In early May several hundred investors, doctors and brokerage research analysts attended a dinner presentation after cocktails offered by the leadership of Myriad Genetics in Manhattan’s midtown. Salt Lake City–based Myriad, best known for its hereditary cancer tests, was in New York to tout new research on its increasingly popular GeneSight product during the American Psychiatric Association’s annual conference.
The APA conference is an important event for Wall Street as well as pharmaceutical companies because of the massive amount of money Americans spend each year on drugs and therapy for treating depression. A March 2017 American Psychological Association article estimated the annual cost of treating depressive disorders at $71 billion and rising; in May Myriad said the total cost of major depressive disorder was $100 billion a year. Thus, taking notice of the latest drug development news could potentially be very lucrative for companies and investors alike.
And following Myriad’s $225 million purchase of GeneSight’s developer Assurex Health in August 2016, managers had a story to tell about their newly acquired diagnostic test, which uses a patient’s genetic profile to guide a psychiatrist in selecting an antidepressant.
GeneSight assesses 12 different genes in a patient to rank medications according to their prospective usefulness in treating the person’s clinical depression. It is but one entry in the emerging field of pharmacogenomics, which is concerned with how someone’s genetics can affect his or her response to drugs. And GeneSight fits into a broader movement that has emerged over the past decade called “personalized medicine” or “precision medicine,” aimed at taking into account individual variation in genetic information, lifestyle and environmental factors when considering treatment options.
So as Myriad CEO Mark Capone sat with brokerage analysts at a table, Dr. John Greden, who directed a nearly six-month-long randomized clinical trial on GeneSight, crisply presented what the company touted as “landmark” research. Greden, the executive director of the University of Michigan Comprehensive Depression Center, was warmly received, especially when he emphasized data about patients’ response to GeneSight and the remission of major depressive disorder.
Adding to GeneSight’s cachet was an Oct. 1, 2015, local coverage decision by Centers for Medicare and Medicaid Services contractor Palmetto GBA guaranteeing full reimbursement of the test’s $2,000 price for Medicare patients. And if the clinical trial concluded with results showing the test’s efficacy, commercial health care plans were expected to quickly start covering the cost of GeneSight for their members.
The day after Myriad reported that its quarterly earnings had greatly improved from the period a year prior, the company’s market capitalization grew more than $319 million, when its stock price surged $4.57.
In many ways the Manhattan presentation was the perfect event, pleasing prescribers and investors alike.
Apart from a few tiny clues, nothing would have indicated to attendees that the event was just for show, a facsimile of how a major life sciences company proceeds when discussing vital research.
If Myriad had truly wanted the medical community to grapple with its research, though, it would have secured a formal slot at the conference to present Dr. Greden’s findings and take questions from guests or researchers without ties to the company.
Instead Myriad held an off-site “satellite symposium,” for an invitation-only audience largely composed of psychiatrists and primary care doctors in private practice who prescribed GeneSight as well as analysts and bankers seeking underwriting and advisory business with Myriad.
Myriad’s staging of the May session was clever in that executives could claim to have shared the trial’s findings at an APA conference, though it was just at a poster presentation, where Dr. Greden talked to whoever walked by. This is scarcely what one would expect after a study of a genetic test that a public company has repeatedly hailed as a “landmark” achievement.
The reason for the dodge might be that from a scientific standpoint, GeneSight is a bust (although it is Myriad’s most important source of sales growth.) In an 1,167-patient trial whose findings the company announced on Nov. 2, 2017, (and published Jan. 4, 2019, in the Journal of Psychiatric Research), GeneSight failed to meet its primary endpoint of demonstrating superiority over established treatments and did not achieve 23 of its 25 secondary endpoints. The primary endpoint in a clinical trial is “the main result that is measured at the end of a study to see if a given treatment worked,” according to the National Cancer Institute.
The patients whose doctors used GeneSight reported a 27.2 percent reduction in symptoms of depression; in contrast, practitioners noted a 24.4 percent reduction for those who received treatment as usual.
Though a 2.8 percent point improvement may seem (narrowly) promising for GeneSight, for psychiatrists considering treating depression among large groups of patients, this difference is statistically indistinguishable from treatment as usual; the study’s p-value was .107. In most medical trials, 5 percent is considered the cutoff for significance. After factoring in the possibility of random or experimental error, no doctor could be certain that any benefit comes from using GeneSight.
So it was a lot safer for Myriad’s stock price that Dr. Greden and company executives spoke to a carefully curated audience about why positive results for two secondary endpoints were more significant than GeneSight’s failure to attain its primary endpoint.
Little of this might be obvious to investors and analysts because Myriad has kicked up enough dust over the years that they appear to have stopped demanding straight answers.
Then again, it’s hard to blame them since Myriad’s revelation that GeneSight’s randomized controlled trial went poorly appeared only in the last paragraph of its November 2017 press release — in a discussion of $185 million in clinical milestone payments owed Assurex’s former shareholders: “That clinical trial milestone payment will not be due because this endpoint did not achieve statistical significance in the entire study population.”
A seven-month Southern Investigative Reporting Foundation investigation of Myriad’s business practices raises the following question: What is the premium an investor should pay for a company’s ability to spin ever more fantastical nonsense — and to brilliantly navigate the opaque line between required disclosure and misdirection?
GeneSight and rival pharmacogenomic tests have caused a fair bit of controversy over the past year.
In April the American Journal of Psychiatry published a critical 14-page review of “the evidence base” behind the marketing claims of GeneSight and three competitors, finding numerous problems with the tests the companies have presented to validate their products. And in July the American Psychiatric Association’s Council on Research released a statement that echoed the journal’s finding.
