma-associated exudative retinal detachments. Trans Am Ophthalmol Soc 2003;101:53–56.
doi:10.1016/j.ijrobp.2006.01.020
CLINICAL INVESTIGATION
Eye
PROTON BEAM RADIOTHERAPY FOR UVEAL MELANOMA: RESULTS OF CURIE INSTITUT–ORSAY PROTON THERAPY CENTER (ICPO) RÉMI DENDALE, M.D.,*‡ LIVIA LUMBROSO-LE ROUIC, M.D.,† GEORGES NOEL, M.D.,‡ LOÏC FEUVRET, M.D.,‡ CHRISTINE LEVY, M.D.,† SABINE DELACROIX, PH.D.,‡ ANNE MEYER, M.D.,§ CATHERINE NAURAYE, PH.D.,‡ ALEJANDRO MAZAL, PH.D.,*‡ HAMID MAMMAR, M.D.,* PAUL GARCIA, PH.D.,‡ FRANÇOIS D’HERMIES, M.D.,§ ERIC FRAU, M.D.,储 CORINE PLANCHER, D.M.,¶ BERNARD ASSELAIN, M.D.,¶ PIERRE SCHLIENGER, M.D.,*‡ JEAN JACQUES MAZERON, M.D.,‡ †‡ AND LAURENCE DESJARDINS, M.D. Departments of *Radiation Oncology, †Ophthalmology, and ¶Biostatistics, Curie Institut, Paris, France; ‡Curie Institut–Orsay Protonthérapie Center, Orsay, France; §Department of Ophthalmology, Hotel Dieu Hospital, Paris, France; 储Department of Ophthalmology, Le Kremlin Bicêtre Hospital, Le Kremlin Bicêtre, France Purpose: This study reports the results of proton beam radiotherapy based on a retrospective series of patients treated for uveal melanoma at the Orsay Center. Methods and Materials: Between September 1991 and September 2001, 1,406 patients with uveal melanoma were treated by proton beam radiotherapy. A total dose of 60 cobalt Gray equivalent (CGE) was delivered in 4 fractions on 4 days. Survival rates were determined using Kaplan–Meier estimates. Prognostic factors were determined by multivariate analysis using the Cox model. Results: The median follow-up was 73 months (range, 24 –142 months). The 5-year overall survival and metastasis-free survival rates were 79% and 80.6%, respectively. The 5-year local control rate was 96%. The 5-year enucleation for complications rate was 7.7%. Independent prognostic factors for overall survival were age (p < 0.0001), gender (p < 0.0003), tumor site (p < 0.0001), tumor thickness (p ⴝ 0.02), tumor diameter (p < 0.0001), and retinal area receiving at least 30 CGE (p ⴝ 0.003). Independent prognostic factors for metastasis-free survival were age (p ⴝ 0.0042), retinal detachment (p ⴝ 0.01), tumor site (p < 0.0001), tumor volume (p < 0.0001), local recurrence (p < 0.0001), and retinal area receiving at least 30 CGE (p ⴝ 0.002). Independent prognostic factors for local control were tumor diameter (p ⴝ 0.003) and macular area receiving at least 30 CGE (p ⴝ 0.01). Independent prognostic factors for enucleation for complications were tumor thickness (p < 0.0001) and lens volume receiving at least 30 CGE (p ⴝ 0.0002). Conclusion: This retrospective study confirms that proton beam radiotherapy ensures an excellent local control rate. Further clinical studies are required to decrease the incidence of postirradiation ocular complications. © 2006 Elsevier Inc. Uveal melanoma, Proton beam radiotherapy, Local control, Survival, Enucleation, Ocular complication.
Uveal melanoma is the most frequent ocular tumor but remains rare with a reported incidence of 6 to 7 cases per 100,000 inhabitants (1, 2). Recently, surgical enucleation was the standard treatment, but various eye-preserving treatment modalities have subsequently been developed, such as local resection (3– 6), plaque brachytherapy (7, 8), external beam radiotherapy with photons using a linear accelerator (9), helium ions (10), and proton beam therapy (11–15). Proton beam radiotherapy is a conservative alternative to enucleation for the management of uveal melanoma, because a comparison between enucleation and radiotherapy
does not reveal any significant difference in terms of overall survival and metastasis-free survival (16). In 1954, at the initiative of Irène Joliot-Curie, a synchrocyclotron particle accelerator was constructed by Philips at the Orsay University campus. It was initially devoted to nuclear physics research, but a phase of experimentation to develop medical applications was initiated in 1987 (17). Proton beam radiotherapy at Orsay, essentially for uveal melanoma, was performed for the first time in 1991 (18 –20). In this study, we will report the results of proton beam radiotherapy for uveal melanoma on a retrospective consecutive series of patients with a minimum follow-up of 2 years.
Reprint requests to: Rémi Dendale, M.D., Department of Radiotherapy, Curie Institut, 26 rue d’Ulm, 75248, Paris Cedex 05, France. Tel: (⫹33) 1-4432-4632; Fax: (⫹33) 1-4432-4616;
E-mail: [email protected] Received April 26, 2005, and in revised form Jan 17, 2006. Accepted for publication Jan 18, 2006.
