Typos Matter in Patent Applications and Organic Chemistry

One of the reasons that I enjoy working on patents is that a patent application requires highly precise language. Despite the typos that still slip past me and onto my blog, I really enjoy writing precise technical definitions, searching for the right word, and hunting down those pesky typos.

This recently was an issue for Enanta Pharmaceuticals, which was trying to get a patent for compounds and methods for treating coronaviruses. In their provisional application from July 20th, 2020, they claimed chemicals including "NHC(O)—C2-C12-alkyl", meaning they were trying to claim those compounds with any formula where there is NHC(O) on one end, an alkyl group on the other end, and somewhere from two to twelve carbons in the middle.

On April 6th, 2021, Pfizer publicly disclosed a protease inhibitor with the same basic formula, but only ONE carbon in the middle. This chemical eventually became a part of Paxlovid, which Pfizer is using to treat COVID-19.

Enanta says that they realized on July 9th, 2021, that, actually, their provisional application contained a typo, and what they meant to write was "NHC(O)—C1-C12-alkyl". They filed a non-provisional application on July 19th, 2021, some of which, after several continuations, issued as patent 11,358,953.

Just last week, on June 23rd, the Court of Appeals for the Federal Circuit affirmed a District Court decision invalidating all the claims from the '953 patent. The decision held that Pfizer was the first to disclose the idea of the compound with only one carbon in the middle. While it is possible to use a provisional application to lock in an early filing date for what might eventually become a patent, in this case, the idea of the single-middle-carbon compound wasn't in Enanta's provisional application. The earliest public disclosure of that compound was Pfizer's disclosure in April of 2021. Because of this, Enanta's non-provisional application doesn't have the earlier priority date, and they couldn't get a patent for it.

The court held that it is not acceptable for someone to argue that they just made a typo, and they should therefore be understood to have come up with a broader idea than they initially wrote down and submitted to the patent office. I love the court's example here: a two-carbon alcohol, which is ethanol, is quite different from a one-carbon alcohol. Ethanol can be pleasantly intoxicating, which is why I homebrew beer and why I used to make and sell beer for a living. Methanol will kill you. What Enanta has done here is similar to disclosing ethanol, and then later claiming that it was a typo and you meant to also claim methanol. Those compounds are not the same. Experts wouldn't assume that it's just a typo distinguishing the two. And the two compounds have wildly different effects.

Part of the issue for Enanta was that they testified that an expert reading their provisional application would "understand that other references to ‘C2-C12 alkyl’ . . . are likely to be erroneous as well.” The court was not convinced that an expert thinking the formula was probably a mistake was sufficient for them to hold that Enanta definitely meant to include the 1-carbon compound in their provisional application.

So, what's the lesson here? Inventors and patent practitioners - make sure your have the science completely nailed down, and then take the time to make your language as perfect as is humanly possible. Brewers - you probably don't need a reminder, but stick with ethanol.

© 2026. All rights reserved.