|Shares Out. (in M):||147||P/E||0.0x||0.0x|
|Market Cap (in $M):||460||P/FCF||0.0x||0.0x|
|Net Debt (in $M):||-10||EBIT||0||0|
Summary – The NASDAQ biotechnology index is up almost 40% YTD. This meteoric climb in the index has created a number of overvalued companies and ample short sale opportunities. This report concerns a company called Navidea (Ticker – NAVB) which I believe to be grotesquely overvalued. What makes Navidea, a company that is developing medical imaging agents such a good short is the fact that it is overvalued even if their product is approved. Since this product was very recently rejected by the FDA on9/10/12, the short is even more compelling. My incredibly generous estimate of NAVB’s fair value is $1.00. That is 68% below where the stock presently trades.
Let’s start by taking a look at Navidea’s capital structure.
The company’s capital structure consists of:
103,495,392 common shares outstanding
19,685,400 shares of common stock upon the conversion of from Series B Convertible Preferred shares
333,333 shares of common stock upon the conversion of Series C Convertible Preferred shares
3,226,000 of stock options
17,500,000 warrants exercisable at an average price of $0.56 per share
All combined for a grand total of 147,264,055 shares fully diluted for a diluted market cap of $460.9 million
The company owes Hercules $6.573 million in principal due December 2014. This note accrues at an annual interest rate of 10%. Navidea also has a lending facility extended to them by their largest shareholder. There is $15 million available on this facility at 10% per annum compounded monthly2.
$5.5 million per quarter (much of this goes to pay management’s salaries)
Navidea’s financial position is terrible and the terms on which their largest shareholder is willing to loan them money reflect this. It seems likely that the company will need to finance and this could serve as a catalyst that brings the shares closer to fair value. Even if the company draws down their credit line completely, it only delays the inevitable financing.
Recent events and regulatory history:
On September 10, Navidea announced that the FDA had issued a CRL(euphemism for rejection) for Lymphoseek3. In their press release, they stated that “the decision was focused on issues with third-party Lymphoseek contract manufacturing, and was not related to any efficacy or safety data filed within the Lymphoseek NDA.” The press release is intended to assure shareholders that all is well and implies that this issue is easily resolvable. As discussed below in greater detail, this is not likely to be the case. The statistical and methodological flaws associated with the Lymphoseek clinical studies have been well documented elsewhere and I agree with much of it4. I will restrict this analysis to regulatory concerns only.
The other aspect of this PR that is troubling to those following the situation closely is “…was not related to any efficacy or safety data filed within the Lymphoseek NDA.” Navidea reported that they were required to complete two studies for approval study 05 in breast cancer and melanoma patients and study 06 in head and neck cancer patients5. NAVB completed study 05 on July, 2009 and reported “positive results”. Strangely, they initiated study 09 in June, 2010 which had a near identical design. If study 05 worked and it coupled with study 06 was sufficient for approval, why would study 09 – a duplicate of study 05 be required? As it turns out, we had to wait until June of 2012 for an answer. A meta-analysis of lymphatic mapping agents including study 05 was published in the journal Clinical and Experimental Metastasis66 where it was revealed that Navidea had excluded 2 sites from their analysis for protocol violations “Subjects from the NEO3-05 sites 05 and 06 were excluded from the analysis…” It also explains the subtle change in language in their regulatory filings where they refer to “protocol compliant” sites. This is yet another example of Navidea not being completely forthright with investors. Sins of omission can be just as damaging as sins of commission.
What about study 06? Study 06 began enrolling in May, 2009. Navidea has stated publicly as late as May 2012 that they still have not accrued enough patients to perform an interim analysis 7 . This seems odd that a 200 patient study could not enroll in less than 3 years until you recognize that Navidea has opened 13 clinical centers but 6 have terminated enrollment8. Why would a study center terminate enrollment? Perhaps Navidea closed the center for poor enrollment. Perhaps the center closed itself for poor enrollment. Perhaps the center was committing protocol violations. The bottom line is that whatever the reason, terminating nearly 50% of the sites in your trial is always bad. Naturally, management has not addressed this beyond speculating that they do not think they need the results of study 06 to secure approval9.
