For those who find math daunting, determining whether the USPTO has correctly calculated patent term adjustment under 35 U.S.C. § 154(b) can create an Excedrin size-headache. Patent term adjustment adds term to the granted patent for any “period of delay” caused by the patent prosecution process which is due to USPTO. The problem is that there can be more than one “period of delay” that goes into the patent term adjustment formula.
A further complication is that 35 U.S.C. § 154(b) doesn’t permit double counting where these “periods of delay” overlap. So when does such “overlap” occur? That was at the center of the controversy in Wyeth and Elan Pharma International Limited v. Kappos involving U.S. Patent No 7,179, 892 (the ‘892 patent) and U.S. Patent No 7,189,819 (the ‘819 patent). In the end, the patentees (Wyeth and Elan Pharma) won the patent term adjustment battle in getting, respectively, 294 and 230 more days of patent term than the USPTO said they were entitled to under 35 U.S.C. § 154(b) for the respective patents.
In Wyeth, two different “periods of delay” were involved in the patent term adjustment determinations which will be referred to (for simplicity) as the A guarantee period and the B guarantee period. The A guarantee period addresses delays in meeting various examination deadlines. The B guarantee period addresses delays occurring 3 years after the filing date of the patent application until the patent issued.
The USPTO disagreed, pointing to the current version of 37 CFR § 1.703(f) which interpreted the B guarantee period to start at the time of the filing of the application, not 3 years after the filing date. In other words, each of the patents would only get the greater of the A or B guarantee period. Accordingly to the USPTO’s formula, that meant the ‘892 patent was entitled to only 462 additional days of term, while the ‘819 patent was entitled to only 492 additional days of term.
What made the current version of 37 CFR § 1.703(f) particularly controversial was a 2004 amendment to this regulation which changed “periods of adjustment” attributable to the “overlap” to “periods of delay.” In explaining the reason for this 2004 amendment, the USPTO said:The language of former [37 CFR] § 1.703(f) misled applicants into believing that [periods of A-delay] and [periods of B-delay] were overlapping only if the [period of A-delay] occurred more than three years after the actual filing date of the application. If an application is entitled to a [B-]adjustment . . . the entire period during which the application was pending before the [PTO] . . ., and not just the period beginning three years after the actual filing date of the application; is the period of delay under 35 U.S.C. [§ 154(b)].
After filing petitions for reconsideration of these patent term adjustments, Wyeth filed suit in the District Court for the District of Columbia (D.C. District Court). The D.C. District Court granted Wyeth’s motions for summary judgment in favor of the “A + B” formula for calculating the respective patent term adjustments for the ‘892 and ‘819 patents. The D.C. District Court found the USPTO was entitled to no Chevron deference in interpreting 35 U.S.C. § 154(b) according to the formula of 37 CFR § 1.703(f). In fact, even if Chevron deference was applicable, the D.C. District Court found the USPTO’s formula to be contrary to the “plain language” of 35 U.S.C. § 154(b): “The problem with the [USPTO]’s interpretation is that it considers the application delayed under [the B guarantee] during the period before [the application] has been delayed” (emphasis in the original).
The Federal Circuit agreed with the D.C. District Court: “a violation of the B guarantee—the one at the heart of the issue in this case—begins when the [USPTO] fails to issue a patent within 3 years after the actual filing date of the application” and “ends when the patent is issued.” In fact, the Federal Circuit characterized the “greater of A or B” interpretation by the USPTO in 37 CFR § 1.703(f) as “strained.”
The USPTO tried to defend its “strained” interpretation using several arguments. First, the USPTO pointed to alleged anomalous results “where a hypothetical patentee appears to receive some type of windfall adjustment under [35 U.S.C. § 154(b)] despite being in a similar position as other applicants who receive no similar adjustment.” But the Federal Circuit bluntly responded that 35 U.S.C. § 154(b) “requires as much,” and that the USPTO’s interpretation would also “result in some imbalanced treatment of similarly-situated patentees” under certain scenarios exemplified by the Federal Circuit.
Second, the USPTO urged support for its interpretation in the second clause of 35 U.S.C. § 154(b)(2)(A) which said “the period of any adjustment granted under this subsection shall not exceed the actual number of days the issuance of the patent was delayed.” Besides finding the USPTO’s argument on this point “unclear,” the Federal Circuit observed that this second clause only took effect if the first clause was satisfied (i.e., to the extent that the “periods of delay” attributable to each of the A and B guarantees overlapped). In other words, combining the A and B guarantee periods and then subtracting the “overlap” after the start of the B guarantee period satisfied with this second clause.
Third, the Federal Circuit found “nothing to rescue” the USPTO’s interpretation in the legislative history of 35 U.S.C. § 154(b). Fourth and not surprisingly, the Federal Circuit found that the USPTO’s interpretation was entitled to no Chevron deference. Instead, the Federal Circuit said: “Because the language of [35 U.S.C. § 154(b)] itself controls this case and sets an unambiguous rule for overlapping extensions, this court detects no reason to afford special deference to the PTO’s interpretation.” In fact, the Federal Circuit completely rebuked the USPTO’s formula in 37 CFR § 1.703(f) for calculating patent term adjustment by saying “this court accords no deference to the [USPTO]’s greater-of-A-or-B rubric.”
In the end, the “A + B formula” for calculating patent term adjustment prevailed in Wyeth. Put differently, the patentees had the correct math for 35 U.S.C. § 154(b).
*© 2010 Eric W. Guttag.