Biotech Patent Strategy

Biotech faces specific § 101 eligibility hurdles not present in mechanical or software patents: NATURAL PHENOMENA RULE: Mayo Collaborative Services v. Prometheus (S.Ct. 2012): laws of nature (natural phenomena) are not patentable even if they are newly discovered; the 'inventive concept' beyond the natural law must be sufficient to transform the claim; IMPACT ON DIAGNOSTICS: diagnostic methods that observe natural correlations between a biomarker and a disease are at high eligibility risk; Ariosa Diagnostics v. Sequenom (Fed. Cir. 2015): method of detecting cell-free fetal DNA in maternal blood — ineligible because the correlation was natural; the conventional PCR amplification steps were not inventive; IMPACT ON GENE SEQUENCES: Myriad Genetics (S.Ct. 2013): isolated naturally occurring genomic DNA sequences are NOT patentable (products of nature); cDNA (complementary DNA, made from mRNA, excludes introns) IS patentable if it differs markedly from naturally occurring DNA; claim drafting now focuses on cDNA, specific modifications, or fusion proteins rather than isolated genomic sequences; PROSECUTION STRATEGIES FOR § 101: step 2A prong 1: if the claim does not recite a natural phenomenon (focus on the technical method steps, not the natural correlation), it may avoid the Mayo analysis; step 2A prong 2 (practical application): claim the diagnostic method as integrated into a treatment decision or treatment step; EXAMPLE: 'detecting X biomarker AND administering drug Y if detected above threshold Z' — the treatment integration may provide the practical application; step 2B (significantly more): conventional steps — PCR; ELISA; standard immunoassay — are NOT 'significantly more'; novel detection steps, specific reagent combinations with unexpected properties, or non-routine combinations are stronger; ANTIBODY CLAIMS: functional antibody claims ('an antibody that binds to epitope X with Kd < 1 nM') are generally § 101 eligible if structural rather than pure functional; specific sequence-defined antibodies are clearly eligible.

Amgen v. Sanofi (2023) fundamentally changed how antibody genus claims must be drafted and supported: THE CASE: Amgen v. Sanofi (S.Ct. 2023): Amgen's patents claimed all antibodies that bind to a specific region of PCSK9 protein AND block PCSK9 from binding to LDL receptors (functional definition of a genus of antibodies for cholesterol treatment); Amgen had identified 26 specific antibodies but claimed the entire genus functionally; HOLDING: § 112(a) ENABLEMENT requires that the full scope of a claim be enabled; for a claim covering potentially millions of antibodies in a genus, Amgen could not enable the full scope; the patent disclosed how to make 26 specific antibodies plus a 'conservative substitution' approach and a 'roadmap'; SCOTUS held: this 'roadmap' approach requiring extensive trial-and-error screening did not enable the genus as a whole; when a patentee claims an entire genus but enables only a 'portion' of it, the claim fails § 112(a); WANDS 8-FACTOR ANALYSIS: enablement is assessed under 8 factors (In re Wands): breadth of claims; nature of the invention; state of the prior art; level of ordinary skill; level of predictability; any working examples; quantity of experimentation necessary; presence of guidance in the specification; for antibody genus claims: breadth is HIGH (all antibodies meeting functional criteria); quantity of experimentation can be HIGH (enormous diversity of antibody space); this often fails the Wands analysis post-Amgen; HOW TO DRAFT ANTIBODY CLAIMS AFTER AMGEN: NARROW STRUCTURAL CLAIMS: sequence-defined antibodies (specific CDR sequences); multiple species claims (each specific antibody + its variants) rather than a single genus claim; FUNCTIONAL + STRUCTURAL HYBRID: antibody with specific CDR sequences PLUS binding specificity and affinity limitations; CONTINUATION STRATEGY: claim 26 specific antibodies in the initial application; use continuations to add claims to newly discovered antibodies as the program develops; EMBODIMENT EXAMPLES: the more working examples disclosed, the better the enablement — front-load the specification with diverse, structurally defined antibody examples.

