Accelerate Diagnostics, Inc.: The Snowball Effect Is Just Beginning
AXDX: Now Entering Accelerating Growth Phase
I think AXDX is finally approaching an inflection point in their unit sales of the Pheno system and hospitals are beginning to feel more pressure to incorporate the Pheno into their blood testing protocol. The signing of Mayo Clinic as a customer confirms the positive outlook for the big, independent ARLG study results which will further increase this pressure. Management was overly optimistic the past few years about unit sales and [repeatedly] learned just how slow and cumbersome the hospital sales process is. In general, hospitals are a mess these days and getting budget approval for even a $1,000 improvement can take over a year, so despite the compelling evidence of better health outcomes and costs savings of the Pheno, its $200,000 per unit price tag (hospitals average 3-4 units) was taking several years to work through the process. Only the well run, forward thinking hospitals were trying it out and, unfortunately for all of us, hospitals that fall in that category are few and far between. Now the rest of the hospitals are coming on board and I expect this snowball effect to only grow from here.
Quick Overview of Accelerate Diagnostics This is not a widely followed stock so it may be new to many readers. The company’s main product is the Pheno system approved by the FDA in early 2017 to ID and test antibiotics (called susceptibility testing) for bacterial infections in blood and is also approved in Europe and the Middle East. The current standard of care involves growing bacterial cultures, then ID them and then susceptibility testing and this is generally a long process. While this 2-3 day process is ongoing, severe patients stay in the hospital under observation and are treated with powerful antibiotics that act on both gram-positive and gram-negative bacterial groups (broad spectrum). These powerful antibiotics can hopefully hold off an infection until a more targeted antibiotic can be found through susceptibility testing but they often disrupt healthy bacteria and prolonged exposure can cause antimicrobial resistance (i.e. MRSA). The Pheno’s ability to do rapid ID is not really that new as other devices have achieved this using MALDI-ToF (https://www.youtube.com/watch?v=7BlqGq0bvS4) and molecular or DNA analysis. The key is in the susceptibility testing to figure out which antibiotic works best as even the same bacteria ID can result in different antibiotic treatments with some strains being more resistant than others. The Pheno does the ID and susceptibility testing all in one step (after the blood culture) and around 40 hours faster than the current standard of care. The science behind this can quickly get beyond my pay-grade but my basic understanding is the Pheno uses a very high-powered microscope (they hired a guy who built super satellite telescopes for DOD and had him basically reverse the direction) and a supercomputer (hired a few former genius Russian hackers and brilliant nuclear fission scientist) to study real-time changes in cells and how they respond to antibiotics. The current device being used for susceptibility testing in most hospitals is bioMerieux’s Vitek 2 system which is much slower.
What Changed? Time and learning, for starters. The evidence for the Pheno was compelling from the beginning but getting through the budget approval for hospitals takes a long time. Also, navigating the hospital budget process from scratch was quite the learning experience for these guys. When I ran into Tom, Larry, and John around Tucson after the Ventana Medical Sale, I would sometimes hear mention of regret for selling Ventana too soon. “If we had just run it for a year or two, we could’ve added a billion to the price tag,” was not an uncommon phrase. This time around, they were committed to not selling too soon. However, I think they learned the hard way that running the business through the launch phase is more challenging than expected. The further deterioration of the US hospital system didn’t help either. When they started selling the Pheno in the US, they expected the sales cycle to only take 3 months! Then, they amended that to 6 months, then 9 months, then add another 3 months to get the IT systems up-to-date, etc... By late 2017, they had the sales cycle extended all the way to 14 months (plus another 3-4 months for implementation)! What do you mean I need a contract with your GPO?! Since then, they’ve learned a few things, contracted with the major GPOs, and collected a mounting pile of evidence that has most hospitals now signing without a lengthy evaluation process and go-live in 4-9 months.
Rental Program: After repeatedly running into the wall of the hospital capital budgeting process, they started testing a rental program near the end of 2Q 2018 and formally launched the program in 3Q 2018. A hospital putting in 4 units, was unable to come up with $700-800k for the devices unless it was budgeted a few years in advance. The rental program successfully solved this issue and now most new contract placements are through the rental program and skip the evaluation process. The real money is in the annuity stream of the reagent sales – more on that later.
