Brief Overview and Investment Thesis
AspenBio Pharma (APPY) - soon to be renamed to Venaxis - is the maker of AppyScore, a unique blood test that is designed to help diagnose acute appendicitis. Recently, management announced that they had met with the FDA and were planning on embarking on a large study to further assess the utility of their AppyScore blood test. In addition, the company will be shedding their other endeavors (focused on animals), revamping their board, renaming the company (to Venaxis), and focusing exclusively on advancing AppyScore to the next level.
A recent thorough analysis by an SA contributor outlined a bull case for this company. Several reasons for a rosy outlook were given including: 1) recent substantial insider buying, 2) upcoming FDA pivotal trial in the third quarter of 2012, 3) promising pilot study of AppyScore, 4) lack of competition, 5) blockbuster market potential, and 6) historically low valuation. I will not rehash these points - but instead, will fabricate a logical argument of potential success based on the data available.
However, as some of the commentators of that article noted, there are many red flags with this company. These include massive loss of shareholder value from numerous share plunges. Much of the devaluations have been due to poor performance of the test in the past. The reason for this was that the test was based on one lab study initially, then two different ones. In fact, there was so much anger in the investment community that the company was sued on two occasions for violation of SEC rules/regulations - both cases have recently been dismissed. There have also been two reverse splits in the last 12 months that have contributed to loss of shareholder value. Because of these ghosts of the past, the stock price has been quite volatile as of late - despite some good news from the company.
The management recently changed the AppyScore to be more accurate. Initially, it was a single lab test. More recently, however, the company modified the test to be a combined value of three separate tests (white blood cell count, c-reactive protein, and mrp 8/14). As will be outlined herein, this new test is much better and has a high likelihood of being a useful and reliable test if the results hold up in the next trial.
Scientific Appraisal of Data
The rest of this article will focus on the most recent pilot study that was conducted. It will synthesize the results based on the company's multiple presentations (found here and here) in the recent past.
In 2011, the company performed a pilot study of the new 3-measure AppyScore at 11 hospitals (details found in above links). They recruited 503 patients who came into the ER with abdominal pain. The prevalence of appendicitis in these patients was 29% (144 patients). They found that the test had a sensitivity of 97%, a specificity of 43%, and a negative predictive value of 97% in that patient population.
From their data, one can extrapolate the following 2x2 table of results:
|Positive AppyScore||Negative AppyScore|
It requires a bit of statistical knowledge and decision analysis to be able to make heads or tails out of these results. For a primer, refer to the Wiki pages (sensitivity and specificity, negative predictive value, and positive predictive value). Most would look at the 43% specificity number and automatically think that the test is no good. However, the power of this test is the high sensitivity: the probability that a patient with appendicitis will have a positive test. This is very high - and some would argue that it is higher than a CT scan. The reason for this is that a CT scan still requires interpretation - and if a non skilled radiologist looks at a CT scan, he/she may not be able to pick up a subtle case of appendicitis. To diagnose a case of appendicitis, the radiologist has to actually see the appendix - it either looks inflamed or it looks normal. If the radiologist cannot see the appendix (as happens in some cases), then the radiologist cannot make the call one way or the other. In contrast, this test puts out a negative or positive value and so it is more reliable from that perspective.
Negative predictive value is the probability that a patient with a negative test result will not have the disease. This value is a direct function of the sensitivity of the test and the baseline prevalence of appendicitis in the ER patients who present with abdominal pain. In this study, prevalence was 29%. The negative predictive value depends on the prevalence. If only 1% of the patients actually had appendicitis, the negative predictive value would be very high - however, if 90% of the population had appendicitis, the negative predictive value would be relatively lower.
Why does all this matter? When a physician evaluates a patient for potential appendicitis, there are 3 outcomes: 1) definitely not appendicitis, 2) definitely appendicitis, 3) maybe appendicitis. In the first two scenarios, the choices are obvious - the first group goes home and the second group gets an operation. The third group needs to be further evaluated with imaging. This is where the power of AppyScore can be substantial. Many of these patients in the 3rd group would undergo CT scanning - which produces a lot of radiation and, as recent additional studies have shown, can cause a marked increase in the incidence of cancer. As a result, it is preferable to avoid CT scans - especially in children.
