Description

Applied Genetic Technologies (AGTC) is a small gene therapy company focused on rare ocular conditions.  Valued at only $63 million ex. net cash, AGTC has clinical programs in 2 conditions that have shown what I would call good-enough data which the market has ignored. In particular, even though one of their trials is not finished, they have already told us that there is enough data to advance it to a late-stage pivotal, but that message might have been lost by investors waiting for catalysts like “report of final trial data” or swamped by heavy ATM use which will find out when the 10-Q is filed.   Their platform technology also has pre-clinical programs in several other areas including partnerships with smaller companies in opto-genetics (therapy + device+ light = good) and ontology (hearing).  While ocular gene therapy makes sense because it does not have broader systemic delivery issues, it has still been a tough space for others.  AGTC itself had a 2015 partnership with Biogen for a disease called XLRP that was cancelled and the rights returned in 2019 after BIIB acquired Nightstar Therapeutics for $877 million to work in the same area.  One would think that was a black mark.  Meanwhile Nightstar’s programs failed in 2021 and have been shut down while AGTC presses forward.  Several other recent small partnerships and out-licensing deals also help to validate the potential of the platform technology.  As its programs progress I believe there will be an upward re-rating, perhaps quite significantly, in AGTC going forward.  I will discuss more detail below. 

Business Description

Like other gene therapy companies, AGTC’s core technology platform involves proprietary promoters and capsids to deliver public domain gene of interest via an AAV (adeno-associated virus) vector.  AGTC manufactures its own vectors and has significant experience in design and selection, particularly for genetic diseases of the eye.  Unlike many newer companies, AGTC has been working in gene therapy since it was founded in 1999.  An earlier iteration of its technology was partnered with Genzyme from 2010 to 2013.    The company went public in 2014 at $12 per share. Since then it stock has been as high as $30 early on and since 2017 has been in the $3.50 to $8 range with occasional high volume movements to the upside and to the downside.  At just over $3.60 it is at the lower end of where it has ever traded.  So what happened since its IPO?  In 2015 it announced a large partnership with Biogen for two indications.  Unfortunately one of the indications did not work in 2018.  In 2019 Biogen almost simultaneously cancelled its partnership and acquired a competitor called Nightstar Therapeutics which was working in the other indication covered by Biogen’s partnership. (As noted above that might not have been the best move as Nightstar’s program recently blew up while AGTC looks promising).  Nevertheless, I think that fact pattern helps to explain the low valuation of AGTC today.  Competitively MeriaGTX (MGTX) is partnered with JNJ (Janssen) in overlapping areas (XLRP and ACHM A&B –see below).  It is slightly behind AGTC and there are other deep in the weeds reasons to prefer AGTC’s approach but there is no doubt that MGTX has far superior resources and it is a horse race.   Also newly public 4D molecular (FDMT), a former AGTC partner (more like a supplier of candidate vectors) from 2015 until AGTC ended the partnership in 2018, also has an earlier stage program in AGTC’s largest indication (XLRP)

First indication: XLRP

The lead and largest indication for AGTC is X-Linked Retinal Pigmentosa (XLRP) a genetic subset of the degenerative blindness condition Retinal Pigmentosa.  First a patient loses their peripheral vision and ultimately a man goes completely blind (only one copy of the gene).  AGTC is targeting the RPGR gene.  There are an estimated 20,000 cases of XLRP in the US and Europe.     

AGTC has reported data from its ongoing Phase 1/ 2 trial that now includes an extension requested by the FDA to gather additional patient response information.  The most recent information was released in May of this year.  At the highest dose levels and utilizing the criteria being used for the next trial, 4 out of 8 (50%) patients responded.  The measurement is an improvement in a retinal sensitivity measurement.   There were also responders at lower dose levels and two out of two are seeing a persistent and durable response after 3-years (this is the longest data of any XLRP program).  The company is also measuring something called the BCVA (Best Corrected Visual Acuity) which is clinically relevant but to its knowledge is not being used by competitive programs.  The company is now launching two additional clinical trials called the Skyline (NCT03316560) and the Vista (NCT04850118).  Skyline is an additional 12 patient Phase 1 / 2 (really an add-on to the existing trial) that will validate an additional outcome measurement that was used in the approval of Spark/NVS’ Luxterna: a mobility maze. It will also be starting a Phase 2/3 called Vista with two dose level groups of 20 each and a 20 patient untreated group.  The company will report masked 3 month data from Skyline in the fourth quarter of this year and 12 month data in 3Q22.  Meanwhile, AGTC expects to provide masked interim 6 month results from Vista in the fourth quarter of 2022.  Management indicated that if the results in 4Q22 are sufficiently exciting, that six month Vista data might be sufficient to file a BLA with the FDA for approval.  

