holding that Medtronic "overrules everything in [Michael v. Shiley, Inc., 46 F.3d 1316 (3d Cir. 1995)] that would prevent a plaintiff from pursuing a cause of action for fraudulent misrepresentation based on common law principles"Summary of this case from In re Orthopedic Bone Screw Products
Argued: June 9, 1998.
Filed: November 17, 1998.
On Appeal from the United States District Court for the Eastern District of Pennsylvania, (D.C. No. MDL 1014).
Michael D. Fishbein, Esq. (Argued), Arnold Levin, Levin, Fishbein, Sedran Berman, 510 Walnut Street, Suite 500, Philadelphia, PA 19106, Counsel for Appellant.
John S. Buresh, Esq. (Argued), Fred M. Feller, Buresh, Kaplan, Jang, Feller Austin, 2298 Durant Avenue, Berkeley, CA 94704.
George P. Noel, Esq., Noel Hackett, 25 East Second Street, P.O. Box 1590, Media, PA 19063, Counsel for Appellee, Buckman Co.
Anthony C. H. Vale, Esq. (Argued), Pepper, Hamilton Scheetz, 18th Arch Streets, 3000 Two Logan Square, Philadelphia, PA 19103-2799, Counsel for Amicus-Appellees, Danek Med. Inc., Sofamor SNC, Sofamor Danek Grp., Warsaw Orthopedic, Sofamor, Inc.
Robert E. Welsh, Jr., Esq., Welsh Recker, 1818 Market Street, Suite 3402, Philadelphia, PA 19103, Counsel for Appellee, Louis C. Bechtle.
Janet L. MacDonell, Esq., Deutsch, Kerrigan Stiles, 755 Magazine Street, New Orleans, LA 70130, Counsel for Appellee, Scoliosis Research Society.
Norman J. Jeddeloh, Esq., Burditt Radzius, 333 West Wacker Drive, Suite 2600, Chicago, IL 60606, Counsel for Appellee, American Academy of Orthopaedic Surgeons.
BEFORE: STAPLETON, COWEN and RENDELL, Circuit Judges.
This case involves thousands of individual plaintiffs-appellants who claim that they suffered injuries resulting from the implantation of orthopedic bone screws into the pedicles of their spines. Over 2,000 individual actions were consolidated for pre-trial proceedings pursuant to the multi-district litigation statute, 28 U.S.C. § 1407. This appeal involves plaintiffs' allegations that one defendant, Buckman Company, Inc. ("Buckman"), made misrepresentations to the Food and Drug Administration ("FDA") serious enough to play a substantial role in the events which resulted in their injuries. The district court granted Buckman's motion to dismiss this state-law "fraud on the FDA" claim, determining that: (1) the claim was "precluded by virtue of the fact that the MDA [i.e., the Medical Device Amendments of 1976, 21 U.S.C. §§ 360c-360k, to the Federal Food, Drug, and Cosmetic Act of 1938 ("FDCA"), 21 U.S.C. § 301 et seq.] does not provide for a private right of action," App. at 157; and (2) plaintiffs could not, as a matter of law, demonstrate proximate causation between the misrepresentations to the FDA and any injuries suffered. App. at 158-59. We will reverse.
The orthopedic bone screw devices that are the subject of this suit are regulated by the FDA pursuant to the MDA. Congress enacted the MDA to address concerns regarding the safety and effectiveness of the wide variety of medical devices introduced into the market. See Medtronic, Inc. v. Lohr, 518 U.S. 470, 476 (1996). The MDA requires classification of medical devices into three categories based upon the risk that they pose to the public. Class I devices present no unreasonable risk of illness or injury and are subject only to general manufacturing controls.
See 21 U.S.C. § 360c(a)(1)(A). Class II devices are potentially more harmful and are subject to federal performance regulations. See id. § 360c(a)(1)(B). Finally, Class III devices "present a potential unreasonable risk of illness or injury" and are subject to the strictest regulation. Id. § 360c(a)(1)(C).
Class III devices may not be introduced into the market until the manufacturer provides the FDA with "reasonable assurance" that the device is both safe and effective. See id. § 360e(d)(2). This "premarket approval" process requires a manufacturer to submit, and the FDA to review, "all information, published or known . . . or which should reasonably be known . . . concerning investigations which have been made to show whether or not such device is safe and effective." Id. § 360e(c)(1)(A). In addition, the manufacturer must provide the FDA with a full statement of design, manufacturing, and marketing plans, as well as any additional information requested by the agency. Securing premarket approval is an arduous and time-consuming process; each submission requires an average of 1,200 hours of FDA review. See Lohr, 518 U.S. at 477.
There are three important exceptions to the requirement that Class III medical devices receive premarket approval before being placed on the market. First, a manufacturer can obtain an Investigational Device Exemption ("IDE"), which allows limited use of an experimental medical device to gather the type of data necessary to support a premarket approval application. See 21 U.S.C. § 360j(g). Though an IDE allows use of a Class III device in clinical trials, it does not permit introduction to the general public.
Second, the MDA allows Class III devices that were in commerce prior to its enactment to remain on the market until the FDA initiates and completes a premarket approval analysis for those "predicate" devices. See id. § 360e(b)(1)(A). This "grandfathering" provision reflects Congress' recognition "that existing medical devices could not be withdrawn from the market while the FDA completed its PMA analysis for those devices." Lohr, 518 U.S. at 477-78.
Third, to prevent manufacturers of grandfathered devices from monopolizing the market while new devices wait for FDA approval, and to ensure that improvements to existing devices are introduced quickly, the MDA allows devices that are "substantially equivalent" to an existing predicate device to avoid the premarket approval process. See Lohr, 518 U.S. at 478; 21 U.S.C. § 360e(b)(1)(B). The procedure a manufacturer follows to take advantage of this exception is known colloquially as the "§ 510(k) process," after the number of the relevant section in the original Act. A § 510(k) application must contain information supporting the device's substantial equivalence to a predicate device, proposed labeling for the device, and any additional information requested by the FDA. See 21 C.F.R. § 807.87. For a device to be approved under the § 510(k) process, the FDA must determine that the new device has the same intended use as the predicate device and that it possesses the same technological characteristics or is as safe and effective as the predicate device. See 21 U.S.C. § 360c(i)(1)(A).
