Cuộc họp báo cáo kết quả kinh doanh Q3 2026 của Arvinas (ARVN): Thời gian duy trì dòng tiền và các cột mốc tiến trình phát triển

Thông tin trọng tâm

• Arvinas kết thúc quý 3/2026 với 558,9 triệu USD tiền mặt, các khoản tương đương tiền, tiền mặt bị hạn chế sử dụng và chứng khoán kinh doanh, giảm từ 685,4 triệu USD vào cuối năm 2025. Ban lãnh đạo dự báo nguồn tiền mặt hiện có đủ để duy trì hoạt động đến nửa sau năm 2028.
• Chi phí phi GAAP hàng quý giảm 25% so với cùng kỳ năm ngoái xuống 53,5 triệu USD. Chi phí R&D giảm xuống 46,9 triệu USD, trong khi chi phí G&A giảm xuống 12,6 triệu USD.
• Việc công bố dữ liệu Giai đoạn I đầu tiên của chất phân hủy BCL6 ARV-393 dự kiến diễn ra tại hội nghị ASH vào tháng 12, tiếp theo là bản cập nhật diện rộng hơn vào giữa năm 2027 bao gồm bệnh nhân u lympho tế bào B dùng liều cao hơn, độ bền phản ứng và dữ liệu ban đầu về việc kết hợp với glofitamab.
• Arvinas dự kiến báo cáo dữ liệu Giai đoạn I ban đầu trên người tình nguyện khỏe mạnh đối với chất phân hủy polyQ-AR ARV-027 trong nửa đầu năm 2027 và bắt đầu cho bệnh nhân teo cơ tủy và hành brain dùng thuốc trong cùng kỳ.
• Dữ liệu Giai đoạn I ban đầu đối với chất phân hủy HPK1 ARV-6723 dự kiến sẽ có trong nửa sau năm 2027. Việc dùng thuốc kết hợp với pembrolizumab được kế hoạch thực hiện vào nửa đầu năm 2027.
• Đối với ARV-102, Arvinas đang ưu tiên nghiên cứu Giai đoạn Ib đối với bệnh liệt trên nhân tiến triển ở EU và Vương quốc Anh vào năm 2027 trong khi đơn đăng ký thử nghiệm lâm sàng (IND) tại Mỹ vẫn đang bị tạm dừng.

Dữ liệu tài chính cốt lõi

Hiệu quả kinh doanh và hoạt động

ARV-393: Sắp tới đợt công bố dữ liệu lâm sàng đầu tiên

ARV-393 là chất phân hủy BCL6 đường uống đang được phát triển cho điều trị u lympho tế bào B và tế bào T. Đợt công bố Giai đoạn I đầu tiên tại ASH sẽ bao gồm tính an toàn, khả năng dung nạp, dược động học, dược lực học và hiệu quả sơ bộ, trong đó phân tích hiệu quả tập trung vào khoảng 15 bệnh nhân u lympho tế bào T.

Hầu hết bệnh nhân trong các nhóm thử nghiệm ban đầu đều nhận mức liều thấp hơn mức phơi nhiễm được dự đoán là có hiệu quả trong các mô hình DLBCL tiền lâm sàng. Dù vậy, Arvinas báo cáo đã ghi nhận đáp ứng trên cả hai phân nhóm u lympho tế bào B và tế bào T. Ban lãnh đạo cho biết tỷ lệ đáp ứng trên khoảng 30% trong bối cảnh tăng liều sớm ở mức thấp này sẽ là tín hiệu khả quan, nhưng dữ liệu về độ bền phản ứng sẽ bị hạn chế do những bệnh nhân dùng liều cao nhất có thời gian theo dõi ngắn nhất.

Bản cập nhật vào giữa năm 2027 dự kiến sẽ bao gồm thêm nhiều bệnh nhân u lympho tế bào B được điều trị ở mức phơi nhiễm dự đoán có hiệu quả, thời gian theo dõi dài hơn và dữ liệu ban đầu từ việc kết hợp với glofitamab. Ban lãnh đạo cho biết nghiên cứu kết hợp sớm này chủ yếu nhằm chứng minh tính an toàn và khả năng phối hợp hơn là tính vượt trội so với chỉ dùng glofitamab đơn trị liệu.

ARV-027: Hướng tới mục tiêu giảm 50% polyQ-AR

Chương trình Giai đoạn I cho ARV-027 đang tiến triển qua nhiều nhóm liều ở người tình nguyện khỏe mạnh. Arvinas đang hướng tới mục tiêu giảm khoảng 50% polyQ-AR, mức độ mà ban lãnh đạo cho biết đã cải thiện sức mạnh và độ bền trong các mô hình tiền lâm sàng trong khi vẫn duy trì một số tín hiệu androgen.

Dữ liệu ban đầu trên người tình nguyện khỏe mạnh dự kiến sẽ có trong nửa đầu năm 2027. Công ty cũng kỳ vọng sẽ bắt đầu cho bệnh nhân SBMA dùng thuốc trong khoảng thời gian đó. Ban lãnh đạo cho biết nghiên cứu trên người tình nguyện khỏe mạnh nhằm mục đích xác định liều dùng cho nhóm bệnh nhân; việc tiến hành trên bệnh nhân SBMA không phụ thuộc vào việc phải chứng minh một ngưỡng phân hủy cụ thể ở người tình nguyện khỏe mạnh trước.

ARV-6723: Đang triển khai nghiên cứu miễn dịch ung thư Giai đoạn I

Arvinas đã khởi động thử nghiệm Giai đoạn I đối với ARV-6723, chất phân hủy HPK1 đường uống của hãng, trong Quý 3/2026. Việc tăng liều đơn trị liệu đang được tiến hành, và việc thử nghiệm liều kết hợp với pembrolizumab được kế hoạch vào nửa đầu năm 2027.

