Written Description

The written description requirement and enablement are both housed in 35 U.S.C. § 112(a) but serve different purposes: WRITTEN DESCRIPTION TEXT: § 112(a) requires that 'the specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same'; ARIAD PHARMACEUTICALS v. ELI LILLY (Fed. Cir. 2010 en banc): the Federal Circuit sitting en banc confirmed that WRITTEN DESCRIPTION is a SEPARATE requirement from enablement; the majority held there are three distinct requirements in § 112(a): (1) WRITTEN DESCRIPTION — the inventor must describe what the invention IS; (2) ENABLEMENT — the specification must teach HOW to make and use it; (3) BEST MODE — the specification must disclose the best known way of practicing the invention; THE WRITTEN DESCRIPTION STANDARD: the standard is whether the disclosure 'reasonably conveys to those skilled in the art that the inventor had possession of the claimed subject matter as of the filing date'; it is assessed from the perspective of a POSITA at the time of filing; THE ARIAD DISSENT: Judge Rader (joined by others) dissented, arguing that written description serves the same function as enablement and that a separate written description doctrine was not supported by the statutory text; the majority was 6-5; despite the close vote, Ariad remains good law; WRITTEN DESCRIPTION vs. ENABLEMENT — THE KEY DIFFERENCE: ENABLEMENT asks: can a POSITA make and use the claimed invention from the specification?; WRITTEN DESCRIPTION asks: did the inventor describe WHAT the invention is — did the inventor have possession?; PRACTICAL IMPLICATION: a broad functional claim might be enabled (a POSITA could make embodiments within it) but lack written description (the specification doesn't show the inventor had possession of the FULL SCOPE of the claim); this is the key insight from Ariad — broad genus functional claims can fail written description even if some genus members are enabled; WHEN WRITTEN DESCRIPTION ARISES: most commonly: (1) continuation applications — new claims filed in a continuation must find written description support in the ORIGINAL specification; (2) broad genus claims — must show representative disclosure of the claimed genus; (3) functional claiming — claims defined by function rather than structure may lack written description if the specification doesn't correlate structure to the function.

Genus claims present the most challenging written description scenarios, especially in biotechnology: THE GENUS CHALLENGE: a genus claim covers a class of compounds, antibodies, nucleic acids, or other biological molecules that share a common feature; example (problematic): 'An antibody that binds to protein X and blocks its activity' — this covers ANY structure that performs this function; THE WRITTEN DESCRIPTION REQUIREMENT FOR GENERA: to describe a genus adequately, the specification must provide: (1) REPRESENTATIVE SPECIES: enough examples of genus members to show possession of the full genus; (2) STRUCTURAL FEATURES: identification of structural characteristics that are common to the genus members and that correlate with the claimed function; (3) BREADTH COVERAGE: not just one or two examples at one end of the genus, but representative coverage across the breadth of the claim; REPRESENTATIVE SPECIES PRINCIPLE: the inventor need not have actually made every genus member; but the specification must describe enough representative species that a POSITA would understand the inventor had possession of the GENUS (not just isolated species); if the inventor only made one or two species, the genus claim is typically unsupported; BLAZE MARKS: the Federal Circuit uses the 'blaze marks' metaphor — like a trail through a forest, the specification must provide blaze marks that guide a POSITA to the specific subset of compounds or biological molecules that constitute the genus; without blaze marks, the genus claim is like claiming ownership of a forest without identifying which trees; ARIAD APPLIED TO ANTIBODY CLAIMS: Amgen Inc. v. Sanofi (S.Ct. 2023): while primarily an enablement case, the Supreme Court's reasoning aligns with written description concerns for broad antibody genus claims; the Court held that Amgen's genus claims covering ALL antibodies that bind PCSK9 at the sweet spot were invalid — the specification disclosed only a limited number of representative antibodies and a functional description of what the antibodies should do; FUNCTIONAL DESCRIPTION IS NOT ENOUGH: describing the genus by what it DOES (function) rather than what it IS (structure) is generally insufficient for written description; example: 'a compound that inhibits HIV protease' — without structural description, this is only a functional claim; the inventor must show they had possession of a class of compounds that actually performs this function; REGULATORY SEQUENCE CLAIMING: in biotechnology, claims to specific nucleic acid sequences or protein sequences are well-supported by written description (actual sequences = structure); claims to CLASSES of sequences are harder to support.

