Conferencia de resultados del 3T 2026 de Sarepta Therapeutics (SRPT): Eleva sus previsiones a medida que Elevidys se estabiliza

Puntos clave

• Los ingresos totales por productos en el tercer trimestre de 2026 fueron de 324 millones de dólares, un descenso del 13% interanual, incluidos 83 millones de dólares de Elevidys y 240 millones de dólares de la franquicia PMO.
• Los ingresos totales fueron de 381 millones de dólares, una caída del 5% interanual, ya que 58 millones de dólares en ingresos por colaboraciones y otros conceptos compensaron parcialmente las menores ventas de Elevidys.
• Sarepta elevó el límite inferior de sus previsiones de ingresos por productos para 2026 en 50 millones de dólares, reduciendo el rango a 1.250–1.300 millones de dólares. La dirección citó la solidez de PMO y la estabilización de la demanda de Elevidys.
• La compañía registró una pérdida operativa GAAP de 9 millones de dólares y un beneficio operativo no GAAP de 23 millones de dólares en el trimestre. El efectivo y las inversiones cerraron el tercer trimestre en casi 1.000 millones de dólares.
• La actividad de formularios de inscripción para Elevidys se mantuvo en línea con la del segundo trimestre. La dirección prevé que los ingresos de Elevidys en el cuarto trimestre se mantengan en términos generales en línea con los del tercer trimestre, si bien reconoce la variabilidad asociada a una terapia génica de administración única.
• Los datos de dosis múltiples ascendentes para SRP-1003 DM1 y SRP-1001 FSHD en dichos programas se esperan para más adelante en 2026. Sarepta se está preparando para el desarrollo orientado al registro, manteniendo al mismo tiempo flexibilidad entre las vías de aprobación acelerada y tradicional.

Datos financieros principales

En los primeros nueve meses de 2026, los ingresos por productos fueron de 983 millones de dólares, un descenso del 34% interanual. Los ingresos por colaboraciones y otros conceptos aumentaron a 530 millones de dólares frente a los 251 millones de dólares, lo que situó los ingresos totales acumulados en el año en aproximadamente 1.500 millones de dólares, una caída del 14%.

El resultado operativo GAAP y no GAAP acumulado en el año totalizó 363 millones de dólares y 507 millones de dólares, respectivamente. Sarepta afirmó que su negocio base generó casi 300 millones de dólares en efectivo tras ajustar 250 millones de dólares en pagos por colaboración a Arrowhead.

Desempeño comercial y operativo

Elevidys se estabiliza en un nivel base más bajo

Elevidys generó 83 millones de dólares en el 3T. Los formularios de inscripción estuvieron en línea con los del 2T y por encima de los niveles observados a principios de 2026. La dirección considera esto como una nueva base de referencia para evaluar la ampliación de su equipo sobre el terreno y las iniciativas de formación para médicos y pacientes.

La compañía está haciendo hincapié en los resultados a tres años de EMBARK y los datos a dos años en pacientes tratados entre los 8 y 12 años. Más de 1.500 pacientes han recibido Elevidys. La dirección señaló que la evidencia a más largo plazo está ayudando a médicos y familias a evaluar el perfil de beneficio-riesgo del producto, aunque los cambios en los hábitos de prescripción requieren tiempo y una interacción continuada.

Sarepta prevé que los ingresos de Elevidys en el 4T evolucionen en línea con los del 3T. La dirección señaló que el período de aproximadamente seis meses entre la inscripción y la infusión implica que los ingresos a corto plazo reflejan en gran medida la demanda generada antes de sus iniciativas comerciales más recientes y de las actualizaciones de datos.

La franquicia PMO sigue siendo la base comercial

La franquicia PMO generó 240 millones de dólares en el 3T. Las nuevas inscripciones de pacientes aumentaron, mientras que la adherencia respaldó la estabilidad de la base comercial de pacientes.

Más de 1.800 pacientes en todo el mundo han recibido las terapias de omisión de exón de Sarepta. La dirección destacó la evidencia en el mundo real a largo plazo de la franquicia, la infraestructura consolidada de reembolso y las relaciones con la comunidad de Duchenne.

Sarepta asume que la terapia competidora para el exón 51 de Dyne recibirá la aprobación en enero de 2027. La dirección señaló que es poco probable que se produzca un cambio generalizado de tratamiento en pacientes estables al principio y prevé que cualquier impacto competitivo sea más visible hacia finales de 2027, sujeto a la ficha técnica definitiva del competidor y a su despliegue comercial.

Los resultados en DM1 y FSHD son los próximos grandes catalizadores de la cartera de desarrollo

Sarepta prevé datos de dosis múltiples ascendentes para SRP-1003 en DM1 y SRP-1001 en FSHD en las próximas semanas. Los estudios tienen como objetivo establecer las relaciones dosis-exposición, el silenciamiento de la diana, los efectos biológicos posteriores y la tolerabilidad.

En el caso del SRP-1003, la compañía está evaluando el silenciamiento de DMPK y la corrección del splicing. La dirección afirmó que una corrección del splicing en el rango del 10%–15% no sería suficiente y tiene como objetivo un efecto significativamente mayor.

Para el SRP-1001, Sarepta busca concordancia entre la exposición muscular, la supresión de la vía DUX4 y los biomarcadores circulantes. La dirección advirtió de que no se debe dar un peso excesivo a los resultados funcionales tempranos debido a la población de estudio reducida y heterogénea y al seguimiento limitado.

Los preparativos de fabricación para posibles estudios de fase 3 están en marcha. Las cohortes de dosis más altas también podrían servir de base para la selección de la dosis definitiva a principios de 2027.

Estrategia clínica en Duchenne

Sarepta decidió no reanudar la inscripción en el estudio ENVISION tras una suspensión de más de un año. La dirección atribuyó la decisión a barreras operativas y técnicas y afirmó que ENVISION no era fundamental para restablecer el acceso a Elevidys para pacientes no ambulantes.

En su lugar, la compañía considera la Cohorte 8 de ENDEAVOR, que evalúa el sirolimus profiláctico como una posible estrategia de mitigación de la lesión hepática aguda, como fundamental para ese objetivo. Se prevé que la inscripción finalice para finales de 2026, con interacciones previstas con la FDA para el 1T de 2027 tras disponer de los datos de seguridad a 12 semanas.

Sarepta también prevé presentar una solicitud suplementaria de licencia para productos biológicos para Elevidys en niños menores de 4 años hacia mediados de 2027. El estudio MIS51ON de fase 3 de rango de dosis con dosis más altas de EXONDYS 51 ha alcanzado el hito de última visita del último paciente.

Previsiones de la dirección

La dirección tiene previsto preservar la flexibilidad financiera de cara al vencimiento de los bonos de la compañía en 2027, al tiempo que financia programas de mayor prioridad, entre ellos DM1 y FSHD.

