Science’s COVID-19 reporting is supported by the Pulitzer Center.
Three unlikely collaborators are at the heart of the fast-moving COVID-19 research scandal, which led to retractions last week by The Lancet and The New England Journal of Medicine (NEJM), and the withdrawal of an online preprint, after the trove of patient data they all relied on was challenged. The three physician-scientists never were at the same institution nor had they ever before written together, but they are the only authors in common on the disputed papers, and the other co-authors all have ties to at least one of them. Their partnership, which seized a high-impact role during a global public health crisis, has now ended disastrously.
The first author for both retracted papers was cardiac surgeon Mandeep Mehra, an eminent Harvard University professor who works at Brigham and Women’s Hospital (BWH) and is known internationally for cardiovascular medicine and heart transplants. He provided the kind of gravitas that can fast-track papers to leading journals. In a statement provided by BWH, Mehra said he had met another of the trio, cardiac surgeon Amit Patel, in “academic and medical circles,” and that Patel had introduced him to Sapan Desai, a vascular surgeon and founder of Surgisphere, the tiny company that supplied the data. Journal disclosures, however, also indicate Mehra received compensation from Triple-Gene, a gene therapy company Patel co-founded to develop cardiovascular treatments.
Desai publicly aspired to combine big data and artificial intelligence (AI) in ways that he said can replace randomized controlled clinical trials. For a brief moment, it seemed that Surgisphere’s enticing data set, said to include nearly 100,000 detailed patient records from about 700 hospitals on six continents, would settle questions about the possible benefits of various drugs—including the controversial antimalarial hydroxychloroquine—for COVID-19 patients.
Patel once apparently headed cardiac surgery at the University of Miami Miller School of Medicine. A university press release announcing his arrival in 2016 is no longer posted on the university website, however, and the school has not confirmed his job duties there. More recently, he has been a volunteer adjunct professor at the University of Utah. But, as STAT first reported yesterday, Patel tweeted on Friday that he had severed his relationship with the university, which a school spokesperson confirmed. In recent years Patel has developed and commercialized experimental stem cell therapies purported to cure heart problems, reverse aging, or treat sexual dysfunction. He is also part of a network of physicians that just launched a trial to use stem cells from umbilical cord blood to treat COVID-19 patients.
Normally co-authors of high-profile papers share subject area expertise or have clear professional ties, says Jerome Kassirer, chief editor of NEJM during the 1990s. He calls the collaboration of the apparently disparate individuals “completely bizarre,” and a red flag that the studies warranted intensive scrutiny that the journals failed to provide.
None of the three co-authors responded to requests for comment. Patel spoke with a Science reporter initially but said he wanted to wait for audits of the Surgisphere data to comment, and Desai’s spokesperson stopped communicating after the retractions. Still, interviews with former colleagues and a long paper trail shed some light on each of them.
Desai had a history of convincing respected researchers of his skill and integrity. One of them, Gilbert Upchurch, department of surgery chair at the University of Florida, wrote last year in a journal commentary that he had never met Desai but had nonetheless mentored him remotely and developed an online friendship with him. Upchurch placed the scientist in a group of “amazing and talented young vascular surgeons.”
Illinois court records show Desai is facing two medical malpractice lawsuits filed last year. He told The Scientist that he “deems any lawsuit naming him to be unfounded.”
Desai has a history of big aspirations and entrepreneurial ventures—some short-lived. His science-fiction blog, corewardfront.com, was meant “to find the most parsimonious route for mankind to establish a meaningful presence in space.” In 2009, he wrote that the site would publish fiction “grounded in facts and reality,” adding, “the scientific method must be followed … religiously.” The blog is no longer published.
As a student, Desai won several small National Institutes of Health (NIH) grants for studies of the vestibular system. He started Surgisphere in 2007, when he was a medical resident at Duke University. Surgisphere’s initial products were medical guides and textbooks, although Desai has said he was working on big data projects for the company from its birth. In 2010, under the firm’s auspices, he founded the Journal of Surgical Radiology—whose editors included researchers with well-established publishing records. It folded in January 2013. Articles from the journal were cited only 29 times in its history, according to Scimago, a journal rating service. Yet an undated Surgisphere web page, no longer accessible online, said the online-only publication had 50,000 subscribers and nearly 1 million page views monthly—which would have placed it in elite company in academic publishing.