A device is "substantially equivalent" to a predicate device if it:
(i) has the same technological characteristics as the predicate device, or
(ii) (I) has different technological characteristics and the information submitted that the device is substantially equivalent to the predicate device contains information, including appropriate clinical or scientific data if deemed necessary by the Secretary . . ., that demonstrates that the device is as safe and effective as a legally marketed device, and (II) does not raise different questions of safety and effectiveness than the predicate device.
21 U.S.C. § 360c(i)(1)(A).
The advantage of the § 510(k) process is significant to manufacturers: review of a § 510(k) application by the FDA requires an average of only twenty hours of agency time, compared to the 1,200 hours required for full premarket approval. See Lohr, 518 U.S. at 478-79.
Buckman is a regulatory consultant to medical device manufacturers, and was retained by the AcroMed Corporation to act as its liaison to the FDA in its attempt to secure marketing clearance for its product. AcroMed initiated the § 510(k) process to secure marketing clearance for its orthopedic bone screw device, known as the Variable Screw Placement ("VSP") Spinal Plate Fixation System, in September 1984. In this application, AcroMed indicated that it intended to market the device for use in spinal surgery. The FDA rejected the request, determining that the VSP device was a Class III device and was not substantially equivalent to any predicate device.
A second device, known as the ISOLA Spine Fixation System, is also at issue in this appeal. Since the facts alleged regarding both systems are substantively similar, our discussion of the VSP system applies with equal force to the ISOLA system.
In September 1985, AcroMed, through Buckman, filed a second application for marketing clearance through the § 510(k) process. The application provided additional information about the VSP device and again indicated its intended use in spinal surgery. The FDA again rejected the application, determining that the device was not substantially equivalent to a predicate device and that it posed potential risks not exhibited by other spinal fixation systems.
In December 1985, AcroMed and Buckman filed two final § 510(k) applications seeking market approval for its product. AcroMed and Buckman split the VSP device into its component parts, renamed them "nested bone plates" and "cancellous bone screws" and filed a separate § 510(k) application for each component. In both applications, a new intended use was specified: rather than seeking clearance for spinal applications, they sought clearance to market the plates and screws for use in the long bones of the arms and legs. AcroMed and Buckman claimed that the two components were substantially equivalent to predicate devices used in long bone surgery. The FDA approved the devices for this purpose in February 1986.
The subject of this appeal is plaintiffs' fraud on the FDA claim against Buckman. Plaintiffs allege that AcroMed, through Buckman, made material misrepresentations in its successful § 510(k) submission to the FDA. Specifically, plaintiffs allege that AcroMed, through Buckman, "sought approval of its VSP plates and screws for use in long bones simply as a pretext in order to market the device for its true intended use in the spine." App. at 58. According to the plaintiffs, Buckman, AcroMed and other defendants contrived a strategy in which they would (1) misrepresent the intended use of the screw system to the FDA, (2) obtain the FDA's approval to market the system for that misrepresented use, and (3) subsequently market the system for a use, spinal surgery, for which they could not obtain the FDA's approval. Plaintiffs contend that, as a result of the misrepresentations to the FDA, a § 510(k) clearance was issued, allowing the devices access to the market they would not otherwise have had.
In 1994, the Judicial Panel on Multi-district Litigation designated the district court as the transferee court for In re: Orthopedic Bone Screw Liability Litigation, MDL No. 1014. Since that time, some 2,300 civil actions have been transferred to the district court. In March 1995 the district court granted a motion brought by AcroMed and other defendants to dismiss plaintiffs' fraud on the FDA claims, determining that those claims (1) were preempted by the MDA, and (2) impermissibly implied a private right of action for violation of the FDCA. See App. at 28 (Pretrial Order No. 12). Relying on our decision in Michael v. Shiley, Inc., 46 F.3d 1316 (3d Cir. 1995), the district court's order purported to "dismiss all fraud on the FDA . . . claims contained in any pleading" in the consolidated litigation. App. at 37.
Subsequently, in June 1996, the Supreme Court decided Medtronic, Inc. v. Lohr, 518 U.S. 470 (1996), in which the Court addressed the scope of MDA preemption for the first time. As a result of that decision, plaintiffs filed amended complaints which re-asserted fraud on the FDA claims against Buckman and other defendants. AcroMed and other defendants filed a motion to reaffirm Pretrial Order No. 12 in light of Lohr. Buckman joined this motion, but also simultaneously filed its own motion to dismiss the fraud on the FDA claims against it. The district court granted AcroMed's motion and reaffirmed Pretrial Order No. 12, though on modified grounds. The district court recognized that Lohr likely invalidated one of the two grounds for its prior decision. Nevertheless, it concluded that Lohr did not completely undermine Pretrial Order No. 12 because Lohr "did not address claims of fraud on the FDA with respect to the absence of a private right of action for violations of the FDCA." App. at 157. The district court determined that allowing a fraud on the FDA claim would essentially create a private right of action for violation of the FDCA and that, "in view of the FDA's exclusive prosecutorial discretion . . . any grant of an implied private right of action would be contrary to the letter and spirit of the statute." App. at 153.
In addition, the district court determined that dismissal of plaintiffs' fraud on the FDA claims was warranted because "the alleged fraud cannot be said to have been a proximate cause of plaintiffs' alleged injuries." App. at 158. The court reasoned that since any misrepresentations were directed only to the FDA and any injuries were caused only by unrelated physicians, the connection between the alleged fraud and plaintiffs' injuries "is so tenuous, if existent at all, that the former cannot be said to be a `substantial factor' in causing the latter." App. at 159 (citations omitted).
After granting AcroMed's motion to dismiss the fraud on the FDA claims, the district court subsequently granted Buckman's identical motion to dismiss "for the reasons stated in Pretrial Order No. 845." App. at 162 (Pretrial Order No. 900). Thereafter, the district court certified Pretrial Order No. 900 as a final order pursuant to Fed. R. Civ. P. 54(b). See App. at 163 (Pretrial Order No. 1067).
Certification was proper for Pretrial Order No. 900 because plaintiffs asserted no other claims against Buckman other than the fraud on the FDA claims. As a result, Buckman is the only proper party to this appeal. Although Sofamor, S.N.C. and Danek Medical, Inc. submitted a joint brief and an appendix as appellees, we treated those documents as submissions of amici curiae and issued an appropriate order to that effect.
