Nature Biotechnology 24, 270 - 273 (2006)
Published online: 8 March 2006; | doi:10.1038/nbt0306-270
Ten years of biotech gaffesJohn HodgsonJohn Hodgson is editor-at-large, Nature Biotechnology and a principal at Critical I Ltd., The Colin Sanders Innovation Centre, Mewburn Road, Banbury, OX16 9PA, UK. john.hodgson@criticali.net Much of biotech's success has been built on lessons learned from mistakes. But the past ten years has also witnessed mistakes many in biotech would prefer to forget.On the occasion of Nature Biotechnology's 10th anniversary, it would seem right and proper to celebrate some of biotech's greatest achievements. But I shall leave the plaudits to others, elsewhere in the journal. This article is about biotech's memorable gaffes, the deeds or events that their doers would rather forget. In a field of human endeavor where so much is about transforming something that is hardly known into plausible products for unserved markets, it would be surprising if no mistakes were made along that tricky way. From the appointment of the wrong executives to the misworded legal document, most biotech companies are replete with everyday errors, the kind of mistakes that could happen to anyone and which are readily and rapidly correctable, almost before anyone finds out. I will be leaving such trifles unconsidered. What I deal with here are the 'bio-bloopers' that mattered, the events that had, or still have, repercussions for the life science industry or for those who depend on it.
This is not a definitive list of the errors and idiocy prevalent in this industry and the gaffes are presented in no particular order. Other goofs should doubtless have been included, and I encourage readers to send feedback on major omissions to the journal.
Gaffe 1: Clinton and Blair joint declaration
Throughout the latter half of the 1990s, biotech investors had been getting rather heated over the prospects of companies involved in genomics (or companies who had merely donned the genomics mantle). The fever mounted towards the end of 1999 when Nature published the first completed DNA sequence of a human chromosome (chromosome 22), when the billionth base of the human genome was sequenced and when Celera Genomics (Rockville, MD, USA) and a consortium of academic partners prepared to publish the completed fruitfly genome.
Meanwhile, investors were throwing money at companies, such as Celera and Human Genome Sciences (Rockville, MD, USA) and Millennium (Cambridge, MA, USA). These three companies each raised close to a $1 billion on the back of follow-on offerings and convertible instruments, whereas others, such as Abgenix (Fremont, CA, USA), Gene Logic (Gaithersburg, MD, USA), Incyte (Palo Alto, CA, USA) and Medarex (Princeton, NJ, USA) raised $250–600 million each. The investment sentiment was that genomics was going to provide the key to designing more effective drugs through an increased understanding of the molecular targets that genomics projects would be able to associate with particular disease conditions.
The genomics revolution, as it was painted, was characterized as a land-grab in which the greatest potential for profitable drug making would lie with those companies who were able to lay claim to the largest slice of the human genome and that success in genome grabbing would come to those who went into the endeavor with the greatest resources. Thus, investors figured that the more they put into a company, the more they would get out.
Their bubble was burst by President Bill Clinton and Prime Minister Tony Blair at a press conference held in the White House on March 14, 2000. The two heads of government called for the information from the human genome sequence to be made fully accessible to all people.
The fallout. The warm, fuzzy joint declaration was broadly interpreted as a move to end the controversy over patents based on sequence-tagged sites, thousands of which were awaiting examination at the US Patent and Trademark Office. Biotech stock values immediately plummeted 20% and the event proved to be more or less the start of a 'nuclear winter' for biotech financing.
In the event, it didn't matter whether or not Blair and Clinton had actually intended to imply that sequence patents should not be granted. The market now understood, as if it had just awoken from a dream, that the value of the human genome couldn't be measured by the height of paper describing sequence-tagged sites or sequence claims. Genomic associations might strip away some of the inherited factors from complex diseases and they might help identify targets for drug development, but the value in the process lay more or less exactly where it always had—in the ability to bring molecules through clinical development and deliver them to patients.
The take-home lesson: just because everyone says that the emperor's new clothes are wonderful, it doesn't mean he isn't just standing there in the buff.
Gaffe 2: the ImClone debacle
There were, it could be said, numerous mistakes made in transforming ImClone's (New York) investigational cancer drug, IMC-C225, into the ImClone/Bristol-Myers Squibb's (BMS; New York) product, Erbitux (cetuximab). The first was the extraordinary lavishness of the $2-billion September 2001 deal under which ImClone and BMS would codevelop and copromote the monoclonal antibody that targeted epidermal growth factor receptor. BMS would give ImClone $1 billion for passing three clinical milestones and then 60% of the revenues on top of that. As the market leader in oncology at that point but with a fast evaporating pipeline, BMS was clearly desperate to buy a slice of Erbitux. The next 'gaffette' was BMS's rather hands-off approach to the codevelopment process. Left to its own devices, ImClone's conduct of the clinical trial of the drug in colorectal cancer was substandard and its approval submission was described as "sloppy" by the US Food and Drug Administration.
