First-in-Class is the Only Class

Map-makers in newly opened territory leave a distinctive legacy. The first map’s lines do not become more accurate as the years pass, but they do become harder to redraw. Later cartographers, with better instruments and clearer sightlines, can survey the same ground more faithfully — but the official boundary lines, the ones that determine who owns what and where the disputes land, are drawn against the first map. The diagnostics version of this is folksier and just as accurate: the first product through the FDA sets the goalposts, and everyone after has to clear those goalposts and bring whatever else they have to offer.

Single-target therapeutic areas in biopharma now work the same way, and the patent record reads the difference between a first survey and a better one cleanly. KRAS-G12C is a good example. Amgen’s sotorasib (Lumakras) received FDA accelerated approval in May 2021, the first inhibitor of any KRAS variant in any indication and the end of a forty-year run during which KRAS was the textbook example of an “undruggable” oncogene. Mirati Therapeutics’ adagrasib (Krazati) followed in December 2022; Bristol-Myers Squibb acquired Mirati in October 2023 for approximately $5.8 billion. Revolution Medicines (Revmed) arrived later, with a chemically distinct entry point — pan-RAS macrocyclic inhibitors (daraxonrasib / RMC-6236) that act on the active, GTP-bound RAS conformation rather than the cysteine-trap mechanism that defined the first two launchers. The press-release version of the field has a tidy ranking: Amgen first, Mirati/BMS second, Revmed chasing.

The patent record disagrees. Read against the same six-dimension specification rubric this series has been describing — abstract specificity, claim-type strength, dependent fallback, prior-art neighborhood, Amgen v. Sanofi enablement, and survivability under Inter Partes Review (IPR) — the current full-text universe of biopharma KRAS / RAS-pathway patents and the top decile of that universe tell a different story. The first surveyor is no longer the one drawing the boundary lines. The second survey is the one most of the field is now being read against.

The KRAS / RAS-pathway biopharma universe carries 154 patents in the post-launch filing window; the top decile distills to fifteen. Of those fifteen, Revmed holds three, including ranks #1 and #2 — both pan-RAS macrocyclic composition-of-matter filings, with the rank #1 filing finishing the head-to-head pairwise tournament against the rest of the top decile at fourteen wins and zero losses. The rubric’s text-grounded justification for the rank #1 filing is the language a BD&L professional would hope to see attached to a foundational claim (but rarely does): direct pan-RAS / KRAS positioning, broad composition-of-matter scope, deep Markush hierarchy, strong enablement support. Foundational composition claims of that calibre are not what fast followers usually file. They are what first-in-class incumbents file. The chemistry, the specification depth, and the prosecution arc all read as a first survey — except the survey was undertaken by the company the conventional ranking puts third. In April 2026, Revmed reported the pivotal Phase 3 RASolute 302 readout in previously treated metastatic pancreatic ductal adenocarcinoma: median overall survival of 13.2 months on daraxonrasib versus 6.7 months on standard-of-care chemotherapy, with a hazard ratio of 0.40 in a disease whose five-year survival is roughly 3%. AACR data presented a week later in the first-line setting (47% monotherapy and 58% combination response rates, with 6-month overall survival above 80%) extended the Phase 3 result earlier into the disease course. Daraxonrasib already holds FDA Breakthrough Therapy Designation, Orphan Drug Designation, and a Commissioner’s National Priority Voucher. The clinical record has now begun to confirm what the patent record had structurally anticipated.

Mirati’s two top-decile patents split. The higher-ranked filing — a tetrahydropyridopyrimidine pan-KRAS platform at rank #3 — is structurally serious; the rubric reads it as commercially validated and highly enforceable, with a calibration haircut for being follow-on to the approved adagrasib franchise rather than disclosing genuinely new platform breadth. The other top-decile filing, at rank #15, is an adagrasib-plus-mTOR-inhibitor combination — useful lifecycle IP around an approved drug, but combination claims are not foundational claims, and the spec’s pairwise record (zero wins, fourteen losses) reflects that. Mirati also appears in the bottom decile, with a crystalline-form filing on the KRAS G12C inhibitor — a polymorph patent of the kind that characterizes a late-life-cycle position. The structural picture is mixed: one new platform claim, one combination, one polymorph. The post-acquisition filing posture looks closer to a defended franchise than to an expanding one.

