Seeking Alpha
Long/short equity
Profile| Send Message| ()  

We have published a more detailed report on Avanir Pharmaceuticals on September 11, 2013. A PDF version of the full 47-page report can be found at gravityresearchgroup.com/reports.

Avanir Pharmaceuticals (AVNR) is a pharmaceutical company currently marketing one FDA-approved drug: Nuedexta. Nuedexta is a combination of two generic drugs (dextromethorphan and quinidine) and is approved to treat Pseudobulbar Affect (PBA), a condition that causes patients to experience bouts of uncontrollable laughing and/or crying. We believe that AVNR stock will collapse in the near future due to the perfect storm of overhanging potential negative events threatening its solvency, including a sharp drop in future revenue and fines resulting from a potential illegal off-label marketing investigation, steep declines in revenue and default on outstanding debt due to loss of this week's patent case, and other events that may adversely affect AVNR's cash flows in the near future.

1) We believe AVNR has engaged in illegal off-label marketing of Nuedexta and that off-label prescriptions have generated a significant portion of AVNR's Nuedexta revenue to date. We have collected strong evidence to support this conclusion, including numerous documented allegations of and allusions to off-label marketing from purported AVNR employees and recorded discussions we have held with former AVNR employees. Below are selected comments in reference to the off-label marketing at AVNR (see the full report for more):

  • "People are marketing off label for "Agitation" and "other behaviors." Many of us have seen it first hand with LNAM [long-term care neuroscience area manager] counterparts. It is a huge concern."
  • "…there is so much off label selling - it's dangerous. Reps will get caught...the company will get caught...good luck...this company is teetering on the edge with one foot in illegal waters and the other on a banana peel"
  • "[Y]ou will be encouraged to promote off label, you will be competing against reps who have zero issue selling off label…"
  • "…practically EVERYONE is selling OFF-LABEL!"
  • "…we have a HUGE culture problem in LTC [long-term care] selling off-lab[el]."

We believe AVNR could face over $100mm in fines and penalties and see its Nuedexta sales plummet following a potential government investigation. We reached this figure through research of settlements, which appear to be based on both the egregiousness of the off-label selling as well as the amount of revenue generated by the drug. For example, Elan's $203mm settlement represented approximately 93% of the total historical revenue it generated with the drug Zonegran. Cumulatively, AVNR has generated over $100mm in gross revenue from sales of Nuedexta.

According to one of our sources, at the time he left AVNR, the off-label marketing of Nuedexta at AVNR was not merely rogue sales representatives, but extended up the chain and was driven by multiple regional sales directors, and at least one national sales director. We believe that off-label marketing contributed to the mysteriously rapid departure of AVNR's former Chief Commercial Officer who arrived soon after Nuedexta was launched and who, as we confirmed with a source, became aware of the issue prior to leaving the company after less than eight months.

Former employees of AVNR have told us that the actual pool of PBA patients in the US is significantly lower than Management represents, making it difficult if not impossible to reach sales targets without engaging in off-label marketing (we believe the pool of marketable patients is at least 75% smaller than the 1.8mm figure quoted by Management).* This situation is eerily similar to what happened at InterMune, where AVNR's current CEO Keith Katkin was previously a Vice President of Marketing. InterMune was found guilty of engaging in off-label marketing while Katkin worked there, and former employees have testified in court that Katkin set sales goals for a drug so high that the only way to attain those sales goals was through off-label marketing.1 The fact that AVNR has experienced extensive turnover in salespeople and 50%+ of Nuedexta pills are being sold through long-term care facilities despite Management explicitly stating that such facilities only constitute 5.6% of total PBA patients also served to heighten our suspicions. There has been a recent rash of off-label marketing cases through long-term care facilities in particular among other pharmaceutical companies, and based on our research and discussions, we believe this is where the majority of off-label marketing of Nuedexta is taking place

2) We believe that AVNR has a very high probability of losing its current patent case, and as a result, will likely face an 80%+ decline in quarterly revenue by the end of FY 2014 and default on its debt due to a clause which lists "[a] commercial generic version of Nuedexta (for the treatment of pseudobulbar affect) becomes available" as an event of default.2

We believe the evidence presented in the court filings creates a clear-cut case in the defense's favor, and a reverse payment ("pay-to-delay") settlement with Par Pharmaceuticals (Par) is the only way AVNR could avoid a ruling invalidating its patents. However, Par has a strong incentive not to accept such a settlement due to 1) the fact that a favorable ruling in the case would lend itself to future cases for non-PBA indications using quinidine/dextromethorphan combinations and 2) anti-trust concerns, which are especially significant in light of the recent June 2013 Supreme Court ruling that reverse payment settlements are subject to anti-trust scrutiny. Having reviewed thousands of pages of court filings and academic/industry publications related to the case and patents in question (relevant parts of which are posted on our website under Appendix X to the report), we believe that AVNR would lose a ruling in the patent case. This would soon trigger default on its debt due to a clause in AVNR's credit agreement that declares the marketing of generic Nuedexta to be a default event.3 While creditors always have the option of allowing the company to continue with regularly-scheduled payments (typically at a much higher interest rate), we believe that would not be feasible in this case because we estimate that generic competition would wipe out AVNR's quarterly revenue from Nuedexta by at least 80% by the end of FY 2014, magnifying AVNR's already-negative operating cash flows.

