HER3-DXd

HER3-DXd (patritumab deruxtecan) is an antibody-drug conjugate (ADC) from Daiichi Sankyo and Merck that targets HER3, a signaling protein found at high levels on many solid tumors. The lead program is a Phase 3 trial in HR+/HER2-negative metastatic breast cancer (breast tumors driven by estrogen or progesterone receptor signaling but without HER2 amplification, the subtype that accounts for roughly 70% of advanced breast cancer), NCT07060807, enrolling roughly 1,000 patients with Merck Sharp & Dohme as the listed sponsor [6]. A Phase 2 readout in that same population (Pistilli et al., Nat Med 2025) showed activity in heavily pretreated endocrine-resistant disease (tumors that have stopped responding to anti-estrogen hormonal therapies), which is the clinical rationale for betting on Phase 3 [2]. The commercial stakes are high. In 2024 Daiichi and Merck withdrew their US accelerated-approval filing (BLA, Biologics License Application, the FDA submission required to market a biologic) in EGFR-mutant non-small cell lung cancer after HERTHENA-Lung02 did not deliver the overall survival result FDA wanted, which shifted the center of gravity to breast cancer and a wider HERTHENA panel of trials across solid tumors [9]. The October 2023 Merck-Daiichi partnership carries up to $22 billion in total deal value across HER3-DXd and two other Daiichi ADCs (ifinatamab and raludotatug deruxtecan), with $4 billion upfront, which gives both sides heavy incentive to find an approval path somewhere [10]. Breast cancer is now the make-or-break indication.

HER3-DXd remains investigational with no approvals anywhere in the world. The furthest-along program is Phase 3 in HR+/HER2-negative metastatic breast cancer, where NCT07060807 is recruiting under Merck Sharp & Dohme sponsorship, with a 1,000-patient target and progression-free survival as the primary endpoint [6]. Primary completion is scheduled for 2028, with interim analyses expected earlier. Beyond breast cancer, Daiichi runs NCT06172478 as a Phase 2 basket in locally advanced or metastatic solid tumors (n=740) [7], and Merck is pursuing gastrointestinal cancers in NCT06596694 (Phase 1, HERTHENA-PanTumor02). The TUXEDO-3 Phase 2 trial reported activity in active brain metastases from both breast cancer and NSCLC in late 2025, published in Lancet Oncology [3][4]. HERTHENA-PanTumor01 produced response-rate data across advanced solid tumors in a smaller signal-finding study [1]. The regulatory history matters: Daiichi and Merck submitted a BLA under accelerated approval for EGFR-mutant NSCLC in late 2023 based on HERTHENA-Lung01 Phase 2 data, then withdrew it in June 2024 after FDA raised issues tied to the confirmatory Phase 3 HERTHENA-Lung02 and manufacturing considerations [9]. No active FDA Breakthrough Therapy or Fast Track designations are currently published for the breast cancer program, which is unusual for an ADC in late-line disease. Public filings do not disclose whether designations were sought and declined or simply not pursued, so the absence is itself a data gap rather than a clear negative.

HER3 is a receptor sitting on the surface of cancer cells. Unlike its cousin HER2, HER3 does not have working kinase activity of its own, so it cannot send growth signals by itself. It has to pair up with HER2 or EGFR to signal, and when it does, it fires up the PI3K-AKT pathway, which keeps tumors alive and resistant to therapy. Tumors upregulate HER3 as an escape route when you block EGFR or HER2, which is why it has been a target of interest for two decades, and why essentially every monoclonal antibody attempt against HER3 (seribantumab, lumretuzumab, patritumab as a naked antibody) failed to deliver clinical benefit [5]. The ADC approach changes the logic. HER3-DXd binds HER3 on the tumor cell, gets internalized, and releases deruxtecan (DXd), a topoisomerase I poison that breaks DNA strands and kills the cell. DXd is the same payload used in Daiichi's Enhertu (trastuzumab deruxtecan), the most commercially successful ADC on the market. A critical feature of DXd is bystander killing: the payload is membrane-permeable, so once released inside a HER3-expressing cell it diffuses outward and kills neighboring tumor cells regardless of their HER3 status. This is the standard mechanistic justification for why DXd-based ADCs can work in tumors with heterogeneous or low target expression, and it is what connects the Phase 2 observation to the Phase 3 design. The thesis: HER3 does not need to drive the tumor, it just needs to be present often enough on some cells to serve as a delivery address, and bystander killing handles the rest. Pistilli et al. (Nat Med 2025) reported single-agent activity in HR+/HER2-negative breast cancer regardless of baseline HER3 expression level, which is both encouraging (broad population) and a worry (no biomarker to enrich responders) [2].

