Genentech v. Iancu – another reminder that narrower, but valid, claims are preferable to broader, but invalid ones

Genentech v. Iancu, Appeal Nos. 2019-1263, 2019-1267, Fed. Cir., March 26, 2020.

Genentech owns US Patent Nos 7,846,441 and 7,892,549, directed to the treatment of cancer with a combination of an antibody and a taxoid, without anthracycline.  The patents share a specification, and the claims of each require that the combination be administered in an amount effective to extend the time to disease progression in the patient being treated.  During prosecution, the Examiner rejected the claims as indefinite, stating that it was unclear what the extension of time to disease progression is relative to.  Genentech responded:

[T]he expressions “extend the time to disease progression” and “response rate” are clear from the specification (see, in particular, page 15, line 15-17; and pages 42-43 and would be readily understood by a skilled oncologist.  Clearly, the combination of an anti-ErbB2 antibody and a taxoid is administered in an amount effective to extend the time to disease progression relative to an untreated patient.

Slip opinion, page 6. Petitioners filed six petitions collectively requesting inter partes review of all claims of both patents.  The Board instituted all six, and construed the term “effective amount” in accordance with the definition argued by Genentech during prosecution, and held the so-construed claims obvious.  Genentech challenged the Board’s constructions on appeal, arguing that the Board “erred in its reliance on this exchange by using it ‘to override the meaning evident from the specification.'”  Slip opinion, page 7.  The Federal Circuit sided with the Board, stating:

…the specifications discuss several drugs and drug combinations that may be viable comparators, including a taxoid, anthracycline/cyclophosphamide treatment, and chemotherapy.  The specifications do not select any one of these as the comparator, nor do they preclude comparison to an untreated patient…

Further, the Court stated:

Genentech did not view this choice as inconsistent with the specification during prosecution.  Instead, Genentech interpreted its own claim language, based on its own specification’s disclosure, as referring to a comparison to untreated patients.

And provided this twist of the knife:

The Examiner even provided Genentech with the very alternative option (“taxoid alone”) for which Genentech now advocates.

Taking the opportunity to provide a reminder about the impact of prosecution history estoppel within a family of applications, the Court held:

Genentech’s comparator choice during prosecution of the ‘441 patent applies equally to the same claim term that appears in the ‘549 patent, which share a specification and is in the same patent family.  Given this relationship between the patents, the Board construed the relevant terms consistently in all six inter partes reviews.  We see no error in the Board’s construction.