News Centre - Article 6

Product-By-Process Claims: A Jurisdictional Comparison

- Tom Boyce, Bill Bennett & Nick Finnie


In some circumstances, particularly in the chemical and life sciences, it can be difficult or impossible to define a product solely by reference to its structural or chemical features. Indeed, in many cases the precise structure or chemical formula may not be completely known at the time a patent is drafted.

A solution in these circumstances is to use a form of product claim that characterises the product by reference to the process that has been used to make the product.  Examples of such claims include:  “Product X produced by Process Y”, “Product X obtained by Process Y”, or “Product X obtainable by Process Y”.  The use of product-by-process claims has been described as a “rule of necessity” when the inventive product cannot be described in any other way (1).  But a wide range of claim formats can be considered “product-by-process” claims, including claims partially limited by process steps or even product claims with functional terms (2). 

Typically, such “product-by-process” claims are thought to extend protection beyond that conferred by process claims alone.  Yet, the same claim may confer distinctly different protection depending on the jurisdiction in which it is prosecuted or enforced.  Crucially, the language chosen can greatly influence the examination and enforcement of these claims across several important jurisdictions.


The EPO considers the terms “obtained”, “obtainable”, and similar terms to be equivalent and directed broadly to the product per se.  A product obtainable by a process is a claim format that, when read literally, should encompass all products with the resulting characteristics imparted by the process.  A claim to a product obtained by that same process simply includes the end point of the process. 

In line with this construction, the European Patent Office examines product-by-process claims as directed to the product per se without limitation as to how the products are actually produced.  The EPO also restricts the use of product-by-process claims to when there is no other information available that could enable an applicant to define the product satisfactorily by reference to its composition, structure or some other testable parameter (3).

For infringement proceedings, the authors understand that all signatories to the EPC will construe the claims for infringement as the EPO does for examination – i.e. as directed to the product per se.


For examination, current USPTO practice construes these claims as directed to the product, irrespective of the process by which it is made.  Examination asks initially if the product itself is novel.  Process limitations play a role only if “the product can only be defined by the process steps by which the product is made, or where the manufacturing process steps would be expected to impart distinctive structural characteristics to the final product.” 

The USPTO focus on the patentability of the product often appears to result in claims reciting whatever structural or physical limitations might be available to at least partially describe the product in those terms.  These are necessarily less than comprehensive in limiting the claim (else there would be no process limitations) but more than what may be required by the EPO.  The net result in prosecution can be a marked distinction in claims that grant in the US from their counterparts before the EPO.  US claims may appear as “hybrids” - reciting structural features along with process limitations. 

Recently, the construction of these claims for assessing infringement was settled in the Abbott Labs v. Sandoz (4).  Sitting en banc on only this issue, the CAFC held that product-by-process claims in the US include the process steps as express limitations of the claim.  Contrary precedent, including Scripps Clinic, (5) has been expressly overruled.  While this decision appears to have settled divided precedent on the construction of these claims in infringement proceedings, it perhaps raises further questions about how such claims should be examined.  Indeed, the dissent in Abbott Labs v. Sandoz suggests that “[f]or the first time, claims are construed differently for validity and for infringement.”   Certainly, the holding of Abbott Labs now puts a premium on the ability to claim without reference to process limitations.

Interestingly, in light of the practice in other jurisdictions, the Court’s majority opinion in Abbott Labs dismissed attempts by counsel for Abbott to finely parse claims based upon a plain language interpretation of “obtainable” as opposed to “obtained”.  The Court found this argument unavailing, at least on the facts of the case.


In Japan, product-by-process claims are generally treated broadly as claims to the product, per se, without limitation as to how the products are actually produced (6).  Thus, the approach in Japan is similar to the approach adopted in the US and in Europe during examination.

It is the authors’ understanding that the Japanese Patent Office does not allow claims written in the “obtainable” format, and construes the “obtained” format broadly as a claim to the product, per se


In contrast to the approaches in the EU, US, and Japan, but in line with Australia’s emphasis on plain language interpretation, Australian patent law makes a clear distinction between the “obtained” and “obtainable” formats of these claims.  A product “obtained” by the process is presumptively limited by the process features.  A product “obtainable” by a process is presumptively to the product itself, unlimited by the process, though characterised by it.

The Australian Examiner’s Manual suggests that the “obtainable” claim format will only be possible where the chemical structure or composition of the product is undetermined, echoing EP practice.  However, in the authors’ experience this restriction is not strictly observed, depending of the facts of the case.

