The Best Bad Deal of All Time? Pfizer/King and the Business Dynamics of Abuse Deterrence

Pfizer’s acquisition of King gave the company a chance to play in a new kind of market: abuse deterrent opioid pain relief formulations. Two years later, Pfizer is moving in the other direction, even as FDA’s actions validate the notion that there is an opportunity—and that the scale of Big Pharma can play a critical role.

May 10 was a bad day for small companies trying to make it big with abuse deterrent opioid formulations.

The day began with Pain Therapeutics Inc. and Durect Corp. releasing copies of a letter they had received from their Big Pharma partner, Pfizer Inc., offering a very ominous sounding update on the status of the long-acting oxycodone formulation Remoxy, which was originally developed by Pain Therapeutics using Durect’s formulation technology.

It ended with Endo Pharmaceuticals Inc. receiving bad news from FDA: the agency rejected Endo’s citizen petition seeking to remove two generic versions of its Opana ER brand on the grounds that they do not match the abuse deterrent features of the new formulation Endo launched last year. (Also see "FDA Crushes Endo’s Hopes, Allows Original Opana Generics To Remain On Market" - Pink Sheet, 10 May, 2013.)

For investors in all three companies, May 10 was a very bad day. In percentage terms, Pain Therapeutics suffered the most, losing almost exactly 50% of its value in a single day of trading. For Durect, the hit was also large: the stock lost a third of its value—and has since drifted below the $1 per share threshold (a potential trigger for delisting from the NASDAQ).

Endo, by far the biggest of the three, suffered the least in percentage terms, losing 5% as the news broke on May 10, and another 4% on May 13 as investors digested the impact of FDA’s decision. By market cap, however, the news was costliest for Endo: the 5% drop represents more than $200 million in valuation for the company in a single day. (See Exhibit.)

Not everyone was disappointed in the news. For Impax Laboratories Inc., the rejection of Endo’s petition means that its generic version of Opana ER can stay on the market (and the company’s shares moved up 8% as a result).

The big winner, though, is Purdue Pharma LP. On April 16, the company succeeded where Endo failed—getting an explicit abuse deterrence claim added to its label, with potential generics to the old formulation blocked as a result. The news from May 10 means that Purdue’s OxyContin will remain the only long-acting opioid with an explicit abuse deterrent formulation claim – and that a potential new competitor from Pfizer will be delayed further if not discontinued altogether. (Investors, of course, can’t share in that joy since Purdue is a privately held company.)

Exhibit 1

Dashed Hopes For Abuse Deterrence

Source: The RPM Report

But what about Pfizer?

For investors, the company’s latest Remoxy setback was greeted as a non-event, with Pfizer shares edging up slightly on May 10.

In fact, the company’s reconsideration of investing more behind the project looked prescient after FDA’s rejection of Endo’s petition: clearly, there is a high bar for getting a marketable claim of abuse deterrence, above and beyond what Pfizer would need to spend just to get to market. So maybe Pfizer’s smartest choice is indeed to stop throwing good money after bad, and just drop Remoxy.

But don’t let the short term trading reaction fool you. Pfizer’s update on Remoxy is only the latest sign that the biggest name in Big Pharma is struggling to capitalize on a long-term opportunity in abuse deterrence.

When 2011 began, Pfizer looked well-positioned to take on Purdue for dominance in the opioid pain relief class, thanks to the $3.3 billion acquisition of King Pharmaceuticals at the end of 2010. That isn’t a huge acquisition by Pfizer standards, but it carried a potentially huge upside if the King portfolio of abuse deterrent technologies paid off.

Opioid pain relievers, after all, are the largest therapeutic class in the US, by dispensed prescription volume. Sales volume isn’t as high, thanks to high rates of generic use.

However, OxyContin is a blockbuster by anybody’s standards, and the potential for abuse deterrent formulations to extend the lifespan of those brands—perhaps in perpetuity—is a very attractive commercial proposition. And, in its decisions on OxyContin and Opana, FDA made clear both that extended product life-cycles can be achieved, but that it won’t be easy. In other words, FDA is setting a high bar, one that is likely to take a company with the scale and resources of a Big Pharma to clear.

So the strategic case for a company like Pfizer to invest in abuse deterrence is stronger than ever. But Pfizer is moving in the other direction, thanks to setbacks with the products and product candidates it thought it had.

Two years after the King acquisition, it looks like Pfizer bought the right strategy—but the wrong products.

Going Backwards in Pain Category

To be fair, Pfizer never described the King acquisition as a bet on the abuse-deterrent model.