INTRODUCTION
780
Protontherapy of uveal melanoma
METHODS AND MATERIALS Patients Patients were referred to the Curie Institut–Orsay Proton Therapy Center by three hospitals (Curie Institut, Hôtel Dieu Hospital, Kremlin Bicêtre Hospital). All patient and tumor data were prospectively recorded in a database. From this database, we retrospectively selected patients with uveal melanoma treated with proton beam radiotherapy and followed during a minimum of 2 years. Not included in this study were patients with a previous local treatment, metastases, extrascleral tumor invasion, or iris melanoma at diagnosis. A total of 1,406 case files were selected for patients treated between September 1991 and September 2001. The clinical diagnosis was established 3 or 4 weeks before radiotherapy by the referring hospital on the basis of clinical ocular fundus examination, A and B mode ultrasound, and fluorescein angiography. Patients with little and anterior uveal melanoma were treated by 125I brachytherapy plaque and were not analyzed in this study.
Treatment procedure The procedure after diagnosis was precisely defined. One to 2 weeks before radiotherapy, patients underwent a surgical procedure for precise demarcation of tumor margins by an ophthalmologist. This surgery consisted of suturing 4 or 5 tantalum clips (diameter: 2.5 mm, thickness: 0.5 mm) onto the sclera around the tumor base. The limits of the tumor base were determined by transillumination. Proton beam radiotherapy was planned and performed at Curie Institut–Orsay Proton Therapy Center 1 to 2 weeks after surgery. The treatment technique has been described previously (18 –21). Briefly, head immobilization during radiotherapy was ensured by a custom-made thermoplastic mask associated with a bite block. EYEPLAN software (22–24) was used for dosimetry based on three-dimensional (3D) reconstruction of the eye, including the tumor. Tumor margins were drawn by the referring center ophthalmologist and were verified by the radiation oncologist. The optimal dosimetric eye position was determined according to the following criteria: total irradiation of the tumor surrounded by a safety margin of 2.5 mm; and reduction of the dose delivered to the lacrimal gland, macula, optic disc, and lens in decreasing order of priority. Beam characteristics were selected to place the 90% isodose 2.5 mm behind the tumor. The lateral collimator border defined the site of the isodose 50%. During treatment, patients were seated on a robot chair facing the beam and were asked to fix a small target light placed at a known angle adopted during the treatment planning process. As the beam remained horizontal and immobile, the patient was moved into the treatment position. The robot chair can be moved in all directions and angles with a precision of about 0.1 mm and 0.1°. A systematic attempt was made to keep the eyelids out of the beam’s field. Before each irradiation session, clip positions were verified by 2 orthogonal Polaroid X-rays, and the chair was then moved to the correct position. The fixed gaze was monitored by a magnifying camera (⫻10). Irradiation was systematically suspended whenever the eye shifted from the tumor position and was resumed only after the eye had returned to the optimal position. A total dose of 60 cobalt Gray equivalent (CGE) (using a relative biologic effectiveness of 1.1) was delivered in 4 fractions on 4 consecutive days. An ophthalmologic examination was systematically performed by an ophthal-
● R. DENDALE et al.
781
mologist during the week of irradiation. No transpupillary thermotherapy and no other local treatments were done on these patients.
Follow-up Patients were regularly followed by their ophthalmologist from the referring hospital 1 month after irradiation, every 6 months for the first 2 years, and annually thereafter. A complete ocular functional assessment, including determination of visual acuity, lid damage, keratitis, cataract, glaucoma, optic neuropathy, maculopathy, retinopathy, and ocular inflammation, was performed 1, 2, 3, 5, and 10 years after irradiation. A liver ultrasound for detection of metastases was performed every 6 months for 10 years and annually thereafter. Enucleation was performed for local recurrence and for ocular complications such as severe glaucoma, major retinal detachment, retinal hemorrhage, disabling eye pain, or for ocular complications preventing good clinical and B-mode ultrasound follow-up.
Statistical methods Survival rates were calculated from the day of irradiation to the date of the event of interest or the date of last follow-up. Five-year local recurrence, metastasis-free survival, overall survival, and enucleation for complication (EC) rates were determined by the Kaplan–Meier method (25) and were expressed with a 95% confidence interval (CI). The log–rank test (26) was used to compare survival distributions between the various putative prognostic parameters, based on univariate analysis of age, gender, eye involved, tumor site in relation to the equator, tumor diameter, tumor thickness, retinal detachment, tumor volume, tumor–macula distance, tumor– optic disc distance, eye volume receiving ⱖ30 CGE (corresponding to 50% of the total dose), lens volume receiving ⱖ30 CGE, ciliary body volume receiving ⱖ30 CGE, retinal area receiving ⱖ30 CGE, and macular area receiving ⱖ30 CGE. For some parameters, such as optic disc area receiving ⱖ30 CGE and length of optic nerve receiving ⱖ30 CGE, the median values were equal to zero. We therefore analyzed the optic disc area irradiated and the length of optic nerve irradiated at the dose ⱖ12 CGE (corresponding to 20% of the total dose) with a median value different from zero. Only parameters found to be significant on univariate analysis were introduced into the Cox model. For multivariate analysis of metastasis-free survival, the “local recurrence” variable was introduced into the Cox model to evaluate the impact of this event on the incidence of metastasis. A p value ⬍0.05 was considered to be significant. The tumor site was defined in relation to the equator line. If the melanoma reached the equator line, it was called “at the equator.” If the melanoma was behind or in front of the equator line without reaching it, it was called “posterior” or “anterior,” respectively.
