Pixium Vision — Pixium seeks to address short Iris II lifespan

Pixium recently reported interim six-month follow-up results on the 10-patient European Iris II clinical trial. The first patient was implanted in early 2016 and the final patient was implanted in early 2017. While the preliminary data showed that patients had moderately improved visual function when the device was in operation (compared to when it was off), the company found that for the first five implanted patients, the device stopped operating correctly at approximately nine to 12 months post-implantation. At this time point, the implant stopped communicating with the companion patient-worn goggles and sensor. This cessation of function occurred much earlier than was originally anticipated (the CE Mark was provided assuming a lifespan of three years). It expects that the remaining patients who have been implanted with the device would eventually reach the same conclusion (product operational failure within a year of implantation). The company has decided to suspend future commercial Iris-II implantations until the issue is resolved.

The firm believes it has identified a cause for the earlier than expected cessation of functionality. Fortunately, Iris II was designed to be explantable (and potentially exchangeable, replaceable or upgradeable) and hence, after explanting the first patient for which this issue was identified in or around May 2017, investigators localised a “kink” or fold in the foil at the edge of the device. It is hypothesized that during the operation of device (after implantation), mechanical forces (from eye movement) can lead to microfractures (or breaks) of the wiring emanating from the 150 electrodes embedded on the device. In addition, there may be moisture accumulation that can lead to electrical shorts and a disruption in communication.

There were some adverse events (AEs) reported in the 10-patient clinical study. One patient had a vitreo-retinal traction occur (was resolved), one had a refixation of the device (was resolved), and two had hypotony (low intraocular pressure; one was resolved at six months and one is ongoing). These AEs (frequency of 0.4/implantation) were largely manageable and generally do not affect the long-term safety or biocompatibility of the device.

Pixium proposes, for all of the 10 participants in the Iris II study and who consent to this new proposal (it is currently unknown how many participants have already provided such consent), to explant the existing Iris II implantations and to then re-implant the device with an additional application of a biocompatible seal (similar to a bandage of sorts) at the most sensitive (or potentially fragile) point of insertion, with the view that this should reduce mechanical load on the device and extend its lifespan. The company is petitioning with French and European regulatory agencies to proceed with this strategy, and expects to monitor the participating patients for approximately another year or longer, to determine whether product lifespan improves. While the company may contemplate resuming Iris II commercialisation prior to obtaining data on the re-implantations of this cohort, we believe it is not likely to be successful in re-launching Iris II until a definitive solution to this problem can be confirmed. And hence we do not expect commercial sales to resume until 2019 at the very earliest. Further, there remains a definite risk that the company’s proposed remedy to address the premature product failure may not succeed in extending the product’s operational lifespan. If the issue cannot be resolved, we are doubtful that an Iris II implant that would only function for nine to 12 months at a time (and thus require re-implantation at such intervals) would be commercially viable.

As stated above, we are pushing back our forecast for the resumption of Iris II sales, to 2019. We also believe that the presently identified issue with Iris II may diminish the commercial appeal of the product for the intended target population, those patients with retinitis pigmentosa (RP) or other severe retinal dystrophies. We have lowered our peak market share assumption from 15% to 10% within this group, and have pushed back our year of peak sales forecast from 2021 to 2022.

For Prima, the company continues to expect to approval from EU regulators to start the five-patient European feasibility study for Prima in Q417 and to have the first human implantation occur before YE17. After obtaining six-month data from study participants, the company plans to start an EU pivotal study (we assume a US pivotal study will commence in parallel). We previously assumed both pivotal studies would start in H218, but Pixium now expects the earliest time for a pivotal study to start will be in early 2019. This pushes back our EU and US Prima launch timeline estimates to 2021 (from H220, previously) and H222 (from 2022, previously). Hence, we estimate that Prima will start generating revenue (in Europe) in 2021.

