In Part 3 of this series, we focus on two “second-wave” ADC programs that have attracted considerable attention over the past few months: raludotatug deruxtecan (R-DXd), a CDH6-targeted DXd ADC being co-developed by Daiichi Sankyo and Merck in ovarian and other solid tumors, and SHR-4849 (IDE849), a DLL3-targeted ADC that has shown high response rates in small cell lung cancer (SCLC).
Following Part 1 (AstraZeneca’s platform strategy) and Part 2 (Enhertu in early HER2-positive breast cancer), this article examines how Merck is using ADCs as cornerstones of its post-Keytruda strategy and how DLL3—a target previously associated with failure—may be redeemed by better ADC design. Basic concepts such as ADC architecture and the properties of DXd payloads will be covered in more detail in the separate “Beginner to Expert | ADC Intro to the Front Line: A Comprehensive Guide to the 2025 Global ADC Arms Race” series.
Merck–Daiichi’s multi-billion-dollar ADC alliance: three pillars beyond Keytruda
In 2023, Merck committed $4 billion upfront and up to $22 billion in total deal value to co-develop three DXd-based ADCs from Daiichi Sankyo, positioning them as key pillars of its post-Keytruda oncology franchise. The candidates are:
- HER3-DXd (patritumab deruxtecan): a HER3-directed ADC in late-stage development for EGFR-mutant NSCLC and other tumors
- I-DXd (ifinatamab deruxtecan): a B7-H3-directed ADC in development for SCLC and other malignancies
- R-DXd (raludotatug deruxtecan): a CDH6-directed ADC being tested in ovarian, kidney and other cancers
Among these, R-DXd currently has some of the most mature data in a clearly defined unmet-need population: recurrent, platinum-resistant ovarian cancer and related entities. At ESMO 2025, Daiichi and Merck presented phase 2 results from the REJOICE-Ovarian01 trial and announced plans to move into a phase 3 monotherapy study.
For Merck, these programs are intended not only to offset the Keytruda patent cliff (beginning around 2028) but also to build combination and sequencing strategies around PD-1 blockade, with ADCs serving as key engines of tumor debulking and immunogenic cell death.
R-DXd in platinum-resistant ovarian cancer: a 50.5% response rate in a difficult setting
The phase 2 part of REJOICE-Ovarian01 enrolled 107 patients with platinum-resistant ovarian, primary peritoneal or fallopian tube cancer, many of whom had received multiple prior lines of therapy, including PARP inhibitors. R-DXd monotherapy was tested at three dose levels, including a middle dose of 5.6 mg/kg that ultimately became the planned phase 3 dose.
Key efficacy findings were:
- Confirmed ORR across all doses (n=107): 50.5%
- Three complete responses and 51 partial responses
- Disease control rate (DCR): 77.6%
- The highest ORR (57.1%) was observed at the top dose, but the 5.6 mg/kg middle dose achieved an ORR of 50%, including two complete responses, and was selected for phase 3 based on the overall risk–benefit profile
In platinum-resistant ovarian cancer, single-agent chemotherapy typically yields ORRs in the 10–20% range. Against that backdrop, an ORR of around 50% with a manageable safety profile is a clear signal of substantial activity. The DXd payload, known from Enhertu and Datroway, provides a potent topoisomerase I inhibition and a strong bystander effect, which may be particularly valuable in tumors with heterogeneous CDH6 expression.
CDH6 as a target: how it differs from FRα
When thinking about ADCs in ovarian cancer, FRα-targeted mirvetuximab soravtansine (Elahere) is the current benchmark. It has demonstrated ORRs in the 30–40% range in FRα-high, platinum-resistant patients and has become an important option in that molecularly defined subgroup.
R-DXd, in contrast, targets CDH6 (cadherin 6), a cell-adhesion molecule overexpressed in subsets of ovarian, kidney and other solid tumors with a distinct expression pattern from FRα. By focusing on CDH6, R-DXd may open up:
- Subsets of ovarian cancer that are low/negative for FRα but high for CDH6
- Additional indications such as renal cell carcinoma and other CDH6-positive tumors
In that sense, R-DXd is not competing directly with FRα ADCs but rather expanding the addressable landscape of ADC-eligible ovarian cancer. At the same time, it leverages the same underlying DXd payload and cleavable linker technology that has been clinically validated in Enhertu and Datroway.
Phase 3 plans and potential combinations with Keytruda
The phase 3 part of REJOICE-Ovarian01 will compare R-DXd 5.6 mg/kg monotherapy versus investigator’s choice chemotherapy in platinum-resistant ovarian cancer, with PFS and OS as key endpoints. A positive outcome would position R-DXd as a strong contender for a new standard of care in this difficult setting.
Beyond monotherapy, Merck has publicly discussed the potential to combine R-DXd with Keytruda in CDH6-expressing tumors. The rationale is to exploit ADC-induced immunogenic cell death and antigen release to enhance the effectiveness of PD-1 blockade. This is conceptually similar to TROP2 ADC–checkpoint inhibitor combinations being explored by multiple companies and could, if successful, broaden the role of both agents across lines of therapy.
SHR-4849 in SCLC: ORR 73%, strong brain-metastasis activity
The DLL3-targeted ADC SHR-4849 (IDE849) represents a striking example of how a previously “failed” target can be revisited with improved ADC design. DLL3 is a Notch pathway inhibitor that is highly expressed in approximately 85% of SCLC tumors while being largely absent from normal tissue—an attractive target on paper, but one whose clinical potential was questioned after the failure of the earlier DLL3 ADC rovalpituzumab tesirine (Rova-T).