Riesgos y aspectos a vigilar

• Las ventas de Elevidys se mantienen por debajo de los niveles del año anterior, y los ingresos trimestrales pueden fluctuar al tratarse de una terapia génica de administración única.
• El plazo de aproximadamente seis meses entre la inscripción y la infusión retrasa el impacto en los ingresos de las iniciativas comerciales actuales.
• El margen bruto del 3T se redujo en 52 millones de dólares en cargos por compromisos «take-or-pay» vinculados a obligaciones de fabricación no cancelables de Elevidys.
• Restablecer el acceso a Elevidys para los pacientes no ambulantes depende de los datos de seguridad de la cohorte 8 de ENDEAVOR y de las posteriores conversaciones con la FDA.
• Los planes de registro para DM1 y FSHD dependen de los próximos datos de knockdown, biomarcadores, seguridad y selección de dosis, así como de las valoraciones de los reguladores.
• El diseño del ensayo para DM1 sigue sujeto a la heterogeneidad de los pacientes, la sensibilidad de los criterios de valoración y la duración del seguimiento. El equipo directivo está revisando datos externos de historia natural y resultados de competidores antes de definir su enfoque definitivo.

Puntos destacados de la sesión de preguntas y respuestas con analistas

• Aprobación acelerada: La dirección considera que la aprobación acelerada podría seguir estando disponible tanto para DM1 como para FSHD si los datos son lo suficientemente sólidos, pero la estrategia final dependerá de las conversaciones con la FDA.
• Selección de dosis: Sarepta tiene la intención de seleccionar la dosis máxima tolerable que ofrezca el mayor grado de knockdown. Se esperan datos de las cohortes de dosis más altas para principios de 2027.
• Diseño del ensayo pivotal para DM1: La dirección señaló que la selección de la población y la duración de la evaluación pueden ser tan importantes como el propio criterio de valoración. Sarepta está evaluando datos de historia natural, hallazgos internos y aportaciones de líderes de opinión clave.
• Vía de Elevidys para pacientes no ambulantes: La compañía prevé dirigirse a la FDA hacia finales del primer trimestre de 2027 con los datos de seguridad a 12 semanas de la cohorte 8 de ENDEAVOR.
• Enfermedad de Huntington: Sarepta prevé obtener datos de prueba biológica de concepto para SRP-1005 hacia mediados de 2027. La dirección afirmó que los resultados preclínicos respaldan la administración sistémica a través de la barrera hematoencefálica y hacia los núcleos cerebrales profundos.
• Plazos de crecimiento de Elevidys: La dirección considera que la franquicia conserva su potencial de crecimiento a largo plazo, pero declinó especificar cuándo podría volver a acelerarse el crecimiento, citando los tiempos del tratamiento y la variabilidad intertrimestral.

Transcripción completa de la llamada de resultados

Transcripción completa de la conferencia de resultados

Comentarios de la dirección

Tamara Thornton

Please go ahead. Thank you, Michelle, and thank you all for joining today's call. Earlier this afternoon, we released our financial results for the third quarter of 2026. The press release, along with our presentation slides and supplementary information, are available on the investor section of our company website. We plan to file our Form 10-Q for the quarter today with the SEC. Joining me on the call today are Mike Severino, our Chief Executive Officer; Dr. Louise Rodino-Klapac, President of Research and Development and Technical Operations; Patrick Moss, our Chief Commercial Officer; and Ryan Wong, our Chief Financial Officer. Additionally, joining us in the Q&A portion of the call are Ian Estepan, President and Chief Operating Officer, and Dr. James Richardson, our Chief Medical Officer.

Before we begin the formal remarks, I would like to note that during this call we will be making a number of forward-looking statements. Please refer to Slide 2 of our presentation to view the formal text of these Safe Harbor statements. These statements involve varying risks and uncertainties, many of which are beyond Sarepta's control. Actual results can materially differ from these forward-looking statements, and such risks can adversely affect our business, our results of operations, and the trading price for Sarepta's common stock. We strongly encourage all listeners to review the company's most recent SEC filings for a detailed description of these applicable risks. Sarepta explicitly states that it does not undertake any obligation to publicly update or revise forward-looking statements or financial projections based on subsequent events. Furthermore, please note that we will discuss non-GAAP financial measures during today's webcast. Complete descriptions and reconciliations of our GAAP to non-GAAP financial measures are included in today's press release and the accompanying slide presentation available to investors on our website. With that, I will now turn the call over to our CEO, Mike Severino.

Michael Severino

Thanks, Tam. Good afternoon, everyone, and thank you for joining us. It has been a little over 90 days since I joined Sarepta, and I want to begin by sharing a few reflections from that time and how they are shaping my priorities for the company. What initially attracted me to Sarepta was a combination that I believe is unique in biotech at our scale: an established commercial business and leadership position in Duchenne, a promising pipeline with the potential to produce differentiated medicines, and the financial strength to invest in that pipeline from our own commercial cash flows. After my first 90 days, I believe those fundamental strengths are very much intact.

At the same time, I joined the company following a challenging period, so I have been deliberate in focusing on evaluating the business objectively and identifying clear competitive advantages, opportunities where we can improve execution, and areas where investment has the potential to generate the greatest long-term value. My observations are informed by interactions with our stakeholders: physicians, patients, investors, and employees. And as Sarepta employees have welcomed me into the organization, I have been impressed with their talent, resilience, deep dedication to serving patients, and shared commitment to advancing the opportunities in front of us.

Turning first to our commercial business, our immediate priority has been to reinforce the strength of the broader Duchenne franchise while stabilizing Elevidys demand. During the quarter, enrollment form activity for Elevidys was in line with the second quarter. We believe that this establishes a new baseline that we can use to assess the effectiveness of our commercial initiatives going forward. We are also bringing a growing body of clinical evidence, including new long-term data and data in the important 8- to 12-year age range, into our conversations with physicians and families, enabling us to put the benefit-risk profile of Elevidys into the appropriate context. Importantly, this quarter also demonstrated the continued strength of our PMO franchise. Based on our strong PMO performance and stabilization in Elevidys, we are raising the low end of our 2026 total net product revenue guidance and narrowing the range to $1.25 billion to $1.3 billion. Patrick will discuss our commercial performance and outlook in greater detail shortly.

Second, I'll turn my attention to the promise of our pipeline. One of my strongest conclusions from my first 90 days is that the potential of our pipeline, particularly in DM1 and FSHD, is not yet fully reflected in how Sarepta is viewed today. These are genetically defined diseases where the underlying biology is well understood. We have shown robust knockdown of disease-causing mutations in a number of preclinical systems, and promising data from our SAD studies support their potential to be differentiated therapies. The next step is to demonstrate that these observations extend to a multiple dose setting and demonstrate consistent, robust knockdown, leading to the downstream biologic effects we believe will predict long-term clinical benefit. Our multiple ascending dose datasets, which we expect to share in the upcoming weeks, are designed to provide the data necessary to move both programs into late-stage development. In anticipation of these data, we are already preparing for the initiation of registrational studies for both programs.

While we are preparing to move forward with a sense of urgency, we will preserve the ability to learn from the emerging competitive landscape and our discussions with regulators to inform the optimal development and regulatory path. Louise will provide more detail on these programs, as well as our progress across the Duchenne clinical portfolio in her section.

Third, I would like to highlight our financial strength, which I view as an important strategic pillar for Sarepta. Revenues from our commercial portfolio provide us with the ability to fund the high-value opportunities in our pipeline while retaining flexibility around how and when we deploy capital. We intend to remain rigorous about where we invest, prioritizing programs where we believe Sarepta has a scientific or commercial advantage, and maintaining discipline across the broader cost structure. Ryan will provide more detail on the quarter, but our continued non-GAAP profitability, cash generation, and balance sheet strength give us the flexibility to invest through the important catalysts ahead while maintaining a disciplined approach to capital allocation.