We have jurisdiction over the final order of the district court pursuant to 28 U.S.C. § 1291. We exercise plenary review over a dismissal for failure to state a claim, and we accept all factual allegations of the plaintiffs as true and draw all reasonable inferences therefrom in their favor. See Trump Hotels Casino Resorts, Inc. v. Mirage Resorts Inc., 140 F.3d 478, 483 (3d Cir. 1998).
Count I of the complaint tendered to us as representative is entitled "Fraud (on the FDA)." App. at 55. It alleges that: (1) "[i]n response to the FDA's inquiry concerning the intended use of the bone plates and bone screws, . . . Buckman, acting as representative of AcroMed . . . intentionally and falsely represented to the FDA that . . . [they were] `intended for use in appropriate fractures of long bones of both the upper and lower extremity and such other flat bones,'" App. at 58; (2) "Buckman . . . intended for the FDA to clear the VSP device components based upon this misrepresentation," id.; (3) "[i]n reliance on this express misrepresentation, the FDA determined . . . that AcroMed's nested bone plates and cancellous bone screws were substantially equivalent to devices marketed . . . prior to May 28, 1976 for the intended use represented by AcroMed" and issued a § 510(k) clearance, id.; (4) "the VSP device was intended exclusively for use in the spine," id.; (5) "the only purpose for which AcroMed's plates and screws were sold was for use in the spine," id.; (6) "[t]he FDA was ignorant of the fact that these devices and device components were intended by AcroMed to be used as pedicle screw fixation devices," id. at 63; (7) "[w]ere it not for these fraudulent acts and statements, the FDA would not have issued § 510(k) clearances for AcroMed's pedicle screw fixation devices for any purpose, the devices would not have been introduced into interstate commerce, and Plaintiff would not have been exposed to the dangerous device which was surgically implanted in the Plaintiff's spine," id.; and (8) "[a]s a direct and proximate result of the wrongful conduct alleged in Count I of this Complaint, Plaintiff has suffered, and will continue to suffer, severe physical harm, including injury to Plaintiff's spine," id. at 63-64.
Count I is thus drafted to track the elements of a common law cause of action for fraudulent misrepresentation: (1) a representation of fact, opinion, intention or law; (2) knowledge of its falsity; (3) an intent to induce reliance; (4) justifiable reliance; and (5) resulting injury. See Restatement (Second) of Torts § 525 et seq.
At the time of the decision here appealed, the bone screw cases were still at the pleading stage. While it is clear that the plaintiffs' "fraud on the FDA" claims are based on state rather than federal law, neither the parties nor the district court have yet addressed the potential choice of law issues and those issues are not before us. We note at the outset, however, that the controlling state law of fraudulent representation in some cases may be different from that in other cases. This is important to our analysis because the district court's dismissal of all "fraud on the FDA" claims was appropriate only if federal law forecloses resort to state law or if the allegations comprising those claims do not state a claim on which relief can be granted under any of the potentially applicable state laws. We first address the issue of whether federal law forecloses resort to state law and then turn to the issue of whether the allegations of the complaints may state a claim under any of the potentially applicable state laws.
The MDA contains the following preemption provision:
Except as provided in subsection (b) of this section, no State or political subdivision of a State may establish or continue in effect with respect to a device intended for human use any requirement —
(1) which is different from, or in addition to, any requirement applicable under this chapter to the device, and
(2) which relates to the safety or effectiveness of the device or to any other matter included in a requirement applicable to the device under this chapter.
Subsection (b) of § 360k grants authority to the Secretary of Health and Human Services to grant exemptions from preemption and is not pertinent here.
The FDA regulations interpreting § 360k of the MDA provide in part as follows:
State or local requirements are preempted only when the Food and Drug Administration has established specific counterpart regulations or there are other specific requirements applicable to a particular device under the act, thereby making any existing divergent State or local requirements applicable to the device different from, or in addition to, the specific Food and Drug Administration requirements. There are other State or local requirements that affect devices that are not preempted . . . because they are not "requirements applicable to a device" within the meaning of section [360k] of the act. The following are examples. . . :
(1) Section [360k] does not preempt State or local requirements of general applicability where the purpose of the requirement relates either to other products in addition to devices (e.g., requirements such as general electrical codes, and the Uniform Commercial Code (warranty of fitness)), or to unfair trade practices in which the requirements are not limited to devices.
(2) Section [360k] does not preempt State or local requirements that are equal to, or substantially identical to, requirements imposed by or under the act.
In Medtronic, Inc. v. Lohr, 518 U.S. 470 (1996), the Supreme Court applied these statutory and regulatory provisions in the context of a suit against the manufacturer of a pacemaker that had been cleared as a Class III device under § 510(k). Lohr, the plaintiff there, asserted common law negligence and strict liability claims based on the design of the pacemaker as well as state law claims based on defective manufacture and mislabeling. The Court held that these claims were not preempted. The Court gave the preemption provision of the statute a narrow reading. It stressed that preemption occurs under § 360k only when (1) there is a state "requirement" "with respect to a device;" (2) there is a federal "requirement" "applicable to the device" under the MDA; (3) the state requirement is "different from, or in addition to" the federal requirement; and (4) the state requirement relates to the safety or effectiveness of the device or some matter included in the federal requirement. See 518 U.S. at 500.
The Court found that issuance of a § 510(k) clearance "did not `require' Medtronic's pacemaker to take any particular form for any particular reason; the agency simply allowed the pacemaker, as a device substantially equivalent to one that existed before 1976, to be marketed without running the gauntlet of the MDA process." 518 U.S. at 483-84. The § 510(k) process thus established no federal requirement "applicable to a device" within the meaning of the MDA. "Similarly, the general state common law requirements [relied upon by Lohr] were not specifically developed `with respect to' medical devices. Accordingly, they [were] not the kinds of requirements that Congress and the FDA feared would impede the ability of federal regulators to implement and enforce specific federal requirements." Id. at 501. The general obligations imposed by the state common law relied upon by Lohr were "no more a threat to federal requirements than would be a state-law duty to comply with local fire prevention regulations and zoning codes, or to use due care in the training and supervision of a work force. These state requirements therefore escape pre-emption, not because the source of the duty is a judge-made common-law rule, but rather because their generality leaves them outside the category of requirements that § 360k envisioned to be `with respect to' specific devices such as pacemakers." Id. at 501-02.
