Pharma’s Cutting Edge

Avastin Distribution Change - Open Letter

I read today’s WSJ Health Blog on Avastin thinking that Genentech had appeased eye specialists by maintaining distribution of Avastin to compounding pharmacies, but after reading the above open letter, I’m confused about Genentech’s actions in the recent past and about what it’s planning to do in the near future.  Take a look and let me know if you share in my confusion.

The decision we communicated [to cease Avastin distribution to compounding pharmacies] was not made lightly. In fact, it was guided by our company’s strong commitment to take actions that are scientifically and clinically sound and in the best long-term interest of patients, while at the same time adhering to government regulations and remaining mindful of the retinal community’s views.

The above open-letter snippet suggests that Genentech was motivated by their concerns around the safety of Avastin formulated for intravenous delivery being delivered ocularly.  They make this explicit later:

Genentech’s decision was not motivated by a desire for increased profits. We did not and do not expect that this change in policy toward compounding pharmacies will lead to any increase in LUCENTIS® (ranibizumab injection) sales. Further, we expect Avastin to be available and that physicians will continue to prescribe it for ocular indications.

So far, so good.  Genentech is making a logical argument.  Essentially it’s saying that, like Hippocrates himself, Genentech aims to first do no harm.  Fair enough, although I think it’s disingenuous of a for-profit company to imply that an important decision can be divorced from a business case, but I’ll leave that for a future discussion.

Then we get the news that FDA had inspected Genentech’s manufacturing facilities (just when, relative to the company’s compounding-pharmacy decision, we’re not told), and Genentech reacted:

…[W]e destroyed four batches of Avastin deemed unsuitable for use in the eye due to a higher visual inspection standard. (These lots would have been entirely suitable for its approved use as an intravenous cancer medication.) The action resulted in the loss of more than 350,000 vials of Avastin with a market value of more than $200 million.

Now I’m lost.  Why would Genentech have to destroy $200 million of Avastin stock intended for use as a cancer therapy (albeit used off-label for other uses, included ocular uses) to appease FDA inspectors?  Is Genentech saying that FDA mandated that Genentech produce a product capable of withstanding the same visual-inspection standards required of ocular drugs and that Genentech relented to this mandate without fighting it?  Is that what Genentech is saying?  Okay, let’s say that’s it for arguments sake.  What happens next?

The letter implies that sometime after this self-imposed drug destruction, Genentech made its compounding pharmacy decision and soon after that met with key opinion leaders to revisit the decision (N.B. no mention was made of FDA’s involvement at this KOL meeting).  Genentech communicates publicly after this KOL meeting that it will continue to supply Avastin to compounders until at least January 2008 and maybe longer if FDA approves. 

In contrast to what was earlier implied by the juxtaposition of the FDA manufacturing/packaging audit and destruction of drug, this latter sequence of events suggests that Genentech’s Avastin destruction was voluntary, not FDA imposed.  Why else would Genentech be able to decide unilaterally to continue supplying Avastin to compounding pharmacies–without FDA approval to do so–beyond November, the original date of its distribution cessation?  If FDA had ordered destruction of Avastin they would have also, it would seem, had to approve continuation of its supply to compounding pharmacies.  Yet, no mention is made of such approval.  Indeed, Genentech states that it will continue shipment of drug to compounding pharmacies after January only with FDA approval, again suggesting that FDA is having some say in how Avastin is distributed (or packaged for distribution to compounding pharmacies).

So, I’m left with a bunch of questions that leave me wondering what really happened here and what is going to happen come February 2008:

What did FDA actually find during its audit and what specifically prompted the destruction of a large amount of Avastin intended for use in oncology?  Was the destruction voluntary or FDA-mandated? What alternatives to destroying the drug were considered?

What changes to manufacturing/packaging were made following destruction of the drug, and how will such changes affect distribution to compounding pharmacies? 

Is an FDA ruling on manufacturing/packaging changes both necessary (as implied in the open letter) and sufficient for Genentech to extend distribution of Avastin to compounding pharmacies beyond January 2008?

Is Genentech planning to change the cost of Avastin for compounding pharmacies?  Will there be distinct ocular-specific and oncology-specific formulation/packaging?

When we get some answers to these questions, I’ll post an update.

Disclosure:  I have no equity or other financial stake in Genentech or its direct competitors.

Hey…sorry for the layoff time here at Pharma’s Cutting Edge.  You can breathe a bit easier now that I’m back.  Pun intended.

I’ve covered Exubera’s peri-approval and launch highlights in these pages, so now that Pfizer has pulled its marketing plug amidst ridiculously low sales of the first inhalable insulin I figured I’d a retrospective of those posts to aid the research of lessons-gatherers near and far.

