Cellectis S.A. (ALCLS) Earnings Call Transcript & Summary

Hello, everyone, and thanks for joining the call where we'll discuss Cellectis's business update and financial results for the third quarter of 2026. Joining me are our Chief Executive Officer, Dr. André Choulika, and our Chief Medical Officer, Dr. Adrian Kilcoyne. Yesterday evening, Cellectis submitted a six-K and put out a press release covering the financial statements for the nine months that ended September 30, 2026, as well as a business update. You can find both the report and the release on our website, cellectis.com. I'd like to remind you that this presentation contains forward-looking statements as defined under applicable securities laws, including statements regarding our strategy, development programs, financial condition, and anticipated cash runway. These forward-looking statements are based on what we currently expect and assume, and they involve a range of risks and uncertainties that might make actual outcomes differ materially from what those statements express or imply. To review those risks and uncertainties, please see our filings with the U.S. Securities and Exchange Commission as well as our annual financial report prepared in French. Unless the law requires it, we assume no duty to publicly update or revise any forward-looking statements in this presentation. I'll now hand the call over to André.

Thanks, Arthur. Morning, everyone — glad you could be with us today. Two months back, we shared the news that Cellectis is strategically transforming itself into a business centered on in vivo gene editing. Since September, the whole team has devoted itself entirely to this new chapter. Our starting point is severe dyslipidemias, a space where the unmet medical need is still substantial. With severe hypertriglyceridemia, triglycerides that stay elevated can put patients at risk of serious complications, acute pancreatitis among them. And when it comes to severe hypercholesterolemia, long-term exposure to very high LDL cholesterol raises cardiovascular disease risk considerably. For a great many of these patients, treatment may involve taking a pill each day or receiving injections repeatedly over the long term, and even then lipid control is not always sufficient. What we're doing is distinct: one shot intended to produce an effect that could last a long time. HEAL-101 and HEAL-201 are in development with that goal in mind. I'll begin with .HEAL-101, our in vivo base editing candidate that targets the APOC3 gene in severe hypertriglyceridemia. In October, at the European Society of Gene and Cell Therapy, we shared a broader preclinical dataset. Following one dose of .HEAL-101, as much as 70% of the APOC3 gene was edited, with activity confined to the liver, and triglycerides were consistently lowered by 76% in a diseased model. All together, these findings give us greater confidence in .HEAL-101 as we keep moving the program toward the clinic. Adrian will take you through the specifics shortly. Work on getting the investigator-initiated trial started in China is proceeding as planned, and we anticipate preliminary clinical data during the second half of 2027. I'll now move to our second lead program, .HEAL-201, which is an in vivo gene editing candidate aimed at the PCSK9 gene for severe hypercholesterolemia. The design goal is durable silencing of PCSK9 without changing the DNA sequence. The preclinical results we have are encouraging. In vitro, using human hepatic cell lines, our PCSK9 epigenetic modulator produced high on-site epigenetic editing, averaging above 90%, and we detected no off-site changes in gene or protein expression. In vivo, it lowered mean plasma PCSK9 levels by over 90% in a humanized liver mouse model. Expanding on what we've achieved, we're getting the next stage ready, which we intend to be phase I, and we anticipate preliminary clinical data during the first half of 2028; at this point, I'd like to comment briefly on our older cell therapy programs. Halting work on lasme-cel and eti-cel was a difficult choice. These programs involve patients and families, along with the investigators and site teams who have participated in them throughout the years. We feel profound gratitude toward all of them. What takes priority for us at present is straightforward: to move .HEAL-101 and .HEAL-201 forward toward the clinic while keeping the financial discipline needed to reach the important milestones that lie ahead. Beyond that strategic emphasis, the alliances we hold with AstraZeneca, Servier, Allogene, and Iovance continue to be a significant driver of value for Cellectis. Having said that, I'll now pass things to Adrian, our Chief Medical Officer, who will walk you through the .HEAL-101 data in greater detail.