The Court also held that "[n]othing in § 360k denies [a state] the right to provide a traditional damages remedy for violations of common-law duties when those duties parallel federal requirements. . . . The presence of a damages remedy does not amount to the additional or different `requirement' that is necessary under the statute; rather, it merely provides another reason for manufacturers to comply with identical existing `requirements' under federal law." Id. at 495.
Based on Lohr, it is apparent that, within the meaning of § 360k, there is no federal "requirement" "applicable to the device" at issue here; nor is there a state "requirement" "with respect to" that device. Moreover, the state common law relied upon does not impose any obligation on Buckman inconsistent with federal law. As plaintiffs stress, 18 U.S.C. § 1001 makes it a crime to make a fraudulent statement to a federal agency and 21 C.F.R. § 807.87(j) requires every pre-market notification to contain a statement that the information contained therein is believed to be truthful.
Neither the district court nor Buckman takes issue with any of these conclusions concerning the effect of Lohr on the "fraud on the FDA" claims in this suit. The district court expressly acknowledged that Lohr was inconsistent with one of the two grounds it originally articulated for dismissing these claims. It nevertheless concluded that Lohr was consistent with dismissal because its alternative ground remained unaffected by Lohr. As the district court put it, "[w]hile Lohr may have affected the portion of [the court's prior decision] that relied on the doctrine of preemption, the portion of the court's reasoning based on the absence of an implied right of action [under the FDCA] remains in tact." App. at 152-53. We disagree.
It is, indeed, well established that Congress has not created an express or implied private cause of action for violations of the FDCA or the MDA. It is also true, as the district court pointed out, that this court, in Michael v. Shiley, Inc., 46 F.3d 1316, 1329 (3d Cir. 1995), found this fact significant in the course of holding that a common law fraud claim against the manufacturer of a heart valve was preempted by the FDCA. We there concluded:
Under § 360k, states may not impose different requirements and thereby reach a different conclusion than the FDA. "[W]here the FDA was authorized to render the expert decision . . ., it, and not some jury or judge, is best suited to determine the factual issues and what their effect could have been on its original conclusions." King [v. Collagen Corp., 983 F.2d 1130,] 1140 [(1st Cir. 1993)] (Aldrich Campbell, J.J. concurring). Under the MDA, states have no authority to police Shiley's compliance with the FDA's procedures. If Shiley knowingly misled the FDA in its PMA application, it is for the FDA to remedy that situation using the authority Congress gave it in the MDA.
Further, permitting a fraud claim based on false representations to the FDA would conflict with our precedent that plaintiffs not bring implied causes of action for violations of the Food, Drug and Cosmetic Act. Gile v. Optical Radiation Corp., 22 F.3d 540, 544 (3d Cir.), cert. denied, ___ U.S. ___, 115 S.Ct. 429, 130 L.Ed.2d 342 (1994). Plaintiffs cannot circumvent this bar by characterizing their cause as a claim for state law fraud.
Michael, 46 F.3d at 1329. As we have noted, the district court recognized that the first reason we gave for our holding in Michael is untenable after Lohr. Nevertheless, it believed that it remained bound by our alternative rationale.
It is important to note at the outset that, contrary to the district court's suggestion, our alternative rationale in Michael, like our primary one, was necessarily based on the doctrine of preemption. Absent preemption by federal law (or a ground for stay or abstention, doctrines not pertinent here), a district court having jurisdiction of the parties and the subject matter cannot decline to enforce liability imposed by the relevant state common law. See, e.g., Polselli v. Nationwide Mutual Fire Insurance Co., 126 F.3d 524, 527 n. 1 (3d Cir. 1997) ("A federal court exercising diversity jurisdiction must apply the substantive law of the state whose laws govern the action."). Cf. Erie R.R. Co. v. Tompkins, 304 U.S. 64, 78 (1938) ("Except in matters governed by the Federal Constitution or by Acts of Congress, the law to be applied in any case is the law of the State."). This is important to understand because "`[t]he purpose of Congress is the ultimate touchstone' in every pre-emption case." Lohr, 518 U.S. at 485 (quoting Retail Clerks v. Schermerhorn, 375 U.S. 96, 103 (1963). Unless Congress intends preemption, there is none. Accordingly, where Congress has expressed itself on preemption in a statute, "Congress' intent, of course, primarily is discerned from the language of the pre-emption statute and the `statutory framework' surrounding it." Id. at 486 (quoting Gade v. National Solid Wastes Management Assn., 505 U.S. 88, 111 (1992) (Kennedy, J., concurring)).
In Michael, we inferred from the fact that Congress created no private right of action for violations of the FDCA that it intended to foreclose suits based on state common law that would provide a remedy for conduct violative of that act. Ironically, a plurality of the Supreme Court in Lohr drew a diametrically opposed inference from the same fact. It viewed Congress' failure to provide a federal remedy as persuasive evidence of an intent not to preempt common law liability for the same conduct:
Under Medtronic's view of the statute, Congress effectively precluded state courts from affording state consumers any protection from injuries resulting from a defective medical device. Moreover, because there is no explicit private cause of action against manufacturers contained in the MDA, and no suggestion that the Act created an implied private right of action, Congress would have barred most, if not all, relief for persons injured by defective medical devices. Medtronic's construction of § 360k would therefore have the perverse effect of granting complete immunity from design defect liability to an entire industry that, in the judgment of Congress, needed more stringent regulation in order "to provide for the safety and effectiveness of medical devices intended for human use," 90 Stat. 539 (preamble to Act). It is, to say the least, "difficult to believe that Congress would, without comment, remove all means of judicial recourse for those injured by illegal conduct," Silkwood v. Kerr-McGee Corp., 464 U.S. 238, 251 (1984), and it would take language much plainer than the text of § 360k to convince us that Congress intended that result.
Lohr, 518 U.S. at 487 (footnote omitted) (plurality).