In April 2007, I waxed philosophic as Exubera was hanging on by a thread:

Recognize in yourself your reluctance to change your opinion as evidence mounts to the contrary; the strength of your unwillingness to change your mind is proportional to the strength of your held opinion.  This is hardly a novel insight, as published evidence supporting it dates to at least the 1950’s.  In behavioral circles, it’s known as attitude strength, or as I like to call it, stubborness.  You might not be able to overcome your stubborness, but you might be more willing to hedge your bets if you can admit to it.

Following the launch of Exubera in July 2006, I set the stage for one of the moore interesting drug launches of the last few years:

This will certainly be one of the more interesting product launches in diabetes care ever. I’m intrigued by competing factors clinicians, patients, and third-party payers will weigh when deciding to use/reimburse for the drug: convenience (multiple injections vs. multiple inhalations; syringes and needles vs. an inhalation device), dose tailoring (extreme flexibility vs. limited flexibility), cost (3x to 5x higher for Exubera), toxicity (an increased risk of lung toxicity and lung function testing before use for Exubera), and Pfizer’s powerful position in the industry.

In May  2006 I extensively reviewed the drubbing Pfizer took from the UK’s nice.  As it turned out, apparently, NICE’s views of the relative quality-of-life benefits of Exubera were shared by doctors and their patients in the U.S:

It’s always a tall order to demonstrate ICE [incremental cost-effectiveness] for a new therapy with similar average efficacy as older therapies. Toss in the fact that the new therapy must cost two to three times as much as older therapies in order to meet profit margin requirements, and economists are faced with a nearly untenable situation. It’s rarely in the company’s best interests today to take extraordinary risks to demonstrate ICE convincingly prior to first launch if the risks aren’t necessary to gain marketing approval and at least Tier 3 coverage in the U.S., understanding that reimbursement in countries like UK is unlikely.

Following approval of Exubera in the U.S., I reviewed the contents of its prescribing information for physicians (i.e. its label):

Patients with Underlying Lung Diseases: Unlike the smoking contraindication, lung diseases present a risk to the patient that has not been quantified. Unstable lung disease is a contraindication due to altered absorption of insulin. The average clinician will be confused between a stable and an unstable lung disease, as am I. The risk management program will need to address this source of confusion. Look for this imprecise labeling as a potential area for product liability issues to arise.

Prior to the review by FDA’s EMDAC in September 2005, I discussed some of the briefing information regarding pulmonary safety:

One has to wonder how Pfizer plans to manage the liability risk of Exubera. They’ve appropriately proposed a raft of post-marketing studies designed to better quantify the clinical risks of the drug in a variety of patients, but at the same time they have suggested that Exubera is safe to use in patients with mild to moderate asthma or COPD, in contrast to the internal FDA pulmonary consultant’s review, which indicated that current data were insufficient to draw such a conclusion. This leads me to believe that Pfizer is planning to fight to avoid a contraindication to use in such patients, clearly an anti-conservative position.

Hey…sorry for the layoff time here at Pharma’s Cutting Edge.  You can breathe a bit easier now that I’m back.  Pun intended.

I’ve covered Exubera’s peri-approval and launch highlights in these pages, so now that Pfizer has pulled its marketing plug amidst ridiculously low sales of the first inhalable insulin I figured I’d a retrospective of those posts to aid the research of lessons-gatherers near and far.

In April 2007, I waxed philosophic as Exubera was hanging on by a thread:

Recognize in yourself your reluctance to change your opinion as evidence mounts to the contrary; the strength of your unwillingness to change your mind is proportional to the strength of your held opinion.  This is hardly a novel insight, as published evidence supporting it dates to at least the 1950’s.  In behavioral circles, it’s known as attitude strength, or as I like to call it, stubborness.  You might not be able to overcome your stubborness, but you might be more willing to hedge your bets if you can admit to it.

Following the launch of Exubera in July 2006, I set the stage for one of the moore interesting drug launches of the last few years:

This will certainly be one of the more interesting product launches in diabetes care ever. I’m intrigued by competing factors clinicians, patients, and third-party payers will weigh when deciding to use/reimburse for the drug: convenience (multiple injections vs. multiple inhalations; syringes and needles vs. an inhalation device), dose tailoring (extreme flexibility vs. limited flexibility), cost (3x to 5x higher for Exubera), toxicity (an increased risk of lung toxicity and lung function testing before use for Exubera), and Pfizer’s powerful position in the industry.