Competitor MGTX also has an XLRP program for which it plans to begin a Phase 2/3 trial later this year (although I could not find it yet on clinicaltrials.gov).  The data from MGTX’s Phase 1 /2  is not apples to apples but it showed improvement as well in its low and medium dose groups.  Its high dose group saw deterioration and adverse events related to inflammation and will not be used in the next trial.  AGTC is advancing a higher dose group that had zero safety issues.  It also believes its capsid is superior and in non-human primates showed a 2 to 3 fold improvement of transgene expression relative to those used by MGTX and the now mothballed Nightstar which could tie to improved efficacy.  The key fact for the moment is that AGTC appears to be ahead of MGTX.    

Second and Possibly Third Indication: ACHM-A and ACHM-B

The second disease that AGTC is targeting is actually two distinct programs going after two discrete genes but I will lump them together here.  The condition is called achromatopsia (ACHM) where a patient has no functioning cones in their eyes.  In addition to being legally blind, ACHM is characterized by extreme sensitivity to light such that even in standard office lighting a patient needs to wear sunglasses or tinted contacts.  There are several genes that separately can cause ACHM when mutated so AGTC has two distinct programs for the two most common genes (call them A which affects 25% and B which affects 50%) which together represent 75% of the 27k patients in the US and EU who suffer from the condition. 

AGTC has completed and just reported on 12 months of data for adults and some pediatric patients in 2 discrete Phase 1 /2 trials (it is awaiting a last pediatric cohort to be reported in 4Q21 to be officially complete).  The therapies were given only in one eye and there were no treatment related severe adverse effects.  The primary endpoint of these trials is the standard visual sensitivity by static perimetry (a test of a person’s ability to see objects of various brightness levels at the end edges of the visual field.  It is also measuring at an important clinical secondary endpoint that measures the most important consideration for patients: light discomfort (it uses a device called OPA for ocular photosensitivity analyzer).  The most recent data was positive but with nuance in that ACHM-B definitely looked good but ACHM-A may or may not turn out to be a dud.  Interestingly, this data was reported very recently on June 24 and the stock price did not move at all although volume spiked (my guess is that the company was highly active in raising capital through their ATM facility).  

In the ACHM-B trial, improvements in the primary retinal sensitivity were reported in the treated eye vs. the untreated eye in four of 11 patients (44%) from the high-dose and pediatric cohorts (which are the cohorts that will be used for the next trial). Improvements in the secondary degree of light discomfort were also observed in six of these 11 patients (66%). It looks like dose might matter as the lower doses were not highlighted.  Now, AGTC is the only company to have reported clinical improvement in ACHM-B and it is the only company looking at light discomfort as a secondary endpoint.  As an aside, several patients also saw improvement in their untreated eye which may have to do with how the brain adjusted cortical function with the therapy.  Since no drug went to the other eye that would be really cool. GeneSight “saw” something similar in its most recent trial with an ocular gene therapy (different / non-competitive indication) that reported in December 2020.   The company is now drafting its end-of-phase 2 FDA briefing packet and getting ready to ramp up production for the next trial (hopefully the last one). 

Meanwhile interpretation of the 12 month ACHM-A results are more complicated.  On the face of it, this trial did not work.  But it may not be game over because scientists noticed that 1 patient out of 20 that did respond had a different mutation in the relevant gene than all of the other patients.  Most patients had various mutations that made a messed up protein that might have interfered with the activity of the new protein made by the inserted gene.  The single responder had a genotype that did not make any protein at all which is similar to how all ACHM B mutations work (where the other trial was successful).  Furthermore, pre-clinical work showed that eyes in younger animals responded better to the therapy even when there were mutations that introduced “debris.”  Thus the upcoming 3 month high dose pediatric data may provide a path forward and we will know at year end.     