RESULTS Ninety-one percent of the 1,406 patients were referred by the Institut Curie ophthalmology department. There were 700 (49.8%) males, and 703 (50%) left eyes were involved. Sixty-one (4.4%) tumors were anterior, 588 (41.8%) were at the equator, and 757 (53.8%) were posterior. Extensive retinal detachment was presented in 371 (26.4%) eyes, but for 4 eyes (0.3%), retinal detachment status was unknown. Other patient characteristics, tumor characteristics, and dosimetric parameters are reported in Table 1. The median follow-up of
782
I. J. Radiation Oncology
● Biology ● Physics
Volume 65, Number 3, 2006
Table 1. Patient, tumor, and dosimetric characteristics (n ⫽ 1406)
Age Tumor diameter Tumor thickness Tumor volume Tumor–optic disc distance Tumor–macula distance Retina area receiving ⱖ30 CGE Eyeball volume receiving ⱖ30 CGE Lens volume receiving ⱖ30 CGE Ciliary body volume receiving ⱖ30 CGE Macular area receiving ⱖ30 CGE Optic disc area receiving ⱖ12 CGE Length of optic nerve receiving ⱖ12 CGE
Mean
Median
Range
61 years 13.3 mm 5.4 mm 0.6 cc 4.3 mm 3.8 mm 31% 2.9 cc 58 cc 41% 48% 45% 2.6 mm
58 years 13 mm 4.8 mm 0.4 cc 3.5 mm 2.4 mm 31% 2.8 cc 58.8 cc 37% 41% 33% 0.8 mm
15–90 years 2.5–24.4 mm 1.2–14 mm 0–3.6 cc 0–21.6 mm 0–23.3 mm 0–74% 0.2–6.6 cc 0–100% 0–100% 0–100% 0–100% 0–12.3 mm
Abbreviation: CGE ⫽ cobalt Gray equivalent.
surviving patients was 73 months (range, 24 –142 months). Nine hundred sixty-six patients (69.4%) were alive with no evidence of disease. Forty-six patients (3.3%) were alive with metastatic disease. Eleven patients (0.8%) were alive with a second cancer. Thirty-one patients (2.2%) were lost to follow-up (25 patients with no evidence of disease and 6 patients with metastatic disease at last follow-up). Local control The 5-year local control rate was 96% (range, 95.4 – 96.6%). Fifty-two local recurrences were diagnosed during follow-up: 36.5% during the first 18 months and 61.5% during the first 3 years after irradiation. The median interval between irradiation and local recurrence was 21 months. The univariate analysis results are reported in Table 2. Two independent unfavorable prognostic factors were identified on multivariate analysis: large tumor diameter (ⱖ13 mm) with a relative risk (RR) of 2.18 (95% CI, 1.21–3.91; p ⫽ 0.003) and small macular area (⬍41%) receiving ⱖ30 CGE with RR of 2 (95% CI, 1.12–3.57; p ⫽ 0.01). The therapy for local recurrence consisted of enucleation for 33 patients (63%), another course of proton beam radiotherapy for 3 patients, 125I plaque brachytherapy for 3 patients, and thermotherapy for 1 patient. Two patients did not receive local therapy because of concomitant metastatic disease. Twenty-six (50%) of the 52 patients with local recurrence developed metastases, and 30 (58%) died. Overall survival Three hundred forty-two patients (24.3%) died during follow-up. Causes of death were melanoma for 242 patients (70.8%), intercurrent disease for 37 patients (10.8%), second cancer for 16 patients (4.7%), and unknown for 47 patients (13.7%). The 5-year overall survival rate was 79% (range, 77.8 – 80.2%). The univariate analysis results are reported in Table 2. Six independent prognostic factors were identified on multivariate analysis: advanced age (ⱖ61 years) with an RR of 2.40 (95% CI, 1.91–3.01; p ⬍ 0.0001), male gender with
an RR of 1.44 (95% CI, 1.17–1.81; p ⫽ 0.0003), large tumor diameter (ⱖ13 mm) with an RR 1.55 (95% CI, 1.17–2.07; p ⬍ 0.0001), median and anterior tumor sites with an RR 1.63 (95% CI, 1.26 –2.08; p ⬍ 0.0001), large tumor thickness (ⱖ4.8 mm) with an RR 1.34 (95% CI, 1.03–1.75; p ⫽ 0.02), and large retinal area (ⱖ31%) receiving ⱖ30 CGE with an RR 1.40 (95% CI, 1.08 –1.8; p ⫽ 0.003). Metastasis-free survival The 5-year metastasis-free survival rate was 80.6% (range, 79.4 – 81.7%). The liver was the first metastatic site in 81.4% of cases. The univariate analysis results are reported in Table 2. Six independent prognostic factors were identified on multivariate analysis: older age (ⱖ61 years) with RR 1.42 (95% CI, 1.12–1.79; p ⫽ 0.004), retinal detachment with RR 1.38 (95% CI, 1.04 –1.73; p ⫽ 0.01), large tumor volume (ⱖ0.41 cc) with RR 1.74 (95% CI, 1.25–2.41; p ⬍ 0.0001), median and anterior tumor sites with RR 1.85 (95% CI, 1.4 –2.5; p ⬍ 0.0001), large retinal area (ⱖ31%) receiving ⱖ30 CGE with RR 1.38 (95% CI, 1.04 –1.83; p ⫽ 0.002), and local recurrence with RR 4.17 (95% CI, 2.71– 6.41; p ⬍ 0.0001). Ocular complications Five-year ocular complication rates are reported in Table 3 for the following complications: maculopathy, papillopathy, keratitis, vitreous hemorrhage, and intraocular inflammation. Some of these ocular complications were not associated with any visual symptoms and were exclusively diagnosed on systematic follow-up examinations. Enucleation for complications (EC) Enucleation performed for local recurrence was excluded from this analysis. Ninety-nine patients (7%) underwent enucleation for complications: neovascular glaucoma with severe pain in 83 patients, loss of vision in 5 patients, and for an unknown cause in 1 patient. Enucleation for complications was performed during the first 2 years after irradi-
Protontherapy of uveal melanoma
● R. DENDALE et al.