So, as I look ahead, my view after 90 days is one of focused optimism. We have work to do and I want us to be measured by execution. But we also have assets that I believe position Sarepta well for its next chapter: a durable Duchenne franchise, potentially differentiated opportunities in DM1 and FSHD, capabilities in rare and neuromuscular disease, and the financial strength to advance those opportunities. Over the coming quarters, I believe the markers of our progress will become increasingly tangible: continued execution in Elevidys, potentially differentiated clinical data from our siRNA programs, disciplined advancement of our highest-valued pipeline opportunities, and consistent financial execution. If we deliver against these priorities, I believe we can strengthen the foundation of the business while positioning Sarepta for meaningful long-term value creation. With that, let me turn the call over to Patrick to discuss our commercial performance. Patrick?

Michael Severino

Thank you, Mike. Today, I will review our third quarter commercial performance, the work we are doing to support treatment decisions for Elevidys, and the continued strength of our broader Duchenne business. Our third quarter resulted in total net product revenue of approximately $324 million, including $83 million of net revenue from Elevidys, and $240 million from our PMO franchise. For Elevidys, we are seeing stabilization in the business, and our exon skipping franchise continues to demonstrate strong and durable performance. The commercial organization is executing well. We continue to hear positive feedback that our field team's efforts are being well received, with physicians and centers appreciating the support, education, and coordination. Key leading indicators for Elevidys and enrollment forms in the third quarter were in line with Q2 and above where we started the year. Willingness to treat remains solid, and recent start form activity supports our current run rate.

We continue to remain focused on execution, identifying patients, and helping physicians and families understand the long-term evidence for Elevidys, including the recently published three-year EMBARK results. Our sales team brought these data to healthcare providers across our sites of care and referral networks this quarter. What we hear in our engagement with HCPs is that the three-year EMBARK data resonate. With more than 1,500 patients now treated, Elevidys has a depth of real-world evidence that is unique in Duchenne. We believe these data are helping physicians and families better appreciate the long-term benefits of Elevidys. For physicians, the data reinforce what they observe in their own patients treated with Elevidys, and some who have had a period of inactivity are returning to prescribe. We believe this evidence is having an impact as HCPs bring it into conversations with Duchenne families. Some families are resuming their treatment journeys after pausing, and eligible patients who have recently made the decision to pursue treatment are doing so with more determination to complete their journeys.

Throughout the quarter, we continued our work to help educate families about Duchenne and the therapies available to them. Further, as new data emerge, we continue to update our patient education resources so families have current, balanced information to support their conversations with HCPs about treatment options. Louise will speak in more detail to the 8- to 12-year-old data recently presented at the World Muscle Society, which show durable two-year functional benefit and manageable safety in patients ages 8 to 12 at the time of treatment. Together with EMBARK, these data reinforce the value of Elevidys across the ambulatory population and help address information needs for families of older ambulatory boys. We continue to work on further building our evidence base with longer-term functional data and ongoing evidence generation. This includes Cohort 8 of ENDEAVOR, which evaluates prophylactic sirolimus treatment as part of an enhanced safety protocol during treatment with Elevidys in non-ambulatory patients with Duchenne. Looking ahead, we believe the trajectory from here will be influenced by the continued evolution of the data.

Turning to our exon skipping franchise, we delivered another strong quarter. New patient enrollments increased and patient adherence to our exon skipping therapies continues to support ongoing stability in our commercial patient base. In September, EXONDYS 51 marked 10 years since its U.S. approval. I've been with Sarepta since that launch. It has been a privilege to be part of this journey from the beginning, building a treatment category where none existed, earning the trust of the Duchenne community, and seeing firsthand what these therapies mean to patients. Over that decade, exon skipping has become a foundational treatment approach in Duchenne. More than 1,800 patients worldwide now have been treated with our exon skipping therapies, with a substantial body of real-world evidence across ambulation, pulmonary, cardiac function, and survival.

That real-world experience is a key reason physicians and families continue to place their confidence in our exon skipping therapies. The Duchenne landscape is evolving, which requires us to continue to leverage the advantages we have developed over the past decade. This includes deep relationships with the Duchenne community, established reimbursement and patient support pathways, and real-world evidence in a disease that progresses over decades, demonstrating the benefit-risk profile of our exon skipping treatments. We believe that the combination of data, expertise, and relationships differentiates Sarepta and positions us to remain a trusted partner as treatment options expand. As pioneers in Duchenne, we will continue to use our expanded and experienced field organization to reinforce the value of exon skipping with physicians, patients, and families.

Turning to our outlook for the fourth quarter, Elevidys revenue in the third quarter was consistent with the guidance we communicated on our last earnings call. We continue to expect Elevidys revenue in the fourth quarter to trend in line with Q3, acknowledging the quarter-to-quarter variability inherent in a 1-time gene therapy. We expect our exon skipping franchise to remain strong in the fourth quarter, continuing the solid performance we have seen across the portfolio. The strength of our PMO business, together with stabilization in Elevidys, supports our decision to raise the low end of our 2026 total net product revenue guidance by $50 million to a range of $1.25 billion to $1.3 billion.

In summary, our performance reflects two important dynamics. First, stabilization of our Elevidys business, supported by a growing body of long-term evidence that is resonating with the community and offers a greater degree of confidence for both physicians and families. Second, our exon skipping franchise continues to provide a durable foundation for our Duchenne business, built on a decade of real-world experience that shows the value of dystrophin restoration versus the natural history of Duchenne. Our priorities remain clear: drive broader adoption of Elevidys among ambulatory patients, build on the strength of our exon skipping portfolio, and support informed treatment decisions through evidence and education. These focused actions, coupled with our deep commitment to patients, will sustain our leadership in Duchenne. Thank you. And with that, I'll turn the call over to Louise.

Louise Rodino-Klapac

Thanks, Patrick, and good afternoon, everyone. We believe the progress the Sarepta team has made these past few months represents significant advancements, positioning us to bring the very best therapies to patients. Beginning with siRNA, it's important to understand why we believe our siRNA approach will deliver the best results for patients. The premise is simple. Rather than applying 1 delivery approach across all tissues, we select the receptor and delivery architecture that is intended to best address the key biological barrier in each disease. Fully understanding the biology of the disease and how to knock down or eliminate the toxic RNA or protein caused by the disease is critical to impacting its root cause. We believe our approach will offer long-term clinical impact and potentially the highest and most durable knockdown, leading to the greatest functional benefits.

More specifically, our key intended differentiators for the DM1 and FSHD programs include reaching enough muscle fibers, achieving sufficient intracellular exposure, and driving meaningful target knockdown. To date, our preclinical and clinical data support the premise and have shown dose-dependent increases in plasma and muscle concentration, no evidence of saturation in muscle uptake, and the ability to potentially achieve deeper target knockdown, including preclinical data demonstrating robust nuclear knockdown through greater tissue exposure. We look forward to sharing with you the results of our MAD study for DM1 and FSHD in the coming weeks. The plan to share these data remains consistent. I'll now provide more color on what success would look like for these readouts.