Also in line with the Australian emphasis on plain language interpretation, these claims will likely be construed by a court using similar principles, looking to the detailed language of the claims in the context of the entire invention described by the specification to determine the scope for infringement determinations.


For most product-by-process claims, the lack of explicit structural or chemical features of the product-by-process can lead to difficulties in enforcement.  While a product-by-process claim may be presumed to cover the product, per se, the issue will often be whether the allegedly infringing product is structurally or chemically identical to the claimed product. 

Some jurisdictions create a statutory presumption that any equivalent product is presumed to have been produced by the claimed process in the absence of evidence to the contrary (7).  This reduces the evidentiary burden upon the patentee when asserting a process claim, particularly in jurisdictions where discovery is limited.  In practice, a plaintiff will likely be required to identify structural or chemical characteristics/properties of the product which result from one or more steps in the claimed process, and which are present in the allegedly infringing product, and which distinguish it from prior art products.   This is certainly now the case in the US after Abbott v. Sandoz and appears to present a substantially increased evidentiary burden on the plaintiff.

Equally, however, where product by process claims are construed as directed to the product per se for infringement (e.g. in the EU, Japan, and, depending on the language of the claim, Australia) it can be difficult for potential infringers to determine whether their product falls within the scope of a product-by-process claim.  The question becomes whether the characteristics imparted by the process claimed can be imparted by a different process.  Of course, the uncertainty surrounding such claims may have a deterrent effect to potential infringers, even if ultimate infringement is not initially clear.

In most jurisdictions, process claims are deemed to also cover the product that has been made by the claimed process (8).  This provides an important means of protection where there are no claims to the product, per se, where the claimed process is performed outside of the patent jurisdiction, and the resulting product is imported into the patent jurisdiction.

Of critical importance, of course, is whether a claim will be construed as a product-by-process claim at all.  A large diversity of claims can be drafted to fit within the class and  “...there may be differing results depending upon the exact wording of a claim at issue” (9). This issue is likely of much greater importance to modern chemical or biological products that to mechanical products.  Careful attention must be paid to how each jurisdiction may construe these claims in examination and infringement.


1. Consider carefully the scope of protection provided by product-by-process claims in each jurisdiction and shape prosecution strategy appropriately.  The same claim will likely be construed differently across jurisdictions.   Of course, where a process results in a product which is both novel and inventive, then product claims should always be included in the application. 

2. Wherever such a product is difficult to define in structural or chemical terms, then product-by-process claims should also be included.  Even where the product can be readily defined in structural or chemical terms, there may be advantages to the patentee to also include product-by-process claims, depending upon jurisdiction.

3. If product-by-process claims are needed, the specification’s description should include claims in at least the “obtained” and “obtainable” formats, with additional options for defining the product included.  The “obtainable” format should be used wherever possible (e.g. Europe and Australia), with the “obtained” format being used in countries where claims in the “obtainable” format are not allowable (Japan), but which nevertheless construe claims in the “obtained” format broadly as being directed to the product, per se

© 2009 - Pizzeys Patent & Trademark Attorneys – AUSTRALIA
Authors – Tom Boyce, Bill Bennett & Nick Finnie

1. Abbott Labs. v. Sandoz, Inc., 2007-1400, -1445 (Fed. Cir., 18 May 2009) (Newman J., Mayer, J., and Lourie, J., dissenting, at *3, *8).
2. Mirabel cited by dissent in Abbott
Case Law of the Boards of Appeal of the European Patent Office, 5th Ed., 2006: pp211-213, paragraphs 6.1 3. and 6.3
4. Abbott Labs. v. Sandoz, Inc., 2007-1400, -1445 (Fed. Cir., 18 May 2009).
5. Scripps Clinic & Research Foundation v. Genentech, Inc., 927 F.2d 1565, 1583 (Fed. Cir. 1991).
6. F. Hoffman La Roche v Otsuka Pharmaceutical Co., Ltd et. al. (Heisei 06 (ne) No. 2857) Tokyo High Court, 17 July 1997.
7. Japanese Patent Law Section 104, Canadian Patent Act Section 55.1
8. See, for example, EPC Article 64(2); Japanese Patent Law Section 2(3)(iii); 35 USC Section 271(g).
9. Abbott Labs. v. Sandoz, Inc., Lourie dissent at *3.