783
Table 2. Univariate analysis Factors Age ⬍61 years ⱖ61 years Gender F M Laterality Left Right Tumor location from equator Posterior At equator Anterior Tumor diameter ⬍13 mm ⱖ13 mm Median tumor thickness ⬍4.8 mm ⱖ4.8 mm Tumor volume ⬍0.41 cc ⱖ0.41 cc Retinal detachment No Yes Distance tumor-papilla ⬍3.5 mm ⱖ3.5 mm Distance tumor-macula ⬍2.4 mm ⱖ2.4 mm Surface-irradiated retina 30 CGE ⬍31% ⱖ31% Volume of irradiated eyeball 30 CGE ⬍2.8 cc ⱖ2.8 cc Volume of irradiated lens 30 CGE ⬍43% ⱖ43% Volume of irradiated ciliary body 30 CGE ⬍37% ⱖ37% Irradiated optic nerve length 12 CGE ⬍0.8 mm ⱖ0.8 mm Irradiated papilla surface 12 CGE ⬍33% ⱖ33% Irradiated macula surface 30 CGE ⬍41% ⱖ41%
Local recurrence RR (p) 1 1.04 ns
Metastatic disease RR (p) 1 1.38 (0.004)
Overall survival RR (p)
Secondary enucleation RR (p)
1 2.28 (0.0001)
1 1 ns
1 1.18 ns
1 1.06 ns
1 1.40 (0.0013)
1 1.21 ns
1 1.17 ns
1 1.17 ns
1 1.2 ns
1 1.12 ns
1 1.51 5.75 (⬍0.0001)
1 3 2.05 (⬍0.0001)
1 2.41 2.03 (⬍0.0001)
1 2.25 2.23 (0.0004)
1 2.30 (0.003)
1 2.81 (⬍0.0001)
1 2.60 (⬍0.0001)
1 2.43 (⬍0.0001)
1 1.91 (0.02)
1 2.83 (⬍0.0001)
1 2.24 (⬍0.0001)
1 6.27 (⬍0.0001)
1 1.94 (0.02)
1 3.11 (⬍0.0001)
1 1.63 (⬍0.0001)
1 5.02 (⬍0.0001)
1 2 (0.01)
1 2.27 (⬍0.0001)
1 1.77 (⬍0.0001)
1 2.57 (⬍0.0001)
1 1.15 ns
1 1.25 ns
1 1.2 ns
1 1.27 (0.038)
1 2.06 (0.01)
1 1.77 (⬍0.001)
1 1.40 (0.0015)
1 1.47 (0.05)
1 1.73 (⬍0.0001)
1 1.61 (⬍0.0001)
1 2.15 (⬍0.0001)
1 2.61 (⬍0.0001)
1 1.48 ns
1 1.91 (⬍0.0001)
1 1.76 (⬍0.0001)
1 3.73 (⬍0.0001)
1 2.21 (0.006)
1 1.82 (⬍0.0001)
1 1.82 (⬍0.0001)
1 4.46 (⬍0.0001)
1 2.22 (0.005)
1 2.15 (⬍0.0001)
1 2.11 (⬍0.0001)
1 4.91 (⬍0.0001)
1 1.25 ns
1 1.11 ns
1 1.09 ns
1 1.08 ns
1 1 ns
1 1.05 ns
1 1.04 ns
1 1.1 ns
1 0.81 ns
1 0.94 ns
1 0.54 (0.032)
1 0.69 (0.0013)
Abbreviations: RR ⫽ relative risk; ns ⫽ not significant; CGE ⫽ cobalt Gray equivalent.