Starting with DM1, SRP-1003 is designed to knock down the toxic DMPK RNA that drives DM1 pathology. In DM1, mutant DMPK transcripts accumulate in the nucleus, sequester MBNL proteins, and disrupt normal RNA splicing. For a therapy to address the underlying biology and treat the disease at its source, it must reach muscle, enter the target cell, knock down nuclear-retained DMPK RNA, and drive downstream splicing correction. Of note, the connection between knockdown and splicing correction and the biological connection to clinical impact and functional benefit is well established in the scientific literature. As presented at the 2026 Oligonucleotide Therapeutics Society in October, we established correction of MBNL-dependent mis-splicing, which means that SRP-1003 is degrading disease-causing RNA within the nucleus. That shows that the therapy degraded pathogenic DMPK mRNA in both the cytoplasm and nucleus in a dose-dependent manner, achieving robust knockdown in each compartment and a nearly 1-to-1 comparable knockdown across cellular compartments.

Clinically, our early SAD data show a dose-dependent increase in plasma concentration and high muscle concentration in the initial cohort. Importantly, our upcoming MAD data, Cohort 3 at 3 mg/kg siRNA and Cohort 4 at 4 mg/kg siRNA, are intended to extend this evidence by evaluating the relationship between dose, muscle exposure, DMPK knockdown, and splicing correction, along with safety and exploratory functional measures. More specifically, we believe SRP-1003's potential best-in-class profile would include the following: dose-responsive muscle exposure and DMPK knockdown; clear downstream splicing correction measured through effects on genes relevant to the pathology of DM1, including the CASI-22 panel, supporting the knockdown in the nucleus; and finally, a safety and tolerability profile that preserves the ability to dose to an optimal level for differentiated pharmacology.

Now moving to FSHD, SRP-1001 is designed to knock down DUX4, the toxic driver of FSHD in skeletal muscle. The objective of our MAD study is to connect muscle exposure with DUX4 pathway suppression and downstream molecular effects while continuing to characterize safety at higher exposure. To that end, we believe a successful readout for our MAD study would include continued evidence of strong muscle delivery and the ability to dose without a dose-limiting safety signal; knockdown of the DUX4 pathway reflected in suppression of the DUX4-related gene panel, together with directional improvement in circulating DUX4-related biomarkers; concordance across pharmacodynamic measures, which can help establish a treatment-related signal from variability in an individual assay; and CK, interpreted directionally and in the context of the broader biomarker package.

While the data will include early functional measures, we would not over-index on functional outcomes in this readout since the study is small. The population was not selected for functional homogeneity at baseline and the available follow-up is not long enough to definitively demonstrate a treatment effect on function. FSHD is a progressive disease and DUX4 knockdown is intended to slow or prevent ongoing muscle damage. So, demonstrating function will therefore require an appropriately sized and controlled study with sufficient follow-up. At this stage, we believe the most credible foundation for longer-term benefit is strong, reproducible DUX4 pathway knockdown supported by concordant biomarkers and a tolerable dose. Preliminary functional assessments can help inform endpoint selection and future study design, but the near-term bar is a coherent biological package showing that SRP-1001 reaches muscle, drives DUX4 pathway knockdown, and produces consistent downstream biomarker effects.

In summary, we believe our FSHD and DM1 programs hold the potential to be best in class by not simply reaching the bloodstream, but by reaching enough muscle fibers, maintaining exposure long enough, achieving sufficient intracellular siRNA concentrations, and driving meaningful target knockdown in the nucleus. Given this potential, we are actively preparing for our registrational studies for both programs. Our development strategy is designed to advance each program efficiently towards registrational development while preserving flexibility as the clinical and regulatory landscape evolves. The upcoming datasets are intended to establish the PK-PD relationship, inform dose selection, and provide the evidence needed to finalize pivotal study design, including patient selection and endpoints. In parallel, Sarepta is advancing operational readiness for late-stage development, including initiating manufacturing for the planned Phase 3 studies with the intent that we are positioned to move quickly following the data and regulatory engagement.

From a regulatory perspective, accelerated approval remains an important potential pathway for both programs, alongside traditional approval, and we are planning in a way designed to preserve optionality for either approach. This could include the possibility of a Phase 2 expansion study to support an accelerated approval pathway, followed by a confirmatory Phase 3 study. Ultimately, the strength of the data and our discussions with the FDA will inform the appropriate regulatory strategy. Our goal is to build a compelling evidence package and pursue the most efficient and appropriate path to bring these potential therapies to patients globally.

Now turning to Elevidys, our gene therapy approved to treat Duchenne. At this year's World Muscle Society meeting in Japan, we were pleased to present important clinical updates, including data that showed 8- to 12-year-old patients treated with Elevidys demonstrated durable and clinically meaningful benefits across the North Star Ambulatory Assessment and timed function measures over two years, compared to a well-matched external control group. As you are aware, the years between 8 and 12 are when these boys begin to rapidly decline. We also presented an analysis of patients treated ages 2 and 3, a total of 26 patients, that demonstrated robust Elevidys dystrophin expression and sarcolemma localization 12 weeks after infusion. The extensive evidence we have generated in this under-4 population is critical to our regulatory strategy to broaden potential Elevidys treatment opportunities for as many Duchenne patients as possible. We're in the process of discussing our approach with the FDA and expect to submit an sBLA for this under-4 population mid-next year.

We were also pleased to have a robust presence at the annual Neuromuscular Study Group Scientific Meeting in San Antonio, Texas. We presented seven posters, including the first results from the Phase 4 ENDURE study, a multicenter prospective observational study in patients receiving Elevidys as part of clinical care for Duchenne. The study included safety in patients who received prophylactic immunosuppression with sirolimus, which provided preliminary real-world data evaluating approaches to liver safety risk mitigation. The data showed that treatment-related acute liver injury rates were consistent with prior reports, and no acute liver injury events were observed among the subset of participants who received prophylactic sirolimus. The totality of these data support what we at Sarepta and others in the community already believe about the positive impact Elevidys continues to have on the trajectory of Duchenne.

Moving now to our ENVISION study. After careful consideration of our clinical development strategy, we've decided not to resume enrollment of the ENVISION study. ENVISION, as you are aware, has been on hold for over a year and we have recruited approximately half of its intended sample size. The feasibility of resuming enrollment in a study after such a prolonged hold introduces operational and technical barriers that we do not believe are practical to overcome at this stage. Notwithstanding our decision to not resume enrollment in ENVISION, our commitment to serve non-ambulatory individuals with Duchenne remains steadfast. This population continues to have fewer treatment options.

ENDEAVOR Cohort 8 remains central to this commitment. To remind you, the purpose of Cohort 8 is to assess prophylactic sirolimus treatment as part of an enhanced safety protocol during treatment with Elevidys in non-ambulatory individuals with Duchenne. Data from Cohort 8 will help determine whether administering sirolimus prior to and after an Elevidys infusion can help reduce acute liver injury, a known risk associated with AAV gene therapy. We expect to complete enrollment of ENDEAVOR Cohort 8 by the end of this year and are planning to engage with the FDA in Q1 2027.

Moving now to EXONDYS 51. MIS51ON is our post-marketing commitment Phase 3 dose-ranging study evaluating higher doses of EXONDYS 51 for Duchenne patients who have an exon 51 skipping mutation. We are pleased to report that we have achieved last patient, last visit in the MIS51ON study. We look forward to communicating our results and updates when available.

