In a long-awaited decision, Eli Lilly's patent (US 5,229,382) covering the schizophrenia drug olanzapine (zyprexa) is valid, enforceable and infringed, according to Judge Richard Young's ruling of the Southern District of Indiana.

Three generics manufacturers challenged the validity of the 1993 patent protecting the basic molecule until 2011. Miami-based Zenith Goldline Pharmaceuticals, Indiana-based Dr. Reddy's Laboratories and Israeli-based Teva Pharmaceutical Industries argued that the patent was not valid because discovery of the drug's molecular structure was obvious and that a previous Lilly patent that expired in 1995 had covered it. The companies also claimed that Lilly misled the patent office and omitted certain test results when applying for the patent.

In the 1970s and 1980s, Lilly produced a novel class of compounds called thienobenzodiazepines — tricyclic compounds with thieno, benzo and diazepine rings fused together. Two patents were issued for this compound class (US 4,115,568 and US 4,115,574,), which have identical technical disclosures, but different claims. The compounds contain halogen atoms; flumezapine was noted as a particularly active compound, but after further preclinical testing this compound was dropped due to toxicology concerns. Olanzepine differs from flumezapine by removal of the halogen atom.

The case hinged on a comparative dog toxicology study that Lilly conducted in 1990, which compared olanzapine to a related and previously patented compound, prior to filing a patent application with the US Patent and Trademark Office (PTO). According to Lilly, the outcome of the study showed that olanzapine had an expectedly superior therapeutic profile, and did not raise cholesterol levels or increase the production of the milk-producing hormone prolactin in the same way as other compounds of the class. The generics companies tried to prove that the dog study was flawed and that olanzepine was not novel and therefore not deserving of unique patent protection.

During the prosecution of the '382 patent, Lilly disclosed that the most closely related prior art was the matter claimed in the '548 patent, but the company did not disclose the '574 patent. However, the patent examiner found this prior art in a search and it was taken into account in the examiner's final decision, and so Lilly cannot be penalized for their omission. Judge Young ruled that the generics companies have not proven by clear and convincing evidence that Lilly concealed prior art nor data from the dog study with the intent to deceive the PTO.