What do I think? I think that FDA told Navidea way back in 2006 that they needed to conduct studies 05 and 06 successfully to receive FDA approval. Navidea conducted study 05 and then realized that if one included all of the centers, that study failed. So they excluded 2 centers from the analysis of study 05, began enrolling an identical study (study 09) and changed the language in their regulatory filings to “protocol compliant” sites. I also think that the selection of the following language in their press release on September 10 was very deliberate “…was not related to any efficacy or safety data filed within the Lymphoseek NDA.” This statement is entirely consistent with the FDA not objecting to study 09 but requesting that the company also submit study 06 with their new drug application.
Turning our attention to the chemistry, manufacturing, and controls (CMC) aspects of Lymphoseek’s history is also interesting for what it reveals. In April 2012, Navidea announced that the FDA had delayed the review decision of Lymphoseek for 3 months. Instead of the decision coming on June 10, 2012, the decision would not come until September 10, 2012. Navidea stated that the problem was related to Chemistry Manufacturing and Control issues (CMC)10, what Navidea failed to state at this time is that they had been dealing withCMC issues since 2006. Once again, analysts ignored this early warning sign that something was wrong with the application and instead argued that a delay meant that approval was all but certain.
Navidea’s management at the end of July announced a lending facility for $50 million at a 10% annual interest rate compounded monthly, 1 month prior to the FDA decision for approval. In the biotech world, when a company raises capital prior to an FDA decision it is usually a sign that management knows that the FDA is set to reject an application and it is their way of ensuring job security. On August 20, 2012Navidea announced a consulting agreement with current board member Eric Rowinsky to, among other things “Assisting with investor relations, roadshows, presentations and financings”11.
On September 10, the FDA denied Navidea’s application ostensibly for Lymphoseek for Chemistry Manufacturing and Control issues12. Navidea’s management claims that the CMC issues are easily resolvable, but I beg to differ. For starters, the FDA does not usually reject an application on easily resolvable issues. Second, the FDA already delayed the application decision and gave Navidea an extra 3 months for CMC issues, yet Navidea could not resolve them in this time frame? Third, they have publicly stated that they have had CMC issues since 200613.
Sell side analysts have perpetuated the myth that the CMCissues will be resolved in short order. This is at odds with the FDA guidance on NDA resubmissions which clearly state that resubmissions that require re-inspection of manufacturing facilities are classified as type 2 resubmissions and have a 6 month PDUFA date. The average time from FDA rejection to the time the company meets with the FDA to discuss the rejection is 2.5 months. Recent examples include MAPPharmaceuticals who received a CRLon 3/26/1214 and had an end of review meeting around 6/25/1215 and Mannkind who received a CRL on 2/25/1216 and had an end of review meeting initially scheduled for 4/15/1217. It will take Navidea a certain amount of time to remedy theCMC issues (if this is in fact possible and if this is the only information the FDA is requesting. I have my doubts). Let’s make the outrageous assumption Navidea can resubmit the NDA the day after they have their post-rejection meeting – mind you, I can find no evidence that such a speedy resubmission has ever happened in the history of FDA. That still means that the drug would not be approved until June, 2013. This resubmission timeline is incredibly generous to Navidea – it could easily be 12-15 months.
I suspect Navidea’sCMCissues are anything but easily resolvable. Given that they have been dealing withCMCissues since 2006, one cannot eliminate the possibility that Lymphoseek will turn into another 16 years of empty promises - in other words, the next RIGscan.
To properly assess the market opportunity, it makes sense to describe the contexts in which lymphatic mapping agents such as Lymphoseek are used. When a patient is diagnosed with breast cancer and scheduled for surgery, the surgeon will typically remove what are called sentinel lymph nodes. These are the lymph nodes that drain the tumor tissue and are also the lymph nodes most likely to contain cancer if the cancer has spread. To identify the sentinel lymph nodes, the surgeon will inject radiolabeled sulfur and a blue dye into the tumor area. The surgeon will then use a hand-held Geiger counter to locate the nearest (i.e. sentinel) lymph nodes and remove them. While these nodes are being removed, the surgeon may also remove lymph nodes that are labeled by the blue dye – easily identified by their color.