ation in 54% of cases and during the first 3 years in 71% of cases. Five-year EC rates were 7.7% (range, 6.7– 8.7%). The univariate analysis results are reported in Table 2. Two independent unfavorable prognostic factors were identified on multivariate analysis: large tumor thickness (ⱖ4.8 mm) with RR 5.42 (95% CI, 2.87– 40.23; p ⬍ 0.0001) and large
lens volume (ⱖ43%) receiving ⱖ30 CGE with RR 2.47 (95% CI, 1.48 – 4.12; p ⫽ 0.0002). Visual acuity Baseline visual acuity and visual acuity at 5 years of follow-up are reported in Table 4. At 5 years, visual acuity
784
I. J. Radiation Oncology
● Biology ● Physics
Table 3. Actuarial ocular complication rates
Maculopathy Papillopathy Glaucoma Cataract Keratitis Vitreous hemorrhage Intraocular inflammation
5-year (%)
Confidence interval
66.5 23.4 28.6 61.8 11.5 13.9 27.5
(63.3–69.7) (20.5–26.3) (26–31.2) (59–64.7) (9.6–13.4) (11.9–16) (24.9–30.1)
remained stable for 38% of patients, decreased for 56% of patients, and improved for 6% of patients. DISCUSSION This series comprises 1,406 patients with a median follow-up of 73 months and a minimum follow-up of 2 years. This is a very homogeneous series in terms of both patient and tumor characteristics and treatment modalities; clip surgery, treatment planning, and treatment delivery were performed according to predefined procedures. These procedures have remained unchanged since 1991, apart from successive versions of EYEPLAN software. This series was comprised of only a few small anterior tumors as a result of the initial patient selection, because small anterior tumors are usually treated by 125I plaque brachytherapy, whereas small posterior tumors are usually treated by proton beam radiotherapy. Surgical access to the posterior segment for placement of episcleral plaques is fairly difficult, and coverage of the tumor base by a plaque may be difficult or incomplete for tumors invading the optic disc, because of the emergence of the optic nerve. Distribution of the intraocular dose delivered by brachytherapy is also much more heterogeneous than that delivered by proton beam radiotherapy. This dose heterogeneity is even more marked for larger tumor volumes. The maximum intraocular dose can exceed 200 Gy for some applications, whereas it is always equal to the prescribed dose of 60 CGE for proton beam radiotherapy. Some published studies have compared the results obtained by plaque brachytherapy and heavy particle (protons or Helium) beam irradiation for tumors with similar characteristics and seem to confirm a higher ocular toxicity after brachytherapy (27, 28). Proton beam radiotherapy is therefore probably a more appropriate treatment option for large posterior tumors, very close to the optic nerve, whereas small anterior tumors can be effectively treated by plaque brachytherapy. On the basis of these selection criteria (small anterior melanomas treated by plaque brachytherapy and other melanomas treated by proton beam radiotherapy), we have not found any difference in terms of toxicity or local control between these two treatment modalities (29). Local control The 5-year local control rate of 96% observed in this series was very similar to the rates reported by others (21,
Volume 65, Number 3, 2006
27, 30 –33), ranging from 90.5% to 96%. The median time to local recurrence was 21 months, and two-thirds of local recurrences were during the first 3 years after irradiation. Munzenrider et al. (31) reported a median interval of 14 months between irradiation and local recurrence. Like others (27), we found that tumor diameter was an independent prognostic factor for local recurrence. The second independent prognostic factor identified by multivariate analysis was a greater macula area (ⱖ41%) receiving ⱖ30 CGE. This means that a better local control rate was observed for eyes receiving a greater macular dose compared to eyes receiving a lower dose. This suggests better local control for tumors situated close to the macula. Three possible explanations that are probably intimately related can be proposed. First, this result could reflect the recruitment bias of our series, which included only a small number of small anterior tumors, treated mostly by plaque brachytherapy. The majority of anterior tumors treated by proton therapy were therefore larger tumors with a higher risk of recurrence. The second explanation could be a more precise definition of tumor margins for tumors situated close to the macula. Tumor margins were determined by two different techniques: The posterior limit of the melanoma was determined by ocular fundus photographs or angiography, allowing precise definition of its relation to the macula, whereas the limits of more anterior tumors are determined by clips sutured onto the sclera over the tumor margins as determined by transillumination. However, the anterior limit of an achromic tumor can be difficult to determine by transillumination. The third explanation is related to the earlier diagnosis of these posterior tumors, which would therefore be associated with less necrosis and possibly more radiosensitivity. Tumors situated close to the macula are more rapidly symptomatic, leading to earlier ophthalmologic examination and diagnosis, whereas a more anterior tumor can remain asymptomatic for a long time. Overall survival and metastasis-free survival The 5-year overall survival rate in our study was 79%, corresponding to the published rates (14, 30, 32, 34), ranging from 70.3% to 78%. The 5-year metastasis-free survival rate was 80.6% and was comparable to the rates reported in other published series, ranging from 74.3% to 91%. This indicated that in 20% of our patients, local treatment alone Table 4. Comparison of VA between initial examination and 5-year follow-up: percentage of patients VA at 5 years Baseline VA
⬍20/200
20/200 to ⬍20/40
ⱖ20/40
Total
⬍20/200 20/200 to ⬍20/40 ⱖ20/40 Total
9.1% 26.4% 18.6% 54.1%
1% 9.6% 10.9% 21.5%
0.6% 4.2% 19.6% 24.4%
10.5% 40.3% 49.2% 100%
Abbreviation: VA ⫽ visual acuity.
