Refrigerated centrifuges, cell culture laminar flow hoods, incubators, fluorescence and light microscopes, polymerase chain reaction machines, spectrophotometers, an ELISA plate reader, a gel documentation system, a gel dryer, water baths, a shaker, a digital retinal camera and a liquid scintillation counter:This is not your common household wish list. But then again, as its originator, UCSF ocular oncologist Joan O'Brien can attest, beating back eye cancer in children requires some special tools.

The National Institutes of Health apparently agreed. In the summer of 2002, the NIH approved O'Brien's five-year $1.7 million grant for studying new techniques to improve the diagnosis, prognosis and treatment of retinoblastoma, the most common eye cancer in children. Equally important to the grant reviewers were O'Brien's plans to gather evidence on the pathway regulated by the retinoblastoma gene -- a pathway that is one of the most commonly altered routes in all cancers.

Such grand objectives are far from the minds of the 350 children followed in O'Brien's Ocular Oncology clinic at UCSF -- the largest such group in the country -- some of whom are too young to understand the seriousness of their condition. "The tumors that characterize retinoblastoma typically arise between the ages of 1 and 5, so it is usually parents who first notice a chalky spot in their baby's eye or draw their pediatrician's attention to an odd kind of reflective quality, like that in a cat's eye." There are more obvious clues in older children. "Some literally start walking into walls," explains O'Brien, who holds the Pearl and Samuel Kimura Endowed Chair in Ophthalmology.

The good news is that more than 98 percent of all retinoblastoma patients are cured by a combination of chemotherapy and laser treatments, or by radiation alone. Blindness or surgical removal of a cancerous eye is much less common than it was several decades ago, although some vision loss typically does occur. But for those with heritable disease that is only the beginning. Sadly, inherited retinoblastoma, caused by mutations to the retinoblastoma gene that resides on chromosome 13, sets the stage for secondary tumors of the face, bones or brain -- tumors that can be disfiguring, difficult to treat and sometimes fatal. Worse, the very radiation used to stop the initial tumor more than doubles the risk that such a secondary tumor will occur. It also produces facial deformities.

Chemotherapy cuts both ways as well: True, it controls retinal tumors, but at the required six sessions, it is toxic to infants and suppresses the immune system, thereby increasing the risk of infection.

Not surprisingly, finding ways to distinguish between the inherited and non-inherited forms of the disease (the latter has a far better prognosis because secondary tumors are less likely) is one of O'Brien's primary objectives. The sheer size of the gene involved makes the task a daunting one. "The retinoblastoma gene has 180,000 bases and 27 exons (regions that are expressed as proteins), with no known hot spots for mutation," says O'Brien. Today's genetic tests, which must be repeated for all 27 coding areas, are cumbersome. Moreover, results can take anywhere from eight months to two years, assuming any insurance company is willing to pay the cost.

O'Brien, who splits her time between the clinic and her research laboratory, has proposed a different kind of genetic analysis, called a protein truncation test. It is cheap and fast, yielding results in a week or less. "There are other benefits, too," adds O'Brien. "The test can identify carriers of the mutant gene and at-risk siblings and, by allowing us to gather data about particular mutations and correlate it with the severity of the disease in large numbers of patients, really individualize our genetic counseling." Thinking globally, such tests may also help to pinpoint genetic modifiers that may be linked to an increased incidence of retinoblastoma in countries such as India and the Philippines.

Of course, knowing the worst is small consolation if treatments do not improve accordingly. O'Brien and her ophthalmology colleagues are working on that as well. As part of the NIH study, they plan to test different chemotherapy combinations separately and in conjunction with the chemopotentiating agent cyclosporin A, and to advocate for clinical trials should any combination prove itself in the laboratory and in their animal model. Cyclosporin will also be tested singly; preliminary findings suggest that it is directly toxic to retinoblastoma cells. If these findings are borne out in O'Brien's research, cyclosporin could become an even more versatile weapon in the therapist's armamentarium.

That would be welcome news to pediatric cancer specialists, who struggle to balance the long-term health of infants and young children with retinoblastoma's immediate threat -- to say nothing of the parents' guilt for passing along a defective gene to their offspring. "I have been working on this problem ever since the retinoblastoma gene was first discovered in 1986," says O'Brien. "But it's still not easy to sit down with parents and their kids, give them the diagnosis and discuss all the options. I hope that we'll have some better news for them soon."