Avastin: accelerated approval example

[Guest guest]

Greetings,

New information about Avastin arrived today, illustrating why interim

endpoints (such as response rate) used to win Accelerated Approval can

SOMETIMES lead to false conclusions - why additional study is needed to see

if the interim measures of efficacy (surrogate endpoints) translates into

real benefit - survival benefit . and if other serious risks from taking the

new drug emerge.

This is not to suggest that Accelerated Approval pathway is not a good way

to more quickly fill the urgent unmet needs of patients with no satisfactory

treatment options - only to remind that the approval is conditional - and

why.

This is not to suggest that the risks that came to light today from the

study of Avastin might not be worth taking - that would depend on the risk

of disease Avastin treats and how well Avastin manages that disease. That

is, the study drug might still provide a NET benefit even if it causes

serious problems . so it depends . and each disease is unique regarding how

much risk is reasonable to accept when treating it.

However, in this case, for the treatment of breast cancer, follow-up showed

NO survival benefit for Avastin when added to standard therapy, . compared

to the standard protocol - and that the added risks from Avastin were not

insignificant. (See the ASCO Post for a lively discussion - which shows (I

feel) that physicians are not always " up " on the pitfalls of observation

guiding clinical practice. http://bit.ly/h3kXlq )

In addition, for Avastin, the risk of heart failure came to light --

information that could have emerged only by comparing Avastin to a placebo -

because in regular clinical practice individual adverse events cannot be

evaluated in this way -- to show a rate of increased risk compared to

something else.

So I think the important question was if the " bar " was set too low for the

accelerated approval of Avastin for breast cancer? And was the height of

the bar determined by good science or in part by political pressures (such

as sponsor-goaded patient advocacy)? For example, was the predetermined

magnitude of the differences in efficacy (needed to win accelerated

approval) too low so that it could not reliably predict a survival benefit?

Finally to the news item:

" Women with advanced breast cancer who were treated with Roche's Avastin

were more likely to develop heart failure than other women, according to an

analysis released Tuesday that raised more concerns about the already

troubled drug.

A team led by Dr. Toni Choueiri of Dana Farber Cancer Institute and Harvard

Medical School in Boston analyzed data on advanced breast cancer patients

from five clinical trials.

They found that 1.6 percent of the women who took Avastin developed

congestive heart failure compared with 0.4 percent of women who got a

placebo.

http://www.msnbc.msn.com/id/40916007/ns/health-cancer/

All the best,

~ Karl

Patients Against Lymphoma

Bcc: undisclosed advisors