Braintree v. Novel: When “a patient” excludes one patient

Background

Braintree Labs is the assignee of U.S. Pat. No. 6,946,149 (“the ‘149 patent”), entitled “Salt Solution for Colon Cleansing,” and the holder of a New Drug Application for SUPREP (magnesium sulfate anhydrous, potassium sulfate, sodium sulfate) oral solution for use in cleansing of stool from the colon in preparation for a colonoscopy. The SUPREP kit contains two six-ounce bottles, which are diluted and administered approximately 12 h apart.

Several months after Braintree received approval of its NDA, Novel Labs filed an Abbreviated New Drug Application containing a paragraph IV certification against the ‘149 patent which is listed in the Orange Book with respect to SUPREP. See id. at 1353. In response to this notice of Paragraph IV certification, Braintree sued Novel in the U.S. District Court for the District of New Jersey asserting infringement of the ‘149 patent under 35 U.S.C. § 271(e)(2). See id.

Initial claim construction—District court

On 19 September 2012, the district court issued a Markman opinion, 2012 WL 4120907 (D.N.J., 19 September 2012), in which it construed four claim terms highlighted in independent claims 15 and 18 of the ‘149 patent:

15. A composition for inducing purgation of the colon of a patient, the composition comprising about 100 ml to about 500 ml of an aqueous hypertonic solution comprising an effective amount of Na₂SO₄, an effective amount of MgSO₄, and an effective amount of K₂SO₄, wherein the composition does not produce any clinically significant electrolyte shifts and does not include phosphate.

18. A composition for inducing purgation of the colon of a patient comprising from about 100 ml to about 500 ml of an aqueous hypertonic solution consisting essentially of an effective amount of Na₂SO₄, an effective amount of MgSO₄, and an effective amount of K₂SO₄, wherein the composition does not produce any clinically significant electrolyte shifts and does not include phosphate.

With respect to the term “clinically significant electrolyte shifts,” the district court rejected constructions put forth by both parties in favor of language from the specification. “Clinically significant electrolyte shifts” was construed to mean “alterations [in] blood chemistry that are outside the normal upper or lower limits of their range or other untoward effects.” Id. at *6.

Summary judgment of infringement—District court

On 18 January 2013, the district court granted Braintree’s motion for summary judgment of infringement. See Braintree Labs., Inc. v. Novel Labs., Inc., No. 11-1341 (PGS), 2013 WL 211252, at *19 (D.N.J., 18 January 2013), vacated and remanded, 749 F.3d 1349 (Fed. Cir. 2014). Novel argued that clinical data provided by Braintree showed that SUPREP causes shifts in electrolyte level outside the normal range in some patients; in response, the court clarified its construction of the term “clinically significant electrolyte shifts,” stating that in its view, the term “necessarily connotes something more than just electrolyte shifts outside normal ranges.” Id. at *6. The court ruled that the phrase “other untoward effects,” found in the specification and in its Markman construction, implied a lack of the severe and life-threatening side effects that were present in some formulations recited in the prior art. See id. These side effects included disturbances in cardiac and renal function, sometimes resulting in death. See id. at *7. Thus, the court concluded, “clinically significant electrolyte shifts” meant “alterations in blood chemistry that are both outside the normal upper or lower limits of their range and accompanied by or manifested as other untoward effects, such as disturbances in cardiac and renal function as disclosed in the patent specification.” Id. (emphasis added).

However, the court’s infringement decision did not rest solely on this clarification. It also based its decision on the term “a patient,” found in the preamble of independent claims 15 and 18, and not originally briefed by the parties. See id. at *8. The court found that the term “a patient” should apply to the clause “wherein the composition does not produce any clinically significant electrolyte shifts,” such that this clause should be read “wherein the composition does not produce any clinically significant electrolyte shifts in a patient.” Id. Agreeing with Braintree that Federal Circuit precedent dictated that the term “a patient” should mean “one or more patients,” the court found that the term “a patient” was a limitation of the claims, and that the evaluation of electrolyte levels recited in claims 15 and 18 were meant to be done on a per patient basis. See id. Because there was no dispute that at least one patient to whom SUPREP would be administered would not show clinically significant electrolyte shifts, and Novel’s product was an exact copy of SUPREP, the court granted Braintree’s summary judgment motion of infringement. See id.

The Federal Circuit vacates in part

Novel appealed the judgment of infringement to the Federal Circuit. On 22 April 2014, a Federal Circuit panel comprised of Judges Prost, Dyk, and Moore issued an opinion vacating the decision in part and remanding to the district court for further proceedings. See 749 F.3d 1349 (Fed. Cir. 2014). The Federal Circuit majority took issue with the district court’s clarified construction of “clinically significant electrolyte shifts,” holding that the patentee’s lexicography must govern the analysis. Id. at 1356. As such, the construction “alterations in blood chemistry that are both outside the normal upper or lower limits of their range and accompanied by or manifested as other untoward effects, such as disturbances in cardiac and renal function as disclosed in the patent specification” was reversed, and changed back to “alterations in blood chemistry that are outside the normal upper or lower limits of their range or other untoward effects,” as the district court had originally construed the term. Id.

