Multiple retinoblastoma in the left eye of a 7-month-old girl. A, In 1998, before chemoreduction. B, In 2003, after chemoreduction and transpupillary thermotherapy, her final visual acuity was 20/25.
Macular retinoblastoma in the left eye of a 3-month-old boy. A, In 1995, before chemoreduction. B, In 2003, after chemoreduction and transpupillary thermotherapy, his final visual acuity was 20/400.
Multiple retinoblastoma in the right eye of a 4-month-old girl. A, In 1995, before chemoreduction. B, In 2003, after chemoreduction and foveal-sparing transpupillary thermotherapy, her final visual acuity was 20/40.
Demirci H, Shields CL, Meadows AT, Shields JA. Long-term Visual Outcome Following Chemoreduction for Retinoblastoma. Arch Ophthalmol. 2005;123(11):1525-1530. doi:10.1001/archopht.123.11.1525
To evaluate long-term visual outcome following chemoreduction.
Interventional case series evaluating 54 eyes of 40 children with retinoblastoma successfully treated with chemoreduction, consisting of a combination of intravenous carboplatin, etoposide phosphate, and vincristine sulfate plus focal therapy without external beam radiotherapy or enucleation. All patients were followed up for at least 5 years. Patient and tumor data were analyzed for their effect on the main outcome measures (final visual acuities of 20/40 or better and of 20/200 or better) using univariate and multivariate regression models. Patients who failed chemoreduction were excluded.
There were 4 eyes in Reese-Ellsworth group I, 7 in group II, 3 in group III, 15 in group IV, and 25 in group V. The mean distance from the posterior tumor margin to the optic disc was 2 mm, and from the tumor margin to the foveola it was 3 mm. After a mean follow-up of 68 months, 27 eyes (50%) had a final visual acuity of 20/40 or better, and 36 eyes (67%) had final visual acuity of 20/200 or better. Of 33 eyes with macular tumor, only 8 (24%) had a final visual acuity of 20/40 or better, and 15 (45%) had a final visual acuity of 20/200 or better. Of 21 eyes with extramacular tumor, 19 (90%) had a final visual acuity of 20/40 or better, and all (100%) had a final visual acuity of 20/200 or better.
The clinical factors that predicted visual acuity of 20/40 or better were a tumor margin at least 3 mm from the foveola and optic disc and an absence of subretinal fluid.
Chemoreduction is the most important treatment modality for globe salvage in patients with retinoblastoma.1- 3 This technique involves the use of intravenous chemotherapy to reduce tumor volume to allow focal therapeutic measures.2- 6 During the past several years, prior studies2- 6 about chemoreduction for eyes with retinoblastoma have emphasized the factors related to treatment failure and the need for external beam radiotherapy or enucleation. Shields and associates7 reported that chemoreduction is successful in tumor control in about 50% of eyes at 5 years. The presence of subretinal tumor seeds at initial examination was the most important factor predictive of tumor recurrence.7 Similarly, chemoreduction is successful in preservation of the eye in 66% of eyes at 5 years.8 Treatment failure requiring enucleation generally occurs in eyes with advanced retinoblastoma with vitreous seeding.8
Following globe salvage, concern focuses on ultimate visual function of the eye. There is limited information on visual results following chemoreduction for retinoblastoma.9,10 In this study, we evaluated visual outcome in patients with retinoblastoma who were successfully treated with chemoreduction and had at least 5 years of follow-up. In addition, we identified factors predictive of good visual outcome of 20/40 or better and of 20/200 or better.
All patients with retinoblastoma who were treated with chemoreduction on the Ocular Oncology Service, Wills Eye Hospital, between July 1, 1996, and December 31, 2002, were identified. Those patients who were successfully treated with chemoreduction, avoided external beam radiotherapy and enucleation, and had at least 5 years of follow-up were included in this study. Institutional review board approval was obtained for this retrospective study.
