Management and Outcome of Retinoblastoma with Vitreous Seeds

Management and Outcome of Retinoblastoma with Vitreous Seeds Fairooz P. Manjandavida, MD,1 Santosh G. Honavar, MD, FACS,1,2 Vijay Anand P. Reddy, MD,1,2 Rohit Khanna, MD, MPH1 Purpose: To report the treatment response of retinoblastoma with vitreous seeds to high-dose chemotherapy coupled with periocular carboplatin. Design: Retrospective, interventional case series. Participants: Consecutive patients with retinoblastoma with vitreous seeds managed over 10 years at a comprehensive ocular oncology center and followed up for at least 12 months after the completion of treatment were included in this study. Institutional review board approval was obtained. Intervention: High-dose chemotherapy with a combination of vincristine, etoposide, and carboplatin in patients with focal vitreous seeds and additional concurrent periocular carboplatin in patients with diffuse vitreous seeds. Main Outcome Measures: Tumor regression, vitreous seed regression, and eye salvage. Results: After excluding the better eye of bilateral cases, 101 eyes of 101 patients were part of the final analysis. All the patients belonged to Reese-Ellsworth group VB, but on the International Classification of Retinoblastoma (ICRB), 21 were group C, 40 were group D, and 40 were group E. The mean basal diameter of the largest tumor was 11.8
4.7 mm. Mean tumor thickness was 7.5
4.0 mm. Vitreous seeds were focal in 21 eyes and diffuse in 80 eyes. Chemotherapy cycles ranged from 6 to 12 (median, 6). Seventy-three eyes with diffuse vitreous seeds received a 15-mg posterior sub-Tenon carboplatin injection (range, 1e13 mg; median, 6 mg). Follow-up duration ranged from 13.4 to 129.2 months (median, 48 months). External beam radiotherapy (EBRT) was necessary in 33 eyes with residual tumor, vitreous seeds, or both. In all, 20 eyes (95%) with ICRB group C retinoblastoma, 34 eyes (85%) with group D retinoblastoma, and 23 eyes (57.5%) with group E retinoblastoma were salvaged. Of 77 eyes that were salvaged, 74 (96%) had visual acuity of 20/200 or better. Twenty-four of 33 chemotherapy failures (73%) regressed with EBRT. None of the patients demonstrated second malignant neoplasm or systemic metastasis. Factors predicting tumor regression and eye salvage were bilateral retinoblastoma and absence of subretinal fluid. Factors predicting vitreous seed regression were absence of subretinal fluid and subretinal seeds. Conclusions: Intensive management with primary high-dose chemotherapy and concurrent periocular carboplatin, and EBRT selectively in chemotherapy failures, provides gratifying outcome in retinoblastoma with vitreous seeds. Ophthalmology 2014;121:517-524 ª 2014 by the American Academy of Ophthalmology.

The management of retinoblastoma has undergone a paradigm change in the recent past. Although the focus continues to be on improved survival and safety of treatment methods, the emphasis is now on eye salvage and optimization of residual vision. A substantial reduction in the frequency of enucleation has been achieved in the last few decades, both for unilateral and bilateral retinoblastoma.1e4 Similarly, the proportion of patients treated with external beam radiotherapy (EBRT) has decreased sharply from 35% in the late 1980s to approximately 7% now.5 Concurrently, there has been an increasing interest in the use of alternative eye- and vision-conserving methods of treatment.2,6e39 Chemotherapy is the current standard of care for retinoblastoma. Chemotherapy can be delivered systemically, regionally, or locally.2,6e14,16e36 Standard triple-drug systemic chemotherapy (vincristineþetoposideþcarboplatin) is most effective in tumors without associated subretinal fluid, subretinal seeds, or vitreous seeds.2,6e14,35 Chemotherapy  2014 by the American Academy of Ophthalmology Published by Elsevier Inc.

provides satisfactory tumor control for Reese-Ellsworth groups I through IV, with treatment failure necessitating additional external beam radiotherapy in only 10% and enucleation in 15%.11 However, 47% of patients with Reese-Ellsworth group V tumors needed EBRT and 53% required enucleation.11 In another study, eye salvage was achieved in 47% of International Classification of Retinoblastoma (ICRB) group D eyes overall, and in only 30% in eyes with diffuse vitreous seeds.35 Cohen et al13 could salvage only 2 of 18 eyes with Reese-Ellsworth group V retinoblastoma with standard chemotherapy alone. Kingston et al6 could salvage 8 of 12 eyes (67%) with vitreous seeds with chemotherapy coupled with EBRT. Overall, 30% to 77% of eyes with vitreous seeds that underwent primary chemotherapy ultimately needed enucleation (Table 1). More recently, the role of superselective chemotherapy is being explored in the management of advanced retinoblastoma.16e21 There is a 64% probability of eye salvage ISSN 0161-6420/14/$ - see front matter http://dx.doi.org/10.1016/j.ophtha.2013.09.011

517

Ophthalmology Volume 121, Number 2, February 2014 Table 1. Outcome of Retinoblastoma with Vitreous Seeds: A Review of the Literature

Authors

Year

No.

7

Murphree et al Gallie et al9 Kingston et al6 Shields et al10

1996 1996 1996 1997

18 22 12 24

Shields et al2

2002

11

2002

54

2009 2011 2011 2012

5 2 9 31

2012 2012 2012 Current study

2 23 12 101

Shields et al

12

Chan et al14 Mallipatna et al31 Shields et al18 Abramson et al20 Kivelä et al24 Munier et al25 Ghassemi and Shields26 Manjandavida et al

Type of Vitreous Seeds Diffuse (n ¼ 18) NA NA Focal (n ¼ 7), diffuse (n ¼ 17) Focal (n ¼ 6), diffuse (n ¼ 5) Focal (n ¼ 32), diffuse (n ¼ 22) NA Diffuse Diffuse Focal (n ¼ 14), diffuse (n ¼ 14), NA (n ¼ 3) NA Diffuse Diffuse Focal (n ¼ 21), diffuse (n ¼ 80)