The defense team for Par and its (former) co-defendants has produced firm evidence that the "composition of matter" concept of combining dextromethorphan and quinidine, and the "method of use" concept of using the resulting compound to treat PBA had both been patented with priority dates before the priority date of AVNR's earliest relevant patent (see US Patent Nos. 5,166,207 and 5,206,248),4 and the impact of low doses of quinidine on dextromethorphan plasma concentrations ("method of use") were documented in a 1999 abstract (the "Yakatan Abstract")5,6 that was published and made publicly available three years prior to AVNR filing for the relevant patents. We believe the data from the Yakatan Abstract in conjunction with the earlier patents makes all three of AVNR's patents invalid due to obviousness based on prior art and/or double-patenting.

Below is our summary of AVNR's patent case, culled from thousands of pages of available court filings from both the plaintiff and defense sides. We believe we present it in a fair and concise way that allows for readers to reach their own objective conclusions. It leaves out certain arguments from both sides that we consider to be frivolous. See Why We Believe the Nuedexta Patents Won't Hold Up in Court for a more detailed discussion of the patent case:

Plaintiff's Case

• AVNR have three patents covering its drug Nuedexta (which contains 20 mg of dextromethorphan and 10 mg of quinidine per pill): the '115 patent, the '242 patent, and the '484 patent. By filing ANDAs and paragraph IV certifications, the defendants are violating all three of them. The relevant claims from the '115 patent are "composition of matter" claims covering the dextromethorphan/quinidine combination, and the relevant claims from the '242 and '484 patents are "method of use" claims covering the use of specific ratios of dextromethorphan and quinidine ("1:0.5 or less" and "1:0.75 or less", respectively) and ranges of doses for dextromethorphan and quinidine ('242 patent: 20 mg/day to 80 mg/day of dextromethorphan and 10 to 30 mg/day of quinidine; '484 patent: 20 mg/day to 60 mg/day of dextromethorphan and 10 to 30 mg/day of quinidine)

Defense's Case

• The relevant claims from all three of the Plaintiff's patents are invalid and/or inapplicable in the case

• The '115 patent has a priority date of June 9, 1992. The asserted "composition of matter" claims of the '115 patent covering the combination of dextromethorphan with quinidine to augment the impact of dextromethorphan are invalid due to double-patenting-those same composition of matter claims were previously claimed through US Patent No. 5,166,207 (the '207 patent)7 which has a June 17, 1991 priority date, and US Patent No. 5,206,248 (the '248 patent), which has a March 27, 1992 priority date. Claims other than composition of matter are irrelevant to the case because the patent specifically addresses indications other than PBA (specifically "chronic or intractable pain, for treating tinnitus and for treating sexual dysfunction"), which is why it was granted by the patent office in the first place given the other two patents had already been filed and approved

• The '242 and '484 patents, which both carry a priority date of July 17, 2002 and cover the treatment of PBA with a combination of dextromethorphan and quinidine, are differentiated from the prior '248 patent (which also covers the treatment of PBA with a combination of dextromethorphan and quinidine) purely due to their dosing claims. The '248 patent does not specify dosing in its claims, and the '242 and '484 patents claim doses that are low relative to the doses used in cases cited by the '248 patent. However, a 1999 publication (the "Yakatan Abstract") 8,9 in conjunction with the '248 patent demonstrated clearly that doses of quinidine/dextromethorphan within the ranges claimed by the '242 and '484 would have been obvious for the purposes of treating PBA.

