Cross-Field Prior Art and Hindsight: U.S. Analogous-Art Doctrine and the EPO Problem-Solution Approach

Executive Summary: This article compares U.S. analogous-art doctrine with the EPO problem-solution approach through Sanofi-Aventis Deutschland GmbH v. Mylan Pharmaceuticals Inc., where the Federal Circuit rejected an obviousness challenge because the petitioner linked a cross-field automotive reference to another prior-art reference rather than to the claimed drug-delivery invention. Under U.S. law, each relied-upon reference must be in the same field of endeavor as the claimed invention or reasonably pertinent to the inventor’s problem, and that threshold remains distinct from the separate requirement to explain why a skilled artisan would combine the references. EPO practice reaches related anti-hindsight concerns through different mechanisms: selection of a suitable starting point, objective formulation of the technical problem, assessment of whether knowledge from another field was realistically available to the skilled person, and the “could–would” inquiry. Although both systems permit reliance on cross-field teachings, neither allows the path to the claimed invention to be reconstructed solely with hindsight; challengers must provide a coherent, evidence-based reason why the skilled person would have considered and applied the remote teaching.

An automobile clutch bearing may seem an unlikely source of prior art for a patent directed to an adjustable-dose drug-delivery device. Yet that was the combination presented in Sanofi-Aventis Deutschland GmbH v. Mylan Pharmaceuticals Inc., where the petitioner relied on an automotive clutch-bearing reference to supply snap-fit engagement grips missing from its other asserted references. The Federal Circuit did not hold that automotive technology was categorically too remote. It reversed because the petitioner compared the automotive reference to another prior-art reference rather than to the challenged patent. Sanofi-Aventis Deutschland GmbH v. Mylan Pharmaceuticals Inc., 66 F.4th 1373, 1377–82 (Fed. Cir. 2023).

That error invites comparison with the European Patent Office’s problem-solution approach. Under EPO practice, an inventive-step analysis ordinarily begins with a selected prior-art starting point, identifies the objective technical problem arising from the differences between that starting point and the claim, and asks whether the skilled person would have arrived at the claimed solution. At first glance, Mylan’s reliance on a problem found in its principal reference might therefore appear to resemble European practice.

The resemblance should not be overstated. The Federal Circuit did not say that Mylan improperly imported European law, and the EPO does not apply the Federal Circuit’s two-part analogous-art test. Nor would a comparison between two prior-art references, standing alone, complete a proper EPO inventive-step analysis. Where a secondary teaching comes from a different or remote field, EPO practice requires an explanation of why the objective technical problem would have made that field, teaching, or common general knowledge realistically available to the skilled person. It then asks whether the prior art as a whole would—not merely could—have prompted the skilled person to modify the starting reference in the claimed manner.

The two systems thus use different doctrinal machinery, but they address a related concern: preventing knowledge of the invention from supplying the missing path through the prior art.

I.             Sanofi and the U.S. Comparator Rule

The patent in Sanofi, U.S. Patent No. RE47,614, concerned a drug-delivery device configured to accommodate different cartridges and permit the setting of different dose sizes. Among other limitations, the claims required a spring washer having at least two fixing elements configured to fix the washer axially and rotationally relative to the device housing. Sanofi, 66 F.4th at 1375–76.

Mylan challenged the claims in inter partes review using three references. Burren disclosed springs in a drug-delivery device. Venezia supplied a spring-washer teaching. De Gennes, an automobile clutch-bearing patent, disclosed snap-fit engagement grips that Mylan proposed using to secure the spring washer. In its petition, Mylan argued that de Gennes, although directed to a clutch bearing, addressed a problem analogous to the problem addressed by Burren—axial fixation and support of two components relative to one another. Its expert repeated the same comparison. Id. at 1376.

The Patent Trial and Appeal Board accepted a broadly framed problem of axially fixing two components relative to one another and held that de Gennes was reasonably pertinent to that problem. Because the parties agreed that the challenged patent and de Gennes occupied different fields of endeavor, reasonable pertinence was the only available route for treating de Gennes as analogous art. Id. at 1376–77.

