By the time Kelly Pfaff got home from driving her son to school that morning, it was too late.
Her husband, John, was supposed to be taking their 4-year-old daughter to school. But the girl and the nanny were still at the Pfaffs' house near San Diego. So were John's wallet, cellphone and wedding ring. John was gone.
Kelly was alarmed, but not surprised. For four years, she had watched her husband, once a successful information-technology executive, avid skier and doting father, spiral inexplicably into despair.
It had started with dark, sulking moods. Then he lost interest in sex. His wife asked him if he was having an affair. "No … Something's just not right down there," Kelly said her husband told her. Panic attacks set in.
He suspected the cause might have been Propecia, the popular Merck & Co drug he had been taking to treat hair loss since around the time his problems started. He quit the pills, but still he couldn't sleep, and he flashed random anger at the children. He started talking about killing himself.
On the morning of March 5, 2013, about 45 minutes before his wife got home, John Pfaff stepped onto the railroad tracks a block away and into the path of a southbound Amtrak train. He was killed on impact.
Kelly Pfaff blames Merck for her husband's death at age 40. In a lawsuit filed in 2015, she alleges that the pharmaceuticals company for years knew but concealed from the public that Propecia could cause the persistent sexual dysfunction and depression that led to her husband's suicide about a year after he quit taking the drug.
John Pfaff wasn't the only man who experienced sexual problems after taking Propecia. His widow's lawsuit was one of more than 1,100 filed across the United States and consolidated in so-called multidistrict litigation (MDL) in federal court in Brooklyn, New York. They accuse Merck of not adequately warning patients of the drug's possible side effects and their duration.
Merck has denied the allegations in court filings and declined to comment further on Pfaff's case. In a statement to Reuters, Merck said it "stands behind the safety and efficacy of Propecia," noting that the drug has been prescribed safely to millions of men since the late 1990s. While the drug's label lists erectile dysfunction and other sexual problems as possible side effects among a small percentage of men, the company rejects allegations that Propecia causes those problems to persist after men stop taking it or that it can lead to mental health issues. Merck says the symptoms themselves could be caused by a variety of other factors.
However, confidential documents reviewed by Reuters accuse Merck of exaggerating the drug's safety record.
Citing internal company communications, these legal briefs filed by plaintiffs' lawyers allege that in revisions to the drug's original 1997 label, Merck understated the number of men who experienced sexual symptoms in clinical trials, and how long those symptoms lasted. Other documents show that Merck knew roughly 20 years ago that sales of the drug would suffer if the public became aware of Propecia's possible long- term effects on men's sexual health.
A redacted section of one plaintiffs' motion, reviewed by Reuters, cites correspondence from a Merck executive in which he objected to what he described as "misleading" information about the incidence of sexual dysfunction in men taking Propecia. That information was placed on the drug's label despite his comments, the court document says, and it remains there today.
Merck said that Propecia's label has always accurately reflected data from the company's clinical trials and that it disclosed all data to the U.S. Food and Drug Administration (FDA). Merck also said the executive's "misleading" comment was taken out of context in the court filing.
The documents reviewed by Reuters were filed under seal or heavily redacted by plaintiffs' lawyers – not Pfaff's – in federal court in Brooklyn. Merck had marked the documents confidential when sharing them with the plaintiffs' lawyers, and those lawyers didn't push to file them openly on the public docket. Judge Brian Cogan allowed the medical secrets contained in the documents to be kept out of public view. Reuters is able to report this confidential information now only after discovering filing errors that left some of it exposed.
Still under seal in Cogan's court are the internal Merck documents on which plaintiffs' lawyers based the allegations they made in the legal briefs Reuters reviewed.
Such court-sanctioned secrecy has become the lethal norm in product-liability litigation in the United States. As Reuters reported in June, judges in large product-liability cases routinely seal evidence relevant to public health and safety. As a result, hundreds of thousands of Americans have been killed or seriously injured by allegedly defective products — cars, drugs, guns, medical devices — while evidence that could have alerted consumers and regulators to potential danger remained under seal.
Court secrecy has become pervasive even though, as a matter of law, court records are presumed to be public. Though exceptions can be made when national security, individual privacy or company trade secrets are at stake, the principle of open justice is rooted in American law. In most U.S. jurisdictions, judges are supposed to weigh a litigant's request for secrecy against the broader public interest in being able to see the evidence, and they must explain on record any decision in favor of secrecy.
They rarely do. Secrecy is convenient — for judges concerned about efficiency, for corporate lawyers concerned about protecting their clients' reputations, and for plaintiffs' lawyers seeking speedy resolution for their clients.