Nghiên cứu lâm sàng sẽ đánh giá tính an toàn, độ phơi nhiễm thuốc, sự phân hủy HPK1 và các tác động sinh học trong khối u. Ban lãnh đạo kỳ vọng có dữ liệu ban đầu vào nửa sau năm 2027. Công ty đã dẫn chứng hoạt tính tiền lâm sàng trên 7 mô hình kháng chất ức chế điểm kiểm soát và sự thoái triển hoàn toàn của khối u khi kết hợp với điều trị anti-PD-1 làm cơ sở hỗ trợ cho chiến lược lâm sàng.

ARV-102: Ưu tiên phát triển tại Châu Âu

Sau khi nhận được tư vấn khoa học từ Ủy ban Sản phẩm Y tế Dùng cho Con người, Arvinas dự kiến sẽ nộp đơn xin thử nghiệm lâm sàng cho nghiên cứu Giai đoạn Ib của ARV-102 trên bệnh nhân liệt trên nhân tiến triển. Việc bắt đầu thử nghiệm tại EU và Vương quốc Anh được đặt mục tiêu vào năm 2027.

Nghiên cứu được đề xuất sẽ sử dụng thời gian dùng thuốc dài hơn so với các nhóm tình nguyện viên khỏe mạnh và bệnh nhân Parkinson 28 ngày trước đó. Nghiên cứu dự kiến sẽ đánh giá tính an toàn, PK/PD, các chỉ dấu sinh học và có thể cả các tiêu chí hướng tới hiệu quả trước khi tiến hành một nghiên cứu lớn hơn phục vụ đăng ký cấp phép.

Arvinas cũng đang tham vấn các cơ quan quản lý Nhật Bản về tiềm năng phát triển tại nước này. Các cuộc thảo luận với FDA vẫn đang tiếp tục, nhưng ban lãnh đạo cho biết các nhà đầu tư nên giả định rằng thử nghiệm Giai đoạn Ib sẽ bắt đầu mà chưa có bệnh nhân tại Mỹ.

ARV-806 cần có đối tác

Ban lãnh đạo nhắc lại rằng chất phân hủy KRAS G12D ARV-806 sẽ chỉ được tiến hành nếu Arvinas tìm được đối tác. Công ty dự kiến sẽ cung cấp bản cập nhật khi và nếu thỏa thuận hợp tác được thiết lập.

Dự báo của ban lãnh đạo

Arvinas kỳ vọng nguồn vốn hiện có sẽ đủ tài trợ cho các hoạt động đến nửa sau năm 2028 và trải qua nhiều cột mốc dữ liệu lâm sàng. Ban lãnh đạo cho biết chi tiêu phi GAAP Quý 3 là mức đại diện hợp lý cho chi tiêu dự kiến trong vài quý tới khi các sản phẩm giai đoạn đầu tiến triển.

Kế hoạch phân bổ vốn hiện tại bao gồm việc tăng liều đơn trị liệu ARV-393, nghiên cứu kết hợp và một số nhóm mở rộng nhất định cho đến hết năm 2027. Chương trình SBMA được tài trợ vốn trong suốt khoảng thời gian dự báo tiền mặt, mặc dù ban lãnh đạo lưu ý rằng thử nghiệm kéo dài vượt quá khoảng thời gian đó.

Các cột mốc quan trọng dự kiến bao gồm:

• Tháng 12/2026: Công bố dữ liệu Giai đoạn I ban đầu của ARV-393 tại ASH.
• Nửa đầu năm 2027: Dữ liệu ban đầu trên người tình nguyện khỏe mạnh của ARV-027 và kế hoạch cho bệnh nhân SBMA dùng thuốc.
• Nửa đầu năm 2027: Dự kiến bắt đầu dùng ARV-6723 kết hợp với pembrolizumab.
• Giữa năm 2027: Cập nhật mở rộng về ARV-393, bao gồm bệnh nhân u lympho tế bào B dùng liều cao hơn và dữ liệu ban đầu khi kết hợp với glofitamab.
• Nửa sau năm 2027: Dữ liệu Giai đoạn I ban đầu của ARV-6723.
• Năm 2027: Mục tiêu bắt đầu nghiên cứu Giai đoạn Ib đối với ARV-102 chữa PSP tại EU và Vương quốc Anh, tùy thuộc vào quy trình xin nộp hồ sơ thử nghiệm lâm sàng.

Rủi ro và các yếu tố cần theo dõi

• ARV-102 vẫn đang bị tạm dừng thử nghiệm lâm sàng tại Mỹ. Ban lãnh đạo không đưa ra lộ trình giải quyết quyết định tạm dừng này và đang tiếp tục chiến lược Giai đoạn Ib tại Châu Âu.
• Bằng chứng về ARV-393 vẫn ở giai đoạn sớm. Dữ liệu hiệu quả ban đầu phần lớn đến từ các mức liều thấp hơn mức phơi nhiễm dự đoán có hiệu quả, trong khi việc đánh giá độ bền và mối quan hệ giữa liều đáp ứng sẽ bị hạn chế do thời gian theo dõi ngắn.
• Khả năng chuyển giao lâm sàng của ARV-027 chưa được chứng minh. Ban lãnh đạo cho biết thời điểm đạt được các lợi ích chức năng ở bệnh nhân SBMA rất khó suy đoán từ các mô hình trên chuột.
• ARV-806 phụ thuộc vào nguồn vốn bên ngoài. Chương trình sẽ không tiến triển nếu không có đối tác.
• Lộ trình đăng ký cấp phép vẫn còn mang tính điều kiện. Việc đưa ARV-027 và ARV-102 vào các thử nghiệm đăng ký cấp phép sẽ phụ thuộc vào kết quả Giai đoạn I hoặc Giai đoạn Ib, các phát hiện về độc tính và trao đổi với cơ quan quản lý.

