Expert Financial Analysis and Reporting

Key Investment Conclusions

In this report, I have made projections for sales, earnings and cash flow for A.P. Pharma (APPA.OB) through 2016. The underlying assumptions and calculations are explained within this report. However, let’s start with the conclusions. My base investment case is that A.P. Pharma will gain approval for its first commercial product APF530 in early 2013, will launch the product in the US with its own sales force in mid-2013 and will partner the product in the rest of the world. Here are my key projections.

If these projections are realized, I think that in 2015 the company could be valued at about 20 plus times estimated 2016 EPS; this results in a 2015 price target of $5.40+. I recommend this stock for sophisticated biotechnology investors who understand the risks of investing in emerging biotechnology companies. There is execution risk for this small company and if my assumptions on approvability of APF530 and the potential for APPA to successfully commercialize it are wrong, the investor could lose most or all of their investment. This risk is counterbalanced by the significant upside opportunity that I have just outlined. Please carefully read the investment risk section that follows. As with other pre-commercial companies in this sector, if you do not understand and are unwilling to accept these risks, this stock is not a suitable investment for you.

Investment Perspective

A.P. Pharma’s lead product, APF30, is being developed for the prevention of chemotherapy induced nausea and vomiting or CINV. It is an injectable drug that uses the company’s Biochronomer™ polymer-based, sustained-release delivery technology to deliver the widely used generic drug granisetron. It increases the time of effectiveness for granisetron from one to five days, which provides important therapeutic advantages.  

The company is in the process of responding to a Complete Response Letter or CRL on APF530 which it received in March, 2010. In that CRL, the FDA expressed no concerns on clinical efficacy and safety of APF530 in its pivotal trial or the statistical analysis used. The FDA asked the company to address a number of CMC issues that were fairly straightforward. Of more concern was a request to conduct a study to rule out the possibility that APF 530 might cause QTc prolongation, a condition which was recently detected with Anzemet, a drug of the same chemical class. On March 26, 2012, APPA announced that it had successfully completed that study which showed no QTc prolongation; this appears to take away the major uncertainty in regard to approval. The other issues in the CRL were less serious and the company believes that it has actions that should successfully address these issues.

The company is on track to refile the NDA in mid-year and assuming a Class II response the PDUFA date would be in early 2013. It has been a routine practice for the FDA to issue CRLs for new drugs at their first PDUFA date with CMC issues being a frequent reason. The second PDUFA date has become the “new” PDUFA date.  It now seems prudent for investors to assume that almost every drug will receive a CRL on its first PDUFA date. A.P. Pharma has already been through this vetting process and appears to have successfully addressed the issues raised by the FDA. This significantly increases the potential for approval in early 2013.

After approval, biotechnology investors are not home free. One of the most troubling situations that has developed over the last few years is the slow launch phenomenon; this has occurred with Avanir’s (AVNR) Nudexta, Human Genome Sciences (HGSI) Benlysta, Auxilium’s (AUXL) Xiaflex and Cadence’s (CADX) Ofirmev. These slow takeoffs reflect physicians’ increased level of concern with safety of new drugs and the slow process involved in getting drugs accepted on formulary. Investors now routinely expect slow launches and hedge funds often short stocks during the launch period. However, I think that APF530 could be the exception to this trend and surprise investors with how quickly it takes off.

The active ingredient of APF530 is granisetron, a product with a long history of safe usage in CINV that is trusted by physicians so that there should be minimal concern about safety and efficacy. There is also a compelling clinical reason to use this drug when the market leading product Aloxi fails to completely halt emesis (vomiting), which is about 30% to 60% of the time. A 1,341 patient Phase III trial showed that APF530 is as effective as Aloxi and offers the physician an alternative when Aloxi fails to establish complete control.  Thirdly, there is a compelling economic reason to use APF530. Both Aloxi and APF530 are reimbursed as Medicare part B drugs. The physician currently receives about $31 per Aloxi injection. However, the peculiarities of reimbursement are such that if APF530 is priced at the same wholesale price as Aloxi, for a period of around 6 to 9 months following introduction, the physician could make $78 for an APF530 injection. This is a significant economic inducement which I will explain in more detail later.

There are difficult challenges. The rapid takeoff I am projecting stems from a pricing aberration under Medicare part B that creates a great economic incentive for physicians to use a new drug.  It is imperative that APPA stages an aggressive launch for APF530 to maximize its market position during this economic window of opportunity. However, A.P. Pharma has no commercial experience, limited infrastructure and a weak balance sheet with only $18 million of cash at yearend 2011. Usually this would rule out a small company like APPA even thinking about marketing a new drug on its own. However, the CINV market is relatively small and can be covered with about 40 reps. I estimate that putting a sales force in place and then launching APF530 would cost about $15 to $20 million and this is within APPA’s reach, even with its limited resources.  

