Federal Circuit Rejects Obviousness Presumption

Patent lawyers love shortcuts the way everyone else loves shortcutsexcept ours come with footnotes, expert declarations,
and the occasional existential crisis over what a “person having ordinary skill in the art” would do before their first
coffee. So when someone says, “Relax, there’s a presumption of obviousness,” you can practically hear the litigation
budgets sigh with relief.

Except… the U.S. Court of Appeals for the Federal Circuit has been reminding everyone that “presumption” is not a magic
spell. In a recent decision involving a pharmaceutical dosing regimen, the court refused to hand the challenger an
automatic win just because numbers in the prior art overlapped with numbers in the claims. In other words: you don’t
get to shout “OVERLAP!” and walk out with the patent’s lunch money.

This article breaks down what it means when the Federal Circuit “rejects” an obviousness presumption, why overlapping
ranges aren’t a universal cheat code, and how this shift affects patent drafting, prosecution, and litigation strategy.
We’ll keep it rigorous, but we’ll also keep it humanbecause if you can’t laugh a little while reading about 35 U.S.C.
§ 103, the statute wins.

Obviousness 101: The Rule, the Myth, and the “It Depends”

Obviousness lives in 35 U.S.C. § 103, which asks whether the claimed invention “as a whole” would have
been obvious to a person of ordinary skill in the art at the relevant time. It’s not about whether the invention is
simple in hindsight (everything looks obvious once you know the answer). It’s about whether the path to the invention
was one a skilled artisan would have taken based on what the prior art taught at the time.

The Supreme Court’s Graham framework sets the factual building blocks (scope/content of prior art, differences,
level of skill, plus objective indicia/secondary considerations), and KSR warned courts not to turn obviousness
into a rigid, formulaic test. Translation: obviousness is structured, but it is not supposed to be robotic.

In litigation, there’s another critical reality check: issued patents are presumed valid, and invalidity must generally
be proven by clear and convincing evidence. So even if a court talks about a “presumption” or burden-shifting
inside an obviousness analysis, the ultimate burden to prove obviousness doesn’t quietly teleport to the patentee.

The “Overlapping Ranges” Shortcutand Why It’s Not Automatic

One common obviousness pattern is the overlapping ranges situation: the prior art discloses a numerical range,
and the claim’s range overlaps or sits inside it. In prosecution (and sometimes in validity challenges), that overlap can
establish a prima facie case: “If you claimed something inside what we already taught, why isn’t that just routine
optimization?”

Federal Circuit precedent (and USPTO guidance) recognizes that overlap can be powerful evidence. But it’s best understood
as a starting point, not a finish line. The key question is whether the claimed range looks like the product of
routine experimentationi.e., the artisan had reason to optimize within the disclosed range and a reasonable expectation
of success in doing so.

And here’s the nuance that matters: overlap by itself is a fact. Obviousness is a legal conclusion drawn from multiple facts.
If the invention is more than “pick a number within the range,” courts are increasingly skeptical of treating overlap as an
automatic presumption that short-circuits the rest of the analysis.

The Headliner: A Dosing-Regimen Case Where “Overlap” Wasn’t Enough

The decision getting the most buzz involved a schizophrenia treatment regimen and an obviousness attack that leaned heavily
on numerical overlap. The challenger argued that because the prior art disclosed dose amounts within or near the claimed ranges,
the court should apply a presumption of obviousness.

The Federal Circuit didn’t buy it. The court emphasized that the supposed presumption depends on factual premises that must
actually be proven in contextlike motivation to modify the prior art and a reasonable expectation of success. The claimed invention
wasn’t just a single value plucked from a known range; it involved a multi-step dosing sequence with clinically significant
choices about loading doses and administration that the prior art did not clearly teach or suggest.

Put simply: when the claim is an integrated regimen (a sequence with timing and dose relationships that matter),
you can’t treat it like a math problem where you circle overlapping numbers and declare victory. The court required a full,
fact-intensive obviousness analysis rather than a shortcut.