More Data and Greater Adoption: The company completed 39,000 tests on 1,850 samples showing significant results and gained FDA approval in Feb 2017. A few early adopters jumped in 2017 but most hospitals were waiting for independent studies to confirm the results and justify the expense in terms of better health outcomes and return on investment. Management now references over 100 papers and publications supporting these objectives. I couldn’t find all 100 but there are many notable independent studies including the following: Univ of Texas Southwester supporting accurate ID and susceptibility results 48.5 hours faster than standard of care (https://open.library.emory.edu/publications/emory%3Atdm8m/); Univ of Tubingen, Germany supporting fast, reliable results reducing ID time by 27 hrs and susceptibility time by 40 hrs (https://jcm.asm.org/content/55/7/2116); Virginia Commonwealth Univ Medical Center supporting accurate ID and susceptibility testing faster than standard of care (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5869823/); St Jude Children’s Research Hospital supporting high degree of accuracy and marked reduction in time to results (https://www.sciencedirect.com/science/article/abs/pii/S0732889317301918?via%3Dihub); University of Arkansas presented at the big IDWeek event in Oct 2018 supporting shortened lab turn-around time and improved clinical outcomes (https://idsa.confex.com/idsa/2018/webprogram/Paper70067.html); University of Arkansas went further in study to show that the Pheno system led to a 3-day decrease in patient length of hospital stay (https://www.contagionlive.com/news/accelerate-pheno-blood-culture-detection-system-found-to-shorten-blood-analysis-time); Peninsula Regional Medical Center found decreased use of broad antibiotics, shortened time to targeted therapy and improved use of antibiotics (AXDX management says they presented at ECCMID and their study also showed a 2 day reduction in hospital stay) (https://www.escmid.org/escmid_publications/escmid_elibrary/material/?mid=68146); Emory University supporting fast and accurate ID and susceptibility testing (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5869849/). The list goes on. I couldn’t find a single study showing bad results for the Accelerate Pheno but if anyone knows of one, please share in the comments section. The big study that everyone is waiting for is the Mayo/UCLA study and I’ll cover this is another section. Some of the top hospitals in the US are now actively using the Pheno including MD Anderson in Houston, St Jude Children’s Research, Moffit Cancer Ctr, Children’s Hospital of Philadelphia, Children’s Hospital of Chicago, Children’s Hospital of LA, etc… As more hospitals adopt the Pheno system and more studies show positive results, the conversation at hospitals is starting to shift from “well that Pheno system would be nice so maybe we’ll get it in a few years” to “oh crap, what is our liability if we don’t have a Pheno system already in place.”
Living in Tucson, I see a constant flow of new hiring going on at AXDX but the recent hire of Jack Phillips is noteworthy. Jack is 54 years old in the prime of his career as President and CEO of Roche Diagnostics North America and decides to leave in order to join AXDX. These guys all worked together back in Ventana and Jack led the growth in the commercial launch when Roche acquired them. He knows the industry very well, has one more good run left in his career, and wouldn’t be making this jump (and buying an extra $1mm of AXDX stock with his personal money) if he didn’t have full conviction in its success. I also know that Larry’s passion is more in the start-up side of things, so wouldn’t be surprised if Jack takes over as CEO at some point.
The ARLG Mayo/UCLA Study This study was completed in early May of this year by Mayo/UCLA with the Antibacterial Resistance Leadership Group (ARLG). This team announced some interim analysis at the ECCMID conference which showed good data but it was still early. The principal investigator did reach out to AXDX in 4Q 2018 and ask permission to use this study to solicit more NIH funds for ARLG and to support a possible bigger study extending it internationally. Also, instead of just releasing the results at some minor conference, they decided to wait until the biggest event of the year, IDWeek in October, to publicly announce the results. Management just announced that Mayo Clinic is now a customer. This is a very positive confirmation that the study results are significant enough to take action and put the Pheno system in place. Mayo provided 400 of the 500 patients used for this study. This is not an easy hurdle to overcome, as the same ARLG group did another big Mayo study published in October 2015 analyzing the Biofire device and while they found some improvements, they basically concluded that using this device did not improve mortality, length of hospital stay, or costs. Mayo Clinic did not implement the Biofire into its lab protocol, however, Mayo just signed a broad contract with AXDX in order to facilitate the Pheno across all of its clinics (up until this point Mayo has used a manual process for susceptibility testing). I believe sales were already reaching a positive inflection point in the 4Q18 but many hospitals were sitting on the sidelines waiting for the results of this study. Putting these pieces together, I think it is very likely that this study will conclude what many other studies have already shown: that time to antiobiotics is accurate and faster using the Pheno AND we should also get some positive secondary endpoints for improvements in mortality, length of hospital stay, and hospital acquired infections.
What Is AXDX Worth? Hard to value at a still early stage, but we’ll take a crack at it. Think razor and razor blade model (or printers and ink cartridges). Larry likes to say they are Gillette and not Schick meaning they are ahead of the curve. Devices are being rented and I don’t think they make much profit there, as the real money is in the reagent annuity stream. Here are the devices under contract so far:
There are currently around 20,000 devices in this category globally with bioMerieux’s Vitek 2 having the largest market share. But if we just look at North America and Europe right now, there are around 4.2 million blood culture samples tested each year. I’m going to assume that they achieve 50% market share over the next 5 years (a little less than what Vitek 2 achieved). Each test costs $200 to run and the gross margins on the reagents is around 70%. This successful scenario values the US/EMEA Pheno business around $50-$60 per share in 2024.
