Skip to Main Content

Cindy Bradley-Graziadei almost didn’t take the test. By the time she saw it advertised, in 2017, she was pretty much past the point of mustering any hope. Her pain had changed everything. She’d had to give up working, hiking, long-distance motorcycling. She could no longer trust her grip, and so drank only out of plastic cups. Sometimes her skin hurt too much to put on clothes. There were days when all she could do was lie in bed, eyes protected by tinted windows and blackout shades. When her husband was still alive, he’d sometimes give her a hug, and she’d cry out, “Don’t do that. That hurts.”

She’d seen doctor after doctor, only to be told there was nothing wrong with her, nothing wrong with her, it was all in her head. She’d gone to psychotherapy. She’d taken various classes of drugs. In Atlanta, she visited a private treatment center, which sucked up $16,000 of her savings, 35 vials of her blood, and made her feel worse, not better. In Tallahassee, Fla., she visited a natural healing clinic, which she was skeptical of and mostly showed how desperate she’d become. She was a police investigator by trade, a firm believer in evidence. Hard to say whether she’d given up on Western medicine or vice versa.

Doctors at the Mayo Clinic had eventually diagnosed her with fibromyalgia. But when she talked over this new test with her husband, he’d said, “It’s not going to hurt to try one more time, try something else.”


The kit arrived by mail: The FM/a Test, made by a company called EpicGenetics. “The first and only blood test to accurately and definitively diagnose the disease,” as company reps had written on Facebook. But what had drawn Bradley-Graziadei in was a different part of their spiel. This blood test was a medical gateway, an entrance exam of sorts for a study “to evaluate a direct, effective treatment.” The therapy itself wasn’t new: It was a century-old vaccine, normally given to prevent certain forms of tuberculosis. But this was a brand-new potential use for it. If the FM/a showed you had fibromyalgia, EpicGenetics explained, you could volunteer for a clinical trial at Massachusetts General Hospital. Couldn’t get much more legit than that.

“It’s one of those tests that unfortunately, it made the commercial area before really good solid studies had been done to validate the efficacy.”

Andy Abril, chair of rheumatology at Mayo Clinic in Jacksonville, Fla.

Her blood was drawn at her local hospital and analyzed at EpicGenetics’ lab in Los Angeles. Not long after, she got a letter with her results, confirming that she did indeed have fibromyalgia. Then, in September 2018, the company sent her an email with the subject line “FMa Treatment Begins soon,” explaining that the trial now had Food and Drug Administration approval and was getting ready to enroll. No need to get in touch, wrote EpicGenetics’ director of client services. “We will contact you when we begin the recruitment period. …”

But the invitation never came. Other fibro patients were telling the same story, three years later. Tracy Bercegeay was so excited when she heard the test advertised on SiriusXM radio early in 2021 that she sent in her application right then and there, from the passenger seat, as her husband drove. It was the Mass. General trial that had convinced her. She tested positive, too. But then, no word about the study. “Crickets,” she said.


EpicGenetics, meanwhile, seemed to be doing well. Its website boasted of “processing more than 30,000 patient test applications.” According to the company, 90% of FM/a tests were covered by insurance. The out-of-pocket price was $1,080 a pop. And Mass. General wasn’t the only renowned medical institution involved. In April, the FM/a Facebook page announced ongoing collaborations with Mayo Clinic and Johns Hopkins, among others.

It was only by putting in a call to Mass. General that a potential customer realized something was up. He was a 73-year-old IT guy named Alan Silverman; the suggestion to contact the hospital had come from a friend who works in public health. So, in the spring of 2021, before he got the test, he phoned the research lab where the trial was supposed to take place. The person on the other end informed him — “in a sort of sharp way,” he recalled — that the lab no longer had any association with EpicGenetics. He called the company back, told them not to charge $1,080 to his credit card after all.

The scientist who runs Mass. General’s immunobiology lab can be hard to get on the phone. Call the number listed on her webpage and you’ll get a chipper-sounding employee who will tell you that Dr. Denise Faustman is very busy. No, she can’t really take calls. When, in June, after a couple of tries, she answered an email about the mystery of the fibromyalgia study, her message didn’t provide much clarity: “There is NO trial since we do not have funding so your question does not make sense.”

Cindy Bradley-Graziadei, who got the FM/a Test in the hope of qualifying for the trial at Massachusetts General Hospital. Colin Hackley for STAT

If the trial was mysterious, it was nothing compared to fibromyalgia itself.

Nobody knows what causes it. It can start with a car accident or a work injury, local pain that colonizes the whole body. Or the catalyst might be an infection or an emotional upheaval. Sometimes there’s no clear trigger. The condition can slink alongside autoimmune illnesses, but most experts don’t think it is itself autoimmune. It’s often accompanied by depression, but it isn’t understood as primarily psychological. Its hallmark is inexplicable hurt, bad enough to take over your life but mired in biological murk.

