Reese–Ellsworth (R–E) group V disease remains the greatest challenge in the successful management of intraocular retinoblastoma.1 In the past, primary treatment of heritable, bilateral R–E group V disease consisted of whole-eye radiotherapy. In the late 1980s, our institution combined four cycles of vincristine, etoposide and carboplatin (VEC chemotherapy) with external beam radiotherapy, in order to improve ocular retention and visual outcome.2 It is now widely acknowledged that external beam radiotherapy delays orbital growth, and secondary malignancy may occur in structures that have received low-dose radiation in childhood. Those at greatest risk of secondary malignancy are children irradiated under the age of 1 year who carry the germ line RB1 mutation.3 Consequently, in the early 1990s, primary radiotherapy was abandoned in London. Since 1995, children with heritable, bilateral R–E group V disease have received six cycles of primary multiagent VEC chemotherapy at 3-weekly intervals.

Chemotherapy can shrink large tumours so they are amenable to further focal treatment (chemoreduction) or delay the need for radiotherapy (chemodelay). R–E groups I–IV show a good response to primary chemotherapy,4 and in this group 85% globe survival has been achieved.5 Small macular tumours greater than 2 mm in diameter show the greatest response to systemic chemotherapy.6 However, chemotherapy is also effective for advanced intraocular retinoblastoma where dramatic early tumour shrinkage is frequently observed, and the presence of a total retinal detachment is not a contraindication to chemotherapy.7 8 The current study examines the success of primary chemotherapy in a select group of children with heritable, group D (R–E group V) retinoblastoma.

METHODS

The retinoblastoma database was used to identify all patients with heritable, bilateral retinoblastoma referred to the Paediatric Ocular Oncology Unit of St Bartholomew’s and the Royal London Hospital between 1995 and 2003. The minimum follow-up period for entry into the study was 1 year. Only children with group D disease (R–E Group V) in at least one eye who had completed six cycles of primary vincristine, etoposide and carboplatin were included in the study. All eyes that underwent primary enucleation were excluded. No adjuvant focal treatment was performed during chemotherapy, but focal salvage treatment was performed, if required, following completion of chemotherapy. The indications for focal salvage treatment included edge recurrence, lack of response to primary chemotherapy, the appearance of new retinal tumours and localised vitreous seeding.

Eyes were classified at diagnosis according to the R–E criteria. Information from the clinical records and Retcam photographs were used to reclassify the eyes according to the International Retinoblastoma classification system. Eyes with diffuse retinoblastoma, secondary glaucoma, extraocular tumour extension or anterior segment involvement were excluded. All tumours occupied at least 50% of the globe and were associated with a total retinal detachment with diffuse vitreous or subretinal seeds. Therefore, they met the criteria for group D eyes.9

Clinical data regarding the patient demographics, tumour response and need for focal salvage treatment were gathered retrospectively. Patient data included the gender of the child. Follow-up data included the timing of salvage treatment or enucleation (in months from completion of chemotherapy), the type of salvage treatment (retinal cryotherapy, chemo-diode, ruthenium episcleral plaque radiotherapy, whole-eye or lens sparing external beam radiotherapy (LSDXT), periocular chemotherapy and second-line systemic chemotherapy), the duration of follow-up and the final visual acuity of the treated eye.

STATISTICAL ANALYSIS

Data analysis was performed using the SPSS 13.0 software package (SPSS, Chicago). The main outcome measures were ocular survival and event-free survival. An event was defined as treatment with either enucleation or external beam radiotherapy. Kaplan–Meier survival estimates for the probability of enucleation and probability of an event were performed as a function of time from completing primary chemotherapy. The relative risk of external beam radiotherapy or enucleation was calculated for unilateral (one eye with group D and the fellow eye with less advanced retinoblastoma) and bilateral group D (R–E group V) heritable retinoblastoma.