Investors must also be aware that there are a very large number of shares outstanding with this company. Because of the CRL, the commercial timing was delayed and the company was forced into some highly dilutive financings. The company reported 200.0 million shares outstanding in its 2011 10-K. However, the highly probable conversion of “in the money” convertible notes, warrants and stock options could increase the share count to 446.7 million by the end of 1Q, 2014.

There are two wildcards that I think are possible for A.P. Pharma that could create significant value for shareholders. The company might be acquired outright. I am not sure, however, how much of a premium it could command over its current price if this were to happen before approval. There is also the possibility that an ongoing study will show that Zofran, a long established market leading drug in CINV, is linked to QTc interval prolongation. It accounts for 29% of the unit market and a warning about this risk could lead to a sharp drop in its usage and significantly enlarge the market potential for APF530. I would note that each 10% of current ondansetron usage that is shifted toward APF530 and Aloxi would increase use and sales of APF530 by 5% according to my assumptions.

Investment Risks

The numbers that I present in this report give the appearance of greater precision than is actually the case. They should be looked at as indicators for trend and magnitude of corporate results, not numbers that are etched in stone. Eventual results could vary significantly from my projections.

FDA regulatory action can be unpredictable and the greatest risk is that APF530 receives another CRL and is delayed. This could delay partnering discussions and create a cash squeeze. The company in this event might have to finance at depressed valuations to keep operations ongoing. This could result in severe dilution to shareholders.

I have assumed partnering deals for APF530 in 4Q, 2012 for parts of Asia and Europe produce milestone payments of $15 million for each market. My thinking is influenced by the $26 million partnering deal that Prostrakan received with its transdermal granisetron patch, Sancuso, for rights to Southeast Asia. My judgment is that APF530 is a superior product. These potential partnering deals would obviate the need for the company to raise capital from the capital markets to launch APF530 in the US. However, the projection of a fourth quarter partnering deal would be prior to the projected approval date for APF530 and potential partners might elect to wait and make sure the product is approved. My projection for cash at the end of 2012 of $36.8 million is dependent on consummating these partnerships. If they are delayed my year end cash balance projection would drop to $6.8 million. This might force the company to raise perhaps $10 million through a small equity offering or bridge loan. 

I am assuming that A.P. Pharma launches APF530 on its own in the US. This is a formidable undertaking for a company that has little commercial infrastructure and no prior experience in selling a drug. There is substantial risk that they may encounter execution problems during the roll out.

APF530 uses APPA’s Biochronomer drug delivery technology to deliver the generic drug granisetron. The intellectual property position is based on patents related to Biochronomer that protect this technology through 2020. I assume that there will be no generic competition until these patents expires. I acknowledge that competitive long acting granisetron products could come to the market though the 505 (b) 2 approval pathways, but they would not be substitutable for APF530 under current FDA practices.

Aloxi’s key composition of matter patent expires in 2015. Recently additional patents covering stability and shelf life were issued which I believe extends the exclusivity of Aloxi through 2024. Generic companies will certainly challenge these new patents and if they prevail, there is a possibility of generic competition to Aloxi in 2015 or years thereafter which might negatively impact the pricing of APF530.

CINV Market Overview

APF530 is targeted at a very promising niche in the market for drugs that treat chemotherapy induced nausea and vomiting (CINV). The efficacy of chemotherapy drugs is based on their ability to target and destroy rapidly dividing cancer cells. However, normal cells that divide rapidly such as those lining the stomach and intestinal tract suffer collateral damage in the process. The induced stress causes the release of the chemical messenger 5-HT3 serotonin in the brain inducing nausea and a signal to vomit (emesis). The understanding of this mechanism of action led to the development of a class of drugs that block the receptor for 5-HT3 serotonin (5-HT3 serotonin antagonists); they have become the mainstay of therapy to prevent CINV.

There are currently four 5-HT3 drugs on the market: Zofran (ondansetron), Kytril (granisetron), Anzemet (dolasetron) and Aloxi (palonosetron). Zofran and Kytril were the pioneers in the market and dominated it for many years. Zofran became generic in 2006 and Kytril in 2008. It is usually the case that when the leaders in a drug category go generic the market shifts to generic competitors, leading to declines in usage and pressure on pricing of the remaining branded competitors in the category. This has not been as true for the 5-HT3 category. To understand why and to understand the potential of APF530, one has to understand some things about market dynamics.