Điểm nhấn phần Hỏi & Đáp với chuyên gia phân tích

• Tiêu chuẩn hiệu quả của ARV-393: Ban lãnh đạo cho biết tỷ lệ đáp ứng trên khoảng 30% sẽ là tín hiệu khả quan trong tập dữ liệu u lympho tế bào T liều thấp ban đầu. Đối với giai đoạn phát triển sau, Arvinas lưu ý rằng dữ liệu đơn trị liệu chất phân hủy BCL6 của đối thủ cạnh tranh đã ghi nhận tỷ lệ đáp ứng trên 50%, tạo ra một điểm tựa cạnh tranh đáng chú ý.
• Lộ trình đăng ký tiềm năng cho ARV-393: Ban lãnh đạo cho biết việc tiến tương đối nhanh tới một chiến lược tập trung vào đăng ký cấp phép có thể khả thi đối với u lympho tế bào T do nhu cầu điều trị chưa được đáp ứng lớn, nhưng quyết định sẽ phụ thuộc vào các dữ liệu lâm sàng mới xuất hiện.
• Chiến lược chỉ dấu sinh học của ARV-027: Công ty đang hướng tới mục tiêu phân hủy khoảng 50% polyQ-AR. Ban lãnh đạo coi mức độ phân hủy mục tiêu trong cơ là chỉ dấu sinh học quan trọng nhất của chương trình vì polyQ-AR chính là nguyên nhân gây ra bệnh.
• Thiết kế lâm sàng của ARV-102: Nghiên cứu Giai đoạn Ib tại Châu Âu nhằm tạo dựng niềm tin thông qua dữ liệu dài hạn hơn về tính an toàn, PK/PD, chỉ dấu sinh học và có thể cả các dữ liệu hướng tới hiệu quả trước khi tiến tới một thử nghiệm đăng ký cấp phép lớn hơn. Một chương trình toàn cầu sau đó có thể sẽ bao gồm cả Mỹ và Nhật Bản.
• Phân bổ vốn: Ban lãnh đạo nhấn mạnh rằng các quyết định chi tiêu sẽ tiếp tục dựa trên dữ liệu, trong đó nguồn vốn được hướng tới các chương trình cho thấy bằng chứng mạnh mẽ nhất về khả năng tạo ra giá trị tiềm năng.

Toàn văn cuộc họp báo cáo kết quả kinh doanh

Toàn văn cuộc gọi công bố kết quả kinh doanh

Phần trình bày của ban lãnh đạo

Operator

Good day, and thank you for standing by. Welcome to the Arvinas Q3 2026 Earnings Call. [Operator Instructions] Please be advised that today's conference is being recorded.

I would now like to hand the conference over to your first speaker, [ Jeff Wells ], Investor Relations. Please go ahead.

Unknown Executive

Good morning, everyone, and thank you for joining us. Earlier today, we issued a press release with our third quarter 2026 financial results, which is available in the Investors and Media section of our website at arvinas.com.

Joining us on the call today, we have Randy Teel, our President and CEO; Angela Cacace, our Chief Scientific Officer; and Andrew Saik, our Chief Financial Officer.

Before we begin, I'll remind you that today's discussion contains forward-looking statements that involve risks, uncertainties and assumptions. These risks and uncertainties are outlined in today's press release and in the company's recent filings with the Securities and Exchange Commission, which I urge you to read.

Our actual results may differ materially from what is discussed on today's call. A replay of this call as well as today's press release and an updated corporate deck will be available on the Investor and Media section of our website.

And now I'll turn the call over to Randy Teel. Randy?

Randy Teel

Thanks, Jeff, and good morning, everyone. I'll start by reiterating our commitment to prioritizing investments in programs where emerging evidence supports the greatest potential for value creation, letting the data drive our capital allocation decisions.

If we execute on that commitment, we believe Arvinas can create important medicines for patients and significant long-term value for shareholders. We are entering an important period of clinical milestones that we believe will further define both our individual programs and the broader potential of our platform.

During the call, I'll provide an overview of our portfolio and plans. Angela will share additional information on our clinical programs, and Andrew will give an overview of our finances and approach to capital allocation.

I'll begin with ARV-393, our oral BCL6 degrader and a great example of a program with strong scientific rationale for differentiation. BCL6 is a key driver of lymphoma biology, and we believe our ability to degrade this classically undruggable transcription factor could offer a differentiated approach to treating B and T cell lymphomas.

We see the potential for our BCL6 degrader to become a backbone of future treatment strategies as both a monotherapy and in combination with standards of care. This could include reducing, delaying or sparing chemotherapy for patients.

During the quarter, as we continued enrolling the monotherapy and combination cohorts of our Phase I trial, we submitted our initial monotherapy dose escalation data for presentation at the ASH conference in December. This will be the first of 2 data readouts for the program with the second readout expected in mid-2027.

The first disclosure will focus on patients with T-cell lymphomas, where there are limited treatment options available. These data support our belief that ARV-393 is an active degrader of BCL6 with the potential to be very competitive as we continue development.

The data we share in 2027 will include a greater proportion of patients dosed at levels expected to reach efficacious exposures in B-cell lymphomas, more robust data on durability and initial data from the combination with glofitamab, the first of our planned combination cohorts.

I'll move next to ARV-027, our first-in-class oral androgen receptor degrader targeting the root cause of spinal bulbar muscular atrophy or SBMA, also known as Kennedy's disease. This degrader targets the polyglutamine repeat androgen receptor or polyQ-AR, which is found in patients with this genetically defined disease.

SBMA is an X-linked disease where the polyQ-AR aggregates and gains a toxic function, driving progressive muscle weakness and functional decline. By eliminating the toxic accumulation of polyQ-AR in muscle, we have the potential to create the first-ever disease-modifying therapy for patients with this rare neuromuscular disease.

The SBMA program is advancing through multiple dose cohorts in its Phase I trial, and we intend to share initial data from the healthy volunteer cohorts in the first half of 2027. We also intend to begin dosing patients with SBMA in the first half of 2027.

If successful, we believe the Phase I data for this program will significantly derisk the program, and we will begin discussions with health authorities about initiating a registration-enabling trial.

I'll turn next to our first immuno-oncology degrader, ARV-6723. By impacting the immune system beyond the ability of current IO therapies, we believe this first-in-class oral HPK1 degrader could become an important treatment for patients in 3 distinct ways: first, following a checkpoint inhibitor; second, in combination with a checkpoint inhibitor; and third, in indications where checkpoint inhibitors have not shown benefit.