We look forward to sharing important updates with you, including data from our DM1 and FSHD MAD studies later this year, and our proof-of-biology data from the Huntington's disease program and data from the ENDEAVOR Cohort 8 study next year. Thank you. And I'll turn the call over to Ryan for an update on our financial performance. Ryan?

Ryan Wong

Thank you, Louise, and good afternoon, everyone. In the third quarter, we executed against our financial strategy, delivering revenue while maintaining expense discipline as we invested in our pipeline. We delivered non-GAAP profit, and the base business continued to generate cash, further strengthening our financial position. Today, I'll walk through highlights from our third quarter results and how we are positioned heading into year end.

Starting with our top-line results, total product revenues in the third quarter were $324 million, a decrease of 13% year-over-year. For the first nine months of 2026, total product revenues were $983 million, a decrease of 34% over the same period prior year. The year-over-year change in both periods was driven by lower Elevidys sales. Collaboration and other revenues were $58 million for the quarter and consisted primarily of contract manufacturing revenues for commercial Elevidys supplies sold to Roche. For the first nine months of 2026, collaboration and other revenues were $530 million compared to $251 million in the same period prior year. All together, total revenues were $381 million in the third quarter and approximately $1.5 billion year-to-date, decreases of 5% and 14% year-over-year, respectively.

Turning to cost of sales and gross margin. The total cost of sales for the third quarter was $189 million, an increase of 25% over the prior year, driven by cost of goods related to contract manufacturing sales to Roche. For the first nine months, total cost of sales was $447 million, which is flat to the same period prior year. Third quarter results included $52 million in take-or-pay charges related to non-cancelable manufacturing obligations where we elected not to manufacture certain inventories for Elevidys. Product revenue gross margins, which are net product revenues less the inventory costs related to products sold excluding to Roche, were 59% in the quarter and 72% through the first nine months. The take-or-pay charge just mentioned impacted margins by approximately 19% in Q3 and 6% year-to-date.

Moving now to expenses. Combined R&D and SG&A expenses in the third quarter were $201 million on a GAAP basis and $169 million on a non-GAAP basis. GAAP expenses decreased 40% year-over-year, driven by the $100 million DM1 milestone payment to Arrowhead in prior-year R&D expense and by lower clinical and manufacturing expenses. For the first nine months of 2026, combined R&D and SG&A expenses were $663 million on a GAAP basis and $558 million on a non-GAAP basis. On a non-GAAP basis, the year-over-year decrease was 61%, driven by the Arrowhead collaboration transaction costs in the prior-year period, as well as the cost reductions from last year's restructuring and R&D pipeline prioritization.

Bringing these results together, we reported a GAAP operating loss of $9 million and a non-GAAP operating profit of $23 million in the third quarter. On a year-to-date basis, we've reported GAAP and non-GAAP operating income of $363 million and $507 million, respectively.

Turning to liquidity, cash investments increased by $54 million in the third quarter to end with a balance of nearly $1 billion compared with $954 million as of year-end 2025. From the start of the year, adjusting for $250 million of collaboration payments made to Arrowhead, our base business has generated nearly $300 million of cash year-to-date.

Looking ahead to the remaining few months of the year, in addition to the updated revenue guidance provided by Mike and Patrick, we are also narrowing full-year guidance for combined non-GAAP R&D and SG&A expenses to between $800 million and $825 million. Taken together, the guidance reflects a higher revenue floor and lower expense ceiling than our previous guidance, reflecting the continued operating discipline of our business. Over the medium term, our capital allocation strategy will continue to focus on preserving financial flexibility as we approach our 2027 notes maturity and funding our high-potential pipeline, including moving our DM1 and FSHD programs forward. And with that, I'll turn the call back to Mike for Q&A.

Michael Severino

Thank you, Operator. Let's go ahead and open up the call for Q&A.

Operator

[Operator Instructions]

The first question will come from Anupam Rama with JPMorgan. Your line is now open.

Unknown Speaker

Hi, this is Anup Patel for Anupam Rama. With your ongoing patient and physician education initiatives regarding Elevidys, what is resonating and what needs more work?

Michael Severino

So I'll start and then Patrick can provide a bit more color. With respect to the commercial initiatives that we put in place, which are both the sales force expansion and the new data that are available to the sales force in the field, I think the key things that are resonating are the long-term data in particular and the data in the 8- to 12-year-old population. As we continue to build that long-term evidence base, showing improved function in a number of domains, it re-centers the discussion around benefit-risk, and we think that's essential because it's important to keep in mind that Duchenne is still unfortunately a progressive, debilitating, and ultimately lethal disease, and showing that long-term benefit is very impactful to physicians when they're making their treatment decisions. Now Patrick, do you want to add anything?

Michael Severino

I would add that with data like this, it does take time and repetition to deliver that message and to change prescribing behavior. And with our expanded team, we're meeting with many physicians more than once. It lets us walk through the data again and again and discuss it in depth. Our focus is really consistent execution and helping patients progress through their treatment journeys.

Operator

The next question will come from Konstantinos Biliouris with Oppenheimer. Your line is open.

Preguntas y respuestas

Konstantinos Biliouris

I have one question on DMD and one on DM1 quickly. On DMD ENVISION study specifically, how does terminating this trial impact the potential return of Elevidys in the market for non-ambulatory patients? And on DM1, what splicing improvement would you consider a success given the high biological noise here and the Novartis failure with an siRNA?

Michael Severino

Okay, so on the question with respect to Elevidys and the ENVISION study, the ENVISION study was not part of our plan to return to the non-ambulatory population. The central part of that plan right now is evidence to show improved benefit-risk with the introduction of sirolimus, which is being studied in ENDEAVOR Cohort 8. ENVISION would not have provided data in time to inform that decision under any case. And given the fact that it's been on hold for more than a year, as Louise mentioned, this was purely a practical decision based on operational and technical considerations. Once we have the Cohort 8 data, as we've said in our prior call, we will discuss them with the agency promptly, and that's what's going to inform the potential path back to the non-ambulatory population. And we remain committed to addressing those data needs through whatever way is needed. But right now, the central piece of that equation is Cohort 8.

With respect to DM1, I think we're in a range with splicing correction where clearly more is better. What Novartis showed was variable, and perhaps in the 10% to 15% range, I think we would conclude that that's not enough. We certainly want to be well above that, but we're in the range where we think more is better. So our aim is to push splicing correction as high as we possibly can. Obviously, the first step in that chain is to get very solid knockdown, robust knockdown. We think the preclinical evidence that we've generated supports our ability to do that, and the ability to drive muscle concentration in the single ascending dose also supports, in our view, our ability to do that. And in the upcoming multiple ascending dose, we want to see that translate into robust knockdown, as I said, and splicing correction that is, I think, meaningfully different than what was demonstrated by the Novartis program.

Operator

The next question will come from Brian Abrahams with RBC Capital. Your line is open.

Brian Abrahams

Are you guys seeing something, I guess, beyond what you've shared before to warrant movement of the siRNAs into Phase 3 and ramp up before fully unblinding? Or is it just a matter of being prudent to maximize your efficiency? And then on a related note, do you think there could be a situation where depending on your data you could still pursue accelerated approval even if that path is not available to others?

