Customize your JAMA Network experience by selecting one or more topics from the list below.
Retinoblastoma is the most common intraocular tumor in children.1 Historically, the treatment of retinoblastoma was enucleation.1 However, improvements in diagnosis and treatment of retinoblastoma have led not only to a survival rate well over 90%,1 but also to an increasing ability to offer globe conservation as a secondary goal of therapy. Globe-conserving therapies include cryotherapy, laser photoablation, hyperthermia, and plaque radiotherapy for smaller tumors, and external beam radiation and systemic chemotherapy for larger tumors not amenable to local control.2
Recently, multiple-agent chemotherapy with etoposide phosphate, carboplatin, and vincristine sulfate combined with transpupillary hyperthermia has been demonstrated to achieve excellent local control of retinoblastoma.3 Further, systemic chemotherapy spares the patient the risk of craniofacial abnormalities and may eliminate or decrease the risk of secondary tumors that have been associated with external beam radiation.4 However, systemic chemotherapy employed for retinoblastoma has been associated with adverse events such as myelosuppression and subsequent infections and need for transfusion of blood products.5 To our knowledge, the incidence of such complications and their effect on treatment remain unknown. This study evaluates the adverse events and resultant alterations in treatment protocol during the treatment of retinoblastoma with systemic chemotherapy.
Report of Cases
The study protocol was approved by the Institutional Review Board of the University of Miami School of Medicine, Miami, Fla. The medical records of all patients with retinoblastoma evaluated at Bascom Palmer Eye Institute, Miami, and treated with systemic chemotherapy at Jackson Memorial Hospital, Miami, between July 1, 1995, and May 31, 1999, were reviewed. Outcome measures included chemotherapy cycle delays, adjustment of chemotherapy doses owing to side effects, delays in planned examinations under anesthesia, unplanned hospital admissions, febrile episodes, myelosuppression, transfusions, and port-related complications.
Five (42%) of the 12 patients were male and all 12 had no family history of retinoblastoma. The mean age at diagnosis was 17 months (age range, 2-66 months); the mean duration of chemotherapy was 5.9 months. One of the 12 patients had extraocular disease.
All of the patients received carboplatin (20 mg/kg for patients <12 months, 550-600 mg/m2 body surface area for patients >12 months) and etoposide phosphate (5 mg/kg if the patient was <12 months, 150 mg/m2 body surface area if the patient was >12 months), 9 (75%) received vincristine sulfate (0.05 mg/kg if the patient was <12 months, 1.5-2 mg/m2 body surface area if the patient was >12 months) and 4 (33%) also received cyclosporine (5 mg/kg per hour bolus for 2 hours before chemotherapy started, then a 1.5 mg/kg per hour infusion over the next 30 hours if the patient weighed <12 kg, adjusted to 4 mg/kg per hour and 1.25 mg/kg per hour, respectively, if the patient weighed 12-30 kg, and adjusted to 3 mg/kg per hour, and 1.0 mg/kg per hour if the patient weighed >30 kg). The 1 patient who had extraocular tumor extension also received intrathecal treatment with a combination of methotrexate, hydrocortisone sodium succinate, and cytarabine. The patients underwent a mean of 7.2 cycles of chemotherapy (range, 4-10 cycles).
Nine patients (75%) had at least 1 cycle of chemotherapy delayed during the course of therapy. The 12 patients underwent a total of 86 cycles of chemotherapy, 17 (19.8%) of which were delayed an average of 9 days (range, 1-30 days). Chemotherapy was often delayed owing to a combination of factors, including neutropenia in 11 (65%) of 17 patients, thrombocytopenia in 6 (35%) of 17 patients, and febrile episode in 5 (29%) of 17 patients. Four (33%) of the 12 patients had doses of chemotherapy reduced secondary to treatment-related morbidity.
During the course of therapy, the 12 patients had a total of 91 examinations under anesthesia scheduled, 16 (18%) of which were delayed an average of 10 days (range, 1-21 days). Seven patients (58%) had at least 1 examination under anesthesia delayed. Causes of delay of examination under anesthesia included neutropenia in 11 (69%) of 16 patients, febrile episode in 7 (44%) of 16 patients, and thrombocytopenia in 4 (25%) of 16 patients.
Nine patients (75%) had at least 1 unplanned hospital admission (mean, 1.7 admission; reference range, 0-4 admissions) during the course of chemotherapy. The reasons for unplanned hospital admissions included neutropenia (5 patients, 60%), fever (4 patients, 50%), and thrombocytopenia (4 patients, 45%) Some patients had more than 1 problem leading to their hospital admission. Five patients (42%) required a transfusion of platelets; 3 patients (25%) needed a transfusion of packed red blood cells during the course of therapy.
Each patient had a central intravenous port placed under general anesthesia at the outset of therapy. Surgical time for port placement averaged 47 minutes (range, 35-90 minutes). Two of the patients underwent unplanned port removal secondary to complications (infection and exposure).
None of the patients required discontinuation of the treatment course secondary to treatment-associated morbidity. Outstanding tumor response was achieved in all patients. No patient in our series required enucleation or radiotherapy owing to failure of the planned treatment with systemic chemotherapy.
Although the use of systemic chemotherapy for retinoblastoma is not amenable to local treatment but is associated with excellent local tumor control, adverse events are likely to complicate the course of treatment. Even though these adverse events are treatable and did not result in the cessation of therapy, physicians administering the treatment and counseling the families of affected patients should be aware of the potential complications and should communicate the risks to the families.
An additional consideration in planning future systemic chemotherapy regimens for patients with retinoblastoma is the incorporation of granulocyte colony-stimulating factor (Neupogen; Amgen, Thousand Oaks, Calif) in the treatment plan. Granulocyte colony-stimulating factor has been demonstrated to have promyelocytic activity and may help to decrease the incidence of myelosuppression, infection, and need for transfusion of blood products in these patients.6
Further information concerning systemic chemotherapy-associated morbidity among patients with retinoblastoma should become available with the results of the international clinical trial funded by the National Cancer Institute to evaluate systemic chemoreduction in the management of patients with large bilateral retinoblastomas.
Corresponding author: Timothy G. Murray, MD, Bascom Palmer Eye Institute, 900 NW 17th St, Miami, FL 33136 (e-mail: firstname.lastname@example.org).
Benz MS, Scott IU, Murray TG, Kramer D, Toledano S. Complications of Systemic Chemotherapy as Treatment of Retinoblastoma. Arch Ophthalmol. 2000;118(4):572–575. doi:
Coronavirus Resource Center
Create a personal account or sign in to: