Big Pharma, Big Data

Source: Financial Times

When Guido Rasi took charge of Europe’s medicines’ regulator in 2011 he inherited an explosive dossier that is now poised to transform drug development.

The furore started in 2007 when two Danish academics pushed the European Medicines Agency for greater transparency on 15 clinical trials. Their pursuit of greater openness set in motion a chain of events that stands to revolutionise the pharmaceutical industry in the months ahead. From the beginning of next year, the EMA will release all information about clinical studies submitted to it by organisations seeking authorisation for new treatments.

As this deadline approaches, heated parliamentary and legal debates are raging to determine how far this new culture of open access should go.

For advocates of change, including many patient groups and academic researchers, greater transparency in drug trials is essential to understand the full risks and benefits of medicines and reduce development costs. Critics argue that a cascade of publicly available data will threaten the confidentiality of patients taking part in trials and undermine the role of regulators.

Many pharmaceutical companies are watching the disputes with trepidation, fearing that the publication of so much trial data could damage their business models by allowing competitors to access the expensive research they undertook to develop their drugs. Chris Viehbacher, chief executive of Sanofi, the French pharmaceutical company, says: “I’m against a data dump on the sidewalk. This is second-guessing the regulators, and it’s not right if we have spent millions of dollars on this data for it to be released to everyone.”

In Britain, parliament’s science and technology committee has launched a probe into whether greater transparency is necessary to stop drug companies “cherry-picking” data that show their products in a favourable light.

The EMA’s decision to release so much data now puts Europe ahead of the level of openness expected in the US. Still, a court case initiated last month has illustrated how the action in Europe has implications for companies from the US and further afield. Two US-based companies — AbbVie and InterMune — are suing the EMA in the EU’s general court in an attempt to prevent it releasing their drug trial data.

This stand-off over data began six years ago, when Peter Gøtzsche and Anders Jørgensen from the University of Copenhagen wrote to the EMA requesting details of clinical trials on which the agency based its decision to approve the weight-loss drugs Rimonabant and Orlistat.

They worried that the drug companies could be concealing the full results of their tests. By potentially exaggerating benefits and playing down side effects, they feared the risk both of harming patients and imposing unnecessary costs on the healthcare system. They sought full details of the protocols describing the trials, the results and the supportive underlying raw data of tests on each patient trying out the medicines.

As they wrote in the British Medical Journal: “The effect on weight loss in the published trials is small, and the harms are substantial.” One — Sanofi’s Rimonabant — was never approved by the US Food and Drug Administration. It has since been withdrawn in Europe over concerns that it triggered suicidal feelings.

After being rebuffed by the EMA on the grounds that such information was commercially sensitive and therefore exempted from the usual EU rights of freedom of information, the Danes appealed to the European Ombudsman, who ultimately ruled in their favour in 2010.

Rather than contesting that opinion, the EMA released the documents in early 2011. Its decision has since triggered a flurry of requests, and officials have responded by making public a number of previously confidential drug authorisation submissions.

Mr. Rasi set EMA’s new tone in November, saying: “We are not here to decide if we will publish clinical trial data, only how.” He plans to release all data as soon as the agency decides to approve or reject a drug.

For medical researchers such as Mr. Gøtzsche, this move is essential. He has won support from others including the doctor and writer Ben Goldacre, author of Bad Pharma, and the “alltrials” campaign supported by a number of patient organisations and others including the BMJ.

They point to the publication in medical journals of only a biased selection of industry-funded clinical trials showing favourable results, and to cases where the scientific questions being studied are altered retrospectively to fit the data better.

Then there are accusations that Merck played down cardiac side effects of its painkiller Vioxx, which was withdrawn from sale globally in 2004, and that GlaxoSmithKline did the same for its diabetes drug Avandia, also subsequently pulled from the market in Europe and North America. The companies denied the allegations.

More recent ammunition has come from a group of academics led by Peter Doshi, who has been seeking fuller disclosure of the trials on Roche’s antiviral drug Tamiflu. After excluding studies they suspect are not independent, they question the drug’s benefit and whether the billions of dollars spent by governments stockpiling it for the 2007 flu pandemic were justified. Roche retorts that the regulators have had access to full clinical trial results even for studies that were not published in medical journals and that the efficacy of Tamiflu has been supported by reviews conducted by independent academics to whom it did release more details.

On the subject of full clinical data transparency, the pharmaceutical industry itself is divided. AbbVie and InterMune, the two companies that have sued the EMA, claim greater release will expose commercial secrets and reduce incentives for investment in developing drugs. That includes competitive details of clinical trial design and plans for their medicines.

Industry experts fear generic drug companies could use the data to accelerate the approval of cheap copies in jurisdictions where patent rules are weaker — or even steal the data and claim it as their own in regulatory submissions.

Paul Stoffels, worldwide chairman of Janssen, part of Johnson & Johnson, cites a different problem. His company’s Zytiga, a new treatment for prostate cancer, is based on an old compound on which the patent was close to expiry. He says he was only able to justify investing in costly clinical trials to support the treatment because of US and EU rules granting an additional period of exclusivity based on the efficacy data. If that information were made public, he argues, such rights would be lost and no company would invest.