This clinical setback was the trigger for the real stupidity. ImClone's CEO Sam Waksal determined to sell his own shareholding in the company before the price plummeted from around $70 at the end of 2001 to just above $15 at the beginning of 2002. This was a clear case of insider dealing that ultimately earned Waksal a 7-year stretch in jail. At the same time, however, he also tipped off domestic goddess, Martha Stewart, the US embodiment of motherhood and apple pie and the eponymous founder of Martha Stewart Living Omnimedia, a business that had once boosted her personal wealth to around $1 billion. Stewart acted on the tip—she held precisely 3,928 ImClone shares which she sold for just under $60 each. It is possible she saved $170,000 (and probably wouldn't even have lost that much if she had hung onto the stock until Erbitux was finally approved in February 2004). She then compounded her culpability by denying to federal authorities that she had ever received a tip. Found guilty on several counts of obstruction of justice, she too ended up in jail and the value of her business dropped nearly $100 million overnight.
The fallout. Greed is universal. But Waksal's and Stewart's actions shook investor confidence in the entire biotech sector. In January 2002, following the 90% nosedive of its own stock price, ImClone's malaise dragged down many other biotech companies (e.g., the Burrill Select biotech index fell by 15% compared with a 1% fall on Nasdaq). Investors had viewed ImClone as a safe bet: it was an established biotech company with a prominent management team, a stellar board of directors, a slam-dunk late-stage product, a promising market and a record-setting deal from a pharmaceutical company for validation. And yet all in the space of a few weeks, investor confidence evaporated. But Erbitux's later approval proved that many of ImClone's fundamentals were sound, suggesting that much of the short selling was based on hysteria rather than evidence.
The other lesson: learn from your mistakes. Don't hide them or compound them.
Gaffe 3: Muddled messages in agbiotech
Gaffes in the agbiotech area could probably fill an entire article on their own; last year's cockup in which Syngenta (Basel, Switzerland) admitted that it has been selling an unapproved transgenic corn seed, Bt10, to US farmers between 2001 and 2004, being a recent example. But industry is not wholly culpable for lukewarm public acceptance and stigmatization of transgenic crop technology. In November 2001, Nature published a paper by David Quist and Ignacio Chapela of the University of Berkeley, California, that purported to show that transgenic DNA constructs had been found in farmers' landrace maize varieties cultivated in remote parts of Mexico, the center of genetic diversity of the crop. This work caused consternation in environmental circles, with renewed calls from activist organizations such as Greenpeace and Friends of the Earth for moratoria and outright bans on transgenic crops. The Mexican government reacted to the paper by amending its genetic modification regulations, making bench researchers liable for breaches of the rules and by banning the export or import of any form of recombinant DNA.
It subsequently transpired that the Quist and Chapela paper was technically flawed. The method of inverse PCR used to detect the transgenes was suspect, and Quist and Chapela were unable to provide samples to enable other researchers to analyze their findings. By April 4, 2002, Nature itself had disavowed the paper (although the authors still stand by their findings) with an editorial note that indicated it would not have published the paper had the criticisms that arose after publication arisen during the review process. A subsequent study of 150,000 samples from Mexican maize failed to find a single sample in which transgenes were detected, this despite the analysis having been carried out by Genetic ID (Fairfield, IA, USA), the favorite DNA diagnostics company of Yogic flyers.
The fallout. With so many of the groups ideologically opposed to transgenic crops able to exploit the media, scare the public and perpetuate myths and conspiracy theories about genetic engineering over the Internet, prestigious journals should be aware of the long-lasting damage resulting from their willingness to widely publicize results that may be contentious or equivocal. Reviewers and editors at journals also have a responsibility to ensure that peer-reviewed data are reliable. Scientific gaffes such as this one, and media distortion of the results of a laboratory study of the effects of Bacillus thuringiensis (Bt) toxin on Monarch butterfly larvae published before it, certainly contributed to the decline of European agbiotech (from 264 field trials in 1997 to 35 field trials in 2002).
