INDIANAPOLIS — Traci Johnson believed it was God's plan for her to leave home to attend a tiny Bible college here — and she prayed every day for the Lord to provide for her tuition.

Then an unusual opportunity presented itself.

Eli Lilly & Co., the pharmaceutical giant headquartered a few miles from Indiana Bible College, was seeking healthy subjects for a live-in clinical drug trial. The 19-year-old freshman told her friends back home in Pennsylvania that the study was her best hope to stay in school.

"Trace, that don't sound right," her friend Colleen Jacoby told her. "I never heard of a human guinea pig."

But the students at the Bible college knew all about the trials. They made perfect subjects for studies requiring healthy people — and they were used often, receiving hundreds, even thousands of dollars for a few weeks work.

If accepted into the study, she could make $150 a day for 49 days — more than a year's worth of her school expenses — for taking a drug known as duloxetine, an antidepressant that had already been given to thousands of people and was on the verge of approval by the Food and Drug Administration.

She had faith that God would find a way. "It was in his hands," she wrote in her diary.

Just before the new semester, a Lilly representative called. Her prayers were answered.

A month later, she was dead.

---

Indiana Bible College is on the outskirts of Indianapolis, in a former hospital. With 260 students, it is a tight-knit community of Pentecostals. TV is banned and girls are required to wear long skirts. It's not unusual to see students drop their knapsacks and form an impromptu circle, praying for a sick aunt or alcoholic cousin.

For Johnson it was a sanctuary.

She grew up in blue-collar Bensalem, Pa., but the center of her childhood was a Pentacostal church in a rough Philadelphia neighborhood. Every Wednesday morning, she walked with the pastor past bars and discount stores canopied by train tracks. They huddled around prostitutes and drug addicts. She prayed so hard for them that tears rolled down her cheeks.

Then last summer, she announced that the Lord had told her to attend Bible school.

"She just went wherever the Lord was leading her," said Kathy DePalma, who ran the Christian day-care center where Johnson had worked.

When friends came to visit in Indianapolis, Johnson chattered about the college's football team, her new church and the young men who had caught her eye.

"I think my highest want right now is the person who Jesus wants me with," she noted in her diary soon after starting school, "someone I can pray with."

In a school essay, she envisioned herself as a preacher's wife, raising her children and sitting in the first pew of church.

All she needed was $3,470 a semester. She had arrived with little money, and her father had recently lost his job as a machinist.

"I REALLY REALLY REALLY need you to open a way for me to pay my bill," she wrote to God in her diary. "Please provide a way."

One Final Trial

A few miles from the Bible college, the neon logo of Eli Lilly glows atop the company's headquarters in south Indianapolis.

In 1972, a Lilly biochemist discovered that a patented chemical, fluoxetine, enhanced the action of the brain chemical serotonin, which affects mood. More testing showed the chemical could dissolve feelings of despair and sadness.

The FDA approved the drug, Prozac, in 1987 and since then, sales have totaled more than $21 billion. But by the late 1990s, the patent on Prozac was about to expire, and the company needed a sequel. Lilly began looking at duloxetine, a patented agent that not only affects serotonin, like Prozac, but also norepinephrine, another brain chemical.

Duloxetine had been shelved in the early 1990s, in part because low doses had no effect on depression. But higher doses, Lilly scientists discovered, relieved depression at least as well as Prozac. Subsequent testing proved the drug also curbed stress-related urinary incontinence.

By 2003 Lilly had a trade name, Cymbalta, and industry analysts were projecting sales of $2.5 billion a year for depression alone, a figure rivaling the high-water mark for Prozac.

The drug already had been tested in 8,500 people, but the FDA wanted one last clinical trial to measure its effect on heart rhythm. It would use doses as much as up to five times that recommended for incontinence, and six times the dose for depression.

Lilly needed 100 healthy females between 18 and 75 (women are more prone to incontinence) for seven weeks.

The inventors of Prozac had their sequel. They just had to complete one last trial.

