The Deals of THIS Decade, Read From the IP

In April 2025, I closed an earlier series about biopharma BD&L best practices with an article called “Deals of the Decades”. Perhaps rashly, I ended with a prediction. Although the 2020s were not yet half spent, it already seemed to me that antibody-drug conjugates (ADCs) — including bi-specifics and tri-specifics — would take top honors as the most successful drug treatment modality of the current decade. The specific deals I pointed to were the Genmab–AbbVie alliance, the first of the mega-ADC deals of the 2020s, plus two acquisitions: Immunomedics (acquired by Gilead for $21 billion in 2020) and Seagen (acquired by Pfizer for $43 billion in 2023).

Fourteen months later, it’s doubtful even Kalshi would dispute this prediction. Pfizer announced a $10.5 billion ADC alliance with Innovent this May. Also in May, BMS and Hengrui launched a $15.2 billion bidirectional co-development alliance for early-stage ADC and bispecific molecules from each company’s pipeline. Regeneron expanded its conditional-bispecific platform collaboration with CytomX at the beginning of this month, lifting total potential deal value across the expanded partnership to approximately $4 billion. AstraZeneca’s continuing extension of the Daiichi Enhertu franchise into earlier-line breast cancer and gastric indications has reinforced the foundational ADC franchise that their 2019 alliance produced. And as this article was being prepared, GSK announced its $10.6 billion acquisition of Nuvalent — three NSCLC programs anchored by the late-stage ROS1 inhibitor zidesamtinib and the ALK inhibitor neladalkib — with the deal explicitly framed as a platform for expansion alongside Ris-Rez, GSK’s B7-H3 ADC in Phase 3 development. The pairing logic — kinase commercial infrastructure acquired alongside an ADC platform readying for launch — is exactly the deal pattern the structural-IP layer underneath this cohort points toward. Across the ADC, bispecific, and tri-specific structural categories, deal-flow validation that was emergent in April 2025 is, as of mid-2026, substantially complete.

What this ADC deal flow does not tell the BD&L professional, however, is how and why a particular valuation attaches to each particular deal. To uncover that layer, one needs to look at specific IP filings within the broad ADC modality space to see which elements are structurally foundational and which are participation-level IP. For example, the Pfizer-Innovent alliance covers a portfolio of ADC programs, but the public deal announcement names none of the underlying patents. The BMS-Hengrui partnership transfers IP and know-how bidirectionally, but the directional question of which IP layers are durable enough to justify the partnership terms is one the trade press can’t answer. The Regeneron-CytomX expansion adds two newly-selected targets to a 2022 base agreement, but the structural-IP question of which targets are foundational and which are exploratory is invisible from the press release. The key value-driving question is one that BD&L professionals have to answer for themselves, against the patent record, and the answer depends on the structural rigor of the underlying IP claims in each instance.

Two target landscapes are useful for reading what the structural-IP layer underneath the successful run of ADC deals looks like in mid-2026. HER3 and TROP2 are both ADC-dominant target classes with substantial competitive depth. Both have foundational filings in the late 2010s, and both have surfaced as anchor targets in the recent deal flow. Measured against the same six-component KC scoring framework that anchored the GLP-1 backtest, the HER3 IP landscape (465 patents in the IP portfolio) produces a top decile dominated by an unusual leading filer: a Juno Therapeutics / Editas Medicine co-assigned CAR-T platform patent — US20220184131A1, ensemble score 4.453 — that reads as the structurally most rigorous filing in the cohort, despite its not being an ADC at all. The top-decile composition behind it includes four Janssen filings (calibrated rank 2, average ensemble 4.279), Yale University’s foundational rank-5 filing, three Genmab filings, two Daiichi continuing-prosecution filings, and four Merus N.V. bispecific filings. The Daiichi and Janssen layers are both lifecycle-extension foundations of the post-Enhertu ADC franchise; the Juno-Editas, Genmab, and Merus layers are the structurally-distinct alternatives that the “deals of THIS decade” prediction implies will be the current decade’s competitive frontier.