More fundamentally, however, Lohr teaches that where Congress has expressed its intention with respect to preemption, we should look primarily to what it said. When it looked to what Congress said in § 360k, the Supreme Court in Lohr concluded that Congress did not intend to prevent a state from providing "a traditional damages remedy for violations of common-law duties when those duties parallel federal requirements." Lohr, 518 U.S. at 495.
Refusing to entertain Buckman's fraudulent misrepresentation claim solely because the statutory scheme does not contain a private cause of action would be the equivalent of finding preemption of state law claims contrary to the clear holding of Lohr. In short, Lohr overrules everything in Michael that would prevent a plaintiff from pursuing a cause of action for fraudulent misrepresentation based on common law principles.
We respectfully disagree in this regard with the conclusion reached by the Court of Appeals for the Seventh Circuit in Mitchell v. Collagen Corp., 126 F.3d 902 (7th Cir. 1997). It there concluded that our holding in Michael survives Lohr. 126 F.3d at 914. While it did not explain its conclusion, one can infer from its earlier decision in the same case, 67 F.3d 1268, that it believes the first rationale for our decision in Michael retains vitality.
Buckman advances a number of preemption arguments in addition to the one adopted by the district court. In each instance, we conclude that Lohr counsels rejection.
Buckman argues that Lohr is distinguishable (1) because it "did not involve claims turning on FDA procedures, but rather claims that the product itself was defective," Buckman Br. at 12, and (2) because the target of the alleged fraud here is "a creature of statute" which has the sole responsibility for enforcing the FDCA and the regulation promulgated thereunder. Buckman Br. at 10.
Buckman and amici (in support of a similar argument) make specific reference to 21 U.S.C. § 337(a) and § 336. Section 337(a) provides:
Except as provided in subsection (b) of this section, all such proceedings for the enforcement, or to restrain violations, of this chapter shall be by and in the name of the United States.
Section 336 provides:
Nothing in this chapter shall be construed as requiring the Secretary to report for prosecution . . . minor violations of this chapter whenever he believes that the public interest will be adequately served by a suitable written notice or warning.
Once again, we look first to Congress' express message concerning preemption — § 360k. In the context of the text of that section as construed in Lohr, Buckman's suggested distinctions are simply unpersuasive. Indeed, given that there must be state and federal requirements with respect to a device, it is harder to argue preemption under § 360k based on FDA procedures than based on FDA substantive requirements for a regulated device.
We do not rule out the possibility that there may be some areas of preemption outside the scope of § 360k based on clear and direct conflicts between the requirements of state law and those of the FDCA. The existence of § 360k, its relatively narrow scope, and Lohr's presumption against preemption of areas traditionally occupied by state law, however, counsel caution in finding implied preemption where no express preemption exists. Here, appellees' argument boils down to a contention that the litigation of suits of this kind is fundamentally inconsistent with the regulatory process established by the FDCA. We see no inconsistency between the FDA having the exclusive prerogative of bringing actions to enforce the FDCA and preserving the right of people in the plaintiffs' position to bring common law fraudulent misrepresentation claims. Moreover, we do not share appellees' apparent perception that litigation of such claims holds the potential for courts and juries second-guessing the FDA. Indeed, there is ample precedent, in related contexts, to support a claim premised on misrepresentations made to a federal agency. See, e.g., Learjet Corp. v. Spenlinhauer, 901 F.2d 198, 202-03 (1st Cir. 1990) (allowing aircraft purchaser's claim against manufacturer based on misrepresentation to Federal Aviation Administration); Stanton by Brooks v. Astra Pharmaceutical Products, 718 F.2d 553, 568-69 (3d Cir. 1983) (allowing patient's claims against drug manufacturer based on misleading the FDA); Hawkins v. Upjohn Co., 890 F. Supp. 609, 612 (E.D. Tex. 1994) (same).
Buckman also asks us to affirm on the alternative ground that the complaints do not state a claim for fraudulent misrepresentation. Three reasons for such a conclusion are urged. First, it is said that the complaints do not, and cannot, allege causation. Second, it is argued that, in the context of a § 510(k) application, a statement of the intended purpose of the applicant, as a matter of law, is not a representation of fact, opinion, or intention. Finally, it is argued that the complaints do not, and cannot, allege reliance by the plaintiffs on the alleged misrepresentation by the FDA.
The district court determined that plaintiffs' fraud on the FDA claims should be dismissed because "the alleged fraud cannot be said to have been a proximate cause of plaintiffs' alleged injuries." App. at 158. In the court's view, "[l]egal causation largely fails because the object of plaintiffs' fraud-on-the-FDA claim is the FDA and not the doctors who performed surgeries on the plaintiffs or the plaintiffs themselves. . . . The causal connection between the alleged fraud committed on the FDA and plaintiffs' alleged injuries is so tenuous, if existent at all, that the former cannot be said to be a `substantial factor' in causing the latter." App. at 158-59. While we are not in a position to canvass all the potentially applicable law, what we know about tort law generally makes us unwilling to say that all of the plaintiffs' claims will fail for want of the kind of a causation that will give rise to liability.
First, the mere fact that the alleged fraudulent misrepresentation was made to the FDA and not the plaintiffs does not necessarily preclude a finding of legally sufficient causation. Section 310 of the Restatement (Second) of Torts states, for example:
An actor who makes a misrepresentation is subject to liability to another for physical harm which results from an act done by . . . a third person in reliance upon the truth of the representation, if the actor
(a) intends his statement to induce or should realize that it is likely to induce action by . . . a third person, which involves an unreasonable risk of physical harm to the other, and
(i) that the statement is false, or
(ii) that he has not the knowledge which he professes.
Another example is provided by our decision in Stanton by Brooks v. Astra Pharmaceutical Products, 718 F.2d 553 (3d Cir. 1983). The plaintiff there, who had suffered brain damage as the result of an adverse reaction to the local anesthetic, Xylocaine, recovered a substantial judgment against the manufacturer of the drug. The plaintiff alleged that the defendant had misled the FDA by withholding information it was required to report concerning adverse effects previously suffered by others. We sustained the judgment against an attack that there was no legally sufficient causal connection between the defendant's misleading conduct and the plaintiff's injury. We concluded:
The jury heard the testimony of four well qualified expert witnesses . . . from which they could infer . . . (1) that had the FDA had these reports, it would have required notice to the medical community (through the package insert or PDR) of the critical information contained in the more than 200 adverse-reaction reports of the incidence of cardiac and respiratory arrest, notwithstanding low dosage of Xylocaine; and (2) that physicians receiving this information would have considered it in deciding how — and whether — to administer Xylocaine to their patients. Resolving every inference in plaintiffs' favor, we cannot conclude that the record was devoid of the requisite minimum quantum of evidence supporting the verdict.