Protontherapy of uveal melanoma
was insufficient. As reported by other authors, patients who developed local recurrence had a poorer prognosis than those who did not, because 50% with local recurrence developed metastatic disease. Our multivariate analysis demonstrated local recurrence to be an independent unfavorable prognostic factor for metastasis-free survival, but it is unclear whether metastases are directly related to local recurrence or whether they reflect a particularly aggressive phenotype of the primary tumor. The prognostic value of a local recurrence event on metastasis incidence has already been described by others (34, 35). Prescher et al. (36) showed that monosomy 3 had a negative prognostic impact on the metastasis incidence and survival rate. Further studies are necessary to identify such tumors, but they would require performing systematic biopsy in the primary tumors. Eye complications and enucleation Proton beam radiotherapy can induce ocular complications, particularly glaucoma (31, 37). According to Char et al. (32), 34.9% of treated eyes developed neovascular glaucoma. In our series, the enucleation rate related to ocular complications was 7%, corresponding to the published EC rates (14, 30 –34, 37, 38), which range from 2.2% to 18% and are related to neovascular glaucoma in 71% to 100% of cases. In our series, 64% of these enucleations were performed during the first 3 years. Egan et al. (34) and Fuss et al. (30) reported EC rates during the first 3 years of 64% and 100%, respectively. Wilson and Hungerford (38) reported a mean interval between irradiation and EC of 28 months. Some authors have reported retinal detachment (37) and tumor– optic disc distance (14, 34) to be independent prognostic factors for EC. In our series, only tumor thickness and lens volume irradiated were found to be significant independent prognostic factors on multivariate analysis. The risk of secondary enucleation is therefore related to a tumor characteristic (tumor thickness) and a dosimetric parameter (dose to lens). Lens protection was not checking in our dosimetric planning. However, we always tried to reduce the dose delivered to the corneal limbus, but dosimetric planning for large tumors, anterior tumors, and tumors invading the ciliary body often required eye positions leading to irradiation of the anterior chamber and therefore the lens. Complications leading to enucleation could be related to irradiation of the anterior chamber, despite the avoidance of the lens itself. There was a high cataract rate in our series, but cataract surgery did not seem to have any impact on the rate of glaucoma and EC (data not shown). Unfortunately, data corresponding to the dose received by the anterior chamber were not recorded in this study. Tumor thickness is a reflection of tumor size, which determines the ocular volume irradiated. The tumor thickness and the size of the collimator are indeed closely related. With a large collimator, it becomes difficult to avoid the irradiation of the anterior chamber, and the irradiated ocular volume is always large. Various methods of reducing treatment toxicity are being investigated. They include total dose reduction, reduction of
● R. DENDALE et al.
785
safety margins, increased fractionation, and concomitant treatment with radioprotectors, anticoagulants, or thermotherapy.
Total dose reduction The initial protocols for ocular proton beam radiotherapy were established with a small number of high-dose fractions, because the clinical data available at that time indicated favorable results for this type of irradiation on skin melanomas (12, 39). Plaque brachytherapy, the first radiotherapy modality for uveal melanoma, used empiric doses between 70 and 140 Gy. These high doses were possible, because no substantial normal-tissue reaction was observed (12). Proton scatter in the tissues is much less marked than during conventional external beam radiotherapy (photons or electrons). The sudden dose drop posterior to the Bragg peak and the use of collimators manufactured according to the tumor profile allow the use of very small safety margins around the tumor, resulting in a much smaller intraocular volume irradiated. The equivalent dose delivered by current proton beam radiotherapy regimens is much higher than that delivered by conventional radiotherapy with 2 Gy fractions. Bentzen et al. (40) reported that the estimated dose to control 80% of tumors measuring 10 mm and 20 mm in diameter was 22 Gy and 28 Gy, respectively, delivered in 9 fractions. The dose of 60 CGE delivered in our series corresponds to 2 to 2.5 times these values. Some authors (41, 42) have recently challenged the concept of the radioresistance of melanoma. In a dose reduction study, Gragoudas et al. (43) reported the 5-year results of a randomized trial on 188 patients comparing doses of 50 CGE and 70 CGE delivered in 5 fractions over 5 days and found no significant difference between the two dose regimens in terms of ocular toxicity or local control.
Reduction of the tumor safety margin A reduction of the safety margin around the tumor would decrease the ocular irradiation volume. However, based on a series of patients with a follow-up of 15 years, Egger et al. (21) reported that localized reduction of the tumor safety margins to avoid irradiation of the optic nerve and optic disc induced a decrease of the 5-year local control rate from 96.3% to 71.3%. Other teams (44) have reported that reduction of tumor safety margins is possible in particular cases by using 3D CT reconstruction and by introducing anatomic modifications in their EYEPLAN dosimetric planning to obtain a better representation of the region of interest concerned by the tumor safety margin reduction. We recently reported (45) the preliminary results of a study combining very localized reduction of the safety margin close to the macula and localized transpupillary thermotherapy and obtained good preservation of visual acuity with no reduction of the local control rate.
786
I. J. Radiation Oncology
● Biology ● Physics
Increased fractionation Late complications are directly correlated with the dose per fraction. During proton beam radiotherapy, 14 to 15 CGE are delivered in each fraction. An increased number of fractions (4 to 8 fractions) with the same total dose of 60 CGE could possibly decrease the ocular complication rate without compromising local control. However, doubling of the number of fractions compared to the current regimen would raise a number of practical and cost problems. This type of fractionation study would have to be conducted very cautiously with strict tumor selection and rigorous discontinuation criteria. Concomitant treatments WR2721 (amifostine) or WR-77913 (aminophosphorothioate) have a known radioprotective effect (46 – 48). Several teams (46, 47, 49, 50) have tested these molecules in patients treated with irradiation for various tumor locations. They have reported an encouraging radioprotective effect on normal tissues. These products must be administered by i.v. injection before each treatment session. Hypofractionated proton beam radiotherapy seems to be particularly suitable for radioprotective therapy. Anticoagulants or platelet aggregation inhibitors can decrease tissue hypoxia by reducing the number of micro-
Volume 65, Number 3, 2006
thrombi (51) and could therefore reduce the toxic effect of irradiation. In fact, tumor and vascular necrosis, vascular thrombosis, and intraocular inflammation after proton beam radiotherapy can be responsible for severe hypoxia of healthy intraocular tissues (52). This hypoxia predisposes to angiogenesis, which can be responsible for the appearance of neovascular glaucoma. A combination of transpupillary thermotherapy and protons could reduce post proton therapy ocular toxicity. Char et al. (53) reported an exudative retinal detachment reduction by thermotherapy after a proton radiation. CONCLUSION We have reported the results of a homogeneous series of 1,406 patients treated for uveal melanoma by proton beam radiotherapy in a single center using the same treatment procedure. The local control rate was good and similar to previously published rates, but ocular complication rates and the secondary enucleation rate without complication during the first year remain significant. The most frequent ocular toxicity was neovascular glaucoma, but enucleation was avoided in the majority of these patients. Future research should try to reduce the ocular complication rate while maintaining the current good local control rates.