In other words, a perfect habitat for stigma. No coincidence that women with fibro outnumber men two or three to one. As chronic pain advocate Dawn M. Gibson put it, the condition has “a misogynist stink around it” — yet another chapter in the long history of women being disbelieved, their experiences dismissed. The result is a sort of mass medical gaslighting. In the U.S., researchers estimate that the condition affects millions of adults. But it often takes years of searching before someone’s fibro gets diagnosed. Some doctors don’t believe fibro exists at all.

Researchers who’ve devoted their careers to fibro describe it as a central nervous system disorder with some endocrine and immune involvement — a vague, multifarious phenomenon often summarized as “hypersensitization.” If your body is a guitar, fibro is an amp, turned up way too loud. Every accidental scritch of the pick, every finger-graze on a string — it’s all suddenly deafening.

But that doesn’t capture the accompanying frustration. Fibro didn’t get a code in the International Classification of Diseases until the 1990s, and its inclusion didn’t settle anything. The criteria then involved pushing “tender points,” checking whether a normally unpainful stimulus felt extreme. Pressing too hard, though, would hurt anybody. “It was supposed to be the amount of pressure that causes blanching of your own fingernail,” said Sam Whittle, director of the fibromyalgia clinic at the Queen Elizabeth Hospital in Adelaide, Australia. “What does that actually mean? I don’t know.” In The Lancet, two clinicians called the diagnosis “a label so easily abused as to have become meaningless.”

“My mentors at Georgetown told me it would be academic suicide to study fibromyalgia as a career, because there was no good science,” said Dan Clauw, the director of the University of Michigan’s Chronic Pain and Fatigue Research Center. He could only get funding through a historical fluke: Gulf War veterans returning in the 1990s with strange, fibro-like symptoms. That meant he could get Department of Defense backing to slide both soldiers and civilians into brain scanners, looking for hints about what was going wrong.

The disorder has crept toward legitimacy, but the “stink” remains. Tender points have largely been dropped. Instead, to diagnose fibro, a physician asks about a constellation of symptoms, evaluating their number and severity. Have you had widespread pain for at least three months? Have you had cognitive fog? Have you been waking unrefreshed? Some, though, still treat it as a diagnostic trash can, a lazy label for any patient with unexplained pain.

“It was a relief that I didn’t have to start over. I know there’s something wrong with me. My primary care doctor knows there’s something wrong with me. I’m not a malingerer… But to have to start over and look for the answer again was really unappealing.”

Cathy Middlekauff, fibromyalgia patient

So even when patients finally do get a diagnosis, it can feel unsatisfying, vulnerable to disbelief. The resulting care often reinforces the sense that their pain isn’t being taken seriously. They’re told to exercise even though it hurts to move. They’re told that they can learn to live with the pain. “They might encounter a physician who says, ‘OK, we can only deal with three of these 18 issues,” said Kevin Hackshaw, the University of Texas at Austin’s chief of rheumatology.

Many keep searching — for relief, but also for confirmation, validation, some kind of certainty. The idea has enticed so many people to Clauw’s brain imaging studies that he’s started answering the question before they can ask: No, you won’t get your scans back to show your doctor that you have a real disease. It sounds harsh, but to Clauw, the evidence isn’t yet good enough for the technique to yield a diagnosis. Same goes for Hackshaw’s research on blood chemistry.

With certain tests, though, that distinction goes unpoliced. Lab-developed tests, or “home-brew” tests, like the FM/a are done in-house, in a single place. A government inspector makes sure the lab procedures are proper — are they measuring the molecules they say they’re measuring — but not that those molecules are proven to be clinically useful. There are thousands of these tests. The FDA had the discretion to vet them and ensure they work as advertised, but in many cases didn’t. Then, in 2020, the Trump administration took away its authority to regulate these tests at all.

Ask a fibro specialist about the FM/a — as patients and doctors often do — and they’ll say it isn’t nearly as definitive as the company makes it out to be. “It’s one of those tests that unfortunately, it made the commercial area before really good solid studies had been done to validate the efficacy,” said Andy Abril, chair of rheumatology at Mayo Clinic in Jacksonville, Fla., who helped start its fibromyalgia program.

Clauw had a similar take. “I give tons of talks,” he said. “If someone asks me the question, I’ll say, ‘There is no diagnostic test for fibromyalgia.’”

Our world is abuzz with medical misinformation. Compared with bleach-as-cure quackery and credulous claims about the anti-Covid power of horse de-wormer, a diagnostic test with insufficient evidence might seem piddling. But the FM/a’s ties to a prestigious institution tell a different, broader story, one about money in medicine and its noxious side effects.

