Data manipulation expert Elisabeth Bik compares the tales of 2 accused biotechs: Cassava and Athira

Alzheimer’s disease research has had a tumultuous year. First, patients finally got a new treatment when the FDA approved Biogen’s Aduhelm, setting off a chorus of criticism and challenges for the company.

Then there were multiple allegations of data manipulation against smaller biotechs—Cassava Sciences being the most dramatic, as well as Athira Pharma. The latter ended up firing its CEO Leen Kawas, Ph.D., after an independent special committee hired by the company determined she had altered images in her 2011 doctoral dissertation and in at least four co-authored research papers during her graduate studies at Washington State University.

As for Cassava, the embattled company and its CEO Remi Barbier continue to rail against “outlandish allegations” made by short sellers. The company faced an FDA citizen petition in August that questioned the scientific integrity of clinical trials conducted for simufilam. Since then, the Journal of Neurology reviewed raw data from the disputed paper and found no evidence of manipulation, Barbier pointed out in a statement to Fierce Biotech.

Fierce Biotech caught up with one of the scientists who spotted irregularities in Cassava’s research after the petition raised those questions. Elisabeth Bik is a private microbiologist and science integrity consultant specializing in image manipulation who received her Ph.D. from Utrecht University in the Netherlands. She’s now based in the U.S.

RELATED: Cassava ploughs on with 2nd phase 3 trial of Alzheimer's med despite federal investigation

Bik has no involvement in the citizen petition and does not hold any stock in Cassava; the issue with the studies was brought to her attention via her Twitter following. She gives her assessment of which company did things correctly and how the other can right its wrongs. 

This interview has been lightly edited for clarity and brevity. 

Fierce Biotech: What are some common issues or problems that you see in your daily work?

Elisabeth Bik: Most of the problems that I see are either in Western Blots—which are those protein molds that sort of look like black stripes or horizontal stripes—or in microscopy images of tissues of cells, sometimes microscopy images of tumors or on mice. So those would be things visible with the naked eye. So those types of images you would mainly find in scientific papers that are in the biomedical/molecular biology field.

FB: Can you tell me what you found when you looked at Cassava’s studies?

EB: There were several statements in this petition, and one of them was about the Western Blots, so I looked at the allegations. I looked up the original photos, and I could confirm that there seemed to be cases where there were bands that look extremely similar to each other. But some of these blots were very low resolution, and they cranked up the contrast a lot so you see these sort of extremely black bands or extremely white background. There's no noise or background to look at, and so it makes it a little bit harder to know for certain if these blots are duplicated or these bands were duplicated. But for sure it looked like they were more similar than I would have expected.

RELATED: Athira CEO Kawas officially out after investigation found she did alter grad research

And I also found in these same papers some additional problems that have not been mentioned in the petition. So some of these bands appeared to be surrounded by a rectangle where the background is a little bit different than the rest of the blocks, suggesting that that band might not have been part of the original photo but was added in later.

I'm not saying that that really happened. But just by looking at these photos, it appears that there was some manipulation of some kind, and one would hope to see the original blots, because that could take away all the concerns.

FB: When you see something like that in a study, what does that mean for the outcome? Does it bring into question the results?

EB: It does, in my opinion. For me, science is about finding the truth. And of course we have the hypothesis of how, what the outcome would be of a particular experiment. But if the experiment has a different outcome, one should not make the results fit the hypothesis; it should be the other way around.

RELATED: Cassava CEO, 'dazed and confused,' takes to biotech defense handbook by blaming short sellers for woes

So when … in a scientific paper, you find evidence or at least a suspicion of photographic manipulation, it might just be a control blot, but it does raise serious concerns about how honest the researcher is.

So for me, even Photoshopping in a very minor experiment raises general questions about all the data in a paper.

FB: Why have we seen so many instances of manipulation concerns in Alzheimer’s specifically?

EB: I don't think it specifically happens in Alzheimer's. I guess Alzheimer's is a particularly interesting topic … a lot of people are interested in finding a drug that works because so far there hasn't been anything on the market; nothing seems to work. It's a disease that affects millions, billions of people. There's an interest in this particular disease, but there is also in cancer and other diseases, and so I think image manipulations happen everywhere.

FB: Is research conducted by biotech companies developing drugs held to the same scientific standards as a public institution?

EB: Yes, it should be. I think any paper that comes out in a scientific journal should be honest, because, like I said, science is about finding the truth. And I feel science should be a field where we can trust each other and we can cite each other's papers and build out our research on previous papers. So it doesn't matter if the authors are working in a biotech company or doing research at a university.

RELATED: Cassava defends Alzheimer's science after citizen petition raises 'grave concerns' about clinical studies

This Cassava research was NIH-funded, so it was sort of an academic researcher working for a biotech company, which makes it even more complex. But it should not matter who does the research. It should be honest and trustworthy.

FB: How do companies or researchers come back from something like this? Can they say, ‘Hey we realized this was a mistake'? How would you advise a company in this situation?

EB: I think Athira did a great job addressing the problem. They said, 'We have a serious problem here.' The CEO was immediately put on leave. And in that particular case, the research she did during her Ph.D., which was the research that was questioned, wasn't directly involved with the Athira research. I think Athira handled this really well, and kudos to them. They didn't dismiss it, they took action, and that's how this should be done. This is the only way to do it.

Cassava, on the other hand, had an immediate response where they completely dismissed the problems that were raised in the citizens petition. They released this statement where it had like facts and fiction. Basically, they called all these allegations fiction, and then they had facts, and some of these facts were clearly, in my opinion, fiction, and it was not written by a person who has any knowledge about molecular biology or about blots in general. It was very dismissive of all the concerns while, in my opinion, the concerns had merit.

RELATED: Lab named by Cassava to back up Alzheimer's data says it wasn't involved, sending biotech's shares into freefall again

The only way to dismiss the concerns is by showing the original blots and saying, 'Yes, they look similar, but here are the high-resolution scans, you can clearly see they're different bands.'

I feel the company has missed an opportunity to show that they care about science. They were very dismissive; they made it look like these were ridiculous statements. In my opinion, that made the case even worse.

Editor's note: Responding to questions from Fierce Biotech about whether Cassava would publicly release the Western blots as Bik suggests, Barbier said the company supports "productive disagreements. That’s how science advances. ... The editor of the Journal of Neuroscience put your question to bed."