US 8772306

I'll search for authoritative information on this specific patent number and any 2026 CAFC docket activity.

Now let me specifically check for any 2026 CAFC (Federal Circuit) docket activity involving this patent.

Let me try to obtain the actual claim text to describe the independent claims accurately.

US Patent 8,772,306 — Summary

Important sourcing note: I searched for this exact number and did not find any 2026 Court of Appeals for the Federal Circuit (CAFC) docket activity involving patent 8,772,306. The Federal Circuit oxybate appeals I could locate (Avadel CNS Pharmaceuticals v. Jazz Pharmaceuticals, Nos. 24‑2274 and 24‑2277, filed Aug 2024, closed May 2025) involve different patents (e.g., US 10,966,931; US 11,147,782; US 10,758,488; US 9,731,963 — NOT the '306 patent). If a 2026 CAFC matter on the '306 patent exists, I could not verify it, and I will not guess. The 2026 New Jersey district-court dockets surfaced (e.g., 2:26‑cv‑01739 / 2:26‑cv‑01740) are district-court ANDA suits, not CAFC appeals, and I cannot confirm from my results that the '306 patent is asserted in them.

Bibliographic data

Abstract (verbatim, condensed)

The invention improves the safety and efficacy of administering GHB or a salt thereof. It was discovered that concomitant administration of an MCT inhibitor — such as diclofenac, valproate, or ibuprofen — affects GHB administration. Diclofenac lowers the effect of GHB (potentially unsafe); valproate increases the effect of GHB (potentially unsafe).

Independent claims — plain language

The patent has 34 claims. Per the PTAB institution decision in IPR2016‑00024 (Ranbaxy v. Jazz), claims 1, 11, 19, 30, and 33 are independent, and all challenged claims are directed to methods of treating sleep disorders by orally administering a reduced dosage of GHB to patients concomitantly receiving valproate. (Notably, although the specification discusses adjusting GHB upward for diclofenac, the granted claims appear to be valproate-focused.)

Independent claim 1 (verified across the PTAB decision, the Ranbaxy/Par expert declarations, and JD Supra/Mondaq coverage):

A method for treating a patient suffering from excessive daytime sleepiness, cataplexy, sleep paralysis, apnea, narcolepsy, sleep time disturbances, hypnagogic hallucinations, sleep arousal, insomnia, or nocturnal myoclonus with GHB or a salt thereof, comprising orally administering to the patient at least a 5% decrease in an effective dosage amount of the GHB or salt when the patient is receiving concomitant administration of valproate (an acid, salt, or mixture thereof).

In plain terms: When a narcolepsy/sleep-disorder patient is also on valproate, give them a reduced GHB dose (by at least 5%) to offset valproate's potentiating effect.

Independent claim 19 was also reproduced as "illustrative" in the Board's IPR2016‑00024 decision, but the full text was not captured in my search results — I cannot state its exact wording with confidence. Based on the specification and the case record, the other independent claims (including claim 19 and others such as 11, 30, 33) are understood to recite analogous dose-adjustment methods with varying numerical limits — e.g., reductions of "at least about 15%" (dependent claim 2), and ranges such as about 15%–30% or about 5%–35%, and specified baseline daily doses (e.g., 4.5 g, 6 g, 7.5 g, 9 g). Treat the exact scope of claims 11, 19, 30, and 33 as unverified here.

Key prosecution/litigation context (grounded)

• The claims were asserted to survive IPR: IPR2016‑00024 (Ranbaxy) terminated–settled (May 23, 2016); IPR2016‑00002 (Par Pharmaceutical) institution denied on the merits (Apr 12, 2016); IPR2016‑00546 (Amneal) institution denied (Jul 28, 2016).
• The claim 1 "at least 5%" reduction, with dependent claim 2 at "at least about 15%," was central to the PTAB's denial reasoning.

Confidence statement

High confidence: patent number, title, inventor (Eller), assignees, application number, abstract, expiration, claim 1 text, and the PTAB outcomes. Lower/flagged confidence: exact filing/issue dates (one-day discrepancies between sources), the full text of independent claims 11, 19, 30, and 33, and the existence of any 2026 CAFC docket — which I could not confirm. No result for the specific number 8,772,306 in a 2026 Federal Circuit proceeding was found.

I'll research litigation involving US patent 8772306. Let me start with searches.

Let me gather details on specific cases.

US Patent 8,772,306 — Litigation Summary

The patent itself (context for the list): US 8,772,306 B1, "Method of administration of gamma hydroxybutyrate with monocarboxylate transporters," inventor Mark Eller, filed April 29, 2013, granted July 8, 2014; original assignee Jazz Pharmaceuticals, Inc., current assignee Jazz Pharmaceuticals Ireland Ltd. It is the original "DDI" (drug–drug interaction) patent in the Xyrem®/Xywav® oxybate family (priority March 1, 2013; anticipated expiration March 15, 2033). Because it is an Orange Book–listed method-of-use patent for sodium oxybate, all known litigation is Hatch-Waxman ANDA/505(b)(2) litigation brought by Jazz, plus one collateral False Claims Act case.

Below, I distinguish (A) cases I could positively verify as asserting the '306 patent, (B) additional cases that Google Patents' litigation data links to this patent number but whose parties/dates I could not independently confirm, and (C) related contested proceedings.

A. District-court cases verified as asserting the '306 patent

B. Additional cases that Google Patents' litigation data links to U.S. 8,772,306 (party/date unverified by me)

These case numbers appear in Google Patents' "Family has litigation" record for the '306 patent, but I could not independently confirm the parties or filing dates, so treat them as leads rather than verified assertions:

• 2:14-cv-05824 (D.N.J.)
• 2:14-cv-06150 (D.N.J.)
• 2:14-cv-06151 (D.N.J.)
• 2:14-cv-07757 (D.N.J.)
• 2:15-cv-01043 (D.N.J.) — appears to involve Amneal (the submittal in this docket contains Amneal's declaratory-judgment counterclaims of non-infringement and invalidity specifically directed to the '306 patent).
• 2:17-cv-04439 (E.D.N.Y.) — the only non-New Jersey venue in the list
• 2:17-cv-05487 (D.N.J.)
• 2:18-cv-00029 (D.N.J.)
• 2:23-cv-03182 (D.N.J.)

A November 2015 IPR real-parties-in-interest filing by Ranbaxy stated the '306 patent was then the subject of four pending litigations: Jazz v. Amneal, 2:13-cv-391 (consolidated); Jazz v. Roxane, 2:15-cv-1360; Jazz v. Wockhardt Bio, 2:15-cv-5619; and Jazz v. Lupin, 2:15-cv-6548 (all D.N.J.). It also noted two further cases (Jazz v. Amneal, 2:15-cv-6562 and Jazz v. Par, 2:15-cv-7580) concerned a related patent—these are family, not '306-specific, so I have not listed them as '306 litigation.

