With any serious disease, it's obviously a good idea to find the best doctor - and the best hospital - you can.

But with ovarian cancer, a rare disease that strikes 25,000 women a year, kills nearly 15,000, and is almost impossible to detect early - it's absolutely essential.

That's because there are often no symptoms in the early stages. In three-quarters of cases, by the time ovarian cancer is diagnosed, it's already spread. Currently, only half of women diagnosed with it are alive five years later.

All of this means it's crucial to get state-of-the-art chemotherapy and specialized surgery, not from a general surgeon or gynecologist, but from a gynecological cancer surgeon who knows how to probe every inch of the abdominal cavity for tiny tumors.

In ovarian cancer, the primary tumor usually, though not always, begins in the ovary itself. It then spreads quickly throughout the abdomen, scattering mini-tumors all over - on the colon, the spleen, the gallbladder, the diaphragm. Even the walls of the peritoneum - the Saran-wrap like tissue that covers all the internal organs - are studded with sprouting tumors.

"When we see inside the abdomen, it's like DOTS candies," says Dr. Linda Duska, a gynecologic cancer surgeon at Massachusetts General Hospital. "It's not just a mass in the ovary, it's diffuse miliary cancer - little, teeny things everywhere." Research shows that the more thorough this initial surgery - which involves a long, vertical incision and can take several hours - the better a woman's chances of survival.

Historically, those odds have been grim. If the cancer is caught early, while the tumor is confined to the ovary, the 5-year survival rate is more than 90 percent. But few cases are caught early because there's still no good screening test. Researchers are working on new tests, including a blood test called LPA.

There is already a blood test for a protein called CA125 that can detect some tumors, but it's notoriously unreliable. It misses some cases and suggests cancer is present when it's not. Ultrasound can spot some cancers, but it, too, raises many false alarms. Even when these two tests are combined with a standard pelvic exam, ovarian cancer is so hard to differentiate from benign cysts on the ovary that 30 women with suspicious findings may be sent to the operating room for every cancer found.

Barbara O'Brien, 54, an Arlington woman, is one of the lucky ones. She was diagnosed three years ago when her cancer was in the earliest stage. But she says she's "one of the few in my support group" whose cancers were caught this early.

If cancer isn't caught until after it has spread to the fallopian tubes, the 5-year survival rate drops from 90 percent to 80 percent. If it has spread to the lymph nodes and abdomen, it drops to 30 percent. Even when symptoms - abdominal swelling, bloating, vague abdominal and pelvic pain, gas - are present, they are so non-specific, a doctor may not suspect ovarian cancer.

Better chemotherapy drugs, however, and equally important, a much better understanding of how best to combine and administer them, are beginning to make a dent in those numbers.

There's no data yet showing that bone marrow transplantion is more effective than standard chemotherapy. But there are several studies showing that giving chemotherapy intraperitoneally - through a tube into the abdomen, instead of through intravenous injections into the bloodstream - may yield some improvement in survival.

The advantage is that the drugs get directly to the tumor, cause less nerve and marrow damage, and trigger fewer side effects. The downside is this treatment can cause severe abdominal pain and may not work against tiny tumors that travel through the circulation to other areas of the body.

Another emerging strategy is to try new drugs early, instead of waiting until a relapse, as is traditionally done. "The hope is that by utilizing more of the new, active agents in ovarian cancer right at the beginning, it may result in more effective killing of tumors and potentially prolong survival," says Dr. Ross Berkowitz, co-director of the Gillette Center for Women's Cancers at Dana-Farber Cancer Institute.

Doctors are also finding new ways to combine drugs so that they attack the tumor through different biochemical pathways and don't exacerbate each other's side effects. "The concept of chemotherapy that works in different ways is critically important," says Dr. Stephen A. Cannistra, program director of gynecological medical oncology at Beth Israel Deaconess Medical Center in Boston.

For instance, platinum-based drugs - either cisplatin or carboplatin - have long been the mainstay of treatment. The drugs insert themselves into DNA and interfere with its replication. Seven out of 10 tumors can be shrunk this way, but the drugs kill only tumor cells that are sensitive to them, and many aren't.

Adding Taxol to platinum drugs yields significantly better survival, notes Dr. Edward Trimble, a specialist at the National Cancer Institute. In part, that's because Taxol works differently, by binding to a cellular structure called tubulin. When it binds, the chromosomes can't pull apart and the cell can't divide.

Another relatively new drug called Hycamtin (topotecan) works in yet another way, by blocking an enzyme called topoisomerase-1, without which DNA can't unwind and the cell can't divide.

Already approved for women whose ovarian cancer has recurred, Hycamtim is now being studied as a first-line treatment. A drug called Doxil works similarly, by inhibiting an enzyme called topoisomerase-2.

Still another emerging strategy is to borrow chemotherapy drugs from other types of cancer. A pancreatic cancer drug called Gemcitabine, for instance, shows enough promise against ovarian tumors that doctors are now designing studies to test it in newly-diagnosed women. Doctors are also trying a lung cancer drug called Navelbine for women with recurrent tumors.