From a competitive perspective, MGTX has programs in ACHM-A and B and is expected to report its first data later this year so it is slightly behind AGTC.  These programs are not highlighted as top 2 ocular programs in the summary investor presentation but they are listed on the pipeline slide.  It looks like in type A MGTX is only dosing pediatric patients. 

Other Programs:

The company also has three small partnerships (one with a clinical asset), one other interesting initiative, and some preclinical efforts. 

First, since 2017 it has teamed with small private optogenetic company Bionic Sight in Retinitis Pigmentosa (degenerative disease with 200,000 patients in the US and Europe of which XLRP is a small subset).  In March of this year Bionic Sight reported that all four of the first four patients in its first Phase 1 -2 clinical trial (which had been delayed by COVID) had the ability to detect light and motion and two were also able to detect direction.     As part of the deal AGTC funded $2.2 million of work and invested over $6 million in two installments and now owns 15.5% of Bionic Sight (~$32.5 million valuation) and has right of first option to license or partner the program after the current trial is completed.

Second, since October 2019 it has collaborated with Otonomy for gene therapies related to congenital hearing loss. It is a 50/50 split of expenses and future economics.  Pre-clinical proof of concept data was presented at a May 2021 scientific conference for a therapy that would address 30% of congenital hearing loss cases (120,000 addressable in the US). It can make genetically deaf mice hear again.

Third, in April 2021 it non-exclusively licensed its cone-specific promoter technology to French company SparingVison for two non-competitive indications (and an option on a third) in return for a future undisclosed upfront fee, milestones and royalties.  If anything this is excellent validation of the usefulness of the AGTC platform.     

Next is what I would call an interesting initiative because it is like a free option.  Going back to October 2018 AGTC abandoned its program in X-linked retinoschisis (XLRS) because its phase 1 / 2 trial did not meet its efficacy endpoints.  (This program had also been part of the 2015 Biogen partnership and its failure probably contributed to the 2019 termination).  Subsequent work suggested that the intravitreal injection delivery method might not have been ideal and that a sub-retinal method might work better.  In April 2021 AGTC licensed the XLRS program to TeamedOn International who will be carrying forward a sub-retinal program with the existing therapy with no cost to AGTC. There is potential upside from developmental milestones and royalties.  XLRS has an addressable patient population of approximately 35,000 patients in the US and Europe.  

Finally, I won’t dwell on its internal earlier stage programs but AGTC has several preclinical efforts in other ocular indications (Dry AMD and Stargardts disease) as well as 2 neurodegenerative conditions FTD and ALS.  There are certainly other groups working in all of these areas. 

Rough Valuation

A very rough order of magnitude estimate of the total undiscounted market opportunity after-tax for its 3 lead programs would be over $80 per share if things go right.  Cut that in half for PV (assume only one treatment ever needed and not everyone will get it the moment it is available) and then another 75% for risk-adjustment/limited penetration due to competition and the resulting price would be $10.50 per share.  XLRP accounts for 57% of the value, ACHM-B accounts for 40% and ACHM-C in this scenario is very small and is only 3% of the value.  Here is how I thought about this so you can use your own assumptions:

First there are now 46.6m shares including full credit for all currently out of the money outstanding options.  From there, I assumed that the $50m ATM would be fully utilized around $3.75 per share for an additional 13.3m shares.  Then I added another 13.3m shares to assume it would cost $100m total to execute net of $9.9m of debt related to the facility build out.  This is a total of 73.2m shares (so an additional 36% dilution).

Next I used US and Europe patient estimates for each of the 3 programs.  20,000 XLRP / 35,000 ACHM-B and then only 875 for ACHM-A (25% of all 70,000 ACHM are Type A but I am assuming only 5% have the no-protein genotype and I assume that the pediatric data does not expand the market).

Looking at the total addressable patient populations I tried to estimate a conservative gross profit per treatment relative to other revenue per treatments (with high gross profits).  XLRP is similar in size/ a little bit larger than the CAR-T Yescarta (20,000 vs 7,500) which costs $300,000 per treatment.  So I used $250,000 here.  Next ACHM-B is nearly twice as large (35,000) so I used only $100,000 per treatment.  Last ACHM-A is far more rare (n=875) so it should probably be between  Yescarta’s $300k and Kymriah’s $480k (300 possible patients) but I also used $250k.  Doing that math is $8.7B.  Assuming a 30% tax rate and applying my share assumption is $83.30 per share.  $41.65 after a 50% haircut for present value and then $10.41 after another 75% to risk-adjust and assume that Meira/JNJ will also be successful and share the markets.  