Michael Severino

So the data that give us confidence to begin preparations for registrational studies come from all the lines of evidence that I mentioned just a moment ago. We have very strong preclinical data and good early clinical data, data from our single ascending dose studies that tell us that we believe we should be able to achieve very high levels of knockdown in both programs, that will translate to the downstream effects, whether it's splicing for DM1 and DMPK knockdown or DUX4 inhibition of DUX4-regulated genes in FSHD, and so we think those data are very strong. The single ascending dose data bolster that confidence because of the ability to achieve high muscle concentrations in FSHD. We've already shown strong knockdown. We have those data coming in the multi-dose setting for both programs, on FSHD and our ability to look at that for DM1. That constellation of evidence is what gives us confidence. And we're going to move forward rapidly. And we're prepared to act when we see the data in the upcoming weeks, as we said. So it's really that constellation of evidence that gives us confidence.

Operator

The next question will come from—

Michael Severino

So I'm sorry, Operator. There was a second part of the question that I forgot to address, which is whether we believe there could be an accelerated approval pathway for our programs, even if that was not available to others. And the short answer to that is the ability to pursue accelerated approval is based on the strength of the data. We would expect to have strong data, and so we would expect to have those pathways available to us. Obviously, we'll discuss our data with regulators as we move forward, but our expectation is that accelerated approval should be on the table for both of those indications.

Operator

Our next question will come from Andrew Tsai with Jefferies. Your line is open.

Lin Tsai

Sticking with DM1 and FSHD, totally understood what to expect on efficacy. On safety, hypothetically speaking, what could be some of the special AEs you're particularly looking out for? I don't think there's a TfR, so there shouldn't be anemia risk, but alpha-v beta-6, could that be linked to any respiratory CV signals? Maybe even for your earlier stage IPF program, but anyway, any color around what you expect on safety would be very helpful.

Michael Severino

The safety profile that we've observed, both preclinically and in the single ascending dose studies, has been very favorable. And we've not identified dose-limiting toxicities or particular toxicities that might, for example, be AEs of interest going forward. As you mentioned, we do not have a transferrin receptor component to our delivery system, and so anemia would not be something that would be expected in our program. And with respect to the integrin targeting and experience with integrin molecules, modulation as a therapeutic program, I think they're very different things. So we know that in some cases, when very high levels of alpha-v beta-6 have been targeted, they have led to pulmonary effects when prolonged suppression is the goal. However, as a delivery vehicle where there's only intermittent exposure, we've observed a very favorable profile and so we would not expect to see similar findings in our program. James, I don't know if you want to add anything further.

Operator

The next question comes from Ellie Merle with Barclays.

Eliana Merle

Just another couple on DM1 and FSHD. Just as you guys are moving quickly into registrational studies, how are you thinking about dose selection? Could you potentially take higher doses forward than the ones that we will get at these upcoming MAD readouts? And then also, you mentioned you plan to move forward rapidly into registrational studies. Could you just give a little bit more clarity in terms of what specifically this means from a timeline perspective when you might meet with the FDA and how you're thinking about that?

Michael Severino

Certainly. I'll start and then I'll let James provide some more color. So with respect to dose selection, in both of these indications, we are in a situation where we believe greater knockdown is better. So we will take forward the highest dose that gives the best knockdown that is tolerable, and all of the evidence to date would make us think that the full dose range that we're exploring will be tolerable based on all the evidence that we've seen to date. And so there are higher dose cohorts in both the DM1 and FSHD programs that are underway that would give their full readout in early next year, as we've mentioned in other settings. And so it certainly is possible that 1 of those could ultimately be the selected dose. However, we do believe we'll see enough evidence and a strong enough indication of dose response with the upcoming data readouts that we'll have a good indication of both the ability and the dose that we will likely move forward into registrational studies. So, James, I don't know if you want to provide any additional detail.

James Richardson

Just to reiterate, I think we will get a large range of exposure out of these upcoming readouts, but as Mike has referenced, the FSHD program has a higher dose cohort that is still in the clinic at the moment. We'll get a read on that early next year. I think putting these two readout points together will give us an opportunity to pick our dose and discuss that with our regulatory agencies in the new year.

Operator

The next question will come from Ritu Barooah with TD Cowen. Your line is open.

Joshua Fleishman

Curious, what needs to be done prior to releasing the upcoming MAD data, and once that data is presented are there any gating items to pivotal trial start? And then one more to tag on to that, how does Novartis's recent miss in HARBOR and potential patient heterogeneity shape the way you think about designing a potential pivotal DM1 study?

Michael Severino

Okay, maybe I will start with the comments on HARBOR, and then Louise, do you want to pick up with the question about what needs to be done prior to release of the data? With respect to HARBOR, obviously we need to see more complete data from HARBOR to have a better understanding of exactly what happened. What is publicly available is still just top-line commentary essentially on the results: the fact that their dose failed to hit statistical significance, although there have been follow-up comments that they believe they're seeing clear evidence of biological effect in multiple secondaries. And we think that's a good thing, and that's consistent with what we already know in the field, which is that DMPK mutations are causal in this disease.

But I would say there are two things that we think about when we look at HARBOR. First is just the degree of DMPK knockdown and splicing correction that one is able to achieve. And as I mentioned earlier, more is better. And the higher up we are on that curve of knockdown and splicing correction, obviously, I think it strengthens our ability to show an impact even in a similarly designed study. But it's also worth looking at those data when they are available, which is going to be in an upcoming meeting. We're hearing it may be MDA in the spring, but we'll have to see if that in fact is confirmed, to see how the endpoint vHOT performed, because there are a number of questions that remain to be answered. Were there patient selection, patient variability issues that may have clouded the ability to get a positive readout in that study, along with potentially a knockdown that was not quite sufficient to demonstrate an effect in that population?

There are design considerations that one could use to address this: baseline enrollment criteria, potentially even stratification if necessary. So we'll keep an eye to all of those things as the data emerge and as we discuss our plans with regulators. But I'd also say we're not locked into any particular study design at this point. We have the flexibility to choose amongst various approaches, both for accelerated approval and full approval, and that will be a focus of our conversations with regulators early next year. So, Louise, do you want to take the question about data readiness?

Louise Rodino-Klapac

Sure. I think in terms of what needs to be done, it's typical for any end of a study right now, which is really data transfer, data cleaning, QC, and then any analysis that needs to be done with that individual dataset. So that's the phase that we're in right now in terms of getting ready for the upcoming readouts.

Operator

The next question will come from Brian Skorney with Baird. Your line is open.

Brian Skorney

I guess I also want to push a little bit on the HARBOR results and how you sort of interpret potential pivotal trial design as you're thinking about structuring that. And kind of given the historical context Sarepta has had, I mean, you've run three randomized controlled studies in a muscle disease, and you're sort of in this regular shift to trying to find the right endpoint, the right timeframe for analyzing the endpoints. And we haven't really seen that come to an effect. So I'm just wondering how you kind of think about de-risking a similar pathway in DM1, where we don't really know as of now, even when we restore muscle normal protein function, how to design a study with an endpoint that can capture an effect with a heterogeneous clinical function. And that time course of measurement doesn't sort of overwhelm any clinical difference. So when you think of your pivotal trial design, is vHOT a good endpoint, is 5-times sit-to-stand a good endpoint, is MDHI—how do you kind of fall on what, assuming you have to do a full clinical study for full approval at some point, what is the clinical endpoint that you're looking at and timeframe for evaluation?