The Federal Circuit reversed. U.S. law recognizes two tests for analogous art: whether the reference is from the same field of endeavor as the claimed invention, regardless of the problem addressed, or whether it is reasonably pertinent to the particular problem with which the inventor was involved. In re Bigio, 381 F.3d 1320, 1325 (Fed. Cir. 2004); In re Clay, 966 F.2d 656, 658–60 (Fed. Cir. 1992). The second test asks whether the subject matter of the reference logically would have commended itself to the attention of an ordinarily skilled artisan considering the inventor’s problem. Clay, 966 F.2d at 659.

Both tests require a comparison with the challenged patent or claimed invention. Analogous-art doctrine defines the body of prior art relevant to the obviousness inquiry; a document’s similarity to another asserted reference does not establish that the document belongs within that body of art. Sanofi, 66 F.4th at 1377–80; see also Donner Technology, LLC v. Pro Stage Gear, LLC, 979 F.3d 1353, 1358–61 (Fed. Cir. 2020).

That distinction is substantive, not semantic. Two references may address similar problems even though one or both would not reasonably have been considered by an artisan confronting the problem addressed by the claimed invention. Conversely, references used in a § 103 combination need not be analogous to one another. Each relied-upon reference must instead be analogous to the claimed invention. See MPEP § 2141.01(a).

Mylan’s presentation did not make that connection. Its petition and expert testimony characterized de Gennes as analogous to Burren. When Sanofi challenged the omission, Mylan continued to insist that Burren supplied the pertinent problem and criticized Sanofi for relying on the objectives of the challenged patent. The Board then effectively worked backward from Mylan’s Burren-centered problem statement to relate the problem to the ’614 patent. Because the petitioner bore the burden of proving unpatentability, the Board could not construct the missing theory for it. Sanofi, 66 F.4th at 1380–82; see also In re Magnum Oil Tools International, Ltd., 829 F.3d 1364, 1375–81 (Fed. Cir. 2016).

The court’s holding does not mean that automotive references are categorically unavailable against medical-device claims. Nor does it require an artisan to remain within rigid industry classifications. It requires the challenger to connect each reference to the claimed invention through one of the two recognized analogous-art tests.

II.          U.S. Doctrine After Sanofi: Breadth, Procedure, and Combination

Although Sanofi supplies a clear comparator rule, the broader doctrine remains flexible. The Supreme Court has cautioned that familiar technologies may have uses beyond their primary purposes and that obviousness should not be confined by formalistic rules. KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–21 (2007). The Federal Circuit similarly has instructed that the scope of analogous art should be construed broadly in light of KSR. Wyers v. Master Lock Co., 616 F.3d 1231, 1237–38 (Fed. Cir. 2010).

The reasonable-pertinence inquiry exists precisely because an ordinarily skilled artisan may seek solutions outside the invention’s immediate field. The relevant problem therefore cannot be defined so narrowly that only references from the same field of endeavor could qualify. Donner Technology, 979 F.3d at 1359–61. In In re ICON Health & Fitness, Inc., for example, the court held that a folding-bed mechanism was reasonably pertinent to problems involving the support and storage of a folding treadmill. 496 F.3d 1374, 1379–80 (Fed. Cir. 2007).

At the same time, the problem cannot be abstracted until it becomes universal. Describing the problem as “joining components,” “preventing movement,” or “retaining one object relative to another” risks making nearly every fastening mechanism pertinent to every mechanical invention. Reasonable pertinence must account for the purposes, structures, functions, operating conditions, and technical constraints that would have influenced where the skilled artisan realistically looked for a solution. See Clay, 966 F.2d at 659; Airbus S.A.S. v. Firepass Corp., 941 F.3d 1374, 1381–84 (Fed. Cir. 2019). Common sense cannot substitute for evidence that the artisan would have considered a materially remote field. Circuit Check Inc. v. QXQ Inc., 795 F.3d 1331, 1335–37 (Fed. Cir. 2015).

