In mid-April 2014, the words ‘scandal’, ‘secret’, ‘drugs’, and ‘Roche’ (the Swiss drug giant) delighted the medical press, which was quick to publish tantalizing news about the ineffectiveness of Tamiflu and the enormous amounts spent by governments in order to stock up on the drug, in preparation for possible flu epidemics.
“From the Cochrane analysis, Tamiflu does not reduce the number of hospitalizations. There were not enough data to see if it reduces the number of deaths… When analysing the five clinical studies which present detailed diagnosis, no benefit is observed.” The evidence indicating this inefficacy had been leaked, bit by bit, through unconventional methods of corresponding with pharmaceutical companies and regulators. The diligent authors were under the impression they were onto a well-kept secret—a kind of secrecy reminiscent of The Da Vinci Code. This made many wonder if the end of drugs was nigh. Aren’t drugs just a system of snatching money from the unfortunate victims of human suffering?
In the first decade of this century, Tamiflu was a pharmaceutical bestseller: seventy governments around the world ordered it, stockpiling reserves to treat 220 million people. The total acquisition cost was nearly $7 billion.
Why spend seven billion dollars on an anti-flu drug? The answer has been tied to the fear of a hypothetical flu pandemic, a possibility that began to concern the US authorities after 1977, and the World Health Organization in the late 1990s. Until 1976, the world had had a mostly reactive approach to influenza worldwide. In 2019, when a new swine flu virus was detected, US authorities took a proactive stance in preventing and preparing for a possible flu epidemic. 
Although many thought at the time that such concerns were absurd, as the flu was often considered a trivial disease, those who knew history also knew that the flu is capable of causing disaster: in 1918, when World War I ended, a flu pandemic broke out, known as the “Spanish flu”. In the two years (1918-1919) that it haunted the entire planet, the Spanish flu killed about 20-25 million people. In comparison, the carnage caused by the First World War totalled about 10 million deaths among soldiers and approximately another 6 million deaths among civilians (to these are probably added several more millions of deaths linked to war-related starvation). Less than two years of Spanish flu had an impact at least as dramatic as the four years of war massacres. Therefore, fears regarding a potentially catastrophic pandemic weren’t exactly out of the blue, although it had been argued that an influenza virus was unlikely to cause another pandemic similar to that of 1918-1919.
In addition, in the last 10-15 years, the persistence of a bird influenza virus on several continents, with the potential to undergo a mutation to trigger a pandemic (given that it has proved to be zoonotic, causing 377 deaths out of 633 human clinical cases, i.e. a mortality rate of about 60%) made the adoption of preventive measures a public health priority. Among these measures was the storing of considerable amounts of Tamiflu.
Looking for data
In the context of these concerns about a potential flu pandemic, the British and Australian governments called on the Cochrane collaboration network to update the organisation’s systematic reviews of Tamiflu. Cochrane systematic reviews are generally considered to be the highest quality evidence in the healthcare field, due to their scientific rigour. Such reviews had been carried out in 2008, concluding that, to some extent, there is evidence to indicate that the drug has the ability to reduce the rate of influenza complications, such as pneumonia, for example.
Dr Keiji Hayash, a paediatrician in Japan (one of the world’s largest Tamiflu users), aggressively challenged the findings of the Cochrane review after some side effects were recorded in children. He pointed out that the statistical analyses in the Cochrane report were based on data gathered in 10 randomized controlled trials, but only 2 of these had been peer reviewed. The other 8 had been published in the scientific papers of some medical conferences (5), one of them was only a summary (1), and the others consisted of data obtained by Roche themselves. They had been compiled into a 2003 article published by several employees of the manufacturing company, a company consultant, and a professor, Laurent Kaiser. “We strongly suppose that the reviewer’s conclusion about the complications was mainly determined by these 8 RCTs [randomized controlled trials], we should appraise the 8 trials rigidly. Without this process it’s difficult to conclude that oseltamivir (Tamiflu) can prevent lower respiratory tract complications,” concluded the Japanese paediatrician.
Tom Jefferson, the leader of a group involved in the Cochrane review, decided to take into account Dr Hayash’s criticism and gain access to complete data on those clinical trials. This proved much harder in practice than it seemed in theory. By requesting information from Roche and authorities in Europe, the United States and Japan, it took the authors several years to gain access to relevant information, due to the fact that Roche was unwilling to make it public. On one occasion, Roche representatives said that researchers already had all the information they needed, and on different occasions, they stated either that they were concerned about the confidentiality of the patients included in the clinical trials, or that they were sceptical about the independence and impartiality of the Cochrane group.
The involvement of several journalists by the authors in the rather contorted correspondence with the pharmaceutical giant led to an increase in the negative public perception of the lack of transparency of data generated in clinical trials. Moreover, a campaign to gain public access to all (or almost all) data generated in clinical trials was launched. Only after this movement of public opinion intensified, and the questions of the Cochrane group and the prestigious British Medical Journal became more and more insistent and widespread, did the company give in and make the necessary data available to researchers.
How effective is Tamiflu?
The results of the Cochrane review do not commend the efficacy of the anti-influenza drug: it reduces the duration of symptoms of the disease by less than a day (about 16 hours), but does not reduce the risk of hospitalization. It does reduce the risk of radiologically unconfirmed pneumonia, although the effect was not significant in the 5 studies that used a more detailed diagnosis of pneumonia. It does not reduce the risk of bronchitis, sinusitis, or otitis media.
