(Editor’s Note: This article covers a stock trading at less than $1 per share and/or with less than a $100 million market cap. Please be aware of the risks associated with these stocks.)
- Abdominal pain accounts for over 10 million visits to USA emergency rooms per year.
- Computed tomography imaging is used to diagnose approximately 35% of these cases, but comes with a heavy radioactive price tag.
- Accuracy of computer tomography imaging is measured at 93-98%.
- Venaxis, Inc. is developing a promising diagnostic tool showing ~97% accuracy in negative test results at a significantly cheaper financial cost and zero radioactive side effects.
Venaxis, (NASDAQ:APPY) has outlined a clear problem within modern hospital computed tomography (CT) imaging use. Risk for this diagnostic procedure is extreme, and constantly understated - especially with children. Studies have shown that a single CT imaging study can be as detrimental as having a career as a radiation worker within the nuclear industry. APPY is developing a new diagnostic tool, called APPY1, to address this significant issue. Scouring of previous articles relating to this tool, as well as the Twittersphere, demonstrates a significant lack of understanding. Thus the purpose of this article is to better communicate the diagnostic use for APPY1. There is a great need within the financial community to best explain APPY1, since many simply do not have proper awareness. But first, some general housekeeping.
The Coming Data
Data are to be released in March 2014. It is critical that every investor know this. As a micro-cap stock, traders who are very risk averse should not hold through these binary events. Only risk what you are willing and able to lose. A bullish case for APPY data remains, but can never be guaranteed. Do not hold or buy if you are not willing to take on a significant amount of risk.
To ease the biotech investor mind, one must be comfortable with the cash position of a prospective investment. When it comes to APPY, the chance of dilution in the short term is very minimal. As of the company Q3 2013 conference call, the cash, cash equivalents and short-term investments totaled approximately $17 million. The burn rate per quarter is around $3 million. From a recent interview on Seeking Alpha with the company CEO, it is estimated that cash will carry the company into early 2015.
The Statistics Behind APPY1
If we are going to understand the diagnostic tool, APPY1, then we need to get our heads around the statistics first. Usually this has investors running, but I'll help by keeping this (relatively) quick and painless. As per the company slides, recent data from 2011 has shown the following statistical significance for APPY1.
Image from: Company Website
Before mulling over the numbers, I find Wikipedia can usually help. Here's what it has to say regarding sensitivity and specificity - and what exactly they mean.
Image from: Wikipedia
Okay. Now we can dissect these results a little better. The reported negative predictive value (also called NPV) for APPY1 is reported as 97%. This means that for 100 negative appendicitis test results, 97 will be a true negative, and 3 will actually have the condition. APPY1 also has a sensitivity of 97%. For 100 people with appendicitis, APPY1 will flag 97 of them correctly. Lastly, the specificity of APPY1 is listed at 43%. Thus out of 100 people, 43 will be listed as a true negative (saved from CT radiation) and 57 will test positive.
To be disheartened by a specificity that is not 90+%, like the sensitivity, would be overly optimistic. The truth remains within diagnostic medicine that high sensitivity and high specificity is difficult, if not sometimes impossible, to achieve. Stay with me.
In a journal article from 2011, Lalkhen and McCluskey can be quoted as:
The relatively crude measures of sensitivity and specificity discussed previously fail to take into account the cut-off point for a particular test. If the cut-off point is raised, there are fewer false positives but more false negatives-the test is highly specific but not very sensitive. Similarly, if the cut-off point is low, there are fewer false negatives but more false positives-the test is highly sensitive but not very specific.
This is the case for APPY1. I mentioned earlier that the test used an algorithm along with three separate variables to determine a result. Let's theoretically pretend that only one variable is used - the white blood cell count. Within a healthy population, the number of white blood cells will vary from individual to individual, likely following a normal distribution. Some people naturally have high levels, some people naturally have low levels. Within the literature, it is well established that appendicitis results in an increase in white blood cells. Thus the white blood cell count for appendicitis positive patients will be higher. It may look like this:
The y-axis is the number of patients, and the x-axis is the measured white blood cell count. As you can see, those with appendicitis have a higher count than the norm. Thus we can set the "cutoff" for our APPY1 test at the higher levels, as pictured above. Anyone testing to the right of this cutoff will result in positive appendicitis from APPY1. However, we can note two things. 1) The lower end of the "appendicitis white blood cell count" distribution does not exceed the cutoff, thus this result will be negative. 2) The higher end of the "normal white blood cell count" distribution DOES exceed the cutoff, thus this result will be positive. Together, these are false positives and false negatives respectively.
If we were Venaxis, Inc., how would we stage this cutoff? Considering appendicitis is dangerous if left unchecked, it is obvious that we want to minimize the number of people with appendicitis who can potentially slip through this cutoff. The approach to this would be to stretch our cutoff further to the left of the graph. In doing so, we will minimize the number of false negatives, but also introduce a great number of false positives. We can start to understand how sensitivity and specificity are in a "tug of war."