Turning to the claim term “a patient,” the majority ultimately construed this term to mean “the general class of persons to whom the patented compositions are directed, i.e., a patient population.” Id. at 1357 (emphasis added). As evidence, the majority cited portions of the patent’s specification that indicated that the object of the invention was “a well tolerated orally administered colonic purgative that … avoided the risks of upset or electrolyte balance in patients.” Id. (emphasis and alteration in original). Because there was some evidence in the record that SUPREP caused some alterations in blood chemistry in some patients that was outside the normal range, the majority concluded that there was a genuine dispute as to whether SUPREP was covered by the claims of the patent, and remanded to the district court for further proceedings. Id.

However, Judge Moore authored a dissent where she disagreed with the majority’s construction of the term “a patient.” Specifically, she thought that the district court was correct, and that “a patient” should be construed to mean “one or more” patients. “[T]here is no plain meaning of ‘a’ that excludes the singular.” Id. at 1365 (Moore, J., dissenting). Judge Moore argued that the specification’s use of the term “patients” when it meant a group was indication that the patentee knew the difference between the terms “a patient” and “patients,” and drafted the claims accordingly. The dissent pointed to other parts of the specification that indicated, e.g., that the invention avoids electrolyte shifts that are clinically significant to the patient. See id. at 1366 (quoting the specification, where it said “One purpose of the present research was to develop a safe, effective, and well tolerated small volume solution made up of a high concentration of poorly absorbable salts that induce colon cleansing catharsis after oral ingestion without clinically significant alternation [sic] of sodium, chloride, bicarbonate, potassium, calcium, and phosphate level and balance or other untoward effects on the recipient” (emphasis in original)). Furthermore, Judge Moore argued that the majority’s concern with finding that a composition that met the claims 1 out of 100 times could infringe was a question of damages, not infringement. See id. And Judge Moore noted that while this was a Hatch-Waxman case, where a de facto injunction would issue (by virtue of a delay of approval of Novel’s ANDA), that was not a reason to rewrite the claims, even though she would dislike that outcome. See id. at 1367. Finally, Judge Moore struggled with the meaning of “patient population,” and asked whether the majority’s interpretation of “a patient” to mean a patient population would result in a claim that is indefinite or unsupported. See id. at 1367 n.3 (Moore, J., dissenting) (“What percentage of people constitutes a patient population? Is the patient population everyone who takes the drug? A statistically significant number of people who take the drug—which would be 99.5%? A majority? And have we now written an indefinite or unsupported claim for the patentee?”).

Braintree’s petition for rehearing and rehearing en banc

Braintree petitioned for a pane rehearing and rehearing en banc to address whether the term “a patient” should be construed to mean “more than one” and to exclude “one.” See Petition for Panel Rehearing and Rehearing En Banc, No. 2013-1438, ECF # 62 (Fed. Cir., 22 May 2014). Braintree proposed two resolutions. First, that the decision be vacated with respect to the construction of the term “a patient” only, and to remand to the district court for a Markman hearing on that term, which was not originally briefed by the parties below. See id. at 1–3. Alternatively, Braintree asked that the court of appeals stay resolution of the case until the Supreme Court’s decision in Teva Pharms. USA, Inc. v. Sandoz, Inc., No. 13-854 (U.S.), which was poised to (and ultimately did) resolve the issue of whether claim construction decisions should be reviewed completely de novo, or whether deference should be given to factual findings at the district court level. See id. at 4.

Braintree’s petition generally tracked the arguments raised by Judge Moore in her dissent. Braintree argued that in the past, the Federal Circuit had only deviated from the general rule that “a” means “one or more” when it established through intrinsic evidence that “a” should mean “one and only one.” See id. at 5. But, Braintree argued, the Federal Circuit had never interpreted “a” to exclude the singular. See id. According to Braintree, the panel majority’s construction was seemingly motivated by concerns of an “absurd result” where it would be possible to practice the claims of the patent while avoiding clinically significant electrolyte shifts in only 1% of the patient population. Id. at 7–8. However, Braintree argued, this was impermissible results-based claim construction, and furthermore, was completely hypothetical, since the record established that many patients taking SUPREP experienced no clinically significant electrolyte shifts. Id. at 8. Moreover, Braintree argued, the reality of drug therapy is that the same drug can and does have different effects in different people: even if only a small portion of patients experienced no clinically significant electrolyte shifts, de minimis infringement would still be actionable under Federal Circuit precedent. Id. at 8–9. Braintree requested to have the claim construction issue sent back to the district court for supplemental briefing on the term “a patient,” rather than accepting the panel’s construction that proceeded on a barren record below. See id. at 7–10.