Note that I valued all other programs at zero and I have assumed no pharma partnerships on the late stage assets. For reference, the 2015 partnership with Biogen on XLRS and XLRP could have been worth $1 billion (was $94m upfront and a $30m equity investment at a 27% premium) to AGTC at an earlier stage of development and coincidentally the size of the combined XLRP and ACHM-B populations is exactly the same total number of potential patients.  Lastly it is irrelevant today but there are meaningful NOLs and tax credits relative to the value of the company (undiscounted after-tax ~$60m). 

If things go wrong, cash per share is roughly $2.36 as of the last Q but going lower due to cash burn – it would be roughly 33% down from current levels before it becomes a negative EV company (which also happens). 

Investment Positives

Efficacy of 50% in XLRP and between 44% and 66% as seen in ACHM-B seems sufficient if each holds up in its next trial since there are no alternatives.  For reference Novartis/Spark’s Luxterna, a $850,000 gene therapy for a different ocular condition, was approved after a 31 patient trial with 21 patients treated and 10 in the control arm.  After one year. it looks like it improved sight more than the control on roughly half the patients (see the clinical trial section of the FDA package insert) although the headline was that “after 2 years almost all study participants saw improved sight.”  At 1 year 90% of the treatment group saw improved sight but 40% of the control group did too.    

The focus on rare diseases means smaller trials and more rapid progress relative to other indications.

Other companies want to work with AGTC: it licensed its promoter technology to SparingVision for non-competitive ocular programs.

A new manufacturing facility announced in 2021 shows the company is maturing into later stage development.

The company is partnered with one of the best facilities for ocular gene therapies that have the luxury of choosing which trials it will participate in (Oregon Health & Science University’s Casey Eye Institute).   

Confusion over XLRS’ delivery method vs. other programs/Biogen partnership cancellation muddies the water. For completeness, it is also worth noting that AGTC had been partnered with Genzyme from 2010-2013 on a first generation wet-AMD therapy that did not move forward. 

Other programs are earlier stage but their existence further mitigates binary risk.

AGTC can tap a new $50m ATM and maybe it has already given that $83m of volume took place between June 23 and June 25 when the ACHM data came out.  We will find out when 2Q21 is reported.  It is possible that overuse of the ATM on those days kept a lid on the stock price when the positive 12-month ACHM data was disclosed.

There may be ways to further improve patient selection for the next trials such as retinal morphology but we do not have sufficient details yet.  This could increase the chance of success even more. 

There are no warrants outstanding.

If the trials fail to show improvement (the endpoints), the long term data may still have commercial value if it shows slowing or halting of progression given that the diseases are degenerative.  

Investment Concerns

Another company (MGTX) partnered with big pharma (JNJ’s Janssen) has 100% overlap with AGTC’s two main diseases.  There are technical differences between the programs and AGTC is slightly ahead but MGTX is well resourced.    

XLRP program not fully funded as of 3-31 with a pivotal trial starting.

If AGTC fully uses its $50 million ATM without regard to letting the stock appreciate (which is the way some other companies generally use it) it could keep a lid on the upside until it is exhausted. 

Trial numbers are very small and this is biotech.

Management turnover: the CFO left this year for an oncology start-up and the CSO just moved to EDIT (a much larger gene therapy company, one of the premiere CRISPR cos). The CMO left in December 2019 to go to another ocular (not gene therapy) company called AsclepiX going after Wet AMD / DME and RVO.  Perhaps turnover is the downside of having a company HQ in fairly remote part of northern Florida vs. a more traditional biotech hub. I don’t have any further insight but on balance it is not a huge vote of confidence. 

Expensive rare disease drugs may see extreme pricing pressure in the future that could limit the attractiveness of orphan indications.  Gene therapies are amongst the most expensive.  

Catalyst

More data / progress in lead programs

Upcoming R&D day on July 22^nd (possibly more details)

Quarterly results to see if ATM was used to bridge capital needs through end of XLRP phase 2/3 trial.