Management

Use of External Beam Radiotherapy (n ¼ 4) (n ¼ 3) (n ¼ 12) (NA)

Enucleation, n (%) 14 77) 11 (50) 4 (33) 8 (33)

Eye Salvage, n (%) 4 11 8 16

(22) (50) (66) (67)

Vision Salvage, n (%)

CRD CRD CRD CRD

Yes Yes Yes Yes

NA 8 (36) NA NA

CRD

Yes (n ¼ 4)

7 (64)

4 (36)

NA

CRD

Yes (n ¼ 4)

16 (30)

38 (70)

NA

CRD CRD, PO topotecan CRD, IAC CRD, IAC

Yes (NA) No Yes (n ¼ 2) Yes (NA)

CRD, IV methotrexate CRD, IV melphalan CRD, IV melphalan HD CRD, PO carboplatin

No No No Yes

2 2 1 8

(40) (100) (12) (25)

3 0 8 23

(60) (0) (88) (75)

NA NA NA NA

1 3 6 24

(50) (17) (50) (24)

1 20 6 77

(50) (83) (50) (77)

NA NA NA 74 (74)

CRD ¼ chemoreduction; HD ¼ high dose; IAC ¼ intra-arterial chemotherapy; IV ¼ intravitreal; NA ¼ not applicable; PO ¼ periocular.

in eyes with vitreous seeds.20 In another study, 67% of eyes in the subgroup of patients with vitreous seeds underwent enucleation after intra-arterial chemotherapy.17,18 Local chemotherapy, intraocular or periocular, provides high intravitreal drug concentration and helps manage vitreous seeds.22e33 However, intravitreal chemotherapy has not yet gained wide acceptance because of the apprehension of possible extraocular dissemination and pending standardization of dose and frequency of administration.22e27 Periocular chemotherapy also results in high intravitreal drug level, and periocular carboplatin injection has been found to be an option in the management of retinoblastoma with vitreous seeds.28e33 Retinoblastoma with vitreous seeds accounts for approximately 30% (280 of 957 eyes) of all patients with retinoblastoma in our clinic,40 and the situation is much worse in low-income countries.37 Clearly, there is a need to improve eye salvage in patients with retinoblastoma with vitreous seeds using cost-effective treatment protocols but without compromising patient safety and survival. Our study involves a large series of such patients and focuses on the role of primary high-dose chemotherapy and periocular carboplatin.

Methods Our retrospective, noncomparative, interventional case series included 137 eyes of 101 consecutive patients with retinoblastoma with vitreous seeds managed with high-dose chemotherapy with or without periocular carboplatin from July 2000 through December 2010 and who had completed at least 12 months of follow-up after the last treatment. After excluding the better eye of patients with bilateral vitreous seeds, 101 eyes of 101 patients were analyzed. Our objective was to study tumor regression, vitreous seed

518

regression, and eye salvage after high-dose chemotherapy alone in eyes with focal vitreous seeds and high-dose chemotherapy and periocular carboplatin in eyes with diffuse vitreous seeds. The study setting was a multispecialty retinoblastoma treatment facility at a tertiary care eye hospital in southern India. Institutional review board approval was obtained. The data were collected prospectively and were analyzed retrospectively. The demographic information, clinical findings, treatments, and visual outcomes were tabulated. Demographic data recorded at the initial visit included patient age at diagnosis (in months), gender (male or female), hereditary pattern (familial or sporadic), laterality (unilateral or bilateral), and the involved eye (right, left, or both). The total number of tumors per eye and grouping by the Reese-Ellsworth classification and ICRB35 were recorded. Each tumor was evaluated for the greatest basal dimension (in millimeters) and the proximity of the nearest tumor margin to the optic disc and foveola using indirect ophthalmoscopy. Tumor height (in millimeters) was assessed by B-scan ultrasonography and, where that was not possible, was estimated clinically. The presence and extent of subretinal fluid (localized, diffuse, and clock hours) and of tumor seeding in the vitreous (focal, diffuse) and subretinal space (focal, diffuse) were assessed. Vitreous seeds and subretinal seeds were defined as focal if located 3 mm or less from the tumor and as diffuse if located more than 3 mm from the tumor. The chemotherapy protocol included a combination of highdose intravenous vincristine, etoposide, and carboplatin (Table 2) in all the patients with vitreous seeds. Every child received a minimum of 6 cycles of high-dose chemotherapy. Chemotherapy was extended up to a maximum of 12 cycles if there were residual tumors or vitreous seeds beyond 6 cycles, vision salvage was possible (determined based on the tumor location), and enucleation (only eye with visual potential) or EBRT (heritable retinoblastoma with age 12 months) were not feasible. Periocular carboplatin was administered specifically in patients with diffuse vitreous seeds that remained diffuse and viable as judged clinically after the initial 2 cycles of high-dose chemotherapy. It was

Manjandavida et al



Retinoblastoma with Vitreous Seeds

Table 2. High-Dose Chemotherapy: Drugs, Dose, and Schedule Drugs

Day 1

Day 2

Vincristine Etoposide Carboplatin

0.025 mg/kg body weight 12.0 mg/kg body weight 28.0 mg/kg body weight

12.0 mg/kg body weight

continued until complete regression of the vitreous seeds or until a decision was taken in favor of enucleation or EBRT. The patients were followed up with examination under anesthesia every 3 weeks during chemotherapy and every 4 to 6 weeks thereafter until complete tumor regression. Subsequently, follow-up evaluation was performed every 3 to 6 months as needed. All focal treatment methods (cryotherapy, thermotherapy, or plaque brachytherapy) used were recorded. Episodes of recurrence and treatment for the same were documented. Details of EBRT (indication, total dose, fractions, outcome, complications), if given, were noted. At the last visit, visual acuity, tumor regression patterns, and local and systemic complications were assessed. The outcome in every case was assessed by an experienced ocular oncologist (S.G.H.). The clinical data were analyzed with regard to the main outcome measures: tumor regression, vitreous seed regression, and eye salvage. The data also were analyzed with tumor recurrence, vitreous seed recurrence, need for EBRT, and vision salvage as secondary outcome measures. The effect of each individual demographic factor, clinical variable, and treatment method on the final outcome was analyzed. All the variables were analyzed as discrete categories. Patient age, intraocular pressure, tumor base, tumor thickness, proximity to the optic disc, and proximity to the foveola were grouped into discrete categories. Data analysis was performed using the Stata 11 statistical package (StataCorp, College Station, TX). For analysis of risk factors for predicting the outcome measures, a univariate and multivariate Cox proportional hazard model was used. The variables that were significant (P < 0.05) on a univariate test and also those considered to be clinically significant were entered into the multivariate logistic regression analysis model. For variables that showed a high degree of correlation, only 1 variable from the set of associated variables was entered. A final multivariate model fitted variables identified as significant predictors (P < 0.05) of the outcome measure in question. Multiple colinearity between variables was assessed using a variance inflation factor, and the proportionality of the model also was tested.

