Read our COVID-19 research and news.

Researchers say a camel virus described in an NEJM paper last week may be the result of lab contamination.

Researchers say a camel virus described in an NEJM paper last week may be the result of lab contamination.

M. Enserink

Research teams clash over too-similar MERS papers

A great story can be told again and again. But scientists working on the deadly Middle East respiratory syndrome (MERS) virus are puzzled by two papers appearing in separate journals that not only tell the same story, but also are based on data from the very same patient in Saudi Arabia.

The double publication—the first in Emerging Infectious Diseases (EID), the other later in The New England Journal of Medicine (NEJM)—has pitted Saudi Arabia's former deputy minister of health, Ziad Memish, against infectious diseases specialist Tariq Madani of King Abdulaziz University in Jeddah, who recently became the Saudi government's chief scientific adviser on MERS. Caught in the uncomfortable middle is German virologist Christian Drosten of the University of Bonn, who helped both and became a co-author on Memish's paper.

To make matters more complicated, several scientists, including Drosten, say the central conclusion of the NEJM paper—that the MERS virus jumped from a camel to a human at a Saudi farm—is flawed and most likely the result of lab contamination.

The case is another example of the complex politics of MERS research in Saudi Arabia, says Ian Mackay, a virologist at the University of Queensland, St. Lucia, in Australia, who first highlighted the similarities between the two papers on his blog on 5 June. "It typifies what MERS has been all about: very poor communication.”

"This is a much bigger issue than just a duplicate publication,” adds Michael Osterholm, director of the Center for Infectious Disease Research and Policy at the University of Minnesota, Twin Cities. "It really is a sign of the overall scientific investigation dysfunction that has occurred to date in Saudi Arabia.”

The paper by Memish and Drosten, which appeared online in EID on 20 March, reported a male MERS patient from Jeddah who had fallen ill in October 2013 after caring for sick animals in his herd of nine camels. The patient, who later died, may have been infected by his camels, the authors concluded, although they couldn’t rule out that both the camels and their owner had been infected by a third source with MERS.

Then on 4 June, NEJM published a paper by Madani and colleagues at his university and King Fahd Medical Research Center that appeared to describe the very same patient. The paper included full genome sequences of the viruses isolated from the camel and the patient, based on which the authors concluded that the man had most likely been infected by his camels.

Madani, who is the chair of a Ministry of Health (MOH) advisory committee on MERS, confirmed to ScienceInsider that the two papers describe the same case. "This patient was my patient at King Abdulaziz University Hospital,” he wrote in an e-mail. The man was admitted to the hospital on 3 November and his condition deteriorated. Madani says that after learning that the herder had taken care of sick camels, he sent a team of veterinarians and virologists to the man’s farm to collect samples on 8 November. Madani also notified MOH, where Memish was working, of his strong suspicion that camels might be the source. By 10 November, Madani says, his group detected the virus in the patient and in one of the camels.

One day later, “the MOH announced that they discovered that a camel was a source of infection to a human being without giving the credit of this discovery to our group or at least to our institution,” he says. Madani says he wrote an official complaint to the president of his university, who advised him to publish the findings as soon as possible.

But Madani’s NEJM paper was still beaten by 2 months by the EID publication that Memish co-authored. Memish says the ministry carried out its own investigation, as was standard procedure, sampling camels and people in the patient's household and in the health care facilities where he was treated. The first camel samples were taken on 9 November, according to the EID paper. "By no means did any practicing physician have the responsibility nor the authority to interfere with the MoH public health investigation or conduct their own investigation without official coordination,” Memish writes in an e-mail to ScienceInsider.

Memish says that nasal secretions from two camels tested positive for MERS repeatedly, and the 11 November announcement from the ministry came about because Abdullah Al-Rabeeah, Saudi Arabia's health minister at the time, ordered him to release the information as preliminary results. The Saudi press agency and the World Health Organization were informed on that day, and the next day Memish sent a report to ProMED, an online outbreak reporting system.