CINV occurs in about 75% of patients given chemotherapy with different regimens causing more or less nausea and vomiting. It usually appears on the first day a few hours after the chemotherapy is administered and can last for five days. There are two phases of CINV, the acute onset during the first 24 hours and delayed onset that occurs over the next four days. Zofran and Kytril are effective in the acute phase, but much less so in the delayed phase because of their short half-lives.

On the first day, CINV drugs are almost always given intravenously in order to achieve effective blood levels at the time the chemo is administered. On subsequent days, it is usually not practical to administer the drugs intravenously as the patient most likely has gone home. This means that for short acting drugs like Zofran and Kytril, an oral dosage form is required for days two through five. For patients who have some nausea and vomiting, they may be unable to swallow the pills and compliance may be a problem. Well controlled studies indicate that Zofran and Kytril are effective in about 42% of patients receiving moderately emetogenic chemotherapy and in 39% of patients on highly emetogenic chemotherapy therapy. Effectiveness or complete response is defined as a patient having no emesis and not requiring rescue medication. A CINV drug can offer some relief without meeting these criteria.

When only Zofran and Kytril were available there was often difficulty in controlling nausea and vomiting in the delayed phase. In 2004, MGI Pharma introduced Aloxi into the market. The advantage of Aloxi is that it has a very long half-life so that with one IV course of therapy, it retains its efficacy through the delayed phase. After Zofran and then Kytril went generic, they were available for about $5 to $10 per injection versus $175 for Aloxi. Nevertheless, the long acting advantage of Aloxi has allowed it to become the market leader with 49% of the unit market.

While Aloxi is the leading drug in the CINV market, it has shortcomings that APF530 can capitalize on. Clinical studies have suggested that Aloxi has a complete response of 46% to 69% in moderately emetogenic chemotherapy and 41% to 62% in highly emetogenic chemotherapy. Hence, there are large numbers of patients who don’t obtain full relief with Aloxi, somewhere in the range of 30% to 60%. When Aloxi doesn’t work, doctors may add corticosteroids or Merck’s Emend to the regimen on the next cycle. Emend works through a different mechanism, blockage of substance P, to control CINV.

There is now an unmet medical need for a long acting drug that has the potential to work when Aloxi falls short. This is one of the market segments that APF530 is targeting. It is long acting like Aloxi in that it can be injected on the first day and lasts five days. Head to head studies in a large phase III involving 1,341 patients showed that it is statistically non-inferior to Aloxi which indicates equivalency. The hope is that many patients who are not treated effectively with Aloxi may be switched to APF530 and controlled in subsequent chemotherapy cycles. It is often the case that drugs of the same chemical class will produce different effects on the same patient. Even though, the large phase III trial comparing Aloxi and APF530 showed equivalent efficacy, physicians could hypothesize that they might have different effects on different patients.

There is another potential advantage for APF530 that was hinted at in the clinical trials. Most cancer patients go through 5 to 15 cycles of treatment during the course of the disease. Aloxi seems to lose some effectiveness as the number of cycles increase, while complete response appears to increase with APF530. More studies are needed to confirm this observation, but it is worth noting and watching. If proven in clinical studies, this would be a huge advantage versus Aloxi.

Medicare Part B Reimbursement is Very Favorable to APF530

Clinical benefits alone could allow APF530 to gain a meaningful share of the CINV market. However, there are also strong economic incentives for doctors to prescribe APF530. The reimbursement of injectable 5-HT3 drugs is covered under Medicare part B regulations which set the pricing for Medicare and is also closely followed by private insurers. Aspects of the reimbursement regulations create economic incentives which favor newly introduced drugs over older ones.

Let me show how this pricing scheme works in the case of Aloxi. The wholesale acquisition price or WAC for Aloxi is about $300, but it is sharply discounted so that its actual average selling price or ASP that physicians actually pay is about $175. The manufacturer is required by Medicare regulations to gather and submit actual billing information from accounts which is used to calculate the ASP. Physicians are then reimbursed for the ASP plus a 6% markup. In the case of Aloxi, the drug is purchased for $175 and reimbursed at $186 resulting in an $11 profit from this factor. In addition, physicians also bill for an administration fee that is roughly $20 per injection. Hence the physician receives $11 plus $20 or roughly $31 per injection of Aloxi.