Surgisphere appears over time to have shifted its efforts into developing a database of hospital records that could be used for research. When the pandemic erupted, Desai declared that his data set could answer key questions about the efficacy and safety of treatments. Speaking about the finding that hydroxychloroquine increases mortality in COVID-19 patients, the main finding from the now retracted Lancet paper, he told a Turkish TV reporter, “with data like this, do we even need a randomized controlled trial?” Soon after, the World Health Organization temporarily suspended enrolling patients for its COVID-19 trial of the drug.
Immediately after the Lancet and NEJM studies appeared, however, critics identified anomalies in the data. And they doubted that a tiny firm—with a scant public track record in AI, few employees, and no publicly named scientific board—could convince hundreds of unidentified hospitals in dozens of nations to share complex, protected, and legally fraught patient data. Ultimately, despite Desai promising repeatedly to allow an independent audit of Surgisphere, the firm refused to release the raw patient data and agreements with hospitals for an audit, so no one could validate the authenticity of its database.
No hospitals have come forward to acknowledge working with Surgisphere. Indeed, NHS Scotland, which is mentioned as a case study on the company’s website, says none of its hospitals worked with Surgisphere and that it would ask the firm to remove an image of a Glasgow hospital from its website.
Science contacted several of Desai’s current or former employees or colleagues. Most would not comment. But Fred Rahimi, an Illinois podiatrist and co-author of a paper with Desai, praises the surgeon as highly capable for salvaging limbs, and easy to work with. Through his publicist, Desai cited Mark Melin, a University of Minnesota, Twin Cities, vascular surgeon, as a supporter. Before the retractions, Melin called Desai “a gentleman of the highest integrity” who has “nothing to cover up.”
But one physician-scientist who worked closely with Desai several years ago, says, “Just about everyone who knew him would say: ‘I just didn’t have a good feeling about him.’ … After they’d been with him, most people dissociated themselves from him,” the scientist says, declining to be named to avoid personal and institutional embarrassment.
In the decade since completing his medical residency, Desai moved from job to job—at Duke, the University of Texas, Southern Illinois University, and two private Illinois hospitals, according to his LinkedIn profile. “You might say we should have stopped him, which now seems obvious,” Desai’s former colleague says. “We should have found a way to get together and say, ‘What’s going on here?’ rather than allowing him to move from place to place. We should have done better as a medical community. We looked the other way.”
Before and after his stint at the University of Miami, which appears to have started in late 2016 or early 2017, Patel’s academic home was the University of Utah. He started as a full-time faculty member at Utah in 2008 and kept that position until he left for Miami. The website for Foldax, a heart valve company that he serves as medical adviser, describes him as a “Tenured Professor of Surgery in the Division of Cardiothoracic Surgery at the University of Utah School of Medicine and Director of Clinical Regenerative Medicine and Tissue Engineering at the University of Utah.”
The university confirmed Patel had tenure there, but says the directorship was an “unofficial title.” And among more than 100 publications listed on his University of Utah profile, nearly two-thirds were actually co-authored by other scientists who share the same surname. The page was removed from the university website after inquiries from Science.
According to the NIH database, Patel has never received funding from the agency. Before the recent COVID-19 papers, one of his most notable publications was a 2016 paper in The Lancet, which reported that extracting stem cells from the bone marrow of a person with end-stage heart failure and then reinjecting them could reduce the number of cardiac events that produced deaths or hospital admissions by 37%. The 126 patient, 31-site, phase II trial was billed in a press release, now not available on the University of Utah website but stored elsewhere, as “the largest cell therapy trial for heart failure to date.” Despite the apparent positive results, the sponsoring company Vericel no longer is developing stem cells for heart disease and, according to its webpage, is focused on “advanced cell therapies for the sports medicine and severe burn care markets.”
