Replimune Group, Inc. (REPL) Earnings Call Transcript & Summary

Welcome, everyone, to the 39th Annual JPMorgan Healthcare Conference. My name is Anupam Rama. I'm one of the senior biotech analysts here at JPMorgan. I'm joined by Tessa Romero and Matt Bannon from the team. The next presenting company is Replimune, and speaking on behalf of the company, we have CEO, Philip Astley-Sparke. [Operator Instructions] With that, I can pass it over to Philip.

Thank you, Anupam. Safe harbor on Slide 2 and starting on Slide 3 for the Replimune overview. Replimune was founded to realize the full potential of oncolytic immunotherapy. Our products are designed to maximally activate a systemic immune response against a patient's cancer, and we believe that our programs will become the second cornerstone of immuno-oncology. RP1, our lead asset, principally targets immune-responsive tumor types and having generated strong data sets in PD-1 naive cutaneous squamous cell carcinoma and anti-PD-1 failed melanoma. We have 2 registration studies ongoing in those settings. We're also pleased to announce that we hope to this quarter have a further study with RP1 initiate in lung cancer. RP2/RP3, intended to treat less immunoresponsive tumor types, and following the generation of strong RP2 single agent data in heavily pretreated patients with immune insensitive tumor types, the combination phase, the Phase I study with Opdivo, is now enrolling. We also announced -- last week, I was very pleased to announce that RP3 is also now in the clinic. Dosing has commenced in single agent portion of the Phase I study. We have brought manufacturing in-house. Our commercial-scale manufacturing facility is now fully constructed and operational, and GMP production is underway. We have also commenced pre-commercial planning activities. We're well capitalized to deliver with $493 million estimate on the balance sheet as of the 31st of December, funding us into the second half of '24. So what is -- a reminder of oncolytic immunotherapy is the use of viruses that replicate in tumor cells and destroy them, do not replicate in healthy tissue. On injection into a tumor mass, the tumor mass is invariably either completely or partially destroyed. And as the virus rips through the tumor, the neoantigens within are exposed to the immune system in an optimal environment of necrotic cell death. This is a major immune danger signal, attracts antigen-presenting cells to the site, internalize those escaped neoantigens, drains the lymph nodes with their prime T cells to destroy uninjected deposits throughout the body. So it's a dual mechanism of action, direct viral-mediated tumor cell lysis followed by the engineering of a full systemic immune response. There are many different viral species being used as oncolytics. We use the herpes virus. We believe that herpes virus has been optimal species for oncolytic use, it being highly lytic and inflammatory and having high carrying capacity to be able to carry in multiple immune-stimulating proteins into the tumor microenvironment to further amplify the immune response. Within the herpes field, not all constructs are created equal. Our construct is a new clinical isolate that's been deliberately selected after a comprehensive screen of 30-or-so isolates and picked for its lytic ability. And then from all of our products and part of our platform, we express a fusogenic glycoprotein in a gibbon ape leukemia virus, which essentially increases lytic and immunogenic potential of the construct, 10- to 100-fold. And these aspects are designed to maximize antigen presentation or so-called Signal 1. We then further express immune-stimulating proteins from this backbone designed to maximize T cell co-stimulation of the antigen presenting cell/T cell interface, so-called Signal 2. And with RP1, we express GM-CSF; RP2, an anti-CTLA-4 antibody; and RP3, in addition to anti-CTLA-4, the ligand CD40 and 4-1BB, which have pleiotropic effects and also induce inflammatory cytokines. We do believe this is the most practical and comprehensive way to activate a systemic immune response against a patient's cancer. In terms of practical, it's an off-the-shelf product, simple to produce and relatively cheap to produce. It is also a relatively well-tolerated modality. On the efficacy side, it is effectively a pan, universal neoantigen vaccine not confined to 1 or 2 neoantigens. If you're using the right viral species like herpes, you're triggering innate immune system pathways. And through the necrotic cell death created through the virus, you're quickly bringing the adaptive side of the immunity. And then further, through expressing these various immune-stimulating proteins, you can layer in multiple additional mechanisms of action. So a quick recap of our pipeline. As I said right at the beginning, we have 2 potential registration studies ongoing in PD-1 failed melanoma and cutaneous squamous carcinoma having generated strong data sets in those settings. We do also have a study ongoing cohort in MSI high cancers, colorectal cancers predominantly and, as I mentioned, about to start the cohort in lung cancer. Switching to RP2, RP3, we've seen very compelling single agent activity with RP2, and RP3 is now in the clinic. We're giving considerable thought to indication prioritization, where to place our bets in terms of latest stage development in immune insensitive tumor types. Our lead indication at the moment is cutaneous squamous carcinoma. We have several studies ongoing in this setting, not as well-known as melanoma, but the minimum addressable population is equivalent in terms of number of deaths. Up until 2 years ago, it was largely an unmet need, but PD-1s, anti-PD-1s are now approved, Libtayo and Keytruda giving response rates ranging from the mid-30s to 50% but still very low complete response rates on label, single digit. Our study, potential registration study, the CERPASS study, is