The CRISPR patent landscape is one of the most significant IP conflicts in biotech history: THE PARTIES: Broad Institute (MIT/Harvard): Feng Zhang — claimed CRISPR-Cas9 in eukaryotic cells (relevant for human therapeutic applications); University of California Berkeley: Jennifer Doudna/Emmanuelle Charpentier — claimed CRISPR-Cas9 in general (prokaryotic + eukaryotic); THE DISPUTE: both parties filed patent applications in 2012-2013; interference proceeding at PTAB; PTAB ruled in favor of Broad (2017): Broad's claims to CRISPR in eukaryotes are patentable over Berkeley's claims to CRISPR generally; no interference-in-fact because eukaryotic CRISPR was not obvious from Berkeley's work; Federal Circuit affirmed (2018 and 2022); CURRENT STATE: Broad Institute holds key patents for CRISPR-Cas9 in eukaryotic cells (relevant for human therapeutics); UC Berkeley holds patents for CRISPR-Cas9 generally; BOTH patent estates are commercially licensed to multiple parties; neither party has an absolute monopoly; MAJOR LICENSEES: Editas Medicine (Broad license); Intellia Therapeutics (Broad + Berkeley licenses); CRISPR Therapeutics (Berkeley license); Beam Therapeutics; Prime Medicine; many others; LICENSING LANDSCAPE: therapeutic use: most CRISPR therapeutics companies have licenses from both estates or have sublicenses through their foundational investors; research use: Addgene distributes CRISPR reagents for academic use under MTAs; commercial agricultural/industrial applications: some agricultural applications may be outside the human therapeutic patents; FREEDOM TO OPERATE: CRISPR FTO is complex — requires analysis of patents from both estates, additional Cas variants (Cas12, Cas13, base editors, prime editors), delivery mechanism patents, and therapeutic target patents; FTO analysis essential before committing significant capital to CRISPR therapeutic programs; BEYOND Cas9: Cas12 (better for DNA detection); Cas13 (RNA targeting); base editing (David Liu/Broad); prime editing (Broad) — each has its own patent landscape.

Biologic patent strategy must be coordinated with regulatory data exclusivity to maximize protection periods: REGULATORY PATHWAY: Biologics are approved under the Biologics License Application (BLA) pathway under the Public Health Service Act (§ 351); biosimilars are approved via the abbreviated pathway (§ 351(k)); BIOLOGIC PRICE COMPETITION AND INNOVATION ACT (BPCIA) — 'PATENT DANCE': §§ 351(l): the biosimilar applicant must provide the reference product sponsor (RPS) with its biosimilar application and manufacturing information within 20 days of FDA acceptance; the RPS then identifies patents that could reasonably be asserted; the parties negotiate a list of patents for immediate litigation; remaining patents preserved for later infringement suits; REGULATORY EXCLUSIVITY PERIODS: 12-YEAR EXCLUSIVITY: a reference product (first approved biologic) receives 12 years of market exclusivity from first approval; NO biosimilar can be approved (or made effective) until 12 years after the reference product's BLA date; 4-YEAR DATA EXCLUSIVITY: FDA cannot even accept a biosimilar application until 4 years after BLA approval; PEDIATRIC EXCLUSIVITY: 6 months added to both periods if the innovator conducts qualifying pediatric studies; PATENT-EXCLUSIVITY INTERACTION: the patent term (20 years from filing) and regulatory exclusivity run concurrently; for a biologic where development takes 10-12 years, the remaining patent term at approval may be 8-10 years; the 12-year exclusivity may overlap substantially with the patent term; PATENT TERM EXTENSION (§ 156): biologics are eligible for PTE for up to 5 years of the regulatory review period; only ONE patent eligible per product; CONTINUATION STRATEGY: file continuation applications throughout development to capture evolving manufacturing processes, formulations, dosing regimens, patient selection biomarkers; these later-filed continuations (which have later filing dates and thus later expiration dates) can extend the effective patent cliff beyond the original composition-of-matter claims.

Biotech claim strategy requires a layered approach covering multiple claim types: COMPOSITION OF MATTER CLAIMS (strongest protection): specific nucleic acid sequences; polypeptides (proteins, antibodies) with defined sequences; isolated cells or cell lines; transformed organisms; novel chemical entities; vectors with specific constructs; CLAIM DRAFTING TIP: include multiple dependent claims at different levels of specificity — from broad genus to specific embodiment; METHOD CLAIMS: method of treating (composition + use = two layers of protection); method of manufacturing; method of diagnosis (watch for § 101); method of screening; CLAIM DRAFTING TIP: method claims should specify enough technical steps to avoid § 101 problems while remaining broad enough to cover competitor approaches; ANTIBODY CLAIM STRATEGY (post-Amgen): Tier 1: narrow structural claims (specific VH/VL sequences); Tier 2: CDR-defined claims (specific CDR 1-3 sequences in heavy and light chains); Tier 3: functional + structural hybrid; avoid pure functional genus claims that require extensive screening to make and use; PROVISIONAL STRATEGY: file provisional before ANY disclosure; include all embodiments, all data, all manufacturing variants; complete description of all antibodies discovered so far; as new data accumulates during the 12-month provisional period, file continuation-in-part OR file non-provisional with updated data; CONTINUATION STRATEGY: parent application: compound + composition; continuation 1: manufacturing process claims; continuation 2: formulation; continuation 3: dosing regimen and patient selection; continuation 4: new therapeutic indications discovered post-approval; FREEDOM-TO-OPERATE: conduct FTO before committing to development — biotech has expensive clinical programs; identifying blocking patents early allows negotiating licenses before having commercial leverage; DEFENSIVE PUBLICATION: for research tools and upstream platform technologies that are difficult to patent but should not be owned by competitors, publish first (creates prior art against competitors while preserving your own freedom to operate).