Medicare part B regulations create a unique situation that is very much to the advantage of new drugs. The pricing regulations are based on historical ASP, but new drugs have no pricing history. Instead, new drugs are reimbursed at markup over WAC prices for roughly two full quarters following introduction. For example, if a drug were introduced on July 1, 2012, it would probably be reimbursed at a markup of 6% over WAC for until December 31, 2012. Reimbursement would then shift to a markup over ASP.

Let me illustrate how APF530 could benefit from this. Let us assume that it is introduced at a WAC of $300, the same as Aloxi. For 6+ months, physicians would be reimbursed at a 6% markup over WAC or $318. APPA during this time could elect to discount APF530 and sell it to doctors at perhaps $260 so that the difference between $318 and $260 or $58 would be profit for the physician. The total reimbursement for physicians would be $58 plus the administration fee of $20 or $78 for APF530; this compares to $31 for Aloxi per my previous example. Past experience suggests that a $30 difference in price can induce physician prescribing behavior and the difference in this case is $47.

This initial discount incentive lasts for six and possibly as long as nine months past the launch. What happens next? Over this time, APPA gives Medicare the necessary data to calculate an ASP. Physicians then begin to bill at a markup over the newly established ASP; Medicare is continually recalculating ASP.  However, this doesn’t immediately eliminate the incentive enjoyed by APF530. For example, if we assume that during introduction period for APF530 that the physician is actually paying $260 per injection, this then becomes the new ASP on which reimbursement is calculated. Remember that the ASP for Aloxi is $175 so that APF530 can still provide a meaningful discount. If the ASP is $260 and APF530 is discounted to $220 per injection, the physician could still realize a profit of $66 per injection versus $31 for Aloxi. Hence, this economic incentive can probably be maintained for some period past the first six to nine months following introduction.  

Market Opportunity

The occurrence of chemotherapy induced nausea and vomiting is significant. There are about 7 million cycles of chemotherapy given each year which results in the sale of 5.1 million vials of injectable drugs. Aloxi accounts for 49% of vials sold; generic Zofran and Kytril for a combined 40% and Emend 11%. The main market opportunity arising from the clinical advantage of APF530 is for use in those 30% to 60% of patients who don’t achieve a complete response with Aloxi in earlier cycles. This doesn’t mean that Aloxi is totally ineffective, but rather that the patient has some nausea and vomiting and may require rescue medications. A second and lesser opportunity would be for physicians who have strong confidence in and are loyal to granisetron. They might be induced to switch to APF530 because of the advantages of its longer acting five day formulation.

The clinical reasons are augmented by the significant economic incentives I previously discussed and these are quite meaningful. There is one example that we can look at to see just how powerful economic incentives can be. Zofran went generic in the fourth quarter of 2006 and as expected, generics quickly eroded sales of the originator product. At the same time, it also caused a quick 25% decline in the use of Aloxi. There is no clinical reason as to why this should occur. However, there was a strong economic reason as for about six months, physicians were being reimbursed at 106% of the prior branded price for Zofran of $100 but were paying $10 for the generic. After six months, the ASP dropped to that of generic and this incentive went away. Physicians quickly returned to Aloxi, which recaptured the entire market share it had lost and continued on its previous strong growth trend.

Physicians using CINV drugs broadly practice in either outpatient clinics or academic and hospital settings. The outpatient clinics are much more attuned to economics while the academic and hospital settings have more focus on price. The generic Zofran experience suggests to me that as much 25% of the Aloxi market might switch to APF530 due importantly to economics. Unlike generic Zofran, there are also strong clinical reasons to use APF530 as I have previously discussed.

I have put together a sales and unit use projection for APF530 based on the following assumptions:

1. APF530 is approved in early 2013 and launched on July 1, 2013.
2. The economic incentive of Medicare part B results in a rapid takeoff and following the experience noted with generic Zofran allows APF530 to capture 25% of Aloxi’s unit use by early 2014.
3. The added clinical benefits of APF530 result in a modest expansion in the use of long acting 5HT3 drugs.
4. The impact of having two sales forces detailing the attributes of long acting 5-HT3 drugs could also capture some market share from the short acting agents.
5. APF530, unlike generic Zofran, has staying power because of its clinical benefits. Aloxi’s loss in market share is permanent.
6. The WAC of APF530 is initially set at $300 per injection. As discounting lowers the ASP over time, it drifts closer to the $175 of Aloxi, but this takes over two years.