C. Related contested proceedings (PTAB — not district-court litigation, but directly on the '306 patent)

The '306 patent was challenged three times at the PTAB; none resulted in instituted review:

• IPR2016-00002 — Petitioner Par Pharmaceutical, Inc. — filed Oct 6, 2015 — Not instituted (merits), decision Apr 12, 2016.
• IPR2016-00024 — Petitioner Ranbaxy, Inc. — filed Oct 7, 2015 — Settlement (per Google Patents).
• IPR2016-00546 — Petitioner Amneal Pharmaceuticals LLC — filed Feb 2, 2016 — Not instituted (merits), decision July 28, 2016 (cited as Amneal Pharms. LLC v. Jazz Pharms. Ireland Ltd., 2016 WL 5222883 (PTAB July 28, 2016)).

The district court litigation brought by Amneal and Roxane referenced the '306 IPR and stayed or administratively terminated '306 proceedings pending it.

D. Collateral False Claims Act case touching the '306 patent

• United States ex rel. Leonardo Sorgi v. Jazz Pharmaceuticals PLC, Jazz Pharmaceuticals, Inc., and Jazz Pharmaceuticals Ireland Limited — D. Mass., 1:21-cv-10891-PBS. Filed under seal May 27, 2021; unsealed July 18, 2024. Relator (a patent attorney) alleges Jazz defrauded the PTO in obtaining the '306 patent (non-disclosure of the 2012 Xyrem label and adverse-event reports; allegedly misleading the PTAB). The U.S. and several states declined to intervene (May/June 2024); amended complaint filed Sept 1, 2024; Jazz's motion to dismiss filed Dec 2, 2024; oral argument April 2, 2025. This is an FCA case about the '306 patent, not an infringement suit.

E. Context (antitrust matters that reference Xyrem patents, including the '306-family)

Reverse-payment/antitrust suits against Jazz over Xyrem settlements — e.g., In re Xyrem (Sodium Oxybate) Antitrust Litigation (N.D. Cal., incl. 3:20-cv-04667), Blue Cross & Blue Shield Ass'n v. Jazz Pharmaceuticals (N.D. Ill., 1:20-cv-03543), United HealthCare Services v. Jazz Pharmaceuticals (D. Minn., 0:21-cv-00737), and the AFSCME/UFCW complaints — reference the Xyrem patent settlements and Orange Book listings generally. The '306 patent is part of the Xyrem Orange Book estate, but I cannot confirm it was a specifically asserted patent-in-suit in those antitrust actions.

Bottom line

US 8,772,306 has been heavily litigated. It has been asserted by Jazz (now Jazz Pharmaceuticals Ireland Ltd.) against at least Amneal, Roxane, Wockhardt, Lupin, Teva, Granules, and Tris Pharma in D.N.J. (plus one apparent E.D.N.Y. case), with the earliest verified assertion in the consolidated 2:13-cv-00391 action and the most recent in the Feb 20, 2026 Tris Pharma suits. It was also the subject of three PTAB IPR petitions (all denied institution or settled) and is the patent at the center of the Sorgi qui tam FCA case in D. Mass.

Caveats: (1) Several case numbers in Google Patents' litigation record (Section B) could not be independently confirmed as '306-specific rather than merely '306-family litigation; some may assert sibling patents (e.g., 9,050,302, 9,486,426). (2) Outcomes for the 2023–2026 cases are not confirmed. (3) I did not obtain every filing date from primary dockets; dates marked with a range are approximate. If you need court-confirmed party names and filing dates for the Section B cases, the PACER dockets for each listed D.N.J. case number would need to be pulled directly.

Sources: Google Patents litigation record for US 8,772,306 (https://patents.google.com/patent/US8772306/en); Ranbaxy IPR2016-00024 real-parties-in-interest filing (ptacts.uspto.gov); PTAB IPR2016-00024 / IPR2016-00002 / IPR2016-00546 records; D.N.J. dockets 2:13-cv-00391, 2:15-cv-01360, 2:15-cv-05619, 2:15-cv-06548, 2:21-cv-14271, 2:23-cv-00329; CourtListener complaint in Jazz Pharmaceuticals Ireland Ltd. v. Tris Pharma (2:26-cv-01739); Univ. Sorgi FCA filings (gov.uscourts.mad.234665); Jazz Pharmaceuticals 10-K/8-K disclosures.

I'll start by reconciling the canonical ODP list against the litigation metadata embedded in the patent text, then search for the specific proceedings.

Now let me pin down the Ranbaxy partial institution, the Amneal petition, and any Federal Circuit activity on the '306 patent.

Let me confirm the Par petition disposition, the panel composition, and check for any later challenges to the '306 family.

Proceedings overview

On-file count: three AIA trials are documented against US 8,772,306 — IPR2016-00002 (Par), IPR2016-00024 (Ranbaxy), and IPR2016-00546 (Amneal) — even though the structured ODP block in my prompt reports none. The ODP feed is stale here; the proceedings list below is sourced from the patent's own embedded litigation metadata (which links all three PTAB case numbers) and corroborated by web search for each docket. Breakdown: 0 active · 0 claims invalidated · 0 claims sustained by written decision · 1 settled (after partial institution) · 2 institution denied on the merits · 0 final written decisions ever issued. All 34 claims remain alive and unamended — but no panel has ever reached the merits on paper, so the patent is untested rather than hardened. The three challengers were all generic ANDA filers (Par, Ranbaxy/Sun, Amneal), not a defensive aggregator, and two of the three walked away at the institution stage.

Sourcing note / flag: the "PTAB proceedings on file" block states the ODP API returns no AIA trial proceedings. That conflicts with the Google Patents litigation metadata embedded in the patent text (PTAB case IPR2016-00024 filed (Settlement); PTAB case IPR2016-00002 filed (Not Instituted - Merits); PTAB case IPR2016-00546 filed (Not Instituted - Merits)), with Docket Alarm's docket for patent:(8772306), and with DrugPatentWatch's Jazz PTAB table. I treat the ODP gap as an indexing lag, not as an absence of proceedings. Canonical source for each case: PTAB E2E, https://ptacts.uspto.gov/ptabweb/.

IPR2016-00024 — Ranbaxy, Inc. v. Jazz Pharmaceuticals, Inc.