It is true that the Yakatan Abstract was published by AVNR employees, however, the waiver that prevents publications by inventors from being considered prior art only applies if the publication was made within a year of the priority date of the patent-in this case, the Yakatan Abstract was published three years before the priority dates of the relevant patents. The Yakatan Abstract showed that 2.5 mg of quinidine combined with 30 mg of dextromethorphan would produce peak blood concentration of 42.75 ng/mL, while 25 mg of quinidine produced a peak blood concentration of 96.9 ng/mL (both calculated solely based off of figures given in the Yakatan Abstract). It also stated that the same level of blood concentration was observed with both 50 mg of quinidine and 75 mg of quinidine, and 25 mg of quinidine yielded 85% of the plasma concentration levels of 50 mg, indicating that 25 mg of quinidine yielded close to the maximum effect in terms of dextromethorphan inhibition, making it obvious that doses higher than 25 mg would not be significantly beneficial to patients. Meanwhile, the '248 patent includes three examples of patients for whom a combination of DM and Q treated emotional lability. Example 2 describes a patient successfully treated for PBA, whose DM blood plasma concentrations "usually averaged between 43 and 55 nanograms per milliliter." 42.75 ng/mL, the peak blood concentration generated by 2.5 mg of quinidine per the Yakatan Abstract, is nearly within the range shown to be effective in treating patients in the '248 patent. Therefore, the combination of these pieces of prior art demonstrate that very low doses of quinidine/dextromethorphan well within the ranges claimed by the '242 and '484 patents would have been sufficient to treat PBA three years in advance of the priority dates of the '242 and '484 patents.

Plaintiff's Rebuttal

• The '207 and '248 patents do not make the composition of matter element of the '115 patent obvious because they do not list specific doses of dextromethorphan and quinidine under their claims, while the '115 specifically claims doses of quinidine of "300 milligrams/day or less"

• The '242 and '484 patents are valid because the Yakatan Abstract and '248 patent are not sufficient to claim obviousness. Focusing on Example 2 of the '248 patent amounts to cherry-picking, and it is unreasonable to think that someone skilled in the art would have motivation to reach a similar conclusion based on the prior art presented

Defense's Rebuttal

• While 300 milligrams or less of quinidine is not explicitly listed under the claims of the '207 and '248 patents, it is obvious based on the descriptions of the '248 and '207 patents. They specifically included the results of trials done using 150 mg and 300 mg doses of quinidine

• Focusing on Example 2 of the '248 patent is not cherry-picking-only two of the three available examples provided blood plasma concentrations, and based on severe negative side effects of overdosing, there was clear incentive to find the lowest effective dose. Example 1 of the '248 patent does not provide the patient's blood plasma concentrations, while Example 3 described a patient successfully treated for PBA with DM blood concentrations that "usually averaged between 130 and 180 ng/ml". There is clear motivation to find the lowest effective dose that could be used to treat patients because the reaction can be severe in the event of an overdose, as was noted in the description of the '248 patent:

"Dextromethorphan is believed to act at the phencyclidine (PCP) binding site, which is part of the NMDA receptor complex…One patient suffered a severe adverse reaction at the 30 mg/day dosage of DM, when quinidine was co-administered. He had not previously suffered any adverse reaction when DM alone was administered. His reaction, which lasted several days, was similar to the hallucinatory and psychotic reactions that are often observed in people who illegally abuse the drug phencyclidine (also known as PCP or "angel dust")."

Combining the learnings of the '248 patent with the Yakatan Abstract, it is obvious that doses well within the ranges claimed by AVNR in the '242 and '484 patents would have been sufficient to treat PBA. Therefore, the patents are invalid because they are obvious based on prior art.

We believe that the recent settlements are irrelevant, and the case that matters is the one between AVNR and Par Pharmaceuticals, which was first to file its ANDA. Generic companies that are first to file generally foot the vast majority of the legal bills in exchange for a period of 180 day exclusivity, after which other generic competitors are permitted to enter. We believe that the aftermath of AVNR losing the patent case spells certain bankruptcy or dilution to pennies per share for investors in the near future

3) Other imminent events will negatively impact AVNR's cash flows and further endanger AVNR's vulnerable cash position. After including the impact of the agreement with OptiNose, AVNR's cash balance as of the end of the last quarter stood at $35mm, with $30mm in debt. With consistent negative operating cash flows at $10mm+, debt beginning to mature in 2014 at the rate of $3mm/quarter, approximately $1mm/quarter in revenue from its docosanol ("Abreva") license disappearing next year due to the terms of the license agreement, and increased costs of $2-3mm/quarter over the coming quarters due to the OptiNose deal, we do not see how AVNR can survive through the next 12 months without significant equity offerings to shore up its balance sheet. Based on the current price, even ignoring the entrance of generic competition and the off-label marketing allegations, we believe these events have not been properly accounted for by investors.