A Great Fundraiser

Lilly's human test clinic, located at the University of Indiana Medical Center, is a resort-like facility with a library, rooftop sundeck and a panoramic view of downtown. "I felt I was on a mini-vacation," reads one testimonial on the clinic's website.

The site touts the drug trials as a great way for schools, churches and community organizations to raise money.

There are hundreds of similar test centers around the country, many of them near college campuses because of the ready supply of students looking for part-time work.

Healthy subjects, free of the ailment for which the drug is designed, are typically used to measure a drug's side effects and health risks. By the time a drug has reached this stage, it has been extensively tested in animals, and the risk of death is considered minimal.

"My test was for a medicine for schizophrenia," said Nasri Ashkar, a 21-year-old senior at Indiana Bible College.

The medicine made him itchy, but "it wasn't a bad experience at all," he said.

Another classmate, 22-year-old Gary Parks, said he had applied to nearly two dozen studies at the clinic and participated in nine. Ticking them off, he quickly lost track: "Something to prevent the spread of cancer … a diabetes one."

Parks made $2,600 in the diabetes study for a week of work, although he recalled that "everybody was throwing up."

Nausea, Parks concluded, was just part of the job.

Before every trial, a Lilly official explained the risks and asked him to sign a consent form, a legal document which protects both the subject and the company. It entitles participants to medical care for health problems arising from the trial, allows them to leave the study at any time and warns of the danger of withholding information from researchers.

Parks said the worst risk he ever heard of during a drug trial was the possibility of fainting from low blood sugar. He wasn't worried. "They have I.V.'s," Parks said, referring to intravenous fluids. "They can bring you back."

Enough students joined the Lilly trials over the years that they became a routine way of making money, like delivering pizzas or parking cars. That worried some members of the Cavalry Tabernacle Church, which all the students attended.

At a dinner with students, one woman told them that "your body is the temple of the Holy Ghost."

The Rev. Paul Mooney, who heads the college and the church, said participating in a Lilly trial was just for "raising pocket money." He said he neither encouraged nor discouraged students from taking part in such trials.

"We're not talking about illegal drugs," Mooney said. "And these programs are very well monitored by the government and so forth, and they have all kinds of guidelines."

An Overdose at 15

Johnson seemed like a good candidate for the duloxetine trial. At 5-foot-4 and 130 pounds, she was physically fit and, by all accounts, reliable and upbeat.

But during her interview, she told screeners that when she was 15, she had landed in the emergency room after overdosing on Tylenol pills and had to have her stomach cleared, according to a psychiatrist studying duloxetine who has read a report on Johnson's case. The doctor spoke on condition of anonymity because the information was considered confidential.

A family friend said the overdose occurred after Johnson's first boyfriend broke up with her, and involved a cholesterol-lowering medication used by her father, not Tylenol.

Johnson denied to screeners she was suicidal then or depressed now, the psychiatrist said.

The researchers accepted her in the study, which was conducted at several sites across the country.

Then they presented her with a consent form that listed side effects common in previous duloxetine trials, including insomnia, nervousness and anxiety. It also noted rarer effects, such as fainting and an occasional feeling of indifference.

She signed the form.

At the time, a controversy was brewing over antidepressants that affect serotonin.

Six months before the trial began, drug maker GlaxoSmithKline sent a letter to doctors in England warning that its drug Seroxat (known as Paxil in the United States) should not be prescribed to people under 18. It reported that in its own pediatric trial, subjects reported side effects including "crying, mood fluctuations, self-harm, suicidal thoughts and attempted suicide" when they stopped taking the drug.

Two months later, drug maker Wyeth cautioned U.S. doctors that its drug Effexor — the FDA-approved antidepressant most similar to duloxetine — increased the risk in teenagers of "suicide-related adverse events such as suicidal thoughts and self-harm."

In December, British health officials effectively banned the use of most antidepressants in children and teens.

Lilly's chief medical officer, Dr. Alan Breier, said in an interview that the company was confident that duloxetine was safe. Previous trials did not reveal a statistically significant number of suicides, Breier said.