Results Of 1067 patients with retinoblastoma managed at the Oncology Service between July 2000 and December 2010, 101 eyes (right eye:left eye, 51:50) of 101 consecutive patients who had vitreous seeds, who had been managed with high-dose chemotherapy, and who were available for follow-up for at least 12 months after the completion of treatment were included in this study. Fifteen patients, 3 in ICRB group C, 7 in ICRB group D, and 5 in ICRB group E, who had been managed with a similar protocol but who had follow-up of less than 12 months (range, 2e10 months; median, 7 months) after completion of treatment were excluded from the study. Twelve of these 15 patients had shown complete regression of the tumor and vitreous seeds at the last available follow-up. One patient each had residual tumor, residual vitreous seeds, or both. The last advice before loss to follow-up was EBRT for the 2 patients with residual tumor and vitreous seeds and enucleation for the patient with both residual tumor and vitreous seeds.

The median age of the patients was 25.2 months (mean, 34.5
31.8 months; range, 3.3e192.8 months). There were 60 male and 41 female patients. Fifteen patients had familial retinoblastoma. There were 18 patients with unilateral and 83 with bilateral retinoblastoma. Of 83 patients with bilateral retinoblastoma, 36 had vitreous seeds in both the eyes at presentation, and the better eye (based on clinical judgment of the size of the tumor and extent of vitreous seeds) was excluded from the analysis. Of the remaining 47 bilateral cases, 26 had undergone enucleation of 1 eye, and 21 were ICRB group A (n ¼ 16) or group B (n ¼ 5) tumors. Twentynine patients had received prior treatment elsewhere for the same eye that was part of this study. Prior treatment included cryotherapy (16 eyes), laser photocoagulation (7 eyes), and 1 to 4 (median, 2) cycles of chemotherapy (6 patients) with various drug combinations. Although 70 patients had white reflex as the major symptom, 17 had reduced vision, and 14 had squint. The median duration of symptoms was 2 months (mean, 3.6
4.9 months; range, 0.0e24.0 months). Initial visual acuity as estimated by age-appropriate means by a specialist pediatric optometrist was worse than 20/200 in 87 eyes and 20/200 or better in 14 eyes. All the patients belonged to Reese-Ellsworth group VB, but according to the ICRB, 21 had group C disease, 40 had group D disease, and 40 had group E disease. Eighty-one eyes had an endophytic tumor: 1 was exophytic, 6 were diffuse infiltrative, and 13 had mixed configuration. A single tumor was found in 35 eyes, whereas 66 eyes had multiple tumors, ranging from 2 to 14 (median, 6). The maximum basal diameter of the largest tumor ranged from 4 to 20 mm (mean, 11.8
4.7 mm; median, 12 mm). The basal diameter was 16 mm or less in 45 eyes, whereas it was more than 16 mm in 56 eyes. Maximum tumor thickness ranged from 2 to 20 mm (mean, 7.5
4.0 mm; median, 8 mm). Twenty-nine eyes had tumor thickness of 5 mm or less and 72 eyes had tumor thickness of more than 5 mm. The closest edge of the tumor was 2 mm or less from the optic nerve in 45 eyes and 2 mm or less from the foveola in 40 eyes. Vitreous seeds were focal in 21 eyes and diffuse in 80 eyes and were located in the anterior vitreous in 20 eyes and in the posterior vitreous in 49; they were diffuse in 32 eyes. Thirty-three eyes had concurrent subretinal fluid and 28 eyes had subretinal seeds. Management consisted of high-dose chemotherapy with a combination of vincristine, etoposide, and carboplatin (Table 2). Chemotherapy cycles ranged from 6 to 12 (mean, 6.3
1.1; median, 6). Seventy-three eyes (33 of 40 with ICRB group D disease and 40 with ICRB group E disease) with diffuse vitreous seeds that remained unresolved with the initial 2 cycles of highdose chemotherapy received a 15-mg deep posterior sub-Tenon carboplatin injection along with each cycle of chemotherapy from the third cycle onward until complete regression of the vitreous seeds. The number of injections ranged from 1 to 13 (mean, 5.8
3.0; median, 6). Eighty-four patients underwent chemocryotherapy 1 to 2 hours before each cycle of chemotherapy. Focal treatments including cryotherapy (42 eyes) and transpupillary thermotherapy (54 eyes) were performed during examination under anesthesia before administration of chemotherapy. Cryotherapy (35 eyes) and thermotherapy (62 eyes) were continued beyond cycles of chemotherapy until complete tumor regression. Follow-up duration ranged from 13.4 to 129.2 months (mean, 50.5
20.9 months; median, 48.0 months). Overall, after high-dose chemotherapy with or without periocular carboplatin and focal treatment, 53 eyes (52%) showed complete regression of the tumor and vitreous seeds, whereas 15 eyes (15%) needed enucleation for residual or recurrent tumor, vitreous seeds, or both with no scope for gain in vision as judged clinically. The remaining 33 eyes

519

Ophthalmology Volume 121, Number 2, February 2014

Figure 1. Retinoblastoma with focal vitreous seeds (A) in the left eye of a 9-month-old child showing (B) complete regression after 6 cycles of high-dose chemotherapy.