Gaffe 4: European 'novel food' regulation
After years of indecision and wrangling between European Union (EU) member state governments, the European Council agreed on regulations for 'novel foods' in May 1997. A novel food was one, the legislation said, that was not "substantially equivalent" to something already on the market. The logical thinkers in science and industry concluded that many genetically modified (GM) foods would be substantially equivalent to existing types and therefore they would escape additional and meaningless labeling. However, the EU Novel Food Regulations had not defined how equivalence or nonequivalence would be determined or indeed how big 'substantial' is.
In the event, substantial turned out to mean that if any smidgeon of protein product or recombinant DNA could be detected, then the food would fail the "substantial equivalence" test. Substantial thus meant "detectable by the most refined genetic or protein test." Improved PCR-based methods, often developed in collaboration with the European Commission's Joint Research Centers in Ispra, Italy and Geel, Belgium, meant that any food over which a GM grain or bean had breathed was no longer substantially equivalent, not withstanding the fact that the genetic modification in question had no material impact on the product. Consequently, all food containing any amount of GM flour or GM rapeseed oil had to be labeled "contains GM ingredients" or "may contain GM ingredients."
The fallout. The opponents of genetic modification were quick to act within the logical void. They tested unlabeled food products and publicized the results. They promulgated boycotts of supermarket chains and retailers throughout Europe and the United States until each of the companies signed up to a GM-free policy. One-by-one these giant corporate pussycats not only bowed to this unreasoned pressure but made marketing capital from it. 'GM-free' labels now shout out emptily from every supermarket floor in Europe.
The take-home lesson: don't expect political fights to be clean and don't expect big business to support logic when markets are at risk.
Gaffe 5: US bans funds for new ES cell lines
In the run-up to the 2001 US presidential election, George W. Bush decided that his pro-life stance had implications for the then-emerging field of regenerative medicine. In 2000, while Bush was governor of Texas and running for the Republican nomination for president, the National Right to Life Committee, the largest US anti-abortion group ran radio advertisements attacking Bush's rival, Arizona senator, John McCain. McCain had drifted from a staunch anti-abortion, anti-embryonic stem (ES) cell position to an anti-abortion, pro—fetal tissue research position. But Bush's campaign conflated the abortion issue with the use of fetal cell lines and this simplistic position won the political day.
After Bush was elected, he announced (in what was painted as a grand compromise) that the US government could pay for stem research only as long as it was conducted on one of 60 preexisting cell lines. This imposition roundly ignored the status of the ES cell field at that point, which was (and still is) that it remains unclear which cell lines will work in any given protocols. The American Society for Cell Biology and the Federation of American Societies for Experimental Biology responded feebly to what some US researchers felt was nothing less than a ban.
The fallout. It is estimated that the US National Institutes of Health spends around $600 million in the stem cell area, only 1–2% of which is now spent on ES cell work. Parts of the rest of the world, particularly the United Kingdom, Singapore, China and Australia have moved (relatively) quickly to capitalize on the US's position, with a combination of clear political support for human ES work and significant scientific programs. In November 2004, California voted through a $3-billion bond to support human ES work in the state, but none of that money has yet been spent. In any case, with so much in this nascent field still dependent on know-how and relatively irreproducible technical skills, some have estimated it could take five years for the United States to regain momentum in the area, even if Bush's ban was to be reversed tomorrow.
With the United States a mosaic of attitudes at best, the world has lost its biggest potential single market for stem cell therapies. Europe, with market-orientation overlaid on a backdrop of variegated religious adherence has deferred market approvals of ES cell therapies to the national level. Without the US single market to propel Europe to greater coherence and common sense, the use of the approach is likely to remain patchy across the Northern Hemisphere.
Gaffe 6: Germany's BioRegio competition
Until 1995, Germany had been the pariah state of the European biotech endeavor. It was the center of protests about plant biotech, chemicals/pharmaceutical giant Hoechst had been prevented for over a decade from developing its recombinant insulin plant in Frankfurt and the nation was highly sensitized by history against the accumulation of genetic information and genetic experimentation. Then, in 1996, Federal Research Minister Juergen Ruettgers launched the BioRegio competition. Seventeen of Germany's regions were awarded $65,000 to prepare proposals that would "accelerate the commercial development of biotechnology in Germany." There would be three winning regions, each of which would receive $30–35 million in direct subventions over five years in order to execute their plans, as well as preferential access to matched funds aimed at biotech startups.