In most pediatric institutions, CT scans are not the first mode of imaging. Almost always, an ultrasound is performed first. Ultrasound is a great test to confirm appendicitis but a terrible test to rule it out. So, if a radiologist sees a big, inflamed appendix on the ultrasound, it is done - patient gets an operation. If however, the radiologist doesn't see this - it does not necessarily mean that the patient doesn't have the disease. Many of these patients would then end up getting CT scans. Ultrasound for appendicitis is a perfect example of a test with a high specificity and a relatively high positive predictive value. However, it has terrible sensitivity because in a large number of patients who have the disease, the result is negative. Nevertheless, despite this major shortcoming, it is utilized universally (by the way, this is a perfect example of an imperfect test that has widespread use - the key is that physicians understand the limits of the test and know how to interpret the results accordingly).
In the latest Aspenbio study, of the 503 evaluated patients, 63 were sent home without imaging and 43 had operations without imaging. These two patient groups fit into the #1 and #2 groups outlined in the second to last paragraph above. The 'maybe' group comprised 397 remaining patients, many of whom had some sort of imaging. The prevalence of appendicitis in this group was similar to the overall group, about 25.9%. Based on the AppyScore, 130 of these patients would have had a negative test result - and of these, only 3 would have had appendicitis (NPV of over 97.6%). [Note: This number is higher than the overall trial results because the prevalence of appendicitis in this group was slightly lower.] The major point here is that with all the imaging modalities used, including CT, ultrasound, etc, still 5 patients with appendicitis were sent home. If this test was used 2 fewer patients with true appendicitis would have been sent home with negative results. This issue underscores the power of this test - it is better than what we have now and it can make decision making easier for the clinician.
Ultimately, this test will be like a d-dimer test for diagnosing pulmonary embolism. This test has high sensitivity (93-95%) but low specificity (50%) - these properties (similar to AppyScore) give it a high negative predictive value: if the test is negative, that means the likelihood of disease is very low. So patients with negative d-dimer are highly unlikely to have pulmonary embolism - easy, simple test that detects a potentially lethal disease. Widespread use!
Applying the same logic to AppyScore, the physician can rest assured that the likelihood of missing an appendicitis is very low - and can avoid doing an unnecessary, and potentially risky CT scan. However, if the test is positive, it doesn't necessarily guarantee that there's appendicitis - in that case, additional imaging will be necessary.
AppyScore is also a fast test and would allow the patient not to linger for hours in the ER while waiting for a CT scan. Early, safe discharge = happy physician who has to deal with one less patient + major cost savings by avoiding hours of waiting. This is an extremely useful test - and if results are confirmed, ER administrators and physicians will flock.
As with any small biotechnology company, there is substantial risk here. The company has a long history of failure and mistakes. The data from the initial pilot studies was also promising - yet the pivotal trials failed to show good outcomes. However, the upside potential remains quite substantial. The newest pilot study had a large number of patients - one has to take a very logical approach in terms of risk/benefit ratio. We will do a simple outcomes analysis below.
The company has obtained the nod from FDA to proceed with a larger study to better quantify the characteristics of this test. The new study will be larger and should better delineate the sensitivity and specificity of this test. If these results are promising, it is highly likely for the product to get approval - and the market for it will be substantial.
Simple Outcomes and Risk Analysis
This is a very simple analysis that (if done correctly) will pay off in the long term (worked for Sarepta Therapeutics). It will fail - for sure - in some instances. But over time, if you stack the chances in your favor, it will pay off well in the end.
This is a rough analysis: The stock is now trading around $2.80. The most recent low is about $1.33, so if the trial fails, the price will fall to around $1.00. The company is currently trading at valuation of about $20 million. In the past (when there was more promise), the company was valued at around $400 million. If the upcoming trial is successful and FDA approval and CE mark are garnered, the stock has the potential to trade at previous valuation (actually probably higher, but let's be conservative), indicating a stock price of $40-50. What is the lowest likelihood of success of this product to make an even trade?
Probability of success x payoff = probability of failure x loss
Let's call probability of success x; we know that probability of failure = 1 - x
($40 - $2.80)x = $1.0(1 - x)
x = 2.6%
So, if you think there's a greater than 2.6% chance that the trial will be successful and the product will be approved, you should take this bet. You can play around with these loss and gain numbers as you want and see what you come up with. The insiders have done this math and a bunch of them have committed to the stock at just less than $2.
Final Words of Caution
Investing in small biotechnology companies is highly risky. I suggest you do your own research before investing.