• Type: Inter Partes Review (35 U.S.C. §§ 311–319)
• Filed: 2015-10-07 (accorded filing date 2015-10-07)
• Status: Terminated-Settled (Google Patents: Settlement). Plain English: Ranbaxy got 16 of the 34 claims into trial and then the parties killed the case 41 days later, before any merits briefing completed.
• Judge panel: Brian P. Murphy, Christopher G. Paulraj, Erica A. Franklin (APJs), Trial Center 1600
• Petition grounds: All claims 1–34 (independents 1, 11, 19, 30, 33) unpatentable under § 103. Art included the Xyrem 2005 label, the Depakote 2011 label, Cagnin, Waszkielewicz and the FDA Guidance; supporting declaration of Dr. David P. Rotella. Ranbaxy additionally argued that the "warning" step of claim 19 and the "recommending" steps of claims 19, 30 and 33 are "printed matter" not entitled to patentable weight (citing King Pharm. v. Eon Labs, 616 F.3d 1267, 1279 (Fed. Cir. 2010); In re Ngai; In re Distefano, 808 F.3d 845, 850 (Fed. Cir. 2015)).
• Institution decision: Partially instituted — 2016-04-12 (Paper 10, Decision to Institute). The panel instituted on claims 19–34 (16 claims) and declined to institute on claims 1–18, holding: "We, therefore, decline to institute an inter partes review of claims 1–18 based on Petitioner's obviousness challenges." On claims 1–18 the panel agreed with Patent Owner that the Xyrem label and Cagnin taught away from co-administering GHB and valproate. On claims 19–34, the panel accepted Ranbaxy's printed-matter argument at the threshold, since Patent Owner "does not address this issue in its Preliminary Response."
• Final Written Decision: None. No FWD was ever entered; no claim was canceled or sustained on the merits.
• Settlement / termination: Terminated 2016-05-23 by joint request after institution (Paper 15, Order Termination of the Proceeding; Paper 17 Notice re: expungement; Paper 18 Notice of Refund). Settlement terms are not public and should be treated as confidential. Ranbaxy's 2016-05-24 fee-refund request confirms the posture: termination "a mere 41 days after the April 12, 2016 institution … as to 16 of the 34 claims," with "[n]o substantive post-institution proceedings" having occurred.
• Appeal: None. With no FWD, there was nothing appealable under § 319.
• Defensive value: The only proceeding that ever moved past institution, and it evaporated on settlement — so the Ranbaxy prior-art combination has never been adjudicated, and Judicial estoppel under § 315(e)(2) never attached (estoppel requires an FWD). That cuts both ways: Ranbaxy/Sun is not estopped from re-running better art, and neither is anyone else. Also worth noting for a defendant: claims 19–34 — the concentration/pH/warning/recommendation claims — were the soft underbelly that got over the § 314(a) threshold on a printed-matter theory, precisely because those limitations read on labeling rather than on dosing.

IPR2016-00002 — Par Pharmaceutical, Inc. v. Jazz Pharmaceuticals, Inc.

• Type: Inter Partes Review
• Filed: 2015-10-06
• Status: Institution Denied (Google Patents: Not Instituted - Merits) — Decision Denying Institution of Inter Partes Review, Paper 12, 2016-04-12
• Judge panel: Not confirmed in the sources retrieved (decision issued out of Trial Center 1600 on the same day as the Ranbaxy institution decision; I could not verify the APJ names from the public record I pulled, so I am not naming them).
• Petition grounds: § 103, all claims 1–34. Par treated the five independent claims as rising and falling together on the same obviousness rationale (Pet. 28–31), relying on the Xyrem 2005 label, the Depakote 2011 label, Cagnin, Waszkielewicz and the FDA Guidance.
• Institution decision: Denied as to all 34 claims, 2016-04-12. The panel's core reasoning, verbatim from the decision: "We determine that Petitioner has not demonstrated a reasonable likelihood of prevailing with respect to any of the challenged claims." Two independent grounds for the denial:
  1. Teaching away — "As an initial matter, Petitioner does not account for the prior art's teaching away of the co-administration of GHB and valproate," pointing to "the explicit black box warning that GHB should not be taken with other CNS depressants, despite Petitioner's acknowledgement that valproate is a CNS depressant," and to Cagnin's ultimate conclusion that "[o]nly GHB discontinuation restored the neurotransmitter balance and caused seizures to stop."
  2. No reasonable expectation of success — even assuming no teaching away, "Petitioner has not demonstrated sufficiently that a skilled artisan would have had a reasonable expectation of success in treating the claimed sleep disorders with such a reduced dosage of GHB." The panel rejected the "obvious to try" framing.
• Final Written Decision: None.
• Settlement / termination: N/A — denied at the § 314(a) threshold, so no trial existed to settle.
• Appeal: None. A denial of institution is not appealable.
• Defensive value: This is the strongest defensive datum on the whole docket for the patent owner, and the weakest for a defendant. A well-supported § 103 attack on independent claims 1 and 11 was held insufficient because the prior art's own warnings discouraged co-administration. Any new petition built on the same Xyrem-label-plus-Cagnin theory will face the identical teaching-away problem. New art (see Strategic summary below) is the way through, not a repackaged version of this petition.

IPR2016-00546 — Amneal Pharmaceuticals LLC v. Jazz Pharmaceuticals, Inc.

• Type: Inter Partes Review
• Filed: 2016-02-02 (accorded filing date 2016-02-02)
• Status: Terminated-Denied / Institution Denied (Google Patents: Not Instituted - Merits) — institution decision 2016-07-28
• Judge panel: Christopher G. Paulraj (APJ, author of the decision), Erica A. Franklin and Susan L. C. Mitchell (APJs); Trial Center 1600
• Petition grounds: § 103, all claims 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34 (per the Patexia claim listing for this docket). Amneal re-used some of the Ranbaxy/Par art but buttressed it with a detailed GHB clearance-pathway analysis (MCT inhibition vs. GHB dehydrogenase inhibition) — a deliberate attempt to cure the "unpredictability" finding that had defeated Par and Ranbaxy on claims 1–18. Amneal argued drug–drug interaction screening was "routine practice."
• Institution decision: Denied. The Board did not institute. The timing aligns exactly with the published procedural history: Jazz's Preliminary Response was due 2016-05-08, and the three-month § 314(b) decision landed 2016-07-28. Amneal's petition did not overcome the same two defects the earlier panels identified — the teaching-away problem and the absence of a reasonable expectation that a reduced dose was the right answer.
• Final Written Decision: None.
• Settlement / termination: N/A — denied at the threshold.
• Appeal: None on record (institution denials are not appealable).
• Defensive value: Amneal — the most sophisticated of the three [Xyrem generics with parallel IPR programs][1] — tried to win on the biochemistry and still lost at the door. That is a meaningful signal that the § 103 case against the dosing-reduction claims (1–18) is genuinely hard, not merely a matter of picking better references from the same 2011-era corpus.

[1]: Amneal/Par were simultaneously running IPR2015-00545, -546, -547, -548, -551 and -554 plus IPR2015-01903 against the '730-family distribution patents — and won those. Do not conflate IPR2016-00546 (the '306 patent, Amneal, denied) with IPR2015-00546 (the '106 patent, Par/Amneal, instituted and won). Same digits, different patent, opposite outcome.

Strategic summary

Claim status. Every claim of 8,772,306 is UNSUSTAINED-BUT-ALIVE: claims 1–18 are UNTESTED on the merits (institution denied in IPR2016-00002 and IPR2016-00546; institution denied in IPR2016-00024); claims 19–34 are UNTESTED on the merits (instituted in IPR2016-00024, then terminated by settlement 41 days later before any response or FWD). No claim has been canceled by the PTAB, and no claim has been upheld by a Final Written Decision. Any statement that "the PTAB upheld claims X–Y" would be false — there is no written decision to cite. The patent expires 2033-03-15 and remains Orange-Book listed for Xyrem®; it is still being asserted (e.g., D.N.J. 2:23-cv-00329, 2:23-cv-01617, 2:23-cv-03182, 2:25-cv-14606, and the 2026 filings 2:26-cv-01739 / 2:26-cv-01740).