To expand upon this, consider the following figure from the research article linked to earlier, by Lalkhen and McCluskey. This is called a receiver operator characteristic curve, and often used to determine a diagnostic test's "worth."
Within this figure we see three distinct curves: A, B and C. The A curve is deemed a worthless diagnostic tool. There is no advantage in "sacrificing" sensitivity for specificity, or vice versa. In essence, a coin toss is just as effective. The B curve is a sample of a traditional diagnostic tool used in practice. Diagnostic tests that function along this curve can be shown to have use. The C curve is a perfect test. It is the ideal, though likely not plausible. The appendicitis test by Venaxis, APPY1, is roughly plotted within this figure. Here you can see that it shows a great use.
Let me touch on one issue, though. Some investors fret that the NPV for APPY1 may not be as high within data set to be released this March. If you have learned anything from what I've explained, you would realize that this NPV is not necessarily important. It is a balancing act between what should be deemed a positive test result and what should be a negative test result. These cutoff points that APPY1 use can always be adjusted to obtain the desired NPV.
Hopefully from this the statistical nature behind APPY1 is starting to show more use. Let's discuss the potential market impact now.
APPY1 in the Market
We're going to look at some bearish arguments here regarding APPY1. The first, and most common, is the fact that CT imaging shows high accuracy. Why would one ever want to use APPY1, or even consider it, when CT imaging exists?
First and foremost, the problem that APPY1 is trying to address is the excessive use of CT imaging. With each CT image comes a dose of radiation. The use of this procedure on children and adolescents has been proven within the literature to have detrimental effects later in life.
Second, a CT image is not as accurate as you may think - or as many may have you believe. With a CT image, one is able to visualize the appendix. Now there are a few problems with this. This image is in 2D and can sometimes be difficult to interpret, since we are trying to infer something from a 3D object in 2D space. Visualizing this appendix becomes even more difficult when we take into account the fact that humans, once again, vary from individual to individual (not surprising). A radiologist must determine if a specific appendix is inflamed. This can become quite difficult when the appendix varies in size, position, and other properties.
This difficulty shows. A CT scan is not perfect. In a study published in Pediatric Emergency Care in 2000, the sensitivity for a CT scan was shown as 53% for appendicitis. Another study lists the NPV for CT imaging at 95%. What this means is if we use APPY1 and get a negative result, this negative result is more likely to be true than the CT counterpart. If you are in the ER, would you rather a negative result from APPY1 or from a CT image? Clearly if you pick CT imaging, you are just being foolish. This is where the power of APPY1 comes from. Any patient who receives a negative appendicitis result can potentially avoid the CT image, an attractive opportunity when you consider it will nuke you with radiation and cost thousands of dollars.
A second argument comes from the all too common, "even if the appendicitis result comes negative, doctors will still want to use a CT image to try and determine what the actual problem is". Well, no. If this were true, then everyone entering the emergency room would be subjected to a CT scan - since doctors want to know what is wrong, right? The study linked to previously from Reich et al. showed a CT usage of 37%. Within the company investor presentation, APPY lists a research paper by Larson et al. in Radiology (2011), where over multiple studies the CT usage ranged from 36-45%. Clearly, the idea of CT imaging as a "must use" is false.
Third, perhaps you may be questioning the adoption of APPY1 in the hospital setting. What inclination would a hospital have for introducing this extra test? I'm sure everyone has seen this slide before.
At 30% market penetration, this puts revenue for APPY1 at the current total market cap for Venaxis, Inc. ($50 million). Now, let me explain what market APPY is hoping to penetrate, exactly.
As we can see, this 30% applies only to people aged 2 - 20 who receive blood tests in the ER for abdominal pain. A blood test is utilized to measure the white blood cell count, as mentioned earlier. It has been shown that alone, a white blood cell count can achieve a negative predictive value of 82%. APPY1 is a blood test that has a negative predictive value of 97% (it uses white blood cell count + other variables to determine a result).
The adoption of this test should then be obvious. It is well within reason to imagine that one blood test can be expanded as a new and better blood test. In this way, APPY1 is not a new diagnostic procedure - but rather an improvement on one that is already well used within the clinical setting. The 30% market penetration of children receiving blood tests has PLENTY of room for expansion (i.e. >30% penetration, applied to a greater number of ages, applied to people who previously did not get blood tests for abdominal pain).
The conclusion for this report is simple. If you are/were a parent, consider your child in the emergency room with abdominal pain. If a blood test were being performed, would you want APPY1 (keeping in mind a blood test with APPY1 would be much more accurate than a standard blood test in diagnosing appendicitis)? If a CT scan was suggested, would you want APPY1 first (keeping in mind that a negative test result from APPY1 is just as powerful, if not more powerful, than one from a CT scan)? These answers should be obvious.
Disclosure: I am long APPY. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.