in 240 patients, where we're comparing RP1 plus Regeneron's Libtayo against Libtayo alone. It's a cross-sharing collaboration with Regeneron, and the primary end point is overall response rate. We've powered the study to show a 15% delta improvement, such -- for example, if the Libtayo arm got 40%, we would win in the mid-50s. But moreover, from what we've seen to date and what we know about the modality, we expect to show a 2 to 3x improvement in the overall complete response rate, which more closely correlates with survival. To date, the data we reported at CSCC has generally waxed and waned slightly, basically delivered an 80% response rate and a 50% complete response rate in CSCC. And those responses have been very durable in nature and obviously a very high rate of actual complete response, very deep responses. And we've seen responses both in systemic disease and in advanced loco-regional disease, which is actually a highlight in this indication. It's actually the primary driver of mortality, it being principally a disease of the head, neck and the scalp and invades vital structures. But this is an example of a full systemic immune response, whereby a patient presented with large lesions in the neck, retroperitoneal node and extensive mets in the bone in the spine. And we injected the one area of the neck with a priming dose of RP1 as mono, and the tumor flattened and started to resolve as did the uninjected one on the other side of the neck. But then led in anti-PD-1 by 16, 24 weeks, the extensive disease, as shown by the CT scan, in the neck had resolved, as had the retroperitoneal node. And the bone mets looked to be static and then were confirmed to be metabolically inactive a year later, and the patient was declared as a complete responder. These are the most recent patients that we have given update on in October over beyond what we had disclosed last June. The top patient there presented with a very nasty ulcerative mass on the foot and tumor burn in the groin area, which was injected. Those lesions resolved. The uninjected lesions on the foot side resolved, and then additional injections were made around the periphery and this patient continues to improve. So a pictogram tells a thousand words. Overall, our responses are very durable in nature, now out to around 600 days. We've only ever had one patient progress, and that patient was in response for quite some time and the responses, as I said before, very deep and durable in nature and give us quite confidence we will meet with success in our potential registration study underway. Switching to anti-PD-1 failed melanoma. This is still a real unmet need. Majority of patients, despite the advance of anti-PD-1, will become primary or have acquired resistance to that therapy. And when they do, treatment options are limited. The response rate to second-line of anti-PD-1 for patients that have truly progressed, confirmed through 2 scans, is very low, mid-single-digit at best. So we have dosed 16 patients in cutaneous melanoma that have failed anti-PD-1 in a very late-stage population of 87.5% stage IVM1b/M1c, an advanced visceral disease. As of last October when we last had a data cut, 9 of these patients had shown some evidence of clinical benefit as defined by at least stable disease with evidence of antitumor activity, of which 5 were full responders, 1 CR and 4 PRs, 4 out of these 5 having failed both anti-PD-1 and anti-CTLA-4 given the current response rate of around about 31%. A further patient is an ongoing surgical CR out to 12 months of disease having not returned, and a further patient remains on study with stable disease. And to give some examples in this setting, this first patient at the top with ipi/nivo failed melanoma, extensive visceral disease, fairly progressing, one coming on to treatment. And we injected a liver lesion shown in the red circle, which has almost resolved away at the last data cut. But you can see the uninjected lesions in the liver and the spleen have completely resolved away, and you can see resolution of a lower mass there in the lung and reduction of a very large mass in the top box of that CT scan. The bottom patient had a very large groin lesion, was essentially homebound, wheelchair-bound and also had extensive mets in the groin nodes, the lung and suspicious bone lesions. We again only injected one mass in the groin area, which resolved away, as did uninjected disease in the lung and the other nodal disease, and the bone mets look static. So it's not actually clear if this patient has any disease at this point in time, not an official CR but a -- what we would call a clinical CR. And very interesting biomarker data from this patient. On the right there in the immunohistochemistry, you can see a classic mechanism for primary resistance to anti-PD-1, which is T cell exclusion from the tumor microenvironment. After layering in our regimen, that immune tolerance is broken and the profound clinical benefit is evident. Again, these plots tell a thousand words, very durable responses in melanoma generally, not just in anti-PD-1 failed but the other melanoma types we've tested in and garnered responses, including mucosal and also in PD-1 naive disease. Again, all but one patient has ever gone into response -- remains into response now beyond 450 days. So the responses that we are engineering are uniformly, very durable in nature. And the spider plot there, I think, is of interest that the red lines are PD-1 failed patients and the green lines are PD-1 naive, and they pretty much overlap, which I think really shows the potential of RP1 in the PD-1 failed setting. And again, we have high expectations that we will meet with success in our potential registrational expansion cohort of 125 patients, and we will be discussing that study design with the FDA towards the end of the quarter. So switching to RP2, which is RP1 that in addition expresses an anti-CTLA-4 