Nerve breaker

The face of the FM/a Test is a physician named Bruce Gillis. As founder and CEO of EpicGenetics, he’s appeared in the company’s ads wearing a crisp monogrammed white coat. In one, from 2020, he declares the “good news” of his “conclusive” test. When patients or doctors reach out, he often responds himself. His company has gone to bat for customers, to get insurers to cover the cost of the FM/a. He’s busy, but he makes time for interviews. He does this for fibro patients. By his account, he’s one of the few true advocates they have.

It wasn’t always so. The story he tells is a story of conversion, from fibro-skeptic to fibro-apostle — Paul on the road to Damascus, transposed into the oh-so-American world of blood-test entrepreneurs. His heavenly ray of light came as a request from an insurance company.

A little over a decade ago, he was working in private practice in Los Angeles, focusing on occupational medicine. He interrogated the effects of heavy metals and other environmental chemicals, sometimes serving as an expert witness in worker’s compensation cases. He examined people filing for disability. Many with fibro can’t keep working and make insurance claims — and that was why the company came to Gillis: to ask whether these stories of pain and fatigue should be believed, whether fibro was a legitimate disease.

“Being a fibro-skeptic, I thought I would find that it wasn’t,” he said. But he had a faculty appointment at the University of Illinois College of Medicine and teamed up with colleagues there to do a pilot study. Of all the biological machinery that might be implicated in head-to-toe pain, they figured, the immune system was a likely candidate. They did a workup of 17 fibro patients and 17 healthy volunteers. “We thought, for sure we weren’t going to find anything different,” he said. “When we got the results back, we literally fell off our chairs. The patients with fibromyalgia had unequivocal immune system abnormalities.”

That prompted a larger study, comparing 110 people diagnosed with fibro to 91 healthy controls. The researchers took blood, spun it down to separate out the white cells, and then chemically prodded them into producing proteins called cytokines, which act as immunological messengers. The concentrations of most cytokines, the team reported, were lower in the fibro patients’ samples than in those from their fibro-less counterparts. The research won an award from the American Association for Clinical Chemistry in July 2012, and was published as a peer-reviewed paper that December. Within months, the FM/a Test was born.

When Gillis began pitching it, though, doctors weren’t impressed: “What I learned was, my goodness, the bias among health care professionals. … I would say 90% of physicians felt that fibromyalgia is a bogus affliction of neurotic, hypochondriacal, hysterical women.”

They didn’t just dismiss the existence of the disorder, he went on. “In contacting rheumatologists, though they didn’t come out and openly say it, they certainly implied it, that they could not support this test, not because it wasn’t accurate, not because it wasn’t peer-reviewed and published, as it had been,” Gillis said. Rather, in his analysis, they had an economic incentive to disregard his test, because that way they could keep doing thousands of dollars’ worth of other tests to rule out other diseases.

But a dozen clinicians and researchers — all specialists in either fibro or diagnostic test evaluation — told STAT that there isn’t sufficient evidence showing that the test accurately identifies who has fibro. Some suspected there was a different economic incentive at play: The money that EpicGenetics could make selling the FM/a.

Through a lawyer, Gillis rejected the premise that his peer-reviewed research — and the resulting test — could be susceptible to critiques that hadn’t themselves been published in scientific journals. It isn’t the first time it’s happened. In April 2013, soon after the FM/a had hit the market, the trade publication GenomeWeb ran a story with the headline, “Despite researcher criticism, EpicGenetics’ fibromyalgia test does $750K in sales in first month.”

It’s easy to understand why the test has such a strong allure. Many patients have, after all, had unequivocally positive experiences with the company. For some, it has helped convince their doctors that they have a real illness. For others, it had helped convince themselves that their fibro diagnosis is correct, that they don’t have to keep searching for explanations.

“It was a relief that I didn’t have to start over,” said Cathy Middlekauff, a 54-year-old in Hagerstown, Md. “I know there’s something wrong with me. My primary care doctor knows there’s something wrong with me. I’m not a malingerer… But to have to start over and look for the answer again was really unappealing.”

FM/a Test TV ad still
A still from a TV commercial that EpicGenetics released for the FM/a Test in 2020. Screen shot via iSpot
FM/a Test TV ad still
The ad emphasized that the FM/a Test is “definitive.” Screen shot via iSpot

EpicGenetics wasn’t Gillis’ first testing company. Nor was it the first time he’d seen a business opportunity in a contentious ailment.

He’d grown up in Libertyville, Ill., now a Chicago suburb full of pharma employees, but back then a place of corn, soy, and dairy — in Gillis’ words, “a real Podunk town.” His mother’s father had emigrated from Ukraine with a sharp mind for business, and worked his way up to owning department stores. But they were lost in the Depression, and Gillis grew up with one parent selling shoes and the other selling bedding. His older brother got a Ph.D. in education, sparking, as he put it, “the proverbial, ‘my first son didn’t become a doctor, maybe my second son will.’”