REFERENCES 1. Vidal JL, Bacin F, Albuisson E, et al. [“Melanoma 92”. Epidemiological study of uveal melanoma in France]. J Fr Ophtalmol 1995;18:520 –528. 2. Scotto J, Fraumeni JFJ, Lee JA. Melanomas of the eye and other noncutaneous sites: epidemiologic aspects. J Natl Cancer Inst 1976;56:489 – 491. 3. Stallard HB. Partial choroidectomy. Br J Ophthalmol 1966; 50:660 – 662. 4. Damato B. Adjunctive plaque radiotherapy after local resection of uveal melanoma. Front Radiat Ther Oncol 1997;30: 123–132. 5. Foulds WS. The local excision of choroidal melanomata. Trans Ophthalmol Soc U K 1973;93:343–346. 6. Kivela T, Puusaari I, Damato B. Transscleral resection versus iodine brachytherapy for choroidal malignant melanomas 6 millimeters or more in thickness: A matched case-control study. Ophthalmology 2003;110:2235–2244. 7. Lommatzsch PK. Results after beta-irradiation (106Ru/106Rh) of choroidal melanomas. Twenty years’ experience. Am J Clin Oncol 1987;10:146 –151. 8. Stallard HB. Radiotherapy for malignant melanoma of the choroid. Br J Ophthalmol 1966;50:147–155. 9. Dieckmann K, Georg D, Zehetmayer M, et al. LINAC based stereotactic radiotherapy of uveal melanoma: 4 years clinical experience. Radiother Oncol 2003;67:199 –206. 10. Char DH, Castro JR, Quivey JM, et al. Helium ion charged particle therapy for choroidal melanoma. Ophthalmology 1980;87:565–570. 11. Lumbroso L, Levy C, Plancher C, et al. [Results of proton beam irradiation for treatment of choroidal melanoma]. J Fr Ophtalmol 2002;25:290 –297. 12. Gradoudas ES, Goitein M, Koehler A, et al. Proton irradiation of choroidal melanomas. Preliminary results. Arch Ophthalmol 1978;96:1583–1591.
13. Gragoudas ES, Goitein M, Koehler AM, et al. Proton irradiation of small choroidal malignant melanomas. Am J Ophthalmol 1977;83:665– 673. 14. Egger E, Zografos L, Munkel G, et al. Results of proton radiotherapy for uveal melanomas. Front Radiat Ther Oncol 1997;30:111–122. 15. Brovkina AF, Zarubei GD. Ciliochoroidal melanomas treated with a narrow medical proton beam. Arch Ophthalmol 1986; 104:402– 404. 16. Seddon JM, Gragoudas ES, Albert DM, et al. Comparison of survival rates for patients with uveal melanoma after treatment with proton beam irradiation or enucleation. Am J Ophthalmol 1985;99:282–290. 17. Rosenwald JC, Bridier A, Sabattier R, et al. [Preliminary dosimetry study of the Orsay synchrocyclotron proton beam with a view to applying it therapeutically]. Ophtalmologie 1990;4:123–127. 18. Lumbroso L, Levy C, Plancher C, et al. [Results of proton beam irradiation for treatment of choroidal melanoma]. J Fr Ophtalmol 2002;25:290 –297. 19. Desjardins L, Levy C, d’Hermies F, et al. [Initial results of proton therapy in choroidal melanoma at the d’Orsay center for Proton Therapy; The first 464 cases]. Cancer Radiother 1997;1:222–226. 20. Schlienger P, Habrand JL, Schwartz L, et al. Initial results with one-year minimum follow-up of the first 146 patients with a uveal melanoma treated with protons at CPO (Orsay). Bull Cancer Radiother 1996;83:212s–214s. 21. Egger E, Schalenbourg A, Zografos L, et al. Maximizing local tumor control and survival after proton beam radiotherapy of uveal melanoma. Int J Radiat Oncol Biol Phys 2001;51:138 –147. 22. Goitein M, Miller T. Planning proton therapy of the eye. Med Phys 1983;10:275–283.