Estoppel landscape — this is the single most important point for a defendant today. Because no IPR ever reached an FWD, § 315(e)(2) estoppel never attached to anyone. Par, Ranbaxy/Sun and Amneal are free to re-petition on grounds they "raised or reasonably could have raised," and — more to the point for a new defendant — you face no estoppel from their filings at all. Conversely, you also get no free ride: nothing in the record is a validity adjudication, so there is no collateral-estoppel or issue-preclusion shortcut, and no IPR record to borrow. The entire prior-art field is open, subject only to the ordinary § 315(b) one-year bar running from service of your own complaint and § 325(e) estoppel from your own prior IPRs.

Where the openings are. The Board's denials were doctrine-driven, not art-driven: teaching away (Xyrem's black-box CNS-depressant warning; the SSADH contraindication; Cagnin's conclusion) plus failure to show a reasonable expectation of success for a reduced dose. Both are vulnerable to a different evidentiary record. Specifically: the antitrust/inequitable-conduct allegations pending against Jazz in the Xyrem MDL (N.D. Cal., No. 3:19-cv-03698 and related) assert that Jazz withheld its 2012 Xyrem label during prosecution and "affirmatively misled the PTAB" by relying only on the 2005 label to argue teaching-away. If that is right, the 2012 label — which changed the message from "should not be used in combination with … other CNS depressants" to "[t]he concurrent use of Xyrem with other CNS depressants … may increase [certain risks] … dose reduction … should be considered" — is (a) § 102/§ 103 art none of the three petitioners deployed as a primary reference, and (b) potential § 282 inequitable-conduct / unclean-hands material. That label change is the most obvious unexploded ground still on the table, and it maps directly onto the exact limitation (dose reduction on concomitant administration) the Board found non-obvious. This is a research lead to verify against the actual file wrapper, not a conclusion — I have not confirmed the 2012 label's precise effective date or its use as a reference in any petition.

Pattern signals. Three separate generic filers petitioned within five months (Par, 2015-10-06; Ranbaxy, 2015-10-07; Amneal, 2016-02-02), all against the same 2013-priority patent, all in Trial Center 1600, all as a byproduct of the Xyrem ANDA litigation — no defensive aggregator (e.g., Unified Patents) appears anywhere in the chain. Jazz's IPR strategy here was pure defense: two successful Preliminary Responses and one fast settlement, with zero appeals. Contrast the '730-family distribution patents, where the PTAB invalidated all of them on 2016-07-27 and a seven-appeal Federal Circuit wave followed — Jazz Pharms., Inc. v. Amneal Pharms., LLC, Nos. 2017-1671, -1673, -1674, -1675, -1676, -1677 and -2017-2075 (Fed. Cir. July 13, 2018), affirming invalidation. That appellate activity is about the '730 family, not the '306 patent — do not let it into a brief about 8,772,306. No Federal Circuit appeal has ever arisen from an IPR of the '306 patent, because none reached FWD.

Recommended next steps

• For a defendant being asserted against today: do not represent to a court or an adversary that 8,772,306 "survived IPR." It never went to a written decision. The accurate framing is: "Three IPR petitions were filed in 2015–2016; two were denied institution (IPR2016-00002; IPR2016-00546); the third (IPR2016-00024) was partially instituted on claims 19–34 and terminated by settlement on 2016-05-23 before any merits briefing. No claim of the patent has been adjudicated by the PTAB." Pull the four primary documents directly from PTAB E2E (https://ptacts.uspto.gov/ptabweb/): IPR2016-00002 Paper 12 (Denying Institution, 2016-04-12); IPR2016-00024 Paper 10 (Instituting, 2016-04-12); IPR2016-00024 Paper 15 (Termination, 2016-05-23); IPR2016-00546 institution decision (2016-07-28). Docket mirrors are at https://www.docketalarm.com/search/?q=patent:(8772306).
• If you are weighing a new IPR: the § 315(b) clock runs from service of the complaint on you. A new petition is not barred by the old dockets, and no estoppel runs to you from Par/Ranbaxy/Amneal. Budget for the teaching-away rebuttal on claims 1–18 specifically — that was the dispositive issue twice. The two denial orders give you the Board's own roadmap of what the petition must affirmatively address: the black-box warning, the SSADOL/SSADH-deficiency contraindication, and Cagnin's ultimate conclusion. For claims 19–34, remember the Board already credited the printed-matter theory at the institution stage — those claims are the cheaper target, though they cover labeling/recommendation conduct rather than the dosing method itself.
• Discovery/validity work-up to prioritize: (1) the prosecution file wrappers of the '306 family to test the inequitable-conduct allegations about the withheld 2012 Xyrem label — the '306 family comprises 8,772,306 (filed 2013-04-29), 9,050,302 (filed 2013-03-15) and 9,486,426 (filed 2015-05-08), all claiming priority to provisional 61/771,557 (2013-03-01); (2) whether the 2012 label was ever placed in any IPR record as an exhibit; (3) whether the '302 and '426 patents have themselves drawn petitions — I did not find any IPR numbers for 9,050,302 or 9,486,426 and am not asserting there are none, only that my searches did not surface them.
• No active PTAB proceedings to track. There are no institution deadlines, oral hearings or statutory § 316(a)(11) FWD due dates to calendar for this patent. All three dockets are closed. (The 2025–2026 D.N.J. cases listed in the patent's litigation metadata are district-court matters, not PTAB trials.)

Confidence flags: the APJ panel for IPR2016-00002 is not confirmed in the sources I retrieved and I have deliberately not named judges for that docket. The IPR2016-00546 institution-decision reasoning I have described by its alignment with the prior two denials and the published procedural timeline; I did not retrieve the text of the 2016-07-28 decision itself, so the "same two defects" characterization is inference, not quotation. Settlement terms in IPR2016-00024 are confidential and I have not guessed at them.

I'll research the assignment chain for US 8772306, cross-referencing the patent text provided with USPTO assignment records and related sources.

Let me pull additional details on the assignment correspondents, litigation timing, and Jazz's corporate status.

Assignment & Ownership Reconstruction — US 8,772,306 B1

Scope note on sourcing: The patent's own Google Patents legal-events table and a PTAB filing (Jazz v. Ranbaxy/Par/Amneal, IPR2016-00002/-00024/-00546, "Notice of Related Matters") together disclose the recorded chain with reel/frame numbers. I was not able to live-retrieve the Assignment Center correspondent-of-record fields (the assignmentcenter.uspto.gov record page is not in the fetched set), so I do not assert a correspondent name for the two core patent assignments. Where I reference an attorney, I say exactly which document it comes from. I have not fabricated reel/frame numbers or correspondents.