He did, in 1974 — and then got a Harvard master’s in public health to boot. But he also inherited his grandfather’s entrepreneurial drive.

One firm he helped lead was focused on environmental health. It evaluated workplace chemical exposures, and in 1989, took part in the cleanup after the Exxon Valdez leaked crude oil into the pristine waters of Alaska’s Prince William Sound.

Another was Cloud Nine, a company that dealt in planes — buying them, selling them, leasing them. Its aircraft flew livers in need of transplantation and customers in need of a private jet.

He was briefly an adviser for a publicly traded company that had pivoted from landscaping to lab-testing cannabis. But the executives never seemed to ask for his advice, and he soon withdrew, as did the CEO, who went back into retirement to focus on writing novels to rescue the reputation of the historical Dracula.

The picture that emerges is of a successful clinician-businessman, whose medical practice was lucrative enough to bankroll his business ventures without any need for other investors. He endowed a financial aid fund at the Harvard school of public health. He financed research in university labs. He flew from L.A. to Boston or Chicago to meet his collaborators. Occasionally, when commercial airlines proved inconvenient, he’d charter one of Cloud Nine’s private planes.

“We realized that to market our test, we’re going to have to do it through the patients.”

Bruce Gillis, EpicGenetics founder and CEO

“He’s really not a scientist,” said Bellur Prabhakar, a professor of immunology and microbiology at the University of Illinois at Chicago, with whom Gillis published two papers. To Prabhakar, he was an M.D. with a solid understanding of advanced biology, the ability to read a scientific paper, and a plethora of ideas, most of which, in Prabhakar’s estimation, wouldn’t work — someone who needed a bench scientist’s help: “He’s a person who thinks of things, that’s for sure. But not all his ideas are going to be testable or provable.”

A dean had introduced Gillis as a well-respected alumnus, and Prabhakar was intrigued by one of his ideas: figuring out whether exposure to a toxic chemical left a telltale molecular imprint in a human cell. Their interactions, while the research was going on, tended to be brief. “Whenever he was in town, he had all kinds of business interests, and I really am not privy to all of them,” Prabhakar recalled. “And he would come and say, ‘Hey, Prab, I’m gonna be in Chicago. Can we can get together for a few minutes?’ And literally, it was always a few minutes. Maybe we had one lunch or one dinner in the entire time I’ve known him. And then he would come to the lab. I think he just wanted to make sure that, you know, he’s not financing something that’s not happening.”

The study was indeed happening, and to Prabhakar, it produced exciting, early-stage results: A specific chemical seemed to predictably shift how certain genes were called into action. The data came from cells in lab dishes, though. “Then you really need to do population-based, extensive studies to validate the data and say, ‘How broadly applicable are these findings to the general population?’” Prabhakar said. But this wasn’t his main area of interest, and he didn’t keep tabs on the testing company that Gillis formed after the papers came out.

The origin story might sound familiar. It, too, had begun with a request from insurers, this one about objectively evaluating injuries attributed to toxic chemicals. So Gillis teamed up with University of Illinois colleagues and published a few papers. From those papers, the Cytokine Institute was born.

Its analyses cost between $6,250 and $12,500 a pop, according to a 2007 article in Business Insurance. Gillis told the journalist that the product could determine with 99.9% accuracy whether someone had been exposed to a substance and whether it had caused injury. Neal Jardine, an attorney who defends companies in these sorts of cases, told the publication about a tire retreader who’d claimed his cancer was caused by chemical exposures on the job, but who dropped his lawsuit after Gillis’ tests showed no link. The technique, Jardine went on, was “objective” and had “the potential to change workers’ comp claims forever.” Gillis told the BBC it had already been used as evidence in 20 cases in California civil court.

Some who worked in the field weren’t convinced. They worried that there weren’t sufficient data to back up the claims the company was making to the press. In a commentary in an occupational health journal, a toxicology professor at the University of California, Berkeley, wrote that it was “scientifically indefensible” to say that one of Gillis’ papers showed a unique molecular or DNA signature. In an American Bar Association newsletter, a lawyer and toxicologist included the Cytokine Institute in a roundup of “un-validated” tests and questioned Gillis’ description. “He was essentially saying that this technique that he had could do more in terms of proving liability or lack of liability than I thought it could,” said Tony Hopp, a Chicago attorney who defends companies in toxic tort cases and co-wrote the piece.

Gillis dismissed these critiques. He told GenomeWeb that the toxicology professor was biased; as was listed in the disclosures published alongside the commentary, he received consulting and expert-witness fees in these sorts of cases from both plaintiffs and defendants. Through a lawyer, Gillis defended the soundness of his work, said that some critiques were based on misapprehensions, and pointed out that his research had been peer-reviewed while there weren’t any peer-reviewed studies that disproved his findings.