Why the Federal Circuit’s reasoning matters

This wasn’t the court saying overlap is irrelevant. It was the court saying overlap doesn’t automatically do the whole job for you.
If you want the benefit of an overlap-based presumption, you still have to show that a skilled artisan would have been prompted to
optimize the regimen in the claimed wayand would have reasonably expected the claimed approach to work.

That distinction is huge in pharma and biotech, where “try a different dose” can range from “routine tinkering” to “please don’t do that,
patients are involved.” When physiology, safety, efficacy, and dosing dynamics enter the room, pure numerics often stop being dispositive.

Contrast Case: When the Presumption Does Apply

To see the boundaries, it helps to compare situations where the Federal Circuit has applied an overlapping-range
burden-shifting framework. In cases like In re Peterson (prosecution context) and later decisions involving composition
ranges, the court has treated overlap as establishing a strong prima facie showingespecially where the claimed range appears
to be a subset of what the art already suggested, and the applicant doesn’t show the range is critical or yields unexpected results.

Likewise, in some post-grant and PTAB settings, the court has faulted the Board for not applying the overlapping-range framework
when the facts genuinely fit the classic pattern (a known composition range, a claimed sub-range, and no persuasive evidence that the
selected slice does something surprisingly different).

So the message is not “presumption is dead.” The message is “presumption is conditional.” Courts will ask: is this really
a routine-optimization range case, or is the claim an engineered combination/sequence where overlap is only one piece of the story?

Strategy Implications: What Litigators Should Do Differently

1) Don’t argue overlap in a vacuumbuild the “why”

If you’re challenging validity, treat overlap as your opening exhibit, not your closing argument. You need a narrative:
what problem was the skilled artisan solving, what options did the art suggest, and why would the artisan have been motivated to
land on this particular range/sequence with a reasonable expectation of success?

2) Expect courts to scrutinize regimen claims like systems, not single dials

In multi-step method claimsespecially dosing regimenscourts may view “numerical overlap” as an incomplete description of the invention.
Timing, sequencing, route/site of administration, and clinical rationale can be the inventive core. When that’s true, overlap won’t
substitute for showing that the prior art taught the claimed structure of decisions.

3) Secondary considerations can still matter, but nexus is the gatekeeper

If you’re defending a patent, objective indicia like unexpected results, long-felt need, and commercial success can be powerful
but only if you can connect them to what the claims actually require. Federal Circuit case law on the
presumption of nexus underscores that you don’t automatically get credit for commercial success when the product includes
significant unclaimed features. If you want secondary considerations to carry weight, show that the product is coextensive with the claims,
or explain why unclaimed features are not driving the results.

Drafting & Prosecution Tips: How to Reduce “Presumption” Risk

1) Explain the technical logic for the rangedon’t just list it

A bare range can look like the result of routine optimization. But a range tied to a mechanism, safety constraint, stability window,
pharmacokinetics/pharmacodynamics relationship, or manufacturing reality looks less like “pick a number” and more like “solve a problem.”
Spell out why endpoints exist and what breaks outside them.

2) Add dependent claims that encode the real invention

If the invention is a sequence, include claim sets that lock in timing relationships, administration sites, patient subpopulations,
or clinically meaningful constraints. A challenger is far less likely to win with “overlap!” when the claim reads like choreography
rather than a single slider.

3) Preserve evidence of unexpected results early

If you have data showing surprising efficacy, reduced side effects, improved compliance, or stability advantages, capture it.
Overlapping-range frameworks are often rebutted by evidence of criticality or unexpected properties. Waiting until litigation to
organize this story is… how you end up paying for very expensive “re-discovery” of your own science.

What This Means for the Bigger Obviousness Landscape

The Federal Circuit’s refusal to hand out an automatic obviousness presumption fits a broader theme: obviousness is deeply factual,
and courts are wary of doctrinal shortcuts that replace proof with slogans.

That theme also harmonizes with the Supreme Court’s direction in KSR: avoid rigid rules, use common sense, but still ground the analysis
in evidencenot hindsight. For range-based patents, the future likely looks like this:

• Classic composition sub-range cases may still trigger a strong prima facie showing.
• Complex method/regimen claims will be analyzed holistically; overlap may be necessary but not sufficient.
• Motivation and expectation of success will keep doing heavy lifting (because courts want real reasons, not retroactive logic).
• Secondary considerations remain available, but the nexus fight will continue to be the bouncer at the door.