Patel left Miami under unclear circumstances, but has retained ties with Camillo Ricordi, an influential stem cell researcher at the University of Miami School of Medicine who is also the founder of a nonprofit called the Cure Alliance. The alliance previously focused on testing whether stem cells derived from umbilical cord blood could treat diabetes or Alzheimer’s, but has now pivoted to “fighting COVID-19,” according to its website. Ricordi is the principal investigator on a multisite trial to see whether the stem cells can treat lung inflammation in severe COVID-19 patients and Patel is listed in various references to the trial as a “key contributor” or “co–principal investigator.” Ricordi did not reply to requests for comments on his relationship with Patel.
Patel recently tweeted that he is “related to Dr. Desai by marriage” but called that “old news” and added, “Despite this I still do not have the information of what happened at Surgisphere.” In addition to apparently connecting Mehra and Desai, Patel had prior connections with other authors of the NEJM paper and the preprint. David Grainger, co-author of the preprint, is a professor of biomedical engineering at the University of Utah and also works with Foldax. Grainger declined to comment.
Timothy Henry, a cardiovascular clinician and scientist at the Christ Hospital in Cincinnati and a co-author on the NEJM article, has written several scholarly articles with Patel, including the 2016 Lancet paper. Henry, who also declined to comment, advises Patel’s Triple-Gene, which develops cardiovascular gene therapy treatments. Henry and Patel advise—and Patel is a board member of—Creative Medical Technology Holdings, a Phoenix company that develops and markets stem cell therapies, including treatments purported to reverse aging and cure sexual disfunction.
Creative Medical’s CaverStem and FemCelz kits are distributed to physicians who use them to extract stem cells from a patient’s bone marrow, then inject the cells into the penis or clitoral area to stimulate blood flow, according to a statement filed with the U.S. Securities and Exchange Commission. (As of the market close Friday, the publicly traded firm’s shares were valued at one-third of 1 cent.) The CaverStem treatments are advertised by the company as successful in more than 80% of patients, based on a 40-person phase I clinical trial that was not randomized or controlled, and on observations of 100 other patients. Phase I trials typically measure safety, not health benefits of a potential treatment.
Science contacted multiple colleagues or co-authors of Patel. None would comment. Before the retractions, two high-profile researchers—Deepak Bhatt, who directs interventional cardiovascular programs at BWH; and Peter Gruber, a pediatric cardiothoracic surgeon at Yale University—endorsed Patel on his LinkedIn page. Bhatt says he doesn’t know Patel and attempted to remove his endorsement after being contacted by Science. Gruber says he overlapped with Patel at the University of Utah about a decade ago, but doesn’t know his work in detail.
In contrast, Mehra—author of more than 200 scholarly articles, editor of The Journal of Heart and Lung Transplantation, and head of the cardiology division of the University of Maryland before moving to BWH in 2012—enjoys considerable support even after the unraveling of the recent studies. “Obviously, you don’t rise to the position he’s risen to without being ambitious, but I’ve never had any indication whatsoever that he would do anything unethical,” says Keith Aaronson, a cardiologist at the University of Michigan, Ann Arbor, who collaborated with Mehra on several studies, including a clinical trial of a mechanical pump for heart failure patients.
Mehra, the first author on both retracted papers, was the only one to issue a personal statement of apology, for failing “to ensure that the data source was appropriate for this use.” BWH and Harvard declined to say whether further investigation of Mehra’s roles in the papers would occur. (Mehra has written papers recently with another co-author of the Lancet paper, Frank Ruschitzka of University Hospital Zürich.)
“I think he just fell into this—perhaps a little naïvely,” says another former collaborator, cardiothoracic surgeon Daniel Goldstein of the Albert Einstein College of Medicine. “Given the amount of data that was in the [Surgisphere] database, it’s just hard to believe someone would [fabricate] something like this.”
Kassirer offers a harsher view: “If you’re a scientist and you’re going to sign on to a project, by God you should know what the data are.”
With reporting by Kelly Servick and John Travis. This story was supported by the Science Fund for Investigative Reporting.