Though his research led to legal strategies that are still being used today, he went on, the Cytokine Institute found “a shallow market,” which he said had to do with “incentives driving toxic tort litigation. … In short, uncertainty was advantageous to plaintiffs and defendants, thus definitive testing was unattractive.”

The company ceased operations in 2010. Its most recent record is a court case. The vice president claimed Gillis owed him money; Gillis claimed his employee had frittered away work hours on a personal project, which turned out to be a movie about “an American veterinarian turned gunrunner” on the Caribbean island of Anguilla. The matter was eventually settled.

Dr. Denise Faustman_01
Denise Faustman, the Mass. General diabetes researcher whom Gillis approached about a fibromyalgia study. Barry Chin/The Boston Globe for STAT

When Gillis put in a call a few years ago to Mass. General, he was pitching himself as a philanthropist. That, at least, was Denise Faustman’s understanding. “This guy came in and wanted to donate money to see if trials could get started in fibromyalgia,” she recalled in a phone interview, after a hospital vice president encouraged her to explain her cryptic email.

Faustman knew a thing or two about philanthropists. She’s a diabetes researcher, and she’d staked her career on a hypothesis derided by some of the usual funders of diabetes research. Philanthropists provided an alternate stream of support. One was Lee Iaccoca, the cigar-puffing, tough-talking auto executive, whose first wife died from complications of type 1 diabetes. Another was a self-described ragamuffin of a South African businessman, who’d made his money in digital manufacturing and didn’t want any other parents to lose children to the disease, as he had. In a way, Gillis just seemed like the latest member of the club.

The trial Gillis was proposing was for fibro, not diabetes — but it involved the same tuberculosis vaccine that had caused a stir around Faustman’s research. In 1902, a veterinarian had picked some bacteria off the infected udder of a tubercular cow. Decades of coddling in the lab softened it: The microbe had enough of its old edge to provoke an immune reaction in a healthy infant, but not so much that it would cause an infection. Known as BCG, or Bacille Calmette-Guérin, it was introduced in 1921 and became one of the most widely used vaccines in the world.

But its effects weren’t limited to TB, and Faustman wondered if it could help in type 1 diabetes. The disease involves white blood cells going rogue and attacking the patient’s pancreas. Normally, those blood cells gone berserk would be kept in check by their own colleagues — the immune system’s internal investigatory team — but that self-policing was somehow disabled. BCG could boost the body’s general anti-infection forces; could it also spur the parts of the immune system responsible for reigning in rogue white blood cells?

To say this hypothesis was unpopular is an understatement. Two of Faustman’s colleagues circulated a nasty letter about her, apologizing to patients for the false hope her research would raise. Newspapers covered the kerfuffle, and gave her a lifelong distaste for the word “controversial,” — which might explain her initial hesitancy to get on the phone with a reporter. The usual diabetes funders gave her proposals a pass. She retorted that their lack of interest revealed them as biomedical mercenaries, who had no interest in an old cheap drug that wouldn’t yield much profit.

Faustman’s results were promising, though, and she’s become a kind of doyenne of the non-TB uses for BCG. That was why Gillis approached her. Now that he’d developed a diagnostic tool, he said, patients were begging him to look into a treatment. He’d initially intended on making his own fibromyalgia vaccine using a different mycobacterium, but the Los Angeles public transportation system took his lab by eminent domain to extend a train line. Having an existing mycobacterium vaccine tested out by an existing lab seemed alluring — and Faustman was game. She put Gillis in touch with hospital administrators to work out the financial details. The deal was signed in March 2017. According to court documents, Gillis pledged up to $8.7 million: a first chunk of $580,000, with more to follow if the project got regulatory approval.

But then, after regulatory approval came through late in the summer of 2018, the next payment never arrived. The timing seemed weird: It wasn’t unheard of for a backer to pull out of a project. But it was odd to suspend a trial after you’d spent enough money for the researchers to get the FDA’s go-ahead, but before a single patient had been recruited. Neither Gillis nor the hospital would say how much money he’d given by then. When Faustman contacted him to inquire about the late payments, the hospital said, he gave different responses: First, he was waiting to hear about other business transactions; then, he explained that his focus had shifted away from BCG. She kept in touch, hoping it might shift back.

Tracy Bercegeay
Tracy Bercegeay got the FM/a test after hearing a radio ad early this year and seeing online material about the Mass. General trial. Courtesy
Alan Silverman
Alan Silverman phoned Denise Faustman’s lab and was told it no longer had any association with EpicGenetics. Courtesy

Talk to any expert in study design, and they’ll tell you that a peer-reviewed paper does not a clinical breakthrough make. It’s the most common caveat in biomedical news: nifty, but not ready for prime time. That is the heart of outside experts’ critique of the FM/a. To them, Gillis’ papers are promising, worthwhile, uncovering the start of a path that researchers should keep following — but the studies don’t go far enough to provide anything close to diagnostic certainty.