The trial's overseers agreed. Dr. Rafat Abonour, head of the University of Indiana board that approved the duloxetine protocol, said he did not recall that suicide was ever mentioned during the review process.

Five suicides had occurred among 4,124 depressed subjects in studies of the drug. One subject had taken only a placebo. In a recent study of about 1,000 depressed people taking duloxetine for up to a year, seven people attempted suicide and seven others reported that they'd thought seriously about it.

Breier said that was less than would be expected in a group of depressed patients. The annual suicide rate in the general population is about one in 10,000. The risk in people diagnosed with depression can be 10 to 30 times higher.

Lilly spokesman David Shaffer said that because data did not link duloxetine to suicide, a history of depression — even a past suicide attempt — would not necessarily disqualify somebody from the trial.

Several people in the trial had struggled with moodiness and despair.

Yullan Valor, a freelance product consultant, was among about two dozen subjects who took part in the duloxetine trial at California Clinical Trials in Glendale. She said screeners seemed interested only in depression suffered during the last seven years. She said they did not ask for details about the depression she said she suffered two decades ago — and so never learned that she had once called a suicide hotline for counseling.

Tatiana Sikic, another participant, didn't tell screeners that five years ago she cut her wrists and took some pills in an attempt to kill herself.

All she said was that she had a history of depression.

"I really needed to be in the study," said Sikic, who was out of work at the time. "And I felt fine."

A Host of Odd Reactions

On Jan. 10, Johnson, along with the Bible college's secretary and another student, entered the enclosed world of Lilly's clinical drug trial.

The experience blends dorm living and medical quarantine. Subjects share meals and TV time, opinions and life stories, forming a makeshift community.

In this study, subjects took duloxetine twice a day. At regular intervals, the medical staff took blood samples and checked heart rhythms.

Each participant took the drug for 20 days: 16 days working up to a dose of 400 milligrams of duloxetine, followed by a four-day weaning period in which the dosage dropped to zero. For the rest of the study, they were given a placebo.

Almost everyone had some odd reaction to the drug. Some could not sleep. Others could not get out of bed. There was constant bickering.

One woman at the Glendale site said she was stunned when she felt an overwhelming urge to run over her husband with the family car as he walked past. Some subjects cut in front of Alzheimer's patients for use of coveted VCRs.

"It was turning into a madhouse," said 38-year-old Carmellia Wright, an actress. "Every minute someone was breaking down crying or laughing."

Two weeks into the experiment, Johnson dropped out of school. Even though she could leave the clinic, she was missing classes and skipped the funeral of her grandfather.

Still, she found time to socialize. And there was a young man, a sophomore, a Christian. "I LOVE his love for you," she wrote to God about the boy in her diary. "God, please work it out."

They quarreled, but on Jan. 28 — a day when she took the maximum 400-milligram dose of duloxetine — they went on a date and parted with a kiss. "So yeah talking about DRAMA!!" she wrote in her diary.

She took 240 milligrams the next day, beginning a withdrawal period when brain chemicals can swing wildly. On Feb. 3, Johnson took 120 milligrams of duloxetine before starting on the placebo.

She seemed fine and baby-sat three nights later, telling one mother she couldn't wait to get back to school. She talked with friends back home and was anxious to be there for the delivery of her sister's baby.

At 3 p.m. the next day she spoke by phone with John Crompton, a church friend from Philadelphia, and told him she felt sick and needed to rest.

Sometime in the next few hours, the young woman took the multicolored scarf that she wore around her waist and looped it around her neck. She tied the other end to a shower rod.

And there she hung, feet dangling close to the floor, until a nurse found her body shortly after 8:30 p.m.

Reassuring Stockholders

The Rev. Joel Barnaby, Johnson's pastor back home in Pennsylvania, broke the news to her parents in their living room.

Reporters barraged Lilly with questions.

At one of the trial sites, in Evansville, Ind., directors shut down the study, sending home all 16 subjects, according to a Lilly spokesman. But enough people remained in the study to ensure the trial was still scientifically valid, the spokesman said.

