This post is from the Reed Smith (and now we should add Cozen) side of the blog only, as Dechert is involved in the litigation to be discussed.

You’ll have to forgive us – we’re weird that way – but we found the opinion in Carter v. APP Pharmaceuticals, LLC, 2013 WL 5532767 (D. Ariz. Aug. 13, 2013), fascinating.  Also mysterious:  why did Carter take two months to show up on Lexis or Westlaw?  We can’t answer that, but we can tell you about the rest of the opinion. Carter deals with a new application of preemption, at least as far as we know.  The product, a solution used to keep catheters unobstructed while in use on trauma patients (and perhaps others), contained the long-approved drug heparin – an anticoagulant.  For about as long, heparin has been known to cause, in some people, a serious, potentially fatal condition:  heparin-induced thrombocytopenia and/or thrombosis (“HITT”).  Heparin’s label contained a lengthy and detailed warning about HITT, but plaintiff alleged that the nurses using the product missed the symptoms and as a result his arm had to be amputated.  2013 WL 5532767, at *1.

Yeah, we know.  The facts sound rather like Wyeth v. Levine, 555 U.S. 555 (2009) – right down to the plaintiff bringing a product liability action on allegations that sound more like medical malpractice.

The defendant’s product was approved by the FDA in a procedure we don’t think has ever come up in a preemption situation in a product liability case – the so-called “paper NDA.”  As described in Carter

[Defendant’s] heparin product was approved by the Food and Drug Administration (“FDA”) under a New Drug Application (“NDA”) filed pursuant to 21 U.S.C. §355(b)(2).  [Defendant] submitted an NDA because the drug’s container was a new mechanism for delivery but the heparin sodium incorporated into [defendant’s] heparin flush is the same as that contained in the product of the reference listed drug (“RLD”) holder [Defendant] states it was not required to, and did not, conduct additional clinical, safety and efficacy studies of the heparin sodium drug to acquire FDA approval. According to [defendant], its reliance on the RLD required that it use a label identical to the RLD holder’s label approved by the FDA. The RLD identified for heparin sodium . . . is a heparin product manufactured by [another company], not [defendant].  According to [defendant], it was required to match its warnings with [that other company’s label, including any updates.

*          *          *          *

The third kind of [prescription drug] application is based on 21 U.S.C. §355(b) (2).  This kind of application has been referred to as a “paper NDA.”  Under [this], a drug manufacturer may file an NDA for a drug that is not entirely new but is not simply a generic version of a branded drug.  The applicant provides certain information that allows it to avoid preclinical and human studies necessary to achieve a full NDA and shows the drug’s approval will not infringe a valid patent.

2013 WL 5532767, at *2, 5 (citations, quotation marks, and footnotes omitted).This “paper NDA” sounds a lot, to our non-FDA-regulatory-trained minds, like a cross between the (preempted – see PLIVA v. Mensing, 131 S. Ct. 2567 (2011)) generic and (unpreempted – see Wyeth v. Levine, 555 U.S. 555 (2009)) innovator drug processes.  Being intrigued, we Googled “paper NDA” for more background.  It turns out (at least according to Wikipedia) that the “paper NDA” actually was a shortcut devised by the FDA for generic drugs before there was any Hatch-Waxman Act.

Apparently, that process wasn’t very user-friendly – with only 15 paper NDAs between 1962 and 1984, when it was essentially superseded by H-W.  An FDA Guidance from 1984 seems to confirm this, stating that for all approvable paper NDAs in the pipeline, current procedures will apply, but anything new (or unapprovable) will be converted to H-W ANDAs.  So, from everything we’ve been able to turn up, it appears that paper NDAs are to ANDAs what transitional devices (devices approved as “drugs” prior to the 1976 Medical Device Amendments) are to PMA-approved medical devices.