The test went on the market after publication of the single study in 2012 — and it didn’t include anyone besides fibro patients and healthy controls. That’s a good way to look for breadcrumbs, but doesn’t show that they form a real trail. “The title is a bit misleading. ‘Unique immunologic patterns in fibromyalgia’ — it’s not unique,” said Uma Sharma, who led fibromyalgia clinical trials at Parke-Davis and Pfizer, and is now chief scientific officer of the research firm MMS Holdings. After all, those same cytokine patterns might be present in people with lupus — also characterized by widespread pain and fatigue — they just weren’t included in the study.

In 2015, after two years of FM/a sales, Gillis published a follow-up paper that began to address this issue. It included two additional patient groups, one with lupus, and another with rheumatoid arthritis. In some ways, the data looked encouraging: Of the patients diagnosed with fibro, 93% tested positive, while 11% of healthy volunteers did. “The signal on this is quite strong,” said Jon Deeks, who leads the Test Evaluation Research Group at the University of Birmingham, in England.

But 31% of those with arthritis and 29% of those with lupus also tested positive for fibro. People can have two of these ailments at once — but the researchers had explicitly tried to exclude anyone with that sort of double-diagnosis. “It actually doesn’t look like it distinguishes well” between these three diseases, said Deeks. In the paper itself, Gillis and his co-authors wrote that it was “exploratory and provided preliminary information.”

Researchers who focus on evaluating diagnostic tests also faulted Gillis’ team for a statistical issue. The FM/a is presented as a score — the result of a fiddly bit of math that combines the levels of different cytokines. Gillis’ team had used one dataset of patients to develop the formula behind the score, and then used the very same dataset to test how well that formula works, rather than using data from similar but different patients. That meant the formula might perform better than if it were analyzing data it hadn’t “seen” before. As Qing Pan, a George Washington University biostatistician, explained after reading the paper, “I have a kid taking the college admission exam, I have him doing the exam questions over and over. If he went on to take exactly the same exam, he’s going to score very high. But that doesn’t tell you his real level.”

Or as Jenny Doust, a University of Queensland epidemiologist, put it, “This is clearly going to overestimate the accuracy and predictive values of the test.”

One of Doust’s greatest concerns, though, had to do with an absence. No study has been published on the FM/a that mimics how medical sleuthing happens in the real world. There, patients come into the clinic with a hodgepodge of aches and anxieties and twinges and tingles that may or may not be linked, that might seem like one illness but turn out to be several, that might linger or fade, that might evade description, like a ghost.

To tell if the test is clinically accurate, Doust explained, she’d want to see a study that worked like this: Give everyone who comes in for fibro-like symptoms the FM/a, then have them diagnosed by a clinician who’s unaware of the test result, and then compare the outcomes from both methods.

That way, you could see how well the test works in people whose symptoms qualify them for the FM/a, but whose cases are less clear-cut. The inclusion criteria for the fibro group in Gillis’ 2015 paper were, after all, stringent: diagnosed for at least a year according to the latest criteria, verdict confirmed by two new doctors before the person is deemed eligible. “It’s not that this is a completely bad study. It’s a study that’s typical for the early stages of test evaluation, before you can say anything about how this test or this combination of biomarkers will distinguish between people with and without fibromyalgia,” said Mariska Leeflang, a University of Amsterdam epidemiologist.

Gillis, through his lawyer, rejected these critiques as “a misapprehension of the iterative process of science” and “nonsensical.” The 2012 study, he said, didn’t purport to compare fibro patients to those with other ailments. The 2015 study, he went on, used measurements widely accepted for their accuracy and sample sizes that were statistically significant: “The dataset was thus well-designed and statistically sound to develop a scoring system.” He added that the research “demonstrated … far more of the individuals with fibromyalgia” tested positive on the FM/a than did those with other illnesses, and that if you separate out the cytokine levels that are combined by the test rubric, those individual levels show “a unique immunologic pattern” in fibro patients. Plus, fibro patients are also distinguishable from patients with lupus or rheumatoid arthritis “though differential diagnosis, which may include a multitude of tests for those conditions,” he said.

Gillis acknowledged that the 2012 and 2015 studies are the only ones that have been published evaluating the FM/a, but added that EpicGenetics has continued to refine its analyses since 2015. From his perspective, skepticism about his test is, for the most part, just an extension of fibro-skepticism. “When you tell me there are people who do not support the test, I guarantee you, the vast majority are the same people who don’t support the diagnosis,” he said in an interview.

Some of his critics, though, were the very people who’d risked their reputations to try to solidify fibro as a diagnosis in the first place. Others had no stake in fibro debates. Yet their critiques are technical — and the company’s business strategy had shifted away from rheumatology conferences and toward TV and radio advertising. When Gillis saw just how reluctant doctors were about recommending the FM/a, he said, “we realized that to market our test, we’re going to have to do it through the patients.”