Inventors

Pattern note: This is a single-inventor patent, so the "all inventors departing within 12 months of filing" heuristic does not apply in the classic sense. There is no multi-inventor team dispersing. The inventor assigned his rights to Jazz promptly (recorded 2013, see below), consistent with an ordinary employee invention-assignment, not a pre-fire-sale departure.

Original assignee

• Entity on the issued patent: Jazz Pharmaceuticals, Inc. (Palo Alto, CA); original assignee per Google Patents/Unified Patents. The Xyrem asset family traces back to Orphan Medical, Inc. (Minnetonka, MN), which Jazz acquired in 2005.
• Current assignee: Jazz Pharmaceuticals Ireland Ltd (wholly-owned subsidiary), per Google Patents and CourtListener/PTAB records.
• Product embodying the claims: Yes. Jazz ships Xyrem® (sodium oxybate) (FDA-approved July 17, 2002) and Xywav® (a mixed oxybate salt, approved 2020). US 8,772,306 is listed for both. So the claims are practiced by a real commercial product, and the patent is Orange Book–listed.
• Primary line of business: Specialty biopharmaceutical company (sleep/narcolepsy, neuroscience, oncology). Publicly traded (NASDAQ: JAZZ).
• Current status: Operating, solvent, publicly traded. This is the parent of the Jazz Ireland assignee, not a licensing vehicle.

Assignment timeline

Reel/frame and recording dates below are as stated in the PTAB "Notice of Related Matters" (IPR2016) and the Google Patents legal-events table. I report the reel format literally as recorded (5-digit reel).

• Executed ~2013-07 / recorded 2013-07-09 — Reel 30836/0953

  • Conveyance: Assignment (inventor → company)
  • Assignor: Mark Eller (inventor)
  • Assignee: Jazz Pharmaceuticals, Inc.
  • Correspondent: Not retrieved — Assignment Center correspondent field not available in the fetched record. (The later IPR "Statement under 37 CFR 3.73(b)" was signed by Francis Dominic Cerrito, Reg. No. 38,100, on 2015-10-27; that is a litigation power-of-attorney filing, not the assignment correspondent, and I do not treat it as such.)
  • Context: Standard employee invention assignment — the sole inventor assigned to his employer as the patent issued. (Google Patents shows a related legal-event date of 2013-07-19, which may reflect execution vs. recordation.)

• Executed 2014 / recorded 2014-07-29 — Reel 33480/0220

  • Conveyance: Assignment
  • Assignor: Jazz Pharmaceuticals, Inc.
  • Assignee: Jazz Pharmaceuticals Ireland Ltd
  • Correspondent: Not retrieved (as above).
  • Context: Internal corporate reorganization — transfer of the patent to Jazz's Irish operating subsidiary, recorded three weeks after issuance (2014-07-08). Aligns legal title with the group's contracting entity; not an arm's-length third-party sale. (Google Patents event dated 2014-08-06.)

• 2015-06-19 — Security Agreement

  • Conveyance: Security Agreement (lien grant)
  • Assignor/Grantors: Jazz Pharmaceuticals International III Ltd, Jazz Pharmaceuticals International Ltd, Jazz Pharmaceuticals Ireland Ltd, Jazz Pharmaceuticals, Inc.
  • Assignee: Bank of America, N.A., as Collateral Agent
  • Correspondent: Not retrieved.
  • Context: Securitization / secured financing — group-wide collateral grant to a lender. Not a title transfer.

• 2021-05-05 — Security Agreement

  • Conveyance: Security Agreement
  • Assignor/Grantors: Cavion, Inc., Celator Pharmaceuticals, Inc., Jazz Pharmaceuticals Ireland Ltd, Jazz Pharmaceuticals, Inc.
  • Assignee: U.S. Bank National Association
  • Correspondent: Not retrieved.
  • Context: Securitization / refinancing — new collateral agent under Jazz's credit arrangement; reflects acquisitions (Celator, Cavion) folded into the collateral pool.

• 2021-05-05 — Release

  • Conveyance: Release by Secured Party
  • Assignor: Bank of America, N.A. (releasing party)
  • Assignee/Released: Jazz Pharmaceuticals, Inc.; Jazz Pharmaceuticals International Ltd; Jazz Pharmaceuticals International III Ltd; Celator Pharmaceuticals, Inc.; Cavion, Inc.; Jazz Pharmaceuticals Ireland Ltd
  • Correspondent: Not retrieved.
  • Context: Lien release accompanying the U.S. Bank refinancing on the same date.

No assignment to any third-party acquirer, NPE, or licensing entity appears on the record. The only true title transfers are (1) inventor → Jazz and (2) Jazz Inc → Jazz Ireland. The rest are secured-financing instruments.

Timeline diagram

timeline
    title Ownership of US 8772306
    2013 : Filed by Jazz Pharmaceuticals Inc
         : Assignment recorded Eller to Jazz reel 30836 frame 0953
    2014 : Patent issued July 8
         : Transferred to Jazz Pharmaceuticals Ireland reel 33480 frame 0220
    2015 : Security agreement to Bank of America as collateral agent
    2021 : Security agreement to US Bank
         : Release of Bank of America lien

NPE / troll-pattern signals

1. Shell-entity transfer — Not present. The only inter-entity transfer (Reel 33480/0220, recorded 2014-07-29) moves the patent from Jazz Pharmaceuticals, Inc. to Jazz Pharmaceuticals Ireland Ltd, an operating subsidiary of the same public parent — not a licensing-only LLC, and there is no registered-agent-service address or single-purpose shell in the chain. Naming is not the basis here; the entity is a documented operating company and the parent of an Orange Book drug.

2. Known asserter in the chain — Not present. Neither Jazz Pharmaceuticals, Inc. nor Jazz Pharmaceuticals Ireland Ltd appears on any NPE/asserter list (Acacia, Marathon, IV, IPNav, Wi-LAN, Conversant, Vringo, Pendrell, etc.). Jazz is the plaintiff in the ANDA litigations, i.e., the operating patentee.

3. Repeat correspondent across the chain — Unclear / no finding. I could not retrieve the Assignment Center correspondent-of-record, so I cannot test recurrence. The only attorney surfaced is Francis Dominic Cerrito (Reg. No. 38,100) on a 2015-10-27 IPR power-of-attorney/§3.73(b) statement — a litigation filing, not an assignment recording. A single appearance in a non-assignment document is expressly not a finding under the stated rule.

4. Cascading transfers — Not present. One title transfer between affiliated companies, plus two security instruments and a release. No chain of LLCs, no ≤24-month cascade of distinct assignees, no shared registered-agent addresses.

5. Pre-litigation transfer — Not present. The Jazz Inc → Jazz Ireland transfer was recorded 2014-07-29, roughly three weeks after issuance (2014-07-08) and contemporaneous with the consolidated ANDA litigation window (Jazz v. Amneal, 2:13-cv-00319, D.N.J.) — timing consistent with an intra-group housekeeping reorg, not an arrangement to create standing/venue for an NPE suit. Jazz litigates as the operating patentee in its home district (D.N.J.).

