Replies to post #16346 on Biotech Values

[zebra4o1]

Great to see some hard numbers on this phenomena: 81% of the phase III's had a lower response rate. But what a strange excercise: doing statistical checks for factors predicting sucess or failure. The answer is so obvious, but maybe not amenable to a simple statistical analysis. I wondered if they checked success rates versus the size to the company.

[AlohaDan]

chemotherapy in advanced solid malignancies

That's a key for your highlite on Phase II vs Phase III.

Don't let the later standup for other drug apps.

[terry hallinan]

DewDiligence,

[extracted from quoted study:]
The vast majority (81%) of phase III studies have lower response rates than preceding phase II studies, with a mean difference of 12.9% among all studies analyzed.

Proving once again that response rate has dubious meaning. Every purveyor of quack remedies has grand stories of tumor shrinkage - normally the cancer just went away according to the pitchman.

At the PRW annual meeting, a shareholder asked about the RR from the gawdawful long Phase I and got hemming and hawing and quibbling from the CEO. An independent director from Harvard stood up and put a merciful end to the pain. (I don't recall his name or connection but I was impressed by the fellow despite his dubious association.)

In a rather lengthy dissertation, the director made the claim that the despised stabilization rate was a far more valid statistic, essentially saying that people who were dying stopped dying. Response rate in advanced stages of cancer seemed something like measuring the health of a forest by the number of downed trees.

As you might guess, the eager flock of MB faithful hungered for word on that RR. With that all would be well. The wait continues for the magical fix. the CEO said one might possibly be forthcoming if they can find a statistician with a handy bent for imaginary math with a sliding scale.

GLGS put a new twist on things with announcement that their newly non-alcoholic elixir showed a response rate with "periods of stable disease" in a Phase I.

[sigh]

I still like that Harvard guy. If it was just a barker's spiel, at least it was a good one though it will cut no ice with the SEC or the Street people.

The theory advanced and "proven" before concerning the decline in RR in Phase III's is that the selected drug from among the contenders is the one that, by chance, had the best RR and subsequently reverts to the mean. It is completely logical and my guess is too pat. When an engineer gets different measures from the same ruler, he looks to a problem with the measure. The biologist looks to a problem with the patient and the Street people look to a problem with the CEO.

Best, Terry