The field-of-endeavor test is similarly broad but tied to the invention as claimed. Relevant considerations include the invention’s embodiments, functions, structures, purposes, and disclosures concerning its use. Bigio, 381 F.3d at 1325–27. The field is not limited to the claim’s point of novelty, the narrowest possible characterization of the invention, or the patent’s particular emphasis within a broader field. Unwired Planet, LLC v. Google Inc., 841 F.3d 995, 1000–02 (Fed. Cir. 2016).

Claim scope also constrains how the relevant art and motivation to combine may be framed. In Axonics, Inc. v. Medtronic, Inc., the Board restricted its analysis to medical leads used for sacral neuromodulation even though the claims covered medical leads more broadly. The Federal Circuit held that the relevant art and combination analysis could not be limited to an unclaimed anatomical context or preferred embodiment. 73 F.4th 950, 958–59 (Fed. Cir. 2023). The question is whether an artisan would have combined the cited teachings to arrive at the claimed invention—not whether the combination would have satisfied requirements peculiar to the principal reference but absent from the claims.

The post-Sanofi cases also clarify how analogous-art issues must be presented in an IPR. In Netflix, Inc. v. DivX, LLC, the Federal Circuit rejected an unduly rigid requirement that the petitioner expressly announce, in formulaic terms, “the field of endeavor is X.” The petition and reply, read as a whole and in context, sufficiently presented the petitioner’s position. Evidence relevant to field of endeavor and reasonable pertinence may overlap, even though the two remain distinct tests. 80 F.4th 1352, 1358–64 (Fed. Cir. 2023).

Corephotonics, Ltd. v. Apple Inc. should be read with particular care. The Board concluded that Apple’s petitions inadequately addressed analogous art because they appeared to compare the asserted references with one another rather than with the challenged patents. On appeal, however, the Federal Circuit identified no procedural error in the Board’s treatment of the issue. Apple’s reply responded to the patent owner’s analogous-art challenge while remaining tied to the same references, disclosures, and underlying obviousness theories identified in the petitions. The court also concluded that Apple could elaborate under the reasonable-pertinence test after initially emphasizing field of endeavor because the two tests were related bases for the same obviousness ground, not wholly distinct grounds of unpatentability. Corephotonics, Ltd. v. Apple Inc., 84 F.4th 990, 1003–10 (Fed. Cir. 2023).

Corephotonics therefore permits responsive elaboration where the reply remains tied to the same references, disclosures, and obviousness theory, but it does not license a petitioner to introduce a materially new obviousness ground. Its result was record-dependent. The court emphasized both the responsive nature of Apple’s submission and the patent owner’s notice and opportunity to address it. A reply may develop an issue placed in dispute; it may not markedly depart from the petition’s theory or replace a missing ground with a new one. See id. at 1003–10; Henny Penny Corp. v. Frymaster LLC, 938 F.3d 1324, 1330–31 (Fed. Cir. 2019).

The nonprecedential decision in Apple Inc. v. Gesture Technology Partners, LLC illustrates the practical consequences of imprecise advocacy. The majority affirmed a Board finding that Apple had not established that a reference was analogous art where Apple used inconsistent formulations of the patent’s field of endeavor and failed to preserve reasonable pertinence. The majority stated that a patent’s field of endeavor generally should not vary depending on the reference under consideration, while recognizing that a patent might have more than one field in some circumstances—for example, where the specification discloses distinct embodiments and the claims at issue are directed to one of them. No. 2023-1494, slip op. at 14–15 (Fed. Cir. June 5, 2025) (nonprecedential).

Judge Prost dissented. In her view, the Board had defined the field too narrowly by focusing on the claimed point of novelty, contrary to Unwired Planet and the rule that analogous art should be construed broadly. She also questioned whether different formulations necessarily were contradictory, reasoning that they might identify complementary points of commonality between the patent and different references. Id. at 1–3 (Prost, J., dissenting). The divided disposition counsels against treating the majority’s field-stability discussion as a categorical rule, particularly in light of the opinion’s nonprecedential status.

