Non-obviousness: when failure equals successOctober 23, 2019

A recent Federal Circuit Court decision demonstrates that evidence of prior failures to achieve desired results can be potent evidence of lack of predictability in a field, resulting in the non-obviousness of using a compound to treat a specific cancer, even though the drug target was identified as effective in a test tube.

The decision, OSI Pharmaceuticals v. Apotex, reversed a decision by the US Board of Patent Appeals and found patentable claims to use of the compound erlotinib to treat non-small cell lung cancer (NSLC), previously identified as a leading cause of cancer deaths.  OSI Pharmaceuticals, LLC v. Apotex Inc. et al, 2018-1925 (Fed. Cir. October 4, 2019).  The OSI patent, US Patent No. 6,900,221, was challenged in an Inter Parties Review proceeding by Apotex and found invalid as obvious by the Board.  The Court found the Board had misread a cited reference as teaching more than substantial evidence supports.  The Board had agreed with Apotex that a review article by Dr. Jackson Gibbs showed that erlotinib has anti-cancer activity against non-small cell lung cancer.  However, only one of the thirty research studies reviewed referenced NSCLC antitumor activity with a different compound, and another reference mentioned erlotinib but not as useful with NSCLC treatment.  Dr. Gibbs confirmed this in a declaration.  This reference was combined with another, Schnur, showing erlotinib as one of 105 compounds that are inhibitors of a family of proteins including epidermal growth factor receptor, which erlotinib targets.  It refers to use in a variety of cancers, including lung cancer.  Further, OSI’s own 10-K SEC filing was cited discussing erlotinib as targeting NSCLC.

After discussing the misreading of the Gibbs reference, the Court continued that there could be no reasonable expectation of success in a highly unpredictable field having high rates of failure.  Noting an over 99.5% failure rate of NSCLC treatment drugs entering Phase II clinical trials, such failures demonstrate the lack of reasonable expectation of success.  These numbers, the Court added, do not take into account the number of drugs that do not make it to Phase II trials.  What is more, the evidence demonstrated that even if a compound was useful for certain tumor types and certain cancers, it is unpredictable that it will be useful in treating other types of cancer.  This is true, the Court held, even though compounds identified as epidermal growth factor receptor inhibitors are targets for cancer treatment.  This alone is not indicative of treatment success, the Court found, and in vitro – test tube – effectiveness is a poor proxy for effectiveness in treating cancer in the body.  Further, the 10-K filings did not show any efficacy data specific to non-small cell lung cancer.

The decision is a precedential decision which provides a guide for subsequent cases and can be reviewed in full at

← Return to Filewrapper

Stay in Touch

Receive the latest news and updates from us and our attorneys.

Sign Up