________________________________________________________________

AVNR also faces two other major issues in selling Nuedexta that we believe have largely been overlooked by analysts and investors: lack of marketability for drugs to treat PBA due to the nature of PBA as a relatively non-threatening symptom rather than a condition in and of itself (in fact, because PBA is generally not even considered to be a disease, the phrase "pseudobulbar affect" is not even consistently used to describe the condition-it is also called pathological laughing and crying, emotional lability, emotional incontinence, involuntary emotional expression disorder, emotionalism, and numerous other names), and competition from antidepressants (which have been clinically shown to treat PBA, and we believe, based on published data, are superior to Nuedexta in terms of efficacy, cost, efficiency, and safety). At the same time, we believe that AVNR's pipeline will provide no relief for investors-instead, it will likely accelerate AVNR's financial distress by increasing R&D costs as well as requiring milestone payments to OptiNose and Concert Pharmaceuticals:

I. AVNR has admitted that it does not have sufficient capital to independently fund Phase 3 trials for the DPN pain indication within its dextromethorphan/quinidine pipeline, and it should be obvious to investors based on AVNR's financial profile. Given that DPN pain is the most advanced indication in its pipeline, it is unreasonable to think that AVNR has sufficient capital to fully fund any other Nuedexta-related indications

II. AVP-825 (sumatriptan delivered via a dry powder nasal spray to treat migraines) is the only drug in AVNR's pipeline that is remotely close to FDA approval, and we believe it will be virtually unmarketable-in terms of efficacy, it is dominated by competitors (among the five triptan competitors we compared it with, AVP-825 showed the lowest therapeutic benefit for patients in its clinical results!10,11) and on top of that, it requires patients to blow air through their mouth to push dry powder through one nostril and out the other. Further, the triptan market is set up in a way that gives an enormous advantage to the cheapest options. While insurance companies will technically cover all available triptan deliveries, we found that they frequently restrict the number of doses patients are allowed to purchase for more expensive delivery systems (such as nasal and subcutaneous delivery). As a result, tablets command over 95% of the triptan market, despite being the slowest-acting and requiring the highest drug exposures! We believe that this, combined with the uncomfortable administration method, will severely restrict the market share for AVP-825

III. While AVP-786 (deuterated dextromethorphan combined with quinidine to produce similar efficacy to Nuedexta with lower levels of quinidine) would have a lower dose of quinidine than Nuedexta, we believe it will present no significant advantage over Nuedexta/generic Nuedexta. Doctors would prescribe antidepressants to PBA patients for whom quinidine presents a risk. The primary risk for prescribing Nuedexta is that quinidine inhibits metabolism of many drugs, including dextromethorphan, antipsychotics, antidepressants, and other drugs that are commonly taken by the type of patients most prone to PBA. It has been shown that even at 2.5 mg, quinidine can produce a 15x increase in plasma concentrations of dextromethorphan compared to patients taking the same dose of dextromethorphan by itself, with the potential for similar effects when combining quinidine with other drugs. Most drug-drug combinations with quinidine other than dextromethorphan have not been tested, leaving a huge question regarding the safety for patients taking multiple drugs that would be impacted by the quinidine inhibition. We believe that, AVP-786 would only be able to reduce the quinidine dose to 5 mg, at best. As a result, we believe AVP-786 will have no advantage in selling over generic Nuedexta and represents a gimmick rather than a true advance in therapy


[*] Based on our research, discussions with medical professionals, and discussions with ex-employees, we believe the total pool of marketable PBA patients is at least 75% smaller than Management currently claims. Numerous articles have been published citing a "CNS Scale" in estimating seven figure PBA patient populations, but unbeknownst to many investors, CNS stands for Center for Neurologic Study and is an organization that stands to benefit financially from any success AVNR has in selling Nuedexta due to an agreement granting CNS royalties on Nuedexta revenue. We believe this has led to perverted "estimates". In December 2008, Management estimated the US population at roughly half of what Management is projecting to investors today ("likely over one million"). The next year, Management increased its estimate to "an estimated two million patients in the U.S."

___________________________________________________________

[1] See Appendix A - InterMune Securities Litigation

[2] Exhibit 10.1, 10-Q filed on August 9, 2012 for Q3 2012

[3] Id.

[4] US Patent No. 5,166,207 (the '207 patent) which has a June 17, 1991 priority date (www.google.com/patents/US5166207), and US Patent No. 5,206,248 (the '248 patent), which has a March 27, 1992 priority date (www.google.com/patents/US5206248)

[5] Yakatan et al., Low dose quinidine inhibition of dextromethorphan metabolism by CYP2D6, Abstracts of the Twenty-Eighth Annual Meeting, American College of Clinical Pharmacology, September 16-18, 1999, J. Clin. Pharmacol. 39, 984 (1999)

[6] See Appendix B - Yakatan Abstract

[7] google.com/patents/US5166207

[8] Yakatan et al., Low dose quinidine inhibition of dextromethorphan metabolism by CYP2D6, Abstracts of the Twenty-Eighth Annual Meeting, American College of Clinical Pharmacology, September 16-18, 1999, J. Clin. Pharmacol. 39, 984 (1999)

[9] See Appendix B - Yakatan Abstract

[10] "Therapeutic benefit" calculated as the difference between primary endpoint results for placebo and active drug groups

[11] See full report for more details and a direct comparison with existing treatments

Source: Alleged Off-Label Marketing, Patent Woes And Other Issues May Spell Disaster For Avanir