In CareDx v Natera, Appeal No. 2022-1027, (Fed. Cir., July 18, 2022), a three judge panel of Judges Lourie, Bryson and Hughes, affirmed the district court’s finding that the claims of U. S. patent nos. 8703652, 9845497 and 10329607 are invalid for failing to survive the Alice/Mayo test for patent eligibility. I subtitled this post using Mathew 7:13-14: “Enter through the narrow gate. For wide is the gate and broad is the road, that leads to destruction.” The appeal to the Federal Circuit, which I wrote about on October 15, 2021, never got on the narrow road that leads to viable diagnostic claims. It may not have been possible to overcome the obstacles that blocked the road, but CareDx managed to hit them all, and ended up with three invalid patents on natural phenomena.

The claims were directed to a method for detecting transplant rejection or organ failure by isolating and genotyping a sample from the subject who received the donation, quantifying the cfDNA, and diagnosing the transplant status for an increase in donor cfDNA over time. An increase indicates possible transplant failure.

Judge Lourie summarized the claims, some of which are more than a page long, this way:

“Here, as in Ariosa, the claims boil down to collecting a bodily sample, analyzing the cfDNA  using conventional techniques, including PCR, identifying naturally occurring DNA from the donor organ, and then using the natural correlation between heightened cfDNA levels and transplant health, to identify a potential rejection, none of which was inventive. The claims here are equally as ineligible as those in Ariosa.”

Let’s take a quick look at how CareDx got onto the broad road. CareRx hoped to avoid Ariosa by arguing that it was doing more than just measuring a biomarker correlated to an existing phenomenon. Problem 1 is that CareDx did not discover the correlation; it just improved on it (or did it?). Louie writes:

“CareDx argues that the patents’ claims are directed not to natural phenomena, but to improved laboratory techniques. CareDx contends that the ‘claimed advance’ is an ‘improved, human-designed method for measuring increases in donor cfDNA in a recipient’s body to identify organ rejection.’ … In particular, CareDx identifies the use of digital PCR, NGS, and selective amplification to more accurately measure the donor SNPs of cfDNA transplant recipients. However, CareDx does not actually claim any improvements in laboratory techniques … Furthermore the specification admits that the laboratory techniques disclosed in the claims require only conventional techniques and off-the-shelf technology.”

In fact, CareDx had at least one claim in the ‘497 patent that recites that the assay detects the donor-specific circulating cfDNA from the organ transplant when the donor-specific circulating cfDNA [makes] up at least 0.3% of the total circulating cfDNA in the biological sample. I presume that this claim limitation was put into the claim so that “improvement”  could be argued, but the limitation is not mentioned in the opinion.

Let’s look at a few other things CareDx encountered on its broad road to legal destruction. The panel looked at every step of the method in isolation. In other words, once CareDx argued “improvement” it was forced to admit that the specification disclosed that all those analytical techniques, such as PCR, NGS and “selective amplification”, would be considered as conventional in the art. CareDx might have relied on some of the decisions finding patent eligibility where physical equipment was necessarily involved, such as XL LLC v. Trans Ova Genetics or Illumina v Ariosa.

The finding of conventionality of individual steps permitted the court and the panel to effectively rule that the method was directed to a natural product, since the devices used to carry it out were given no weight. Therefore, the patents failed to pass Step 1 of Mayo/Alice. Could it have been argued, if that was the case, that the equipment used to carry out the method was arranged in a novel sequence? (Also, is someone going to argue that PCR involves replicating small amounts of DNA to afford useful amounts? – This is accomplished by the hand of man.)

These are minor thoughts, CareDX should left the word “diagnostic” out of the claims and the specification. This is certainly no more of a diagnostic test than the Mayo range-finding step was. It is presently clear that in the life sciences, recognition of the utility of a naturally occurring correlation is not enough to avoid patent ineligibility. Of course, and this is cold comfort to CareDx, would it have helped to get this method into the safe harbor of methods of medical treatment? In other words, the first step could recite the actual transplantation step and/or the final step of the process could recite some sort of medical intervention. Narrower claims might have returned CareDx to the narrow path of patent life.