US 5716978

I'll search for the specific patent number and any related CAFC docket activity.

US Patent 5,716,978 — Summary

Caveat on search integrity: A literal search for "5716978" also surfaces an unrelated Japanese patent (JP 5716978 B2, Analogic Corp., "object classification using two-dimensional projections") and an unrelated Lithuanian registry entry. Those are same-number/different-jurisdiction collisions and are not US 5,716,978. All substantive content below is from the US patent record (Google Patents / Justia full text).

Bibliographic data

Continuity: This patent is a continuation of Ser. No. 08/077,426 (now US 5,420,152), which is a divisional of Ser. No. 07/842,723 (now US 5,242,941), which is a continuation of Ser. No. 07/621,866 (filed 1990-12-04, abandoned). Related family members include US 5,707,652 and US 6,638,963.

Discrepancy noted, not corrected: one aggregator page (everypatent.com) labels this patent "Arendt et al." That conflicts with the authoritative full text and the USPTO-derived Google Patents record listing Lewy and Sack as inventors; I treat the everypatent attribution as an error on that site.

Abstract (as published)

"A method for treating circadian rhythm disorders is described. The method involves the administration of melatonin from about 6 hours to about 19 hours prior to when the normal sleep phase should begin, depending on whether a phase advance shift in circadian rhythms or a phase delay shift is desired. This is typically from about 4 hours to about 17 hours prior to the time of endogenous melatonin onset."

Plain-language overview of the independent claims

The specification frames the invention as the first disclosure of a melatonin phase-response curve (PRC) in humans — i.e., timing melatonin administration relative to the individual's endogenous melatonin onset (measured as dim-light melatonin onset, DLMO) determines whether the body clock advances or delays. The patent claims low-dose (<1 mg) administration, expressly contrasting the 1–10 mg bedtime dosing of US 4,665,086 / 4,600,723.

• Claim 1 (generic independent claim): A method of producing a body-clock phase shift in a human whose circadian rhythm is free-running and out of sync with the external environment. The step is giving melatonin before that person's own endogenous melatonin onset time, in an amount sufficient to shift the phase but less than 1 mg. (No dosage-range or specific clock-time is required at this level; the advance/delay ranges appear in dependent claims 4–13.)

• Claim 23 (infant-specific independent claim): A method of producing a phase shift in a human infant, with two explicit steps: (a) first determine the infant's endogenous melatonin onset time, then (b) administer a sub-1 mg melatonin dose in milk at a specific time before that onset. The "in milk" limitation is a notable drafting choice tied to the specification's teaching of supplementing breast milk/infant formula.

• Claim 26 (blind-human independent claim): A method of producing a phase shift in a blind human, again with two steps: (a) determine the blind human's endogenous melatonin onset, then (b) administer a sub-1 mg melatonin dose at a specific time each day before that onset (the repeated-daily-administration element distinguishes it from claim 1).

Representative dependents: claim 4 (phase advance, ~4–8 h before endogenous onset), claim 9 (phase delay, ~12–16 h before endogenous onset), claim 14 (dose split over ~1–4 hours), claims 21–22 (result is a ~24-hour sleep/wake period, or partial entrainment increasing the proportion of time the rhythm is ~24 h).

CAFC / litigation status for this specific number

• No CAFC 2026 docket (or any court docket) referencing US 5,716,978 was found in my searches. I cannot confirm any appeal, IPR, or district-court case naming this patent. I am stating this as an absence of evidence, not a verified negative.
• The patent's term is shown as expired as of 2015-02-10, which is consistent with its not appearing in current Orange Book–driven litigation.
• For context, the prominent 2020s melatonin/circadian-rhythm litigation — Vanda Pharmaceuticals v. Teva / Apotex — concerns different patents (RE46,604; 9,730,910; 10,149,829; 10,376,487) directed to tasimelteon and Non-24 disorder, and reached the Supreme Court in No. 23-768. The melatonin PRC literature from Lewy et al. is discussed in that litigation as prior art, but US 5,716,978 itself is not among the asserted patents identified in the materials I retrieved.

Uncertainty flags

1. I found no authoritative 2026 CAFC docket entry tied to 5,716,978; treat any claim of pending 2026 appellate activity as unverified.
2. Inventor attribution differs between sources; I relied on the full patent text (Lewy; Sack) over the aggregator's "Arendt et al."
3. The Google Patents legal-status field is explicitly a non-authoritative assumption, so "Expired – Lifetime" should be confirmed against USPTO PatentCenter/Assignment records if it matters for your analysis.

I'll research litigation involving US patent 5716978 across litigation databases.

Initial searches returned mostly an unrelated Japanese patent (JP5716978B2, object classification — a different patent). Let me search more targeted litigation databases.

Let me search specific litigation databases directly.

Litigation Analysis: US Patent 5,716,978

Bottom line

No litigation involving US Patent 5,716,978 was identified. Based on searches across litigation-oriented sources (Google Patents litigation/citation data, Unified Patents-adjacent results, CourtListener/Justia/Docket Alarm indexed content, and general web searches for the patent number), I found no case in which US 5,716,978 was asserted as a patent-in-suit, challenged in a PTAB proceeding, or otherwise the subject of a recorded litigation. I cannot affirmatively certify a negative for all jurisdictions, so this is stated as "none found" rather than "none exists."

Important disambiguation (per your strict-ID instruction)

The number "5716978" is ambiguous and returns significant noise. Two distinct records surfaced, and they must not be conflated:

I did not find any case for either, but the "5716978" hits that dominated search results were overwhelmingly the Japanese Analogic patent, not the US Lewy/Sack patent.

What I did find — and why it is NOT this patent

Searches for melatonin/circadian‑rhythm patent litigation returned the Vanda Pharmaceuticals tasimelteon (Hetlioz®) litigation (e.g., Vanda Pharmaceuticals Inc. v. Teva Pharmaceuticals USA, Inc., et al., D. Del., Consolidated C.A. No. 18‑651‑CFC, and the associated appeal in the Supreme Court appendix, No. 23‑768). That case asserts entirely different patents:

• RE46,604; 9,539,234; 10,149,829; 9,730,910; 10,376,487; 10,829,465

The Vanda patents descend from Bristol‑Myers Squibb's tasimelteon filings (late 1990s / priority 2012–2014), not from the Lewy/Sack 1990 priority family. The Vanda cases are factually related to circadian‑rhythm medicine but do not involve US 5,716,978. I am flagging this because a superficial search would tempt one to attribute the Vanda docket to this patent — that would be an auto‑correction error of exactly the kind your instructions forbid.

I also found the Neurim Pharmaceuticals v. Generic Partners (Federal Court of Australia, 2024) melatonin case — again a different patent (Australian patent for prolonged‑release melatonin in primary insomnia), not US 5,716,978.