Based on promising monotherapy and combination preclinical data, we initiated a Phase I trial in the third quarter. Monotherapy dose escalation is ongoing, and we plan to initiate dosing in combination with pembrolizumab in the first half of 2027.

HPK1 is a protein that dampens multiple parts of the cancer immunity cascade. And we believe that by removing it, our degrader has the potential to harness the body's natural tumor immunity to fight cancer. We look forward to providing initial data from this program in the second half of 2027.

Rounding out our clinical progress, I'll next provide a brief update on current plans for our LRRK2 degrader, ARV-102. Following scientific advice from CHMP, we're now prioritizing the initiation of a Phase Ib trial in the EU and the U.K. Next month, we plan to submit clinical trial applications for a Phase Ib trial in patients with progressive supranuclear palsy or PSP. Trial initiation is targeted for 2027.

In parallel, we're in consultation with Japanese health authorities to inform potential development in Japan as part of a long-term global clinical strategy. In the U.S., we continue discussions with the FDA about the ongoing clinical hold of our IND. Our current objective is now to include U.S. sites in future trials after completing the Phase Ib trial in the EU and the U.K.

Last month, we reported additional and novel biomarker findings from our Phase I trial, including improvements in an objective and quantifiable measure of impaired eye movement associated with Parkinson's disease. The totality of evidence from our Phase I trial strengthens our belief in the potential of LRRK2 degradation to create novel clinical breakthroughs for patients with devastating neurodegenerative diseases.

What's common across all of our programs is the scientific premise that by degrading the proteins that cause disease, we can create transformational treatments to address unmet patient need. We are pursuing targets where removing the protein has the potential to change the treatment landscape and improve outcomes for patients.

The progress we're making today is laying the foundation for the next wave of innovation at Arvinas and advancing our mission of delivering transformative medicines to patients.

With that, I'll turn the call over to Angela. Angela?

Angela Cacace

Thank you, Randy. As we approach key pipeline milestones, I'll highlight how our preclinical data support our differentiated science, inform clinical development, and position our portfolio to deliver value.

Starting with ARV-393, our strategy is to establish its potential both as a single agent and in combinations that could expand treatment options for patients with lymphoma. We've shown strong single-agent activity in preclinical DLBCL models with enhanced activity, including complete tumor regressions when combined with chemotherapy, biologics and investigational oral agents. The results with R-CHOP and bispecific antibodies support the combination strategy we are advancing in the clinic.

The first Phase I disclosure will provide an early look at safety and tolerability, drug exposure, target engagement and antitumor activity. These initial cohorts were treated at doses below those predicted to be efficacious in preclinical DLBCL models.

Even at these early dose levels, we have observed responses across NHL subtypes, including B and T-cell lymphomas. The disclosure will focus on safety, PK and PD from all patients and preliminary efficacy from early cohorts in patients with T-cell lymphoma where treatment options remain limited. These data will help guide dose selection and expansion plans.

Our next update expected in mid-2027 will include more patients with B-cell lymphoma treated at doses in the predicted efficacious range with longer follow-up, along with initial combination data with glofitamab.

Turning to our polyQ-AR degrader, ARV-027, our Phase I study is designed to establish safety and tolerability and determine whether the drug degrades the androgen receptor in skeletal muscle. Importantly, we are seeing profound preclinical activity in muscle strength and endurance, 2 of the most fundamental manifestations of SBMA disease progression.

In an aggressive mouse model of SBMA, ARV-027 induced polyQ-AR degradation in muscle, and we are excited by the data showing rescue of muscle strength, endurance and survival.

Our therapeutic goal builds on the internal data and literature evidence showing that approximately 50% reduction in polyQ-AR in muscle improved mouse model strength and restored treadmill endurance. We believe that the forthcoming multiple ascending dose data in healthy volunteers and patients with SBMA will help determine whether ARV-027 produces the biologic effects needed to support the advancement of ARV-027 into a registrational trial.

These results, together with chronic toxicology data and regulatory interactions will guide dose selection and our SBMA clinical trial design. We look forward to sharing initial data from our Phase I trial in the first half of 2027.

Now let's turn to our first-in-class immuno-oncology asset, ARV-6723, our oral HPK1 degrader now in a first-in-human Phase I trial. By removing HPK1, we aim to eliminate both its kinase and scaffolding functions to release the brakes on innate and adaptive immunity more fully than kinase inhibition alone.

Specifically, our preclinical data demonstrate 3 compelling findings: first, robust single-agent antitumor activity across 7 checkpoint-resistant models, where a clinical HPK1 inhibitor and checkpoint therapy showed no activity. Second, complete tumor regressions with anti-PD-1 combination treatment. And third, intratumoral immune effects extending beyond T-cell activation, including enhanced interferon signaling and remodeling the myeloid compartment.

This broader remodeling of the tumor microenvironment may help explain oral ARV-6723's differentiated activity and its potential to address tumors that respond poorly to current immunotherapies.

Later this week, at the Society for Immunotherapy of Cancer Annual Meeting, we will present additional preclinical data reinforcing this differentiation and supporting our monotherapy and combination development strategy. Our Phase I monotherapy trial will assess safety, drug exposure, HPK1 degradation and biologic effects in tumors.

Measures of immune activation, including NK and T cell activation will enable us to translate the promising preclinical biology to patients and further guide clinical development.

Moving to ARV-102. One of the central goals of the program has been to establish a comprehensive understanding of how LRRK2 degradation affects biologically relevant pathways associated with neurodegenerative disease. The Phase I data presented at MDS conference last month expanded our biomarker evaluation and further informed our understanding of the molecule's pharmacology.

The data showed ARV-102 achieved brain penetration and retained its potent and deep reduction of LRRK2 in the cerebrospinal fluid or CSF of patients with Parkinson's disease.

ARV-102 also reduces endolysosomal and neuroinflammatory markers in the CSF that have been shown to be elevated in neurodegenerative diseases. Importantly, we presented novel data showing improvements in an objective measure of impaired eye movement associated with Parkinson's disease.