Appreciate that, André, and morning, everybody. I'll start with a quick refresher on the mechanics of .HEAL-101. It's an editor based on TALE technology, and its job is to perform a single C-to-T conversion inside the APOC3 gene. That change creates a stop codon, which shuts down APOC3 protein production, and it does so without cutting DNA or producing a double-strand break. At ESGCT we shared data that built on our September disclosures and addressed four central questions. The first: what level of efficiency does .HEAL-101 achieve at its target? The second: how precise is the editing? The third: does that editing produce a functional effect? And finally: at which location does that activity take place? In a human liver cell line, .HEAL-101 achieved mean APOC3 editing of roughly 90%, with minimal indel formation and no detectable DNA double-strand breaks at the target site. Furthermore, no detectable editing occurred at the leading putative off-target sites, and this editing led to a pronounced decrease in APOC3 protein secretion. In a humanized liver mouse model with normal lipid levels, one intravenous dose yielded mean APOC3 editing of about 65%, with editing reaching as high as 70%. This was accompanied by reduced circulating APOC3 and lower triglycerides, showing the path from gene editing to protein reduction and finally to the lipid phenotype. Critically, ALT levels did not rise meaningfully relative to untreated controls; in a human APOC3 transgenic mouse model of hypertriglyceridemia, which better mirrors the disease context, HEAL-101 lowered APOC3 levels by 67%. That corresponded to a 76% decrease in triglycerides, matching the efficacy we initially reported in September. Equally important, editing was observed in the liver, while no meaningful editing was found in the other tissues we examined. Such liver-restricted activity aligns with the intended profile of our liver-directed LNP delivery strategy and is a key consideration as we move HEAL-101 toward the clinic. Collectively, these preclinical findings demonstrate a uniform profile: editing that is both efficient and precise, a functional impact on triglycerides, and editing activity confined to the target organ. As André noted, our aim for HEAL-101 is to create a therapy given in a single dose that could deliver sustained triglyceride reduction. The launch of the investigator-initiated study in China remains on schedule, and we anticipate preliminary clinical data during the second half of 2027. We look forward to providing further information about the trial as it moves forward. I'll now hand the call to Arthur, our Chief Financial Officer and Chief Business Officer, who will present our third quarter financial results.

Thanks, Adrian. I'll take you through the financial results for the third quarter of 2026 now. As André pointed out, this quarter is one of transition. In line with our strategic transformation, the exit from the Cellectis lasme-cel and eti-cel clinical programs is moving forward as planned. Enrollment is closed, site closeout work is in progress, and patients who received treatment will keep being monitored under the follow-up requirements specified in the protocols. Past that, work to reshape how the company runs is still underway. As part of that, management currently expects to reduce worldwide staffing by up to 70% relative to where it stood before the transition, spread across 2026 and 2027, provided the relevant French Works Council consultation wraps up and the necessary labor law obligations in France and the U.S. are satisfied. Going forward, we expect Cellectis's cash outlays will predominantly fund advancing our pipeline — HEAL-101, HEAL-201, and potential new product candidates among them — as well as one-time restructuring expenses. As of September 30, 2026, cash equivalents, restricted cash, and fixed-term deposits booked as current and non-current financial assets came to $149 million, compared with $211 million on December 31st, 2025. The $62 million swing was driven mostly by $37.9 million in supplier payments, $37.5 million in payroll-related payments covering salaries, bonuses, and social charges, $7.3 million of lease liability payments, a $4.0 million repayment under the PGE loan, and $1.9 million in capital expenditures — partially offset by $21.4 million in cash received from customers and $5.4 million in interest received from our financial and cash equivalent instruments. For the nine months ended September 30th, 2026, consolidated revenues and other income totaled $19.9 million, compared with $67.4 million for the nine months ended September 30th, 2025. The drop of $47.5 million compared with a year earlier came primarily from how much activity took place under the research plans of the AstraZeneca joint research collaboration agreements during that period. For the nine months finishing September 30th, 2026, consolidated R&D expenses reached $108.8 million, versus $69.1 million across the same nine months ending September 30, 2025. That $39.7 million rise stemmed chiefly from a $34.9 million effect tied to the strategic transformation we announced in September 2026 — comprising a one-off impairment loss of $26.6 million, $6.2 million in termination benefits, and $2.1 million of costs connected to closing out the lasme-cel and eti-cel programs. If we set aside the one-time items, R&D spending was up $4.8 million in 2026 compared with 2025. The main drivers were a $1.4 million rise in non-cash stock-based compensation expense and a $2.1 million increase in purchases and external expenses, which was largely the result of higher clinical development costs incurred before the strategic transformation. Consolidated SG&A expenses came to $17.5 million for the nine-month period ended September 30th, 2026, compared with $15.0 million for the nine-month period ended September 30th, 2025. That $2.5 million increase was mainly attributable to a $1.1 million impact of the strategic transformation announced in September 2026, comprising $0.9 million of termination benefits. If we set aside the one-off items, SG&A costs rose by $1.4 million, largely because stock-based compensation expense came in higher. Now on to our financial results: for the nine months that ended September 30th, 2026, consolidated net gain came to $5.2 million, versus a net financial loss of $25.6 million during the nine months that ended September 30th, 2025. That $30.8 million swing comes from financial income rising by $10.8 million alongside a $20.0 million drop in financial expenses. The $10.8 million rise in financial income was mostly driven by a $14.3 million increase in non-cash gains recognized from fair value measurements, chiefly reflecting a gain on the fair value measurements of the tranche A, B, and C warrants issued to the European Investment Bank, EIB, during the period, versus a fair value loss in the same period of 2025. That was partly offset by a $2.7 million decline in interest income earned on cash equivalents, and financial assets, plus a $1.0 million reduction in foreign exchange gains. Financial expenses dropped by $20.0 million, and that decline was primarily the result of a $23.7 million decrease in foreign exchange losses — mostly from the US dollar strengthening against the euro — plus a $5.9 million reduction in losses from fair value measurements, since the tranche A, B, and C warrants issued to the EIB generated a fair value gain during the period, partly offset by a $9.7 million increase in other financial expenses, largely attributable to the expected credit loss allowance connected with a financial warranty. For more detail on our consolidated net loss attributable to shareholders for the nine months ended September 30th, 2026, please see our press release. As André said, after the realignment is completely in place per plan, our cash runway will extend into the second half of 2028. Notably, we expect that to carry us past the first clinical data anticipated from HEAL-101 in the second half of 2027 and from HEAL-201 in the first half of 2028. Our priorities are straightforward: carrying out our strategic transformation with financial discipline and moving HEAL-101 and HEAL-201 forward toward their important clinical milestones. With that, we're happy to take your questions.