Practical Examples: How the Arguments Change

Example A: The “It’s Just a Sub-Range” attack

Prior art teaches: “Use 10–100 mg.” Claim recites: “Use 40–60 mg.” If the art suggests optimizing dose for a known outcome and there’s no evidence
of criticality, courts and the USPTO often treat this as a strong obviousness setup. The patentee’s best response is typically:
(1) prove unexpected results at 40–60 mg, (2) show the art teaches away from that window, or (3) show a constraint the art didn’t appreciate.

Example B: The “Numbers Overlap, Therefore Obvious” bluff

Prior art mentions multiple dosing values, but the claim requires a specific sequence (e.g., a high first loading dose followed by a lower second dose,
with particular timing and administration constraints). Here, the challenger must do more than map values; they must show why the skilled artisan would
have assembled the sequence and reasonably expected it to succeedespecially if the regimen departs from conventional practice.

Conclusion: Presumptions Don’t Replace Proof

“Presumption of obviousness” can be a useful label for a familiar patternoverlap suggests routine optimization. But the Federal Circuit’s recent
rejection of an automatic presumption underscores a core principle: obviousness is not a spreadsheet exercise.

If the invention is truly a classic overlapping-range scenario, challengers may still build a strong case. But when claims reflect integrated choices
sequences, constraints, clinical reasoning, or technical tradeoffscourts will demand the full obviousness analysis: motivation, expectation of success,
and careful treatment of real-world evidence.

In patent law, shortcuts are allowed. But they’re not guaranteed. And the Federal Circuit just reminded everyone: you still have to show your work.

Field Notes: of Lived-Through-It Experience (Without Naming Names)

If you’ve ever sat in a war room while someone cheerfully says, “Don’t worryoverlapping ranges, easy obviousness,” you already know how this movie
sometimes ends: with a judge asking a painfully simple question that collapses the whole shortcut, like, “Why would anyone do that?”

Here’s the pattern I’ve seen play out again and again in range/regimen fights. The challenger starts with a clean, elegant chart:
prior art range on the left, claim range on the right, a satisfying shaded overlap in the middle. Everyone nods. Someone says,
“It’s just optimization.” And then the science people in the room start making the faceyou know the facebecause “optimization” in the real world
often means “try ten things, watch nine fail, and spend months figuring out which failure mode is least catastrophic.”

In mechanical or electrical cases, routine optimization can be genuinely routine. You tweak a parameter, test performance, iterate. But in chemical,
biotech, and especially dosing regimens, the parameter you tweak is attached to a human body, complex kinetics, and safety constraints. The prior art’s
“range” might be more like a vague neighborhood than a clear street address. And a method claim isn’t just “pick 50 mg”; it’s “pick 50 mg, then follow
it with another dose on a schedule, possibly through a particular route, to achieve a specific clinical outcome.” That’s choreography, not a dial.

The most effective obviousness challenges I’ve watched succeed were the ones that treated overlap as a clue and then built the story: a real motivation
grounded in what practitioners were actually trying to solve, plus evidence that the claimed tweak would predictably work. The weakest challenges were
the ones that assumed overlap was self-explanatory. Judges are allergic to hindsight, and nothing screams hindsight like “Obviously you’d choose the exact
numbers that turned out to be best.”

On the defense side, I’ve seen patentees win by doing something surprisingly basic: explaining the “why” with discipline. Not marketing. Not vibes.
The why. Why that first step must be high. Why the second step must drop. Why a different sequence fails. Why the art wouldn’t have combined those ideas.
And when they had dataespecially unexpected resultsthey treated it like a keystone, not an afterthought. But the best teams also respected the nexus
problem: they tied outcomes to the claimed features, not to the product’s entire greatest-hits album of unclaimed improvements.

The Federal Circuit’s stance feels familiar because it matches what works in practice. Overlap can open the door. It doesn’t get you through security,
past the velvet rope, and onto the dance floor. For that, you still need the full ID: motivation, expectation of success, and evidence that holds up
when hindsight is gently removed from the room.