Nerve breaker

By the time Gillis stopped writing checks for the Mass. General trial, in the late summer of 2018, it had already proved useful to EpicGenetics. In late November 2018, the company put out a press release, announcing a 50% increase in requests for the FM/a since Faustman’s trial had gotten approval. In 2020, watchers of “CSI: Miami” saw an ad in which Gillis declared, “With a positive test, you can volunteer for an FDA-approved clinical trial for an investigational new treatment to reverse the disease and eliminate your symptoms.” In 2021, the company was still responding to patient inquiries with an email that mentioned the possibility of participating in the Mass. General trial. Though it was changed after STAT started inquiring about it, in June the EpicGenetics website said that “those who are 18 years and older who are FM/a Test positive are invited to participate” in Faustman’s study.

That surprised Faustman. “The program’s on hold, he knows it’s on hold, so that’s a little bizarre,” she said. She was also taken aback by the company’s claim that if you had a positive FM/a, then you could participate in her study. She was adamant that at no point in planning the trial was an FM/a test a requirement for participating. In order for a study to be at all beneficial, she explained, it had to use widely accepted diagnostic criteria, so that its results would be comparable to other researchers’, so that doctors would know the participants represented their patients. Otherwise, she said, “you’d kind of be laughed at.”

Gillis’ lawyer pushed back in an email to STAT. “If Dr. Faustman communicated that the Mass. General would not require a positive test score on the FM/a Test, she was mistaken,” he wrote. He provided a page from a document related to the trial showing that “cytokine lab results” were part of screening patients’ eligibility for the study. He also wrote that, in a meeting with Gillis this August, two senior researchers at Mass. General “acknowledged that the intent was to use the FM/a Test to determine eligibility.”

Through a spokesperson, the Mass. General senior researchers told STAT that they had been at that meeting but had not said anything about the FM/a being part of the study’s eligibility criteria.

By email, Faustman explained that the cytokine measurements were indeed part of the patient screening for the trial, but that didn’t mean relying on EpicGenetics. “The F/M test is a tool that certainly could be used to assess such levels, but it is not a prerequisite for participation,” she wrote in an email, adding, “Many vendors measure cytokines, and cytokines can also be measured internally in our labs at the MGH.”

Because the funding stopped before recruitment began, to her, the trial remained “kind of imaginary.”

Chris Freeman
Chris Freeman, a fibromyalgia patient in Edmonton, Alberta, has been trying to save money to pay for the FM/a Test. Courtesy
Cathy Middlekauff
Cathy Middlekauff was relieved to get a positive result on the FM/a test: “I know there’s something wrong with me. My primary care doctor knows there’s something wrong with me,” she said. Courtesy

Even in non-Covid times, clinical research is often painfully slow. Recruitment can be hard and funding fickle. Delays and failures are, in a way, normal. The pandemic only made the disruptions worse. Researchers often worried about participants being exposed to the new coronavirus — and Covid-19 is the reason that Gillis invokes most often when asked why he stopped funding the Mass. General trial. But it isn’t the only one. He’s said he realized that patients taking immune-suppressants couldn’t safely get a BCG vaccine, and he didn’t want to exclude them from his work on developing a fibro treatment. He’s said Faustman hadn’t responded to his concerns about other adverse effects. He’s said that Faustman had settled on a sample size that was too small for him. “I have patients by the thousands,” he said — and he wanted the trial to be open to all of them.

Always, he insisted that he was doing what was best for the fibro community. Now, for instance, he said he and his colleagues are doing preclinical research on other treatments that might pose fewer concerns and be accessible to a wider swath of patients, though he declined to give specifics. At one point, as he was being asked to help disentangle the various reasons he’d given for starting to fund a BCG trial and then stopping abruptly, he interrupted.

“You have to be aware of something,” he said. Then his voice became grave. “There’s a significant number of patients who commit suicide who have fibromyalgia, because of the desperation and the chronic, unremitting symptoms,” he said. “I get reminded about that by people weekly, saying they can no longer cope. That’s a major burden.” Because of that, he went on, he’s working as fast as he can.

In April 2021, Bradley-Graziadei was fed up, and she sat down to give EpicGenetics a piece of her mind. “I called your California facility a number of times over the first couple of years and then gave up,” she typed. The company had assured her that everyone who tested positive and wanted to participate in the study would be contacted, she went on. “Here I am at the same address, phone numbers, and e-mail address I’ve had for almost 19 years.”

Her fibromyalgia hadn’t gotten any better; in fact, it had gotten worse. She’d lost her husband to an aortic aneurysm in February 2019, and her mother to Parkinson’s and diabetes complications in July 2020. Bradley-Graziadei’s pain and fatigue were at their most extreme. The trial she’d pinned her hopes on had never materialized. “Thanks for nothing,” she wrote.