All data were collected retrospectively. The demographic information, clinical findings, focal treatment measures, and visual outcome were tabulated. Demographic data recorded at the initial visit to the Ocular Oncology Service were patient age at diagnosis (in months), race (African American, Hispanic, Asian, or white), sex, hereditary pattern (familial or sporadic), laterality of involvement (unilateral or bilateral), and the involved eye. The total number of retinoblastomas per eye and Reese-Ellsworth classification of each eye were recorded. Each retinoblastoma 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 ophthalmoscopic techniques, as well as for thickness (in millimeters) as measured by A-scan and B-scan ultrasonography. The presence and extent of subretinal fluid and of tumor seeding in the vitreous and subretinal space were assessed. Vitreous seeds were defined as local if located less than 3 mm from the tumor and as diffuse if located at least 3 mm from the tumor. We defined macula as retina within a radius of 3 mm from the foveola. The drugs and chemotherapeutic protocol used were reported in previous studies3,7,8 and included intravenous carboplatin, etoposide phosphate, and vincristine sulfate. The patients were followed up on the Ocular Oncology Service with examination under anesthesia on a monthly basis after initiation of chemoreduction until the disease was under complete control. Subsequently, follow-up was provided every 2 to 4 months as needed. All focal treatment methods (cryotherapy, laser photocoagulation, thermotherapy, or plaque radiotherapy) were recorded. If amblyopia treatment was applied, the type and duration of amblyopia treatment and the compliance of the patient were recorded. At the last visit, Snellen visual acuity, tumor regression patterns, and systemic complications were assessed.
The clinical data were analyzed with regard to the 2 main outcome measures of final visual acuity of 20/40 or better (reading vision) and final visual acuity of 20/200 or better (navigation vision). The effect of each individual demographic datum, clinical variable, and treatment modality on the final visual outcome was analyzed by a series of univariate logistic regression analyses. All variables were analyzed as discrete categories except patient age, intraocular pressure, tumor base, tumor thickness, proximity to the optic disc, and proximity to the foveola, which were analyzed as continuous variables and subsequently grouped into discrete categories to derive cutoff values. The correlation among the predictor variables was determined by Pearson product moment correlation. The variables that were significant on a univariate level (P<.05) were entered into multivariate logistic regression analysis using forward stepwise modeling (conditional). For variables that showed a high degree of correlation, only 1 variable from the set of associated variables was entered at a time in subsequent multivariate models. A final multivariate model fitted variables identified as significant predictors (P<.05) in the stepwise model and variables considered clinically important for visual acuity outcome.
Of 528 patients with retinoblastoma managed on the Ocular Oncology Service, 246 patients were treated with chemoreduction as part of the management. Forty patients (16%) were successfully controlled without enucleation or external beam radiotherapy and had 5 or more years of follow-up. Fifty-four eyes of 40 children were included in the study.
Demographic features of the patients are given in Table 1. Each eye was grouped according to the Reese-Ellsworth classification and included Reese-Ellsworth group I in 4 eyes (7%), group II in 7 eyes (13%), group III in 3 eyes (6%), group IV in 15 eyes (28%), and group V in 25 eyes (46%). There were 135 tumors among 54 eyes, and the mean number of individual retinal tumors per eye was 3 (median, 2; range, 1-9). Of the largest tumor in each eye, the mean pretreatment retinal tumor base was 12 mm (median, 12 mm; range, 1-21 mm), the mean tumor thickness was 7 mm (median, 6 mm; range, 1-20 mm), the mean distance from the posterior tumor margin to the optic disc was 2 mm (median, 0 mm; range, 0-20 mm), and the mean distance from the tumor margin to the foveola was 3 mm (median, 1 mm; range, 0-20 mm). At the initial visit, retinoblastoma seeding was clinically apparent in the vitreous in 26 eyes (48%) and the subretinal space in 30 eyes (56%). Vitreous seeds were local in 23 (88%) of 26 eyes and diffuse in 3 eyes (12%) and involved a mean of 3 clock hours (median, 1 clock hour; range, 1-12 clock hours). Subretinal seeds involved a mean of 5 clock hours (median, 4 clock hours; range, 1-12 clock hours). Subretinal fluid was present in 32 eyes (59%) and involved a mean of 60% (median, 65%; range, 10%-100%) of the retina and extended for a mean of 9 clock hours (median, 12 clock hours; range, 1-12 clock hours).
The patients received chemoreduction as already detailed using carboplatin, etoposide, and vincristine and for a mean of 6 cycles (median, 6 cycles) and they were subsequently followed up for a mean of 68 months (median, 64 months; range, 60-96 months). Focal treatment for retinal tumor or seeding in the subretinal area included cryotherapy in 34 eyes (63%) and transpupillary thermotherapy in 35 eyes (65%). The type of regression pattern in the main tumor included type 1 in 25 eyes (46%), type 3 in 21 eyes (39%), and type 4 in 8 eyes (15%).