In May  2006 I extensively reviewed the drubbing Pfizer took from the UK’s nice.  As it turned out, apparently, NICE’s views of the relative quality-of-life benefits of Exubera were shared by doctors and their patients in the U.S:

It’s always a tall order to demonstrate ICE [incremental cost-effectiveness] for a new therapy with similar average efficacy as older therapies. Toss in the fact that the new therapy must cost two to three times as much as older therapies in order to meet profit margin requirements, and economists are faced with a nearly untenable situation. It’s rarely in the company’s best interests today to take extraordinary risks to demonstrate ICE convincingly prior to first launch if the risks aren’t necessary to gain marketing approval and at least Tier 3 coverage in the U.S., understanding that reimbursement in countries like UK is unlikely.

Following approval of Exubera in the U.S., I reviewed the contents of its prescribing information for physicians (i.e. its label):

Patients with Underlying Lung Diseases: Unlike the smoking contraindication, lung diseases present a risk to the patient that has not been quantified. Unstable lung disease is a contraindication due to altered absorption of insulin. The average clinician will be confused between a stable and an unstable lung disease, as am I. The risk management program will need to address this source of confusion. Look for this imprecise labeling as a potential area for product liability issues to arise.

Prior to the review by FDA’s EMDAC in September 2005, I discussed some of the briefing information regarding pulmonary safety:

One has to wonder how Pfizer plans to manage the liability risk of Exubera. They’ve appropriately proposed a raft of post-marketing studies designed to better quantify the clinical risks of the drug in a variety of patients, but at the same time they have suggested that Exubera is safe to use in patients with mild to moderate asthma or COPD, in contrast to the internal FDA pulmonary consultant’s review, which indicated that current data were insufficient to draw such a conclusion. This leads me to believe that Pfizer is planning to fight to avoid a contraindication to use in such patients, clearly an anti-conservative position.

In my last post, I  discussed what I believed would be the basis for debate and concern among the panelists at Wednesday’s FDA DMEDP Advisory Committee–the benefit to risk balance of the drug in light of modest efficacy and risk of suicidality.  This was indeed the topic of much of the committee’s deliberations.  As you now know, the committee voted unanimously (14-0) against approval of rimonabant due to a perceived unfavorable benefit to risk balance.  They also voted unanimously that Sanofi-Aventis hadn’t sufficiently characterized the safety profile of the drug.

What does this mean?  It means that this FDA division and its outside advisers, and probably much of CDER along with them, is taking the position that they want real, sustained evidence of benefits when risks during chronic use of a drug are small in number but potentially fatal.  Stated differently, risk tolerance is low among this regulatory body, and it wants to see relatively more evidence of benefit to counterweigh a risk of a serious adverse event than in the past (I’m thinking 2 to 3 years ago).  With rimonabant, they looked at a drug intended for chronic use that was associated with modest, reversible benefits on weight and markers of cardiovascular disease; no long-term cardiovascular outcome evidence; tolerance issues that promoted drug discontinuation, and an uncommon but potentially fatal adverse effect.  In the current regulatory climate, rimonabant never stood a chance.

Perhaps DMEDP and its advisers would have thought differently had rimonabant been a drug developed to treat a disease without the social baggage of obesity.  In this country, many people, including many health professionals, still consider obesity a condition caused solely by sloth and overindulgence that should be treated by willpower alone.  But I don’t think so.  Rather, I believe that this division and its advisory committee members are simply willing to be more paternalistic than in recent history, in light of missteps made by prescribers, pharmacists, sponsors and others that have led to a series of recent, high-profile drug recalls and new prescribing warnings for marketed drugs.  In essence they’re saying:  “The American public is either unable or unwilling to responsibly manage its use of new drugs.  Until that situation changes, we must intervene by taking away the tough decisions…by limiting the choices that are available to it.”

I haven’t seen FDA place a lot of the blame on itself for the inability or unwillingness of prescribers and users of new drugs to manage their use responsibly, but most senior folks at FDA must know that through their inability to demand creation and dissemination of effective prescribing information, and through their inability to demand creation of and to enforce effective new-drug risk management plans, and through their unwillingness to require the types of large trials needed to define risk before marketing that they shoulder part, perhaps most, of the blame.  None of this has to do with the needed resources for postmarketing surveillance being requested in the new PDUFA negotiations; I’m only speaking of what needs to happen before a drug reaches the market.  For that, FDA has said it has the resources it needs to do the job well.  So, why hasn’t it?