Michael Severino

Well, I'll let James comment on the clinical endpoint. The one thing that I would say is that the further up that curve we are with respect to our ability to knock down the toxic mRNA and improve splicing, the greater our ability will be to show improvement on whatever endpoint we ultimately select as long as, of course, it is disease-related, and we will be picking endpoints that are very clearly disease-related. But James, do you want to talk more specifically about endpoint and study design?

James Richardson

Yes, absolutely. I appreciate the reference to our experience in this area. We are obviously experienced neuromuscular developers with a lot of in-house expertise. And I think 1 very clear learning is it's not the endpoint in itself, but it is the population in which you test this endpoint and the duration over which you're testing it. So I think without revealing our inner workings, we are going through a process of reviewing the access to external data that we have from a natural history perspective, our own data, and interacting with KOLs in the field in order to find that most responsive endpoint or endpoints across the appropriate time period and in the appropriate population.

Operator

The next question is going to come from Joe Schwartz with Leerink Partners. Your line is open.

Joseph Schwartz

Mike, you came from AbbVie where development is run to big pharma standards, and you've said that you want Sarepta judged on execution. Sarepta's history is mixed on clean wins. So I'm wondering, after almost 100 days, what's your honest view of Sarepta's trial design and execution capabilities? And specifically, is there anything that you would change in order to give DM1 and FSHD programs the best shot at being able to succeed based on not only just choosing the patients that you enroll and the endpoints you study, but setting up statistics and training raters and picking sites and everything else that goes into good clinical trial hygiene? Can you give us some insight into your view now that you've been there a while?

Michael Severino

Sure, I'd be happy to do that. And as you point out, I've run clinical development programs around the industry for many years now across a wide range of companies, across Amgen, across AbbVie. And a couple of things I think are transferable across disease states. One of the most important is if you want to have a positive study, you need to have the most substantial effect size you can. And that's why our focus on the ability to deliver, the ability to knock down whatever species we're talking about, whether it's DMPK or DUX4, is so critical. And then a number of study design elements become essential as well. These endpoints require experienced clinical sites to administer effectively, to administer consistently, and to get the sort of quality data that we need.

I think we've gained significant experience as a company over the last many years in how to conduct studies in neuromuscular disease and how to pick those sites, and how to work with those sites to make sure we get quality data. With respect to our study design and statistical capabilities, they are as good here as they are anywhere. I think we know how to design studies well and I have a lot of confidence in the siRNA programs going forward, so I think we'll put together the various pieces of what it takes to run a very effective and what I believe will be a successful registrational Phase 3 program, as we get these data in-house, as we discuss with regulators the overall path to approval, whether that's accelerated initially and in the longer term full approval. But I'm confident that we have the pieces here that we need to do that successfully.

Operator

The next question will come from Yanan Zhu with Wells Fargo. Your line is open.

Yanan Zhu

It's a two-part question on FSHD and DM1. Regarding FSHD, I think you mentioned you look for concordance across PD markers in order to distinguish signal from noise or variability in individual assay. I was wondering, could you elaborate on that? Is there a higher hurdle in terms of signal-to-noise for FSHD as compared with DM1, given that comment wasn't associated with DM1? And then perhaps just quickly wanted to clarify, are you going to support vHOT for DM1? And if you do, how should we interpret that data?

Michael Severino

Yeah, I'll ask Louise to comment on the question with respect to FSHD. With respect to DM1 and vHOT, we will show the functional measures we've collected. Those include vHOT. So that will be part of the data that we expect to release. What I would say is these are relatively small cohorts, or very small cohorts, with respect to the functional endpoints. They are well-sized with respect to the biomarker endpoints, the ability to judge target knockdown, and the impact that has on downstream biologic effects, but they are quite small compared to a Phase 2 expansion or certainly a Phase 3 study. And so that's why we have made the point that 1 should not over-index on functional measures very early on in a clinical program.

With respect to consistency of measures, whether they're biomarkers in FSHD or in DM1, before I hand it over to Louise, what I would say is just it's a general point that we always want to see consistency of measures across a chain of events that we think are causally connected. And that's 1 of the ways that we assess the strength of evidence. Louise, would you like to add any additional color around FSHD in particular?

Louise Rodino-Klapac

Sure. For FSHD, so again we're looking for concordance across our PK, so our muscle and serum concentration, our DUX4 target genes, and then the circulating biomarkers. And so with FSHD, we have the opportunity to look at circulating biomarkers, which supplements the biopsy data, which is a single time point versus circulating biomarkers. So that was really what the comment was about. Again, we're looking at the totality of the evidence and looking for that consistency across from knockdown all the way down to eventual functional improvement.

Operator

The next question will come from David Hoang with Deutsche Bank. Your line is open.

David Hoang

Wanted to pivot a little bit here and ask about EXONDYS and get your latest thoughts on the potential impact if a competitor exon skipper for exon 51 DMD is approved by the January PDUFA date and how you think about the durability of EXONDYS, how much switching could actually occur next year. And then if I could just have a follow-up on your base case, let's say, for the regulatory path forward for restoring a non-ambulatory label for Elevidys, once Cohort 8 data from ENDEAVOR is in hand, where do you kind of go from there and what are your expectations in terms of what the FDA may ask for, if anything else?

Michael Severino

Okay, so with respect to potential competition for exon 51 and the exon skipper, we're prepared for that. Our baseline assumption is that Dyne will get their approval in early '27, in January '27. We believe we are in a strong position to compete in a competitive marketplace. We have a very long track record of safety, a well-described safety profile, significant long-term real-world evidence that shows improvements in a number of areas, including overall function, cardiac function, pulmonary function, compared to matched controls, and we think that positions us very well.

And the other thing that I would say is it takes time in a marketplace like this to establish the appropriate commercial footprint, the reimbursement codes, navigate the reimbursement landscape, and that typically can take a period of months to several months. Based on all those factors, we think wholesale switching of stable patients is not likely early on, and we would expect to see impact when we see it more towards the back end of '27. And at this point, I think we need to see a little bit more, see what their label looks like before we give any more specific color, but that's certainly the way we're thinking about it.

With respect to the non-ambulatory population for Elevidys, what we've said on past calls I think is still very much the case. Discussions with the FDA on a potential path to return to the non-ambulatory population are best had with data in hand. So our plan is to approach regulators, approach the FDA in the first quarter, towards the end of the first quarter when we have the 12-week safety data from Cohort 8 in hand, and we'll have those discussions. And that's what will inform the specific steps from there. Thank you.

Operator

The next question will come from Salveen Richter with Goldman Sachs. Your line is open.

Tommie Reerink

Just a quick pipeline question on Huntington's. In the context of recent siRNA data in the field, maybe speak to what you're targeting in terms of the HTT knockdown with SRP-1005 and your confidence in the overall brain distribution and safety profile here.

Michael Severino

So with respect to our Huntington's program, first of all, we're very encouraged by all the data that we've generated to date. In our Huntington's program, I think the preclinical data very clearly demonstrate that we are not only able to cross the blood-brain barrier, but in the manner in which we're doing that, we get to the deep brain nuclei that 1 needs to impact to demonstrate clinical benefit in Huntington's disease. We see that through the early studies up through NHP studies, and in a separate program, Arrowhead has shown the ability to cross the blood-brain barrier using the same delivery technology that we're using. So we think that is something that adds additional enthusiasm from our perspective about our ability to hit the target that we want to hit.