6. Bankruptcy fire-sale — Not present. No Chapter 7/11 proceedings for Jazz or the assignee; the two 2015/2021 instruments are consensual secured financings with Bank of America and U.S. Bank, not court-ordered sales.

7. Privateering — Not present. No NPE asserts this patent on Jazz's behalf; Jazz itself is the named plaintiff against generic filers.

8. Defensive aggregator — Not present. The chain does not terminate at RPX, AST, LOT, Unified Patents, or OIN.

Litigation/IPR context (for the assertion picture): IPRs on the '306 patent were filed by Par Pharmaceutical (IPR2016-00002, filed 2015-10-06, institution denied 2016-04-12), Ranbaxy (IPR2016-00024, filed 2015-10-07, settled 2016-05-23), and Amneal (IPR2016-00546, filed 2016-02-02, institution denied 2016-07-28) — i.e., generic ANDA challengers, the signature of operating-company assertion, not NPE assertion. District-court suits naming the patent include Jazz against Lupin (2:21-cv-14271) and Alkem (2:23-cv-03182) in D.N.J.

Verdict

Operating-company assertion.

The chain shows a single inventor assigning to his employer (Reel 30836/0953) and an intra-group transfer to Jazz's Irish operating subsidiary (Reel 33480/0220, recorded 2014-07-29); the remaining recorded instruments are ordinary secured-financing liens (Bank of America 2015-06-19; U.S. Bank 2021-05-05) and the corollary release — no third-party or shell assignee, no known asserter, no cascading LLC transfers, and no pre-litigation transfer. Jazz Pharmaceuticals is a publicly traded manufacturer whose Xyrem/Xywav products embody the claims and who asserts the patent directly against generic ANDA filers, so this is classic operating-company assertion, not NPE activity.

Verify at: https://assignmentcenter.uspto.gov/ (search patent number 8772306) — mirror at https://assignment.uspto.gov/patent/index.html. Confirm the correspondent-of-record on Reels 30836/0953 and 33480/0220, which I could not retrieve directly and therefore did not report.

I'll research US Patent 8772306 and identify the prior art references cited against it.

Let me look for the specific "References Cited" on the face of the patent and the IPR/reexam prior art.

Let me retrieve the patent's front-page "References Cited" and the full prior-art list from the litigation/IPR record.

US 8,772,306 B1 — Prior-Art Analysis

1. Patent identification (as verified)

Important scope caveat before the reference list. The claims are method-of-treatment claims directed to adjusting the GHB dose when the patient also receives an MCT inhibitor — chiefly a ≥5% decrease with valproate (claim 1), formulation + warning + "recommending reducing the dose ≥15%" claims (e.g., claim 19), plus diclofenac (increase), ibuprofen (discontinue), aspirin embodiments, and pharmacy-distribution/REMS claims.

Two consequences you must keep in mind:

1. None of the prior art asserted against this patent was a pure § 102 anticipation. Every published challenge (PTAB IPR2016‑00002, ‑00024, ‑00546; DJ/ANDA pleadings) was framed as § 103 obviousness. So the table below maps each reference to the claim(s) it was aimed at, but the honest legal characterization is "obviousness-mapped," not "anticipatory." I flag any reference that could arguably be a § 102 reference on its face.
2. The front-page "References Cited" (56) list is dominated by applicant-submitted IDS art on GHB formulations, pharmacy-management/REMS systems, and general GHB pharmacology — i.e., largely background, not the art that actually attacked the claims. The legally most relevant art is the small group of valproate/GHB-metabolism references developed in the IPRs. I separate the two.

2. Group A — References on the face of the patent (56) "U.S. Patent Documents"

Retrieved from the patent's front page. OCR of the printed front page is imperfect, so number↔name pairings are as they appear; I flag the one internal inconsistency.

U.S. Patent Application Publications cited (face): 2003/0171270 A1; 2006/0018933 A1; 2008/0293698 A1; 2009/0137565 A1; 2010/0112056 A1; 2010/0160299 A1; 2011/0237664 A1; 2012/0076865 A1 — GHB formulations, salts, and REMS/distribution disclosures.

Non-patent "Other Publications" on the face (general GHB pharmacology, largely background): Laborit 1973; Ladinsky 1983; Lammers 1993; Lapierre 1988/1990; Lee 1977; Lettieri & Fung 1978; Maitre 1990/2005; Mamelak 1973/1977/1979/1981/1989; Modi 2007; Nema 1997; Palatini 1993; Roth & Giarman 1966; Scharf 1985; Scrima 1987/1989/1990; Strong 1984; van den Bogert; Vickers 1969; Yamada 1967; Wu 2004; Cash 1999; Waszkielewicz 2004; Kuriyama 1971; Dimitrijevic 2005; Banerjee 1995; Hechler 1991; Smolders 1995.

§ 102 assessment of Group A: None of these front-page references, standing alone, anticipates any of claims 1–34. They disclose GHB formulations, salts, distribution systems, and GHB pharmacology generally — not the claimed step of adjusting the GHB dose because of concomitant valproate/diclofenac. At most, Cook (6,780,889 / 7,262,219) and the Reardan REMS patents supply structural/administrative elements that were combined in § 103.

3. Group B — The legally most relevant prior art (the actual § 103 challengers)

These are the references assembled by Ranbaxy, Par, and Amneal in the 2015–2016 IPRs, and in the parallel New Jersey ANDA litigation. All were asserted in § 103 combinations; the PTAB twice refused institution (IPR2016‑00002, IPR2016‑00024), and denied institution in IPR2016‑00546 — precisely because no reference taught a reduced GHB dose for concomitant valproate.

Reference asserted against the later family member (contrast): In the international search for WO 2021/209956, the examiner cited US 8,772,306 B1 itself (X category, "claims 1, 8, 10") — i.e., your patent now functions as prior art against later Jazz filings, not the reverse.

4. Bottom line on § 102 anticipation

• No reference in the '306 file was asserted as a § 102 anticipation of any claim. All challenges were § 103, and the PTAB denied institution twice (Ranbaxy IPR2016‑00024 and Par IPR2016‑00002) on claims 1–18, holding that (a) the art taught away from GHB+valproate co-administration, and (b) there was no reasonable expectation that a reduced GHB dose would treat the claimed sleep disorders. Institution was granted only for claims 19–34 (which recite "warning" and "recommending" steps), on the "printed-matter" doctrine — not § 102.
• The closest thing to a § 102-style single-reference attack was Maitre (1997) + the Xyrem PI, but even that required combination, i.e., § 103.
• If you want the single most on-point references for a validity/invalidity assessment, they are: Maitre (1997), the Xyrem prescribing information/titration schedule, Cagnin, Hechler (1997), Kaufman (1987), Vayer (1987/1988), and Cook U.S. 6,780,889 (for the claim-19 formulation).