At the time, MERS had been sickening and killing people on the Arabian Peninsula for more than a year, but little was known about how exactly it spread. Several investigations had shown that many camels harbor antibodies against the virus, suggesting a past infection. The Jeddah patient seemed to offer an opportunity to investigate this link and possibly even prove camel-to-human transmission of the virus, a huge step forward. So the two rival teams continued with their parallel investigations trying to prove the same thing.

Madani's colleague Esam Azhar, the first author on the NEJM paper, asked Drosten, who specializes in coronaviruses, for help in early November. Drosten sent the team reagents and suggested how to proceed with the investigation, but he repeatedly refused to be a co-author, he says, because the researchers in Jeddah would not send him samples to test in his own lab. “I agreed with them: Advice yes, reagents sure, but I won’t take responsibility for the validity of the data with my own name.” In fact, Drosten soon became convinced that something was wrong with the data they were sending him.

The Saudi scientists said they had isolated a virus from the patient as well as from one of the camels, grown both samples in monkey kidney cells, and sequenced their genomes, which turned out to be 100% identical. They had also sequenced the virus directly from samples taken from the patient and the camel—and those two were also 100% identical. But the two sets of sequences differed in two positions.

That's impossible to explain, Drosten says; a virus can change slightly when put into cell culture, but why would the camel and the human virus show exactly the same two changes when cultured, changes never seen before? Drosten suspected that contamination had happened, and that what the researchers called the camel virus was actually the human virus as well. (Thomas Briese, a virologist at Columbia University, agrees that it seems unlikely for the mutations to appear twice but says he cannot exclude that possibility.) Frustrated, Drosten stopped working with Madani’s group in mid-December, he says.

Meanwhile, Memish says he had sent samples to the Wellcome Trust Sanger Institute in Hinxton, U.K., which sequenced the complete human virus and managed to sequence some small fragments of the camel virus. In January, Memish also enlisted Drosten, whose team managed to sequence eight further fragments of the camel virus. Together the pieces represented about 15% of the camel virus genome and these parts were almost identical to the genome sequenced from the patient. Antibody tests also suggested that the camel had had an acute MERS infection at the time the patient became infected. Although questions remained, Memish and Drosten decided to publish what they had. Meanwhile, Madani's team decided to submit a paper as well.

The paper by Drosten and Memish reached the finish line much faster; EID published it online on 20 March. Madani's paper went through several revisions and didn't get into print until last week.

Now that it's out, Drosten says he's standing by his criticism. Adding to his suspicions about contamination is that the NEJM authors report finding the virus at very low concentrations in the camel. “It seems strange that they managed to get a virus isolate from that at all,” Mackay says. Azhar says that the samples used to isolate the virus may have contained higher concentrations than those used to first detect the virus.

Virologist Bart Haagmans of Erasmus MC in Rotterdam, the Netherlands, says he has submitted a letter to NEJM to point out this issue, as well as what he claims are errors in the phylogenetic tree in Madani's paper. “I think it’s a pretty weird paper overall,” he says.

In a statement, NEJM says that there is a "substantive difference" between the two papers, and that its editors were not aware of the EID paper. "Had they been aware, they would have requested an acknowledgement in the NEJM article," the statement says. (The journal has now added the acknowledgement itself.) The statement did not address the alleged problems in the paper.

Memish lost his job on 2 June, for reasons that are unclear. Six weeks earlier, Al-Rabeeah had already been replaced as health minister by Adel Fakieh. Osterholm is optimistic that the shake-up is a sign of greater commitment to transparency in Saudi Arabia. In fact, just a day after Memish was fired, the country announced 113 new MERS cases, including 92 deaths, which had occurred in the previous months and had gone unreported.

But Mackay remains unpersuaded that Saudi Arabia will be more forthcoming. "Nothing has changed, the new transparency is just more opacity," he says. The government hasn't explained where the 113 new cases came from, for instance: "There was no data attached whatsoever, how they were tested, dates, ages, anything that would be interesting to an epidemiologist,” he says. "Just putting out this number with a death rate of 80% is crazy."

*Clarification, 11 June, 11:02 a.m.: This story has been updated to clarify a few details in the research described in the EID paper.