Finally, Braintree raised the possibility of this decision effectively requiring a new method of establishing infringement. Before the panel majority’s decision, Braintree argued, “a claim drafter, patentee, or other interested party had no reason to doubt that infringement could be demonstrated by reference to one or more patients; there was no need to define a specific ‘patient population.’” But following the decision, Braintree argued, based on Judge Moore’s dissent, “the proof required to establish infringement in such cases likely will require parties to present and district courts to consider statistical analyses.” See id. at 12–13.

BIO and PhRMA’s amicus brief

The Biotechnology Industry Organization (“BIO”) and the Pharmaceutical Research and Manufacturers of America (“PhRMA”) also filed an amicus brief in an effort to obtain clarity on the construction of the term “a patient.” See Brief of the BIO and PhRMA as Amici Curiae in Support of Rehearing or Rehearing En Banc, No. 2013-1438, ECF # 72 (Fed. Cir., 5 June 2014). The brief attempted to highlight the absurdity of construing “a patient” to exclude the singular. For example, the brief states that “the claim refers to ‘the colon of a patient.’ A population obviously does not share a single colon.” Id. at 4. The brief also argued that this decision led to a slippery slope where every construction of the word “a” was potentially implicated, e.g. “a cell,” “an animal,” or even “a telecommunications device.” One interpretation of each of these could be “a population.” Id. at 4–5. Further complicating matters was the lack of guidance as to what constituted a population—would a “population” of wires include both thick titanium wires and thin steel wires? Id. at 5.

Bio and PhRMA also argued that the majority’s rationale ignored basic science. For example, would a cancer drug that only worked in a small fraction of the population be denied protection? And would an analysis need to determine which patients in a study actually benefitted from the drug, and which benefitted from a placebo effect? Id. at 6–7.

Novel’s rebuttal

Novel submitted a response. See Defendant-Appellant’s Response to Petition for Rehearing and Rehearing En Banc, No. 2013-1438, ECF # 81 (Fed. Cir., 3 July 2014). In response to Braintree’s petition and the amicus brief, Novel argued that the clear language of the patent showed that the claimed invention was meant to be directed to a population, and precluded a composition being covered by the claims if it exhibited any clinically significant electrolyte shifts. See id. at 2–5. Novel contended that as long as even one patient showed a clinically significant electrolyte shift, the composition did not meet the claims. See id. Furthermore, Novel argued, the intrinsic evidence discloses that the intention of the inventor was to find a new treatment applicable to a population of individuals. For example, the patent states that “its objective was to find a well-tolerated orally administered colonic purgative that was as effective as the well-known hypertonic phosphate ravages [sic], that avoided the risks of upset of electrolyte balance in patients.” See id. at 6.

Novel also distinguished Braintree’s de minimis argument as misplaced. In previous de minimis infringement cases, each administration would contain a minuscule amount of infringing product. See id. at 7. But in a situation like the one envisioned by the Federal Circuit majority, 99% of administrations would be completely free of infringement, but Braintree would still obtain an injunction because of the 1% of administrations that might infringe. See id. at 7–8. To hold that the inventors only intended to prepare one patient for a colonoscopy, rather than a population, would be the truly absurd result.

Finally, Novel noted that the cases relied on by Braintree to ostensibly show that the term “a” must be construed as “one or more” all referred to constituent parts of a product or process. See id. at 9–10. Not so with “a patient,” which is not a component of an overall device or method. See id.

Denial of rehearing and cert. petition

On 1 August 2014, the Federal Circuit issued a per curiam order denying Braintree’s petition. See Order on Petition for Rehearing En Banc and Petition for Panel Rehearing and Rehearing En Banc, No. 2013-1438, ECF # 85 (Fed. Cir., 1 August 2014). Braintree subsequently filed a petition for certiorari on 30 October 2014 with the U.S. Supreme Court, arguing that this case presented the same question as Teva v. Sandoz, namely, whether claim constructions should be reviewed de novo, or deference should be given to factual findings made by the district court. See Braintree Laboratories, Inc. v. Novel Laboratories, Inc., 2014 WL 5489479 (U.S.). Braintree argued that had a clear error standard been applied, the Federal Circuit would have had to defer to the district court’s construction of “clinically significant electrolyte shifts,” because that determination was made after the court made factual findings about the credibility of the expert testimony that was introduced, and would have remanded for findings on the construction of the term “a patient.” Id. at *2.