The data showed that reductions in LRRK2 and CSF after treatment with ARV-102 is correlated with changes in ocular motor function and in synaptic biomarkers associated with enhanced plasticity and cognitive resilience. These findings strengthen our confidence in ARV-102's potential to advance treatment for neurologic diseases such as Parkinson's disease and progressive supranuclear palsy.

As we continue development, we are building on these results with biomarkers and functional measures to connect target degradation with downstream biologic effects and potential patient benefit.

In closing, we are excited about the potential of each of these programs with key data insights providing evidence to guide their advancement. We look forward to sharing our progress over the next several quarters.

With that, I'll turn the call over to Andrew to review our financial results.

Andrew Saik

Thanks, Angela, and good morning, everyone. I'm pleased to provide financial highlights for the third quarter of 2026. As a reminder, detailed financial results for the third quarter are included in the press release we issued this morning.

We are entering a period of meaningful clinical data with an innovative portfolio, clear milestones and the financial resources to deliver impactful clinical data that will help us make decisions on our key programs.

At the end of the third quarter, we had $558.9 million in cash, cash equivalents, restricted cash and marketable securities on the balance sheet compared with $685.4 million at the end of 2025. With our healthy balance sheet and projected cash into the second half of 2028, we are well positioned to continue developing our promising oncology and neurology programs.

I am happy to announce that we booked our first royalty under the license agreement with Rigel in the third quarter. The royalty of $0.5 million can be found in the revenue section of our P&L and the associated Yale royalty can be found on the cost of license revenue line.

General and administrative expenses were $12.6 million for the third quarter compared to $21 million for the same period of 2025. The decrease of $8.4 million was primarily due to decreases in personnel and infrastructure-related costs of $4.2 million, costs related to developing our commercial operations of $1.8 million, and professional fees of $1.5 million.

Research and development expenses were $46.9 million in the third quarter compared to $64.7 million for the same period of 2025. The decrease of $17.8 million was primarily driven by decreases in our program expenses of $7 million, non-program-specific expenses of $3.5 million and a decrease in compensation and related personnel expenses of $6.5 million, which are not allocated by program.

Non-GAAP R&D expenses were down $12.6 million compared to the same period last year, representing a reduction of 22%. Non-GAAP G&A decreased by $5.4 million or 37% compared to the prior year period.

Non-GAAP expenses for the quarter were $53.5 million compared to $71.5 million in the prior year. The 25% year-over-year reduction in expenses reflects reduced spend across the organization, including spend on development and reductions in G&A. Spend this quarter is a good representation of what our spend is expected to be for the next few quarters as we continue to progress our early-stage pipeline.

We've maintained a disciplined approach to portfolio decisions and capital allocation throughout the year, and our cash runway into the second half of 2028 allows us to fund operations through multiple key data milestones.

Our commitment to investors is simple. We will allocate capital carefully, communicate transparently and let the data determine where we invest. By executing on that commitment, we will be well positioned to create important medicines for patients and significant long-term value for shareholders.

We believe this creates a compelling opportunity to build long-term value as the next wave of clinical evidence emerges and demonstrates where protein degradation can offer meaningful benefits for patients.

With that, I'll turn the call over to Randy for closing remarks. Randy?

Randy Teel

Thanks, Andrew. As we look ahead, we have a number of meaningful clinical milestones approaching in the coming months, beginning with the first of 2 data disclosures for our BCL6 degrader planned for December, followed by a second disclosure in mid-2027.

We're also on track to share Phase I data for our polyQ-AR degrader in the first half of 2027 and initial data from our HPK1 degrader in the second half of 2027. And we are also working tirelessly to gain approval for the initiation of a Phase Ib clinical trial with our LRRK2 degrader in patients with PSP in 2027.

I'll leave you with what I believe is most important. Arvinas has already accomplished a big win, moving from a scientific promise to the first ever FDA-approved PROTAC degrader. What's ahead is to show that this was just the beginning of what we expect to be multiple successes in bringing important new options to patients in need.

Unknown Executive

Thanks, Randy. Operator, will you please open the queue?

Operator

[Operator Instructions] Our first question will be coming from the line of Edward Tenthoff of Piper Sandler.

Phần hỏi đáp

Edward Tenthoff

Good to hear you guys. Nice update across the pipeline. Two quick questions, if I may. First on the clinical hold on 102. Just remind us sort of where we are on resolving that with FDA.

What steps need to be done? And then maybe a quick one for Andrew. Can you remind us what the parameters are for the Yale royalty, both in terms of percentage, and does that apply to all of the following programs?

Randy Teel

Thanks, Ted, and good to hear you as well. Yes. So for 102, as we said today, the plan will be to move forward with the Phase Ib CTA process in Europe. With respect to the U.S., we still have ongoing conversations with the FDA to work on getting off that clinical hold.

At this point, investors should assume that the Phase Ib will begin without U.S. patients. But that's not to say that later in the development plan, we couldn't bring the U.S. back in. It's hard to say a lot with the conversations ongoing with the FDA, but that's where we are now.

We feel confident that we can execute that Phase Ib in Europe as we get through the CTA process, if that's successful. But that's how we'll proceed for now. Andrew, on the Yale license agreement.

Andrew Saik

Yes, thanks for the question. So yes, on the Yale agreement, we have a royalty. It's a very -- it's a slightly low royalty of about 1% on royalty revenue. So that's going to be ongoing into the future.

That does not apply to all of our programs. I don't believe that we've disclosed which of our programs are subject to the Yale royalty and which are not. I think we'll do that as we get closer to commercialization on the other programs because it's a complicated agreement that we have with Yale.

So -- and then when we get milestones and upfront, it's a slightly higher number. You could probably go back and look at what we paid to Rigel. I'm not going to disclose that number, but it's slightly higher, still low single digit. Is that helpful, Ted?

Edward Tenthoff

Great, guys.

Operator

And our next question will be coming from the line of Jeet Mukherjee of U.S. Bancorp BTIG.

Jeet Mukherjee

Congrats on the progress. Two questions. First, are you looking for an approximate 50% reduction in AR among healthy volunteers for your polyQ-AR degrader program? And what endpoints will be evaluated in the SBMA cohort you'll be dosing first half of next year?