718 F.2d at 569 (footnote omitted).
We have earlier reviewed the specific allegations of the representative complaint in this matter regarding causation. Plaintiffs' theory is that, without Buckman's fraudulent representation, the FDA would not have given AcroMed a § 510(k) clearance and that, absent a § 510(k) clearance, AcroMed would not have access to the market to promote its devices for use in spinal surgery and plaintiffs would not have suffered injury. Given the terminology chosen, it appears to us that the district court viewed the causation standard for unintentional torts as the applicable one and regarded the yielding of plaintiffs' physicians to AcroMed's promotion as an intervening cause. It may be that plaintiffs' theory would be so viewed under the law of some jurisdictions. It is not clear to us, however, that it would be so viewed under all of the potentially applicable law. Buckman's representation to the FDA regarding intended use is alleged to have been intentional. Moreover, while it is not alleged that harm to plaintiffs was intended, plaintiffs do claim to be able to show that Buckman's representation was a necessary element of a conspiracy to get doctors to do exactly what plaintiffs' physicians did. See, e.g., Restatement (Second) of Torts § 435 B ("Unintended Consequences of Intentional Invasions") and § 442 B ("Intervening Force Causing Same Harm as That Risked by Actor's Conduct").
We do not hold that plaintiffs have or have not alleged a legally sufficient causal nexus. That issue can be resolved only after the controlling law has been identified in each case. We do hold that the district court erred in determining at this stage that all of the plaintiffs have failed to allege a legally sufficient causal nexus.
Buckman insists that a statement of intended use in connection with a § 510(k) application is not a representation of a fact or intent but, rather, a request for authority. In its view, the sole interest of the FDA in the applicant's intended use is to enable it to limit the authority granted to that which is sought and considered by the agency. In a related argument, Buckman contends that FDA reliance on the alleged misrepresentation is "negated in this case by . . . the nature of the clearances that were given to AcroMed," Buckman Br. at 23, i.e., they were limited to long bone use only.
We, of course, acknowledge that a statement of intended use in connection with a § 510(k) application has utility in communicating the authority sought and in permitting the FDA to appropriately limit any authority granted. At the same time, we believe that such a statement is a representation of fact that is necessarily important to the FDA as such.
The use to which a device will be put is of critical importance in the regulatory scheme. A device that presents no risk of injury when employed for one purpose may present great risk of injury when employed for another. For this reason a § 510(k) clearance will be issued only if a device is a substantial equivalent of a pre-existing device used for the same purpose. Moreover, a device cleared under § 510(k) cannot lawfully be marketed for a use other than the use specified in the clearance. Thus, if such a device is marketed for another purpose, the manufacturer may be civilly and criminally sanctioned and the marketing enjoined. See 21 U.S.C. §§ 331, 332.
According to Buckman, the FDA requires no representation from a § 510(k) applicant regarding its marketing intentions and reads its representation with respect to "intended use" as no more than a description of the authority sought. Implicit in this view is the assertion that the marketing plans of a manufacturer seeking access to the market are of no interest to the FDA. If this be taken as an assertion of fact by Buckman, it is inappropriate at this stage because it conflicts with the allegations of the complaint. Plaintiff Donald G. Patterson, for example, alleges that "Were it not for [the misrepresentations that the applicant's intended use was in repairing fractures of the long bones], the FDA would not have issued 510(k) clearances . . . for any purpose, the devices would not have been introduced into interstate commerce, and Plaintiff would not have been exposed to the dangerous device which was surgically implanted in the Plaintiff's spine." App. at 63.
To the extent Buckman's argument is an assertion of law, we find no basis for it in the statute, in the regulations, or in any of the authorities cited in the briefing. To the contrary, such an assumption is fundamentally at odds with the statutory scheme. Under that scheme, the FDA is a gatekeeper charged with the responsibility of protecting the public from unreasonable risks of injury from medical devices. A manufacturer of a new device has no access to the market for any purpose absent some kind of FDA approval. Because the degree of risk presented by any device is a function of the use to which it is put, the intended use is thus a central focus in the FDA review. Indeed, there is nothing more material to its determination of whether to grant access to the market.
It is undoubtedly true, as Buckman insists, that a manufacturer unable lawfully to secure a § 510(k) clearance for one purpose may legally be entitled to a § 510(k) clearance for the same device when put to a different purpose. The allegation here, however, is that AcroMed had no lawful access to the market for any purpose and that it secured such access only by representing that it intended to market its device for one use when, in fact, it intended at the time of the application to market that device solely for another, more hazardous use. If a manufacturer in AcroMed's position is successful in its scheme, we know that, when its marketing campaign is detected post-issuance, the FDA is entitled to an injunction stopping that campaign. Under Buckman's view of the law, however, if the FDA is astute enough to uncover the scheme during its review, it is powerless to deny access to the market in the first place. We decline to accept a proposition so fundamentally at odds with the agency's gatekeeping assignment.
The situation here alleged bears no resemblance to one in which an applicant for a § 510(k) clearance intends to market its device for the use represented, subsequently markets for that use, and, at the same time, responds to direct requests for information from physicians concerning off-label use. As a result, we do not share our dissenting colleague's concern that recognizing a statement of intended use in an application for § 510(k) clearance as a representation of the applicant's marketing intentions will chill the dissemination of information relevant to off-label use.
Finally, Buckman argues that on the facts alleged there was no reliance by the plaintiffs on Buckman's statement to the FDA concerning intended use. This is true. This does not, however, negate the possible existence of a claim upon which relief can be granted. As we have earlier noted, § 310 of the Restatement (Second) of Torts suggests that when someone makes a fraudulent misrepresentation for the purpose of inducing action by A which, if taken by A, will involve an unreasonable risk of injury to B, B may recover for resulting physical injury even though he was unaware of the fraudulent misrepresentation. It is possible that the controlling law in one or more of these cases takes a similar approach.