antibody. This isn't just to limit systemic toxicity of anti-CTLA-4. There is a strong efficacy rationale to marry up, maximizing Signal 1 antigen presentation on the MHC through the lytic action of our virus with ensuring there's no negative feedback loop through the CTLA-4 pathway. So the single agent data with RP2 showed that similar side effects to RP1, really mainly erythema at the injection site and febrile reactions. This modality is generally very well tolerated. Patients can often go back to work following treatment the next day. But we also saw compelling single agent responses in immune insensitive tumor types, including mucoepidermoid carcinoma, uveal melanoma and esophageal cancer. We saw a further interesting patient who won't be an official responder because it did progress, but after progression, nontarget lesion started to resolve. This is a patient with a microsatellite stable colon cancer, another immune insensitive tumor, major unmet need. This patient initially presented with a high disease burn in the lung, spleen, liver and one other site. Again, these -- the data cut of October, these responses are all deep and durable and ongoing. And in terms of the actual patient examples, mucoepidermoid carcinoma is a salivary gland cancer, untreatable. Checkpoint blockade drug does nothing. This patient would have had a very bad prognosis leading to death and at the moment is not going durable complete response. Over 4 months, no disease, superficial disease resolved away. It was confirmed as a CR 8 months through PET showing the tumor no metabolic activity. This is a response in uveal melanoma, ocular, eye cancer, melanocytes in the eye, where the eye is often surgically removed but often typically metastasizes to the liver, which is often actually the only site of metastases. Once metastasized to the liver, 1-year survival is 10% to 15% as generally a fairly immune insensitive tumor, single percentage response rates to single agent checkpoint blockade a little bit better, maybe in the teens in combination. But regardless, this patient had failed ipi/nivo when coming on to a study with extensive liver mets. So we injected the liver lesion in the red circle. And you can see that nearly resolved away, and the uninjected ones in the yellow circles having either completely resolved away or also resolved away. This patient continues to improve at each subsequent scan. This is an esophageal cancer patient treated with single agent RP2, came on to study having failed anti-PD-L1, a kinase inhibitor and 4 rounds of chemo, actually had more than 1 liver lesion. It's not a particularly good scan. But you can see that lesion resolving away -- that was injected and abscopal effect in the abdominal lymph node. So that's a summary of our clinical data. We also want to highlight, as I said at the beginning, that we have completed the construction of our manufacturing facility. We only broke ground at the time of the IPO 2.5 years ago. And I think it's testament to really the experience of the team that this facility is already operational and having completed tech transfer from the contract manufacturing organization. GMP manufacturing is underway, and we have had a positive meeting with the FDA in terms of comparability to this contract material, the suite of tests that we will have to run in order to use this as a launch material. So my final slide is really a summary of news flow over the next 2 years. In 2021, we expect to present in some entirely new data sets over and above what we have presented in the past, which is predominantly focused on Phase II data sets in melanoma and cutaneous squamous cell carcinoma as well as Phase I all-comers data set. So over the next year, we'll have new defined data sets in PD-1 failed lung cancer, very much an unmet need. And we never dosed a lung cancer patient. But the reason why we decided to proceed with this is we have seen that we can inject lung mets from other tumor histologies and see effects both the injected lung lesion and in uninjected lung lesions. So we are intrigued to see how this cohort does over the course of this year. We also announced at the end of last year that we plan to go into the PD-1 failed set in cutaneous squamous cell carcinoma, which is kind of intuitive given the responses we've seen in anti-PD-1 failed melanoma and in anti-PD-1 90 CSCC. So we have high expectations that we will also -- our products will also be a benefit in this setting. We have talked about -- the past about CSCC transplant patients. These are patients that develop the disease off of being immunosuppressed for organ graft, transplants and anti-PD-1 is contraindicated due to the risk of loss of graft. This actually gives us an opportunity to generate, in an ethical way, more single agent data with the -- with RP1. In RP2 plus Opdivo, part of the Phase I study is enrolling and enrolling well. So expect towards the middle of the year to be able to report that data in that cohort. And we're excited to have got RP3 into the clinic and should be in a position to report our single agent data in the second half of the year. And of course, during '21, we'll also have additional updates on all the studies we've talked about during 2019. And then in 2022, which isn't too far away now, we expect to have a primary readout from our 2 potential registration studies in PD-1 naive, CSCC and anti-PD-1 failed melanoma as well as the combination phase of the RP3 Phase I all-comer study and could also have, obviously, other readouts from studies we plan to initiate in the interim. We have a full indication prioritization analysis underway as we map out where we'll take RP2 and potentially RP3 in terms of late-stage development and development with registrational intent. And we are in a strong position in terms of capital, having cash 2 years beyond reaching or hopefully reaching all these milestones with runway into the second half of '24.