Sponsors, for their part, also shoulder significant blame for the rise of paternalism at FDA.  Perhaps they just missed seeing the writing that has been on FDA’s wall for the past few years, but I think they saw it quite clearly and chose in some cases to ignore it, hoping it would go away, and in others to address it weakly, with a wink to their investors.  Now they must overcome this guilty-until-proven-innocent mentality to market new drugs for chronic disease states, not to mention the formulary and insurance coverage hurdles awaiting new drugs once approved for marketing.  The regulatory landscape won’t change for quite a while.  If anything, we haven’t seen the top of this wave.

Sponsors, if you hadn’t heeded prior warnings, you’ve now been clearly warned:  It’s up to you and you alone to demonstrate that your new chronic-use drugs work as intended, using highly relevant and generally accepted measures of effectiveness; that they can be used safely, where safely means that the risk of use is very clearly outweighed by the benefit, even if that means you must find ways to limit risk by restricting the user base and creating redundant methods of safeguarding their proper use.  New chronic-use drugs must also be priced so as to be fiscally viable to large purchasers, and their value must not be a matter of faith but rather of overwhelming evidence.  Which other companies are in late stages of development with CB1 antagonists?  That would be Pfizer, Merck, and Solvay/BMS.  We’ll see soon if they have been paying attention.

Sanofi-Aventis made many of the right moves with rimonabant.  They misstepped by not gathering safety information in a way that could be used to create a sound risk-management strategy. There is time for them to make amends, in part, by gathering more evidence of both efficacy and safety.  Should S-A be able to generate evidence from the CRESCENDO outcomes study already in progress that very obese users (i.e. BMI >40) experience a meaningful reduction in their risk of cardiovascular events while using rimonabant, they will go a long way towards tipping the scales in favor of benefits.  In the meantime, S-A must collect more evidence from Europe that shows the drug can be used safely in the “real world”.  This, combined with some way of predicting suicidality through the use of simple baseline measures, other than or in addition to weight and history of depression, should suffice to create a viable risk management strategy.  Whether S-A would want to launch in the U.S. with a restrictive prescribing strategy and its attendant black-box warnings is another question. 

As you might have read in papers like the International Herald Tribune, Exubera, the first marketed inhaled protein therapeutic–recombinant human insulin–isn’t exactly flying off pharmacy shelves. 

Now that sales are flagging, analysts of all stripes are writing that they long ago knew this was going to happen.  This Monday-morning quarterbacking is known in behavioral circles as hindsight bias, or, as I like to call it, bullshit.  What nearly every stock analyst and pundit wrote while Exubera was in Phase 3 testing was that Exubera was going to be a big success.  Nearly every clinical opinion leader I heard from (there were many) said the same thing. 

What happened once Exubera was launched, its lackluster sales, isn’t nearly as interesting as how Pfizer and industry observers reacted to what was happening in the two years prior to launch.  Within industry, there are lessons here for market research, scientific communications, investor relations and executive management of all functions.   Chief among these lessons is this:  If you care about predicting customer responses to new products, and I’m sure you do, you cannot rely on inferences drawn from surveys, interviews, direct observations, product analogs, etc from a single point in time.  You must follow trends in attitudes and actions as information becomes available, and you must do so continuously during development and after launch.  A corollary is that you must carefully plan how you seek and disclose attitude-influencing information. 

Which leads me to my second lesson:  Keeping emerging product information secretive can be important for competitive reasons, but secrets are antithetical to the tasks of predicting market responses to products and to setting financial market expectations; a balance is therefore required.  With a product like Exubera, there were opportunities for disclosing to interested parties, namely key opinion leaders, selected patients, and investors, certain competitive information earlier and more fully than was done.  Here, I’m thinking specifically about the emerging product profile (i.e. safety, tolerability, efficacy, convenience, and usage) and Pfizer’s data and arguments supporting the cost-effectiveness of the product, which I’ve reviewed here previously.  If you’re a Pfizer employee, it’s now time to review whether your internal and external information disclosures were sufficient and timely to meet your market research, commercialization preparation, and financial market goals.  Be brutally honest with yourselves and take corrective policy action if deficiencies are found.  You may need different policies for different classes of products (e.g. a pioneering product vs. a fast follower).

For observers, and investors particularly, my suggested take away is this:  Recognize in yourself your reluctance to change your opinion as evidence mounts to the contrary; the strength of your unwillingness to change your mind is proportional to the strength of your held opinion.  This is hardly a novel insight, as published evidence supporting it dates to at least the 1950’s.  In behavioral circles, it’s known as attitude strength, or as I like to call it, stubborness.  You might not be able to overcome your stubborness, but you might be more willing to hedge your bets if you can admit to it.