Table 2 summarizes visual outcome in the 54 eyes of 40 patients after a mean follow-up of 68 months. Final Snellen visual acuity was 20/40 or better in 27 eyes (50%) and 20/200 or better in 36 eyes (67%) (see also Figure 1). There were 33 eyes with macular tumors and 21 eyes with extramacular tumors. In the eyes with macular tumors, final visual acuity was 20/40 or better in 8 eyes (24%) and 20/200 or better in 15 eyes (46%) (see also Figure 2). In the eyes with extramacular tumors, final visual acuity was 20/40 or better in 19 eyes (90%) and 20/200 or better in 21 eyes (100%). In addition, 27 patients (87%) with bilateral retinoblastoma had a final visual acuity of 20/200 or better in at least 1 eye, and 23 (74%) had a final visual acuity of 20/40 or better in at least 1 eye. Occlusion treatment was successfully used in 10 (25%) patients, while 30 patients (75%) did not fully comply with occlusion treatment. Of the 10 patients who complied with occlusion treatment, 9 had a final visual acuity of 20/200 or better, and 6 had a final visual acuity of 20/40 or better (see also Figure 3). The univariate and multivariate predictive factors for Snellen visual acuities of 20/40 or better and of 20/200 or better are given in Table 3 and Table 4. There were no systemic complications after chemoreduction treatment.
Chemoreduction is an important treatment alternative for retinoblastoma. Enucleation is used less frequently in bilateral and unilateral cases. In a review of 1196 eyes with retinoblastoma treated on the Ocular Oncology Service, Wills Eye Hospital, bilateral retinoblastoma was managed with primary enucleation of at least 1 eye in 96% of cases between 1974 and 1978, but this decreased to 64% of cases between 1989 and 2001.11 Similarly, primary enucleation for unilateral retinoblastoma decreased from 96% in the 1970s to 75% in the 1990s.11 This decreasing trend of primary enucleation is owing to earlier diagnosis and improvements in conservative methods, especially chemoreduction.
Most previous studies regarding chemoreduction for treatment of retinoblastoma have focused on local tumor control, avoidance of external beam radiotherapy and enucleation, and complications of treatment. Shields and associates8 reported that chemoreduction was successful in tumor control and globe salvage in 85% of eyes classified as Reese-Ellsworth groups I through IV and in 47% of eyes classified as Reese-Ellsworth group V. Following chemoreduction, if retinal tumor recurrence was larger than 9 mm in thickness and 15 mm in base, if the subretinal seed recurrence numbered greater than 20 seeds scattered throughout the fundus, or if subretinal or vitreous seeds involved 2 or more confluent quadrants of the fundus, the patient underwent external beam radiotherapy or enucleation.8 With improved globe salvage, questions regarding visual function arise.
Historically, several studies12- 20 during the past 2 decades have focused on visual outcome in eyes following external beam radiotherapy for retinoblastoma. In a review of 232 eyes of 116 patients with bilateral retinoblastoma treated with external beam radiotherapy, Migdal13 reported that, by age 8 years, 26% of eyes had a visual acuity of 20/40 or better OU and 50% of patients achieved a final visual acuity of 20/40 or better in at least 1 eye. Hall and coworkers14 evaluated the visual outcome in 102 eyes of 74 patients with retinoblastoma treated with external beam radiotherapy and found that 58% of eyes had a visual acuity of 20/40 or better, 31% had a visual acuity between 20/50 and 20/200, 9% had a visual acuity of 20/400 or worse, and 2% underwent subsequent enucleation after treatment.
There is little reported on visual outcome following chemoreduction, but it appears similar to that following external beam radiotherapy. Lueder and Goyal12 evaluated visual function after laser hyperthermia and chemotherapy in a 9-month-old girl with macular retinoblastoma and reported normal visual behavior and a normal spatial-sweep visual evoked potential 1 year after treatment. Desjardins and associates10 reviewed visual outcome in a heterogeneous group of 37 eyes of 31 patients with retinoblastoma managed conservatively with chemoreduction and multiple focal measures. They observed a Snellen visual acuity of 20/200 or less in 10 eyes, a visual acuity between 20/200 and 20/40 in 4 eyes, and a visual acuity of 20/40 or better in 23 eyes.10 Macular involvement by the tumor and maculopathy secondary to focal conservative treatment were the most important factors causing decreased visual outcome.10
In our series of 40 patients with at least 5 years of follow-up, 27 eyes (50%) had a visual acuity of 20/40 or better, 9 (17%) had a visual acuity between 20/50 and 20/200, and 18 (33%) had a visual acuity of 20/400 or worse. In addition, 23 (74%) of the patients with bilateral retinoblastoma had a final visual acuity of 20/40 or better in at least 1 eye, and 27 (87%) had a final visual acuity of 20/200 or better in at least 1 eye. Extramacular tumors without subretinal fluid or seeds were associated with 20/40 or better visual acuity (Table 3). The presence of vitreous seeds did not affect visual outcome.