Okay. Thanks for that, Philip. And maybe if you want to introduce the broader team on the line. We can come in with the Q&A. [Operator Instructions]

Yes. So thanks, Anupam. So I'll introduce Rob Coffin, our President of R&D and Founder; Jean Franchi, our Chief Financial Officer; and Pamela Esposito, our Chief Business Officer.

Great. So maybe we'll start out with RP1 in the Phase II CERPASS study in CSCC. Maybe you can give us an update on how the enrollment curve is shaping up. And I know the data is expected to be in 2022 but how we should think about the enrollment curve here.

Sure. Rob?

Well, we're not providing granularity on exactly where we are in enrollment. We are guiding that we're on track for the primary output in 2022, which means that we need to complete enrollment of the trial around the end of the year, if a bit after. However, it would be expected, as with all clinical trials of any size, that the enrollment curve would be relatively hockey stick-shaped, including as over time, we bring on additional sites in countries which aren't immediately open at the beginning of the trial. And it is the case we have seen some effects of COVID on enrollment in this trial, which means we have taken mitigating steps of increasing the number of countries and sites to complete the trial, which will exaggerate that hockey stick shape. But we're confident we'll meet the time line as previously indicated.

Got it. And then in CERPASS, I guess what's your assumption for how the control arm performs, Libtayo. I mean there's a wide range of single agent activity there. I think in the mid-30s to low 50s. And is a 15% delta in improvement in ORR be clinically meaningful in the eyes of physicians? And how are you thinking about CR rate differentiations?