Two additional distinctions complete the U.S. framework.

First, establishing that a reference is analogous does not establish a motivation to combine it with another reference. Analogous-art doctrine determines whether the reference may be considered in the obviousness analysis. The challenger must still identify the proposed modification, explain what the secondary reference contributes, and articulate why an ordinarily skilled artisan would have made that modification. Assertions that two references are analogous, without more, are insufficient. Sisvel International S.A. v. Sierra Wireless, Inc., 82 F.4th 1355, 1364–65 (Fed. Cir. 2023).

Second, U.S. law ordinarily assigns no independent legal significance to characterizing one document as the “primary” reference and another as the “secondary” reference. Those labels generally describe the presentation of the asserted combination: one reference supplies the structure being modified, and another supplies the proposed modification. Where the underlying factual inquiries are clear, the challenger need not separately prove that it selected the uniquely correct primary reference. Schwendimann v. Neenah, Inc., 82 F.4th 1371, 1382–84 (Fed. Cir. 2023); In re Mouttet, 686 F.3d 1322, 1333 (Fed. Cir. 2012).

The resulting U.S. rule is flexible but demanding. No particular incantation is required, and cross-field references are not disfavored merely because they originate in another industry. But the challenger must fairly present why each reference belongs within the art relevant to the claimed invention, separately identify the proposed combination, and explain why the artisan would have made it.

III.      The EPO’s Problem-Solution Framework for Cross-Field Teachings

Article 56 of the European Patent Convention provides that an invention involves an inventive step if, having regard to the state of the art, it is not obvious to the person skilled in the art. European Patent Convention art. 56. The EPO ordinarily applies that standard through the problem-solution approach:

1. identifying the closest prior art or another suitable starting point;

2. determining the objective technical problem; and

3. considering whether the claimed solution would have been obvious to the skilled person.

European Patent Office, Guidelines for Examination in the European Patent Office pt. G-VII, §§ 5–5.3 (2026) [hereinafter EPO Guidelines].

The EPO does not employ a freestanding two-part analogous-art test corresponding to Federal Circuit doctrine. Questions resembling analogous-art concerns instead arise throughout the structured analysis. Is the asserted starting point a realistic technical springboard? What objective technical problem arises from the claim’s differences over that starting point? What knowledge and technical fields are attributable to the skilled person confronting that problem? And would the prior art as a whole have prompted the skilled person to make the claimed modification?

IV.          A Suitable Starting Point

The EPO Guidelines ordinarily favor a starting point having a similar purpose or effect, or belonging to the same or a closely related technical field. EPO Guidelines pt. G-VII, § 5.1. That does not mean there must always be one uniquely correct “closest” reference. Several valid starting points may exist, and an inventive-step objection may succeed from one suitable route even though another document might also qualify as close prior art. Case T 1742/12, On-Demand Instantiation/Raytheon, ¶¶ 6.5–6.6, 10.3 (Tech. Bd. App. June 22, 2016).

This differs from U.S. practice under Schwendimann. U.S. law generally does not impose a separate requirement that a challenger justify its selection of a uniquely correct primary reference. In EPO practice, the selected starting point has greater analytical significance because it defines the differences to be evaluated and materially influences the objective technical problem.

A remote starting point is not automatically impermissible, but it may expose the analysis to hindsight. The recent decision in T 610/24 provides a contemporary illustration. The Board observed that although an inventive-step attack need not begin with the strictly closest prior art, starting from a more remote document often leads to an ex post facto analysis and the formulation of a “hindsight problem”—one the skilled person would not realistically have conceived from that reference. Case T 610/24, Checking the Chemical Compatibility of Pumps and Chemicals/Ecolab, point 4.4 (Tech. Bd. App. Dec. 16, 2025).

T 610/24 is recent and was not published in the EPO Official Journal. It is best used as an illustration of the established anti-hindsight concern, not as the foundation of the analysis. The more stable foundation remains the Guidelines and the Boards’ established decisions concerning suitable starting points, objective technical problems, and cross-field knowledge.