With the suspension of Iris II sales and the uncertainty as to whether the present issue surrounding premature device operational failure will be resolved before 2019, we are lowering our Iris II probability of success estimate to 40% (from 70%, previously). For Prima, as stated above, we have pushed back our launch timing estimates by about six months. We note there are significant differences in device design and functionality between Prima and Iris II (subretinal layer implantation location for Prima vs epiretinal layer for Iris II), and most notably, the signal communications approach applied between the devices and their accompanying goggles headset (Prima uses a wireless photovoltaic approach, versus the wiring and cabling employed on Iris II). Despite these differences, the product longevity issues raised on the Iris II highlight the possibilities for unanticipated or premature operational disruptions to any delicate bionic vision system (BVS) device (that interacts with human tissue), due to mechanical or other unexpected stresses or factors. Further, Prima has not yet received authorisation from European regulators to commence a feasibility study; recall that the company had guided in H216 that this milestone would have been likely to occur by YE16. Accordingly, we believe it is appropriate to reduce our probability of success estimate for Prima to 10% (from 12.5% previously).

Given the above changes, we now obtain a pipeline rNPV (enterprise value) of €63.0m, down from €129.3m, previously. After including €5.4m in net cash at H117 (€14.9m gross cash minus €1.4m in conditional advances and €8.1m in long-term debt), we obtain an equity valuation of €68.5m, or €5.12 per share (compared to €10.08, previously).

We reiterate that Prima remains the dominant value driver for Pixium, as the age-related macular degeneration (ARMD) market is much broader and more potentially lucrative than the far more narrow RP market that Iris II is targeting. For instance, if we were to remove future Iris II sales from our forecasts and valuation, our pipeline rNPV would decrease by only €5.9m (to €57.1m). While the firm’s announcements and issues on Iris II do not explicitly or specifically affect the Prima programme, as stated earlier, we believe that they do justify an increase in the risk factor discount applied to Prima.

We continue to assume a 2017 operating cash burn rate (excluding net interest) of €13.4m and we now estimate a 2018 burn rate of €13.6m (versus €15.1m, previously). Lower projected marketing (since Iris II sales are suspended) and R&D (since the anticipated start of Prima pivotal studies is being pushed into 2019) costs drive the lower 2018 burn rate assumption.

While we believe Pixium has enough gross cash to maintain operations into H118, we expect it will draw the last remaining €3m tranche from the €11m Kreos financing facility in H217 to strengthen its runway and augment financial flexibility. For the same reason, we also continue to assume that Pixium will raise an additional €10m in H217 and €15m in 2018. Our model assumes that, strictly speaking, Pixium would only require €10m in funding to maintain a positive cash balance at YE18, but to maintain operational flexibility and avoid the need to pursue funding methods of last resort, we assume, as is the case for most of our coverage universe, that the firm will seek to always maintain at least six months of financial runway in terms of cash and equivalents on-hand. Hence, our model assumes the firm will raise €25m in total, through YE18, in funding beyond the Kreos financing facility.

For illustrative purposes only, we have added our forecast funding requirements to long-term debt. However, it is possible that Pixium may be required to raise equity instead of debt, to meet its funding needs. Given the recent correction in Pixium’s share price following the announcement of premature Iris II implantation failure, we note that at the current share price, a €10m equity raise would dilute current shareholders by about 20%, and a €25m equity raise would dilute shareholders by about 39%.

Beyond this, we continue to estimate that Pixium will raise €30m in 2019, and €40m in 2020. As above, we have added these forecast funding requirements to long-term debt. We forecast that this funding should enable the completion of registration-enabling Prima clinical studies in both the EU and the US, to meet our estimated launch timelines. This timing for these projected funding requirements can be delayed to a certain extent, should the firm plan to complete US Prima pivotal studies over a longer time horizon than we currently anticipate (ie leading to a potential US launch in or after 2023). Our financial and valuation models do not include the potential dilutive impacts of future equity offerings. We assume that Pixium will only start to become cash flow positive on a sustainable basis once Prima is launched (in 2021).