To prospectively determine the prevalence of high-risk histopathologic features such as choroidal involvement, optic nerve invasion, and scleral and anterior segment involvement in patients with unilateral retinoblastomas who had undergone enucleation and to estimate the event-free survival (extraocular or metastatic disease) and survival of patients with unilateral retinoblastoma with and without high-risk features.

Patients with metastatic retinoblastoma have a very poor outcome [7]. Several studies have identified risk factors which may be associated with the development of metastatic disease including post-lamina optic nerve involvement, choroidal invasion, and scleral and anterior segment involvement [8–10]. Tumor cells in the optic nerve posterior to the lamina cribrosa also confer a poorer prognosis.

Although “massive” involvement of the choroid is considered a poor risk factor, recent data suggest that choroidal involvement alone does not have a negative effect on outcome but when associated with optic nerve involvement seems to have an adverse influence on the outcome. There are fewer studies addressing scleral and anterior segment involvement.

The presence of the above risk factors either singly or in various combinations had prompted previous investigators to use chemotherapy as prophylaxis. Studies by Uusitalo et al. and Honavar et al. have shown that chemoprophylaxis is effective in reducing the occurrence of metastases in patients with retrolaminar optic nerve invasion and massive choroid invasion [11, 12]. The data from these studies are confounded by (1) different criteria for initiation of chemoprophylaxis used and (2) different chemotherapy regimens used.

This prospective study defined specific criteria for initiation of post-enucleation chemotherapy with the goal of preventing metastases and improving patient survival [13]. In addition, the chemotherapy regimen was consistent across all participating centers. It was anticipated that such a study would provide a foundation for future research by providing an estimate of the outcome associated with a uniform post-enucleation chemotherapy regimen for patients with uniformly defined histopathologic risk factors and by providing an estimate of the outcome associated with enucleation alone for patients without the defined histopathologic risk factors requiring chemotherapy. More importantly, for the first time information was gathered about the true prevalence of such high-risk histopathologic features present in the majority of the patients with unilateral retinoblastoma diagnosed in North America.

Patients with the high-risk features listed in Table 21.2 received chemotherapy consisting of 6 cycles of standard-dose carboplatin, vincristine, and etoposide (Chap.​ 11) given once every 4 weeks (Table 21.2). All other patients were treated with enucleation alone.

Patients with at least one high-risk feature for which adjuvant therapy was indicated were nonrandomly assigned to receive a single-arm adjuvant therapy regimen. All other patients were treated with enucleation alone. All patients were followed for the development of metastasis, extraocular disease, MDS/secondary leukemia, and death. The event-free survival distribution was compared to historical series according to treatment arm (adjuvant therapy or enucleation alone) [14, 15].

Patients were entered on this study from February of 2005 until May 2010, and the study has since been closed to new patient enrolment. Patients from across the United States and India were entered in this trial. Over 300 eyes were reviewed by central histopathologic review in a standardized fashion. This process was highly successful and led to a significant number of patients having their pathology reclassified. At present, the two cohorts are being monitored for differences in event-free survival and overall survival.

Using a backbone of neoadjuvant 2-agent (vincristine/carboplatin) systemic chemotherapy (chemoreduction), together with local ophthalmic therapy, the primary aim of this trial was to estimate the event-free survival rate at 2 years. An event was defined as additional chemotherapy, enucleation, external beam radiotherapy (EBRT), or death from any cause. A secondary aim was to estimate the response rate to vincristine and carboplatin after an initial single cycle of chemoreduction prior to implementing standardized local ophthalmic therapy and correlate with event-free survival.

The standard therapies to treat retinoblastoma, enucleation, and EBRT are associated with significant morbidity [16–18]. The prevalence of second malignancies following the hereditary form of retinoblastoma remains higher than that for any other pediatric malignancy, an effect worsened by the use of external beam radiation therapy [18, 19]. To avoid the associated morbidities of these therapies, and to utilize, now standardized, local ophthalmic therapies in a greater number of patients, research has been directed towards chemoreduction—using chemotherapy to reduce tumor volume in order to increase the efficacy of local therapies. In single-institutional studies, small, Group B tumors have been shown to respond to vincristine, carboplatin, and etoposide, and also to carboplatin and vincristine [20]. It is important to demonstrate that etoposide can be omitted from the treatment of these tumors since it increases the risk of infectious complications, and possibly, secondary leukemia [21].

A total of 6 cycles of chemotherapy with standard-dose vincristine and carboplatin (Chap.​ 11) was administered. Response to chemotherapy was determined following the first cycle of vincristine and carboplatin. Local ophthalmic therapy was delivered prior to the 2nd through 6th cycle as clinically indicated. Patients whose disease remained stable were continued on therapy; those who developed progressive disease at any time were treated at investigator discretion. Central review of Retcam images by three ophthalmologists was performed at diagnosis to confirm eye group.

This single-arm trial compared the event-free survival following 6 cycles of 2-drug therapy with the event-free survival expected under the standard 3-drug therapy [2]. An event was defined as the need for non-protocol therapy, including (1) any systemic chemotherapy other than or in addition to vincristine and carboplatin as defined in the protocol, (2) enucleation, (3) external beam radiation, (4) or death from any cause. For patients with bilateral disease, the need for non-protocol therapy of either eye was defined as a failure at the patient level.

The study has since been closed to new patient enrolment having met a stopping criterion defined in the protocol. Currently enrolled patients are being monitored for event-free survival as defined above.

The primary aim was to determine the event-free survival (EFS) at 12 months for eyes with Group D intraocular retinoblastoma treated with systemic high-dose carboplatin/etoposide/vincristine (CEV), subtenon carboplatin, and local ophthalmic therapy. An event was defined for each eye individually as the need for non-protocol chemotherapy, enucleation or external beam radiation, or death. Secondary aims included determination of the event-free survival at 12 months for Group C eyes treated with this regimen and to describe the toxicities, patterns of failure, and predictors of failure from findings at the diagnostic eye exam and the response status at end of therapy.