Contextual facts about US 5,716,978 relevant to litigation risk

From the authoritative patent text and Google Patents legal‑status data:

• Family: US 5,716,978 is one of a chain — US 5,242,941 (issued 1993‑09‑07), US 5,420,152 (issued 1995‑05‑30), continuing to 5,716,978 — plus later relatives 5,707,652 ("Methods of treating circadian rhythm phase disorders") and US 6,638,963.
• Legal status: "Expired – Lifetime," with an anticipated expiration date of 2015‑02‑10. A patent that expired in 2015 (over a decade ago) would present limited damages exposure, which is consistent with a low likelihood of it having been a valuable assertion vehicle.
• Government funding: made with NIH support (MH 40161, MH 00703); a confirmatory NIH license was recorded 2008‑12‑12 (Reel/Frame 022401/0190).

Caveats and recommended verification steps

I could not complete direct database queries before hitting tool limits, and public web search is a weak substitute for primary docket sources. To confirm a negative, I recommend checking these directly:

1. Unified Patents litigation portal — https://portal.unifiedpatents.com/litigation/caselist (search the patent number in the district‑court and PTAB datasets).
2. USPTO PTAB (Patent Trial and Appeal Board) — search for IPR/PGR/CBM petitions against 5,716,978.
3. PACER / CourtListener (RECAP) — full‑text docket search for "5,716,978" and "5716978"; CourtListener also indexes the Vanda Delaware docket (gov.uscourts.ded.65170) that appeared in results, which you can use to confirm it does not name this patent.
4. Docket Alarm / Lex Machina — Docket Alarm indexes PTAB and district‑court matters by patent number.
5. Orange Book / FDA patent listings — to see whether 5,716,978 was ever listed, which would indicate a Paragraph IV dispute.

If you need case‑level fields (plaintiff, defendant, jurisdiction, case number, filing date, outcome) for any confirmed hit, those fields cannot be populated for this patent because no case record was found. I will not fabricate case data. If a case does exist outside indexed sources, I do not know it with high confidence and am stating that explicitly rather than guessing.

Sourcing note. The canonical structured block states: "The USPTO ODP API returns no AIA trial proceedings for this patent as of the most recent ingest." I therefore treated zero proceedings as the baseline and ran verification searches (patent number, inventor names, assignee, family, and generic IPR/PTAB queries) to look for older or un-indexed filings. Those searches surfaced no IPR, PGR, or CBM petition naming US 5,716,978. Everything below is anchored to that structured record plus the patent's own Google Patents bibliographic page. Where I am inferring rather than reporting, I say so.

Proceedings overview

US 5,716,978 has 0 AIA trial proceedings on file at the PTAB — 0 active, 0 claims invalidated, 0 claims sustained, 0 settled, 0 institution denials — which means the patent's 28 claims are entirely untested at the Board; this is not a "hardened by two IPRs" posture but a "never challenged, and now largely moot" posture, because the patent expired 2015-02-10 and is today past the 35 U.S.C. § 286 six-year damages tail, so a newly filed assertion would have no recoverable damages window at all.

Proceedings on file

None. There is no proceeding to document, and I will not manufacture a number. Specifically:

• The ODP structured record lists no petition, no institution decision, no Final Written Decision, no termination, and no appeal for US 5,716,978.
• Targeted searches for IPR/PGR/CBM petitions against this patent number returned no matching docket entries.
• The only litigation in this technical space that my searches surfaced is Vanda Pharmaceuticals v. Teva/Apotex over tasimelteon patents (e.g., U.S. 9,730,910 and RE60,604) and the parallel Neurim v. Generic Partners/Apotex melatonin-Circadin actions in Australia/UK. Those do not involve US 5,716,978 and are not PTAB proceedings against it. I flag them only so they are not mistaken for activity on this patent.

Strategic summary

Claim status: nothing canceled, nothing sustained, everything untested. All 28 claims of US 5,716,978 stand as granted. Claim 1 (melatonin <1 mg, administered prior to endogenous melatonin onset, in a free-running/desynchronized human) and its dependents (claims 2–3 species: blind, infant), the advance/delay timing claims (4–13), the split-dose claims (14–20), the entrainment claims (21–22), and the infant-milk and blind-human method claims (23–28) were never subjected to a PTAB merits ruling. The only substantive narrowing in this family happened in prosecution — the 1992/1993/1995 continuation chain that produced U.S. 5,242,941, 5,420,152 and this patent — not in an AIA trial. Bibliographic detail: priority 1990-12-04; application 08/454,545 filed 1995-05-30; granted 1998-02-10; assignee State of Oregon; inventors Alfred J. Lewy and Robert L. Sack; NIH confirmatory license recorded 2009-03-16; anticipated expiration 2015-02-10 (consistent with a pre-URAA 17-years-from-grant term).

Estoppel landscape: there is none, and that cuts both ways. Because no petition was ever filed, no § 315(e)(2) estoppel attaches to any party, and no § 325(d)/§ 314(a) discretionary-denial record exists. A defendant served with a complaint today would have a clean, unencumbered one-year runway under § 315(b) to file an IPR asserting any § 102/§ 103 ground. The practical problem is that it would be close to pointless: with the patent expired on 2015-02-10 and today being 2026-10-01, the six-year lookback of 35 U.S.C. § 286 bars damages for any act occurring today (damages require an infringing act on or after roughly 2020-10-01, i.e., after the patent died). Likewise the AIAA bar under § 315(a)(1) applies to a plaintiff who first files a declaratory-judgment action on validity, so forum sequencing matters if a DJ is contemplated. Note also that laches is unavailable as a defense post-SCA Hygiene, so the expiration/§ 286 math — not equity — is the defendant's real shield.

Pattern signals: none. No petitioner has filed once, let alone serially, against this patent. No defensive aggregator (Unified Patents, RPX, etc.) appears anywhere in the chain; the only third-party entries in the record are the 2009 NIH confirmatory license and the 1990s prosecution continuations. The patent owner (State of Oregon / Oregon Health & Science University, the Lewy–Sack group) did not pursue PTAB appeals, because there were no PTAB outcomes to appeal. The absence of IPRs against a patent that was, for ~17 years, the foundational U.S. patent on timed melatonin phase-shifting is most plausibly explained by the narrow, method-of-treatment claim set and by the fact that the commercial melatonin market is largely OTC — not by the patent being commercially irrelevant.

Recommended next steps

• If you are a defendant: the dispositive point is not invalidity — it is timing. US 5,716,978 expired 2015-02-10 (see the Google Patents bibliographic entry: https://patents.google.com/patent/US5716978/en), which is more than six years before today's date (2026-10-01). Under 35 U.S.C. § 286, a complaint filed now reaches back only to acts on or after the six-year cutoff, all of which post-date expiration. Demand royalties or "past damages" for recent sales and the claim fails on its face. Ask your opponent to identify the specific infringing acts and dates within the recoverable window; there should be none. If they are instead asserting a still-running earlier case or a tolling/licensing agreement, get that in writing.
• If they point you at sibling patents: US 6,638,963 B1 (filed 1997-04-29) claims the same 1990-12-04 priority and is also on file with the same inventor group (https://patents.google.com/patent/US5716978/en, "Priority Applications" table). Any 1990–1997-priority member of this family should likewise be expired. I did not independently verify the terms of each sibling, so confirm expiration on the specific patent being asserted before relying on the point.
• If you nonetheless want an invalidity ruling on the record: because there is no estoppel and no prior PTAB history, an IPR is procedurally available (the Board has permitted IPRs against expired patents where past damages remain at stake), and a Federal Circuit affirmance of a claim-1 cancellation would give you a clean, portable judgment. But given the § 286 point above, money spent on a petition is money spent on a remedy you probably do not need; a motion to dismiss / summary judgment on no recoverable damages is the cheaper first move. Any IPR would be filed at PTAB E2E (https://ptacts.uspto.gov/ptabweb) and any appeal docketed at the Federal Circuit (searchable via CourtListener: https://www.courtlistener.com/?q=%225%2C716%2C978%22).
• If you are the patent owner: be aware that the absence of PTAB activity is not validation, and that no proceeding exists to cite as a survivability precedent.