Randy Teel

Yes. Thanks for the question, Jeet. I'll start and pass to Angela. And just as a review here, we're dosing healthy volunteers now. We'll get to SBMA patients ideally in the first half of next year. The short answer to your question of degradation of 50% is yes, but I'll let Angela dive into the detail on that.

Angela Cacace

Sure. So we're targeting 50% reduction in polyQ-AR. And the goal here is really to retain some androgen tone while proving that we can reduce 50% androgen receptor, which in our preclinical models and also in the published literature led to rescue of very important endpoints like muscle function and endurance in those models. So that -- I hope that explains why. Thank you for the question.

Jeet Mukherjee

And just one follow-up, if you don't mind. Just on ARV-102, how well do improvements on ocular motor function correlate to overall disease progression or motor symptom improvement in Parkinson's?

Angela Cacace

That's a great question. So this is a novel endpoint that we measured, which is the amplitude of psychotic [ hypometria ]. It does have concordance with the UPDRS based on published literature, but it remains to be proven how well that concords and translates into clinically meaningful improvement.

That said, we are encouraged by our data showing 2 different synaptic endpoint changes, one being CSF improvement in meaningful biomarkers like neuronal pentraxin-2 that is associated with cognitive resilience, and also, again, the psychotic eye movement, which is a measure of the basal ganglia synaptic activity.

Randy Teel

The other thing I'd add, Jeet, is -- and I think this is -- you asked about 027 and 102, and I think the point is similar for both is that doing work to explore potential biomarkers to use in the trials or with patients over time is really important.

These are both rare diseases, both in their first trials for approval will likely be rating scale-based trials. But that doesn't mean at the same time, we can't be exploring biomarkers that may correlate over time. That's true for both 102 and 027.

And on the 027 front, the most important biomarker is the degradation of polyQ-AR itself, which in this case is actually driving the disease. So the biomarker here will be important over time, and we think there's opportunities to pursue them across both programs.

Operator

And our next question will be coming from Yigal Nochomovitz of Citigroup.

Yigal Nochomovitz

I was just wondering if you could provide a quick update with respect to the partnering prospects for ARV-806, your KRAS G12D degrader?

Randy Teel

Yes. Thanks for the question, Yigal. So we had announced that, that program will only move forward with a partner. So we'll look to provide an update on that if and when a partner is found.

That program, which we had announced that earlier this year, we'll have some data coming next month -- this month at a conference, so you can look for that as well. But again, that program is only going to move forward with a partner. So we'll stick to providing guidance for the programs that we continue to push forward in Phase I as wholly-owned programs.

Yigal Nochomovitz

Okay. And then with regard to ASH, so you're in what you described as sort of below. Give us a little more commentary there with respect to what we might expect in terms of efficacy. Are you seeing efficacy at a lower dose than you may have initially predicted? And what kind of clues might you have with that early signal that would support moving higher?

Randy Teel

Yes. Thanks for that as well. Yes. So the first disclosure for the 393 program, the BCL6 degrader will be at the end of the year, as you suggested. We'll have another in the middle of next year.

So we really think of it as a 2-parter. The most important thing to show in a first data disclosure is the translation of the preclinical work into the clinic. So it will always start off with things like safety and PK and PD tolerability will be really important, especially for a program like this that over time, the development plan will include both monotherapy and combinations.

So we have said before that we need to make sure we see good PK. We have also disclosed that we began dosing at quite low doses, well below exposures that we would expect to be efficacious. We want to see good degradation of BCL6. This is a transcription factor that's rapidly resynthesized, so you need to see solid degradation of BCL6.

And on the efficacy front, we're going to focus the first disclosure on T-cell patients. If you are a second-line plus patient with nTFHL right now, you're getting something like a 35% to 45% response rate with drugs that are on the market now despite there being no specifically approved therapies for second-line T-cell patients like this, which is actually a good reminder of the unmet need in the space.

So with therapies on the market getting something like 35% to 45% in a first dose escalation study with low doses like we're in now, we and the investigators would be pretty excited with anything over really a 30% response rate. Certainly would look for that to go higher as we dose higher.

But in an early setting like this with most patients dosed at doses below where we were expecting efficacy, that would be a reasonable bar to set. I hope that helps, Yigal.

Operator

And our next question will be coming from the line of Jonathan Miller of Evercore ISI.

Jonathan Miller

Congrats on the progress. I'll follow up on Yigal's question on ASH. Given that you're primarily going to disclose efficacy in T-cell patients where your competitor in the degradation space is very focused on B-cell patients, I was wondering how we should think about read-through from one indication to another and how we should think about the relative potency of your molecule relative to the competitor? And how we should think about moving beyond T-cells into B-cell indications for the next year?

Randy Teel

Yes. Great question, Jonathan. Thanks for it. So as we said, we'll focus the disclosure in the middle of next year on the B-cell patients. We're bringing them in from the efficacy perspective.

And the reason for that is pretty simple, right? The modeling that we talk about when we talk about dosing below exposures that are expected to be efficacious, that is really done in B-cell, LBCL models. And so that's where we believe that if we take the patients that we have now and you cut down the end by the number of patients that are dosed at a level that you'd expect efficacy with various indications with response-evaluable capabilities or factors, that number gets to be pretty small.

And we want to make sure that the data that we are presenting is informative for investigators and investors. So that's why we focused the first disclosure on T-cell, and we'll wait on the B-cell disclosure for full disclosure on B-cell when we have more patients at higher doses and the ability to show durability and also the ability to show some combinability with glofi as that trial is ongoing, and we'll have more patients that we can share by the middle of next year.

With respect to translatability, we have seen responses in both B-cell and T-cell patients at lower doses. We said that even a year ago. So I think that there could be some transferability, but it will be more informative to wait until we show the data in B-cell patients where you can really get a good sense with more patients, higher doses, more durability.

Jonathan Miller

Great. And if I can sneak in one more on LRRK2. I was wondering if you could talk more medium to long term about how this Phase Ib in EU and U.K. is going to translate to potential registrational paradigms. What's your updated thought on time line to getting into registration?