The visual prognosis of retinoblastoma involving the macula has generally been regarded as poor.13- 20 In the evaluation of 116 patients with bilateral retinoblastoma treated with external beam radiotherapy, Migdal13 reported that 63% of patients with retinoblastoma located outside the macula had a visual acuity of 20/40 or better, while 32% of patients with retinoblastoma located in the macula had a visual acuity of 20/40 or better. Similarly, Hall and coworkers14 found that 78% of eyes with a visual acuity of 20/400 or worse had tumors involving the macula, whereas 33% of eyes with a visual acuity of 20/40 or better had tumors involving the macula. These studies may be difficult to compare because the authors were not clear as to the exact region of the macula. As already noted, we defined macula as retina within a radius of 3 mm from the foveola. We observed that 8 (24%) of 33 eyes with macular tumors had a final visual acuity of 20/40 or better; however, 19 (90%) of 21 eyes with extramacular tumors had a final visual acuity of 20/40 or better. In multivariate analysis, sparing of the macula was a predictive factor for visual acuities of 20/40 or better and of 20/200 or better. Currently, we spare the fovea from consolidation treatment such as transpupillary thermotherapy and cryotherapy to allow for best vision.21
Occlusion therapy can benefit children with retinoblastoma even if it involves the macula. Watts and coworkers9 evaluated the results of occlusion treatment in 15 children with macular retinoblastoma. They observed that 80% of children in whom visual acuity could be recorded had a mean of 5 Snellen lines improvement in visual acuity after occlusion therapy. In their cohort, 73% of children had a final visual acuity of 20/200 or better, and 53% had a final visual acuity of 20/50 or better. We observed that 10 children were compliant with occlusion therapy: 9 of these children had a final visual acuity of 20/200 or better, and 6 had a final visual acuity of 20/40 or better. Once the tumors are controlled, we encourage occlusion therapy for all children with vision-threatening retinoblastoma.
Others have investigated results of occlusion therapy for eyes with organic disease.22- 24 Kushner22 evaluated 61 children with amblyopia associated with structural abnormalities and reported that 56% of children with macular pathologic conditions and 21% of children with optic disc pathologic conditions had improvement in visual acuity after amblyopia therapy. Similarly, Bradford and coworkers23 evaluated 51 children with unilateral structural abnormalities and found that 21% of children with optic disc abnormalities and 42% of children with macular abnormalities had a final visual acuity of 20/80 or better following occlusion therapy. These reports suggest that even eyes with pathologic conditions in the macula or optic disc can retrieve some visual acuity with occlusion therapy.
Our study has several potential limitations. As a tertiary referral center, we may have managed more advanced retinoblastoma patients, especially those in whom only 1 eye remained, in an effort to save the only eye. This might have resulted in poorer visual outcome. With longer follow-up and larger numbers of patients, predictive factors for good visual acuity might be altered. There may be unrecognized variables regarding clinical features of patients and tumors, as well as focal treatment measures that have an effect on final visual outcome. In our study, we assessed visual outcome using Snellen visual acuity. Further studies evaluating the visual field, color vision, dark adaptation, and cognitive function will be necessary to assess the final visual effects of retinoblastoma following chemoreduction.
In summary, for patients with retinoblastoma successfully treated with chemoreduction without enucleation or external beam radiotherapy, visual acuity can be preserved, mostly in those with extramacular tumors. Long-term visual acuity was 20/40 or better in 27 eyes (50%) and 20/200 or better in 36 (67%).
Correspondence: Carol L. Shields, MD, Ocular Oncology Service, Wills Eye Hospital, Thomas Jefferson University, 840 Walnut St, Philadelphia, PA 19107 (firstname.lastname@example.org).
Submitted for Publication: January 28, 2004; accepted January 19, 2005.
Financial Disclosure: None.
Funding/Support: This study was supported by the Paul Kayser International Award of Merit in Retina Research, Houston, Tex (Dr J. A. Shields); Macula Foundation, New York, NY (Dr C. L. Shields); and the Eye Tumor Research Foundation, Philadelphia, Pa (Dr C. L. Shields).
Previous Presentation: This study was presented in part at the American Academy of Ophthalmology 2003 Annual Meeting; November 16, 2003; Anaheim, Calif.