NCT07060807 is a Phase 3 open-label randomized study of HER3-DXd (dosed 5.6 mg/kg intravenously every three weeks, the same regimen used across the HERTHENA program and matched to T-DXd) versus investigator's choice chemotherapy in HR+/HER2-negative metastatic breast cancer, following progression on prior CDK4/6 inhibitor-based regimens [6]. CDK4/6 inhibitors are a class of targeted drugs including palbociclib, ribociclib, and abemaciclib that are now standard first-line therapy for hormone receptor-positive advanced breast cancer, so this trial addresses patients who have already exhausted the first-line targeted option. Enrollment target is 1,000 patients, primary endpoint is progression-free survival (PFS, the time from randomization until the tumor grows or the patient dies) by blinded independent central review, with overall survival as a key secondary. The chemotherapy comparator basket (capecitabine, eribulin, vinorelbine, gemcitabine) is the current standard in this setting, so a clean PFS win would be clinically meaningful. The design concern is biomarker strategy. Patients enroll regardless of HER3 expression, which maximizes commercial opportunity but puts pressure on the drug to show benefit across the full population rather than in an enriched subgroup. The bystander-killing property of DXd (see mechanism) supplies the biological argument for the biomarker-agnostic approach, since even patients with low or heterogeneous HER3 may still see tumor cell death through diffusion of payload from neighboring HER3-positive cells. That argument is biologically coherent but is not an efficacy safety net. Compare that to the T-DXd story in HER2-low breast cancer, where Enhertu's DESTINY-Breast04 win rested on IHC-defined HER2 expression. Pistilli's Phase 2 used membrane HER3 H-score (a continuous IHC scoring metric combining staining intensity and the fraction of positive cells, rather than the 0/1+/2+/3+ categorical scale typically used for HER2) as the exploratory biomarker and found no clean cutoff separating responders from non-responders, which Daiichi has used to justify the biomarker-agnostic approach but which also means the Phase 3 runs without an enriched subgroup to fall back on [2]. The Phase 3 record on clinicaltrials.gov indicates HER3 IHC is being collected prospectively for subgroup analysis, though not used for eligibility. Primary completion is 2028. The clinicaltrials.gov record shows the trial is actively recruiting but still in early enrollment, so the first regulatory-grade readout is more likely 2028-H1 than sooner, barring an interim PFS hit.

Our model estimates a 36% chance this drug is eventually approved. It starts from the historical base rate for Phase 3 drugs in this area (about 48%), then adjusts using ten facts about the trial and sponsor. What moves the number most: it is helped by more secondary endpoints than usual, its light or open-label blinding, and larger-than-typical enrollment for this phase; it is held back by weak or limited earlier-phase results. The other facts land near average for this stage, so they leave the estimate roughly where the base rate put it.

Four distinct failure modes stack. First, efficacy. The Phase 2 ORR (overall response rate, the percentage of patients whose tumors shrink by a prespecified amount, typically 30% or more by RECIST criteria) of roughly 30% in HR+/HER2-negative breast cancer (Pistilli et al.) is respectable but not Enhertu-caliber, and the chemotherapy comparator in Phase 3 produces 10-15% ORRs with real PFS numbers, so the delta must be large enough to clear statistical and clinical thresholds without a biomarker screen [2]. Second, the HERTHENA-Lung02 precedent. The Phase 3 NSCLC program together with manufacturing issues drove a BLA withdrawal in June 2024 [9]. ADCs in post-chemotherapy settings often face crossover and subsequent-therapy effects that blunt OS, and that same risk exists in the breast cancer trial given OS is a key secondary. Third, safety. The DXd payload class carries a documented interstitial lung disease (ILD) signal, with grade 5 events reported for Enhertu and seen across the deruxtecan family. Pistilli et al. reported treatment-related ILD/pneumonitis in approximately 7% of HER3-DXd-treated patients at any grade, with grade 3 or higher events in roughly 1-2%, broadly in line with T-DXd's DESTINY-Breast04 profile (approximately 10% any grade, approximately 1% grade 3 or higher) [2]. That is a managed liability rather than a program threat on current evidence, but it requires active monitoring and dose management, and the signal in a Phase 3 population with 1,000 patients may expose rarer severe events not captured in Phase 2. Fourth, commercial. Datopotamab deruxtecan (Daiichi's TROP2 ADC, approved by FDA as Datroway on January 17, 2025 for pre-treated HR+/HER2-negative metastatic breast cancer) already occupies the same post-CDK4/6 setting [11]. If HER3-DXd approves after Datroway and after Enhertu's HER2-ultralow expansion, Daiichi has to argue sequencing or combination rather than displacement, which complicates uptake even in a positive outcome scenario.

Worth watching, but the signal to catch is not just the Phase 3 PFS readout. The earlier data points matter. TUXEDO-3 brain metastasis results in Lancet Oncology (late 2025) and the ongoing HERTHENA basket trials will show whether HER3-DXd has a differentiated niche that chemotherapy and existing ADCs cannot touch, with CNS penetration the most defensible claim [3][4]. The combination study with trastuzumab deruxtecan (NCT07701941) matters too: a positive signal there reframes HER3-DXd as a backbone combination partner rather than a solo asset, a stronger commercial story than displacing chemotherapy outright [8]. On the money side, Merck's roughly $65 billion revenue base and Daiichi's Enhertu franchise (multi-billion-dollar product growing double-digits) give the program institutional staying power even if the breast cancer Phase 3 wobbles [11]. The 2023 Merck-Daiichi deal priced the three-ADC package at up to $22 billion with $4 billion upfront [10]. The public disclosures do not break out HER3-DXd-specific milestones from those tied to ifinatamab or raludotatug deruxtecan: the economics are bundled across all three ADCs. Asset-level valuation of HER3-DXd therefore requires an assumption about how Merck allocated the upfront and the remaining roughly $18 billion in contingent milestones across the three programs, and the NSCLC BLA withdrawal has already impaired at least one lane of that bundle. On patent timing, patritumab's underlying antibody composition-of-matter coverage runs into the mid-to-late 2030s on the US side, with ADC-specific linker-payload patents and method-of-use filings extending further, giving roughly a decade of exclusivity post-approval before biosimilar risk becomes material (the ADC format itself raises the technical barrier to biosimilar entry well above a conventional monoclonal). Check back when NCT07060807 reports its first interim analysis (likely 2027) or sooner if a Fast Track or Breakthrough Therapy designation is announced on the TUXEDO-3 brain data. Until then, treat as a 30-35% probability bet with asymmetric upside given the breadth of HERTHENA development.