Collected for research blog post:

Dew:

DNDN – I’ve seen several recent posts on various message boards asserting that Provenge is a shoo-in to be reimbursed for off-label use in early-stage prostate cancer. For the most part, these posts are clueless about the way cancer drugs in the US are reimbursed.

Without explicit support in the NCCN compendium (or one of the three other compendia used by CMS), reimbursement by Medicare is highly unlikely. If Medicare does not reimburse for off-label use of Provenge, it’s unlikely that private insurance will do so.

For Provenge to be included in the NCCN compendium will require compelling efficacy data from at least a large phase-2 study.


Clark:
_______________________________________
For Provenge to be included in the NCCN compendium will require compelling efficacy data from at least a large phase-2 study.
________________________________________

I honestly have no idea what the standard here is. Would stat sig on P-11 "Time to Distant Mets" qualify? Randomized trial, earlier stage, but blind is broken already. Does it have to be the primary endpoint of the trial? Would Time to Distant Mets be a viable endpoint? ...
?


Dew:
DNDN re off-label use of Provenge in early-stage PCa:
Quote:
________________________________________
Would stat sig on P-11 "Time to Distant Mets" qualify? Randomized trial, earlier stage, but blind is broken already. Does it have to be the primary endpoint of the trial? Would Time to Distant Mets be a viable endpoint?
________________________________________
Tough questions to answer. My opinions, for what they’re worth:

• 1. Hitting a secondary endpoint would not carry much weight on its own. (As you know, a secondary endpoint technically has not been met unless the primary endpoint is met and all higher-ranking secondary endpoints, if any, have been met.)

• 2. (Somewhat contradictory to #1): The p-value on the primary endpoint need not be stat sig to garner consideration if the miss is small and the clinical (as about to statistical) significance of the primary efficacy metric is indisputable.

• 3. Premature unblinding of a trial is not a fatal flaw.

• 4. Novel endpoints such as time to distant mets may be acceptable if the strong consensus of practicing clinicians is that the metric in question is clearly consequential for patients (i.e. if the clinicians think it’s intuitively obvious the metric is correlated with survival or improved QoL).

In short, the NCCN thought process tends to be more patient-oriented and less math-oriented than the FDA’s.


David Miller:

For Provenge to be included in the NCCN compendium will require compelling efficacy data from at least a large phase-2 study.

Correct. And even if one believes PROTECT is of sufficient size, it won't provide that answer on an allowable endpoint until well in to next decade.

Guys who want it off label could fork over the significant bucks for it, but I doubt that will be noticeable in the overall sales figures.

The first likely label "extension" won't technically be an extension but a mention of a positive relationship between Taxotere and Provenge (along the lines of the Petrylak retrospective data from 01/02a). A prospective trial designed in 2010 and launched in 2011 could get data by 2012 (or earlier if the Cougar trial validates the CTC marker). That would likely improve sales among oncologists. If adoption among urologists is heavy, however, even this 'extension' will also not be not very significant in terms of overall sales snce most guys will have had Provenge anyway by the time they get to the uro onc.

I should note that label extensions are not necessary to get Provenge to blockbuster status in North America, IMO of course.



Dew:
________________________________________
DNDN: I should note that label extensions are not necessary to get Provenge to blockbuster status in North America, IMO of course.
________________________________________
There may be a significant number of men who will want to use Provenge in early-stage disease even if they have to pay out of pocket to get it. For this reason, DNDN would probably be wise not to overplay their hand and set too high a price.