In explaining the acquisition to investors, the company noted that it brought with it three distinct assets: an animal health product line to fit with one of Pfizer’s side businesses; the Meridian auto-injector business that produces EpiPen for sale by Mylan NV; and a portfolio of pain products to enhance Pfizer’s position with Celebrex. And, Pfizer stressed at the time, the deal would be immediately accretive to shareholders. (Also see "King of Pain: Pfizer's Move into Abuse Resistant Opioids and the Virtues of Massive Scale" - Pink Sheet, 1 Nov, 2010.)

Still, of the three distinct business lines, the pain portfolio stands out as a distinct failure.

The animal health business line is a success story – at least on Wall Street. After consolidating King, Pfizer’s next move was spinning out the beefed up animal health line, a move applauded by investors on its own terms and as a potential first step towards a possible even-more-dramatic breakup of the company.

The EpiPen product line has also performed well; Pfizer reported first quarter sales of $72 million, up almost 25%.

As for the King pharmaceutical product line—well, suffice it to say that EpiPen is the only former King brand that qualifies for a sales breakout in Pfizer’s quarterly report.

The off-patent Skelaxin line exceeded $200 million in revenue in 2012, and the muscle relaxant brand continues to face just one generic competitor. However, sales worldwide dipped below $50 million in the first quarter of 2013, and so it is no longer reported separately by Pfizer. However, it is still probably the biggest legacy King brand in Pfizer’s portfolio.

That isn’t what anyone predicted at the time the deal closed.

In buying King, Pfizer got the rights to Remoxy, an already marketed abuse deterrent formulation Embeda, and a unique product line based on “aversion technology” licensed from Acura Pharmaceuticals Inc. It also brought the Flector patch, a transdermal formulation of the generic analgesic diclofenac.

Two years later, Pfizer still has the Flector patch (which has never achieved the $50 million quarterly sales threshold).

As for King’s portfolio of abuse deterrent products, the only brand that Pfizer currently markets is, essentially, a fallback brand: an immediate release oxycodone formulation called Oxecta, which uses a crush resistant technology developed by Acura. It was developed based on feedback from FDA after the “aversion” formulation—which includes niacin to cause an unpleasant flushing reaction to deter abuse—was rejected by the agency. (Also see "A Painful Decision for King and Acura" - Pink Sheet, 1 Jul, 2010.)

The decision to pursue a single ingredient formulation was clearly expedient to get to market. But the Acura strategy depended on the idea that the niacin co-formulation could be applied to many different ingredients and build out a complete portfolio of abuse deterrent products. King claimed to maintain confidence in that approach despite the initial rejection. Pfizer, however, reached a different conclusion, returning all rights to Acura’s “aversion” pipeline in July 2012.

But at least Oxecta is on the market. King’s only marketed abuse deterrent therapy, Embeda, was recalled by Pfizer soon after the King deal closed, due to stability problems, and has not yet been reintroduced. During its first quarter earnings call, Pfizer said it hopes to submit a supplemental application this year to allow a re-launch in 2014.

Remoxy Reconsidered

And then there is Remoxy.

Pfizer may never have expressed blockbuster expectations for the brand, but Wall Street always held out hope. Tellingly, the first question to Pfizer on the conference call announcing the King deal was about how much due diligence the company did on the Remoxy formulation – underscoring investors’ sense that the product offered the biggest upside to the transaction.

At the time, the expectation was that Remoxy would be coming to market in the near future; the refiling of the application came in December 2010, shortly after the King acquisition was announced in October and just before it closed in January 2011. But instead of receiving a chance to go head-to-head against OxyContin, Pfizer received a “complete response” letter in June 2011.

At the time, Pfizer cited manufacturing issues as one contributor to the CR – an ominous enough issue, given the problems that had taken Embeda off the market.

But as recently as its first quarter 2013 conference call, Pfizer still talked about Remoxy as a relatively near-term opportunity. Pfizer Primary Care President John Young reported that Pfizer “had a productive meeting with the FDA in March” to discuss the application, and “we believe we have a path forward.”

That optimism was punctured when the company received the minutes of its FDA meeting, and prompted the letters to Durect and Pain Therapeutics on May 10.

The letter offered two items of bad news for Pfizer’s partners. First, after follow-up communication from FDA regarding the second “complete response” letter issued for Remoxy in June 2011, Pfizer now believes a new clinical trial will be necessary to support approval of the formulation, and hence the earliest resubmission date will be mid-2015.

Second—and even worse for the two companies—Pfizer made clear that it is by no means certain the company will bother to do the new trial.