Hello, and thanks for taking our questions. We have a couple of things to ask. Regarding .HEAL-101, how much triglyceride lowering would be sufficient to proceed into global development? Would it need to approach the roughly 80% reduction that plozasiran shows, or would a 50%-60% reduction suffice? And for .HEAL-201, what LDL-C reduction did you observe together with the PCSK9 suppression above 90%? It would be helpful if you could address both of those.

Thanks for the question, Amin. This is Adrian, and I'll handle the first part. Our expectation is that the efficacy we observe will be at least as good as existing treatments, and that's what we anticipate. There's one more aspect we want to highlight. The first is how much of a percentage reduction you get, and the second is what share of patients actually reach the target — that second piece matters enormously. For a few reasons, we're focused on getting as many patients as possible below that 500 threshold. To start, the dangers associated with hypertriglyceridemia really come from having levels above 500. That point is essential. In the high-risk population and in people with prior acute pancreatitis, once levels drop below 500, the risks are basically gone; that's another thing worth keeping in mind. We also recognize that in the 200 to 500 range, looking across any outcomes, we simply can't detect a genuine benefit. Alright, so there are two things here. First, we aim to get as many patients as possible below target. That said, looking at the data package we have right now, we see no reason to expect anything less than at least parity with the therapies currently on the market. Also, just to remind everyone, the .HEAL-101 program is being conducted in China. Regulations there have shifted quite significantly this year, and the program is backed by a strong preclinical package. As for the PCSK9 90% reduction, we have not yet disclosed the LDL-C reductions publicly. That information will be available from us in June. We will definitely follow up with you on that.

That helps a lot, thanks.

Thanks a lot for fielding our questions, and well done to the team on the quarter's progress. We've got a couple of quick items from us. First, I'd like to revisit HEAL-101. I guess we want to tap management's thinking a little on why APOC3 was picked rather than ANGPTL3. As I understand it, gene editors are going after both targets. What I'm wondering is what made APOC3 the choice and what appealed to you about it. And then, briefly, regarding 101 and the work of getting the study going, do you already know which sites you'll use for the investigator-led trial in China? Also, could you share whatever detail you can about the plans for dose escalation? Are you expecting to start at a dose that's efficacious, and how many dose levels might you be able to escalate through? Thanks so much.