Amgen, the first-in-class launcher, holds zero patents in the top decile and four in the bottom decile. Three of the four are intermediate-synthesis filings around the manufacturing chemistry of the sotorasib molecule; the fourth is a sotorasib dosing-regimen filing. The rubric reads all four as competent and useful — process IP, lifecycle IP, dosing-regimen IP — and as bottom-decile by the only metric that matters in a single-target field, which is foundational claim weight. The first surveyor is now filing the equivalent of property-line refinements to the original survey, not new surveys. That is what a first-in-class commercial position looks like on the patent record three to five years post-launch when the foundational composition claims are already in force and the rest of the patent strategy is dosing, formulation, and process. It is neither a failure nor an indictment. It is the mature filing posture of a company whose foundational position was established at the original launch and whose subsequent IP is securing rather than extending it.

The structural question for any BD&L professional looking at the field is which of these three pictures the next decade’s litigation, opposition, and IPR work will be drawn against. Amgen’s foundational sotorasib position is real and will not be displaced; the composition claim is in force and a generic challenge would not arrive on a meaningful timeline. But the next composition-of-matter claim that defines KRAS therapeutics — pan-RAS, pan-KRAS, G12D, G12V, or first-in-class on any non-G12C variant — is not going to come from Amgen’s current filing layer. That filing layer is dosing and process. The pan-RAS composition layer is being written by Revmed. The pan-KRAS platform layer is being contested between Mirati’s tetrahydropyridopyrimidine filing and several adjacent commercial filers — Treeline, Astellas, Genentech/Roche, Incyte, Alterome, Ranok — none of whom has a marketed asset and several of whom are filing claims that, if granted, will define the freedom-to-operate (FTO) landscape any subsequent entrant must clear.

This is what filing date, claim distinctness, and FTO margin look like as structural determinants in a single-target field. Filing date got Amgen to market first. It did not give Amgen the foundational position over the next mechanism class within the same target. Claim distinctness is what gave Revmed that position — by entering with active-state pan-RAS inhibition rather than cysteine-trap G12C inhibition, and prosecuting composition-of-matter claims that do not collide with the first surveyors’ lines. The BMS acquisition gave the Mirati franchise commercial reach but did not, by itself, refresh the foundational filing layer; the post-acquisition top-decile filings are platform extension and combination, not new foundational claims. The “fast follower with method-of-use claim and an enablement problem” version of Revmed that the conventional read of the field would predict is not what the patent record shows. The patent record shows Revmed as the first-in-class filer for the next era of KRAS therapeutics, with Amgen and Mirati / BMS as the in-class incumbents on the prior generation.

This reading does not, on its own, settle the commercial questions. BMS will extend the Mirati franchise on the strength of an approved drug, an established prescriber base, and a balance sheet that no fast follower can match. Amgen retains the foundational sotorasib position and the manufacturing footprint that comes with three years on market. Revmed’s commercial trajectory in pancreatic cancer still depends on regulatory submission and review, payer access, formulary work, manufacturing scale-up, and label progression across additional RAS-driven indications. What the patent record arbitrates — and what the Phase 3 RASolute 302 result has now corroborated clinically — is which company’s claims will define the boundary lines that the next decade’s KRAS deals, partnering, in-licensing, FTO, opposition, and IPR work will be drawn against.

A BD&L professional preparing a comparable-deals analysis for a KRAS-adjacent asset, or a partner-of-choice analysis for a fast follower entering the field, has to ask the question the press-release version of the field cannot answer. Whose survey are the boundary lines drawn against now? In KRAS-G12C, three years after first-in-class approval, the answer is not the first-in-class company. The first survey is the survey everything later is reconciled to — but only until the second survey is the one the courts, the patent offices, and the deal counterparties decide to use.

---

This is the fourth of five articles on what biopharma BD&L practitioners can read from the patent record now that the FOIA-disclosed contract record has ended. Here’s the start of the series .

The next article looks at the prosecution and defense work the patent record itself preserves — opposition, Inter Partes Review, third-party patent acquisition, multi-jurisdictional family management — as the observable signal of whether a partner is one whose IP capability matches the deals it is signing.