New treatment for the pediatric eye cancer retinoblastoma

Oct 17 2005

Investigators at St. Jude Children's Research Hospital have demonstrated in laboratory studies a new treatment for the pediatric eye cancer retinoblastoma that appears to be more effective than the current standard therapy, and more likely to prevent the recurrence of this cancer. A report on their work appears in the October 15 issue of Clinical Cancer Research.

Retinoblastoma is a tumor that arises in the cells of the retina, the light-sensitive tissue at the back of the eyeball, following mutation of the gene RB1. It is the third most common form of cancer in infants. Left untreated, the cancer spreads and is fatal. Retinoblastoma strikes about 250-350 infants and young children each year in the United States.

The St. Jude study suggests a new approach to treatment that could save both lives and vision, according to Michael A. Dyer, PhD, an assistant member of the department of Developmental Neurobiology. The new treatment might also prevent recurrence of retinoblastoma in children whose cancer is very advanced, he noted. Dyer is the senior author of the Clinical Cancer Care report.

The study showed that combination therapy with topotecan and carboplatin is superior to the standard triple-drug therapy using vincristine, carboplatin and etoposide. The new combination eliminates the use of etoposide, which is known to increase the risk of the children getting a form of leukemia called acute myeloblastic leukemia. The findings also suggest that vincristine, which is used in the standard combination therapy, contributes little to the treatment of retinoblastoma, and therefore can be eliminated from therapy.

Dyer's group was prompted to look for new treatments because of the difficulty in managing this cancer. Patients with retinoblastoma usually undergo enucleation (removal of the diseased eye); and children with retinoblastoma in both eyes often must undergo anticancer therapy in order to avoid the loss of both eyes and blindness.

But despite the critical need for new treatments that are more effective and have less severe side effects, researchers have been limited in their ability to identify new drugs and drug combinations that will improve the outcome of retinoblastoma. One reason for this problem is that there are not enough patients for researchers to enroll in large clinical trials designed to investigate new treatments, Dyer said. Therefore, it is critically important that any new treatment being considered for clinical trials have already demonstrated in the laboratory a high likelihood of success in clinical trials. Until last year when researchers at St. Jude developed the first "knockout" model of retinoblastoma and the two other models presented in this work, there were few options for preclinical studies on retinoblastoma. A knockout model is one that lacks one or more specific genes; knockout models are developed in order to study specific human diseases caused by gene defects.

The investigators made their findings during a sequence of experiments. First, they tested the ability of drugs and drug combinations to kill retinoblastoma tumors removed from patients and grown in tissue cultures. Then the team determined how much drug to use, and how often to use it, by doing pharmacokinetic studies. Finally, the team tested the drugs in laboratory models of the cancer they had developed for this purpose. Pharmacokinetics is the study of the rate of absorption, distribution, breakdown and excretion of drugs from the body.

"Our models of retinoblastoma behave very much like the tumors that occur in children, so our investigation of potential new treatments for this cancer gave us very relevant and useful information, said Nikia Laurie, Ph.D. the paper's first author, who did much of the work on this project.

The two retinoblastoma models the St. Jude team developed were a xenograft model of retinoblastoma, in which a cell line from a tumor removed from a child grew naturally and was available for testing new drug combinations; and a genetic model, in which the team injected a virus carrying a gene that induced from one to five individual tumors in the retina. Importantly, these tumors spread through the blood vessels of the retina, which is particularly important for future studies that could test cancer drugs that work by blocking formation of blood vessels, according to Dyer.

The St. Jude team plans to use their two new models as a "high-throughput" screen to test many new drugs and drug combinations rapidly after determining what doses to use based on culture and pharmacokinetic studies. Treatments that appear to be effective will then be tested in a previously developed "knock-out" laboratory model. This model lacks three critical genes, which causes retinoblastoma. "The knock-out model closely mimics the course of the disease in children, Dyer said. "So it's the perfect final screening tool to identify new treatments likely to benefit children in clinical trials."

Dyer previously reported the development of this knock-out model in a paper that appeared in the June 14, 2004, issue of Cell Cycle. Other authors of the study include Johnathan K. Gray, Jiakun Zhang, Mark Leggas, Mary Relling and Clinton Stewart (St. Jude); and Merrill Egorin (University of Pittsburgh [PA] Cancer Institute).