RESULTS

Between 1995 and 2003, there were 26 children with familial, bilateral retinoblastoma who had group D (R–E group V) disease in at least one eye at presentation. The analysis was limited to 15 children (30 eyes) with a minimum of 1-year follow-up. Three of 15 children had bilateral group D (R–E group V) retinoblastoma, and the remaining 12 had group D (R–E group V) in one eye with group A, B or C (R–E group I–IV) in the other eye. In total, 18 of 30 eyes had group D disease. The median age at presentation was 13 months (range 3 to 45 months), and the male-to-female ratio was 3:2. The median follow-up interval was 30 months (range 15 to 50 months). Seventy per cent of children required hospital admission for the side effects of multiagent chemotherapy; the reasons for hospital admission are listed in table 1. There was no childhood mortality or incidence of secondary malignancy.

Primary chemotherapy alone was successful in achieving complete tumour control in two (11%) of 18 eyes with group D disease. An early chemotherapeutic response was seen in both eyes, and extensive retinal detachment resolved after two cycles of chemotherapy. Thirty months later, both eyes remained disease-free with a visual acuity of 6/12 and 6/6, respectively.

Seven (39%) of 18 eyes with group D disease underwent enucleation. The median time from completion of chemotherapy to enucleation was 9 months (range 4 to 25 months). Using Kaplan–Meier analysis, the ocular survival was 67% at 2 years (fig 1). Following failed primary chemotherapy, two eyes were enucleated for subretinal retinoblastoma and one for vitreous base relapse. One eye was removed for multifocal new tumours that appeared 15 months following chemotherapy. Three eyes were enucleated for vitreous seeds. The seedlings were seen at 7, 12 and 23 months following primary chemotherapy. The eye with vitreous seeds at 23 months underwent multiple salvage treatments including lens-sparing external beam radiotherapy (LSDXT), episcleral plaque radiotherapy and orbital floor carboplatin.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Kaplan–Meier estimates ocular survival following primary chemotherapy and focal salvage treatment for group D eyes. At 2 years (24 months), the ocular survival for group D eyes was 67%.

The remaining nine (50%) of 18 group D eyes were preserved by focal salvage treatments. The successful salvage treatments are detailed in fig 2. Using Kaplan–Meier analysis, the event-free survival was 34% at 2 years (fig 3). The number of salvage treatments per eye ranged from 1 to 3. The chronological order of salvage treatment in months postchemotherapy was tabulated (tables 2, 3). Two eyes underwent episcleral plaque radiotherapy to control localised vitreous seeds at 16 months postchemotherapy. One eye with diffuse vitreous seeds underwent whole-eye radiotherapy, second-line chemotherapy and cryotherapy. Tumour control was achieved; however, the eye suffered ocular hypotony. Six eyes had LSDXT, which was successful in preventing loss of the eye in four out of six cases. In all four successful cases, LSDXT was the only salvage treatment necessary. Final visual acuity following successful LSDXT ranged between 6/24 and 6/60. Children with group D (R–E group V) disease in both eyes were five times more likely to receive LSDXT than children with group D (R–E group V) disease in one eye and less advanced disease in the fellow eye (relative risk 5; CI 1.2 to 20.5). However, the enucleation rate was not affected by the severity of retinoblastoma in the fellow eye (relative risk 0.8; CI 0.2 to 3.3).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Successful salvage treatment following primary chemotherapy for group D eyes. DXT, external beam radiotherapy; LSDXT, lens sparing external beam radiotherapy.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Kaplan–Meier estimates of event-free survival following primary chemotherapy and focal salvage treatment for group D eyes. Events included enucleation and external beam radiotherapy. At 2 years (24 months), the event-free survival for group D eyes was 34%.

In summary, primary chemotherapy alone achieved complete tumour regression in 11% of group D eyes. Using all methods of salvage treatment, including external beam radiotherapy, the ocular survival at 2 years was 67%. However, the event-free survival at 2 years was 34% for group D eyes with heritable retinoblastoma.

DISCUSSION

The success of chemotherapy for intraocular retinoblastoma is largely dependent on the extent of the tumour and the presence of subretinal and/or vitreous seeds.9 10 The majority of publications report the success of primary chemotherapy in a heterogeneous population of eyes with non-heritable and heritable cases, and include subgroup analysis for group V eyes.4 5 7 11 This series only includes eyes with heritable, group D (R–E group V) disease and is, therefore, directly comparable with ocular survival data from the same centre.

In 1996, Kingston and associates2 reported ocular survival for bilateral, heritable R–E group V disease treated with four cycles of primary vincristine, etoposide and carboplatin (VEC chemotherapy) combined with whole-eye radiotherapy (40–44 Gy in 20–22 fractions). The ocular survival using this technique was 58%. In the current study, the 2-year ocular survival following six cycles of primary VEC chemotherapy combined with subsequent focal salvage treatment was 67%. This is surprisingly comparable with the matched series from 1996,2 but only five of 15 children (seven eyes) in the current study underwent external beam radiotherapy. Most children received lens-sparing radiotherapy, which minimises the degree of orbitofacial hypoplasia.12 This study demonstrates that the current protocol of six cycles of VEC chemotherapy has reduced the need for external beam radiotherapy in group D (R–E group V) heritable retinoblastoma, but the 2-year ocular survival of 67% remains similar to 10 years ago.

More recently, the concept of event-free survival has been introduced to record the successful treatment of retinoblastoma.13 External beam radiotherapy is included as an event because of its associated morbidity and mortality. Freidman et al4 and Brichard et al11 reported event-free survival of 39% at 13 months and 38% (median follow-up 21 months), respectively, for R–E group V eyes. All eyes completed six cycles of VEC chemotherapy with focal adjuvant treatment. In contrast, Shields and associates5 reported an event-free survival of 0% at 5 years using the same treatment protocol for R–E group V disease. It is possible that the period of follow-up in previous studies4 11 may not have been sufficient to detect all tumour recurrences, although most subretinal seeds and recurrences occur within the first 3 years.10 The more likely explanation for this discrepancy in event-free survival is the laterality of the disease. All eyes with 0% event-free survival were all unilateral R–E group V eyes. There is a much lower threshold for enucleation in unilateral advanced compared with bilateral advanced retinoblastoma.14 This highlights the difficulty in interpreting results when heritable and non-heritable retinoblastoma outcomes are compared, even using the same classification system.

The International Retinoblastoma classification system has been shown to predict success following chemoreduction.15 In the current study, both classification systems were used so that comparisons could be made with past and future treatments. The evidence base for different types of treatment is not robust. We found that six cycles of standard dose VEC chemotherapy alone with subsequent focal salvage treatment achieved an event-free survival of 38% at 12 months and 34% at 24 months for group D heritable retinoblastoma. All events occurred within 15 months of completion of six cycles of primary chemotherapy, indicating that the period of follow-up in this study (median 30 months) was sufficient to detect recurrences. Following the same chemoreduction regime, but with focal adjuvant treatment, Shields and associates15 reported their event-free survival using the new classification system. They found, in a subgroup analysis, that group D eyes had an overall event-free survival of 47%. The subgroup analysis included non-heritable retinoblastoma, which does not share the same predisposition for the development of new retinal tumours.

The systemic and local treatment of intraocular retinoblastoma varies widely throughout the world. Focal adjuvant treatment is applied to consolidate the effects of chemotherapy using either laser thermotherapy or cryotherapy. However, group D retinoblastoma can be too extensive for focal laser treatment, even after two cycles of chemotherapy. Event-free survival may be improved by administering local chemotherapy.16 17 In a small pilot study, subtenon carboplatin combined with high dose of systemic VEC chemotherapy achieved an event-free survival of 58%, with a median follow-up of 24 months, for group D eyes.17 The role of local chemotherapy is currently being evaluated in the Children’s Oncology Group (COG, USA ARET 0231) trial for group C and D disease.13 In this trial, children will receive six cycles of systemic high dose VEC chemotherapy combined with subtenon carboplatin on cycles 2–4. The primary aim of this trial is to determine the event-free survival at 12 months for group D eyes.13 The COG trials will provide some important answers, but more collaborative studies are needed to determine if local delivery of chemotherapy, such as intra-arterial or intravitreal injection, will improve the survival of group D eyes.

Advances in the treatment of retinoblastoma with systemic chemotherapy have resulted in a significant reduction in childhood mortality.18 However, despite a new treatment regime, ocular survival for group D (R–E group V) eyes has not changed. This study demonstrates that primary chemotherapy alone is successful in only 11% of cases, and focal salvage or adjuvant treatment is necessary to improve ocular survival. External beam radiotherapy still plays an important role in ocular salvage for group D eyes.