(33%) underwent fractionated stereotactic EBRT for residual or recurrent tumor, vitreous seeds, or both with the dose ranging from 3900 to 5000 cGy. Complete regression was achieved in 24 eyes (73%) after EBRT, whereas 9 eyes (27%) needed enucleation for residual or recurrent tumor, vitreous seeds, or both. Of 21 patients with ICRB group C retinoblastoma, 18 (86%) had complete regression of the tumor and vitreous seeds with high-dose chemotherapy (Fig 1), whereas 3 patients (14%; tumor in 1 patient, vitreous seeds in 1 patient, and both in 1 patient) for whom chemotherapy failed needed EBRT. After EBRT, 2 patients showed complete regression, whereas 1 patient needed enucleation for massive recurrence of retinal tumors and vitreous seeds. In all, 20 eyes (95%) with ICRB group C retinoblastoma were salvaged. All of them had visual acuity of 20/200 or better. Of 40 eyes with ICRB group D retinoblastoma, 28 eyes (70%) had complete regression of the tumor and vitreous seeds with either high-dose chemotherapy alone (4 of 7 eyes) or high-dose chemotherapy with periocular carboplatin (24 of 33 eyes; Fig 2). Five eyes (12.5%) in which chemotherapy failed with no scope for visual recovery as assessed clinically underwent enucleation for residual retinoblastoma (tumor in 1 eye, vitreous seeds in 1 eye, both in 1 eye) or recurrent retinoblastoma (tumor in 1 eye, vitreous seeds in 1 eye). Seven eyes (17.5%; 3 eyes with residual tumor, 2 eyes with residual vitreous seeds, and 2 eyes with both) in which chemotherapy failed underwent EBRT. Of these, 6 eyes (86%) showed complete regression, whereas 1 eye (14%) finally had to undergo enucleation for residual tumor and vitreous seeds. Seven

of 40 patients with ICRB group D retinoblastoma, who showed resolution of diffuse vitreous seeds with an initial 2 cycles of high-dose chemotherapy, did not receive periocular carboplatin. Four of them had complete tumor and vitreous seed regression with 6 cycles of high-dose chemotherapy, whereas 3 regressed after EBRT for residual tumor. Of 33 patients with ICRB group D retinoblastoma with residual diffuse vitreous seeds after an initial 2 cycles of high-dose chemotherapy who received periocular carboplatin injections, 26 patients (79%) showed complete regression of vitreous seeds. Two of them, however, underwent enucleation for residual retinal tumor. In all, 34 eyes (85%) with ICRB group D retinoblastoma were salvaged. Thirty-three (97%) had visual acuity of 20/200 or better. Of 40 eyes with ICRB group E retinoblastoma, 7 eyes (17.5%) showed complete tumor and vitreous seed regression with highdose chemotherapy with focal treatment and periocular carboplatin. Sixteen eyes (40%) showed complete vitreous seed regression with high-dose chemotherapy and periocular carboplatin; however, 9 eyes needed further treatment for residual retinal tumor. Ten eyes (25%) in which chemotherapy failed with no scope for visual recovery as assessed clinically underwent enucleation for residual retinoblastoma (tumor in 2 eyes, vitreous seeds in 2 eyes, both in 1 eye) or recurrent retinoblastoma (tumor in 1 eye, vitreous seeds in 1 eye, both in 3 eyes). Twenty-three eyes (57.5%; 6 eyes with residual tumor, 4 eyes with residual vitreous seeds, and 13 eyes with both) in which chemotherapy failed underwent EBRT. Of these, 16 eyes (70%) showed complete regression, whereas 7 eyes (30%) finally had to undergo enucleation for residual tumor and

Figure 2. A, Retinoblastoma with massive vitreous seeds in the only remaining (left) eye of a 3-year-old child with bilateral retinoblastoma. B, After 6 cycles of high-dose chemotherapy and periocular carboplatin injections, the tumor and the vitreous seeds show complete regression.

520

Manjandavida et al



Retinoblastoma with Vitreous Seeds

Table 3. Outcome of Retinoblastoma with Vitreous Seeds: Univariate and Multivariate Analysis with Factors Predicting Tumor Regression, Tumor Recurrence, Vitreous Seed Regression, Vitreous Seed Recurrence, Need for External Beam Radiotherapy, Eye Salvage, and Vision Salvage Cox Univariate Analysis

Outcome Measure Tumor regression

Tumor recurrence

Vitreous seed regression

Vitreous seed recurrence

Need for EBRT Eye salvage

Hazard Ratio (95% Confidence Interval)

Risk Factor

P Value

Bilateral retinoblastoma Absence of subretinal fluid Absence of subretinal seeds Tumor basal diameter 16 mm Tumor height 5 mm Distance from optic disc >2 mm Tumor basal diameter 16 mm Distance from optic disc 2 mm Absence of periocular carboplatin injection Absence of EBRT Bilateral retinoblastoma Absence of subretinal fluid Absence of subretinal seeds Tumor basal diameter 16 mm ICRB group E Endophytic tumor Absence of periocular carboplatin injection ICRB group E Bilateral retinoblastoma Absence of subretinal fluid Absence of subretinal seeds Tumor basal diameter 16 mm Tumor height 5 mm Distance from fovea >2 mm

0.009 0.002 0.001 0.006 0.027 0.045 NS NS NS

2.81 3.44 3.48 3.22 3.78 3.42

0.025 0.036 0.012 0.017 0.050 NS NS NS

2.92 2.46 2.83 2.71 2.28

NS 0.012 0.003 0.004 0.010 0.022 0.050

3.10 3.84 3.50 3.54 10.80 3.40

Cox Multivariate Analysis

P Value

Hazard Ratio (95% Confidence Interval)

(1.29e6.13) (1.60e7.43) (1.63e7.45) (1.38e7.47) (1.17e13.49) (1.02e11.44) NA NA NA

0.041 0.013 NA NA NA NA 0.050 0.019 0.010

2.78 (1.31e9.83) 3.59 (1.04e7.45) NA NA NA NA 8.19 (1.01e67.91) 34.19 (1.80e646.92) 12.82 (1.73e95.24)