We're aiming for very substantial, high-performance huntingtin knockdown. I think at this point, it's hard to give an exact target in terms of what we would expect in CSF knockdown, for example. But the important point to consider here is that by the nature of our delivery system, delivering systemically and getting across the blood-brain barrier, we essentially know that we're getting not only to the surface that interfaces with CSF, as 1 would do in an intrathecal program, but actually getting to those deep brain nuclei where the molecule would really have to do its work. And so we look forward to generating the data next year to prove biologic activity. We'd expect to have those data sometime around mid-year.

James Richardson

Yes, I need to reiterate that point, Mike, that I think it's, for the field, positive to see therapies driving mHTT knockdown in the CSF, but if you're giving your therapy in the CSF and measuring it in the CSF, then you haven't really demonstrated that you're getting it to the area where the disease is active. In contrast, the point Mike made, that when you're getting it systemically measured in the CSF, we know that we are really hitting the disease where it's causing that neurodegeneration.

Operator

The next question will come from Tazeen Ahmad with Bank of America. Your line is open.

Unknown Speaker

Can you hear me now? I'm sorry. Hello? Can you hear me now?

Operator

Yes. You are coming through clear now. You may continue.

Unknown Speaker

As you think about how your PMO franchise could coexist with your gene therapy franchise in DMD, do you have any updated thoughts on that, and whether or not over time you think you would have insurers be more willing to reimburse for both at a more consistent level?

Michael Severino

So Patrick, would you like to take that question?

Michael Severino

So the insurance companies, as they evaluate our therapies, we talk about our long-term evidence for Elevidys and, of course, with our PMOs, and we also point to the real-world evidence. The insurance policies that are established today are really silent on the real-world evidence for the most part, moving from Elevidys back to a PMO. It's something that the team continues to keep an eye on, and we do engage with prescribers and the payers that have questions around the benefits of both treatment of Elevidys and a PMO. And so those conversations are ongoing, and we continue to provide those payers with additional information as it's developed.

Operator

And the next question will come from Lachlan Hanbury-Brown with William Blair. Your line is open.

Lachlan Hanbury-Brown

So you mentioned that you want to approach the FDA with the Cohort 8 data in hand in Q1 to then sort of figure out the path forward. Just wondering, can you remind us on the interactions you've had with them about this in the past when you've spoken to them about Cohort 8, maybe what they want to see out of it, if there are any particular safety or if they want any biomarkers or anything?

Michael Severino

Yes, James, do you want to go ahead and take that?

James Richardson

Yes, thanks for the question. So we discussed the trial design with the FDA in the latter half of 2025, when we were also going through the safety label change negotiations. So the primary outcome and the trial design are agreed. This is a trial to look at the impact of sirolimus on the mitigation of ALI. Primarily, we'll be taking biopsy data. I think that will be interesting more broadly because we've seen preclinically some evidence that sirolimus may increase the rates of micro-dystrophin expression, but that's not going to be central to our FDA discussion and wasn't discussed during our agreements around the protocol.

Operator

[Operator Instructions]

The next question will come from Andy Chen with Wolfe Research. Your line is open.

Jason Thomas Bouvier

Just wanted to quickly ask about your outlook for Elevidys revenue growth in 2027 and wanted to touch a little bit on REGENXBIO potentially launching in the second half of '27 targeting younger patients. So are you seeing any treatment deferrals or anticipating market share loss, and does this affect your 2027 outlook and the strategic importance of the planned under-4 label expansion?

Michael Severino

Okay, there are a couple parts to that question, so I'll start. Patrick, you may want to add a bit later. With respect to REGENXBIO, I think we still need to hear more from them about the discussions that they're having with regulators later on this year to understand what a likely path forward would be for them and when a submission and approval might be appropriate. Our view is that there should be a consistent bar for programs in this area, particularly when there is a full approval, such as we have in this space, which generally includes placebo-controlled datasets that are more substantial than what others have presented publicly before. So, for example, we have treated a very large number of patients pre-approval and approximately 1,500 patients, more than 1,500 patients post-approval. And so, it's important to keep in mind that the dataset that would be valuable for benefit-risk for a program that's much earlier like REGENXBIO.

Having said that, part of your question was do we see treatment deferrals? There, my point would be that there is an urgency to treat in this condition. We know that these children have unfortunately a progressive, debilitating, and ultimately lethal condition, and that as time passes, there's increasing muscle destruction that is irreversible. And so there really is an urgency to treat. And when there's a fully approved product that has shown the long-term benefit that Elevidys has shown, we think that's very meaningful to physicians and to patients and their families. And so that's not something that we're seeing significantly.

So the under-4 population, that's an important population to us for xactly the reason that I just described, that with time, there's progressive muscle destruction, and therefore, treating earlier, we believe, is important. We believe our ability to treat the under-4 population is strongly supported by the evidence that we've generated in the 4-and-above population and the correlation that we've shown between dystrophin expression, micro-dystrophin expression with our therapy, and long-term clinical benefit. So it's a population that is important to us, regardless of the competitive considerations, because we think it's the right thing to do for patients. It does open up, if successful, an important additional population for us, but that's the way we're thinking about it.

Operator

I am showing no further questions at this time. My apologies, someone just came into the queue. So our next question is going to come from Gil Blum with Needham. Your line is open.

Jonathan Wolleben

I was wondering with the Elevidys trajectory given the six-month lag that you guys have previously talked about from enrollment to infusion and the relatively flat enrollment forms from second quarter to third quarter. Should we be expecting Elevidys revenues to remain relatively stable then through the first quarter of 2027? And when would you guys expect sales growth to begin re-accelerating? And are there any early indicators that would suggest inflection could occur sooner?

Michael Severino

So this is Mike. I'll start, and then I'll hand it over to Patrick. As we described, the dynamic is such that with the six-month period journey from enrollment form to infusion, what you're seeing in terms of results in Q3 and in Q4 in the future will be demand that was essentially generated earlier in the year, at a point before we had our commercial initiatives in place, before the long-term data and the 8- to 12-year-old data that I described were in place. And so we think of this as essentially a clean baseline. We are expecting that we have stabilized both enrollment form activity and Elevidys demand overall going forward.

We're going to want to see longer-term trends before we give more color on when we would expect to see growth return in the franchise. We do believe that there is the potential for long-term growth, but again, particularly given the variability inherent in a 1-time gene therapy, we want to see some longer-term trends before we give too much more color on that. But our long-term view of the potential of Elevidys has not changed. And there are a number of upcoming data readouts and milestones that could influence that, such as the Cohort 8 data. We talked about the under-4 submission that we're planning next year. So there are a number of things that will determine the overall trajectory there. But again, our view of the long-term potential and the ability to grow Elevidys remains unchanged. It is a little bit early to talk about the specific timing. Patrick, I don't know if you want to add any additional color.

Michael Severino

Great, thanks, Patrick. I would just add that we're going to continue our execution of our commercial initiatives as you've laid out, including patient identification, our physician education, and our treatment education.

Operator

I am showing no further questions in the queue. This concludes today's conference call. Thank you for participating. You may now disconnect.

This live transcript is auto-generated without human intervention or review.