I did not independently re-verify the issue dates of every face-of-patent U.S. document from primary USPTO records in this session; the numbers and the § 103/§ 102 characterization above are drawn from the patent's front page and the IPR/litigation record. If you need certified bibliographic data (exact issue dates for Group A items A5–A28), I'd pull each from USPTO Patent Center/Global Dossier rather than rely on front-page OCR.

I'll ground this analysis in the actual prior-art record for the '306 patent rather than speculating. Let me pull the cited references and the IPR grounds.

I have the IPR record. Let me confirm the identities of the key prior-art exhibits so I attribute them accurately.

Obviousness Analysis — U.S. Patent 8,772,306 under 35 U.S.C. § 103

Sourcing note / caveat. The authoritative full text you supplied for the '306 patent does not contain a discrete, machine-readable "Prior Art" citation list (the page text you pasted runs from the bibliographic block through the specification and is truncated mid‑"Detailed Description"). I therefore grounded the prior-art universe below in the actual prior-art record of this patent — namely the references relied on in the IPR petitions against the '306 patent (Ranbaxy IPR2016‑00024; Par IPR2016‑00002; Amneal IPR2016‑00546) and the PTAB decisions on those petitions. Those references are the same ones that populate a Google Patents "Prior Art / Citations" listing, so this is the correct corpus to analyze. Where I could not verify a reference's exact identity, I say so.

Important cross-reference flag. The prior generated summary stated the '306 has 34 claims with independents 1, 11, 19, 30, 33. The search record confirms this and adds that PTAB denied institution as to claims 1–18 but instituted as to claims 19–34 (the "warning/recommending" claims), reasoning turning on the printed-matter doctrine. That nuance changes the obviousness posture between the two claim groups and is developed in §V below. I also flag one tension: the earlier summary called IPR2016‑00024 merely "terminated–settled"; the record shows it was instituted in part (claims 19–34) and then terminated on settlement — the settlement did not vindicate the claims on the merits.

I. Legal framework

Under Graham v. John Deere, obviousness is assessed from (1) the scope and content of the prior art, (2) the differences between the prior art and the claims, (3) the level of ordinary skill, and (4) objective indicia. Under KSR Int'l v. Teleflex, a combination is obvious where the elements were known, the combination was "no more than predictable use of prior-art elements according to their established functions," or where the claimed approach was "obvious to try" from a finite number of identified, predictable solutions. A reference teaches away where it "criticizes, discredits, or otherwise discourages" the modification (In re Fulton). Critically for a method-of-treatment claim, the challenger must show a reasonable expectation of success — that the POSA would have expected the selected doses to be effective for the claimed therapeutic purpose (DePuy v. Medtronic).

Level of ordinary skill (POSA). As defined in the Par petition and adopted by the Board: a person with a Ph.D., Pharm.D., or M.D. and ~5 years' experience treating neurologic disorders (including narcolepsy, cataplexy, EDS), or a clinical pharmacologist with ~3 years advising on drug dosing in light of drug–drug interactions.

The claimed subject matter (to be tested).

• Claim 1 (independent): A method of treating EDS, cataplexy, sleep paralysis, apnea, narcolepsy, sleep-time disturbances, hypnagogic hallucinations, sleep arousal, insomnia, or nocturnal myoclonus with GHB or a salt, comprising orally administering to the patient at least a 5% decrease in an effective dosage amount of GHB/salt when the patient is receiving concomitant valproate (acid, salt, or mixture).
• Claims 2–4 etc. (dependent): ≥ about 15% reduction; specific ranges (5–10%, 10–15%, … 45–50%); divalproex sodium as the valproate species; some claims add aspirin.
• Claim 19: narcolepsy; administer GHB salt formulation 350–750 mg/mL, pH 6–10; determine if patient is also on valproate; warn of a potential interaction; recommend reducing the dose ≥ 15%.
• Claims 30/33: recommend a ~20% decrease in the starting dose, to an adjusted dose of ~3.5–4 g (claim 30) or ~3.6 g/night (claim 33).

II. Scope and content of the prior art (of record)

(Reference-identity caveat: I could not, within the search budget, positively confirm which publications Ranbaxy designated Ex. 1003/1004 and the "Cook patent." The substantive teachings are quoted from the petitions and decisions; the bibliographic handles for those specific exhibits should be verified against the exhibit list before being relied on formally.)

III. The differences between the prior art and the claims

The gap the patent had to bridge is narrow on its face:

1. Treating the enumerated sleep disorders with GHB/salt — fully disclosed by the Xyrem 2005 PI and the clinical literature.
2. Oral administration — disclosed.
3. Concomitant valproate — the Depakote label, Cagnin, and the dehydrogenase literature place a valproate-treated patient in the same clinical space as a GHB-treated patient.
4. A ≥5% dose decrease — the single point of novelty. The prior art supplies the reason (valproate raises GHB exposure / potentiates CNS depression) and the mechanism for executing a reduction (the PI's own titration schedule and the "reduce by one-half in hepatic impairment" instruction).

So the entire §103 question collapses to: would a POSA, aware of the valproate–GHB metabolic interaction and of the PI's dose-adjustment machinery, have found it obvious to give a modestly reduced GHB dose (≥5%) to a valproate patient?

IV. Obviousness combinations and motivations to combine

Ground 1 — Xyrem 2005 PI + Depakote 2011 Label + Cagnin

• Combination: A POSA treating narcolepsy/EDS with Xyrem who learns the patient also takes Depakote (valproate).
• Motivation: Both labels identify CNS/respiratory depression as the shared hazard; the Xyrem PI expressly contemplates co-use with CNS depressants only with caution and directs dose consideration. Cagnin supplies a concrete, severe adverse outcome (resistant seizures/psychosis) from the combination. A POSA seeking to keep the patient on both therapies has an articulated reason to mitigate additive CNS depression by lowering the GHB dose.
• Result recited: A ≥5% oral GHB reduction during concomitant valproate.

Ground 2 — Ground 1 + Weiss (and the GHB-dehydrogenase references)

• Combination: Add Weiss/Cash/Bernasconi/Hechler, teaching that valproate inhibits GHB dehydrogenase, raising GHB levels.
• Motivation (stronger): This converts the rationale from generic "additive sedation" to a specific pharmacokinetic mechanism — valproate raises systemic/brain GHB, so the prudent, mechanism-driven response is to lower the exogenous GHB dose to keep exposure in the therapeutic window. Weiss's caution to use such combinations "with great care" is naturally read as dose management. This directly supports claim 1's ≥5% reduction and the range claims.