V.             The Objective Technical Problem

The objective technical problem is derived by identifying the differences between the claim and the selected prior art, determining the technical effects attributable to those differences, and formulating the problem in light of those effects. EPO Guidelines pt. G-VII, § 5.2. The objective problem may differ from the problem stated in the application and may require reformulation in light of the prior art developed during examination or opposition.

The problem must not contain elements of or pointers toward the claimed solution. A problem framed as “providing snap-fit engagement grips that axially and rotationally fix the washer” would effectively disclose the path to the answer. A more neutral formulation might concern secure retention, compact assembly, dimensional accommodation, or improved cartridge positioning, depending on the technical effects actually established by the claim’s distinguishing features.

Case T 422/93 illustrates how problem formulation also affects the relevant skilled person and the knowledge attributed to that person. The Board held that the skilled person should be identified from the objective technical problem arising from the closest prior art. The analysis may not define the skilled person as a specialist in the field of the claimed solution merely because the solution ultimately came from that field. Nor does the skilled person’s basic knowledge automatically include the knowledge of a specialist in another field when the closest prior art provides no indication that the solution should be sought there. Case T 422/93, Luminescent Security Fibres, 1997 O.J. E.P.O. 25, headnotes 1–3 (Tech. Bd. App. Sept. 21, 1995).

That is not the same as Sanofi’s rule that each reference must be analogous to the claimed invention. But the two doctrines respond to a similar hindsight danger: neither permits the known solution to determine retrospectively what problem the skilled person supposedly was addressing or what expertise that person supposedly possessed.

VI.          When Cross-Field Knowledge Becomes Realistically Available

Established EPO authority recognizes that the skilled person may draw from neighboring, broader, general, and, in appropriate circumstances, remote technical fields. The inquiry is integrated into the problem-solution approach rather than treated as an independent fourth step.

In T 176/84, the Board held that the relevant state of the art includes not only the invention’s specific field, but also neighboring or broader fields in which the same or a similar problem arises and of which the skilled person would be expected to be aware. Case T 176/84, Pencil Sharpener, 1986 O.J. E.P.O. 50, headnote (Tech. Bd. App. Nov. 22, 1985).

T 195/84 extended that principle to general technical problems. Solutions developed in a non-specific, general field may form part of the general technical knowledge attributed to skilled persons in particular fields when the application seeks to solve that same general problem in a specific setting. Case T 195/84, General Technical Knowledge, 1986 O.J. E.P.O. 121, headnote (Tech. Bd. App. Oct. 10, 1985).

Even a genuinely remote field is not categorically excluded. T 560/89 held that an artisan confronting a problem in one field might consider another, non-neighboring field where the same problem was widely known and a meaningful relationship existed between the materials used in the two fields. Case T 560/89, Filler Mass, 1992 O.J. E.P.O. 725, headnote (Tech. Bd. App. Apr. 24, 1991).

These authorities should not be reduced to a categorical taxonomy. Labeling two fields “neighboring” does not itself decide inventive step, just as labeling them “remote” does not end the inquiry. The important question is whether the objective technical problem, viewed without knowledge of the claimed solution, would have made the other field, its teaching, or the relevant common general knowledge realistically available to the skilled person.

Thus, the fact that an automotive clutch bearing and a drug-delivery device both use components that resist axial movement does not by itself establish that the respective fields are technically related. The analysis would need to consider the similarity of the actual retention problem, the applicable loads and tolerances, manufacturing constraints, materials, safety requirements, operating environments, and functional objectives.

VII.       The “Could–Would” Inquiry

Finding the secondary teaching does not complete the EPO analysis. The third stage asks whether the prior art as a whole would—not merely could—have prompted the skilled person, faced with the objective technical problem, to modify or adapt the starting prior art so as to arrive at the claimed subject matter. EPO Guidelines pt. G-VII, § 5.3.

