Replies to post #23752 on Biotech Values

[DewDiligence]

>The second is questionable and very problematic from a statistical standpoint unless their SAP with the FDA says something like "open the database when at least XXX patients" have had the event. Most SAP's I've read specify a specific number. (Others here can chime in if the SAPs they've done/read are different from my experience.)<

The SPARC Satraplatin trial is one case where the company (GPCB) increased the event trigger in order to delay the unblinding.

[terry hallinan]

BSR_David,

PLEASE, PLEASE, PLEASE take what happened to CTIC investors as a caution against believing "longer is better."

I believe it.

Just can't seem to get out of my head that living longer is good for cancer patients.

AGEN's enrollees in its pivotal RCC trial are way late (the trigger point is based on progression as well as death).

There is a complication.

The trigger point was reached last fall.

So what the....?

The excuse or reason, whichever grabs you, is that there are questions about protocol violations with some enrollees being metastatic. X-rays are being studied by outside consultants.

Possible to see lots of intrigue involved or extreme care.

Often it is best to take the darkest possible view of things no doubt.

Still the triggering number is way late in any event. About a year late from recall.

That should be good. No guarantee, of course. There is the usual problem with lying stats.

In the end you have to make a judgment on the people involved doing the right thing.

And that is a tough call always.

Best, Terry

[jellybean]

David, do you cover Telk in your newsletter?

[poorgradstudent]

>the only prospective statistically significant positive conclusion you can gain from those trials is you should go to Russia to be treated if you have lung cancer.<

I thought the stellar trials showed that if you had cancer, you were better off to have had it outside the US.

But if you wanted to be treated for cancer, then you were better off in the US.

Bottom line is that observations like the above are descriptive of poorly run trials.

Another interesting one was:

If you're going to be treated for melanoma using genasense, you're better off in Australia.

Briefing documents are fun.

[poorgradstudent]

re: Telk

This is one of those annoying situations that demonstrates an investor's lack of power.

By giving the unnecessarily drawn out 12-16 week timeframe for data analysis, Telik guaranteed that they could announce all the results at once regardless of when the trials end.

It comes down to whether or not management can be trusted. Problem is, telik has had it easy up till now. No adversity.

The possibilities i see:

1) they are truthful regarding the event accruals,

2) they've continually amended their SAPs to push back the trigger numbers in assist-1 and 2, thereby making their comments that "the trigger has not been hit" technically true but clearly misleading.

3) they're outright lying about the triggers not being hit.

And I agree about the ctic caution in essence; it has merit. Problem is that the "longer is better" observation is pretty much a necessity for these types of trials to be successful. It's just not sufficient. But that's the fun part right?