The TROP2 reading is structurally different and analytically informative. The V3 top decile is dominated by six Kisoji Biotechnology filings — a Quebec-based biotech whose binding-agent IP scores at ensemble 4.525 across six related family publications, occupying tournament-calibrated ranks 1 through 6 — followed by the Immunomedics (post-Gilead) inheritance layer (three top-decile filings on sacituzumab govitecan / IMMU-132 family work), the Boehringer Ingelheim tri-specific binding-molecule layer (two top-decile filings at ensemble 4.24), and Daiichi’s continuing TROP2 ADC prosecution. The pattern that emerges in both landscapes — and which is invisible from the deal flow alone — is that the foundational-IP layer is no longer concentrated in the major pharma incumbents. Kisoji is a non-US, non-major-pharma originator whose binding-agent IP outscores the Immunomedics franchise Gilead acquired for $21 billion. The Juno-Editas CAR-T platform IP outscores most of the major-pharma ADC continuation prosecution in HER3. The structural-IP layer underneath the current cohort reveals where the next deals will originate, not just where the next deals will land.

One observation worth holding onto specifically. The top-decile composition in both HER3 and TROP2 cohorts includes filings from originators that are neither Western major pharma nor Chinese biopharma — Kisoji (Quebec), Hummingbird Bioscience (Singapore), Adagene (Singapore-anchored ADC and bispecific platform), Alligator Bioscience (Sweden), and Antikor Biopharma (UK). The post-2020 structurally-foundational ADC and bispecific IP layer is being filed by a wider geographic cohort than the major-pharma deal flow alone suggests, and the BD&L professional who is valuing a partnership against an ADC-relevant target needs to read the structural-IP layer to know which mid-tier and emerging biotechs are positioned to lead the next round of large licensing deals. The CytomX-Regeneron pattern — small-biotech conditional-platform IP licensed to major pharma — is the inverse of the major-pharma franchise-acquisition pattern. Both patterns will continue to drive the ADC deal flow through the back half of the decade.

The six-component KC scoring rubric, the tournament-calibrated top-decile rankings for HER3 (465 patents) and TROP2 (347 patents), the per-patent sub-dimension breakdowns, and the methodology documentation — including the prior-art-landscape anchor language calibrated specifically for the post-2020 ADC, bispecific, and CAR-T modality competition — are written up in full at rDNA.ai, along with the back-tested top deciles and the underlying scoring traces. Applied here, the structural-IP analysis converts my mid-decade prediction from a directional observation into an actionable read of where, within the validated ADC modality classes, the structurally foundational IP actually sits — which is the diligence layer the BD&L professional needs to bring to the deal negotiation.

Hail-Mary predictions made mid-decade are rarely confirmed by deal flow fourteen months later. Perhaps my April 2025 prediction was less of a “climb out on a limb” than the latest of a long-running series of best-practices observations. What has happened since — the Pfizer-Innovent, CytomX-Regeneron, BMS-Hengrui, and AstraZeneca-Daiichi continuing-extension deal flow — has helped my “limb-climbing” transform into an analytical baseline. The next key prediction, whatever shape biopharma deal flow takes for the 2030s, will be observable in recent foundational IP and either confirmed or refuted by the particular deals that follow. For now, my ADC + bispecific + tri-specific “deals of THIS decade” prediction stands — the structural-IP layer underneath it is readable, and the BD&L professional doing diligence on the next round of mid-tier and emerging-biotech partnerships now has a methodology that reads the foundational-IP question independently of the press-release framing.

---

---

This is the ninth article in the rDNA.ai biopharma BD&L series and closes this first introduction to the post-FOIA BD&L toolkit. These past four articles have read the V3 KC scoring framework against five representative biopharma IP landscapes — GLP-1, CLDN18.2, TYK2, HER3, and TROP2 — and against the surrounding deal flow that these IP portfolios supported. The next layer of the toolkit — full prosecution-history reading at the dependent-claim and amendment level, integrated against the deal-IP linkage — is in development. The series will resume when that work is ready to share.

Here’s the start of the series .