The classic authority is T 2/83, which distinguishes technical possibility from an actual reason to proceed. The question is whether the skilled person would have made the modification in expectation of some improvement or advantage, not merely whether the person possessed the technical ability to do so. Case T 2/83, Simethicone Tablet, 1984 O.J. E.P.O. 265, headnote II (Tech. Bd. App. Mar. 15, 1984).

That inquiry resembles the U.S. requirement for a reason to combine, although the doctrines are not identical. Under either system, identifying all claim elements somewhere in the prior art is insufficient. In the EPO analysis, the challenger must show why the skilled person confronting the objective technical problem would have used the secondary teaching to modify the starting reference in the claimed manner.

VIII.                  Applying the Two Frameworks to Sanofi

The U.S. Analysis

Under U.S. law, the threshold issue was whether de Gennes belonged within the prior art relevant to the claimed drug-delivery device. Because de Gennes was not in the same field of endeavor, Mylan needed to establish that it was reasonably pertinent to a problem addressed by the ’614 patent.

A properly developed theory would have begun by identifying the relevant problem or problems from the claims, specification, and perspective of the ordinarily skilled artisan. The formulation could not be confined artificially to every detail of the preferred drug-delivery embodiment, because reasonable pertinence must leave room for useful teachings from other fields. But neither could it be reduced to the universal problem of fixing two components relative to one another.

The analysis then needed to compare de Gennes directly with the claimed invention’s problem. That comparison might have addressed whether both technologies involved compact retention structures, similar axial and rotational forces, comparable assembly constraints, or similar requirements for secure engagement within limited space. Expert testimony might have established that engineers designing drug-delivery devices routinely consulted general mechanical-fastening literature or that snap-fit connections were common general knowledge for the relevant retention problem.

If that evidence established reasonable pertinence, de Gennes could enter the obviousness analysis. The separate motivation-to-combine inquiry would then ask why an ordinarily skilled artisan would have selected de Gennes’s particular snap-fit geometry and incorporated it into the Burren–Venezia device, with a reasonable expectation that the modification would perform the claimed function.

Mylan instead argued principally that de Gennes addressed a problem analogous to Burren’s. That reference-to-reference relationship might have been relevant to explaining the proposed combination, but it did not establish that de Gennes was analogous art to the claimed invention. The omission was dispositive because Burren and Venezia did not supply the missing fixing elements without de Gennes. Sanofi, 66 F.4th at 1380–82.

The EPO Analysis

The EPO analysis would be organized differently. Assuming Burren were selected as a suitable starting point, the first substantive task would be to identify the claim features absent from Burren and determine the technical effects attributable to those differences.

Those effects might include secure axial and rotational retention of the spring washer, compactness, ease of assembly, reliable cartridge positioning, or accommodation of dimensional variation. The evidence would determine which effects were actually achieved across the claim’s scope and therefore could support the objective technical problem.

The problem would then need to be formulated without embedding the snap-fit solution. A drug-delivery-specific problem might make automotive clutch-bearing technology appear remote. A more general problem concerning compact mechanical retention could make general snap-fit teachings or common mechanical-engineering knowledge realistically available. But the level of generality could not be selected merely because it produced a convenient route to de Gennes.

The next issue would not be whether de Gennes passed a formal European “analogous art” test. It would be whether the objective technical problem would have made the relevant field, teaching, or common general knowledge realistically available to the skilled person. If compact snap-fit retention was a routine response to a general mechanical problem, T 195/84 might support attributing that knowledge to the skilled person. If the drug-delivery problem shared meaningful technical conditions with neighboring mechanical assemblies, T 176/84 might support consideration of those fields. If de Gennes remained genuinely remote, T 560/89 and T 422/93 would require a more concrete explanation of why the problem would have brought its teaching within the skilled person’s realistic consideration.

Finally, the could–would inquiry would ask whether the prior art as a whole would have prompted the skilled person to use de Gennes’s particular snap-fit arrangement in the Burren-based device. It would not be enough that the engagement grips physically could be transplanted. The record would need to explain why the skilled person would expect that structure to solve the objective problem while remaining compatible with the drug-delivery device’s dimensional, manufacturing, operational, and safety constraints.

