When Alejandra was a 16-year-old teenager back in 2007, she had the aspirations of any girl her age. She attended an escuela secundaria in Medellín, one of Colombia's largest cities. Schoolwork was interspersed with as many hours as she could squeeze in hanging out with friends at favorite haunts throughout the city.

Then her mother, Yolanda, started to lose her memory. The quiet but conscientious woman would say hello to a visitor and, moments later, would repeat the same greeting again—then again. By her mid-40s, Yolanda had developed early-onset Alzheimer's disease. For Alejandra, it meant that her adolescence had come to an end. Like it or not, she had to take on the chief responsibility of full-time care for her increasingly helpless mother.

Alejandra, now 24, has since moved to the nearby municipality of Copacabana, where she shares a concrete block apartment on an upper floor with her aunt and two uncles, nine-year-old daughter, Luna, and 17-year-old sister, Carolina, who dropped out of high school to help. Her mother can no longer talk nor walk; she spends much of her time bent over in a chair. One of the uncles, age 51, also has dementia.

Every day the girls cook for the two Alzheimer's patients. They feed their mother and Uncle Albeiro through tubes. They wash them by hand and carry them to bed. They repeat the same routine day after day, with no break for birthdays or holidays. “I had hopes,” Alejandra says. “I had plans. I wanted to study. I wanted to be a nurse. So many plans that I haven't been able to follow through with.... I already feel like I'm becoming old.”

Alejandra had a presentiment of her fate long before her mother was formally diagnosed. As a child, she remembers watching her mother tend to her grandmother, who also had Alzheimer's. Many people in this corner of the world share Alejandra's experience. She and her family are among more than 5,000 members in 26 extended families throughout the Colombian province of Antioquia who are at high risk of a rare genetic form of Alzheimer's. Paisa, the label for the mutation that causes the illness, is a regional nickname for the people of Antioquia. Located on chromosome 14, the mutation has been traced back to the time of the conquistadors, in the 16th century. When a copy of the altered gene is inherited from either the mother or father, the child is assured of getting the disease at an early age.

Familial Alzheimer's, as this form of the disease is called, accounts for only about 1 percent of the more than 35 million Alzheimer's cases worldwide. (It has gained a high public profile recently because of Julianne Moore's Oscar-winning role in Still Alice as a woman suffering from early-onset Alzheimer's.) Among the extended families near Medellín, however, it is disturbingly common. The Paisa mutation is present in more than 20 percent of the 5,000-plus members of the 26 families. A carrier will most likely get the disease before age 50.

The high predictability of Alzheimer's among the people of Antioquia has now begun to attract the attention of specialists from around the world. For years repeated attempts to develop a treatment for Alzheimer's have fallen short. Frustrated by seeing drug after drug fail, scientists have come to the conclusion that it may be too late to stem progression of the disease once symptoms have appeared. Instead they have begun to focus on prevention. Rather than treating people who already show signs of dementia, it may be necessary to administer candidate drugs to individuals who are still healthy and then continually check to see whether they stay that way or develop Alzheimer's.

Running such trials in the general population would be time-consuming and expensive because of the difficulty of forecasting when, or if, the disease will develop—uncertainties not shared by the Colombian families. The families who carry the fateful mutation have been recruited for a clinical trial to test a drug to see if it can fend off Alzheimer's. Having lived with the harsh reality of the disease for centuries, they have emerged as a vital link in the search for preventive treatments.

The vision

The shift to an emphasis on prevention in the fight against Alzheimer's builds on the most important advance in research on the disease in recent years. The ongoing clinical trial is deploying magnetic resonance imaging and positron-emission tomography—brain scans that combine with spinal taps to seek out telltale signs of Alzheimer's. Using these technologies, researchers can watch changes occurring in the brain of a person destined to develop dementia sometimes decades before a formal diagnosis of the disease is made.

These new tools can help determine what will happen if a drug, perhaps even one that failed in previous testing, is tried in patients years before they become forgetful. If the brain changes associated with Alzheimer's do not show up on scans—and the patient does not exhibit any cognitive changes—the drug may be helping to stave off the disease. The predictability of Alzheimer's in middle-aged Antioquians with the Paisa mutation has fueled the growing interest in recruiting them for clinical trials to test prevention strategies.