Cynthia “Cindy” Bradley-Graziadei
Bradley-Graziadei during her first year as a patrol officer with the Hollywood, Fla., police department. Colin Hackley for STAT

Within a few hours, Gillis wrote back himself: “There is no reason to lose hope. Not only do we have FDA approval for a clinical treatment trial using the BCG vaccine, we are also working on two other Mycobacterium treatment protocols.” Then he explained his concern that fibro might in fact be protective against the immune overreaction that sometimes kills Covid-19 patients. “Therefore, we decided that there was a risk of reversing a person’s fibromyalgia if it meant that we would be increasing their risk for the potentially life threatening effects of a COVID infection. We are consequently working on new treatments that will focus on the symptoms of fibromyalgia.”

It was, word for word, the same message the company had posted in reply to Facebook comments from other angry patients. Included was the link to a peer-reviewed paper Gillis had published with University of Illinois colleagues on fibro and Covid. But nowhere did it say that Gillis had paused the BCG study himself before Covid-19 was known to exist. He’d been using an aborted trial to sell an unproven test to people who were desperate after years of being told their disease was in their heads. With a philanthropic pledge to a prestigious hospital, Gillis may have managed to buy his test a booster pack of credibility.

The same money that can yield evidence and life-altering treatments, it turns out, can also end up eroding trust and deepening patients’ alienation. The medical system hasn’t inspired much faith in the fibro community to begin with. If you’ve spent year after year seeing doctor after doctor, trying test after test and treatment after treatment, an unfulfilled promise can seem less like a one-off and more like yet another data point in a pattern.

As Sam Whittle, the rheumatologist in Adelaide, pointed out, fibromyalgia often keeps people from working, diminishing socioeconomic status. Pain, fatigue, and fibro-fog affect the energy you have to vet possible tests and treatments. It reduces your ability to have a social life, and your support structures begin to fall away. “They are vulnerable to people who will sell them things they can’t afford,” Whittle said.

This isn’t abstract. Chris Freeman, a fibromyalgia patient in Edmonton, Alberta, said that he’s been saving up for the FM/a, in the hope that it will help convince his community his illness is real. But he’s on disability, his wife lost her job during the pandemic, and they have two kids. “When you’ve got to put food on your table and pay bills, that comes first,” he said.

When Bridgett Monville, of Waterford, Wis., realized that the trial wasn’t happening, she and her husband discussed possibly arranging a trip to Mexico. “We talked about going to a country where TB vaccines are given regularly just to try the treatment,” she said. They even got as far as trying to figure out how they’d convince a Mexican doctor to give her a century-year-old vaccine that’s mostly given to newborns. Eventually, they scrapped the plan.

Gillis has other ongoing collaborations — and he says these are yet more signs that his work on the FM/a has been reviewed by others in the medical community and that it passed muster in their eyes. As with Mass. General, the collaborators’ view is more nuanced. Some of them have accepted the FM/a at face value.

“Candidly, I have not, on a real granular level, looked at the data and how closely it correlated, so I won’t posit myself as being an expert on reviewing all that literature, but what I have read is a very compelling story, that it is very useful in fibromyalgia,” said Nathan Goldin, of Virginia Urology Associates, who is collaborating on research with Gillis, when asked about the concerns other researchers had raised about the test.

But clinician-researchers at the Mayo Clinic in Rochester, Minn., had reached out to Gillis specifically because they wanted to do their due diligence before they felt comfortable using the FM/a. They’ve since embarked on a research project to test how well the FM/a works. They wouldn’t have bothered if they thought Gillis’ two papers were worthless, they explained, but they also wanted to check the test’s performance themselves.  “I wanted to see whether or not this blood test holds any water,” explained Arya Mohabbat, a rheumatologist at Mayo and former practice chair of its fibromyalgia clinic. “I just want to know: Is this even worth anything, or is it a waste of time?”

From Bradley-Graziadei’s perspective, that question was answered a long time ago. “It’s almost cruel. It is cruel,” she said. “I’m not saying that this doctor is a bad person. I’m not saying that at all. I just know that I got my hopes up that one last time, and it won’t happen again. Luckily the insurance covered the test, so I didn’t lose any money over it.”

For now, she’s been going to physical therapy and grief counseling. Besides that, she prays. She can’t get down on her knees — that hurts too much — but she talks to God as she goes about her routine. When she’s laying down or sitting or driving, she’ll speak aloud, asking Him to wrap her in His loving arms and protect her, to help her through the pain.

“Dr. Gillis actually responded to me when I wrote that comment: ‘Don’t give up hope,’” she recalled. “Well, right now, I’m just trying to survive every day.”

Graphics by Mike Reddy for STAT

Create a display name to comment

This name will appear with your comment