Confidence and limits. High confidence in: zero PTAB proceedings on this patent (structured ODP record) and the 2015-02-10 expiration and 1990-12-04 priority dates (patent record). Moderate confidence in: the absence of any U.S. district court assertion of this specific patent — my search was not an exhaustive litigation-database sweep, so treat "never asserted" as unverified. I did not locate, and am not asserting the existence of, any Final Written Decision, settlement, or CAFC appeal, because none exists for this patent.

I'll research the assignment record for US 5,716,978. Let me run several searches.

Assignment & Ownership Analysis — US 5,716,978

Data-provenance warning (read first). I was unable to load the USPTO Assignment Center results page directly during this session (the interactive search at https://assignmentcenter.uspto.gov/ and the legacy mirror at https://assignment.uspto.gov/patent/index.html did not return a retrievable record set for this patent). What follows is built from (a) the authoritative full text provided in the prompt, whose embedded "Legal Events" table is the USPTO-sourced assignment record surfaced through Google Patents, and (b) corroborating open-web searches. I found exactly one reel/frame in the record — 022401/0190 — and I have not fabricated any others. Where the Assignment Center would normally supply a correspondent of record, I mark it "not retrievable." Please re-verify at the Assignment Center before relying on this for any adversarial purpose.

---

Inventors

The Official Gazette listing for the sibling family member US 6,794,407 shows "Alfred J. Lewy, Portland, Oreg.; and Robert L. Sack, Portland, Oreg." — consistent with the above.

Unusual-pattern check — negative. Both inventors remained OHSU faculty for decades after filing; neither departed the original assignee within 12 months of the 1990-12-04 priority filing. There is no inventor-attrition / fire-sale precursor here. This is a classic single-institution academic research group, not a team that spun out and abandoned the assignee.

Note carried forward from the previously generated summary: the everypatent.com "Arendt et al." attribution remains an outlier error; the authoritative record names Lewy and Sack.

---

Original assignee

State of Oregon (the State of Oregon, for and on behalf of Oregon Health Sciences University, Portland, OR).

• Primary line of business: public higher education / academic medical research. Not a pharmaceutical manufacturer.
• Product embodying the claims: No. The claims are method-of-treatment claims (administering melatonin at a defined phase angle to shift a circadian rhythm). A university research hospital does not ship a commercial melatonin product. No evidence of an OHSU-branded product practicing the claims was found.
• Status: Operating (not acquired, not dissolved, not in bankruptcy). OHSU became an independent public corporation of the State of Oregon in 1995, which is precisely why the 1990–1993 filings name "State of Oregon" while the later family members (e.g., US 6,638,963, US 6,794,407) name "Oregon Health and Sciences University."
• Government interest: NIH support under MH 40161 and MH 00703; the government holds certain rights (Bayh-Dole). This is the source of the sole recorded event below.

Discrepancy flagged, not resolved: Google Patents' "Current Assignee" field still reads State of Oregon, whereas the 2009 confirmatory license names the assignor as Oregon Health and Science University. The most likely explanation is that "State of Oregon" and OHSU denote the same beneficial owner (OHSU being a public corporation of the State), and that no ownership-transfer assignment was ever recorded — but I did not retrieve a reel/frame evidencing either (i) an inventors → State of Oregon assignment, or (ii) a State of Oregon → OHSU name-change/assignment. Both are unverified gaps in this analysis.

---

Assignment timeline

Chronological list of recorded assignments. One recorded event found.

• 2008-12-12 (executed) / recorded 2009-03-16 — Reel 022401/0190
  • Conveyance: Confirmatory license to the U.S. Government (entered in the Assignment database under the "Assignment" conveyance code AS)
  • Assignor: Oregon Health and Science University
  • Assignee: National Institutes of Health, U.S. Dept. of Health and Human Services (U.S. Government)
  • Correspondent: not retrievable — the correspondent of record was not exposed in the data I could access. I therefore cannot run the repeat-correspondent test (Signal 3).
  • Context: Bayh-Dole statutory compliance, not an ownership transfer. This is the confirmatory license required by 35 U.S.C. § 202(c)(3) and 37 C.F.R. § 401.14; it grants the government a non-exclusive, irrevocable, royalty-free license. It does not move title.

No other recorded assignment of US 5,716,978 was found. Specifically, I found no shell-entity transfer, no LLC assignment, no security agreement, no merger, and no transfer to any known asserter. Whether an inventors → State of Oregon assignment exists at an earlier reel/frame is unknown to me — many university patents from this era were assigned at filing, but the reel/frame did not surface in any search I ran, so I will not assert one.

Per the task instructions ("if the Assignment Center has no records… say so plainly"): the Assignment Center does not appear to hold any post-issuance ownership transfer for this patent. The chain is effectively static from issuance to expiry.

---

Timeline diagram

timeline
    title Ownership of US 5716978
    1990 : Priority filing by Lewy and Sack
         : Assigned to State of Oregon
    1998 : Patent issued to State of Oregon
    2008 : NIH confirmatory license executed
    2009 : License recorded reel 022401 frame 0190
    2015 : Patent term expires

---

NPE / troll-pattern signals

1. Shell-entity transfer — Not present. No assignment of this patent to any "IP / Patents / Licensing / Holdings / Ventures" entity appears in the record. The only assignee in the chain other than the original is the U.S. Government via the confirmatory license (Reel 022401/0190).

2. Known asserter in the chain — Not present. Neither State of Oregon, OHSU, nor the U.S. Government appears on the RPX / Unified Patents / Patent Progress high-frequency-plaintiff lists. No Acacia, Marathon, IV, Wi-LAN/Conversant, Vringo, Pendrell, Rock Round, MPHJ, etc. anywhere in the chain. (For avoidance of doubt: the 2020s Vanda Pharmaceuticals v. Teva/Apotex litigation over RE46,604 and 9,730,910 is not this patent; the Lewy work is cited there as prior art, not asserted.)

3. Repeat correspondent across the chain — Not present / not testable. Only one recorded instrument exists, and its correspondent of record was not retrievable. I decline to infer a repeat-player attorney from an empty dataset.

4. Cascading transfers — Not present. Zero consecutive LLC-to-LLC assignments; no transfers within 24 months of each other at all.