Lilly assured stockholders that the suicide would not delay the drug's release later this year. The company also reported Johnson's earlier pill-swallowing episode to the FDA and the scientists who were studying duloxetine for Lilly.

While company officials declined to comment in detail on Johnson's death, they said they did not believe the duloxetine contributed to it — and that the reasons behind her suicide were a mystery. She left no note.

At the Glendale test site, clinic staffers tried to ease worries by telling subjects that Johnson had tried to commit suicide before, and that problems with money and other matters had pushed her over the brink, several subjects said.

"The psychiatrist told me that she had a history of depression and that she had just broken up with her boyfriend," Wright said.

Lilly asked test subjects to sign new consent forms and started daily psychological evaluations. It also doubled the weaning period from the drug to eight days. The new forms disclosed the suicide, saying that "at this point the sponsor believes that this event was not caused by duloxetine or the study."

Most of the Glendale subjects stayed in the trial, but they were infuriated to discover in an online news article that they were being paid less than their counterparts in Indiana.

To quell revolt, the clinic raised their pay to match the $150 a day in Indiana.

Convergence of Beliefs

Those closest to Traci Johnson blame her death on the drug — and the lure of money.

"This is a terrible spiritual breach of Christian ethics," Barnaby said. "Christians should never have to experiment with psychotropic medicine as a means to make money."

Johnson believed that faith would protect her — faith that God had led her down a path to $7,000, that other people at the Bible college wouldn't participate in an unsafe study, that a company as huge as Eli Lilly would not let anything happen to her.

Lilly officials believed that their data ruled out a link between duloxetine and suicide, and that it wasn't necessary to tell subjects about the suicide controversy.

Even after studies of antidepressants involving thousands of people, the debate over their risk still rages.

Most psychiatrists say antidepressants are more likely to prevent suicide than trigger it. "Looking at the analysis the British regulators did, I see no significant difference between the drugs and placebo when it comes to suicide risk," said Dr. David Brent, a psychia- trist at the University of Pittsburgh who studies suicide prevention.

But some researchers point out that suicide is inherently difficult to study. It occurs too rarely to provide reliable data, and too many factors, often deeply personal, can spark a plunge into depression.

They say Lilly would have to conduct a thorough investigation of each suicide during the duloxetine trials to determine whether the drug was involved. Company officials, who knew few details about the previous suicides, said they now plan to study all six deaths.

The concerns over antidepressants were already serious enough that on March 22 the FDA warned that some patients could become suicidal when they first started antidepressants or during withdrawal. The agency urged the makers of 10 drugs currently on the market to include labels alerting doctors and consumers to danger signs such as anxiety, hostility and agitation in patients of all ages.

The FDA is still reviewing duloxetine. If the drug wins approval, regulators will likely advise that it, too, carry the warning.

Painful Questions

On a blustery day in February, 17 students rode a bus from Indiana to Pennsylvania with Pastor Mooney for Johnson's funeral.

Among them was the young man from the Bible college who Johnson had been dating. He has refused to talk about her death. The night she died, he had been scheduled to join an unrelated study at the Lilly clinic.

He never showed up.

More than 300 mourners gathered at her old church. "Traci Johnson died last Saturday night, by no fault of her own," Barnaby said from the pulpit.

The worshipers could not reconcile their memories of Johnson with her suicide. "I was surprised because she had the Holy Spirit," said Ernest Copple, 79, an official at the Indiana church Sunday school.

To fathom the possibility that Johnson freely chose suicide would be to challenge the very foundations of her faith. She would never forsake God's most precious gift.

"We know if you take your own life, you don't go to heaven," said Jacoby, her friend from home. "We wouldn't do that."

The day after the suicide, Pastor Mooney invited Dr. Michael Turek, the top Lilly investigator in the Indianapolis duloxetine trial, to church. Students who attended said Turek expressed Lilly's sympathy and answered questions.

One student asked if they should still participate in trials.

That, Turek replied, was up to them.