In a phase 1 study presented at WCLC 2025, SHR-4849 was evaluated in 71 patients with relapsed or metastatic SCLC and other DLL3-positive neuroendocrine tumors. The results were impressive:
- ORR: 73% (52/71)
- DCR: 93%
- In patients receiving SHR-4849 as second-line therapy, ORR rose to 77% and DCR to 97%
- Among patients with brain metastases, intracranial responses were observed in 15 of 18, underscoring activity in a particularly hard-to-treat subgroup
Given the extremely limited options and poor prognosis in relapsed SCLC, especially with brain metastases, these numbers are remarkable. They suggest that DLL3 remains a highly relevant target when paired with the right payload and linker chemistry.
Learning from Rova-T: how design differences reshape safety
Rova-T initially generated excitement in DLL3-high SCLC but ultimately failed in the phase 3 TAHOE trial due to a combination of unacceptable toxicity and lack of survival benefit. In TAHOE, 64% of patients experienced grade 3 or higher adverse events, and a substantial number died while on treatment. The experience raised concerns that DLL3-targeted ADCs might inherently be too dangerous.
The SHR-4849 phase 1 data suggest otherwise. Among 100 patients evaluated for safety:
- 48% experienced grade 3 or higher adverse events, notably lower than the 64% seen with Rova-T
- The most common toxicities were hematologic (neutropenia, leukopenia, thrombocytopenia), with nausea and decreased appetite also observed but generally mild
- Treatment-emergent adverse events leading to drug discontinuation occurred in only 2% of patients, compared with about 7% with Rova-T
Importantly, the pattern of severe, often lethal toxicities that plagued Rova-T—such as pneumonia and malignancy-related complications—was not prominent with SHR-4849. Discussants at WCLC highlighted the role of payload and drug-to-antibody ratio (DAR) in this divergence.
Rova-T used a highly potent pyrrolobenzodiazepine (PBD) payload with a DAR of 4, whereas SHR-4849 employs a topoisomerase I inhibitor payload with a DAR of 8. Despite carrying twice as many payload molecules per antibody, SHR-4849 appears to have a more manageable safety profile—suggesting that the qualitative nature of the payload and its release kinetics may matter more than the sheer amount of drug attached.
What R-DXd and SHR-4849 share: “second-wave” targets and refined design
R-DXd and SHR-4849 illustrate several broader trends in the ADC field:
- Both focus on “second-wave” targets—CDH6 and DLL3—that are strongly expressed in specific tumors but relatively restricted in normal tissues.
- Both address indications with severe unmet need and limited effective options: platinum-resistant ovarian cancer and relapsed SCLC.
- Both rely on next-generation payloads, particularly topoisomerase I inhibitors and DXd-type designs, rather than the older, more toxic payloads that defined early ADCs.
For SCLC in particular, where immunotherapy has improved outcomes but not transformed the disease, the emergence of a DLL3 ADC with high ORR, strong intracranial activity and a manageable safety profile is potentially practice-changing—assuming that ongoing development confirms these early results.
Strategic and business implications: Merck and emerging players
From a strategic standpoint, R-DXd is one of three DXd ADCs that Merck believes could collectively generate multi-billion-dollar annual revenues for each company by the mid-2030s. As Keytruda approaches loss of exclusivity, these agents are designed to provide new growth pillars across solid tumors.
SHR-4849, meanwhile, highlights the emergence of China-based innovators as serious ADC players. If the program advances successfully, it will compete not only with future DLL3-targeted assets from Western pharmas but also with other modalities targeting SCLC and DLL3-positive neuroendocrine tumors. The SCLC market is relatively small in absolute terms, but its high unmet need and the possibility of label expansion make it strategically significant.
Implications for Japanese researchers and companies
For Japanese academia and industry, R-DXd and SHR-4849 underline two key points:
- First, the “second wave” of ADCs will be defined as much by target biology as by chemistry. Identifying targets like CDH6 and DLL3—where tumor-specific overexpression is clear and normal-tissue exposure is limited—requires strong pathology and omics capabilities, areas where Japanese institutions are already strong.
- Second, failed targets are not always dead targets. Rova-T’s failure did not condemn DLL3; it highlighted flaws in payload choice and overall ADC design. Revisiting such targets with improved payloads, DAR and dosing strategies can yield very different outcomes.
In practice, this means that Japanese players do not necessarily need to compete head-on in TROP2 or HER2, where the first wave is already crowded. Instead, they can create value by discovering and characterizing new second-wave targets, or by designing better ADCs against targets that others have prematurely abandoned.
My Thoughts
Looking at R-DXd and SHR-4849 side by side, I see them as powerful case studies in how ADC innovation is shifting from “what target can we hit?” to “how do we design the right payload and format for a given target and disease context?” DLL3 was very nearly written off as a failed target after Rova-T, yet SHR-4849 shows that with a more appropriate topoisomerase I payload and a re-optimized DAR, you can recover high efficacy with a much more acceptable safety profile. R-DXd, for its part, demonstrates how a validated payload-platform like DXd can be extended to new biology—CDH6—and new tumor types, while fitting neatly into Merck’s broader strategy to manage the Keytruda patent cliff. For Japanese innovators, the lesson is not to chase every hot target or modality, but to choose a few areas where they can combine strong biology, smart chemistry and thoughtful clinical design. That includes being willing to re-examine “failed” ideas from the past with new tools and better data. If approached in that spirit, the second wave of ADCs could offer significant opportunities not just for global giants, but also for focused, high-quality players who know exactly where they can add unique value.
This article has been edited by the Morningglorysciences team.
Related Articles
Comments