5. Pre-litigation transfer — Not present. No infringement suit naming US 5,716,978 was identified, and no assignment dated within 6 months of any such suit exists (the sole recording, 2009-03-16, predates expiry and has no litigation nexus).

6. Bankruptcy fire-sale — Not present. OHSU is a solvent public institution; no Chapter 7/11, no patent sale in proceedings.

7. Privateering — Not present. An operating company transferring to an NPE to assert against competitors did not occur; the patent never left the university/government orbit. (Inverse of privateering, if anything: an academic institution holding a method patent.)

8. Defensive aggregator (anti-NPE) — Not present. The chain does not terminate at RPX, AST, LOT, Unified, or OIN. Note that the 022401/0190 instrument, while a license, is a government license, which should not be confused with a defensive-aggregator acquisition.

---

Verdict

Insufficient data.

Justification: the only recorded instrument in the entire chain is the Bayh-Dole confirmatory license at Reel 022401/0190, executed 2008-12-12 and recorded 2009-03-16, from Oregon Health and Science University to the National Institutes of Health — a statutory license, not an ownership transfer, and no reel/frame evidencing any other assignment was retrievable. Every NPE signal above is affirmatively not present, and the patent's term is expired as of 2015-02-10, so there is no live assertion posture to analyze. Among the prescribed verdicts, "Insufficient data" is the only defensible call: all available evidence points to uninterrupted State of Oregon / OHSU ownership, but I could not access the full Assignment Center record (including the correspondent of record) to close the loop, and I will not upgrade a negative finding into a positive NPE determination without a reel/frame to cite.

Verification links:

• USPTO Assignment Center — https://assignmentcenter.uspto.gov/
• Legacy Assignment Search (search by patent number 5716978) — https://assignment.uspto.gov/patent/index.html
• Google Patents legal events (source of Reel 022401/0190) — https://patents.google.com/patent/US5716978/en

Recommended follow-up if this matters operationally: (1) confirm whether an inventors → State of Oregon assignment exists at an earlier reel/frame; (2) retrieve the correspondent of record on 022401/0190; (3) confirm the State of Oregon → OHSU name-change/assignment status, since Google Patents still lists the current assignee as "State of Oregon" while the 2009 license names OHSU as assignor.

I'll search the USPTO and related databases to verify the patent and identify its prior-art references.

Let me gather details on the remaining cited references.

Prior-Art Analysis — US 5,716,978 (Lewy & Sack, "Methods of treating circadian rhythm disorders")

Sourcing note. I confirmed the target record directly. All prior-art listings below come from the citation face of US 5,716,978 itself (Google Patents full-text: https://patents.google.com/patent/US5716978/en and Justia mirror https://patents.justia.com/patent/[5716978](/patent/5716978)), supplemented by targeted lookups of the individual references. Per the strict-ID rule, I am again flagging that searches for "5716978" also return JP 5716978 B2 (Analogic Corp., "Object classification using two-dimensional projections") — a different patent in a different jurisdiction that is not prior art to and not the same as US 5,716,978. It is excluded.

---

1. What the examiner actually cited

The patent carries 8 patent citations and 40 non-patent citations. Priority date is 1990-12-04; the application (08/454,545) was filed 1995-05-30; issued 1998-02-10. Examiner of record: Peter G. O'Sullivan (per Justia examiner index, which lists the sibling US 5,707,652 under the same examiner).

A. Patent citations (as they appear on the face of US 5,716,978)

(Note the literal transcription: the patent text also refers in its Background to "U.S. Pat. Nos. 4,665,086 and 4,600,723." Item 3, US 4,665,086, is a Monash sibling of US 4,600,723.)

B. Key non-patent citations (of the 40)

• Arendt et al., 1984, "The effects of chronic, small doses of melatonin given in the late afternoon on fatigue in man: A preliminary study," Neurosci. Lett. 45:317–325.
• Arendt et al., 1985, "Some effects of melatonin and the control of its secretion in humans," CIBA Found. Symp. 117:266–283.
• Arendt et al., 1987, "Some effects of jet-lag and their alleviation by melatonin," Ergonomics 30:1379–1393.
• Armstrong, 1989, "Melatonin and circadian control in mammals," Experientia 45:932–938.
• Armstrong, 1991, "Treatment of sleep disorders by melatonin administration," Adv. Pineal Res. 6:263–274.
• Gwinner & Benzinger, 1978, J. Comp. Physiol. 126:123–129 (melatonin entrains bird activity/rest cycle).
• Mallo et al., 1988, "Effects of a four-day nocturnal melatonin treatment on the 24 h plasma melatonin, cortisol and prolactin profiles in humans," Acta Endocrinol. 119:474–480.
• Underwood, 1986, "Circadian Rhythms in Lizards: Phase Response Curve for Melatonin," J. Pineal Res. 3:187–196.
• Lewy et al., 1992, "Melatonin shifts human circadian rhythms according to a phase-response curve," Chronobiol. Int'l 9:380–392 (the inventors' own enabling paper — not prior art against themselves).
• Others: Claustrat 1992, Petrie 1989, Samel 1991, Nickelsen 1991, Skene 1989, Wirz-Justice 1990, Parry 1987, Rosenthal 1988, Deguchi 1978, Lewy & Markey 1978.

---

2. Reference-by-reference analysis

Below I give full citation, dates, description, and the claim(s) each reference could potentially reach — with an explicit § 102 vs. § 103 call. Important framing: for a reference to anticipate under § 102 it must disclose every element of the claim, arranged as claimed, in a single reference. The distinguishing elements of the independent claims of US 5,716,978 are (i) a sub-1 mg dose, and (ii) timing relative to the individual's endogenous melatonin onset (DLMO), in (iii) a free-running/desynchronized human. No cited reference discloses (i)+(ii) together. As developed below, this set is predominantly § 103 obviousness art, not § 102 anticipation art.

1) EP 0 126 630 A1 / B1 — Monash University (Short, Armstrong et al.)

• Full citation: EP 0 126 630 A1, "The use of melatonin for the manufacture of a medicament," Monash University; A1 published 1984-11-28, B1 granted 1987-09-16; priority 1983-05-18.
• Description: Broadest Monash disclosure. Teaches treating disturbances of circadian rhythms of bodily performance/function — explicitly night-shift transition and transmeridian travel ("jet lag") — by administering melatonin, with the administration schedule "varied to suit particular needs and circumstances."
• Potential § 102 claim(s): Claim 1 is the only candidate (melatonin to treat a circadian-rhythm disturbance). But the reference does not disclose the sub-1 mg limitation or DLMO-relative timing, and addresses disturbed/entrained rhythms (jet lag), not the free-running rhythm recited in claim 1. Strictly, no single-reference anticipation. Best treated as § 103 art, and even then it is the primary "general teaching" reference rather than the closest art on the two limitations that matter.

2) US 4,600,723 — Monash University (Short et al.)