(1.15e7.48) (1.06e5.69) (1.26e6.36) (1.20e6.14) (1.01e5.30) NA NA NA

NA NA 0.018 0.050 NA 0.050 0.045 0.040

5.76 6.92 4.15

NA (1.28e7.50) (1.60e9.24) (1.48e8.29) (1.35e9.33) (1.42e82.15) (1.01e11.60)

0.015 0.026 0.005 NA NA NA NA

2.75 (1.22e6.23) 2.76 (1.13e6.77) 3.50 (1.45e8.45) NA NA NA NA

3.05 2.63

NA NA (1.21e7.68) (1.01e6.96) NA (1.01e33.40) (1.04e46.00) (1.03e16.79)

EBRT ¼ external beam radiotherapy; ICRB ¼ International Classification of Retinoblastoma; NA = not applicable; NS ¼ not significant.

vitreous seeds (n ¼ 3), recurrent tumor (n ¼ 2), or recurrent vitreous seeds (n ¼ 2). In all, 23 eyes (57.5%) with ICRB group E retinoblastoma were salvaged. Twenty-one (91%) of these had visual acuity of 20/200 or better. Overall, 77 of 101 eyes (76%) achieved complete regression, 53 after high-dose chemotherapy (chemotherapy alone in 22 eyes, chemotherapy with periocular carboplatin in 31 eyes), and 24 with additional EBRT. Twenty-four of 33 eyes (73%) in which chemotherapy failed were salvaged with EBRT. Seventy-four of 77 salvaged eyes (96%) achieved visual acuity of 20/200 or better at the final follow-up. Of these, 46 eyes (60%) had visual acuity of 20/40 or better, 14 eyes (18%) had visual acuity of 20/50 to 20/80, 14 eyes (18%) had visual acuity of 20/ 100 to 20/200, and 3 eyes (4%) had visual acuity worse than 20/ 200. Twenty-four eyes (24%) needed enucleation, either for failure of chemotherapy (n ¼ 15) or for failure of chemotherapy and EBRT (n ¼ 9). None of the patients demonstrated a second malignant neoplasm or systemic metastasis during the follow-up period. Factors predicting tumor regression were bilateral retinoblastoma (P ¼ 0.009), absence of subretinal fluid (P ¼ 0.002), absence of subretinal seeds (P ¼ 0.001), tumor basal diameter of 16 mm or less (P ¼ 0.006), tumor thickness of 5 mm or less (P ¼ 0.027), and distance from optic disc of more than 2 mm (P ¼ 0.045) by the univariate analysis; and bilateral retinoblastoma (P ¼ 0.041; hazard ratio [HR], 2.78; 95% confidence interval [CI], 1.31e9.83) and absence of subretinal fluid (P ¼ 0.013; HR, 3.59; 95% CI, 1.04e7.45) by the multivariate analysis (Table 3). Factors

influencing tumor recurrence were tumor basal diameter of more than 16 mm (P ¼ 0.050; HR, 8.19; 95% CI, 1.01e67.91), distance from optic disc of 2 mm or less (P ¼ 0.019; HR, 34.19; 95% CI, 1.80e646.92), and absence of periocular carboplatin injection (P ¼ 0.010; HR, 12.82; 95% CI, 1.73e95.24) by the multivariate analysis (Table 3). Factors predicting vitreous seed regression were bilateral retinoblastoma (P ¼ 0.036), absence of subretinal fluid (P ¼ 0.012), absence of subretinal seeds (P ¼ 0.017), and tumor basal diameter of 16 mm or less (P ¼ 0.050) by the univariate analysis; and absence of subretinal fluid (P ¼ 0.018; HR, 3.05; 95% CI 1.21e7.68) and absence of subretinal seeds (P ¼ 0.050; HR, 2.63; 95% CI, 1.01e6.96) by the multivariate analysis. Predictors of vitreous seed recurrence were ICRB group E disease (P ¼ 0.050; HR, 5.76; 95% CI 1.01e33.40), endophytic tumor (P ¼ 0.045; HR, 6.92; 95% CI, 1.04e46.00), and absence of periocular carboplatin injection (P ¼ 0.040; HR, 4.15; 95% CI, 1.03e16.79) by the multivariate analysis (Table 3). Need for EBRT correlated with ICRB group E tumor (P ¼ 0.015; HR, 2.75; 95% CI, 1.22e6.23) by the multivariate analysis (Table 3). Factors predicting eye salvage included bilateral retinoblastoma (P ¼ 0.012), ICRB group C tumor (P ¼ 0.022), focal vitreous seeds (P ¼ 0.022), absence of subretinal fluid (P ¼ 0.003), absence of subretinal seeds (P ¼ 0.004), tumor basal diameter of 16 mm or less (P ¼ 0.010), tumor thickness of 5 mm or less (P ¼ 0.022), and distance from the fovea of more than 2 mm (P ¼ 0.050) by univariate analysis; and bilateral retinoblastoma (P ¼ 0.026; HR, 2.76; 95% CI, 1.13e6.77) and

521

Ophthalmology Volume 121, Number 2, February 2014 absence of subretinal fluid (P ¼ 0.005; HR, 3.50; 95% CI, 1.45e8.45) by the multivariate analysis (Table 3). No factor had a statistically significant correlation with vision salvage. Complications of chemotherapy included transient anemia (n ¼ 11), pancytopenia (n ¼ 6), and thrombocytopenia (n ¼ 4). All these patients received fresh blood transfusion (n ¼ 11; range, 1e4 transfusions; median, 2 transfusions), platelet transfusion (n ¼ 4; range, range, 1e6 transfusions; median, 3 transfusions), or granulocyte colony-stimulating factor injection (n ¼ 3; range, 1e2 injections; median, 2 injections). The complications of radiotherapy were dry eye (n ¼ 4), radiation keratopathy (n ¼ 3), cataract (n ¼ 4), and radiation retinopathy (n ¼ 1). Except for localized conjunctival scar, there was no major complication of periocular carboplatin.