The lymph nodes that the surgeon removes are immediately examined by a pathologist. If there is no cancer in these nodes, the patient’s wound is closed and their surgery is over. If cancer is identified, the surgeon will remove many of the lymph nodes in the axilla. While this has the effect of increased morbidity such as swelling in the affected arm, it improves survival at 8 years by 1.5%. Not a big difference but a big difference if you are the patient.
According to the national cancer institute (NCI), there were 226,870 breast cancer diagnoses made in 201218. Not all of these diagnoses result in a sentinel lymph node biopsy (SLNB). For example, ductal carcinoma in situ may be sufficiently early stage so as not to warrant a SLNB. Breast cancer in men – about 1% of cases – also does not warrant a biopsy. On the other end of the spectrum, patients who have been diagnosed with metastatic disease at the time of presentation do not need a lymph node biopsy since the physician already knows that the disease has spread. In aggregate, about 70% of breast cancer diagnoses will require a SLNB.
According to the NCI, there are 76,250 cases of melanoma diagnosed in the USannually19. In melanoma, where it can be more difficult to identify the sentinel nodes a similar fact pattern prevails and approximately 70% will require a SLN biopsy. Taken together, the total addressable number of cases of breast cancer and melanoma in which a SLNB is warranted is 212,000. Now let’s examine how these SLNB’s are currently performed.
The current standard of care for lymph node mapping is sulfur colloid and blue dye20. Pharmalucence sells sulfur colloid and while the list price is $159 per vial, although most radio-pharmacies purchase the product for significantly less than this21. NAVB has partnered Lymphoseek with Cardinal. Cardinal has a 60% market share in radiopharmaceuticals. Navidea has stated that the partnership is such that they expect a 40% EBIDTA margin on Lymphoseek sales (source: conference call transcript8/7/12).
So where does that leave NAVB regarding sales? Let’s make some more outrageous assumptions. Let’s assume that NAVB prices this product at a 4X premium to Sulfur colloid. This is a stretch given that there is not a shred of data to suggest that Lymphoseek is in any way better than Sulfur colloid. Let’s also assume that Lymphoseek captures 100% of the market share the day it is launched. Of course, this means that they capture 100% of the 60% of the market that Cardinal controls. Finally, let’s assume that management’s guess of 40% EBITDA margin is accurate.
Eligible procedures – 130,000
Price per use - $600
Market share – 60% (100% of 60%)
Total sales – 76.64MM
EBITDA margin – 30.6MM
NAVB likely cannot be profitable on this level of EBITDA because they spend 22.5MM per year currently and their expenses will increase upon a commercial launch. We therefore cannot value them on an EPS multiple basis or a free cash flow basis – both are negative. I am assuming NAVB pays no taxes because they have large NOLs. So at what multiple of EBITDA is this of value to an acquirer? Let’s be generous and assume 5X. That implies a valuation of 153MM or $1 per share. Since Navidea has not drawn on their loan facility, I will not deduct the 30MM that that would be required to be paid from the 153MM proceeds. That assumes that Lymphoseek has a 100% probability of approval.
At this point it is worth summarizing what assumptions are embedded in the $1.00 per share fair value estimate:
1. Navidea has an end of review meeting with the FDA within 2-3 months
2. Navidea is able to fix in a few weeks theCMCissues plaguing them since 2006
3. Navidea resubmits their NDA the next day
4. Navidea receives a class 2 resubmission and a 6 month PDUFA target date
5. The FDA approves the drug in the absence positive results from study 06
6. Navidea launches the drug
7. Navidea captures 100% of the market on the day of launch
8. Navidea prices the drug at a 400% premium to the list price of the standard of care
9. Navidea does not finance or draw on their loan
10. Navidea’s cash lasts them until FDA approval
11. Navidea’s management finds a buyer of this asset at 5x EBITDA (this would involve a rational decision)
If all of this comes to pass, then Navidea is worth precisely $1 dollar per share. Is your head spinning too?
Navidea has three other programs in their pipeline. The primary goal of discussing these programs is to emphasize management’s ineptitude with respect to selecting which clinical programs to pursue. Moreover, NAVB has no internal R&D capability and is therefore limited to developing products that they can in-license.
