Combination therapies are often used to treat medical conditions that have proved difficult to cure with single-agent therapies, including, for example, tuberculosis, leprosy, cancer, malaria and AIDS. According to the 2013 annual meeting of the American Society of Clinical Oncology, cancer drug combinations represent the future, because single-drug therapy often fails as cancer cells develop resistance to the treatment.

Patent protection is vital for the development of combination therapies, because the process can require years of extremely costly research and clinical testing for drug developers. Combination use with other drugs is also a common strategy for pioneers who are seeking ways to protect new investments made in their previously marketed drugs. The Federal Circuit's decision in Novo Nordisk A/S v. Caraco Pharmaceutical Labs. Ltd., No. 11-1223 (Fed. Cir. June 18, 2013), however, indicates that patent protection for combination therapies likely will be significantly more difficult to obtain in the future.

The patent at issue in Novo Nordisk was directed at the treatment of noninsulin-dependent diabetes mellitus using a combination of repaglinide (marketed as Prandin in the United States) and metformin (sold under the trade name Glucophage). The Federal Circuit held that the claimed combination was obvious, despite evidence of synergistic effect. Repaglinide falls within the functional class called "insulin secretagogues," which work by stimulating insulin release from pancreatic beta cells. The Federal Circuit noted that combination therapy using metformin and another secretagogue, glyburide, was well known in the art to produce beneficial and even synergistic results. Therefore, "it is reasonable that an artisan seeking to combine a known insulin sensitizer (like metformin) with a new insulin secretagogue (like repaglinide) would base his expectations upon prior art sensitizer/secretagogue combinations."

Despite the Federal Circuit's reasoning, it is noteworthy that there are two subclasses of secretagogues: meglitinides (such as repaglinide) and sulfonylureas (such as glyburide, disclosed in prior art). Repaglinide differs from the sulfonylureas in its molecular structure, profile of action and excretion mechanism. In addition, even within the subclass of sulfonylureas, only some sulfonylureas formed synergistic combinations with metformin. Nonetheless, the Federal Circuit held that the claimed combination was obvious based on the general mechanism of action for all secretagogues, saying that "[r]epaglinide and sulfonylureas are both insulin secretagogues, and…both treat diabetes by stimulating the pancreas to release insulin." This common mechanism of action shared by glyburide (prior art) and repaglinide (claimed invention) outweighed Novo Nordisk's evidence of "unexpected" synergistic results. Synergistic results were in fact acknowledged by the court, but found to be predictable, based on prior art.


The U.S. Supreme Court in KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007), has opened the door for a finding of obviousness based on the "obvious to try" rationale — if there are a finite number of identified, predictable solutions with a reasonable expectation of success. Novo Nordisk now imposes a higher standard for nonobviousness of combination therapies, as there is no clear evidence on record showing a finite number of predictable solutions.

Synergy is often not predictable, and not necessarily shared by all members of a drug class. This was noted by Judge Pauline Newman in her dissent, which pointed out that only some sulfonylureas formed synergistic combinations. Nonetheless, the Novo Nordisk majority held that a general knowledge of synergy from some members of a drug class is sufficient to meet the "predictable" solution requirement. Thus, unpredictability based on the evidence that synergy is not a general property shared by all members of the class may not be sufficient to overcome prima facie obviousness. Under Novo Nordisk, synergistic results among some members of the class were at least enough to make the combination obvious to try.

Furthermore, structural differences may not be sufficient to support unpredictability either, as repaglinide and sulfonylureas clearly belonged to two distinct subclasses of secretagogues. They only shared the same end result (or the same functional class) — stimulation of insulin release; and that was deemed sufficient for a finding of obviousness.

Synergy between different classes of cancer drugs has been investigated based primarily on their known mechanisms of action. For example, the therapeutic effect of certain alkylating agents (such as cyclophosphamide) plus purine analogs (such as fludarabine) has been investigated. These studies likely have created a high bar for patenting combination therapies using other alkylating agents and purine analogs. Under Novo Nordisk, if two classes of compounds are generally expected to have some beneficial or synergistic effect, combination therapy using one compound from each class is likely obvious. Thus, evidence of synergism, without more, may not be sufficient to overcome a rejection based on obviousness, because such an effect arguably is expected based on each drug's functional class.

Under these circumstances, nonobviousness would require that the synergy be truly unpredictable, or the results be greater than an expected level of synergy. For example, one might have to show that a specific combination achieves superior results as compared to the whole class in general. Parallel comparative studies may be required to show that a specific combination produces pharmacological activities that are significantly higher than the rest of the members of the class or other combinations of drugs from the same class.

However, relying on superior results to overcome prima facie obviousness may force an enablement issue. For example, if superior results are derived from a limited number of experiments, the results may be treated as not commensurate with the full scope of the claims. Alternatively, superior results may be derived from in vitro data, and may be treated as not credible if there is evidence to indicate that the in vitro results are not reasonably predictive of in vivo efficacy, which is increasingly more often the case.

Strong evidence of unpredictability in combination therapy — other than structural differences or lack of synergy in some members of the class — is also desirable. For example, it has been reported that certain ratios of combined drugs can be synergistic while other ratios of the same agents may be antagonistic, implying that the most efficacious combinations may be those that use certain agents at reduced doses. Such evidence may allow an applicant to claim combination therapy based on an optimal dosage — or a range of effective doses — for each drug. However, the applicant should be mindful to include all clinically meaningful doses, if possible, to avoid easy design-around by competitors.

Judith L. Toffenetti is a partner at McDermott Will & Emery. She is focuses on life sciences and biotechnology. Cynthia Chen is an associate who focuses on patent prosecution and portfolio development and management in the fields of biotechnology and pharmaceuticals.