Braintree’s petition was denied on 8 December 2014. See Braintree Labs., Inc. v. Novel Labs., Inc., 135 S. Ct. 764, 190 L. Ed. 2d 630 (2014).

Remand to the District of New Jersey—Braintree wins

The construction of “a patient” now being finalized, the parties turned to the question of how to apply this unusual construction to the facts of their case. Specifically, how would one show that clinically significant electrolyte shifts were happening in a population of patients?

On 6 January 2015, Braintree and Novel filed short five-page briefs on the question of whether SUPREP causes clinically significant electrolyte shifts in a patient population. See Trial Brief of Plaintiff Braintree Labs., Inc., No. 11-1341, ECF # 430 (D.N.J., 6 January 2015); Defendant Novel Labs., Inc.’s Trial Brief, No. 11-1341, ECF # 431 (D.N.J., 6 January 2015). Braintree set out to prove that statistical analysis was the proper method to analyze whether clinically significant electrolyte shifts were taking place in a population, specifically, a statistical analysis of mean electrolyte levels, and that SUPREP’s mean electrolyte values remained within the normal range. See ECF # 430 at 1. Novel, on the other hand, wanted to show that individual patients were experiencing clinically significant electrolyte shifts, and that those patients were a patient population, i.e. they were members of “the general class of persons to whom the patent compositions are directed” See ECF # 431 at 1-2. Through this, Novel hoped that the court would find that SUPREP is not a composition that does not cause any clinically significant electrolyte shifts, and thus falls outside the claims of the patent. See id. at 2.

On 1 June 2015, Judge Sheridan found for Braintree. After defining “patient population” to include “a person that a doctor has said has a need for a colonoscopy,” but who does not have a contraindication for SUPREP, the court found that Braintree’s experts were more credible than Novel’s. See Braintree Labs., Inc. v. Novel Labs., Inc., No. 3:11-CV-01341-PGS, 2015 WL 3492936, at *13 (D.N.J., 1 June 2015) at ¶¶ 3, 20. The court adopted Braintree’s expert’s analysis, which measured changes in the mean response, rather than Novel’s expert’s analysis, which measured the percent of the population outside the normal range. See id. at ¶¶ 32-67. In particular, Braintree’s expert found that after administration of SUPREP, none of the mean serum electrolyte levels in any of the SUPREP clinical trials fell outside the upper or lower limits of the normal range for any of seven different electrolytes. Id. at ¶ 56. As such, the court found that the results of his analysis showed that SUPREP did not produce any “clinically significant electrolyte shifts” in the patient population under the Federal Circuit’s claim constructions. Id. at ¶ 67.

Judge Sheridan found that SUPREP met every limitation of claims 15 and 18 of the ‘149 patent, that Novel’s ANDA product, if sold, would infringe those claims, and would induce infringement of the method of use claims 19, 20, and 23. Id. at ¶¶ 68-69. The injunction against Novel’s products was also reinstated after a motion by Braintree. See Final Order of Amended Judgment, No. 11-1241 (PGS) (LHG), ECF # 467 (D.N.J., 25 June 2015).

Further developments and conclusion

On 2 June 2015, Novel appealed to the Federal Circuit, arguing that (a) the trial court did not decide whether SUPREP causes “untoward effects,” (b) the trial court misapplied the claim construction, and (c) the trial court did not consider invalidity under section 112. See Principal Br. for Appellant Novel Labs., Inc., No. 2015-1700, ECF # 33 (Fed. Cir., 9 September 2015) at 2-4. The appeal is currently stayed pending modification of the judgment in the district court. See Order, No. 2015-1700, ECF # 90 (Fed. Cir., 10 August 2016). Regardless, one thing is clear—the construction of the term “a patient” to mean “patient population” is now the law in the Braintree case, and may inform claim construction of that term for years to come. However, to date, courts are refusing to extend Braintree beyond its facts. See Allergan Sales, LLC v. Sandoz Inc., No. 2:12-CV-207-JRG, 2016 WL 1224868, at *15 (E.D. Tex., 29 March 2016) (“Braintree does not establish a rule that the term ‘a patient’ (or ‘a person’) always refers to ‘a general class of persons to whom the patented compositions are directed, i.e. a patient population.’ … Rather, it simply stands for the proposition that this was the appropriate construction based on the facts in that case.”); Recro Tech. LLC v. Actavis Labs. FL, Inc., No. CV 14-1118-GMS, 2015 WL 9590585, at *1 n.1 (D. Del., 29 December 2015) (declining to construe “a subject” to mean “the general class of persons to whom the claimed compositions are directed,” and stating that “the court is not convinced that the Federal Circuit intended Braintree to affect a massive change in the law of claim construction.”) Time will tell whether Braintree is an aberration based on particular facts or a harbinger of things to come.