In what appears to be a response to Kahn’s article by the USPTO, the July 31 Washington Post article reports that USPTO officials state that “they do not require applicants to delineate how well any genetic tests—including those looking for an increased risk for diseases—work by race. A search of more than 3 million patents over the past 20 years found only 37 that included any reference to race.”
One USPTO spokesperson further stated that “[w]e recognize that race is a fuzzy concept,” also noting that “[i]t’s the genetic marker that’s important, not the racial group.”
It is noteworthy that the USPTO, at least as a matter of policy, has this viewpoint. As I see it, an important consideration in personalized medicine is an individual’s specific genetic profile, but not necessarily one’s race or ethnicity. In fact, personalized medicine runs the risk of missing its mark altogether if it relies only on race, a category that is dubious when it comes to generalizing as to specific biomarkers and actual genetic profiles.
In this sense, I share Kahn’s concern if it is indeed accurate that USPTO examiners (rightly or wrongly in light of USPTO policy) are “requiring applicants to include racial categories.” Especially if examiners assert that personalized medicine claims must recite racial categories in order to be enabled, such assertions have the danger of being inaccurate scientifically and legally, among other things.
As noted by Kahn, a designation of race involves numerous social and cultural aspects, in addition to subjective considerations of how one sees themselves. Such aspects often have little bearing on whether a particular group has a specific genetic profile, and certainly cannot speak to the genetic profile of every individual in the group. To the extent that a claim reciting a specific genetic profile (e.g., X polymorphism in gene Y) lacks enablement, it does not help the situation to recite a race element.
Even if a rare examiner suggests (incorrectly, as concluded by the BPAI in at least two relevant cases) that a patent claim requires a race category element in order to be enabled, this does not mean “the USPTO is forcing race into biotech patents,” as asserted by Kahn. As the Washington Post article points out, the USPTO recognizes that “[i]t’s the genetic marker that’s important, not the racial group.” Absent evidence that the USPTO is mandating (and training examiners to implement) policies consistent with Kahn’s hypothesis, a random rogue examiner does not mean the sky is falling.
This all said, Kahn’s article prompted me to conduct my own review of relevant claims allowed by the USPTO. I wanted to see for myself whether, as Kahn asserts, “it is evident that the practice of requiring patent applications to introduce race into their biotech patents has become routinized at the USPTO.”
A USPTO official stated to the Washington Post that “[a] search of more than 3 million patents over the past 20 years found only 37 that included any reference to race.” Even assuming examiners have, in fact,“forced” race elements into claims in every one of the 37 cases (an unlikely scenario), 37 patents fails to indicate a “routine” situation, at least to me.
My own preliminary search on the USPTO website revealed more than 37 patents reciting a race element, but not many more. For example, I conducted a search on the USPTO website for issued patents having claims reciting “African American” or “Asian” or “Hispanic” or “Caucasian.” This search received 128 “hits.” A quick review of these hits indicated that many patents were related to plants, insects, or subject matter otherwise irrelevant to personalized medicine.
Looking at the first 50 hits of this search, I uncovered 19 relevant U.S. patents, which are outlined in the Table. A review of the other 78 (of 128) hits uncovered 16 additional relevant patents (including U.S. Pat. No. 6,716,518 discussed in Kahn’s article).
In other words, this single search uncovered 35 relevant patents. I also conducted a few other preliminary searches using terms such as (Jewish or Filipino), (descen$ or ancestry), (treat$ or diagnos$ or screen$) and/or (human or patient or “biological sample”). Relatively narrow searches using such terms uncovered five additional patents, bringing my total to 40.
Before relying on this number, however, I should note that I conducted additional broader searches, but hit numbers were large (>500), with many hits being irrelevant. For example, adding the term “Japanese” to my initial search (i.e., “African American” or “Asian” or “Hispanic” or “Caucasian” or “Japanese”) resulted in 1161 hits, a significant jump from 128. Substituting “European” for “Japanese” in the search resulted in 727 hits. Substituting terms such as “Black” or “White” resulted in hits greater than 20,000. I did not review these larger search results. Thus, my searches discussed herein are not conclusive of the total number of U.S. issued patents containing claims that reference race.
Looking at the Table, however, a few things pop out. First, the USPTO has issued, and continues to issue, personalized medicine patents with claims that recite race elements. Second, in most such cases, independent claims recite the race element in question, not merely a dependent claim. Third, the claims recite specific genetic parameters in addition to the race element, which may render the race element irrelevant to patentability (including enablement).