A pivotal figure in this quest is Francisco Lopera, a 63-year-old neurologist who had begun studying the families of Antioquia long before anyone thought they would become important to Alzheimer's research. Lopera spent his teenage years in Yarumal, a town from which many of the families with the Paisa mutation hail. He remembers neighbors from his boyhood who suffered from dementia in middle age. After completing postdoctoral work in Belgium in the late 1980s, he became convinced he could accomplish more as a researcher in Colombia than if he were to stay to pursue a career at a European research institution. By then, he had found the first family with what he suspected was a genetic form of the disease. In 1987 he returned to Colombia, where he took a position as a neurologist at the University of Antioquia and resumed his study of families with early-onset Alzheimer's.

Lopera, today head of the neuroscience group at the university, has built extensive genealogies of the 26 affected families over the decades. His research has presented obstacles that he would not have encountered if he had stayed in Europe. Maintaining routine contact with the various families sometimes required a military convoy to pass through regions where fighting between the Colombian army and guerilla groups persisted. For a time, about 15 years ago, it even became too dangerous to make the trip at all.

The initial curiosity about families with what appeared to be a genetic form of Alzheimer's evolved into a full-fledged research endeavor. Lopera, together with Kenneth S. Kosik, then at Harvard Medical School, Alison Goate of Washington University in St. Louis and other U.S. investigators, pinpointed the exact location of the Paisa mutation on a gene on chromosome 14 in 1995.

The new focus on prevention results from the overwhelming failure rate of potential treatments tested on Alzheimer's patients. Of the 413 clinical trials from 2002 to 2012, more than 99 percent failed. The few that have received regulatory approval can sometimes provide temporary relief from symptoms, but ultimately they cannot stop the loss of memory or other cognitive abilities.

This lack of success has forced drug companies and academic researchers to contemplate clinical trials while patients are still healthy. By the time the earliest symptoms of Alzheimer's appear, current thinking holds, neurons may have already started to die and communication links from one brain cell to another no longer function. When memory problems start to appear, no drug may be able to rescue a patient from the disease.

The 26 extended Colombian families are ideal for studying preventive treatments because the more than 1,000 Paisa mutation carriers among them provide a large enough pool of potential study participants to obtain meaningful results in a clinical trial. And because familial Alzheimer's is so predictable in this group, researchers can work backward by 10 to 15 years to calculate when to start giving a drug to stop the disease.

The existence of these extended families gained enough visibility that it drew interest from the Banner Alzheimer's Institute in Phoenix, which then brought the group to the attention of major drug companies in 2010 to convince them to contemplate prevention trials. Banner partnered with Genentech and the University of Antioquia on a trial of crenezumab, a monoclonal antibody that is intended to bind and help to remove toxic beta-amyloid protein fragments in the brain. The effort has received more than $100 million from Genentech, the Banner Alzheimer's Foundation and the U.S. National Institutes of Health.

The trial

Called the Alzheimer's Prevention Initiative, the trial is unusual in several respects. It is being conducted outside the safe confines of a major medical center in Boston or San Francisco. And neither Lopera nor the University of Antioquia had prior experience with clinical trials of any kind—let alone putting in place the still evolving testing protocols for determining whether a drug is effective in stopping a disease, in some cases 15 years before symptoms appear. “Normally nobody believes that you can do this kind of ambitious project in Latin America,” Lopera says. Banner “believed in us, and that was very fortunate because we've shown that we can work with them in a serious way. And this project got started successfully because of that confidence.”

In late 2013 the research team began administering the drug to study participants, most of whom are in their 30s and 40s. More participants are still being recruited, but the eventual goal is to administer the drug subcutaneously to 100 carriers of the Paisa mutation. (There are also two placebo groups, made up of 100 carriers and 100 noncarriers.) Each subject will undergo a testing period of five years. Brain scans and spinal taps will determine whether the drug has halted the accretion of beta-amyloid. Psychological tests will look for cognitive decline.

If crenezumab was found to alter the course of the disease, it would constitute a breakthrough in Alzheimer's medicine. Trials would then begin to determine if the drug worked in healthy elderly people who do not have the Paisa mutation but whose brain scans show that pathological changes have very recently begun to occur.

