AlmaGuzman had been in Chicago's Cook County Hospital for almost a week. Her children had brought her to the emergency room because her abdomen was growing, and they thought that at the age of 48 she was too old to be pregnant.

Her evaluation had been intensive. She was not pregnant. A CT scan had shown that her abdomen was filled with fluid. The list of potential causes for such a condition can be long, but her internists had ruled out the easy ones: Alma did not have any of the metabolic disorders that cause cirrhosis of the liver; she did not have long-standing hepatitis, though her blood tests revealed she had been exposed to hepatitis years before; and her heart, kidneys, thyroid, and blood vessels were quite healthy. Plus, her chest X ray was normal. Scanning through her test results on the hospital computer before going to see her, I thought everything seemed fine.

But one glance at her told me otherwise. She looked 58, not 48, her face was pinched and sallow, and her huge abdomen seemed to spill away from her and cover the mattress. She tried to sit up, but the volume of her belly stopped her. She struggled just to turn to me and shake my hand.

We spoke in Spanish. She was depressed because her doctors believed she had ovarian cancer. I had been called in because I'm a gynecologic oncologist, trained to examine and diagnose women with pelvic cancers. Alma was trying to cope with the idea she might be dying. Her answers to my questions were terse, toneless, and barely audible. Her breath was short because of the fluid weighing against her diaphragm. She said she had not lost weight, but while her body had grown, the flesh had fallen away from the rest of her. Her appetite had disappeared over several months, and she had cramps after she ate, although she did not vomit. No one in her family had ever had cancer.

My exam was unrevealing, but even large tumors can remain hidden in a massive abdomen. Although her lab tests were mostly normal, a tumor marker in her blood, known as CA-125, was almost 400, a high level often associated with ovarian cancer.

Her scans were puzzling. Usually, ovarian cancer begins as a mass in the ovary that sheds cells into the abdomen. Fluid accumulates in the abdomen when those cast-off cells choke lymph vessels, preventing normal drainage. But Alma's scan showed no mass. And she didn't fit the profile: Ovarian cancer is a disease more common in developed countries, and until three years ago Alma had lived in a Sonoran village in Mexico. Most women with ovarian cancer have few or no children and thus have ovulated more than women who have carried pregnancies. One theory suggests that when eggs burst through the ovarian surface month after month, mutations can occur as the ovary repairs its surface. Alma was married at 16 and had raised six children, including the four daughters who hovered around me as I worked with her. There was another puzzling aspect of her case: a low-grade fever.

As Alma suspected, a diagnosis of advanced ovarian cancer is usually hopeless. The disease is especially insidious because symptoms are few and vague until the cancer is advanced. The absence of an ovarian mass didn't rule out cancer in this case. Many women who present with findings like Alma's have peritoneal cancer, a tumor related to ovarian cancer that arises from the abdominal lining adjacent to the ovaries rather than from the lining of the ovaries themselves. Peritoneal cancer also can produce a rise in CA-125. The only certain way to establish the diagnosis is to open the abdomen surgically and remove the ovaries and any visible tumor.

Few patients want surgery if they can avoid it. And there was still a possible explanation for her condition that we couldn't rule out yet.

"Have you known anyone with tuberculosis?" I asked Alma.

"¿Como no?" she responded. "In my village, everyone knows somebody with tuberculosis." Her husband had died from it. She had tested positive for TB exposure years before, and she had even taken medicines for it.

"For how long did you take the medicine?" I asked, knowing that even the shortest courses of antituberculosis therapy take months.

"A few weeks," she said, shrugging. "My husband was sick. The children were growing up. There was no money. I felt fine. I've always felt fine, until now."

The residents were intrigued by my suggestion. They drew off quarts of Alma's fluid and sent it for testing. Unfortunately, Mycobacterium tuberculosis, the TB-causing bacterium, takes weeks to grow on special agar plates. And more-sensitive DNA tests for the bacterium take up to two weeks. The old-fashioned microscopic evaluation of the fluid sample failed to turn up either tubercle bacilli or cancer cells. That was not a surprise because both are often too scarce to detect. If Alma had TB, the bacteria could be infecting the peritoneal lining. The resulting irritation would cause fluid to accumulate. Peritoneal tuberculosis can follow a lung infection, so we tried to collect sputum. But Alma wasn't coughing, and even putting a tube into her lung to collect bronchial secretions didn't turn up any tubercle bacilli. A specimen of urine centrifuged to concentrate the sediment also proved negative.

Cancer remained the likely diagnosis, so we scheduled Alma for surgery. Her face clouded over as I described the procedure—not only the incision and the organs we might remove but also the substantial risk in a weakened patient. There are waiting lists for all but emergency surgery at Cook County Hospital, so while we waited for Alma's turn, we tried one last test: a biopsy of the uterine lining. In it, under the microscope, we saw the classic granulomas—nodules filled with immune cells—that are the hallmark of tuberculosis.

In women, pelvic and peritoneal tuberculosis are uncommon infections. They are thought to develop when a TB lung infection allows tubercle bacilli into the blood, and blood-borne bacteria implant in the fallopian tube and uterus. Infertility may be the only symptom, and the infection may progress silently for years.

Fortunately, treatment for peritoneal tuberculosis is less radical than ovarian cancer therapy. Unless the diagnosis is in question, opening the abdomen is not recommended. A peritoneal infection cannot be removed surgically, and members of the operating team may become infected by bacilli in fluid droplets aerosolized during the procedure. Instead, treatment consists of combinations of antituberculosis drugs for a year or more. Alma's therapy began immediately, and her swelling slowly diminished. Now, several months later, she can walk again, and her face has filled in. When she meets me in the hallway on her way back to the medicine clinic, she clutches my hands, her face bearing a smile as broad and beaming as the sun of her homeland.