An even more high-tech solution was reported recently by Tayyaba Hasan, a biochemist at the Massachusetts General Hospital Laser Center, and others, in the Journal of the National Cancer Institute. Hasan's team studied ovarian cancer that was resistant to cisplatin.

The researchers hooked together a drug called a monocloncal antibody (designed to find its way to markers on ovarian cancer cells) with a light-sensitive molecule called a chlorin.

A laser light activates the chlorin, which then destroys the cancer cells - but not normal cells - in the immediate area. "The exciting finding," says Hasan, is that this approach was 13 times more effective than standard chemotherapy alone. Other researchers, including a team at the University of Pennsylvania, are pursuing a similar approach.

And even that's just the beginning. In a collaborative effort, researchers at the Dana-Farber, MGH, and Brigham and Women's Hospital are freezing bits of ovarian cancer tissue in hopes of making individually-tailored vaccines. The idea is to kill the cells, insert genes that make an immune-boosting protein called GM-CSF, then re-inject the cells back into the patient.

Other researchers are trying gene therapy to beef up production of cancer-fighting proteins produced by a gene called p53. Still others are working on SERMS, or selective estrogen receptor modulators, to block hormonally-driven cancers. And others, including researchers at New England Medical Center, are conducting trials of a monoclonal antibody called OvaRex to help the immune system attack ovarian cancer cells.

There is no question that ovarian cancer is still a horribly stubborn disease. But the research is beginning to pay off.

Carolyn Mostecki, 54, a professional gardener in Gloucester, appears to be in remission after six years of treatment. She took an experimental drug called Taxotere, a cousin of Taxol, but thinks Tibetan herbs have helped, too.

Alice Rouff, 60, a restaurant hostess from Ashland who was diagnosed 10 years ago, is also optimistic. "I'm totally fine now," she says, though she's scared to use the word "cure."

"And every day, I make a good day."

SIDEBAR:

FIGURING A WOMAN'S RISK FOR CANCER OF THE OVARIES

There are no definitive ways to prevent ovarian cancer, but some factors may reduce or increase risk.

Last week, for instance, Italian researchers reported in the Journal of the National Cancer Institute that taking a drug related to vitamin A (fenretinide) may protect against ovarian cancer, in part by triggering apoptosis, or cell death.

In general, scientists believe that the more ovulatory cycles a woman has in her life, the greater the risk, and the fewer cycles she has - whether they are interrupted by pregnancy, birth control pills or breastfeeding - the lower the risk.

Every time a woman ovulates, there is microscopic damage to the surface of the ovary where the egg pops out. Usually, this damage is quickly healed, but cells must work overtime to repair it. During this repair, researchers theorize, there is an increased risk of genetic mutations, which may lead to cancer.

Studies show that a full-term pregnancy, during which there is no ovulation, reduces the ovarian cancer risk by 50 percent; subsequent pregancies offer additional protection. Breast feeding, which can also inhibit ovulation, reduces ovarian cancer risk, too, but this data is less convincing.

There is good evidence, though, that oral contraceptives, which keep the pituitary gland from triggering ovulation, decrease ovarian cancer risk by about 50 percent if they're taken for a total of five years, not necessarily continuously.

And what of the interplay between infertility and ovarian cancer? That's dicey. If infertility means there's no full-term pregnancy, that increases risk just as it would in a fertile woman who never had a baby. On the other hand, if infertility is caused by lack of ovulation, as it can be, that could reduce risk, though this hasn't been proved.

Fertility drugs such as Clomid and Pergonal have been suspected in some cases of ovarian cancer. Some evidence suggests that ovulation-inducing drugs may increase risk, particularly if the drugs don't work and a woman never gets the risk-reducing benefits of pregnancy. On the other hand, a recent California study found no such association.

Curiously, tubal ligation, in which a woman's fallopian tubes are tied to prevent eggs from getting from the ovaries to the uterus, may reduce the risk of ovarian cancer by 30 percent, for unclear reasons. One theory is that ligation blocks potentially carcingenic substances from travelling from the vagina, cervix or uterus up to the ovaries.

In support of this, scientists point to several studies suggesting that talcum powder, which some women put on diaphragms or on genital skin, can raise ovarian cancer risk.

Some women, including some Ashkenazi Jews, also have mutations in BRCA1 and BRCA2 genes that increase risk of both breast and ovarian cancer. In general, women have about a 1.4 percent chance of getting ovarian cancer over a lifetime; in women with one close family member who has had the disease, the risk rises to 5 percent. If more close family members are affected, it rises more.< Doctors suggest that women who test positive for the mutations or have a strong family history of the disease consider having their ovaries removed surgically in hopes of preventing ovarian cancer. Even after such surgery, however, it may be possible to develop a related cancer in the peritoneum, the tissue that lines the abdomen.