Protontherapy of uveal melanoma
23. Perret C. The therapy planning program EYE. Swiss Institute for Nuclear Research, Internal Report; 1987. 24. Sheen M. EYEPLAN proton therapy planning program. Bebington, Wirral, UK: Claterbridge Centre for Oncology, Internal Report; 1995. 25. Kaplan E, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1971;53:457– 481. 26. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 1966;50:163–170. 27. Wilson MW, Hungerford JL. Comparison of episcleral plaque and proton beam radiation therapy for the treatment of choroidal melanoma. Ophthalmology 1999;106:1579 –1587. 28. Char DH, Quivey JM, Castro JR, et al. Helium ions versus iodine 125 brachytherapy in the management of uveal melanoma. A prospective, randomized, dynamically balanced trial. Ophthalmology 1993;100:1547–1554. 29. Desjardins L, Lumbroso L, Levy C, et al. [Treatment of uveal melanoma with iodine 125 plaques or proton beam therapy: Indications and comparison of local recurrence rates]. J Fr Ophtalmol 2003;26:269 –276. 30. Fuss M, Loredo LN, Blacharski PA, et al. Proton radiation therapy for medium and large choroidal melanoma: Preservation of the eye and its functionality. Int J Radiat Oncol Biol Phys 2001;49:1053–1059. 31. Munzenrider JE, Verhey LJ, Gragoudas ES, et al. Conservative treatment of uveal melanoma: Local recurrence after proton beam therapy. Int J Radiat Oncol Biol Phys 1989;17:493– 498. 32. Char DH, Kroll SM, Castro J. Ten-year follow-up of helium ion therapy for uveal melanoma. Am J Ophthalmol 1998;125: 81– 89. 33. Courdi A, Caujolle JP, Grange JD, et al. Results of proton therapy of uveal melanomas treated in Nice. Int J Radiat Oncol Biol Phys 1999;45:5–11. 34. Egan KM, Ryan LM, Gragoudas ES. Survival implications of enucleation after definitive radiotherapy for choroidal melanoma: An example of regression on time-dependent. Arch Ophthalmol 1998;116:366 –370. 35. Harbour JW, Char DH, Kroll S, et al. Metastatic risk for distinct patterns of postirradiation local recurrence of posterior uveal melanoma. Ophthalmology 1997;104:1785–1792; discussion 1792–1783. 36. Prescher G, Bornfeld N, Hirche H, et al. Prognostic implications of monosomy 3 in uveal melanoma. Lancet 1996;347: 1222–1225. 37. Foss AJ, Whelehan I, Hungerford JL, et al. Predictive factors for the development of rubeosis following proton beam radiotherapy for uveal melanoma. Br J Ophthalmol 1997;81:748 –754. 38. Wilson MW, Hungerford JL. Comparison of episcleral plaque and proton beam radiation therapy for the treatment of choroidal melanoma. Ophthalmology 1999;106:1579 –1587.
● R. DENDALE et al.
787
39. Habermalz HJ, Fischer JJ. Radiation therapy of malignant melanoma: Experience with high individual treatment doses. Cancer 1976;38:2258 –2262. 40. Bentzen SM, Overgaard J, Thames HD, et al. Clinical radiobiology of malignant melanoma. Radiother Oncol 1989;16: 169 –182. 41. Fowler JF. Radioresponsiveness of malignant melanoma to fractionated irradiation. Int J Radiat Oncol Biol Phys 1984; 10:439 – 440. 42. Noel G, Simon JM, Valery CA, et al. Linac radiosurgery for brain metastasis of melanoma. Stereotact Funct Neurosurg 2002;79:245–255. 43. Gragoudas ES, Lane AM, Regan S, et al. A randomized controlled trial of varying radiation doses in the treatment of choroidal melanoma. Arch Ophthalmol 2000;118:773–778. 44. Chauvel P, Iborra N. Uveal melanoma relapses: The controversy about safety margins. PTCOG. Paris; 2004. 45. Dendale R, Lumbroso-Le Rouic R, Delacroix S, et al. Transpupillary thermotherapy and protontherapy association for posterior choroidal melanoma [Abstract]. PTCOG. Paris; 2004. 46. Bourhis J, De Crevoisier R, Abdulkarim B, et al. A randomized study of very accelerated radiotherapy with and without amifostine in head and neck squamous cell carcinoma. Int J Radiat Oncol Biol Phys 2000;46:1105–1108. 47. Antonadou D, Throuvalas N, Petridis A, et al. Effect of amifostine on toxicities associated with radiochemotherapy in patients with locally advanced non-small-cell lung cancer. Int J Radiat Oncol Biol Phys 2003;57:402– 408. 48. Belkacemi Y, Rat P, Piel G, et al. Lens epithelial cell protection by aminothiol WR-1065 and anetholedithiolethione from ionizing radiation. Int J Cancer 2001;96:15–26. 49. Athanassiou H, Antonadou D, Coliarakis N, et al. Protective effect of amifostine during fractionated radiotherapy in patients with pelvic carcinomas: Results of a randomized trial. Int J Radiat Oncol Biol Phys 2003;56:1154 –1160. 50. Reddy VN, Ikebe H, Giblin FJ, et al. Effect of radioprotective agents on X-ray cataracts. Lens Eye Toxic Res 1989;6:573– 588. 51. Salomon O, Rosenberg N, Steinberg DM, et al. Nonarteritic anterior ischemic optic neuropathy is associated with a specific platelet polymorphism located on the glycoprotein Ibalpha gene. Ophthalmology 2004;111:184 –188. 52. Lumbroso L, Desjardins L, Levy C, et al. Intraocular inflammation after proton beam irradiation for uveal melanoma. Br J Ophthalmol 2001;85:1305–1308. 53. Char DH, Bove R, Phillips TL. Laser and proton radiation to reduce uveal melanoma-associated exudative retinal detachments. Trans Am Ophthalmol Soc 2003;101:53–56.