The Propecia case in federal court in Brooklyn has followed this familiar pattern. Judge Cogan has, without explanation, allowed Merck and plaintiffs' lawyers to keep information submitted in court confidential.
Some of these documents slipped through cracks in the wall of secrecy. One was inadvertently entered into the public record, staying in the open for a year before being sealed, but in the meantime, it made its way into an obscure public filing, where Reuters found it. The other was faultily redacted, making it possible for this reporter to read it. The contents of both are reported here for the first time, more than two years after the first was filed in Cogan's court.
Had he known the additional information about Merck's clinical trials, Nelson Novick, a dermatology professor at Mount Sinai School of Medicine in New York, said he would have been more cautious when prescribing the drug for young men. "I would have appreciated being apprised of this information earlier," Novick said.
Novick said any medical information about Propecia still being kept secret in court should be unsealed. "It goes without saying that the more information a physician has, the more he can share with the patient and the more informed the decision-making process becomes," Novick said.
Use of the drug remains widespread. It is now sold as Propecia and, since 2013, in generic versions under the chemical name finasteride. Last year alone, finasteride was prescribed for hair loss more than 1.6 million times in the United States, according to healthcare data company IQVIA. U.S. President Donald Trump has taken it, according to summaries of his past two annual physical exams.
From 2009 to 2018, the FDA received about 5,000 reports of sexual side effects or mental health side effects – and in many cases, both – occurring in men who took Propecia. Of those, about 350 reported suicidal thoughts, and about 50 said a patient committed suicide. The data do not show whether the symptoms resolved after stopping the drug.
Pfaff is among a small group of plaintiffs who haven't accepted settlement offers from Merck. She said she is pursuing her lawsuit not just for monetary damages, but also to bring to the public's attention anything Merck knows about mental health problems, particularly suicidal thoughts, that may be associated with Propecia. "If I can save a family from losing their husband, if I could just save one, that would mean the world to me," she said.
Judge Cogan did not respond to questions about why he has allowed the court filings to remain secret.
Merck developed finasteride in the 1980s. The drug reduces a testosterone-related hormone in certain tissues, including skin and the prostate gland. Merck initially obtained FDA approval in 1992 to market finasteride in a five-milligram pill called Proscar to treat enlarged prostate. Five years later, the FDA approved a one-milligram version, which Merck named Propecia, to treat male pattern baldness.
From the beginning, Merck knew that Propecia could cause sexual dysfunction. In three clinical trials Merck conducted prior to receiving FDA approval of Propecia, 3.8% of the hundreds of men who took the drug experienced sexual side effects, including erectile dysfunction or decreased libido.
That result has appeared on Propecia's label since the drug hit the market in 1997. The original label, based on a year of clinical trial results, also said the symptoms resolved in "all men" once they stopped taking the drug.
That's a crucial point for doctors and patients when considering a drug for treating a cosmetic problem. Even a slight risk of sexual dysfunction could give some doctors pause, but they would be reassured if a patient could reverse the problem by no longer taking the pills, said Dr Jerry Avorn, a professor at Harvard Medical School.
The distinction was also important to Merck, according to a sealed document Reuters was able to review because it was filed publicly elsewhere. A 1999 internal marketing study cited in the document found that 40% of men who had heard of Propecia were aware of potential sexual side effects, and that such knowledge would prevent half of those men from taking it. The court document does not say how many men were surveyed.
The sealed document also cites a 2016 sworn deposition in which a plaintiffs' lawyer asked former Merck marketing vice president Paul Howes: "So you knew internally that if these sexual adverse events were prolonged or lengthened or never went away, that that would be something that would impact sales in a negative way. Right?"
"Yes," Howes said.
Howes, who held that marketing job from 1998 until 2001, declined to comment for this article.
In 2002, more than four years after Propecia hit the market, Merck revised the drug's label to reflect the results of a study that expanded upon two of the drug's three original clinical trials. The original trials were one-year studies; the extended trials covered four additional years. All were conducted by Merck in the mid- to late 1990s.
The revised label retained the original label's statement that 3.8% of the 945 men taking Propecia in the first year of research experienced sexual side effects. But the company added that, among the 323 men who took the drug for all five years of the extended study, "the incidence of each type [of sexual side effect] decreased to no more than 0.3% by the fifth year of treatment."