The long-standing relationships that Lopera and his colleagues have cultivated with the 26 affected families over the course of decades have been key to recruiting and retaining patients for the trial. Lopera, with his long, gray hair parted down the middle, brings an avuncular presence to his dealings with family members. Last November, when he learned that a few journalists visiting Medellín planned to publish the surnames of patients they had met, he was adamantly opposed even though they had received permission. Lopera explained how doing so might tarnish a family name, even for those who do not carry the mutation. It might then become difficult for family members to get insurance or find a marriage partner.

The Colombian families, their memories of sick relatives stretching back multiple generations, have embraced the trial wholeheartedly. Shehnaaz Suliman, project head for the study at Genentech, expects that trial participants, many of whom must travel long distances to the hospital, are more likely to show up to receive routine drug injections than the average group of enrollees in a study being carried out in the U.S. or Europe.

Her conviction may have been reinforced by Antioquia residents such as Hugo, a 40-year-old who was working last November tending horses for wealthy owners in the town of El Retiro. Hugo makes the 20-mile trip every two weeks to Medellín to receive his treatment: either crenezumab or a placebo. Neither he nor the medical personnel who administer the injection know which it is.

Even if he turns out to be in the placebo group, Hugo says he understands why these clinical trials are important. His father and grandfather died of Alzheimer's, and four of his uncles, two of whom are now deceased, have had it. Hugo remembers how his father, once stricken, would obsessively polish family members' shoes all day and would become frantically anxious if he lost sight of his wife for just a second. “It's hard,” Hugo says. “It's a legacy that comes from very long ago, and you have to face up to it.”

The trials have provided a shred of hope for Hugo and his relatives. “We're doing fine because, with Dr. Lopera, we're hoping that there will be a positive result with these treatments and that there may be a cure for our children.” As he speaks, Hugo sits on a bench in a picnic area near the stables. Dogs bark loudly. He locks one arm with one his nieces, who had taken the arm of her sister at her side.

Despite the burden of the family's medical history, Hugo and several family members who gather one gray November day seem to be a model of psychological resilience. They needle one another about forgetting small details and how that might mark an early sign of the disease. I ask Hugo's sister, Gudiela, age 47, who is also in the trial, if she worries about getting dementia. “Truthfully, no,” she says without missing a beat. “The person who's got to worry is the person who is going to take care of me”—a response that is met with prolonged guffaws.

The first indication of whether the Colombian trial is helping people who carry the Paisa mutation will not come until 2018—the trial ends in 2021. There is no guarantee for crenezumab. It failed in a U.S. clinical trial in patients with mild to moderate Alzheimer's in mid-2014. Further analysis of the results showed that it might have provided some benefit in the early stages of the disease. For just this reason, researchers are moving ahead with the Colombian trial to determine what will happen if the drug is administered long before the first symptoms appear.

Even if crenezumab fails in Colombia, all may not be lost. The trial is the best test to date of the 30-year-old beta-amyloid hypothesis, the idea that toxic protein fragments are the cause of the pathology of Alzheimer's. If the trial does not show any benefit for Hugo, Gudiela and other members of the 26 families—and if other prevention trials just starting in the U.S. are also a bust, the scientific research establishment will have to start turning its attention to possible alternatives to the beta-amyloid hypothesis. They may need to test drugs that attempt to counter the buildup of toxic proteins other than beta-amyloid, and they may also look for chemical agents that protect neurons or deal with the biochemical processes underlying brain inflammation.

Even then, the infrastructure built in Medellín to conduct clinical trials on crenezumab may still serve its original purpose by examining new approaches to prevent the disease. The ability to predict when the disease will occur in a given patient among the Paisa mutation group may continue to be an invaluable asset to drug researchers for future trials. Medellín is now emerging as a major international center for investigating Alzheimer's—and a possible venue for a significant breakthrough—a status that may persist for years to come.

The city may have the needed staying power because the social structure of these families—and the reality of dementia in parents, grandparents, uncles and aunts—will likely keep attracting participants for clinical trials. Family members, moreover, exhibit a natural inclination in caring for the sick, a contrast to the depersonalized medicine too often practiced in the U.S. and Europe. Hugo remembers taking short walks with his father down the block, holding his hand to make sure he did not stray and then be unable to find his way home.

The close-knit family structure ensures that a rural laborer, often accompanied by relatives, will show up reliably at the hospital every two weeks to receive a drug or a brain scan. The curse of the conquistadors could turn out to be a blessing in disguise in discovering a pivotal medical advance that can benefit millions worldwide destined to receive a dreaded diagnosis.