Watch out for mice

What will happen to the medicine for two million dollars after the recall of the key article

Polina Loseva, N+1

Recently, the journal Nature Biotechnology withdrew a 2010 article about how gene therapy for spinal muscular atrophy was tested on laboratory mice. Since then, the article has been cited 557 times and mentioned in at least 59 patents, and based on its findings, clinical trials have been conducted and the drug "Zolgensma" has been registered. Now there are "numerous inaccuracies" in the article: the data on the survival of mice turned out to be questionable. We are finding out whether this will affect the fate of patients who have already received a shot of "Zolgensma" or are just waiting for it.

Zolgensma is a pioneering medicine. In particular, it is one of the first gene therapy drugs against rare hereditary diseases. Most of the currently approved methods of gene therapy treat blood diseases, such as hemophilia. Zolgensma targets the nervous system: with the help of adenovirus particles, it delivers a healthy copy of the SMN1 gene to motor neurons. And thus should mitigate (and ideally even cure) the symptoms of spinal muscular atrophy (SMA) — genetic pathology, in which a group of spinal cord neurons gradually die off, and skeletal muscles, without receiving a signal from them, contract poorly and eventually atrophy.

Zolgensma was not the first drug for SMA, Spinraza had already entered the market before it. The latter works on a different principle: it does not deliver a new gene, but corrects protein production. Therefore, the effect of it is temporary, and injections need to be repeated. "Zolgensma" became the first medicine, which is enough to inject just once.

And since one injection was supposed to solve the patient's problem for life, Novartis assigned its drug a record price at that time — $ 2.1 million. So "Zolgensma" became the most expensive medicine in the world for three years, the palm was taken away from her only in 2022.

During these three years, "Zolgensma" was registered in more than forty countries, and the account of patients who received the "golden" injection went to thousands. And only now scientists have doubts about whether this drug was tested fairly enough initially.

Questions for mice

The 2010 work is a typical test of the new method on mice. The researchers worked with a model line of animals devoid of both copies of the SMN gene. The experimental group was divided in half, each half received an injection of adenovirus particles: either experimental — with the SMN gene, or control — with the green fluorescent protein gene.

After that, the scientists checked that the gene really got to the nerve cells, that it triggers protein synthesis, and the cells themselves work (that is, generate impulses) and establish contact with the muscles. Then the animals were observed: how they gain weight, whether they quickly turn over into a comfortable position, and most importantly — how long they live. And at the observation stage, something went wrong.

The survival schedule that the researchers cite in the article looks very optimistic. While 11 control mice did not live even a month, 11 experimental mice confidently held on longer. One of them died on the 97th day for an unrelated reason, four researchers were euthanized around the same time to assess the state of neurons. And the remaining six, judging by the schedule and the text of the article, were already 250 days old by the time the article was submitted to the magazine, and they were not going to die.

On the left is a graph with "inaccuracies". Green indicates the survival of the control group, red indicates the experimental group. On the right — body weight: gray indicates healthy mice, red — mice with SMA after treatment, green and black — mice with SMA that received the fluorescent protein gene or no treatment at all (Kevin D Foust et al. / Nature Biotechnology, 2010).

This graph, despite the small sample, means (although it does not prove definitively) that the therapy can be considered safe and effective: the experimental group not only lives longer than the control, but also does not die from possible side effects in the long term.

It turned out that the long-term perspective cannot be judged by these data. On October 6, 2022, Nature Biotechnology withdrew the original article and published an explanatory note. It follows from it that in 2021, the authors of the article themselves contacted the editorial board of the journal and reported that they had found "inaccuracies" in their data on the survival of mice. The data themselves are not published on the site, but it follows from the text of the note that there were inaccuracies in how the authors measured the life expectancy of mice and included animals in the sample. After corrections, it turned out that only one mouse (and not six at all, as stated) survived 250 days after treatment.

As a result, the journal decided to withdraw the article from publication — because "the number of inaccuracies undermined the editorial board's confidence in this work." The authors of the work did not agree with this decision, but they have not commented on it publicly yet. But the management of Novartis It stated that it agrees with the review of the article and has already removed links to it from its advertising materials.

Mistakes of the past

"With "Zolgensma" This is the second case," says Ilya Yasny, head of the scientific expertise of the LanceBio Venture venture fund. — These are traces of previous sloppy work." Three years ago, there were already disputes around the medicine, very similar to what is happening now.