BioRegio worked like a dream. All over Germany, life science companies sprang into being. Around 150 started in 1997 alone and by 2000, Germany had over 600 firms. A very high proportion of them attracted investment, with seed investors urged on by matching state and federal investments and loans that could triple the venture capitalist's initial outlay. The global investment boom in 1999/2000 led to the formation of a German public market for technology stocks, the Neue Markt, and for some, at least, the virtuous investment cycle could be completed.
But by 2001, an increasing number of German companies were on the slide. Having peaked in March 2000, the market lost over half its value in the next six months and eventually lost 97% of its value before the Deutsche Borse phased it out. With the public market gone, any lingering doubts that venture capitalists in Germany would invest in more startups in the absence of further federal or state incentives were dispelled. The fledgling companies founded in the BioRegio years would have to fend for themselves.
The fallout. Virtually all the venture capital invested in German biotech since 2000 has gone into companies founded between 1997 and 2000 as venture investors have protected their own interests (or attempted to cover their tracks, depending on your point of view). The take-home message is that it is easy to dispense free money to scientific enterprises but creating sustainable businesses around a core technical idea requires investor persistence and careful management, preferably by people who have some relevant experience. This lesson was learned anew in the United Kingdom, Germany, Denmark and Sweden, all of which were apparently unaware that the United States had been through it all several years before. The Netherlands, Portugal, Spain and many other European nations are now making the same mistakes.
Gaffe 7: the British Biotech saga
In May 1996, a company with no products on, or even near, the market became the fourth highest valued biotech stock behind Genentech, Amgen and Chiron. In that month, British Biotech's (Oxford, UK) stock price all but reached $60, making the company temporarily worth nearly $2.5 billion. All of the investors' chips were being placed on British Biotech's lead compound, marimastat, an orally available matrix metalloproteinase inhibitor (MMPI) that had shown some promise in metastatic pancreatic cancer. A lethal combination, then, of a company with no market track record and a lead compound in a new drug class with a poorly understood mechanism of action.
The fallout. The inevitable unpleasantness happened, but in a rather protracted fashion in the first few months of 1998. A long-running feud broke out involving the company's senior management team, the company's head of clinical research, Andrew Millar, and institutional investors in the company. The senior management team refused to unblind ongoing trials of marimastat, despite indications that the drug was not performing well. By May 1998, things got worse as Millar was dismissed from his post after taking his rather pessimistic views of the clinical chances of marimastat and a second British Biotech drug, lexipafant for acute pancreatitis, to the Financial Times newspaper. British Biotech tried to counter this bad publicity through a 31-page statement issued to shareholders, although the statement was made somewhat less convincing by the resignation of its CEO, Keith McCullagh, on the same day. By this time, the company's stock price had fallen to 1.5% of its earlier peak value.
The lesson: live by volatility, die by volatility.
Gaffe 8: Hoechst Marion Roussel passes up Velcade
Velcade (bortezomib) is now a significant drug for Millennium Pharmaceuticals (Cambridge, MA, USA). In multiple myeloma, sales of the drug ($50.9 million) in the third quarter of 2005 were up 35% over the same quarter in the previous year. Millennium now has the drug in a variety of phase 3 trials in solid cancers, including prostate, lung, breast and ovarian cancer, as well as in hematological cancers, such as leukemia. But the product hasn't always been that successful, and it hasn't always been Millennium's.
The drug started life as PS-341, a molecule made in 1995 by chemists at a Cambridge, Massachusetts company called Proscript and designed to interact with the proteasome. In November 1995, Proscript signed a $38-million development deal with Hoechst Marion Roussel (HMR; which later merged with Rhone-Poulenc Rorer to become Aventis (Strasbourg, France)) around ProScript's small molecule proteasome inhibitors. That deal was one of a number of relationships that HMR formed in the early 1990s recognizing the paucity of its innovative machinery and development pipeline. However, HMR's acquisitive phase didn't last long. In February 1997, ProScript reacquired all rights to the proteasome inhibitors as HMR began 'rationalizing' its biotech relationships in preparation for the merger with Rhone-Poulenc Rorer.
The fallout. This was definitely a double-edged sword for Proscript. On the one hand, it enabled the company to bank a valuable asset: PS-341 turned out to shrink tumors in mice with lung cancer and later, in 1998, successfully came through human toxicity trials funded by $75,000 from junk-bond king Michael Milken's Prostate Cancer Foundation.