Mylan’s comparison of de Gennes with Burren therefore would not necessarily have been irrelevant under EPO practice. Once Burren was selected as the starting point and the objective technical problem properly formulated, the relationship between Burren’s technical needs and de Gennes’s teaching could form part of the analysis. But the comparison would not be sufficient by itself. It would need to fit within the complete chain from the claim’s distinguishing features, to their technical effects, to the objective problem, to the knowledge realistically available to the skilled person, and ultimately to a reason the skilled person would have made the claimed modification.

The careful comparative conclusion is therefore not that the EPO necessarily would have reached the same result as the Federal Circuit. The EPO might have found the claim obvious if snap-fit retention was established as common general knowledge for the objectively identified problem and the prior art would have prompted use of de Gennes’s configuration. It might have found an inventive step if the relevant constraints were specific to drug-delivery devices and the starting art supplied no realistic reason for the skilled person to consider clutch-bearing technology.

The central point is narrower: a bare reference-to-reference comparison would not complete either analysis.

IX.       Practical Implications

For U.S. petitioners and invalidity challengers, Sanofi calls for a direct and disciplined presentation. Each relied-upon reference should be connected to the challenged patent through field of endeavor or reasonable pertinence. The field and problem should be grounded in the claims, specification, references, expert evidence, and perspective of the ordinarily skilled artisan. When both analogous-art routes are supportable, preserving both may reduce the risk presented by an uncertain field definition. The submission should then separately explain the proposed modification and why the artisan would have made it.

Netflix means that the argument need not be expressed through ritualistic language. Corephotonics means that a responsive reply may elaborate on a fairly presented theory under appropriate record-specific circumstances. Neither case makes the initial petition irrelevant. A petitioner that leaves the challenged patent out of its analogous-art comparison and later proposes a materially different combination, rationale, or ground remains exposed to waiver and new-theory objections.

For patent owners, the first diagnostic question should be whether the challenger compared the remote reference with the patent or merely with another prior-art document. The second should be whether the asserted problem has been framed at an appropriate level of abstraction. A formulation tied to every detail of the preferred embodiment may be too narrow; a formulation such as “retaining components” may be too broad to say anything meaningful about where the skilled artisan would have looked. Patent owners should also keep analogous art analytically distinct from motivation to combine. A challenger can fail either because the reference was outside the relevant art or because the proposed use of an otherwise analogous reference was unsupported.

Before the EPO, the corresponding advocacy should be expressed within the problem-solution framework. The party challenging inventive step should explain why the starting point is a realistic technical springboard, derive the objective technical problem from established differences and effects, and avoid incorporating the solution into the problem. Where the secondary teaching originates elsewhere, the analysis should show why the objective problem made that field, teaching, or common general knowledge realistically available. It should then establish a “would,” not merely a “could.”

Transatlantic practitioners should resist equating the systems’ terminology. A U.S. primary reference is not necessarily the EPO’s closest or suitable starting prior art. U.S. reasonable pertinence is not simply the EPO neighboring-field inquiry under another name. And the EPO problem-solution approach does not eliminate the need to justify why a skilled person would have used a teaching from another field.

X.          Conclusion

Sanofi does not establish that engineering knowledge must remain confined within industry boundaries. Cross-field references may be highly relevant, and both U.S. and EPO law recognize that skilled persons may draw upon knowledge outside their immediate specialties when the technical problem provides a legitimate path to it.

What Sanofi establishes is that the necessary connection cannot be supplied through a comparison that omits the claimed invention. Under U.S. law, each relied-upon reference must be analogous to the claimed invention, not merely analogous to another reference. Under EPO practice, a cross-field teaching must emerge from a suitable starting point, an objectively formulated technical problem, knowledge realistically available to the skilled person, and a showing that the prior art would have prompted the claimed modification.

The frameworks are different, but their central caution is much the same: prior art may travel across technical fields; the reasoning that brings it there cannot be reconstructed only after the invention is known.