Others argue rival companies could analyse the results to modify, accelerate or abandon their trials on similar drugs, saving costs and eroding the competitive advantage of the original developer. Supporting this view is the fact that four-fifths of all the freedom of information requests made to the EMA have come from other companies, consultants or lawyers, with just a handful from independent medical researchers or non profit groups. AbbVie’s request came from UCB and InterMune’s from Boehringer Ingelheim. Both of those companies are working on competitor drugs.

Not everyone agrees on the dangers. GSK has come out in favour of transparency, after being sharply criticised in 2008 by the UK’s drug regulator for failing to flag an unpublished study that identified suicidal feelings in children taking its antidepressant Seroxat. Late last year, Sir Andrew Witty, the chief executive, voiced support for a full release of data. “There was a history in the industry of fear around collaboration especially, openness of information and a mistaken judgment that actually by being more open and more transparent around data somehow that would destroy the fundamental business model,” he said.

But he and others are concerned about two further factors. The first is the danger of breaching patient confidentiality. All patients must give consent to take part in drug trials, while trusting companies to safeguard their personal data. The fear is that — especially for rare diseases — it could prove extremely difficult to ensure results are anonymous. That could make patients vulnerable to discrimination by insurance companies, employers or direct marketers.

Prof Sir Michael Rawlins, the outgoing head of the National Institute for Health and Clinical Excellence, the UK government’s medicines watchdog, told politicians last month: “Getting patients’ agreement can be very complicated, but it is not fair on patients to opt out by saying that we have patient consent.” He argues that access to patient-level data are also unnecessary and that the debate should focus on ensuring broader access to comprehensive clinical trial results. While all such studies are increasingly submitted to regulators to examine safety and efficacy, they are not all passed on to bodies such as his that assess cost effectiveness.

The second concern raised is the danger of misuse and misinterpretation of patient data, without the safety valve of a regulator. Full public release allows access to anyone, from medical academics with flawed analyses to vaccine denialists. Both could scare patients away from valuable medicines. “Patients could get very confused with the information coming out,” says Elmar Schnee, chief executive of Cardiorentis, a biotech company. “I hope this transparency gets stopped. People could do a lot of stupid things with it.”

The UK Academy of Medical Science, the US Institute of Medicines and the Wellcome Trust are in discussions with industry and academia about an agreement to release patient-level data only selectively to legitimate researchers to minimise abuse.

While the risks of full disclosure of patient data are unclear, the benefits also remain untested. The US FDA — which does not release patient-level data — employs its own scientists to trawl through and raise any concerns with companies. The EMA typically does not. So far, Mr. Doshi’s studies of Tamiflu have identified potential concerns but no smoking guns. Mr. Gøtzsche says he has recruited a new student to analyse the data on the two drugs that he received from the EMA in 2011 but has not yet been able to study it in detail.

In future, it seems likely that greater access to legitimate researchers equipped with the right skills will provide greater insight into drugs and help lower the costs of developing medicines for everyone. But, particularly in the shadow of the litigation now under way, the shorter-term emphasis may focus on consistent and comprehensive release of clinical trial results, not the underlying data.

In the long term, Mr. Gøtzsche argues, “the more we share the data, the more efficiently the drug companies will be able to develop new drugs, which will benefit everybody. We must leave the dark ages of secrecy that we have suffered for too long.”

Research: Drug companies look to crowdsourcing for discoveries

The vast amount of scientific information available online, and the growing sophistication of computer tools to “mine” it, have the potential to make significant medical advances far beyond the limits of any researcher or company.

With patients showing an increasing willingness to share personal medical data and discuss their conditions, specialist social media networks and groups such as PatientsLikeMe are being used to find recruits for clinical trials. Others, such as 23andme, which sells consumer genetic tests, have been able to use the data collected from their clients to identify patterns that may have been overlooked elsewhere. Last year it identified and filed a patent on a genetic variant linked to Parkinson’s disease.

Philippe Wolgen, head of Clinuvel, an Australian biotech company, says Facebook allowed his company to better understand patients’ feedback as it tested a new treatment for the rare condition of erythropoietic protoporphyria, or light intolerance.

Innocentive, originally developed by Eli Lilly and now a standalone company with clients in and beyond the pharmaceutical industry, uses “crowdsourcing” to solve problems. Like the X Prize, it posts a challenge on the internet with a cash reward for anyone able to resolve it.

InfoCodex has developed a computer program that automatically analyses medical literature for overlooked information. In a recent test, it sought new biomarkers — biological measures such as indicators in the blood — for diabetes and obesity by scanning 120,000 published academic papers and internal documents at Merck.

The result, according to Carlo Trugenberg, the company’s founder, was a number of potentially interesting new approaches that could accelerate drug discovery. They were so promising that Merck, the client, decided it would not make them public. In the fiercely competitive world of drug development, transparency is still sometimes one-way.

(c) 2013 The Financial Times Limited