Ground 3 — Ground 2 + Xyrem titration schedule ("routine optimization")

• Combination: Apply the PI's own 1.5 g titration increments to a compliant dose reduction.
• Motivation & arithmetic: Reducing a 9 g/night dose by one 1.5 g increment → 7.5 g/night (16.7%); 7.5→6 g (20%); 6→4.5 g (25%); 4.5→3 g (33.3%). Every one of these exceeds claim 1's 5% floor and lands inside the dependent-claim ranges (e.g., claim 2's "≥15%," claim 4's 15–20%/20–25% bands). The PI's instruction to "reduce the starting dose … by one-half" in a metabolism-compromised state (hepatic impairment) supplies a concrete template. Claim 30/33's ~20% starting-dose reduction to ~3.5–4 g tracks both the 1.5 g decrement and the 3.5 g/night dose actually used in Cagnin. Under KSR, selecting one of a finite number of predictable reductions is "obvious to try."

Ground 4 — The "warning/recommending" claims (19, 30, 33) and printed matter

• For claims 19/30/33, the only additions over claim 1 are informational steps — "determining … valproate," "warning of a potential drug/drug interaction," and "recommending reducing the dose ≥15%/20%."
• Applying In re Distefano, informational content lacking a functional/structural relation to the invention is given no patentable weight. Strip the printed-matter limitations and these claims reduce to "administer GHB (at a stated 350–750 mg/mL, pH 6–10 concentration) and reduce the dose by ≥15–20% when the patient also takes valproate" — i.e., the same method as claim 1 with different numbers. That is why the PTAB instituted on claims 19–34 while denying on 1–18: the warning/recommendation language, once discounted, leaves an obvious dosing method over Grounds 1–3 (the 350–750 mg/mL and pH 6–10 formulation parameters are conventional Xyrem/Xywav ranges discussed in the specification itself and in U.S. Pat. Nos. 8,263,650 / 8,324,275).
• Aspirin-dependent claims: Ranbaxy argued the art taught aspirin increases free valproate, so a POSA would be even more motivated to reduce GHB in an aspirin-plus-valproate patient — an additional, independent reason to reach those dependent claims.

Motivation-to-combine summary

The Xyrem and Depakote labels share an overlapping patient population (narcolepsy, epilepsy, bipolar disorder co-occur and share anticonvulsant/CNS-depressant therapies); both flag CNS depression as the key safety risk; the scientific literature provides a specific mechanism (GHB-dehydrogenase/SSADH inhibition raising GHB); the Xyrem PI provides the tool (titration increments; halve-in-hepatic-impairment rule); and FDA DDI guidance provides the expectation that such adjustments are routine. That is a textbook KSR motivation chain.

V. The counterarguments that actually defeated claims 1–18 (and where they are vulnerable)

The PTAB denied institution as to claims 1–18 for two reasons — each of which is an evidentiary failure in the petitions rather than a legal bar:

1. Teaching away. The Board found the Xyrem black-box warning ("should not be used in combination with … other CNS depressants") plus Cagnin's "only GHB discontinuation restored … balance" criticizes/discredits co-administration, and that the petitions never explained why a POSA would ignore the warning and co-administer "even at a reduced dose" (In re Fulton; KSR).

   • Vulnerability: This defense depends on which Xyrem label is treated as prior art. The 2012 label revision — Jazz's own document — retracts the flat "should not be used" language and substitutes "dose reduction … should be considered." A challenger who properly frames the 2012 label as §102(b) prior art (published Dec. 2012, before the March 1, 2013 priority date) neutralizes the teaching-away argument: the sponsor's own contemporaneous labeling endorses dose modification rather than abstinence. (This is precisely the theory advanced in later Jazz/Amneal litigation alleging the 2012 label was withheld during prosecution and from the PTAB.)

2. No reasonable expectation of success / unpredictability. The Board credited Jazz's evidence that GHB is cleared by multiple pathways, that valproate's MCT-inhibition effect can increase renal clearance (lowering GHB) while its dehydrogenase-inhibition effect raises GHB, and that the petitions never showed a POSA would expect a reduced dose to remain effective for the claimed sleep disorders without side effects (DePuy). The Board rejected "routine optimization" because of this unpredictability.

   • Vulnerability: The Board's own framing ("at the very least further experimentation") is the KSR "obvious to try" standard — routine PK titration of a drug whose PI already mandates titration. And a challenger can rebut the unpredictability theme with the direction of the interaction: valproate's net effect (per the 2012 label and the Drugs.com/label-derived summary: a ~25% mean increase in systemic exposure) is upward, which supports a downward dose adjustment.

Net assessment: On the current record, the strongest §103 case is against claims 19–34 (instituted; printed-matter reasoning strips the warning/recommendation limitations) built on Grounds 1–3. Claims 1–18 are the harder target: they require affirmatively overcoming (a) the teaching-away effect of the 2005 label/black-box warning and (b) the unpredictability of the net metabolic effect — the two grounds on which Ranbaxy and Par failed. Introducing the 2012 label and a quantitative net-exposure (≈25% increase) teaching materially strengthens the obviousness attack on claims 1–18 as well, because it both removes the teach-away and supplies the dose-reduction rationale the Board found missing.

VI. Objective indicia to weigh (Graham factor 4)

Relevant secondary considerations on this record cut both ways and should be addressed:

• Unexpected results / skepticism: Jazz's PTAB evidence of unpredictable, bidirectional valproate effects is the counter-indicium the Board credited.
• Commercial success / long-felt need / licensing: Xyrem/Xywav are blockbuster products with extensive Orange-Book listing and serial ANDA litigation — but commercial success must be nexused to the claimed dose-adjustment; much of the success is attributable to the underlying GHB therapy (disclosed in the prior art), not the ≥5% reduction step.
• Industry recognition/copying: The eventual adoption of "reduce by at least 20%" in the Xyrem/Xywav labels (as reflected in the Drugs.com professional monographs) arguably reflects the claimed method's uptake, which cuts toward obviousness-in-practice but can also be argued as praise/copying.

VII. Conclusion

• Claims 19, 30, 33 (and dependent claims 20–34): Strong §103 case. Combining the Xyrem PI (titration + halve-in-hepatic-impairment), the Depakote label, Cagnin, and the GHB-dehydrogenase literature (Weiss/Cash/Bernasconi/Hechler), with the printed-matter limitations discounted under In re Distefano, renders the dosing/warning method obvious. The PTAB's own partial institution on this group reflects that.
• Claims 1–18: The obviousness case is real but was not carried on the petitions as filed, because they failed to (i) neutralize the 2005-label teaching away and (ii) establish a reasonable expectation of success for a reduced dose. Those two gaps are curable — chiefly by using the 2012 Xyrem label and the quantified net PK effect (~25% exposure increase) — which would materially strengthen the §103 challenge to claims 1–18.
• Confidence: High on the claim text, the reference set, the motivation chains, and the PTAB's stated denying rationales (all quoted from the decisions/petitions). Flagged lower confidence: the exact bibliographic identities of Ranbaxy Ex. 1003/1004 and the "Cook patent," and which specific IPR docket formally instituted on claims 19–34 (record says institution on 19–34 occurred alongside the April 12, 2016 denials on 1–18, and the Ranbaxy matter then settled). No 2026 CAFC docket on the '306 patent is confirmed by the record.