In May 2011, Cassandra Caton, an 18-year-old with honey-colored hair and the soft features of a child, suddenly went blind in her right eye. Five months later, an ophthalmologist noticed something disturbing. A large growth in the back of her eye had ripped her retina, destroying her vision.


He sent her to Washington University in St. Louis, a three-hour drive from her sparsely furnished apartment in the working-class town of Sedalia, Mo.

And there, Ms. Caton, mother of a 2-year-old daughter, wife of a chicken factory worker, got almost incomprehensibly bad news. The growth was cancer, a melanoma, and it was so huge it filled her eyeball.
“Am I going to die?” Ms. Caton asked. “Is my baby going to have a mommy in five years?”

It is a question that plagues cancer patients. Doctors try to give survival odds based on a tumor’s appearance and size, but often that is just an educated guess.

But Ms. Caton had a new option, something that became possible only in this new genetic age. She could have a genetic test of her tumor that could reveal her prognosis with uncanny precision. The test identifies one of two gene patterns in eye melanomas. Almost everyone in Class 1 — roughly half of patients — is cured when the tumor is removed. As for those in Class 2, 70 to 80 percent will die within five years. Their cancers will re-emerge as growths in the liver. For them, there is no cure and no way to slow the disease.

No test has ever been so accurate in predicting cancer outcomes, researchers said.

The data from studies of the test are “unbelievably impressive,” said Dr. Michael Birrer, an ovarian cancer specialist at Massachusetts General Hospital. “I would die to have something like that in ovarian cancer.”
While for now the ocular melanoma test is in a class by itself, cancer researchers say it is a taste of what may be coming as they continue to investigate the genes of cancer cells. Similar tests, not always as definitive but nonetheless able to give prognostic information, are under development or starting to be used for other cancers, like cancers of the blood.

Having a prognosis allows people to plan their lives, but most do not want to know if they have a gene for an incurable, fatal illness, like Huntington’s disease or early onset Alzheimer’s.

The eye test raises a similar choice, with an added twist. This is not a test offered to healthy people, but to patients who have just gotten the news that they have cancer. The results will either give them reassurance that they will survive the cancer — or near certainty that they will die from it.

Can patients in the throes of getting this terrifying news really make an informed choice about whether they want the test? Are they able to understand at such a fraught time that, for now at least, there is nothing that can save them if they get the bad prognosis?

Some doctors do not offer the test, reasoning that there is little to be gained.

But other doctors, including J. William Harbour of Washington University, who developed the test (but does not profit from its use), encourage patients to have it. And probably because of the way he describes it, Dr. Harbour says his patients almost always want it.

Ms. Caton was no exception. Without the test, doctors would have had to guess her outcome based on the size of her tumor. And the conventional wisdom is that people with growths as large as hers have a slim chance of surviving. But perhaps, her doctors hoped, the genetic test would come up with a different answer.


Heralding the Future

Dr. Harbour, a genial and burly man with salt-and-pepper hair, has a way about him that relaxes patients, makes them feel everything will be O.K.

“I give them as much information as I think they can handle,” Dr. Harbour says.

And he’s an optimist. The ocular melanoma test is just the beginning, he believes, of a new understanding of that cancer — and perhaps other cancers as well — and why they spread.

About 2,000 people a year, or about 5 percent of melanoma patients, have ocular melanoma, a tumor of the dark brown melanocytes that form a sheet much like a photographer’s backdrop behind the retina. Those with very large tumors are most likely to have a bad prognosis, but patients with small tumors also can have the deadly type.

Often there are no symptoms; the tumor may be discovered by an ophthalmologist during a routine exam. Other patients, though, lose vision or see flashing lights or a sea of floaters in an eye, all signs of damage to the retina as the tumor encroaches.

Most get radiation, a highly radioactive disc placed on the surface of the eye that destroys the tumor in a few days and then is taken out. But those with huge tumors, like Ms. Caton, must have their eye removed.
Ocular melanoma specialists had long noticed that some patients did well and the rest did not, but Dr. Harbour wanted to know why.
This article has been revised to reflect the following correction:
Correction: July 12, 2012

An article on Tuesday about a genetic test that very accurately predicts the outcome of patients with ocular melanoma misspelled the name of a class of drugs that are being tested against a cancer of white blood cells and perhaps could be used to slow the spread of ocular melanoma. The drugs are histone deacetylase inhibitors, not histone deacylase inhibitors. The article also left the incorrect impression that valproic acid is unrelated to that class. Valproic acid, an older drug used to treat epilepsy and other disorders, subsequently was found to be a histone deacetylase inhibitor.

This article has been revised to reflect the following correction:
Correction: July 9, 2012

An earlier version of this article misstated the given name of an ovarian cancer specialist at Massachusetts General Hospital who praised the accuracy of a genetic test to determine the course of ocular melanoma. He is Dr. Michael Birrer, not Matthew. An earlier version also misstated the five-year mortality rate from Class 2 eye melanomas. It is 70 to 80 percent.