THE PREECLAMPSIA PUZZLE.

In June, 2000, Ananth Karumanchi, a thirty-one-year-old kidney specialist at Beth Israel Deaconess Medical Center, in Boston, read an article in Nature about preeclampsia, a poorly understood disorder that affects about five per cent of pregnant women. In the developing world, preeclampsia is one of the leading causes of maternal death; it is thought to kill more than seventy-five thousand women each year. In the United States, where treatment is more readily available, few women die of the disease, but complications--including rupture of the liver, kidney failure, hemorrhage, and stroke--can cause lasting health problems. (In rare cases, patients with preeclampsia develop seizures or lapse into a coma; this is called eclampsia.) The only cure is delivery. "If a woman develops preeclampsia near term, then she is induced to have a delivery or undergoes a Cesarean section," Benjamin Sachs, the chief of obstetrics and gynecology at Beth Israel Deaconess, told me. "In most cases, as soon as she is delivered we know she will get better. But, if preeclampsia develops early in the pregnancy, then we have a huge challenge, because we have two patients: the mother and the baby. If you deliver the baby early to spare the mother, then you put the baby at risk for the complications of prematurity; if you wait, then the mother can have severe complications and go on to eclampsia."

Karumanchi had treated several pregnant women with the disorder for hypertension and kidney failure, and he was curious about the disease. The article in Nature described a study by researchers at the University of Reading, in England, who claimed that they had found elevated amounts of a protein called neurokinin-B in the blood of eight women with preeclampsia. The researchers reported that when they injected rats with high doses of neurokinin-B the animals' blood pressure increased. (High blood pressure is a hallmark of the disease.) But the increase was fleeting, and the researchers did not say whether the rats developed other symptoms of preeclampsia, such as protein in the urine and edema (swelling, typically of the face or the limbs). Karumanchi was taken aback. "There was no rationale as to why this protein would be increased in preeclampsia," he told me. "I said to myself, 'Wow! If this paper can make it into Nature' "--arguably the world's premier scientific journal--" 'I am sure we can do better work than that.' "

Karumanchi has a full face and a mop of black hair that is graying at the temples, and he wears aviator glasses. He was born in Mayurnathapuram, a village in southern India, and as soon as he completed his medical degree, at the University of Madras, he sought a residency in the United States, only to discover that positions at prestigious programs are rarely given to foreign-trained doctors. In 1996, after spending three years at Henry Ford Hospital, in inner-city Detroit, he obtained a fellowship to study kidney disease at Beth Israel Deaconess. (I'm on the staff of the hospital, but before writing this article I had never met Karumanchi.) For four years, he worked under Vikas Sukhatme, the chief of nephrology at the hospital, as part of a group of researchers studying the role of angiogenesis--blood-vessel formation--in cancer. But, after reading the article in Nature, he decided that he was more interested in figuring out what caused preeclampsia. He didn't know precisely what he would be looking for, but he was confident that he could obtain tissue from women with preeclampsia, since the disease is so common. "There are placentas being thrown in the waste can every day in the hospital," he told Sukhatme. "So why don't I apply some advanced molecular techniques and try to find what's coming from the placenta that might cause preeclampsia?"

Among medical researchers, obstetrics is often regarded as a dead end. "An enterprising young physician-researcher who seeks to make his name in a field faces huge hurdles if he wants to work with pregnant women," Sachs told me. When a pregnant woman takes a drug or undergoes a medical procedure, her fetus may be affected in ways that are difficult to measure or to predict, and, as Sachs pointed out, a fetus cannot consent to participate in a study. "Our ability to truly understand what goes on in the fetus is poor," he said. "You can't predict physiologically how a fetus is going to respond to some treatment given to the mother. So people are very hesitant to do this kind of research, and the committees that protect human subjects are, by and large, gun-shy." The memory of thalidomide, the sedative that was given to thousands of pregnant women in the nineteen-fifties and sixties and caused severe birth defects--including stunted limbs--still shadows the field, and pregnant women are understandably reluctant to volunteer to ...