They are beginning clinical trials in China for the Pheno in the next few months and expect sales to begin in 2021. The Chinese market is almost as big as the US, Europe, and Middle East combined. Given the growing pile of evidence, I think approval in China is likely. I don’t understand the nuances of the China market as well, so let’s only assume 30% market share for China. That adds another $30-$35 per share in value.
The company is beginning US trials on using the Pheno for severe pneumonia respiratory infections this quarter. One of the main challenges in this market is getting a consistent sample as mucus can be harder to work with and there is no clear gold standard approach to this as currently samples are just smeared on slides. I saw them do a live demonstration of a centrifuge device that separates out the bacteria (which is more dense). For the trials, they have the added step of then taking the capsule from the centrifuge device to the Pheno 2 (more on that later) but the hope is to consolidate it to just one machine eventually. The respiratory infection market is not as big as the blood infection side of things and they haven’t given any indication of pricing and margins. We should get trial results around the middle of 2020 and we can see how much value that might add. They are making some progress on the urine infection side of things but aren’t ready for trials yet (also, there is more competition in this space that seem further ahead than AXDX like GeneFluidics and BacterioScan). They are also making some progress on quick ID of slower growing bacteria (i.e. joint infections) using a gene sequencing approach and running a small electric current through the protein strand to ID (a passion for John, the chairman, who had a personal experience with slow ID/AST of a joint infection that almost killed him).
The company currently has 6 different possible models for the Pheno 2. They all have an upgraded supercomputer inside (processing multiple terabytes of data real-time) and use more of a holographic microscopy approach which can handle much more volumetric samples (i.e. urine and mucus). The different models mostly vary on shape and size with all of them being more vertically condensed to fit better in small lab spaces.
Short Selling AXDX is a popular short selling target right now. The main reason for this, to my understanding, is the limited volume of float (insiders own almost half the shares) and the lack of a large institutional investor base make it easier to manipulate the trading price. My firm participates in securities lending for AXDX shares through a few big brokers which gives us some additional insight into this. From our contacts in trading operations, we hear that 6-7 different hedge funds are trying to play this game at the same time. Who knows how long they can go at this as they have been very active since March of this year and have even ramped that up recently. The last time they played this game and were this active was from around July 2017 to October 2017 and then a big short squeeze shook them out in 4Q 2017. Given the growing pile of evidence that will lead to greater adoption of the Pheno, I think another short squeeze is in the making. I wouldn’t want to be the last hedge fund holding the bag.
Summary The evidence is overwhelmingly that the Pheno provides fast and reliable ID and susceptibility testing for blood infections. The evidence that the Pheno results in better health outcomes and reduced hospital stays is growing and confirmation of this by the Mayo/UCLA/ARLG study will be very significant. Sepsis is a major issue with studies showing it impacts over 30 million people per year and getting faster susceptibility results means the patient gets the right antibiotic treatment quicker. Hospitals are finally feeling pressure to move more quickly in adopting the Pheno to reduce liability risks. Observation hospital stays are not reimbursed by Medicare so as additional studies confirm reduced length of stay, then hospitals will have more confidence in return on investment calculations favoring the Pheno. I believe a 50% market share in the US/EMEA is within reach over the next 5 years which would value AXDX in the $50-$60 range. China should come into the fold in 2021 and has the potential to add significantly to that value. Additional value may come from respiratory infections, joint infections, and urine infections but it is still too early to bank on those yet. Insiders are smart, successful businessmen who have put their own money into this company and currently own 47.6% of the company and continue to buy more. The two largest owners, Jack Schuler and John Patience, are 78 and 71, respectively and word on the street in Tucson is they turned down a takeout offer in the mid-$30s in mid-2017 but would accept something in the $50 range. The snowball effect continues to build, bolstered by more independent studies and adoption by top hospitals, and I don’t think they’ll have to wait too much longer.
This information does not constitute investment advice or an offer to invest or to provide management services and is subject to correction, completion, and amendment without notice. Harrison Wealth Management makes no warranties and is not responsible for your use of this information or for any errors or inaccuracies resulting from your use. The opinions offered herein are for illustrative and educational purposes only and not personalized recommendations to buy, sell or any security. As I have no knowledge of individual investor circumstances, goals, and/or portfolio concentration or diversification, readers are expected to complete their own due diligence before purchasing any stocks mentioned. This material represents an assessment of the market environment at a specific time and is not intended to be a forecast of future events or a guarantee of future results. Harrison Wealth Management may recommend some of the investments mentioned in this article for use in its clients' portfolios.