The written description requirement plays a critical role in continuation and continuation-in-part (CIP) practice: CONTINUATION APPLICATIONS: a continuation application has the SAME specification as the parent; new claims in the continuation must be supported by the ORIGINAL specification's written description; CRITICAL RULE: claims added in a continuation that are NOT supported by the original specification's written description DO NOT get the parent application's priority date; they get only the continuation's filing date as their effective filing date; THIS MATTERS FOR: prior art: intervening prior art (between parent filing and continuation filing) can be used against new claims if the claims lack written description support in the original spec; validity: a continuation claim broader than what the original spec describes can be invalid for lack of written description even if the continuation is filed before any prior art; PRACTICAL EXAMPLE: parent spec filed in 2018 describes a blue laser diode using material A; continuation filed in 2020 adds claims to 'any laser diode' (not limited to material A); the broad claim is NOT supported by the 2018 spec; intervening prior art from 2019 can be used against the broad continuation claim; CIP APPLICATIONS: a CIP adds NEW MATTER to the specification; the new claims receive the CIP's filing date as their effective filing date ONLY IF the new claims rely on the new matter; claims that ARE fully supported by the ORIGINAL parent specification get the PARENT'S priority date; claims that REQUIRE the new matter get only the CIP's filing date; TRACING WRITTEN DESCRIPTION SUPPORT: patent counsel must trace every claim limitation back to the specification (original or CIP addition) to determine the correct priority date for each claim; priority claim analysis = written description claim-by-claim; NEW MATTER BARS: 35 U.S.C. § 132 prohibits adding new matter by amendment; § 112(a) written description ensures that broadened claims in continuations are grounded in the original disclosure; STRATEGIC USE: continuation practice allows applicants to add new claims as competitors enter the market; but ONLY claims that are adequately described in the original specification get the benefit of the original filing date.

The best mode requirement is the third requirement in § 112(a), distinct from written description and enablement: BEST MODE TEXT: § 112(a) requires the specification to set forth 'the best mode contemplated by the inventor or joint inventor of carrying out the invention'; SUBJECTIVE-OBJECTIVE TEST: best mode analysis has two prongs: (1) SUBJECTIVE: did the inventor contemplate a best mode at the time of filing? (subjective — looks at the inventor's mind); (2) OBJECTIVE: if so, was the specification adequate to enable a POSITA to practice the best mode? (objective — POSITA perspective); WHO MUST DISCLOSE: only the NAMED INVENTORS must disclose their best mode; the best mode of co-workers, collaborators, or licensees who are not named inventors need not be disclosed; AIA CHANGES TO BEST MODE: the Leahy-Smith America Invents Act (AIA) of 2011 eliminated best mode as a ground for PATENT INVALIDITY in litigation (35 U.S.C. § 282(b)(3)(A)); HOWEVER: best mode is STILL required in the patent application (§ 112(a) unchanged); violation of best mode can still result in REJECTION during prosecution; violation of best mode is still an issue for applications filed BEFORE the AIA effective date (March 16, 2013); RELATIONSHIP TO WRITTEN DESCRIPTION: written description asks whether the inventor described the invention broadly enough; best mode asks whether the inventor disclosed their PREFERRED WAY of practicing it; they overlap but are conceptually distinct; HOW BEST MODE AFFECTS PROSECUTION: an examiner can reject claims for failure to disclose best mode; this is rare because examiners typically cannot know what the inventor's best mode was unless it appears from the file wrapper or public information; INTERNATIONAL CONSIDERATIONS: many foreign patent systems do not have a best mode requirement; practitioners filing internationally from a US application should be aware that best mode disclosure in the US application may or may not be required or useful in foreign filings.

Effective prosecution strategy for written description requires careful specification drafting: STRUCTURAL vs. FUNCTIONAL DISCLOSURE: describe the invention both STRUCTURALLY and FUNCTIONALLY; structural disclosure is more robust for written description; if claims are defined by function, the specification must correlate specific structures to that function; REPRESENTATIVE SPECIES FOR GENUS CLAIMS: when claiming a genus, disclose multiple representative examples across the breadth of the genus; for chemical genus claims: at least 5-10 working examples with varying substituents; for antibody claims: multiple distinct antibody sequences that demonstrate the breadth of the claimed genus; for software genus claims: multiple algorithmic implementations; WORKING EXAMPLES vs. PROPHETIC EXAMPLES: actual working examples (reduction to practice data) carry more weight; prophetic examples (written in present tense or otherwise indicating they have not been performed) are acceptable but riskier; label prophetic examples clearly; BLAZE MARKS STRATEGY: provide explicit structural, chemical, or biological features that define the genus and distinguish genus members from non-members; give the POSITA 'blaze marks' — guideposts that allow them to determine whether a new compound falls within or outside the genus; FUNCTIONAL LANGUAGE RISKS: avoid claiming exclusively by function without structural support; if functional language is in claims, the specification must contain examples of specific structures that perform the function across the claim scope; GENUS-SPECIES HIERARCHY: use a layered disclosure: broad genus → subgenus → specific species; this supports both broad claims and fallback narrower claims; PRIORITY DATE PLANNING: for continuation practice, ensure the ORIGINAL specification contains written description support for all claims you might want to add in continuations; think about what competitors might design around before filing; SCOPE CALIBRATION: the scope of the genus claim should be commensurate with the scope of the written description; claiming an unlimited genus based on a handful of species at one end is a red flag.