RIGscan is a radio-labeled antibody that NAVB has been developing since 199622. After 16 years, NAVB has still not advanced the RIGscan product into a phase 3 trial that the FDA requested in 1998. Unlike Lymphoseek for which there is clinical data from other studies to suggest that identifying lymph nodes and adjusting the intra-operative treatment of the patient leads to improved outcomes, there is no such data for the use of an imaging agent in colorectal cancer. Therefore in addition to proving that RIGscan identifies cancerous nodes and is safe, NAVB must also show that this leads to improved outcomes. Absent this data, why would any surgeon expose themselves and their patients to unnecessary radiation? In fact, the FDA does not even believe RIGscan is useful, don’t take my word for it though, in 1998 the FDA and the EMA questioned23 “how the finding of additional tumor provided clinical benefit that altered patient management or outcome.”
Demonstrating improved outcomes will require a very large clinical trial. For example, the NASBP32 study which showed that axillary lymph node dissection improved outcomes in patients with breast cancer enrolled 5,611 women and took 5 years to complete24. In many cancers such as colorectal and prostate, surgeons remove some lymph nodes as a matter of course.
I speculate that the unwillingness of NAVB to conduct an outcome study is the reason that NAVB has been unable to agree on a trial design with the FDA for RIGscan. In fact, NAVB has claimed to be in discussions with the FDA over a trial protocol for RIGscan since 1998. Surely there would have been agreement by now if the design of such a trial were a simple matter. This speculation is supported by the fact that in 1997 both the FDA and the European regulator rejected RIGscan.
AZD4694 – is a labeled probe that NAVB proposes can assist in the diagnosis of Alzheimer’s disease. It is in phase 2 studies. In many ways it is analogous to Amyvid which was developed by Avid Pharmaceuticals. Avid was acquired by LLY for 300MM and potential earn outs on 11/8/201025. The problems with this program are two-fold. The first and most glaring problem is that a diagnosis of Alzheimer’s disease is not a challenging one to make. If a physician is not sure if the patient has Alzheimer’s he may prescribe a short course of donepizil. If it works to improve symptoms, then AD is the likely diagnosis. If it does not, then establishing a diagnosis of AD is less relevant because the mainstay of treatment does not work.
The second problem is that the target to which this diagnostic is proposed to bind - called amyloid – does not appear to be relevant to the disease process. We can infer this from the spectacular failures of Elan’s bapineuzumab and Lilly’s solanezumab26, 27. Both of which bind to amyloid and clear it from the body. Both of which failed in large, well-run phase 3 studies. Lilly’s purchase of Avid was likely motivated by the hope that they could expand the use of solanezumab into earlier stages of AD where patients are not yet aware of their symptoms but one can detect amyloid in the brain. Lilly does not break out sales of Amyvid but they are almost certainly de minimus given that they break out the sales of individual products that sell less than 30MM per quarter28. Given the failure of every single program that has targeted amyloid to treat AD, Lilly may be regretting their decision to purchase Avid.
Navidea enrolled its first patient in the phase 2 trial of NAV4694 on 9/18/1229. Navidea expects this study to complete in December, 2013. As far as I am concerned, this entire development program is NPV negative and highlights management’s poor stewardship of shareholder capital and refusal to admit when a product has failed.
Parkinson’s disease program – Not content to waste money on one neurology imaging agent, NAVB announced on 7/31/12 that they in-licensed a radiolabeled imaging agent to “assist in the diagnosis of Parkinson’s disease”30. After many years of lacking good therapeutics, neurologists have become extremely adept diagnosticians. For the non-physicians in the audience, a diagnosis of Parkinson’s disease does not require imaging. A test dose of the standard treatment for Parkinson’s disease carbidopa/levodopa typically clarifies any diagnostic uncertainty that might remain.
Summary – Navidea is grotesquely overvalued even under impossibly optimistic assumptions. It remains unclear as to why the market has not recognized this. Perhaps this reflects the naiveté of the retail investors who own the shares. Another possibility is that investors do not appreciate that the share count is much higher than many sources report. I doubt that it will take an approval and abysmal sales to force the shares to approach fair value though that is clearly a risk.