Randy Teel

Yes. Good question. Thanks. So the thinking now, and Jonathan, you'll remember that we had maybe a year ago talked about the potential for 2 trials, a Phase Ib in the U.S. and a more global registrational trial.

What that's folding into now is the Phase Ib trial first that, as we said, will be in Europe and then look to move to registration after that. So it's a little early to give too much in terms of specifics for that trial, but we'll want to be dosing patients with PSP now for a longer time period.

I'll remind you that we had dosed in the healthy volunteer and PD patients only for 28 days. So this would be longer and enough time to get a good sense of safety, PK/PD as always, and also perhaps some efficacy-oriented endpoints as well, where the goal of that Phase Ib trial is to get the confidence to move into a registrational trial that's larger and later on. So that's how it plays out now.

Going back to the remarks we made on the call, there could be an opportunity at that point to make it a more global development program, potentially including Japan and the U.S. and other geographies. But we'll need to get through the Phase Ib trial first. And as we said on the call, we'll look to start that trial in 2027 after getting through the CTA process in the EU and the U.K.

Operator

Our next question will be coming from the line of Nick Lorusso of TD Cowen.

Nicholas Lorusso

Moving back to 393 again. So you provided the bars for the early data at ASH, which is very helpful. But looking to the data mid next year, what are you looking for there with the potentially higher doses as well as the glofi combo? And could you quickly move to a registrational path following these data?

Randy Teel

Thanks for the questions, Nick. So I'll answer the second one first. Yes, I think that's possible with respect to T-cell patients, where as we've talked about, there's a pretty significant unmet need.

With respect to how we think about what does good look like, which I know is really your question, this is a program that was really our first undruggable, right? So moving into BCL6 after the early success with AR and ER was an opportunity to show what a degrader can do in a space where no degraders had previously moved into -- sorry, where no programs for BCL6 had moved into the clinic.

The inhibitors haven't made it there. So the question of what does good look like for BCL6 is starting to get answered by our competitors in the space already, which is great news because it shows that the target we picked a while ago has some efficacy.

So with respect to what does good actually look like, that bar is going to get set, right? It's a little early to talk about it now. But right now patients on a monotherapy BCL6 degrader, which you've seen some other players have had response rates of over 50%, while we don't think this is a winner-take-all setting, we certainly do see it as an important setting to show that we can be competitive in.

I'll make a specific comment on the glofi combo. The glofi combo that we're running, glofi is quite an active agent. So the objective of the glofi combo will really be around showing combinability.

Certainly, the efficacy can't be worse than glofi already has, but we would expect good responses from glofi already. And in a small dose escalation trial, that is not the place to be looking for a better performance than glofi. So really, the outcome there that we're looking for is combinability, and of course, continued translation of safety, PK, PD.

Angela Cacace

And then just to add to what Randy said, the strong B-cell combination rationale also is due to the enhanced CD20 expression that we see where we would expect to see enhanced combinability with rituximab, so R-CHOP, and then also the bispecifics like glofitamab. So really excited about establishing B-cell response depth and durability.

Operator

And our next question will be coming from the line of Etzer Darout of Barclays.

Etzer Darout

Just a couple of follow-up ones. Will the decision to enter SBMA patients with ARV-027 be contingent on hitting the muscle degradation threshold you noted earlier in healthy volunteers? And then for the PSP Phase Ib study with ARV-102, just trying to understand, is the start in EU, U.K. primarily differences in the requirements for chronic toxicology data for the asset?

Is that the real driver here in terms of being able to start the study there versus U.S.? Any thoughts there would be helpful.

Randy Teel

Thanks, Etzer. The first one around SBMA patients, no. I think the way I understand your question is, will it be gating in the healthy volunteer studies to begin the SBMA patients, do we need to see a certain level of degradation? Really, the answer is no.

The multiple dose portion of the healthy volunteer study is really intended to find a dose, a single dose that we can then continue the Phase I in patients with SBMA. It's a rare disease. So it is not super easy to find a lot of patients to be using in a Phase I study, especially when, as we've said, that Phase I study will be pretty derisking and should provide the evidence from a biomarker perspective to move into a registrational study.

We would love to be saving patients to move into a registrational study before too long. Anything else on that, Angela, before we go on to the 102 that you think to add?

Angela Cacace

I think that's right, Randy. And I think going to a single site to run a Phase I study in healthy volunteers and move to SBMA, it's really sensible given the experience of the single site that we're using for the study for biomarker collection.

Randy Teel

Great. And then on the question around ARV-102 and the differences in the different regulatory discussions, really hard to comment on those ongoing discussions. But what you should assume, Etzer, is that each of those conversations is multifactorial.

And based on the fact that we're saying we'll move through the CTA process with Europe, you can assume that we've gotten scientific feedback that we think provides a path forward to those CTA conversations and starting a trial in the next year, whereas in the U.S., it looks like it will take a bit longer.

We want to make sure that we are moving the program to get to proof of concept as rapidly as we can. We think that's really important to show the value of the program. We have confidence that we can do that with studies in Europe. So that's what we'll focus on.

Operator

Question will come from Tazeen Ahmad of Bank of America.

Unknown Analyst

This is [indiscernible] on for Tazeen. Maybe a question on ARV-027. Maybe just what are you guys looking for on safety and tolerability in the healthy volunteer cohort just as you work towards that around 50% degradation level?

And then longer term, if you're able to achieve that level in SBMA patients, just curious on how you're thinking about the -- how long it would take to see translation into functional benefit. If there's any read-through that you guys have from your mouse work that informs that timing?

Randy Teel

Thanks for the question. Angela, please dive in.

Angela Cacace

Yes. So ARV-027 is not unlike any other molecule being tested for safety. We'll be looking at all aspects of safety and tolerability in that Phase I dose escalation. Hitting that 50% mark comes from our preclinical data where we did hit 50% reduction in polyQ-AR.

So removal of that disease driver rescued strength, endurance and survival in that model, and also it has been shown in other models to do the same. So it's very difficult to translate time in terms of clinical benefit.