• Full citation: US 4,600,723, issued 1986-07-15; priority 1983-05-18; Monash University.
• Description: The flagship Monash jet-lag patent. Teaches 1–10 mg melatonin taken at destination bedtime and again on premature mid-night awakening. Worked examples show ~5 mg doses. The specification of US 5,716,978 expressly characterizes this as "merely reinforc[ing] the (usual) rise in endogenous melatonin which occurs near the time of sleep onset," attributable to melatonin's soporific properties rather than true phase-shifting.
• Potential § 102 claim(s): Claim 1 (method of treating a circadian-rhythm disorder with melatonin). Fails anticipation because: dose is an order of magnitude higher than "less than 1 mg"; administration at bedtime is at/after DLMO, not "prior to" it; and the subject is jet-lagged/entrained, not "free-running and out of synchrony." Analysis is conditional at best; in reality § 103.

3) US 4,665,086 — Monash University

• Full citation: US 4,665,086, issued 1987-05-12; priority 1983-05-18; Monash University.
• Description: Sibling of US 4,600,723 with the identical title (from the Monash University Patents Collection, bridge ID US4665086). Same jet-lag teaching.
• Potential § 102 claim(s): Same as item 2 — claim 1 candidate only; fails on the same three limitations. § 103 art.

4) US 4,687,763 — Massachusetts Institute of Technology

• Full citation: US 4,687,763, "Composition and method for increasing levels or release of brain serotonin," filed 1985-02-25, issued 1987-08-18, MIT.
• Description: Directed to brain serotonin (tryptophan/5-HT augmentation), not to melatonin phase-shifting per se. Cited as general art about indoleamine/serotonin neurochemistry relevant to melatonin's biosynthetic pathway.
• Potential § 102 claim(s): None. It does not disclose administering melatonin to shift a circadian phase. It is background/general-art only (relevance at most under § 103 as evidence of the melatonin–serotonin relationship).

5) US 5,163,426 — Brigham and Women's Hospital (Czeisler, Allan, Kronauer)

• Full citation: US 5,163,426, "Assessment and modification of a subject's endogenous circadian cycle," app. 07/066,677, priority 1987-06-25/26, issued 1992-11-17, Brigham and Women's Hospital.
• Description: The landmark "constant routine" patent. Claims methods of assessing endogenous circadian phase/amplitude (by eliminating masking effects) and modifying it using timed bright light (~9,500 lux) and scheduled darkness at critical phases, optionally via a van der Pol oscillator model. Expressly aimed at jet lag, shift work, advanced/delayed sleep phase. It does not administer melatonin.
• Potential § 102 claim(s): Two aspects are relevant. (a) Claims 23 and 26 each require the step of "determining the time of … endogenous melatonin onset." US 5,163,426 teaches determining endogenous circadian phase, but by temperature/constant-routine methods and light stimuli — it does not teach determining melatonin onset, and it does not administer melatonin. (b) The general "phase-shift the body clock by timed stimulus" concept is shared. No anticipation; at most § 103 art on the "determining phase first" concept (and even there the inventors' own Lewy & Markey 1978 assay is the enabling method actually used).

6) US 5,167,228 — Brigham and Women's Hospital

• Full citation: US 5,167,228, "Assessment and modification of endogenous circadian phase and amplitude," issued 1992-12-01; priority 1987-06-26.
• Description: Continuation-family member of the Czeisler/Kronauer work; light-based phase/amplitude modification.
• Potential § 102 claim(s): Same as item 5. Light-based; no melatonin, no DLMO determination. No anticipation; § 103 general art.

7) US 5,176,133 — Brigham and Women's Hospital

• Full citation: US 5,176,133, "Assessment and modification of circadian phase and amplitude," issued 1993-01-05; priority 1987-06-26.
• Description: Further family member; light/dark scheduling to reset circadian phase, including reduction of amplitude toward a singular point for rapid shifts.
• Potential § 102 claim(s): Same as items 5–6. No anticipation. Relevant to the "phase-shifting is achievable" state of the art (§ 103) and to the inventors' own framing that prior art had only demonstrated light-driven phase shifting in humans.

8) WO 1994/007487 A1 — Massachusetts Institute of Technology (Wurtman)

• Full citation: WO 94/07487 A1, "Use of melatonin to induce sleep," MIT, filed 1992-10-01, published 1994-04-14. (Equivalent EP 0 663 825 B1.)
• Description: Double-blind, placebo-controlled 5×5 Latin-square study administering 0.1, 0.3, 1.0 or 10 mg oral melatonin at 1145 h and measuring sleep latency, temperature, and serum melatonin. Claims use of melatonin (including low oral doses) to induce sleep.
• Critical date problem: published 1994-04-14, i.e., after the 1990-12-04 priority date. For claims entitled to the 1990 priority (e.g., claim 1, whose "less than 1 mg" concept the specification supports), this reference is not § 102(a)/(b)/(e) prior art at all, and it is not an anticipatory reference. Its only potential hook is the infant claims (claims 3/23–25) if those limitations (infant; "in milk"; entrainment of infants) are held not supported by the 1990 disclosure and instead entitled only to the 1993/1995 filing dates — in which case WO 94/07487 could be § 102(b) art against claim 23. But WO 94/07487 discloses sleep induction in adults, not infants or milk, so even then there is no anticipation. Net: no § 102 claim, even setting aside the date defect.

Non-patent references — the ones that come closest

Arendt et al., 1987, Ergonomics 30:1379–1393 ("Some effects of jet-lag and their alleviation by melatonin")

• Description: Orally administers melatonin 4–6 hours before the normal bedtime (plus mid-night dose) to reduce jet lag. The US 5,716,978 specification explicitly addresses this reference and asserts its data do not support actual circadian re-entrainment.
• Potential § 102 claim(s): Claim 1 candidate (timing "prior to" sleep). Fails anticipation: 4–6 h before bedtime is not tied to endogenous melatonin onset; the dose (typically ≥1 mg) is not "less than 1 mg"; subjects are jet-lagged travelers, not free-running. § 103 art — and notably, this reference is why the "relative to DLMO" and "free-running" limitations carry the patentability weight.

Mallo et al., 1988, Acta Endocrinol. 119:474–480

• Description: 8 mg melatonin, 1 hour before bedtime, four days; slight phase advance noted three days after cessation.
• Potential § 102 claim(s): Claim 1/claim 4 candidate (phase advance). Fails: 8 mg ≫ 1 mg; 1 h before bedtime ≈ at DLMO, not "4–8 h prior"; no free-running subject. § 103 art.

Armstrong, 1989, Experientia 45:932–938

• Description: Rat studies; effect of melatonin on the sleep/wake rhythm depends on time of administration relative to the sleep/wake cycle, greatest when given several hours before the effective start of the nocturnal activity cycle. No phase-delay arm and no PRC.
• Potential § 102 claim(s): Species matters. Claims 1–28 are human methods (and claims 3, 23–25 infants). A rodent study cannot anticipate a human method unless it enables the human practice. Also it lacks phase-delay and the sub-1 mg / DLMO limitations. § 103 art on the "timing matters" concept.