A 60-page motion plaintiffs' lawyers filed in Cogan's court alleges that Merck's 0.3% figure on the revised label underreports the number of men who experienced sexual side effects during the extended study. The passages of the motion pertaining to the study were redacted. However, Reuters was able to view the blacked-out material after copying it from a digital version of the motion and pasting it into a document in a different format.
The motion cites an email in which a Merck analyst tells executives that 23 men taking Propecia experienced sexual side effects during the final three years of the expanded study. That phase comprised 922 men who took the drug for varying lengths of time, according to a 2002 article Merck published in the European Journal of Dermatology.
The label, as revised in 2002, omits the experiences of nearly all of those men, reporting only on sexual dysfunction in men who took Propecia in the first year of research and in those who took it continuously for all five years. Merck didn't include the experiences of men who finished the study before the fifth year or who were given placebo doses earlier in the study. The revised label also omitted information about six men who dropped out of the study during the final three years due to sexual side effects.
It is impossible from the numbers Merck has published to tell what percentage of men experienced sexual dysfunction over the full five years.
It is "not fair or standard practice" to report results from only a portion of a multi-year study because it systematically underestimates the risk of any side effect, said Avorn, the Harvard pharmacoepidemiologist who reviewed the court filing and Merck's statements at Reuters' request.
Before the 2002 label change, Dr Keith Kaufman, clinical head of Propecia, discussed different ways of interpreting clinical trial data, the plaintiffs' court filing says. The 0.3% figure is "totally misleading" because, by the fifth year, "you have weeded out the dropouts with the sexual [adverse experiences]," the motion quotes him as saying. The legal filing does not specify who received Kaufman's correspondence.
Kaufman referred Reuters' questions to Merck, which said his remarks were unrelated to the language on the label. Merck said that he was referring to a flawed report submitted by an outside expert and that the company settled the litigation before it could provide proper context for Kaufman's comments.
Plaintiffs' lawyers filed Kaufman's original correspondence in court under seal, and it thus cannot be viewed. Merck declined to waive its claims of confidentiality to share that filing and the Merck analyst's memo with Reuters.
The redacted motion argues that Merck not only underreported the number of men who experienced sexual dysfunction while taking Propecia, but also concealed the duration of those problems. Citing the Merck analyst's memo, the motion says that of the 23 study subjects who experienced sexual side effects during the last three years of the study, seven continued to experience symptoms when they completed it. The symptoms of nine others resolved after they stopped taking the drug, but the time it took was unknown, the motion says. And of the six men who dropped out of the study, it says, one still had symptoms at least 66 days after stopping treatment.
Merck didn't know if some of these men's symptoms resolved, Kaufman said in a deposition.
None of this data described in the plaintiffs' motion was included on the revised 2002 label. Instead, Merck made a small but significant change to the label's language: Symptoms stopped in "men" who went off the pills, the label now said, rather than "all men."
Merck told Reuters that it followed up with patients who dropped out "in accordance with the protocols for the study." It said it has "consistently maintained that the available evidence does not establish that finasteride … causes sexual dysfunction which persists after drug discontinuation."
Merck said it provided its five-year trial data to the FDA, which approved the revised label.
The FDA declined to answer questions about what Merck shared with the agency or how it evaluated specific information the company submitted about its Propecia clinical trials. It said it "takes very seriously its role in continuing to monitor and regulate the safety of drugs both before and once they are on the market."
Dr Michael Irwig, an endocrinology professor at George Washington University who has studied Propecia, reviewed the faultily redacted court filing for Reuters. He said Merck's numbers "look much better" by excluding men who dropped out, and the difference reflects an overall lack of transparency regarding subjects who experienced sexual side effects, he said.
In 2008, after Swedish regulators investigated reports that sexual side effects continued in men after they stopped taking the drug, Merck changed Propecia's label in that country to warn that erectile dysfunction had been reported to persist after stopping the drug.
The same year, Kelly Pfaff told her husband about Propecia. John's hair had started to thin on top. "For whatever reason, it bothered him," Kelly Pfaff said. She said she encouraged him to look into Propecia because she knew a friend was taking it and had hair "like a Chia Pet." She now regrets her advice. "I have to sleep with that every night," she said.
The label prominently warned pregnant women to avoid handling the drug. Kelly was pregnant with their daughter at the time, and she said John warned her not to touch the pills. She said he didn't mention any other risks.
Reports of depression in men taking Propecia were added to the U.S. label in 2010. This disclosure appeared in the "Postmarketing" section of the label, below the much more favorable description of Merck's clinical trial results added in 2002.