Most of these findings are not really that surprising: the summary of product characteristics (the official document describing the properties of the drug for health professionals, doctors, and pharmacists) mentions that the reduction in the duration of symptoms amounts to about a day in the case of the influenza A virus, and 0.7 days for the influenza B virus. The reduction in the risk of hospitalization was not included in the European summary of product characteristics (meaning that it was not taken into account for product authorization), and, regarding the risk of pneumonia, the European document mentions only a slight reduction in the risk of developing “specified lower respiratory tract complications (mainly bronchitis) treated with antibiotics” (the reduction, from 12.7% to 8.6%, was challenged by the Cochrane review).
So, were the billions of dollars spent and the tons of Tamiflu stored in preparation for the flu pandemic a futile exercise? The authors of the Cochrane review and Fiona Godlee, editor of the British Medical Journal, think they were. Neither the European Medicines Agency (EMA) nor its US counterpart, the Food and Drug Administration (FDA), have issued an official position. However, when interviewed by the press, Sabrina Spinosa, from the EMA, said that “the review does not raise any new concerns” and that the agency maintains its point of view on the relationship between risks and benefits specific to the drug.
The Infectious Disease Society of America (IDSA) issued an official statement where they continued to recommend the use of neuraminidase inhibitors (oseltamivir, zanamivir) for the treatment of influenza. In that statement, not only are the data not considered surprising, but it is also mentioned that the Cochrane analysis included both influenza virus-infected and non-infected persons with influenza-like illness, an approach that is not appropriate for determining Tamiflu’s neuraminidase treatment efficacy.
The IDSA statement emphasizes that the main utility of these drugs is, most likely, treating patients with severe conditions, or at higher risk for influenza complications. No placebo-controlled RCTs are available for NAI treatment of hospitalized influenza patients; there is evidence from several observational studies (studies considered less rigorous and less likely to prove the usefulness of a drug) that show the ability of these drugs to reduce serious risks, including death, especially if administered within the first two days after the onset of the disease. The Center for Disease Control and Prevention (CDC) in the United States also issued a statement that they still recommended treatment with antiviral drugs for the flu, especially in high-risk patient groups.
The end of drugs?
The odyssey of gaining access to information about clinical trials and their seemingly disappointing results (which were mostly previously known) has been read from many perspectives by various analysts, experts, and ordinary people. Some found the existence of “legislative loopholes” that allowed pharmaceutical companies to not disclose all clinical data on drugs, and thus there is the need to continue and step up efforts and campaigns to change legislation in order to ensure transparency.
Others have found incompetent authorities that waste impressive amounts of resources for nothing. Still other level the often-made accusation of greed at the pharmaceutical industry, pointing out that it has nothing to offer other than hope, because the drugs they sell are really just well-promoted placebos. The proponents of all kinds of alternative therapies (from homeopathy to acupuncture, Ayurvedic medicine or bio-resonance), argue that conventional medication is useless, expensive, and even dangerous, since it has side effects. So, is this the end of drugs?
That would be hard to believe. Beyond the fact that the usefulness of neuraminidase inhibitors still remains disputed, to generalize from the inefficacy of a single drug to the more than one thousand currently available would be a serious error of elementary logic. The benefits of medications can vary widely, from controlling a symptom (only slightly better than a placebo) to significantly improving quality of life—and even saving lives.
Although current drugs are far from perfect, pharmacology has never had better resources in the history of mankind, to which we have documentary access. In fact, in the last century, our lifespan has increased considerably, mainly thanks to medicines. Moreover, in the last few decades, along with impressive advances in the fields of molecular biology, immunology, genetics, and other medical-biological sciences, progress has been even more pronounced.
We can’t cure AIDS yet. But, if 20-30 years ago the diagnosis of AIDS was a death sentence, today in Western nations, thanks to antiretroviral drugs, an HIV-infected patient at the age of 20 generally has a life expectancy of 70 years or more; that is, almost the same life expectancy of a young person without HIV (albeit there are some differences depending on the various individual risk factors).
Progress in oncology has been limited, but not negligible: data from the UK show that while in the 1960s only a quarter of children diagnosed with cancer survived the disease, today, three-quarters of them win the fight against cancer. In adults, five-year survival rates for brain tumours or colon cancer nearly doubled from the same reference period (1960s); survival rates from leukaemia have tripled compared to 40 years ago. In 1970, only five out of ten women, once diagnosed with breast cancer, survived for more than 5 years. Today, their number has increased to eight out of ten.
Hepatitis C, a major cause of cirrhosis and liver carcinoma, is becoming almost completely curable due to progress in recent years, and the lives of patients with psoriasis or rheumatoid disorders can be greatly improved with new biological therapies (albeit not without risks).
In the last 20 years, the remarkable advances in genomics and proteomics technologies make the idea of personalized medicine and therapy, adapted to the unique genetic profile of an individual, even more feasible, even if we cannot estimate how great the distance is between concept and mass implementation. In any case, this distance is decreasing so that, rather than the end of drugs, we may be seeing—as one BBC journalist put it—the beginning of a new era of drugs.