Discussion Chemotherapy with focal consolidation is used extensively in the primary management of retinoblastoma.2,6e13,34e36 Standard triple-drug chemotherapy (vincristine 0.05 mg/kg body weight, day 1þetoposide 5 mg/kg body weight, day 1 and day 2þcarboplatin 18.6 mg/kg body weight, day 1) is most effective for tumors without associated vitreous seeds.2,6e13,34e36 Success, defined as eye salvage, is reported in 85% of treated patients when the tumor is less advanced (Reese-Ellsworth groups IeIV) and only 47% when it is more advanced (Reese-Ellsworth group V).12 Although only 10% of eyes in groups I through IV required additional EBRT for eye salvage, 47% of eyes in group V did.12 Gunduz et al36 reviewed 105 eyes with retinoblastoma treated with chemotherapy and found that eye salvage was achieved in 87% of eyes classified as Reese-Ellsworth groups I through IV and in only 36% of eyes classified as Reese-Ellsworth group V. In their series, 21% of eyes with Reese-Ellsworth groups I through IV tumors and 31% of eyes with Reese-Ellsworth group V tumors needed additional EBRT in an attempt to salvage the eye. Rates of regression of retinoblastoma and eye salvage with standard triple-drug chemotherapy have been suboptimal for ICRB group D and E tumors.2,6e13,34e36 Apart from chemotherapy, the other treatment options for tumors with vitreous seeds are enucleation or EBRT. Enucleation may not be a reasonable option if it is the only remaining eye of a patient with bilateral retinoblastoma or if there is visual potential. Although EBRT is an option, it has limited success in the presence of vitreous seeds and has a high risk of a second malignant neoplasm, and hence is not preferred as the primary management method. A recent trend is to use EBRT in conjunction with chemotherapy.14,34 In one study, ICRB group E retinoblastoma managed with chemotherapy in combination with low-dose EBRT showed less need for enucleation than eyes treated with chemotherapy alone.14 Of the 64 eyes in the chemotherapy group, 34% were enucleated after chemotherapy alone and 20% were enucleated after chemotherapy followed by a therapeutic dose EBRT. Of the 12 eyes in the chemotherapy plus low-dose EBRT group, only 17% were enucleated.14 However, with follow-up being limited, the study was not designed to address the issues of a second malignant neoplasm, which remains a concern specifically in heritable retinoblastoma.15 Providing EBRT to

522

every patient with advanced retinoblastoma as part of the treatment protocol adds to the cost and needlessly exposes several patients to the risks of radiation. It does make sense to consider only those in whom primary chemotherapy for EBRT has failed. Superselective intra-arterial chemotherapy has generated tremendous excitement.16e20 Reported results show rates of eye salvage of 33% to 67% in the presence of vitreous seeds.17,18,20 Although early reports are impressive, potential vascular complications and the risk of metastasis are valid concerns.16e20 Suboptimal concentration of chemotherapeutic agents in vitreous results in the persistence of vitreous seeds. Although there have been recent reports of direct intravitreal chemotherapy for the management of vitreous seeds, such an approach has not yet gained wide acceptance because of apprehension regarding possible extraocular dissemination of intraocular retinoblastoma and pending long-term results.22e27 The high frequency of administration, 6 injections on average every 7 to 10 days, is also a major concern.22e27 Intravitreal melphalan and methotrexate in varying doses have provided 40% to 87% eye salvage.22e27 Escalation of the systemic dose of etoposide and carboplatin (high-dose chemotherapy; Table 2) possibly results in higher intraocular drug concentration.40 The role of highdose chemotherapy in the management of retinoblastoma has not been examined extensively in the literature. Chemocryotherapy breaks the blooderetinal barrier and enables higher intravitreal concentration of carboplatin when injected systemically.35 Periocular carboplatin and topotecan injection also results in higher intravitreal drug level.28e33 Transscleral penetration of posterior sub-Tenon carboplatin leads to augmented vitreous concentration. The results of a phase I/II clinical trial of periocular carboplatin in the treatment of intraocular retinoblastoma found carboplatin to be beneficial.28 Honavar et al evaluated the role of posterior sub-Tenon carboplatin injection in addition to intravenous chemotherapy in the only remaining eye of patients with intraocular retinoblastoma with vitreous seeds (Honavar SG, et al. Periocular carboplatin in the management of advanced intraocular retinoblastoma. Paper presented at: XII International Congress of Ocular Oncology, September 1e5, 2005; Vancouver). In comparison with a control group in whom only intravenous chemotherapy was used (30% eye salvage), they found a better outcome (70% eye salvage) in the group in which deep posterior sub-Tenon carboplatin was used, thus demonstrating the efficacy of posterior sub-Tenon carboplatin in patients with vitreous seeds. In the current study, periocular carboplatin injection was initiated only in those eyes in which diffuse vitreous seeds persisted beyond the initial 2 cycles of high-dose chemotherapy. This included 33 of 40 eyes with ICRB group D tumors and all the 40 eyes with ICRB group E tumors. Of these, 26 eyes (79%) in ICRB group D and 16 eyes (40%) in ICRB group E showed complete resolution of diffuse vitreous seeds after initiation of periocular carboplatin injection. Our multivariate analysis showed that periocular carboplatin injection was protective against vitreous seed recurrence (Table 3). However, our study is limited by the absence of a control group to compare and conclusively establish the beneficial role of periocular carboplatin. The ongoing