Great. Well, I can start on that. I believe there are three questions in there. You may need to remind me of a few of them as we work through them, Jack. The first, why APOC3 versus ANGPTL3? Since our indication was hypertriglyceridemia, the first thing we did was actually extensive market research to see what the marketplace requires. Specifically for hypertriglyceridemia, the general feeling, based on that extensive market research across cardiologists, endocrinologists, internists, was that optimal reduction in triglycerides is important. The overall sense was that when it comes to specifically going after triglycerides, APOC3 remained better than ANGPTL3. So it looked like the correct indication and the correct target for hypertriglyceridemia specifically. Regarding the China sites, yes, we have picked where we will conduct this work in China; we don't yet have permission from the Chinese to make that public within China, but as soon as we do, we'll share those sites with you. It's enough to say we're quite confident about recruiting at the pace we need, since we can reach a very large patient population. The trial's PI is likely the leading treater in this area in China and has run every other gene therapy program in the past, so the experience is extensive. They wanted to take on this program specifically because the technology appeals to them—they see it as potentially safer given that it makes no nicks in DNA. The third question concerned dose escalation, which matters a great deal to us. When it came to picking locations in China, the PI was selected because they wanted to begin at a safe dose while also being strongly focused on starting at a dose that actually works. They've built their own modeling tailored to these kinds of treatments, which lets them examine our very robust preclinical data and run thorough modeling so we can begin at a somewhat higher initial dose. Clearly, we still anticipate escalating the dose—this is a phase I study. Even so, we do expect the starting dose to be effective. It might not deliver optimal efficacy, yet it does set up what the expectations are for the data readout in 2027. After we begin dose escalation in Q1, we anticipate that the first cohort will show a certain level of efficacy, and we would expect that by the end of the year you will begin to see how durable that effect is, which is a proposition unique to Cellectis. In fact, if patient recruitment starts in Q1, by year end we should be beginning to observe durability of effect of the nine-month type.

Hi Jack, André here, and I want to add one thing: when it comes to reducing triglycerides, APOC3 is a much more effective metabolic intervention than ANGPTL3, which does work, though not to the same degree. What we're aiming for isn't a product that brings down triglycerides and LDL; it's essentially about trying to effectively treat the potential risk of pancreatitis. In our view, this target is the best adapted one, and there's a clear POC with ASOs and RNAs on it, which is very valid. On this side, there's a real unmet medical need for a very drastic lowering, given the metabolic action.

Great. Thanks so much for all the color.

Oh, great. Thanks for taking our questions. Regarding the APOC3 program, is it possible for you to compare how much APOC3 and triglyceride reduction you get from your base editing approach versus the nuclease editing approach? I'm just wondering if there might be any potential differences. Could you also speak to the LNP you're using and how confident you are about liver safety? Thank you.

Well, thank you for the question, Yanan. This is André. The effectiveness of either making base editing versus cutting DNA with a double strand break, is like the deepness of action, is very similar with a slight upside for base editing. The key distinction is basically this: replication has to happen within the cell, and since liver cells divide quite actively, you end up with highly effective base editing. We chose in vivo gene editing using base editors and epigenetic modifiers because cutting DNA carries the danger of so-called non-homologous end joining, which can introduce a mutation into the gene — and there's a meaningful chance, perhaps very, very small but still present, of translocation, chromatolysis, or loss of heterozygosity. What that means is you might lose an arm of a chromosome, or the chromosome could recombine at another location, and so on. With a deaminase alone, one nucleotide gets swapped for a different one, and replication follows. So efficiency ends up comparable — perhaps marginally higher — using the base editor, yet the danger is far smaller, and that's really the chief justification for employing a nuclease or double-strand break, for example a TALEN (the N means nuclease): basically to perform gene insertions, gene replacements, or deletion of a sizable chunk. If, however, your goal is simply to knock out a gene, you can achieve that via methylation, via deacetylation, or basically through base editing that introduces a mutation either within the coding sequence or at splice sites.

Thanks.

I'm not sure that responds to what you asked.

Yeah, that helps. Thanks, André. What about the safety of LNP for the liver?

Yes, it's certainly a safety approach.

So yeah, like I mentioned, we do have NHP data as well as mouse data. It's enough to say that what we've seen on liver toxicity has been quite reassuring.

Great, appreciate the color.