Underwood, 1986, J. Pineal Res. 3:187–196 (PRC for melatonin in the lizard Sceloporus occidentalis)

• Description: First (non-mammalian) melatonin phase-response curve.
• Potential § 102 claim(s): None as anticipation — lizard, not human; the specification itself touts the inventors as "the first to disclose a PRC for melatonin in a human, and perhaps in any mammalian species." This is the strongest § 103 "obviousness of extending a PRC across species" reference, not § 102.

Gwinner & Benzinger, 1978, J. Comp. Physiol. 126:123–129

• Description: Daily melatonin injections entrain the activity/rest cycle in birds.
• Potential § 102 claim(s): None (avian). § 103 background on melatonin as an entraining signal.

Arendt et al., 1984, Neurosci. Lett. 45:317–325 (small doses of melatonin in late afternoon → fatigue)

• Description: Small chronic doses given in the late afternoon alter fatigue/alertness.
• Potential § 102 claim(s): Close on the "small dose, late-afternoon administration" fact pattern, but it reports a symptom (fatigue) effect, not a measured phase shift of an endogenous rhythm, and does not tie dosing to DLMO in a free-running subject. § 103 art; could be cited against claims 4/6–8 if the timing windows were argued to be obvious from "late afternoon."

US 5,707,652 / WO 95/05819 and later family (Lewy) and US 5,449,683 (MIT, Wurtman sleep) — these appear in the "Similar Documents"/"Cited By" lists. The family members are the inventors' own work and are not prior art to this patent; US 5,449,683 (sleep induction) is a counterpart of WO 94/07487 and carries the same post-1990 date defect.

---

3. Claim-by-claim § 102 map (best single-reference candidate)

---

4. Bottom line

1. No cited reference anticipates any claim of US 5,716,978 under § 102. The independent claims each carry at least one limitation absent from every reference: the sub-1 mg dose, the DLMO-relative timing (as opposed to bedtime-relative), and the free-running/desynchronized subject. The BWH/Czeisler references use light, not melatonin; the Monash references use higher doses at bedtime for jet lag; WO 94/07487 is both date-defective against the 1990 priority and directed to sleep induction.

2. The cited art is § 103 material, and it is the reason the claims issued. The two limitations that distinguish this patent — low dose and phase angle relative to endogenous melatonin onset (DLMO) — are exactly the features the specification says the inventors were first to discover (the human melatonin PRC). The examiner (P. G. O'Sullivan) cited the Monash and Czeisler families as the closest art but did not find a single-reference disclosure of those features.

3. Date-sensitivity flag (do not auto-correct): WO 1994/007487 A1 (published 1994-04-14) and US 5,449,683 are after the 1990-12-04 priority date and, for claims entitled to that date, are not valid § 102 prior art. Their appearance on the face of the patent does not convert them into anticipatory references.

4. Species/enablement caveat: Underwood (1986, lizard) and Armstrong (1989, rat) cannot anticipate the human claims; they are at most § 103 art about extending melatonin-PRC reasoning across species.

5. Residual uncertainty: I did not retrieve prosecution-history (file-wrapper) Office Actions, so I cannot confirm which references the examiner actually applied as § 102 rejections versus mere citations; the citation face alone does not distinguish "cited" from "applied." If you need the actual rejection grounds, pull the file wrapper for app. 08/454,545 via USPTO PatentCenter. Everything above reflects the reference disclosures as I could verify them, not the examiner's stated rationale.

Primary sources: https://patents.google.com/patent/US5716978/en · https://patents.justia.com/patent/5716978 · http://data.epo.org/gpi/EP0126630B1 · https://patents.google.com/patent/[US5163426A](/patent/US5163426A)/en · https://patents.google.com/patent/WO1994007487A1 (EP 0 663 825 B1)

I'll ground this analysis in the prior-art references listed on the patent's face and verify key teachings before mapping § 103 combinations.

Obviousness Analysis — US 5,716,978 under 35 U.S.C. § 103

Scope note. This analysis builds on the previously generated sections (bibliography, litigation, PTAB, assignment) and does not repeat them. It uses the prior-art citations appearing on the face of US 5,716,978 (Patent Citations (8) + Non-Patent Citations (40)), as instructed, together with bibliographic verification of what those references actually teach.

Date discrepancy flagged, not resolved. The task header states "Current Date: April 26, 2026," while the system date and the previously generated sections use 2026-10-01. This does not affect any § 103 conclusion (it only shifts the § 286 damages discussion already covered), but the two dates should not be treated as interchangeable.

Practical posture (carried forward). The patent expired 2015-02-10. Any § 103 analysis today is academic or defensive (e.g., supporting a § 282 defense in a case involving acts inside a 2015-adjacent damages window, or supporting a collateral estoppel argument). That framing matters because a challenger in 2026 has no PTAB incentive and no realistic damages exposure.

---

1. Governing law and the prior-art cutoff

This is a pre-AIA patent. The application (08/454,545, filed 1995-05-30) is a continuation in a chain reaching Ser. No. 07/621,866, filed 1990-12-04. Assuming § 112 support in the 1990 parent (the specification is substantially identical across US 5,242,941 → 5,420,152 → 5,716,978), pre-AIA § 102(a)/(b)/(e) and § 103(a) apply, and the critical date for § 102(b) is 1989-12-04 (one year before the 1990-12-04 filing).

This date filter is the single most important step, because many of the references listed on the patent's face post-date the invention and are therefore not § 102(b) art at all.

Consequence. A challenger may not rely on Claustrat 1992, Nickelsen 1991, Samel 1991, Armstrong 1991, or WO 94/007487 to fill any gap — and, critically, may not rely on Lewy et al., Chronobiol. Int. 9:380-392 (1992), which is the published human melatonin PRC itself. That is a major structural point in favor of the patent and is discussed in § 6 below.

Contingency worth flagging: if any limitation in the 1995 claims (e.g., the "< 1 mg" cap or the CT 0 / CT 8 windows) were held to lack written-description support in the 1990 parent, the effective date shifts to 1992-02-25 or later, and several 1991 references become § 102(b) art. The specification's Examples appear common to the family, so I do not assume this, but it is the one path by which the post-1990 references become usable.

---

2. What the prior art would have to teach

Two internal inconsistencies should be noted because they affect the § 103 exposure:

• Claim 28 recites a "delay" window of "about 6 hours to about 19 hours prior to endogenous melatonin onset," which overlaps claim 4's advance window (4–8 h). A person practicing claim 4 at ~6 h would literally perform claim 28. Claim 28 is therefore both the broadest and the most internally strained claim.
• The specification's preferred delay window (10–14 h before onset) does not match claim 9's 12–16 h. Claim 9 covers a band the specification does not describe as optimal.

---

3. What the key references actually teach

Monash family (US 4,600,723; US 4,665,086; EP 0 126 630). Melatonin (1–10 mg) to minimize disturbances in circadian rhythms of bodily performance/function — i.e., jet lag — taken at destination bedtime and again on mid-night awakening. These are the references the '978 specification characterizes as producing mainly soporific/hypnotic, not chronobiologic, effects.

