Vanda v. Roxane Labs. – Are Two Natural Laws Better Than One?
As you will recall, in Prometheus v. Mayo, the Supreme Court held that a claim reciting a natural law had to have other non-conventional steps to pass muster under s. 101. The natural law in Mayo was the correlation between the concentration of the metabolite of the immunosuppressive drug and the efficacy of the drug – unpleasant side effects at the high end of the recited range and lack of efficacy at the low end. The Mayo Court wrote: “We need not, and do not, decide whether were the steps at issue here less conventional, these features of the claims [the administering, determining and recognition steps] would prove sufficient to invalidate them.” It was unnerving that the invalidated claims were method-of-treatment claims.
In his opinion in a Hatch-Waxman litigation involving the anti-schizophrenia drug Iloperidone, Vanda Pharms., Inc. v. Roxane Labs., Inc., Civil Action No. 13-1973-GMS; 14-757-GMS (D.Del., 2016), Judge Sleet upheld the validity of the compound and the method claims in Reissue Patent No. 39,198 and US Patent No. 8,586,610.
The court began by finding the compound claims unobvious. The method claims had been attacked as failing s. 101. I assume that the original method of treatment claims are in the ‘198 patent, which is about to expire. Iloperidone had been the subject of clinical trials by Novartis, but development had been abandoned because the drug unpredictably caused a serious heart irregularity called QTc.
Vanda developed a companion diagnostic test and filed “regimen type claims” that were written as method of treatment claims. (I don’t know why, if a claim to a drug is unobvious, any method of using it is not unobvious as well. In re Pleuddemann, 910 F. 2d 823 (Fed. Cir. 1990). In fact, Judge Sleet found that the method claims were unobvious following a full-scale s. 103 analysis. At the worst, he effectively ruled that the claims were directed to a new use of an “existing drug”, a claim form endorsed as patent eligible in Mayo dictum.)
But I digress. The method claim in question (I will discuss claim 6) was drawn to a method for treating a patient with Ilopridone comprising (a) genotyping a biological sample from a schizophrenic patient to determine if the patient has a CYP2D6 “poor metabolizer” genotype. Such patients are administered doses of the drug that are less than 12 mg per day that was found to lower their risk of QTc. Patient found not to have the poor metabolizer genotype are administered 12-24 mg/day of the drug.
Roxane argued that the method patent “embodies two laws of nature: (1) that mutations in the CYP2D6 genes can alter its enzymatic activity (2) that a patient’s CYP2D6 enzymatic activity affects their metabolism of Iloperidone” and that Vanda just added a conventional dose adjustment to reduce the risk of a side effect. But, in this case, in addition to pointing out that claim 6 is not an attempt to patent “the relationship between Iloperidone, CYP2D6 metabolism and Qtc”, Vanda brought many of its successful non-obviousness arguments to bear on this question.
Vanda argued that the ‘610 method patent was not the result of routine and conventional testing, e.g., that clinical study design is not routine or conventional. Vanda argued that adjusting the doses of similar drugs that are poor metabolizers did not eliminate the QTc side effect and that the original examiner allowed the claims as not barred by the Mayo test after the dosing limitations were added to the claims.
Judge Sleet agreed with Roxane that the claims depend on the relationship between the drug, CYP2D6 metabolism and the QTc side effect, and then, in part 2 of the Mayo test, looked for an additional step that would “transform the claims, making them valid.” Here the Judge was convinced that there was a combination of elements “sufficient to ensure that the claims amount to significantly more than just a natural law.”
The Judge was swayed by the argument that the dosage step does not apply to all patients as in Mayo, but only the population that exhibited the specific mutation:
“The dosage step requires applying genetic tests in a high specified way. Moreover, the process of using the genetic test to inform the dosage adjustment recited in the claims was not routine or conventional and amounted to more than an mere instruction to apply a natural relationship.” Importantly, the Judge gave weight to the recognition of the meaning of the correlations in the claim as involving “the human ingenuity that comes from applying a natural discovery in a way that achieves a ‘new and useful end’”[citing Cellzdirect, Diehr and Alice].
Finally, he distinguished Mayo-type preemption from the present case that would not preempt biological sampling or genotyping.
I can only hope that this decision helps establish that the use of one test to divide a population of patients into two or more groups (of responders vs. non-responders), and then treating the groups differently will pass step 2 of the Mayo test. The ‘610 patent does not expire until 2027, so this decision may well be appealed to the Fed. Cir.