Our holding is a narrow one. We do not hold that any of the plaintiffs have stated a claim under state law upon which relief could be granted. Neither we nor the district court is in a position to do that before identifying, with the help of the parties, the governing law or laws. Rather, we hold that (1) the plaintiffs' "fraud on the FDA" theory of liability is not so at odds with traditional principles of tort law that Buckman is entitled to a dismissal of all claims against it at this stage; and (2) if the state law of fraudulent misrepresentation applicable in one or more of these cases would impose liability on Buckman in the circumstances alleged, that law would not be preempted by the MDA. The judgment of the district court will be reversed, and the case will be remanded for further proceedings consistent with this opinion.
Nor do we hold, as suggested by the dissent, that any of the plaintiffs have stated a claim under the FDA's regulations. Plaintiffs do not purport to state such a claim and, as we have earlier explained, controlling precedent bars recognition of a private cause of action under the FDCA, the MDA, or their implementing regulations.
The majority endorses a claim of "fraud on the FDA" under circumstances that will expose manufacturers to fraud liability for seeking desirable innovations in a product's use, distort the penalty scheme established by the FDCA and its regulations, and generate substantial liability when manufacturers respond to doctors' widely accepted practice of purchasing medical products for off-label uses.
To appreciate how much this case departs from previous cases involving "fraud on the FDA," it is important to see what all agree is not fraud in this case. When Buckman sought approval from the FDA for using its screws on long bones, Buckman did not present any false evidence about the similarity of the screws to the already approved long bone screws, and there is no dispute that AcroMed's screws are "substantially equivalent" to the pre-existing screws. The FDA's determination that AcroMed's screws should be approved for long-bone use has never been questioned.
The basis for the majority's conclusion that Buckman committed fraud on the FDA is that in seeking long-bone approval AcroMed and Buckman really intended to market the screws for spinal use. That is, once the screws could be sold legally for long-bone use, AcroMed planned surreptitiously to induce physicians to buy the screws for spinal use. Clearly, if AcroMed actively marketed the screws for spinal use in violation of FDA's limited approval for long-bone use, AcroMed (and Buckman to the extent it participated in the plan) would have violated the FDA's marketing rule and could be held accountable by the FDA. A violation of the marketing rules, however, is not fraud on the FDA; it is an ordinary transgression of an agency's regulatory rules. AcroMed and Buckman also could be held liable by doctors and, in most states, their patients if AcroMed and Buckman told doctors that the bone screws were approved by the FDA for spinal use or made other material misrepresentations. But the statements made to doctors are not the basis for the majority's theory of "fraud on the FDA." To the contrary, the majority finds fraud in Buckman and AcroMed's statement to the FDA of the intended purpose of the screws in their application for long bone approval. It is on this point that the majority creates dangerous confusion.
The mere fact that AcroMed or Buckman really wanted approval for the spinal use of the screws does not make their request for long-bone approval fraudulent. Seeking approval for one use should not preclude a manufacturer, via the penalty of fraud liability, from seeking approval for another use for the same product. Expanding the uses for existing products will often be a very desirable source of progress and innovation in medical products. A manufacturer may even deliberately start off by seeking approval for a conventional use while waiting to gain approval for a different, novel and potentially highly profitable use of the product. Such a strategy could have the beneficial effect of increasing competition in the market for the conventional use while at the same time providing revenue to make it easier to negotiate the arduous FDA approval process. By reading the phrase "intended use" in the application as a declaration of what the manufacturer wants ultimately to seek with the medical device, rather than simply as a request for a specific FDA approval (with nothing implied about future requests), the majority creates an interpretation that is economically unrealistic and is likely to hinder technological growth.
The legal concept that the FDA was "defrauded" because a company desired approval for a different use from the one requested and ultimately approved is particularly hard to sustain in this case. Before the FDA approved the screws for long bones, it had twice previously rejected Buckman and AcroMed's applications for spinal use. Those prior applications made unmistakably clear to the FDA that Buckman and AcroMed wanted as their first choice approval for spinal use of the screws. There was no deception about AcroMed's ultimate goals. The FDA most assuredly knew what clearance was sought and what clearance was granted or denied.
The majority may believe that what distinguishes this case from one where a company simply seeks approval for one use while really wanting approval for another (and hence what prevents this decision from flatly prohibiting innovation), is that Buckman and AcroMed obtained long bone approval allegedly with the intention of illegally marketing the screws for off-label use. Such a theory, however, stretches the "intended use" statement into an all-purpose guarantee that an applicant will not violate other FDA rules. Under the majority's position the only basis for finding that AcroMed and Buckman committed fraud on the FDA is that they allegedly intended to violate the FDA's regulations prohibiting marketing for off-label uses. It is hard to see why that is not in effect creating a private right of action for a violation of the FDA regulation, and it is difficult to reconcile that conclusion with a point the majority itself made in footnote 8: "controlling precedent bars recognition of a private cause of action under the FDCA, the MDA, or their implementing regulations." Given the demanding detail of the regulatory process, and the wide variety of violations ranging from the innocuous to the serious, recognizing a state cause of action for violations of FDA regulations will greatly distort the penalty scheme established by the statute. The penalties attached to a violation of the FDA's regulations will often be substantially increased, and enforcement of violations will no longer be controlled by the FDA's prosecutorial discretion.
The majority's position is problematic in another closely related respect. The FDA and the medical community have long recognized the importance of doctors' off-label uses of medical products (uses not indicated on the label or approved by the FDA). See James M. Beck Elizabeth D. Azari, FDA, Off-Label Use, and Informed Consent: Debunking Myths and Misconceptions, 53 Food Drug L.J. 71, 72 (1998). Doctors may, in the exercise of their professional judgment, decide that based on current medical research an unapproved use of a product best serves a patient's needs.
The article explains that: "It is an accepted principle that once [the] FDA determines that a drug or device can be marketed, a physician's discretionary use of that product (the practice of medicine) is not restricted to the uses indicated on FDA-regulated labels. Off-label use is widespread in the medical community and often is essential to giving patients optimal medical care, both of which medical ethic[ists], [the] FDA, and most court recognize." Beck Azari, supra, at 72. The article points out that the FDA stated in a 1982 Drug Bulletin that: "`unapproved' or more precisely `unlabeled' uses may be appropriate and rational in certain circumstances, and may, in fact reflect approaches to drug therapy that have been extensively reported in medical literature. . .
