Finding that a Patent Trial and Appeal Board (PTAB) assessment of reasonable expectation of success was not supported by substantial evidence, the US Court of Appeals for the Federal Circuit reversed the PTAB’s obviousness determination. OSI Pharm., LLC v. Apotex Inc., Case No. 18-1925 (Fed. Cir. Oct. 4, 2019) (Stoll, J).

OSI’s patent claims priority to 2000, and the relevant claims recite methods of treating non-small cell lung cancer (NSCLC) with a prior art compound, erlotinib. The PTAB found the claims invalid in an IPR proceeding over the Schnur reference in combination with either the Gibbs reference or OSI’s 10-K filing. Schnur discloses 105 compounds, a group in which erlotinib is a preferred embodiment, as inhibitors of epidermal growth factor receptor that are useful for treatment of a variety of different cancers, including lung cancer. However, Schnur does not teach erlotinib for treatment of NSCLC in particular.

Gibbs generally states that erlotinib and another compound called ZD-1839 appear to have activity against NSCLC, and cites two conference abstracts as supporting that proposition. One abstract discusses ZD-1839 for use in NSCLC, and the other discusses erlotinib for other kinds of cancer besides NSCLC. The PTAB found a “clear inference” that erlotinib has activity against NSCLC, even though neither reference actually provides that teaching and despite a sworn declaration from Dr. Gibbs that he was not aware of any abstracts or articles published as of the time of his article that contained such a teaching. The PTAB concluded that the combination of Schnur and Gibbs taught treatment of NSCLC with erlotinib with a reasonable expectation of success.

The OSI 10-K discloses that erlotinib was entering Phase II clinical trials, and the PTAB concluded that a skilled person would therefore understand the 10-K to teach that Phase I clinical studies and preclinical animal studies had already been successfully performed with erlotinib. Given that the claims at issue are generic to “mammals” and do not specify treatment of humans, the PTAB concluded that a skilled person would understand the OSI 10-K as teaching erlotinib for treatment of NSCLC with a reasonable expectation of success. The PTAB therefore found OSI’s patent claims obvious. OSI appealed.

The Federal Circuit reversed, concluding that the PTAB misinterpreted the cited references as teaching more than was supported by substantial evidence. The Court found that the Gibbs reference and abstracts cited therein did not expressly teach erlotinib for treatment of NSCLC, particularly in view of Dr. Gibbs’s sworn declaration. The Court also found that the OSI 10-K did not actually disclose any information about erlotinib’s effect on NSCLC or specify that any of the underlying preclinical data pertained to NSCLC as opposed to other forms of cancer.

As to the PTAB’s reasonable expectation of success finding, the Federal Circuit pointed to evidence showing that developing therapies for NSCLC is extremely unpredictable, including a statistic that more than 99.5% of NSCLC drugs that entered Phase II between 1990 and 2005 never obtained US Food and Drug Administration approval. In view of this real-world evidence and the fact that none of the cited references disclosed even in vitro “test tube” data, let alone data from an animal, to the effect that erlotinib is useful for NSCLC treatment, the Court found that the references “provide no more than a hope” that erlotinib could be effective for treatment of NSCLC.

Practice Note: The Court was careful to state that its holding does not mean that efficacy data is always required for a reasonable expectation of success, and expressly stated that its decision is limited to the particular facts of this case. However, this decision could make it more challenging to invalidate method of treatment claims, particularly where the real-world evidence indicates that treatment for a particular condition is unpredictable.