In May the Journal of American Medical Association Psychiatry published a two-page opinion from a trio of high-profile psychiatry researchers, Drs. Barbara Sommer, Bruce Cohen and George Zubenko, that explored the marketing of pharmacogenomic tests. They concluded that the current array of pharmacogenomic tests, which assess a small panel of about a dozen genes, are of limited use in treating major depressive disorder. As an illness, depression is simply far too complex, with a spectrum of underlying causes; many have no genetic component at all, they argued.
Dr. Bruce Cohen elaborated on his observations in an interview with the Southern Investigative Reporting Foundation. “There is no scientific basis to order [pharmacogenomic] tests at the moment; they are a complete waste of money,” he said, adding that when even companies conduct their own research to back up their marketing claims, “they do the studies wrong anyhow.” He declined to name any specific companies.
The most problematic aspect to pharmacogenomic tests is that they are aimed at finding something that simply doesn’t exist in the case of major depressive disorder, according to Dr. Cohen: “There are no genes that determine [patients’] risk [with] or their response to medicine,” he said, adding, “No small panel of genes is going to tell you whether you have most diseases, let alone depression. Schizophrenia, for example, has around 8,000 different genes associated with it.”
Dr. Cohen said he fears that doctors, who often don’t have the latest training in psychiatric genetics, are ripe targets for “company sales reps who are very pushy.” Physicians might “mistakenly” use these tests to make scientifically dubious treatment decisions, he said.
Those sales representatives will now have to be mighty pushy to market this expensive product as the prospects of insurance coverage dim. GeneSight sports a lofty $2,000 price tag, but just 15 percent of the 313,000 tests purchased last year were reimbursed at full price, with only $100 repaid for the rest. On average the reimbursement for the test was about $412.
And one brokerage research analyst — a member of a group of professionals rarely accused of skepticism — is not shying away from voicing concerns about some of these issues.
Barclays Capital U.S. life sciences tools and diagnostics analyst Jack Meehan assigned an “underweight” rating to Myriad’s shares. Formally this means that he sees the company’s stock price as underperforming relative to its peers’; informally, it means shareholders should sell their Myriad stock or avoid it altogether. He wrote that insurance plans will be reluctant to include coverage for GeneSight.
In research reports issued on Sept. 5 and Nov. 7, Meehan and his colleagues argued that GeneSight’s failed trial and critical reception by an influential subset of academic researchers pose a formidable barrier to its coverage by non-Medicare health insurance plans. Asked to discuss this research further, Meehan declined to comment, citing his company’s policy.
On Jan. 4 Meehan held a conference call about GeneSight’s trial with his brokerage firm clients and Dr. Charles Nemeroff, a University of Texas Medical School professor and a co-author of both the American Journal of Psychiatry article and the APA statement.
Dr. Nemeroff described the trial as unsuccessful. “The most salient and most important finding in this study is the fact that it’s a failed study,” he said, adding that GeneSight’s benefit for patients, as measured in the trial, “wasn’t even close to being significant.”
During this call, Myriad chose to defend GeneSight’s merits in a highly unusual fashion, however. Its director of clinical development, Bryan Dechairo, spoke up on the call 30 minutes in and after reading a prepared statement, started peppering Dr. Nemeroff with questions; he even tried to query him about a 2006 medical study mentioned in passing. Dr. Nemeroff, who had been politely answering all Dechairo’s questions, quietly informed him that the premise of his last one “doesn’t hold water.”
Two portfolio managers told the Southern Investigative Reporting Foundation that they had never seen a public company’s representative do something like this on an analyst’s client call.
That evening, Scott Gleason, Myriad’s investor relations chief and its head of corporate strategy, sent a select group of money managers and brokerage analysts an email that drew attention to what the company alleged amounted to errors in Dr. Nemeroff’s remarks. Two hours after Meehan’s call, Myriad hosted its own call and reiterated its argument that what matters most to patients is remission from symptoms of depression.
(A decade ago Dr. Nemeroff encountered controversy of his own, although it was not related to his research or views. On September, 17, 2008, Sen. Charles Grassley made Dr. Nemeroff’s undisclosed consulting arrangements with large pharmaceutical companies the centerpiece of a hearing. Emory University then removed him as chairman of its medical school’s psychiatry department. Dr. Nemeroff did not respond to several emails and a phone call seeking comment.)
For his part, Meehan, along with his colleagues at Barclays Capital, has since September raised questions following the conclusion of GeneSight’s unsuccessful trial about whether the product is worthwhile enough for doctors to rely upon. He has floated the idea that the Centers for Medicaid and Medicare could re-examine the decision to reimburse the cost of GeneSight for Medicare patients.
When the Southern Investigative Reporting Foundation contacted Dr. Elaine Jeter, who led the assessment of GeneSight for Palmetto in 2015, she said she viewed her team’s coverage approval for the product as conditioned on a successful randomized clinical trial. Asked if this was a formal condition of the Centers for Medicare and Medicaid Services’ decision or merely her personal view, she declined to answer, saying she had retired from Palmetto two years ago.
One Palmetto official who does have current oversight responsibility for GeneSight, Molecular Diagnostic Services chief Dr. Paul Gerrard, responded to a question about his agency’s initial approval like this: “In the case of GeneSight, a number of studies were done, all supporting a similar conclusion. As such, the studies reviewed in the [local coverage decision] collectively provided sufficient evidence to support the clinical utility of the test for specific uses.” He didn’t further discuss the results of the randomized clinical trial.
Myriad spokesman Ron Rogers declined to comment on a series of questions sent via email. But he felt the need to ask, “Which hedge funds [is the Southern Investigative Reporting Foundation] working for or with related to this inquiry?”
Editor’s note: In the initial version of this article, a paragraph discussing GeneSight’s clinical trial contained an inaccurate description of the clinical significance of the results. This has been corrected and the story updated.