A biotech fraud scandal

Earlier this week, the genetic analysis company Sequenom announced that it had fired its CEO and other employees over dodgy data for the company's blood test for Down syndrome. The test was meant to be launched last June, but the data supporting it must now be verified by independent investigators before it's deemed trustworthy.

"…the test data and results in the company's Trisomy 21 program included inadequately substantiated claims, inconsistencies and errors. Due to deficiencies in the company's disclosure controls and procedures, in a number of instances such test data and results were reported to the public in the company's press releases and other public statements," the company said in a release.

Though there's plenty being said about academic fraud, we don't really hear much about blatant fabrication of data in the biopharma industry. To some extent, this makes sense -- being that much closer to marketing a drug or device provides that much more incentive for the data to be real. But is there more data "mishandling" going on in industry than we hear about? And are the mechanisms for detecting, reporting and punishing fraud as strong as in academia -- where, of course, plenty of people think they are not strong enough?

-Alla Katsnelson, news editor, The Scientist

In my experience, massaging of data occurs as often in large pharma as in academia, but review of data is much more rigorous. Bad data get caught early in pharma and those who produce it are isolated from important projects.

To amplify on the earlier response, outside regulatory scrutiny (like FDA) along with the possibility of job loss, are higher barriers to data manipulation in industry. Also, ongoing monitoring of adverse reaction reports and legal liability associated with products in wide use increase the probability of eventual discovery of fraud, which is also a stronger deterrent in industrial cases than in academia. I doubt that personal integrity is higher in one area than another, but the scrutiny and corrective mechanisms can be quite different.

Robert, I totally agree, as one who's worked in both milieux.
Also in pharma (can't speak for biotech) one is part of a team (biologist, chemist, toxicologist, etc) and one presents data at staff meetings, often weekly. Helps to keep folks on the straight and narrow, knowing there's quite a bit of oversight.

My experience in industry is similar to that of Lisa's. The main difference in industries such as pharma is that the data undergoes much more rigorous scrutiny, usually by colleagues working downstream and on some occasions by internal or external auditors/reviewers. Fabrication and falsification of data also carry much more severe penalties for the employers, who must bear the costs of litigation, loss of patents and other serious costs (both economic and non-economic). As a result, when individuals who engage in such practices are discovered, they are usually dealt with swiftly and in no uncertain terms. Acts of misconduct also tend to trickle up the management chain as well, depending on both the severity of the act and who discovers it. In most companies, the line management is held responsible for the acts of their subordinates, and justifiably so. That doesn't seem to be the case in academia.

I think we're talking about apples and oranges here. Academic fraud occurs for totally different reasons and in completely different environs than biotech/pharma fraud, and those reasons make it easier to catch academic fraud - but I would be very surprised if fraud weren't equal in both venues.

Academic fraud is at its heart about prestige, grant money, and career advancement by individual researchers, and it takes place in a setting where experimental data are often (though not always) generated in isolation from oversight, making it very easy and rather tempting to fudge results that aren't going to support a cherished (or important) hypothesis. The researcher may not even be cognizant of committing fraud but may justify the fudging based on his or her belief that the results are somehow not truly reflective of what's going on (self-deception at its worst). The goals of such fraud are essentially personal and short-lived: to gain advancement, or adulation, or some other form of personal recognition/recompense. But the nature of academia is that recognition mandates dissemination of findings, which leaves them open to peer scrutiny... and that's where the house of cards usually collapses. When a dozen colleagues working on the same problem pick apart and examine every last bit of a fraudster's published data (and not just once, but many times), that fraudster has much less chance of squeaking by the intense level of interest.

Pharma/biotech fraud, on the other hand, is related to economic pressure to obtain results that will support a high-value product for a large-scale, long-term business interest. It usually involves a conscious decision to suppress concerns about a product by not one but a variety of people involved in the corporate heirarchy of the company. Research data are rarely publicized, and then generally only if they are positive or neutral with respect for the intended product (although truly spectacular failures do sometimes see the light of day). There also seems to be considerably less effort made to make studies statistically robust or representative. All of which makes it more difficult to pinpoint areas of data fudging... until the product has been on the market for 5 years and increasing problems show up in the peer-reviewed literature, forcing a re-examination of the original data... and we find that the biotech/pharma in question knew all about the adverse effects but downplayed or squashed the study data (Vioxx, for example... but it's hardly the only one.) But where the problems with the product are limited, the fraud may well slip by undetected--if the number of people affected are few or if the effects are minor (skin rash, joint pain, stuff that could be caused by just about anything) then data showing that these problems were known but not reported could easily get buried and no one ever the wiser... keeping in mind that particularly with a highly successful product, there's incentive to keep any adverse impacts downplayed as much as possible until the product leaves patent.

I wonder whether the difference in the cultures between academia and industry is still the major factor that hinders more research collaborations between the two, and how it can be ameliorated? Anyone?

To Elizabeth, Well said! I have an active interest in academic fraud. Would like to communicate with you--Ed hipsavers@msn.com

One of my CEO was famous for saying "get my drug to market or get it out of my pipeline". He was right - there is good motivation for taking bad data a a blessing in pharma (stops you spending huge development efforts only to be found out later).

I wish every pharma CEO was as pragmatic as that one, but I don't think it's always cut and dried like that. Some drugs have such fantastic potential for profit that there's incentive to keep it IN the pipeline even when the benefits are only moderate and/or side effect profile is what many people would consider unacceptable. The one benefit to working for pharma, though, is that until the product gets to the thumbs-up-or-down phase of development, the scientists themselves have little incentive to fudge their data, as several of the prior commenters noted. It's once the product is in the range of the marketing department that you start to see "spin" applied.

I think you've hit the nail on the head, Elizabeth. Distortion in publications from big pharma comes, IMO, from the participation of Marketing in publications planning. In my experience, the basic science done in big pharma is of much higher quality than the basic science done at universities. It should be. We are almost all PhDs, as opposed to mostly graduate students, planning of experiments is much more thorough, and repeated vetting of data and interpretations is the norm. The high quality of basic science in big pharma is one of the reasons why I work there.

Oh, it's totally clear to me that pharma fraud is largely related to marketing efforts - there are too many controls in place for the scientists to fudge their data, too many people looking over their shoulders and not enough incentives to fake it (lots of disincentives too). Most of the "fraud" isn't even so much pure fraud, i.e. fabricating data, as it is suppressing data that reflects unfavorably on the product. That's why I said it was apples and oranges - the situations that lead to dishonest behavior are completely different. Similar pathologies, but ultimately not the same disease because the causes are unrelated. Kind of an ethical diabetes, type 1 and type 2!

Does anyone know of an expert legal witness on these matters?-ed

There is numerous data "mishandling" going on in both Chinese industry and academia, especially plants with tradditional Chinese medicine-based drugs. Science reported Chinese fabricated paper data is just the tip of the iceberg. http://www.sciencemag.org/cgi/content/full/323/5919/1280

This case reminds me of one of the many biotech start-ups. It seems that this last highly-publicised scandal is doing some good.
_____________________________________________________________
The latest Fluorescent microscope in Toronto, Microscopy Accessories in Toronto, and Microscopy Accessories in Toronto