Meanwhile, the FDA had begun an inquiry after Merck changed its label in European countries to warn about persistent sexual dysfunction. In April 2012, the agency approved another change to the U.S. product label. The agency said at the time that "clear causal links" between the drug and sexual dysfunction had not been established. But, for the first time, the Propecia label acknowledged reports that a range of men's sexual problems persisted after the men stopped taking the drug.
A patient advocacy group, the Post Finasteride Syndrome Foundation, petitioned the FDA in 2017 to either withdraw Propecia from the market or add a black box warning for sexual and mental health side effects. The FDA has not yet responded to the request.
The 2012 label change came as John Pfaff was hurtling toward his final crisis. He quit Propecia that spring, four years after starting treatment. Kelly Pfaff said her husband's decision to go off the drug wasn't prompted by Merck's label change, but more a process of elimination of possible culprits. She and her husband entered therapy together.
In a rambling email to his colleagues, he abruptly quit his job in January 2013. Two months later, he was dead.
Kelly Pfaff's lawsuit cites a peer reviewed study by Irwig, the academic endocrinologist, that found overlap between persistent sexual dysfunction and mental illness among 61 former finasteride users. In the study, 64% of men who experienced sexual problems after they stopped taking the drug also demonstrated moderate-to-severe depression, compared to none of a control group of men with hair loss who hadn't taken the drug. Forty-four percent reported having thoughts of suicide, compared to none in the control group.
Irwig found these rates to be higher compared to other studies of depression among men with sexual problems. He acknowledged he did not have specific comparative data on bald men experiencing sexual dysfunction. Other published studies have found that Propecia can lead to decreased levels of neurosteroids that act as natural antidepressants.
Some dermatologists have questioned whether selection bias tainted Irwig's findings and suggested that publicity and lawsuits could be artificially inflating the number of people reporting Propecia side effects. Merck said the study is "flawed" due to several factors, including a small sample size.
Nicole Rogers, a dermatology professor at Tulane University, said she counsels men on the possibility of side effects and tells them harmful impacts will go away if they stop."For people who have a long family history of [hair loss], who are we to tell them they can't do it?" Rogers said.
After reviewing the information Reuters found on the Merck clinical trials, Rogers said the sexual problems the men reported could have been caused by other factors, including smoking.
Rogers's assessment echoed Merck's position. "Premature hair loss itself, the very condition for which Propecia is prescribed, is associated with low self-esteem, poor body image, and depression," Merck's lawyers wrote in a 2017 court filing. "Rather than attribute their sexual difficulties to the common reasons why young men with premature hair loss experience these problems, Plaintiffs instead assign blame to a drug that, once discontinued, is no longer pharmacologically active in the body."
Sales of Propecia climbed steadily through the 2000s, peaking at $447 million in 2010. Soon after, Merck's patent expired, but use of finasteride remained strong as cheaper generic versions hit the market.
By then, hundreds of men were suing Merck over Propecia. In 2012, their lawsuits were consolidated before Judge Cogan.
As part of the discovery process, the early phase of litigation when opposing sides request information from each other, the court issued a routine protective order, allowing each side to designate discovery material as confidential before sharing it. The plaintiffs then filed a small handful of those documents in court under seal, citing the protective order as justification for the secrecy.
Once evidence hits the courthouse, however, appellate courts have ruled that judges are required to conduct their own analysis of whether the secrecy outweighs the public interest in transparency. Cogan never did so.
In 2016, Cogan selected four of the Propecia lawsuits to go to trial first. The proceedings marked the first time that some of the evidence described in this article might have been publicly aired.
But Merck settled hundreds of the Propecia lawsuits in April 2018. The first trials were canceled, and the information remained under seal.
Merck agreed to pay a lump sum of $4.3 million, according to a Merck securities filing, to be divided among plaintiffs who settled. In exchange, plaintiffs agreed to drop their claims against the company and not to discuss any confidential court documents, including any that were "inadvertently" filed in public and later placed under seal, according to a copy of one of the settlement agreements reviewed by Reuters.
Fewer than 25 U.S. cases remain pending, including Pfaff's. Her lawyers signed the protective order and are now reviewing some of the same discovery material Merck provided to other plaintiffs' lawyers. She says she hopes the case will give the public a full view of what Merck knows about Propecia.
Pfaff, who has moved with her children to Park City, Utah, said she thinks her husband would still be alive had he known how long Propecia's possible side effects could persist. "We do crazy things for vanity. But if we are going to do that, we should know what the risks are," Pfaff said. "I'm pretty sure if John knew he would poison his body, he would be bald."