So obviously, as you say, there's a range of expectations for anti-PD-1 therapy. There's various different data sets out there. But I do think whatever the control arm shows, within that range or even slightly above that range, bear in mind what we're already seeing in CSCC in combination with Opdivo that we should have good confidence that assuming in combination with cemiplimab is similar that we have the headroom to show the needed benefit from a statistical perspective. Discussion with both investigators and, in fact, the FDA in relation to the design of the trial indicated that the statistical objectives of the trial, which are a minimum delta, as Philip said, of around 15% would be deemed to be clinically meaningful. However, as with all data packages, it's dependent on the totality of the data, in particular durability as well as response rate per se. And obviously, increasing the CR rate would also be thought to be particularly clinically meaningful because CR is most likely to contribute to greatly extended survival. And there again, from the data we have already with nivo, we are expecting to move from the low single to roughly 10% maybe range of CR at the time of the data cut, which would be expected for single agent Libtayo to two or threefold that -- based on our current data with nivo. So we think we have good confidence that we have a well-designed study, which should easily enable us to show clinical benefit.

And we got a portal -- e-mail portal question here on RP1. Maybe could you comment if CERPASS and the PD-1 refractory melanoma studies ongoing are going to be used for registration?

Certainly, that's the intention. The CERPASS trial was designed with registration and intent in the first instance and was discussed with the FDA on that basis who have indicated support to the design, assuming successful and -- based on the totality of the data. The melanoma single agent trial was set up to -- sorry, single arm trial was set up with the intent of registration. And while all our advisers indicate that single arm of a trial of that size with the expected efficacy hurdle or objective should be appropriate and as Philip indicated, we don't yet have FDA formal buy-in to that, but we are having a Type B meeting within pretty short order to get formal buy-in to that design.

Got it. Philip, in your presentation for -- related to RP2, you mentioned multiple times that you guys are taking the process right now for indication selection. Maybe you could walk us through kind of like what the push/pull levers are and how you think about what could emerge as one or few of the indications that get prioritized here.

Sure. I mean obviously, we'll be predominantly data-driven. That's part of the reason why we expanded the RP2 plus Opdivo part of the Phase I study from 12 patients to 30. But at the same time, we are in conjunction with a consultant doing a full analysis, looking everything from market potential to the size of study and have to run what the end points might be, how long it might take to read out, what line might be in a real comprehensive analysis. So that's the sort of framework answer to the question. So I put a little bit of meat on the bones. We don't quite yet know what's going to bubble up to the top, but we do think it is a real interest that we have seen in multiple tumor types, we can inject liver lesions and see the injected and uninjected ones in the liver and beyond respond. And obviously, there are many tumor types that metastasize to the liver, large tumor types, and treating those liver mets is particularly problematic.

Got it. I'm going to pass it over to Tess to ask some questions on RP3.

Yes. Guys, this is Tess. Yes. So I know the RP3 program recently initiated. We got that off the ground. It sounds like we're going to see some data this year from that program. I guess, do you think you guys will follow kind of a similar playbook as you followed for RP1 and RP2? Just remind us of kind of what you're thinking in terms of size and scope of data for that initial data set later this year.

So the RP3 Phase I trial is really very similar, if not identical, to the RP2 Phase I trial, which has a dose-rising phase which is relatively modest in size, which goes through 2 dose-rising level cohorts, which obviously will expand to larger cohorts if any DLTs were observed. And then that -- after determining the RP2, the -- based on that single agent data, injecting RP3 into both superficial and deep tumors through imaging guidance, we will then enroll a 30-patient cohort in combination with anti-PD-1 therapy. So it has standard Phase 1 inclusion criteria of all-comers who failed standard of care. And if there were no DLTs and no cohort expansion, the numbers of patients would be similar to what we showed with RP2 in October but obviously maybe larger if there were any reason to expand the cohort.

Okay.

I think we're about at the time here. So I just wanted to thank you guys to -- for this productive session. And I hope you guys have a good rest of the conference.

Great. Well, thank you for the invitation to the conference.

Thanks, Anupam.

Thanks.

Thank you.