But what is exciting is that as we move to the clinic, we'll be looking at muscle integrity. We're taking a close look at natural history data and are grateful to patients as well as NITO for providing us with their data to take a closer look at clinical endpoints that we'll be looking to translate in our registrational SBMA trial. So I'll just -- I think that covers it.

Randy Teel

I'll add one thing, which -- because I don't think we said this on the call or reminded folks of this. ARV-027 degrades both wild-type and polyQ-AR roughly equally well. But in patients with SBMA, they only have the mutant form of polyQ-AR.

So it's a X-linked disease where patients only have one kind. But that's why we think that the studies that we're doing in healthy volunteers first should be quite predictive of what we can see in patients with SBMA.

With respect to can we get 027 to be orally bioavailable, get into muscle and degrade target protein where we need to degrade it, and then degradation of the wild-type should predict degradation of polyQ, which, as we've said, is a pretty clear marker since that's actually the protein causing disease.

Operator

And our next question will come from the line of [ Lee Wojcik ] of Cantor Fitzgerald.

Unknown Analyst

Appreciate the update. Maybe a follow-up on 393 ASH update. How many T-cell lymphoma patients should we expect? Are there enough patients to perhaps assess dose response? And what about durability given many agents here have shown quite modest PFS?

Randy Teel

Yes. Thanks, Lee. So with respect to T-cell patients, as we look at efficacy, I think mid-teens or so with the same caveat that clearly, the patients that have been dosed at the highest levels, which is where you'd be more likely to see efficacy, will have the shortest follow-up just by definition in a dose escalation trial.

So the durability information in this update will be fairly limited, and you should expect to see more next summer as we bring in both T-cell and B-cell patients. We will show some. But to my point, the patients dosed at the highest levels will be the ones by definite with the shortest follow-up, I think mid-teens or so for T-cell patients.

Unknown Analyst

And I also have a question on your HPK1 degrader program. Just curious how much proof-of-concept data do you plan to generate before you refine the next steps? And then just how does the next-gen IO agents like PD-1, VEGF, impact your thinking around where you might fit in?

Randy Teel

Yes. Thanks for that question, Lee. I think the short answer is not much. I mean, we've begun that Phase I trial of HPK1 degrader and think that, that's a program that could have applicability well beyond what PD-1 therapies have done.

So the PD-1s have been great at getting over T-cell exhaustion, but there are multiple points in the immune cascade that we think HPK1 acts as a depressor and so degrading should have a benefit. So that's a trial ongoing, expected to get to its first data in the second half of next year and is going to be entirely in patients that are post-checkpoint inhibitors.

Haven't seen a lot of success there with other agents, but we think that we can see that with HPK1. But that -- the proof that we have there to move to that step has really been in the preclinical work where we've overcome something like 7 immune checkpoint inhibitor resistant models and shown better data in -- versus both PD-1 therapies as well as a prior HPK1 inhibitor. And you'll see more of that data this week at [ SITC ], if you see our poster there.

Lee, I didn't answer your question about next generations, PD-1, VEGF and so on. Look, we think an oral new access of immune -- sorry, of an IO therapy could have a lot of applicability regardless of what the other next generations are doing. It's not really -- I don't really see it as a competition there.

This is a different target with a different set of functions in the body that we can be acting on. And we intend to start combinations with pembro, like we said in the first half of next year, but you should not assume that's the only combination we can do.

Angela Cacace

And preclinically, we'll be disclosing data in the not-too-distant future around other combinations that you mentioned, and other activities that we see within the tumor microenvironment that we think are very exciting for ARV-6723 as an oral agent.

Operator

And our next question will be coming from the line of Kripa Devarakonda of Truist Securities.

Unknown Analyst

This is [ Alex ] on for Kripa. A question on capital allocation. What clinical milestones are fully funded with the current runway into the second half of 2028? And what trials and expansion cohorts does that include?

Andrew Saik

Yes. Thanks for the question. Look, we're fully funded first off on BCL6 through next year, right? So the ascending monotherapy, the combination is all funded. We've also funded some of the expansions where we're going to start looking at combinations.

I'm not going to get into it because I don't think we've given our full development plan on that, but we plan on expanding and testing combinations starting next year in a number of therapies. Those are all funded.

We have funded SBMA through our guidance period. Obviously, the trial goes a little bit further than we have the cash guidance. So we'll have to tidy that up at some point in the future. But I think as those are the major programs, that's probably where I'll stop with the guidance.

We're very careful to make sure that we have cash through to the next key readouts, right? And with cash into the second half of 2028 and all those funded, we think that we're there. So hopefully, that helps.

Unknown Analyst

Yes, that's great.

Operator

[Operator Instructions] Our next question will be coming from the line of Derek Archila of Wells Fargo.

Unknown Analyst

This is [ Jacob ] on for Derek. Just one from us on LRRK2. Should we be expecting ocular motor and biomarker data similar to what you've done in PD to also be generated from the PSP Ib study? And if so how should we be thinking about it with respect to either comps or natural history data?

Randy Teel

That's a really good question, Jacob. I'll let Angela chime in on the details on the ocular motor work, which is really interesting.

Angela Cacace

Yes. So particularly for PSP, it's an entirely different earlier onset of this palsy gaze that occurred. And so we're actually returning our digital biomarkers to what we think are more important movement and falling endpoints in that disease.

So we'll be using those endpoints. Certainly, eye tracking is something that will be studied, but also looking, again, at endolysosomal function markers as well as inflammatory markers, and importantly, translating the pathologic tau reductions that we observed that are the principal hallmark driving PSP. So we're excited about looking at those endpoints as well as the synaptic markers that are key for progression of PSP, including neuronal health markers like neurofilament light chain.

So just to give you an idea of some of the additional markers that we're looking at in that PSP trial relative to what we looked at in PD.

Operator

And I would now like to turn the call back to Randy Teel for closing remarks.

Randy Teel

Thank you very much, operator. Thanks, everyone, for joining this morning. Thanks, especially also to the patients and physicians and all of our team members here at Arvinas that make all this work possible. And we look forward to providing updates to the community in the coming months. Thanks all very much.

Operator

Thank you for participating. You may now disconnect.