In the following table, I have studied historical sales trends for the leading CINV products and used the assumptions just described to make projections as to how APF530 might fit into this market. The numbers in this table are based on injections given during the first day of treatment. I have also divided the market into 5-HT3 antagonists and Emend as the latter is usually given when the 5-HT3s do not achieve control over the nausea and vomiting. A summary of the results arising from these estimates is shown below:

I have assumed that Aloxi and APF530 constitute the long acting segment of the 5-HT3 market and that most of the sales of APF530 come from gaining market share from Aloxi. I have projected a slight increase in the growth rate of long acting agents due to enhanced detailing effort following the introduction of APF530. I have assumed that the ASP for APF530 starts much higher than Aloxi and then drifts lower to $175 in 2016, when it reaches par with Aloxi. The results of my calculations lead to sales projections for 2013, 2014, 2015 and 2016 of $45.5 million, $199.6 million, $184.9 million and $182.3 million. These calculations are shown below.

There is one other element to the market opportunity that needs to be pointed out. There is currently a fourth 5HT3 antagonist on the market called Anzemet (dolasetron). When studies linked its injectable formulation to QTc prolongation, the number of vials sold annually dropped from 473,000 (10% of the market) to about 45,000. In September of 2011, the FDA warned that Zofran (ondansetron) might also increase QTc prolongation. A study to determine if this is the case was undertaken and results have yet to be published. If it is linked to QT prolongation, its 1.7 million of annualized vials (29% of the market) could be shifted elsewhere.

Those institutions currently using ondansetron are probably price sensitive and would be inclined to switch to generic granisetron. Because of the need to have at least two drugs on formulary in the event that one doesn’t work, they are most likely using generic granisetron and ondansetron as their two preferred agents. If ondansetron were to be sharply restricted, they would probably have to add Aloxi or APF530 or both to the formulary. It is difficult to judge how positive this might be. However, I would note that each 10% of current ondansetron usage that is shifted toward APF530 and Aloxi would increase use and sales of APF530 by 5% according to my assumptions.

Sales and Earnings Projections

A.P. Pharma is still pondering whether to launch APF530 in the US on its own or to partner. It is almost a certainty that the product will be partnered in Europe and parts of Asia. I have built a P&L for the go it alone strategy in the US as follows:

The other alternative is to partner APF530 in the US as well as in Europe and parts of Asia. I have run sample calculations for this scenario that indicate that the economic returns for partnering are less favorable, about half or two thirds of the go it alone strategy. This is to be expected as profits have to be shared between partners. However, partnering does have some allure as it significantly reduces the economic and execution risks.

Financial

In calculating the burn rate for 2012, I project that S; G & A in 2012 will increase from $3.5 million in 2011 to $5.4 million in 2012 as the company begins to prepare for the potential launch of APF530. R&D expenses are projected to be up slightly to $9.4 million. This results in a burn rate of about $14 million in 2012 which compares to $11 million in 2011.

If the company decides to launch APF530 on its own, it believes that it can cover the compact oncology market with 40 reps. Each rep results in fully loaded costs of about $200,000 per year or $8 million for 40 reps. The company won’t begin hiring reps until approval is gained which I project to be in early 2013. I am layering this cost on top of the intrinsic burn rate of current operating expenses. I would also expect that other costs associated with the launch could cost $10 million in 2013. This means that all of the operating costs could increase to $32 million in 2013 versus $15 million in 2012.

I am projecting revenues of $46 million in 2013. Even with the added cost of the new reps and launch expenses and the need to fund accounts receivable and inventory to I am projecting, I think that cash flow for the year will approach breakeven. Beyond 2013, I project very meaningful cash generation.

The company ended 2011 with $18 million of cash. Through the four quarters of 2012 and through 1Q, 2013, I estimate that the burn rate will be $20 million. The company clearly doesn’t want to let cash run down to too low a level so there will be the need for additional funding. I am projecting that partnering deals for Asia and Europe will bridge the gap. If there are delays in partnering, the company may have to raise cash from debt or equity in lieu of the partnering money.

The Share Count

As of December 31, 2011, there were 200.0 million shares outstanding. However, there are a substantial number of new shares that are likely to be issued due to the exercise of options, warrant exercises and conversion of convertible notes. By early 2014, this could result in an additional 246.7 shares being issued. The fully diluted share count in 2014 could reach 446.7 million. The details are shown below:

The warrants were issued as parts of PIPE deals that were done in October 2009 and July 2011. The company also has outstanding $1.5 million of convertible notes that can be converted at the holders’ option into 37.5 million shares. These same note holders at their option before May 2013 can provide $3.0 million of additional financing in return for 75.0 million shares. I think that it highly likely that all of the options and warrants will be exercised by late 2013 with the possible exception of the 2.0 million options exercisable at $1.27 and the 4.0 million warrants exercisable at $0.88.