Manjandavida et al



Retinoblastoma with Vitreous Seeds

Children’s Oncology group study (NCT00072384) “Systemic Vincristine, Carboplatin, and Etoposide, Subtenons Carboplatin, and Local Ophthalmic Therapy in Treating Children with Intraocular Retinoblastoma” is aimed at systematically evaluating the role of periocular chemotherapy in cases with intraocular retinoblastoma ICRB group C and D.39 Local drug toxicity and inflammation causing limitation of ocular motility, preseptal cellulitis, and optic atrophy are some of the complications described after periocular chemotherapy.29,30 We have reason to believe that many of these complications may be related to the injection technique. With our technique of deep posterior sub-Tenon delivery of periocular carboplatin alternating in the superonasal, inferonasal, and inferotemporal quadrants in between the recti muscles and sparing the superotemporal quadrant to minimize the risk of damage to the lacrimal gland and consequent dry eye, we did not find any significant complication except for localized conjunctival fibrosis with no functional implications, in contrast to what is reported in the literature (Honavar SG, et al. Periocular carboplatin in the management of advanced intraocular retinoblastoma. Proceedings of the XII International Congress of Ocular Oncology, September 1e5, 2005; Vancouver). Improved methods of drug delivery such as fibrin sealant32 and use of nanoparticle carboplatin33 are expected to decrease the risk of adverse effects. We have adapted high-dose chemotherapy for retinoblastoma with vitreous seeds and concurrent periocular carboplatin specifically in eyes with diffuse vitreous seeds as our primary management strategy. Chemocryotherapy is performed before each cycle of chemotherapy. Focal treatmentdcryotherapy, transpupillary thermotherapy, or bothdis provided as appropriate throughout the treatment period. Prompt enucleation is performed after chemotherapy in patients with residual or recurrent tumor or vitreous seeds with no scope for vision recovery as judged clinically. Selectively, those with residual or recurrent tumor or vitreous seeds and with visual potential undergo stereotactic EBRT. Those in whom EBRT has failed and who have no visual potential undergo enucleation. With this approach, we had a 95% rate of eye salvage in eyes with focal vitreous seeds and a 71% rate of eye salvage in those with diffuse vitreous seeds. The eye salvage rates were 95% in ICRB group C, 85% in group D, and 57.5% in group E. Overall, we achieved 77% eye salvage: 53% with primary chemotherapy and focal treatment with or without periocular carboplatin and an additional 24% with EBRT in patients who had residual or recurrent retinoblastoma after chemotherapy. Significantly, 96% of those who had eye salvage had vision salvage as well. Factors predicting tumor regression and eye salvage were bilateral retinoblastoma and absence of subretinal fluid by the multivariate analysis. Factors predicting vitreous seed regression were absence of subretinal fluid and subretinal seeds. The systemic complications of high-dose chemotherapy were transient. None of the patients demonstrated a second malignant neoplasm or systemic metastasis. However, we advise caution, careful patient selection, and detailed discussion with the family regarding the risks versus potential benefits, specifically in ICRB group E eyes. Eyes with no visual potential as judged clinically and with the tumor obscuring the optic disc may be considered for

enucleation. Primary enucleation continues to be the treatment of choice in ICRB group E eyes with anterior segment tumor infiltration and neovascular glaucoma. Our study has the inherent limitations of a retrospective data set. However, we have the advantage of a large series of patients who underwent protocol-based management at a single center. The results of our study point to the benefits of primary high-dose chemotherapy for retinoblastoma with focal vitreous seeds and high-dose chemotherapy with concurrent periocular carboplatin for those with diffuse vitreous seeds, with impressive eye and vision salvage rates. In patients in whom chemotherapy has failed, EBRT used selectively provides additional benefit.

References 1. Shields JA, Shields CL, Sivalingam V. Decreasing frequency of enucleation in patients with retinoblastoma. Am J Ophthalmol 1989;108:185–8. 2. Shields CL, Honavar SG, Meadows AT, et al. Chemoreduction in unilateral retinoblastoma. Arch Ophthalmol 2002;120: 1653–8. 3. Epstein JA, Shields CL, Shields JA. Trends in the management of retinoblastoma: evaluation of 1,196 consecutive eyes during 1974 to 2001. J Pediatr Ophthalmol Strabismus 2003;40: 196–203. 4. Berman EL, Donaldson CE, Giblin M, Martin FJ. Outcomes in retinoblastoma, 1974e2005: the Children’s Hospital, Westmead. Clin Experiment Ophthalmol 2007;35:5–12. 5. Broaddus E, Topham A, Singh AD. Survival with retinoblastoma in the USA: 1975e2004. Br J Ophthalmol 2009;93: 24–7. 6. Kingston JE, Hungerford JL, Madreperla SA, Plowman PN. Results of combined chemotherapy and radiotherapy for advanced intraocular retinoblastoma. Arch Ophthalmol 1996;114:1339–43. 7. Murphree AL, Villablanca JG, Deegan WF III, et al. Chemotherapy plus local treatment in the management of intraocular retinoblastoma. Arch Ophthalmol 1996;114: 1348–56. 8. Shields CL, DePotter P, Himmelstein BP, et al. Chemoreduction in the initial management of intraocular retinoblastoma. Arch Ophthalmol 1996;114:1330–8. 9. Gallie BL, Budning A, DeBoer G, et al. Chemotherapy with focal therapy can cure intraocular retinoblastoma without radiotherapy. Arch Ophthalmol 1996;114:1321–8. 10. Shields CL, Shields JA, Needle M, et al. Combined chemoreduction and adjuvant treatment for intraocular retinoblastoma. Ophthalmology 1997;104:2101–11. 11. Shields CL, Honavar SG, Meadows AT, et al. Chemoreduction plus focal therapy for retinoblastoma: factors predictive of need for treatment with external beam radiotherapy or enucleation. Am J Ophthalmol 2002;133:657–64. 12. Shields CL, Honavar SG, Shields JA, et al. Factors predictive of recurrence of retinal tumor, vitreous seeds, and subretinal seeds following chemoreduction for retinoblastoma. Arch Ophthalmol 2002;120:460–4. 13. Cohen VM, Kingston J, Hungerford JL. The success of primary chemotherapy for group D heritable retinoblastoma. Br J Ophthalmol 2009;93:887–90. 14. Chan MP, Hungerford JL, Kingston JE, Plowman PN. Salvage external beam radiotherapy after failed primary chemotherapy

523

Ophthalmology Volume 121, Number 2, February 2014

15.

16.

17.

18. 19. 20.

21. 22. 23. 24. 25.

