« Shoxing | Main | Damaged Damages »

October 18, 2008

Term Limits

One might naively imagine that the patent office would encourage patenting, and seek to offer its clientele, namely inventors and patent prosecutors, every advantage. Today's USPTO is quite the opposite. In this episode, the patent office is caught stealing patent term, unlawfully limiting the duration of patent grants.

Wyeth v. Jon Dudas, USPTO (D.C. 07-1492)

35 U.S.C. § 154 [is] the statute that prescribes patent terms. Section 154(a)(2) establishes a term of 20 years from the day on which a successful patent application is first filed. Because the clock begins to run on this filing date, and not on the day the patent is actually granted, some of the effective term of a patent is consumed by the time it takes to prosecute the application. To mitigate the damage that bureaucracy can do to inventors, the statute grants extensions of patent terms for certain specified kinds of PTO delay, 35 U.S.C. § 154(b)(1)(A), and, regardless of the reason, whenever the patent prosecution takes more than three years. 35 U.S.C. § 154(b)(1)(B). Recognizing that the protection provided by these separate guarantees might overlap, Congress has forbidden double-counting: "To the extent that periods of delay attributable to grounds specified in paragraph (1) overlap, the period of any adjustment granted under this subsection shall not exceed the actual number of days the issuance of the patent was delayed." 35 U.S.C. § 154(b)(2)(A). Plaintiffs claim that the PTO has misconstrued or misapplied this provision, and that the PTO is denying them a portion of the term Congress has provided for the protection of their intellectual property rights.

Exactly so. But first, a bit more history.

Statutory Scheme

Until 1994, patent terms were 17 years from the date of issuance. See 35 U.S.C. § 154 (1992) ("Every patent shall contain . . . a grant . . . for the term of seventeen years . . . of the right to exclude others from making, using, or selling the invention throughout the United States. . . ."). In 1994, in order to comply with treaty obligations under the General Agreement on Tarriffs and Trade (GATT), the statute was amended to provide a 20-year term from the date on which the application is first filed. See Pub. L. No. 103-465, § 532, 108 Stat. 4809, 4984 (1994). In 1999, concerned that extended prosecution delays could deny inventors substantial portions of their effective patent terms under the new regime, Congress enacted the American Inventors Protection Act, a portion of which -- referred to as the Patent Term Guarantee Act of 1999 -- provided for the adjustments that are at issue in this case. Pub. L. No. 106-113, §§ 4401-4402, 113 Stat. 1501, 1501A-557 (1999).

As currently codified, 35 U.S.C. § 154(b) provides three guarantees of patent term, two of which are at issue here. The first is found in subsection (b)(1)(A), the "[g]uarantee of prompt Patent and Trademark Office response." It provides a one day extension of patent term for every day that issuance of a patent is delayed by a failure of the PTO to comply with various enumerated statutory deadlines: fourteen months for a first office action; four months to respond to a reply; four months to issue a patent after the fee is paid; and the like. See 35 U.S.C. § 154(b)(1)(A)(i)-(iv). Periods of delay that fit under this provision are called "A delays" or "A periods." The second provision is the "[g]uarantee of no more than 3-year application pendency." Under this provision, a one-day term extension is granted for every day greater than three years after the filing date that it takes for the patent to issue, regardless of whether the delay is the fault of the PTO.1 See 35 U.S.C. § 154(b)(1)(B). The period that begins after the three-year window has closed is referred to as the "B delay" or the "B period". ("C delays," delays resulting from interferences, secrecy orders, and appeals, are similarly treated but were not involved in the patent applications underlying this suit.)

1 Certain reasons for exceeding the three-year pendency period are excluded, see 35 U.S.C. § 154(b)(1)(b)(i)-(iii), as are periods attributable to the applicant's own delay. See 35 U.S.C. § 154(b)(2)(C).

The extensions granted for A, B, and C delays are subject to the following limitation:

(A) In general.--To the extent that periods of delay attributable to grounds specified in paragraph (1) overlap, the period of any adjustment granted under this subsection shall not exceed the actual number of days the issuance of the patent was delayed.

35 U.S.C. § 154(b)(2)(A). This provision is manifestly intended to prevent double-counting of periods of delay, but understanding that intent does not answer the question of what is double-counting and what is not. Proper interpretation of this proscription against windfall extensions requires an assessment of what it means for "periods of delay" to "overlap."

In other words, the statute was poorly written, not clear as to method.

As a basis for determination, the court adopted a hypothetical example used by Wyeth, illustrated below, showing "A delays," which are Office delays beyond statutory requirement, either in examination or other Office tasks. A "B delay" covers the period beyond "3-year application pendency," in other words, where it took more than 3 years to grant a patent. A one-day extension is granted for each day past the 3-year B period.

As shown, the PTO considered the "B period" to be "the entire period during which the application was pending before the office." Thus, the "B delay" would overlap any "A delay." Since overlap is denied by statute, the PTO would deny any "B delay." In writing, the PTO stance may appear arguable, but graphically, as above, it looks ludicrous.

Wyeth argued that ""A period" and "B period" overlap only if they occur on the same calendar day or days."

The PTO's interpretation would give the example patentee a three-year extension, while Wyeth's method would give four.

Judge James Robertson appeared sympathetic to the PTO, though one can wonder the degree to which it was feigned.

The PTO's interpretation appears to be driven by Congress's admonition that any term extension "not exceed the actual number of days the issuance of the patent was delayed," and by the PTO's view that "A delays" during the first three years of an applications' pendency inevitably lead to "B delays" in later years. Thus, as the PTO sees it, if plaintiffs' construction is adopted, one cause of delay will be counted twice: once because the PTO has failed to meet and administrative deadline, and again because that failure has pushed back the entire processing of the application into the "B period." Indeed, in the example set forth above, plaintiffs' calendar-day construction does result in a total effective patent term of 18 years under the (B) guarantee, so that - again from the PTO's viewpoint -- the applicant is not "compensated" for the PTO's administrative delay, he is benefitted by it.

But that was not the way the law was written.

If the outcome commanded by that text is an unintended result, the problem is for Congress to remedy, not the agency.

So the USPTO was thwarted. But already thousands of patents may have been denied their proper term. Alas, the practical impact of the Wyeth ruling is likely quite limited, as 37 C.F.R. § 1.705(d) requires that "request for reconsideration of the patent term adjustment indicated in the patent must be filed within two months of the date the patent issued."

The court considered a moot point that otherwise has considerable bearing in the deference that a federal agency, in this instance, the USPTO, is given to interpret statute, the so-called Chevron deference. The point was moot "because it cannot be reconciled with the plain text of the statute." But whether Chevron deference applies to the PTO is very important in whether the agency may be challenged in its rule-making. This issue is crucial in the Tafas/GSK rules imbroglio, where the PTO is attempting to limit claims and continuations. The PTO was denied in DC district court, ruling that the patent office may only make procedural changes, not ones of substance. Hence the new limits the PTO proposed were tossed as substantive, not merely procedural. That ruling is now on appeal.

Since at least 1996, the Federal Circuit has held that the PTO is not afforded Chevron deference because it does not have the authority to issue substantive rules, only procedural regulations regarding the conduct of proceedings before the agency. See Merck & Co. v. Kessler, 80 F.3d 1543, 1549-50 (Fed. Cir. 1996).

Chevron deference does not apply to the interpretation at issue here.

This case also covered in Patent Docs.

Posted by Patent Hawk at October 18, 2008 11:46 AM | The Patent Office

Comments

Post a comment




Remember Me?

(you may use HTML tags for style)