Underwood, J. Pineal Res. 3:187-196 (1986) — a melatonin phase-response curve. Single melatonin injections in free-running lizards produced shifts "whose magnitude and direction are a function of the time of the melatonin injections, relative to activity onsets"; injections 6–15 h after activity onset elicited phase advances, other phases produced phase delays. (PubMed; Wiley). This is the closest thing in the pre-1990 record to the patent's central insight.

Armstrong, Experientia 45:932-938 (1989) — melatonin is "a potent synchronizer of rat circadian rhythms" with "a direct action on the circadian pacemaker" (PubMed). A later review notes the rat data were incomplete: injection in the late subjective night "did not consistently induce phase-shifts," with advances in only ~1/3 of rats at CT 22 and delays in ~1/5 at CT 18. This cuts both ways and is the patent's best "failure of others" evidence.

Arendt et al., Ergonomics 30:1379-1393 (1987) — 5 mg melatonin. Pre-flight dosing at 18:00 local time, then bedtime dosing (22:00–24:00) post-arrival. Reported faster resynchronization of endogenous melatonin and cortisol rhythms. (Cochrane characteristics summary). Notably, 18:00 with a typical DLMO near 21:00 is roughly 2–4 h before endogenous melatonin onset — i.e., Arendt already dosed before DLMO, but at 5 mg, not < 1 mg.

Arendt et al., Neurosci. Lett. 45:317-325 (1984) — "chronic, small doses" (2 mg) at 17:00 for four weeks; increased evening fatigue after four days and phase-advanced the endogenous melatonin rhythm. With a habitual DLMO near 21:00, this is roughly 3–4 h before onset, i.e., inside claim 4's window (again, at > 1 mg).

Mallo et al., Acta Endocrinol. 119:474-480 (1988) — 8 mg one hour before bedtime → "a slight phase advance three days after cessation."

Brigham & Women's Hospital family (US 5,163,426 / 5,167,228 / 5,176,133). Teaches (i) assessing the subject's endogenous circadian phase (constant routine / ECP minimum, or inference from normative data or habitual sleep/wake pattern) and (ii) applying a stimulus at a critically chosen phase derived from that assessment, with the express statement that "the intervention stimulus chosen may be of any nature (i.e., a pharmaceutical agent or other therapy)" (US 5,167,228). The later continuation US 5,304,212 claims stimulus "preceding said subject's desired bedtime to thereby delay" and "following said subject's desired wake time to thereby advance" (US 5,304,212).

References embedded within the cited references (each independently pre-1989 printed publications, so available as art even though not listed on the '978 face):

• Gwinner & Benzinger (1978) — daily melatonin injections entrained the activity/rest cycle in pinealectomized starlings (i.e., a free-running system entrained by timed melatonin).
• Redman, Armstrong & Ng, Science 219:1089-1091 (1983) — daily melatonin entrained free-running rat activity rhythms.
• Cassone, Chesworth & Armstrong, J. Biol. Rhythms 1:219-229 (1986) — dose-dependent entrainment of rat rhythms by daily melatonin.
• Arendt et al., Lancet 2:772-773 (1988) — melatonin synchronized "a disturbed sleep-wake cycle in a blind man."
• Davis & Mannion, Am. J. Physiol. 255:R439-R448 (1988) and Reppert & Schwartz, Science 220:969-971 (1983) — entrainment of pup/fetal circadian rhythms by maternally administered melatonin.

Bibliographic discrepancy flagged (not auto-corrected): the '978 specification cites "Gwinner and Benzinger J. Comp. Physiol. 126:123-129 (1978)." Verification shows J. Comp. Physiol. 126:123-129 (1978) is Gwinner's single-author pinealectomy paper, while the Gwinner & Benzinger entrainment paper is J. Comp. Physiol. 127:209-213 (1978). Both are pre-1989; the substance cited (entrainment of a free-running rhythm by daily melatonin) is in the 127 paper. Separately, the specification body cites Arendt 1985 CIBA as "266-283, 1986" while its own citation list gives "1985."

---

4. Combinations that would render the claims obvious

Combination A — "Known agent + known PRC + known timing direction" (targets claims 1, 4–8, 21–22)

References: US 4,600,723 / US 4,665,086 (Monash) + Underwood 1986 + Armstrong 1989 + Arendt 1987 + Gwinner & Benzinger 1978.

Rationale / motivation to combine:

1. Monash supplies every element except timing: melatonin is a known, well-tolerated agent for circadian disturbance in humans.
2. Underwood 1986 supplies the concept that melatonin's effect on a free-running circadian system is directionally time-dependent (advances at one phase, delays at others) — and expressly states that its PRC "allows predictions as to the effects of continuous exogenous melatonin … as well as on the mechanism of entrainment."
3. Gwinner & Benzinger 1978 and Redman 1983 (both within Armstrong 1989's citations) supply entrainment of a free-running rhythm by daily melatonin — the functional heart of claims 1, 21, 22, and 26's "specific time each day."
4. Armstrong 1989 supplies the mechanism (direct action on the pacemaker) and the "several hours before the activity cycle" timing generalization.
5. Arendt 1987 supplies the clinical teaching that earlier-than-bedtime (18:00) dosing is what drives resynchronization, i.e., the advance direction.

KSR hook: "If a technique has been used to improve one device, and a person of ordinary skill … would recognize that it would improve similar devices in the same way, using the technique is obvious." Here, a POSA confronting jet lag/shift-work/free-running blindness, armed with a melatonin PRC in one free-running vertebrate (lizard), a synchronizer literature in rats and birds, and a clinical protocol already dosing melatonin hours before bedtime, would predictably set the timing by the patient's own circadian phase rather than by the clock. The element most cleanly met by Arendt 1987 is claim 4's "about 4 to 8 hours prior to endogenous melatonin onset" — 18:00 dosing against a ~21:00 DLMO.

Weakness: every reference in this combination uses 1–10 mg; nothing teaches or suggests < 1 mg except by optimization (see Combination C), and nothing expressly frames the timing variable as endogenous melatonin onset rather than clock time or bedtime.

---

Combination B — "Assess the phase, then dose at phase-selected time" (targets claims 1, 23, 26–28 — the strongest combination)

References: US 5,167,228 / US 5,163,426 / US 5,176,133 (BWH) + Monash + Underwood 1986 + Armstrong 1989.

Rationale / motivation to combine:

1. The BWH specification expressly contemplates that the "intervention stimulus" may be a pharmaceutical agent or other therapy — not just light — and that the appropriate administration time is selected based on the individual's assessed circadian phase, which may be inferred from normative data or the subject's sleep/wake pattern.
2. A POSA looking for a non-light stimulus would immediately select melatonin: Monash/Arendt taught it, Armstrong taught it acts directly on the pacemaker, and Underwood taught that its direction of shift depends on circadian phase.
3. Because BWH teaches deriving the administration time from the subject's own phase, the claimed step "(a) determining the time of the human's endogenous melatonin onset" (claims 23, 26) and the relative-time limitations (claims 4–13) follow directly. The melatonin assay to do so was already published (Lewy & Markey, Science 201:741-3, 1978, cited in the '978 specification itself).
4. Claims 23 and 26 become near-mappings: BWH supplies "determine phase → administer stimulus at phase-derived time"; Monash supplies "melatonin"; Redman/Cassone/Gwinner supply "at a specific time each day"; only the dose and the "in milk" vehicle remain.