May 24, 2019 FDA (US Food and Drug Administration) approved "Zolgensma" for use — that is, checked the results of tests on mice and humans and made sure that the drug is safe and effective. On June 28, representatives of AveXis, the company that invented and developed the drug, contacted the FDA, after which it was bought by the pharmaceutical giant Novartis. They reported that inconsistencies were found in the data included in the application for the FDA. The errors were found in one of the technical experiments, which was not part of the preclinical tests. The medicine began to be produced according to a new protocol, and it was necessary to check that it also works on animals, as well as the first trial batch. And this check was carried out carelessly.

The FDA went to investigate — and indeed, there were inconsistencies. The problem was that the researchers inaccurately recorded the results of observations of mice, and unverified data got into the report. Somewhere the described life expectancy differed from the real one by a day, somewhere by two, and somewhere by as many as 19 days.

The situation looked suspicious: the company reports violations just a month after its product was approved. Novartis explained this by saying that its employees discovered the problem two months before the drug was approved, but after that they conducted their own internal investigation. The investigation was kept secret so that the researchers themselves would not guess anything and did not change the documents, and therefore it went slowly. And it ended after the approval of "Zolgensma".

As a result, the dispute was resolved without prejudice to the patients. The management of Novartis dismissed the director of science of Avexis, Brian Kaspar, as well as the deputy head of the company for scientific affairs Alan Kaspar, Brian's brother. And promised to continue to inform the FDA about all new violations if they are found. And the FDA, in turn, decided not to withdraw the drug from the market: experts concluded that the inaccuracies found "do not affect patient safety or the effectiveness and quality of the product."

And what about the children

The questions to the preclinical studies that forced the editorial board of Nature Biotechnology to withdraw the article are echoes of the same problem. Of course, we are talking about different experiments: in one case, they checked whether the drug works in principle, and in the second — whether it stopped working after changes in the production process. But the "inaccuracies" are absolutely the same: again, inaccurate data, unaccounted-for mice, and the main author of the article is the same Brian Kaspar.

Such a story is rather rare for large pharmaceutical companies. "I don't remember [other] such precedents,— says Yasny. — Usually, if falsifications are detected, it is manipulation of clinical research data, and then the drug is not allowed on the market. And here the violations are small and do not affect the benefit and risk profile, but unpleasant for the reputation."

However, for practitioners and the reputation of the medicine does not seem to have changed. "For me, as a clinician and a doctor who has his own practice, this article is not very alarming," says Svetlana Artemyeva, head of the Department of Neuropsychiatry at the Veltischev Institute of Pediatrics and Pediatric Surgery. According to her, the data from clinical trials in humans is quite enough to assess both safety and efficacy, regardless of the results in mice.

At the same time, there is still quite little data on the survival of patients with SMA after treatment with Zolgensma. Since the medicine is young, even the very first test participants who received it, doctors have been watching for about 6.5 years. And it's too early to judge how treatment will affect their lives in adolescence and adulthood. In addition, at the first stages of the tests, as it often happens, the sample was small — only 15 people.

Artemyeva says that, unlike mice, it is not possible to achieve one hundred percent survival of children who are treated with Zolgensma. The medicine does not begin to act immediately, so the prognosis depends on how competently the doctors lead the child after the injection of "Zolgensma". "When we start treating a severe child with type I SMA," she says, "we recommend continuing respiratory support anyway, monitoring feeding so that there is no aspiration, because it takes time to restore the nervous system."

In some cases, according to her, it is not possible to properly provide respiratory support — but it is not necessary to attribute this to the side effects of the drug. Therefore, Artemyeva believes that the review of the article on the results obtained in mice will not undermine confidence in the drug, and it will continue to be prescribed to patients.

Ilya Yasny also believes that inaccuracies in the article — even the one that served as the foundation for clinical trials — will not affect the fate of "Zolgensma". "The building will stand," he is sure, "because more than 2,300 children have already been treated, and clinical data have priority over preclinical ones. The trajectory of development of treated children is far ahead of historical control: they begin to walk, to say what could not be expected from the untreated."

The FDA has not yet responded to the review of the article, but representatives of Novartis do not fear damage to the company. And if there really are no consequences for "Zolgensma", then it will be possible to state: it only seems that modern pharmacology stands on the shoulders of mice. As soon as the next building is completed and the medicine appears on the shelves, mice from the foundation can be safely removed — without prejudice to the rest of the structure.

Portal "Eternal youth" http://vechnayamolodost.ru