Starpharma : The Third Installment

Over the past two weeks I copped quite a bit of email flack from devout Starpharma (ASX : SPL) shareholders in response to my last (two) posts. Here is a pithy example:

Long Tail, you are just a downramper and a f—–g idiot. You claim to know about science but you can’t see good results when they are shoved in front of your face. Starpharma is a great company and if they say that their data is exciting, I believe them and not you. Even if I don’t understand all the details. Your site a farce and you should just shut the f—ck up.

Gee. Thanks… er mate.

Frankly, the lax culture of scientific integrity that we have in this country in relation to ASX biosciences is appalling. It starts “at the top” with the leadership of ASX-listed companies that care far more about putting out some news to satisfy a few impatient whiners on HotCopper, then caring about whether the news is any good or not. And, like a bunch of lemmings, everyone else follows – including people that should know better and care more than they apparently do. For example, several people forwarded me an extract from Bioshares edition 626 in which the company is reported as follows:

In results announced this week, the ovarian cancer model study delivered some interesting findings, including complete tumour regression at 60 days, convincingly outperforming Herceptin alone and Kadcyla, an antibody conjugate which combines Herceptin with the tubulin inhibitor DM1. All DEP-antibody treated mice were alive at the end of the 60 day study period, whereas all other mice in the study were dead by day 50.

… and…

[Jackie] Fairley said this result was “very exciting”. She said the data had been previously shown to Roche, the marketers of both Herceptin and Kadcyla, who were “blown away” by the results. She also said that was true of all companies that had been shown the data.

I bet Roche was blown away.

I would be blown away too if someone had run a lightweight pre-clinical study in a single xenograft model that suggests – incredibly strongly – that a drug that has been previously shown to have considerable effect on tumour growth doesn’t actually work better than saline. I mean seriously, take a look for yourself:

SPL's data from their ASX announcement is dubious because the anti-Her2 treatment arms and saline (control) have the same tumour growth profile in a model that is known to response well to both Herceptin and Kadcyla. Why is this?

SPL’s data from their ASX announcement is highly dubious because the anti-Her2 treatment arms and saline (control) have the same tumour growth profile in a model that is known to response well to both Herceptin and Kadcyla. Why is this? Please explain?


Trastuzumab (Herceptin) has three proposed mechansisms of action in non-drug resistant Her2+ cells. Two out of three of those mechanisms are likely not available in a SKOV3 / NOD-SCID model, though the dominant mechanism (Her2 signalling) works perfectly well and does impact tumour growth. Trastuzumab-DM1 (Kadcyla) does essentially the same thing as Herceptin, but also adds a conjugated cytotoxin (the DM1) that notionally does its business when the mAb is internalised into the cell. In xenograft models, Kadcyla has been found to be comparatively more potent than Herceptin, including in SKOV3 models.

So then why has Starpharma released a data set that suggests that Herceptin, Kadcyla and saline (aka salt water) have the same level of efficacy?

Well there are a few reasonable explanations.

The first is that the company actually wanted to compare the DEP technology to a sub-therapeutic administration schedule of the Roche products, because that’s what the data fairly clearly suggests they did. If the comparator drug(s) don’t work better than saline, it means either model isn’t responsive to the drug(s) or there isn’t enough drug to perturb the model. It’s just logic. Since the model is generally known to respond to the drug(s), a non-therapeutic regimen is a reasonable hypothesis.

But then if that is case, don’t claim that your drug beats “leading” drugs in your shitty model, because that’s not what the data suggests.

The second explanation is that the company genuinely believed that 20mg/kg dose of comparator drug(s) once a week is a treatment dose. So maybe Starpharma’s scientists just really aren’t that crash hot and were so busy high-fiving each other about the Herceptin/Kadcyla growth curves they just forgot about the controls. Or maybe they didn’t bother to do their research and properly understand what sort of loading and maintenance doses are really required to make the SKOV3 model work with Herceptin. Surely if you are going to compare your drug against the “leading” drugs you would at least try to reproduce the experiments that show how the “leading” drugs work? You know, a little experimental preparation?

The truth is, there is so much that is scientifically underwhelming about what was presented – the lack of real controls for the DEP product (i.e. with and without drug loading), a very small number of animals, a single xenograft model (normally you would do several different types), lack of histology, lack of mechanistic validation (like where do the dendrimers go?) … it’s hard to imagine that any scientist worth his or her salt would let data like that go out into the public domain with these sort of claims attached. It is my earnest hope that there are more than a few glum looking Starpharma scientists walking around the corridors of the company feeling stupid at the moment.

You should be feeling stupid.

Yes you.

And if those Starpharma scientists are not feeling stupid, they should be feeling angry with their management team for prematurely hyping data. They should be angry that their company has a culture of releasing half-assed data, before it is ready (to vamp an AGM meeting no less), which doesn’t make any sense and that only serves to undermine their scientific credibility and that of the company. They should also be angry that there hasn’t been either a retraction of the data or a clarification. If there is more data, or there is an error in the data, just fix it. But don’t think that a lot of glowing media soundbites obfuscates the fact that the data, as presented, is crap.

I don’t mind the DEP technology. I can see a game plan for it. The fact that AstraZeneca wants to do more with the company is good news and very encouraging for shareholders. But for goodness sake, if you want to be a world-class pharmaceutical company, start acting like one and show some scientific integrity.

… and if you made a mistake, just own it.

Image credit : Gratisgraphy.


4 thoughts on “Starpharma : The Third Installment

  1. Well said. There are far too many bio and tech stocks that play to the mining culture of share investment in Aus. It would be a good presentation to take the top and bottom 10 bio stocks of 10 years ago and plot out the results, showing the SP on phase 2 and 3 clinical trial results. It may assist newer investors to gain some insight into the risks and rewards.


  2. Thanks Chris. Whilst I’m a Spl believer I did think it a little odd that the other drugs weren’t significantly better than salt water.


  3. Chris, an excellent commentary. It’s disappointing that some investors don’t want to read another view of the data and only believe with fanatical fervour what has been disclosed in a press release. You may have mentioned this in your other posts but the real issue for me with the announcement is that this is just a bog-standard xenograft study. We all know these days that xenograft studies (in immunocompromised mice using immortalised cells that only partially reflect properties of ‘real’ cancer cells) are flawed in many ways and poorly predict clinical efficacy. Some useful information can be gleaned from these experiments but for more clinically relevant data, patient-derived xenografts (using real human cancer tissue) would give an improved indication of efficacy compared with standard-of-care drugs. Such models are now available for ovarian cancer.
    I actually like SPL but I agree this announcement reflects poorly on their technology.


Say something useful (or at least interesting)...

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s