For claims 23–25 and 26–28, the practical question is not whether the architecture was known (it plainly was) but whether the < 1 mg dose and the specific windows were.

---

Combination C — "Routine optimization / finite predictable options" for the < 1 mg limitation (targets claim 1 and every dependent)

References: Arendt 1984 ("chronic, small doses") + Arendt 1985 (CIBA; "some effects of melatonin and the control of its secretion in humans") + Mallo 1988 + the physiological-level context inherent in Mallo/Arendt.

Rationale / motivation to combine:

1. The specification itself concedes the mechanism of the prior art's clinical benefit: high-dose bedtime melatonin acts "mainly [through] the soporific, hypnotic and sleep-inducing properties" (see Arendt 1984, 1985 and the Monash examples). A POSA seeking a chronobiologic rather than hypnotic effect has an express design incentive to lower the dose to a level equal to physiologic plasma concentrations — exactly the stated purpose of the claimed range.
2. Arendt 1984's own framing ("chronic, small doses") shows the art was already moving toward lower doses.
3. Under In re Aller / In re Boesch / MPEP 2144.05, optimizing a known working dosage range is prima facie obvious absent evidence of criticality, and the "finite number of identified, predictable solutions" rationale of KSR applies to a single scalar (dose).

This is the weakest link in the challenger's case — see § 6.2.

---

Combination D — Infant claims (23–25)

References: Davis & Mannion 1988 + Reppert & Schwartz 1983 (both within Armstrong 1989) + Monash + Arendt 1988 (blind-man synchronization) by analogy.

Motivation: Davis & Mannion expressly teach that circadian rhythms of pups can be entrained by melatonin given to the mother, and Reppert & Schwartz teach maternal coordination of the fetal biological clock. A POSA seeking to correct an infant's free-running/erratic sleep schedule would therefore (i) treat the infant, (ii) dose by the infant's own phase, and (iii) use the most conventional pediatric vehicle — milk or formula — which the specification itself identifies as convenient. The "in milk" limitation is a delivery-vehicle limitation with no asserted unexpected property, and is vulnerable as a mere matter of obvious choice of carrier.

---

5. Claim-by-claim prima facie vulnerability

---

6. Where the § 103 case is genuinely vulnerable (the non-obviousness side)

6.1 The "fill-the-gap" reference is post-priority and cannot be used

The human melatonin PRC — the patent's asserted point of novelty — was published as Lewy et al. 1992, after the 1990-12-04 invention and by the inventors themselves. It is therefore not § 102(a), (b), or (e) art. Under pre-AIA § 103, a challenger may not combine a pre-1990 genus-level teaching (lizard/rat PRCs) with a post-1990 species-level human disclosure to supply the missing limitation. The fact that a human PRC had to wait until 1992 is itself evidence that the 1990 discovery was not in the art. Note the contrast: in Vanda Pharms. v. Teva/Apotex (D. Del.; Fed. Cir. 2023; cert. app. No. 23-768), a court applying a 2012 priority date found the human melatonin PRC to be settled background ("no dispute that it works") — but that is 22 years after this patent's priority, and the Lewy 1992 work is what made it settled.

6.2 The dose finding cuts against the challenger

The patent's own Example 2 reports that 0.5 mg produced a cumulative phase advance equivalent to 5 mg. If credited, that is an unexpected result in the Graham sense: the art uniformly taught 1–10 mg (Monash), 2 mg (Arendt 1984), and 8 mg (Mallo 1988), so going below 1 mg ran against the art's implicit dose floor. A challenger must show the "criticality" of the < 1 mg boundary was predictable, and the cited art does not supply a dose-response for phase shifting at sub-milligram levels.

6.3 Express teaching away on the delay claims

Arendt 1987's statement that morning dosing is undesirable for initiating a phase delay (sedation during the day) is a concrete teaching away from claims 9–13, 19–20, 25, and 28. Because those claims are directed precisely at morning/sleep-offset dosing, a challenger must overcome In re Gurley-style teaching-away.

6.4 Failure of others

Armstrong 1989's rat PRC was incomplete (no consistent late-subjective-night phase shifts; advances in only ⅓ of animals). The '978 specification expressly leverages this: "these authors were unable to demonstrate phase-delay shifts or a phase-response curve." A robust, reproducible human PRC — and in particular a phase-delay limb — was not something the art had in hand.

6.5 The "biological parameter" problem

Claims 23 and 26 require determining the subject's own endogenous melatonin onset before dosing. Readings of DLMO were impractical at scale in 1990; whether a POSA would have had a reasonable expectation of implementing this in routine practice is contestable. (The 1992 lewy PRC study required serial plasma sampling with GC-MS.)

---

7. Bottom line

A prima facie § 103 case exists against the core architecture of claims 1, 4–8, 21–22, 26–27, and 28, built principally on:

• Combination A — Monash (melatonin for circadian disturbance) + Underwood 1986 (lizard melatonin PRC) + Armstrong 1989 (direct pacemaker action; free-running entrainment per Gwinner 1978 / Redman 1983) + Arendt 1987 (pre-bedtime, pre-DLMO dosing for resynchronization); and
• Combination B — the Brigham & Women's patents (assess phase → apply any stimulus, "a pharmaceutical agent or other therapy," at a phase-selected time) + Monash + Underwood.

The claims most likely to survive are the phase-delay claims (9–13, 19–20, 25, 28), because the entire citable pre-1990 art taught evening/bedtime dosing for a delay and Arendt 1987 expressly discouraged morning dosing. The claims most likely to fall are the result limitations (21–22), the broad delay range of claim 28 (which overlaps the advance zone), and — if the dose-optimization rationale is accepted — claim 4, which Arendt 1984/1987 already reach on timing.

Net assessment: roughly a coin-flip. The decisive variable is the "less than 1 mg" limitation. If a factfinder credits the art's uniform 1–10 mg teaching as a floor and credits the 0.5 mg ≈ 5 mg equivalence, the dose limitation supplies non-obviousness for claim 1 and everything depending from it. If the < 1 mg boundary is treated as routine optimization of a known dosage range (In re Aller; KSR), several claims fall — but the delay claims remain protected by teaching-away.

Limits and confidence. High confidence in: the prior-art dates and the pre-AIA framework; the substantive content of Underwood 1986, Armstrong 1989, Arendt 1987, Arendt 1984, Mallo 1988, and the BWH patents as quoted above. Moderate confidence in: the exact content of the 1989–1990 BWH filings (I verified the "pharmaceutical agent" language in the family specification via US 5,304,212, a later continuation filed 1992-01-10; a challenger must confirm that language appears in US 5,163,426 (1987), US 5,176,133 (1989), or US 5,167,228 (1990-05-09) for it to carry a § 102(e) date — I did not complete that verification). I have not located any actual § 103 determination on US 5,716,978; consistent with the previously generated PTAB and litigation sections, I found none, and I am not asserting that any such ruling exists.