“Given the years of delay, additional cost incurred to bring the program to this point and development work left in the program, there will be much to consider when we seek management endorsement” to proceed with the project, Pfizer VP-Global Commercial Development Alan Litwack wrote. “No final decision has been made at this time,” he said.

For glass-half-full types, at least Pfizer hasn’t given up yet. But in a best case scenario, Remoxy approval is now pushed back to 2016—fully nine years after the NDA for the product was originally filed in 2007.

FDA Setting High Bar For Claims

FDA’s May 10 letter to Endo will likely give Pfizer even more to think about.

In denying Endo’s request for an explicit abuse deterrence claim, FDA made very clear that it will not treat all enhanced formulations equally. While Purdue cleared the bar with OxyContin, Endo fell short – and, FDA hinted, may never be able to get a claim with the existing formulation.

The rejection suggests two key concerns:

• First, FDA sees the Opana ER formulation as technically inferior to the new OxyContin
• Second, Endo has not generated nearly as much post-marketing data as Purdue to justify a claim.

Both considerations have important implications in considering the potential for building a long term strategy in abuse deterrence.

Obviously, getting the technology right is going to be an important first step. And, based on the letter to Endo, it sounds like FDA does not think all crush-resistant formulations are created equal.

However, the more important message may be that FDA isn’t likely to make any decisions based on technical attributes of a formulation: good, high quality data—and extensive post-marketing experience—are going to be the key elements of successful applications to add abuse deterrent claims. And Endo’s failing may therefore have less to do with the attributes of the product, and more to do with the company’s less well developed data set.

The agency discussed the comparative attributes of the two formulations in its letter to Endo, taking as an excuse Endo’s assertion (via a supplement to its citizen’s petition) that FDA should treat Opana the same way it treated OxyContin, since the two formulations “have virtually identical abuse-deterrent properties.”

In hindsight, it was a mistake for Endo to press that argument, since it underscores one potential concern the agency will face under the emerging framework for abuse deterrent claims: FDA is crafting a policy that would allow several different “tiers” of claims for sponsors based on the properties of different formulations and the strength of evidence about a real world impact.

The policy will only work if the agency trusts that sponsors will educate prescribers appropriately about the distinctions being drawn by the various claims—and not use approval of any type of claim to support a general assertion of reduced abuse potential.

Endo’s claim also highlighted an important challenge FDA faces in the opioid space: defending its incremental decisions politically, given the intense interest in Rx abuse from several key members of Congress.

It wasn’t just Wall Street that interpreted the agency’s decision to block generic OxyContin as a precursor of similar action on Opana ER: During an appropriations hearing shortly after the OxyContin decision, House Appropriations Committee Chair Hal Rogers (R-Ky.) urged FDA Commissioner Hamburg to take similar action to remove Opana generics. Hamburg’s response was non-committal, with her promising to update Rogers as soon as a decision was made. (Also see "Commissioner Hamburg as Congressional Witness: Mixed Success With GOP Thus Far" - Pink Sheet, 3 May, 2013.)

She noted FDA’s strong interest in curbing abuse—but also the importance of ensuring that formulations actually work as claimed.

A Clear Line Between OxyContin and Opana ER

In its response to Endo, FDA notes Endo’s claim that Opana and OxyContin are “virtually identical” in their formulation technology, and responds bluntly: “We disagree.”

The agency stresses that both formulations “have been the subject of independent, extensive consideration by agency experts over the course of many months,” and that “our decisions take into account the totality of the evidence for the particular drug at issue, and must be made on a case-by-case basis.”

Accordingly, any attempt by Endo to draw parallels between the two “is misplaced,” the agency declared.

FDA, however, goes on to draw some comparisons—all of which are unfavorable to Endo.

“We note that there are differences in the products and the available data such that it is reasonable to draw different conclusions,” the agency says. FDA concluded that OxyContin has demonstrated a decreased potential for both intranasal abuse and abuse via injection in the new formulation, with carefully delineated discussions of the level of in vitro, clinical and post-market data to support each aspect of those conclusions. (Also see "OxyContin’s Abuse-Deterrent Claims Strong, But Stronger Claims Possible" - Pink Sheet, 22 Apr, 2013.)

For Opana, however, the agency notes that, while the new formulation does make crushing more difficult,  “data from in vitro and pharmacokinetic studies show that [the new formulation’s] extended-release features can be compromised, causing the product to ‘dose dump,’ when subjected to other forms of manipulation such as cutting, grinding, or chewing, followed by swallowing.”

The new formulation can also “be prepared for insufflation (snorting) using commonly available tools and methods.” Last but not least, it “can be readily prepared for injection,” FDA says.