However, losing its key pharma collaborator meant that by 1999, Proscript was out of cash and without a clear development path for its products. Having put in around $13 million since its foundation in 1992, Proscript's venture capitalists, HealthCare Investment and Dillon Read Venture Capital, sold the company to another of its portfolio companies, Cambridge, Massachusetts-based LeukoSite, for shares and cash worth $2.7 million. PS-341 thus became LDP-341, but only for four months. At that point, still in 1999, Millennium stepped in for LeukoSite as part of its product acquisition strategy and LDP-341 became MLN341. It still took another 4–5 years of clinical development before Velcade got approval in multiple myeloma but that's a relatively short time in clinical development.
Take-home message: there's gold in them there molecules, but sometimes the detail gets lost in the strategy.
Gaffe 9: Lax protection of personal biological data
The pioneering work of Alec Jeffries at the University of Leicester, UK, brought the world 'genetic fingerprinting' and the tool has been an important part of criminal forensics ever since. Jeffries' restriction fragment length polymorphism (RFLP) method is an impressive way of establishing identity, of demonstrating to a high degree of certainty that a biological sample (saliva, sperm, blood, hair) found at a crime scene was left by a particular person and therefore that that person was at the scene, despite the alibis they may have constructed.
The authorities contend that these kind of data are far too valuable in the fight against crime simply to throw them away, and they are stored in national DNA data collections. The UK's National DNA Database (NDNAD) is the largest of these in the world, holding DNA profiles from almost 3 million people (roughly 10% of the male population as 98% of the records are from men). The data help solve around 15 murders, 45 sexual offences and 2,500 other crimes per month on average.
The fallout. The problem is that DNA has proved so useful in criminal investigations that the boundaries on its use now raise serious civil liberties arguments. Thanks to police and government lobbying, the world's DNA data banks contain not only the RFLP patterns of criminals but also those of suspected criminals, government employees and even whole swaths of the general public that match the description of suspects in serious crimes such as murder or rape. In 2004, in the United Kingdom and the United States, the databases were allowed to include DNA profiles of people merely arrested on suspicion of an offense. As a result of the 2004 'Justice for All Act,' the Federal Bureau of Investigation's (FBI) Combined DNA Index System (CODIS) can now contain DNA profiles of virtually anyone charged with any crime.
Furthermore, the limits on the types of data held and the uses to which they can be put are being inexorably extended. The principal database content is still RFLP profiles, but the original biological samples are also now being banked. This means that there are in essence no limits on the range of genetic markers that could be analyzed. Through partial matching of RFLPs, the databases have already been used for 'familial searching,' allowing police to seek blood relations of those whose samples are held and processed. Using a wider range of genetic markers, forensics moves away from identity and towards population characteristics that indicate ethnicity, facial features, stature and other individual features. The UK authorities have already used these characteristics to narrow down the search for a murderer to people whose geographic origins were in a small West Indian community. In the United States, DNA Print Genomics (Sarasota, FL, USA) now offers a service called DNAWitness to "determine genetic heritage from DNA samples obtained from crime scenes, narrowing the potential suspect pool to a more focused group of likely candidates." Although such tools are both technically feasible and may well be a useful forensic aid to police investigation, they send a dubious message. Crime is committed by individuals, but products like DNAWitness point the finger at specific ethnic groups.
Lesson: an objective technological approach has been transformed before our very eyes into a potentially powerful force for social erosion. The public may respect genetic diagnostics' contribution to crime fighting but it will have questions about government promises to limit the uses of similar approaches in health care.
Coda
With gaffes ought to come culpability. Someone is surely responsible for the howlers that cost investors and companies millions of dollars in lost revenue or market capitalization. And in a few of the cases outlined above, a culprit can be identified and punished by imprisonment or disavowal. But what appears to be more common in biotech is that—despite the abundance of brainpower, regardless of the ample sprinkling of PhDs, MDs and MBAs, and notwithstanding the layers of regulatory oversight imposed—there is a huge capacity for mass self-deception. Almost every national biotech sector started as a sprinkling of startup companies, the vast majority of which quickly ran out of money without achieving anything. And yet every subsequent nation starts lemming-like down the same doomed path. In every investment boom we start off with the same unrealistic expectations and exhibit the same symptoms of shock and disbelief when economic reality dawns, or when we find out that bench researchers are not necessarily the right people to run businesses after all. Each product developer overlooks the obvious obstacles met by innovative products and sees only the theoretical therapeutic upsides.
Although we can never entirely condone the irrational and the unreasonable, perhaps biotech can only exist because we are all willing to indulge in a mass suspension of disbelief, to relax our critical facilities to allow ourselves to believe—despite all the reasons to the contrary—that we can create something that simply did not exist before.
Published online: 8 March 2006.
|