Valid new uses for drugs already on the market are often first discovered through serendipitous observations and therapeutic innovations. . ." Beck Azari, supra, at 77. With regard to bone screws in particular, the article quotes the following statement by the FDA in a 1993 publication titled Food and Drug Administration, Update on Pedicle Screws: "In practice, surgeons often use orthopedic screws which FDA has cleared for other purposes . . . as pedicle screws. Such use of medical devices for non-approved purposes has traditionally been regulated by the hospitals in which the physicians practice and not by the FDA." Beck and Azari, supra, at 77.
Permitting off-label uses may seem anomalous given that the FDA's labeling and marketing restrictions prevent companies from actively pursuing an off-label market. Moreover, the fact that the FDA lacks authority to regulate the practice of medicine might imply that off-label uses amount to an end-run around the agency's authority. But the FDA's own position on off-label use reveals that off-label uses are a valuable part of the practice of medicine. The difficulty not confronted by the majority is that there is a very real tension created between endorsing off-label uses but preventing marketing for those uses. Close questions will be raised about whether a manufacturer was simply responding to doctors' desirable demand for off-label uses, or whether a manufacturer was impermissibly trying to create an off-label market and circumvent the FDA's regulation. As matters stood before this decision, the FDA was left to police the boundary between permissible and impermissible responses to off-label demand. The majority's theory, however, now throws into the jury box the question of when unacceptable "marketing" has taken place.
See Washington Legal Foundation v. Kessler, 880 F. Supp. 26, 28 n. 1 (D.D.C. 1995), Washington Legal Foundation v. Friedman, No. 94-1306, 1998 WL 456372, at *3 (D.D.C. July 30, 1998) and supra note 1.
Consider the problem of deciding when impermissible "marketing" has taken place in manufacturer-sponsored continuing medical education ("CME"). After manufacturers began in recent years financially underwriting more CME seminars and materials, the FDA developed, in response to Congressional hearings and ensuing confusion over an FDA concept paper, a twelve-factor test to decide when a manufacturer-sponsored CME seminar violated the FDA's rules on marketing. See Washington Legal Foundation v. Friedman, No. 94-1306, 1998 WL 456372, *4 (D.D.C. July 30, 1998) (FDA unconstitutionally restricted commercial speech because less restrictive means of achieving the government's interest were available). As is illustrated by both the complexity of the FDA's test and its efforts to hone its response, difficult and subtle balancing is called for in evaluating manufacturers' roles in the creation and distribution of information about off-label uses of drugs and medical devices. The problem is not easily dismissed. Once it is accepted that off-label uses are desirable, it is difficult to maintain that doctors should be shielded from truthful information concerning when and how to use a product for an off-label use. Patients will benefit by having their doctors informed about off-label uses.
Recent legislation which, once effective, will supersede the FDA's restriction on distributing printed material underscores the incongruity of the majority's position. The Food and Drug Modernization Act of 1997, Pub. L. No. 105-115, 111 Stat. 2296 (to be codified at 21 U.S.C. § 360aaa, et seq.) permits manufacturers to distribute information about off-label uses of drugs and devices provided certain requirements are met. Among the requirements for a drug, for instance, is that the manufacturer must have submitted an application for approval of the new, off-label use. Thus, manufacturers are allowed to submit applications for new uses and to distribute certain information about the off-label use while an application for the off-label use is still pending. On the majority's theory, will the manufacturer have committed fraud on the FDA, notwithstanding the new statutory provisions, if the manufacturer secures approval for one use, distributes materials on off-label uses in compliance with the statute, and does so while its primary objective is to obtain approval for the off-label use pending with the FDA?
Contrary to the majority's position, the problem of allowing juries to determine when a manufacturer has improperly advertised an off-label use is not only that it will impose too stringent a rule, chilling dissemination of valuable information about off-label uses; it is also that jury decisions lack the consistency of an agency-enforced rule. Even apart from the harm of restricting beneficial information about off-label uses, and the need for a consistently enforced rule wherever the line is drawn, a jury's assessment of liability is problematic in a third sense: Massive liability will be imposed where the FDA would not find any misconduct, and the cost of medical devices will be needlessly raised. It begs the question simply to respond, as the majority does in footnote 7, that a manufacturer will not face liability when it actually markets for the "intended use" listed in the 510(k) application and answers requests for information about off-label uses. The problem remains: the majority permits liability for fraud on the FDA when a jury concludes that a manufacturer conducted impermissible marketing, demonstrating that the manufacturer "defrauded" the FDA in its statement of intended purpose. Presumably, the majority's response is not that juries' interpretations of "marketing" will pose limited problems because a manufacturer can insulate itself from liability for fraud on the FDA by the simple act of making sales for the intended use listed in the 510(k) application. That seems like a very wooden distinction, for why should those sales be so important from the standpoint of liability for fraud on the FDA? Isn't the problem illegal marketing? If the majority does not require any showing of illegal marketing, how can liability be justified? The effect of the rule will be to prohibit manufacturers from making otherwise entirely legal responses to demand for off-label uses, responses that the FDA itself accepts as desirable. Or even worse, the rule will very directly prohibit innovation in the uses of drugs and devices by triggering fraud liability when an application is filed for approval of a new use. Once it is admitted that the majority's theory focuses on illegal marketing, we are thrown back on all the problems listed above. Nowhere does the majority demonstrate the need for its new theory of fraud on the FDA. The plaintiffs can be adequately compensated by their doctors under existing law for any malpractice demonstrated; the defendants can be held liable for fraud if they told doctors they had FDA approval that they did not; and the FDA remains free to prosecute when it feels impermissible marketing has taken place. Manufacturers also can be held liable under product liability law.
In sum, when juries are permitted to displace the FDA's judgment about whether a manufacturer has engaged in improper marketing, they will fail to provide a consistent standard, inhibit valuable exchange of information on off-label uses, and needlessly raise the price of drugs and medical devices. It seems very unlikely that Congress intended a state cause of action to intrude so much both in the enforcement of the FDCA's regulatory scheme and in the severity of the penalties attached to a violation.
I most respectfully, but vehemently, dissent.