In fact, a footnote at the end of FDA’s letter points out a worrisome trend toward more injection of Opana ER with the new formulation: FDA sees “the troubling possibility that the reformulation may be shifting a non-trivial amount of Opana ER abuse from snorting to even more dangerous abuse by intravenous or subcutaneous injection.”  Woodcock says that a “greater (and rising) percentage of Opana ER abusers are abusing Opana ER via injection since the replacement” of the original formulation with the new one.

The footnote raises the ominous possibility that the reformulation of Opana ER may end up being deemed worse than the original by FDA.

However, the broader message of the letter to Endo focuses on the quality of the company’s database to demonstrate real world reductions in abuse—not the quality of the formulation technology itself.

One factor beyond Endo’s control is time: FDA notes that Endo’s post-market data can only be considered “preliminary,” compared to the more “mature” data submitted by Purdue. The agency, however, also questions the lack of good clinical abuse liability studies submitted by Endo, and also the methodology of its post-market analyses.

Here, the contrast with OxyContin is telling. FDA notes that Purdue provide in vitro data, a clinical abuse potential study and post-marketing data to support a claim of reduced intranasal abuse potential. In contrast, “Endo’s petition and supplements do not describe a clinical abuse potential study to assess the ability to insufflate or clinical preferences.”

In addition, “the preliminary data from the Opana ER post-marketing investigations have significant limitations and are not as mature as the OxyContin post-marketing investigations, which themselves were only supportive, and not conclusive, of reduced intranasal abuse.”

Quantity and Quality of Data Counts

An optimistic reading of FDA’s response would suggest that Endo could seek reconsideration of the agency’s finding in 12-18 months – when there will be as much post-marketing experience with the new formulation of Opana ER as Purdue currently has with OxyContin.

The availability of generics using the old formulation could even accelerate that timeline: unlike OxyContin, there is a head-to-head comparison possible in actual real-world abuse rates between the new and old formulations of Opana. If abuse monitoring indicates that abusers “prefer” the generic, that would help Endo build its case.

However, there is a much less optimistic interpretation as well: the agency seemingly has significant reservations about Endo’s commitment to doing the work necessary to demonstrate abuse deterrence no matter how much time it has to build a case.

The agency’s denial letter includes a number of cases where FDA suggests Endo is cherry-picking from the limited data it has—and is failing to heed the agency’s requests to conduct more studies than it seems to be doing.

In discussing Endo’s supportive data for crush resistance, FDA notes Endo focuses on data showing that using a commercial pill crusher does not affect the bioavailability of the opioid ingredient, “other data…show that other forms of manipulation can compromise [the] extended release features, causing it to dose dump.”

The discussion of intranasal abuse data is more damning, with FDA citing January 2011 review documents that recommend a study of whether a pill grinder would produce a powder that can be snorted. “Endo’s petition and supplements do not discuss such a study and only refer to manipulation using simple tools such as a hammer, pill crusher and two spoons, which are not sufficient to fully assess” the risk of abuse “by experienced drug users.”

The agency thus dismisses Endo’s conclusions from an abuse liability study that focused on “tools and methods other than grinding.”

As for Endo’s claims of diminished risk of abuse by injection, FDA rejects those out of hand. The new formulation “can be readily prepared for injection, despite Endo’s claim,” the letter states. “In addition, certain data suggest that [the new formulation] can more easily be prepared for injection” than the old one. Once again, FDA cites 2011 review documents to support that conclusion.

The most damning comments, however, may be in the agency’s short review of Endo’s post-marketing data on real world abuse patterns.

FDA starts by noting that the data include only 2-3 quarters of experience. That can only be corrected by time.

However, the agency suggests its concerns go much farther than that.

The letter suggests frustration that Endo attempted to make any claims at all based on such limited data: “The data from the post-marketing investigations Endo relies on are inconclusive and, as the company repeatedly acknowledges, ‘preliminary.’”

FDA notes in particular that it has questions about Endo’s decision to set as a baseline the abuse rates for Opana in 2011. Abuse rates for Opana spiked in 2011, which Endo suggests may have been caused by the launch of the abuse deterrent OxyContin formulation. However, “it is too early to say whether the 2011 abuse rates represent the appropriate baseline risk or an aberration.”

In addition to the very short time period available for study, Endo’s analyses “suffer from significant additional deficiencies (including small sample sizes, likely misclassification of drug exposure, and possibly artificially elevated baseline abuse rates), such that it is not possible to draw meaningful conclusions based on them.”

In other words, FDA may not be waiting for more mature data—but a more mature sponsor.