26. 27.

for bilateral retinoblastoma: rate of eye and vision preservation. Br J Ophthalmol 2009;93:891–4. Kleinerman RA, Tucker MA, Tarone RE, et al. Risk of new cancers after radiotherapy in long-term survivors of retinoblastoma: an extended follow-up. J Clin Oncol 2005;23: 2272–9. Abramson DH, Dunkel IJ, Brodie SE, et al. A phase I/II study of direct intraarterial (ophthalmic artery) chemotherapy with melphalan for intraocular retinoblastoma: initial results. Ophthalmology 2008;115:1398–404. Shields CL, Bianciotto CG, Jabbour P, et al. Intra-arterial chemotherapy for retinoblastoma: report no. 1, control of retinal tumors, subretinal seeds, and vitreous seeds. Arch Ophthalmol 2011;129:1399–406. Shields CL, Bianciotto CG, Jabbour P, et al. Intra-arterial chemotherapy for retinoblastoma: report no. 2, treatment complications. Arch Ophthalmol 2011;129:1407–15. Wilson MW, Haik BG, Dyer MA. Superselective intraophthalmic artery chemotherapy: what we do not know. Arch Ophthalmol 2011;129:1490–1. Abramson DH, Marr BP, Dunkel IJ, et al. Intra-arterial chemotherapy for retinoblastoma in eyes with vitreous and/or subretinal seeding: 2-year results. Br J Ophthalmol 2012;96: 499–502. Eagle RC Jr, Shields CL, Bianciotto C, et al. Histopathologic observations after intra-arterial chemotherapy for retinoblastoma. Arch Ophthalmol 2011;129:1416–21. Seregard S, Kock E, Trampe E. Intravitreal chemotherapy for recurrent retinoblastoma in an only eye [letter]. Br J Ophthalmol 1995;79:194–5. Suzuki S, Kaneko A. management of intraocular retinoblastoma and ocular prognosis. Int J Clin Oncol 2004;9:1–6. Kivelä T, Eskelin S, Paloheimo M. Intravitreal methotrexate for retinoblastoma [letter]. Ophthalmology 2011;118:1689. Munier FL, Gaillard MC, Balmer A, et al. Intravitreal chemotherapy for vitreous disease in retinoblastoma revisited: from prohibition to conditional indications. Br J Ophthalmol 2012;96:1078–83. Ghassemi F, Shields CL. Intravitreal melphalan for refractory or recurrent vitreous seeding from retinoblastoma. Arch Ophthalmol 2012;130:1268–71. Seregard S, Singh AD. Retinoblastoma: direct chemotherapeutic drug delivery into the vitreous cavity. Br J Ophthalmol 2012;96:473–4.

28. Abramson DH, Frank CM, Dunkel IJ. A phase I/II study of subconjunctival carboplatin for intraocular retinoblastoma. Ophthalmology 1999;106:1947–50. 29. Schmack I, Hubbard GB, Kang SJ, et al. Ischemic necrosis and atrophy of the optic nerve after periocular carboplatin injection for intraocular retinoblastoma. Am J Ophthalmol 2006;142: 310–5. 30. Mulvihill A, Budning A, Jay V, et al. Ocular motility changes after subtenons carboplatin chemotherapy for retinoblastoma. Arch Ophthalmol 2003;121:1120–4. 31. Mallipatna AC, Demaras H, Chan HS, et al. Periocular topotecan for intraocular retinoblastoma. Arch Ophthalmol 2011;129:738–45. 32. Van Quill KR, Dioguardi PK, Tong CT, et al. Subconjunctival carboplatin in fibrin sealant in the treatment of transgenic murine retinoblastoma. Ophthalmology 2005;112:1151–8. 33. Shome D, Poddar N, Sharma V, et al. Does a nanomolecule of carboplatin injected periocularly help in attaining higher intravitreal concentrations? Invest Ophthalmol Vis Sci 2009;50:5896–900. 34. Shields CL, Ramasubramanian A, Thangappan A, et al. Chemoreduction for group E retinoblastoma: comparison of chemoreduction alone versus chemoreduction plus low-dose external radiotherapy in 76 eyes. Ophthalmology 2009;116:544–51. 35. Shields CL, Mashayekhi A, Au AK, et al. The International Classification of Retinoblastoma predicts chemoreduction success. Ophthalmology 2006;113:2276–80. 36. Gunduz K, Gunalp I, Yalcindag N, et al. Causes of chemoreduction failure in retinoblastoma and analysis of associated factors leading to eventual treatment with external beam radiotherapy and enucleation. Ophthalmology 2004;111:1917–24. 37. Chantada GL. Retinoblastoma: lessons and challenges from developing countries. Ellsworth Lecture 2011. Ophthalmic Genet 2011;32:196–203. 38. Wilson TW, Chan HS, Moselhy GM, et al. Penetration of chemotherapy into vitreous is increased by cryotherapy and cyclosporine in rabbits. Arch Ophthalmol 1996;114:1390–5. 39. Meadows AT, Chintagumpala M, Dunkel IJ, et al. Children’s Oncology Group (COG) trials for retinoblastoma. In: Singh AD, Damato BE, Pe’er J, et al., eds. Clinical Ophthalmic Oncology. Philadelphia, PA: Saunders-Elsevier; 2007:491–5. 40. Reddy VA, Honavar SG, Shome D, et al. Demographics, clinical profile, management and outcome of retinoblastoma. J Clin Oncol 2008;26(Suppl):Abstr 10051.

Footnotes and Financial Disclosures Originally received: December 1, 2012. Final revision: September 5, 2013. Accepted: September 8, 2013. Available online: October 28, 2013.

Manuscript no. 2012-1803.

1

Ocular Oncology Service, L. V. Prasad Eye Institute, Hyderabad, India.

2

Ocular Oncology Service, Centre for Sight, Hyderabad, India.

Presented in part at: American Academy of Ophthalmology Annual Meeting, October 22e25, 2011, Orlando, Florida.

524

Financial Disclosure(s): The author(s) have no proprietary or commercial interest in any materials discussed in this article. Supported by the Hyderabad Eye Research Foundation, Hyderabad, India. Correspondence: Santosh G. Honavar, MD, FACS, Ocular Oncology Service, Centre for Sight, Road No 2, Banjara Hills, Hyderabad 500034, India. E-mail: santosh. [email protected].