Well, PMA preemption has been extended to transitional devices by every case (not all that many) that has considered the issue.  See Brooks v. Howmedica, Inc., 273 F.3d 785, 795-96 (8th Cir. 2001) (en banc) (pre-Riegel v. Medtronic, Inc., 552 U.S. 312 (2008); finding PMA preemption under Medtronic, Inc. v. Lohr, 518 U.S. 470 (1996)); Scott v. Pfizer Inc., 249 F.R.D. 248, 254-55 (E.D. Tex. 2008); Mathis v. E.I. DuPont de Nemours & Co., 2008 WL 162156, at *6 (W.D. La. Jan. 16, 2008); Rousseau v. Depuy Orthopaedics, Inc., 2006 WL 3716061, at *8 (W.D. La. Dec 13, 2006) (pre-Riegel; finding PMA preemption under Lohr).

How did “paper NDA” preemption fare in Carter?  Not very well, actually.  We think that the opinion resembles the First Circuit’s decision in Mutual Pharmaceutical Co. v. Bartlett, 678 F.3d 30 (1st Cir. 2012), rev’d, 133 S. Ct. 2466 (2013), in that the court stuck its anti-preemption head in the sand and refused to apply the impossibility rationale of Mensing to any fact pattern the Supreme Court had not ruled on – at least, that’s what we think is meant by “clear authority” in this quote:

The undisputed facts do not support Plaintiff’s argument that [defendant] filed a full NDA. . . .  The Court therefore concludes the undisputed facts show [defendant’s] heparin flush product was approved pursuant to the . . . “paper NDA” application process under 21 U.S.C. §355(b)(2) as [defendant] contends. The Court further concludes, however, Plaintiff’s claims are not preempted based on Mensing.  [Defendant] has not shown based on clear authority that its heparin flush product is a generic drug or that it was required to “match” the [RLD] label in all respects regarding its product.

Carter, 2013 WL 5532767, at *5.  That’s the sum total of the discussion of preemption in Carter – the defendant was right, and the plaintiff wrong, about the process that was at issue, but without unspecified “clear authority,” the opinion simply refuses to recognize preemption, no matter how close the analogy, and without explaining why.

Even though the defendant did not prevail on preemption in Carter – very unconvincingly – the opinion was interesting because it turned us onto a new application of preemption, which doesn’t happen very often.

But a novel preemption application wasn’t the end of the story in Carter.  Far from it.

The plaintiff also had significant warning causation problems, again stemming from the nature of the product – where the defendant’s label was identical other heparin products.  Indeed, we first came across Carter in the course of writing our post on physicians not reading warnings (see Arizona).  The plaintiff deposed no fewer than five − count ‘em, five – physicians involved in plaintiff’s care in a futile search for warning causation evidence.  But all five testified that they had never read, relied upon, or even seen the label for the defendant’s product:

• #1 – “no recollection” of defendant’s label; relied upon the PDR, “attending seminars, and reading medical journals”; “kept informed” about HITT.  2013 WL 5532767, at *7.
• #2 – “no recollection of seeing the [defendant’s] heparin label”; relied upon “on previous medical experience,” “on in-house pharmacists, the PDR, and on-line information sources”; “was aware of the risks of patients developing [HITT] following administration of heparin.”  Id.
• #3 – “never heard of [defendant] or seen the [defendant’s] label”: “relied on medical literature and pharmaceutical databases”; “was familiar with the causes and effects of [HITT].”  Id
• #4 – “never reviewed a heparin or [defendant’s] heparin package insert”; reviewed “medical journal articles”; “had advanced knowledge of [HITT]” and “was aware of [HITT]’s diagnostic criteria.”  Id.
• #5 – “not seen [defendant’s] label.”  Id.

Thus, the court in Carter concluded that “[t]he undisputed evidence shows these physicians did not rely on the [defendant’s] label regarding Plaintiff’s care.”  Id. The plaintiff in Carter did not get a pass on warning causation the way he did with preemption.  The heeding presumption had been rebutted:

Plaintiff’s alleged heeding presumption evidence does not establish a question of material fact regarding causation but even if the presumption is applied, Defendants have rebutted the presumption. Plaintiff’s physicians as discussed above did not read the [defendant’s] label that accompanied the heparin flush product. Any additional warning not to use the product to avoid the risks of developing . . . HITT would have been redundant of information already in the [defendant’s] label. Plaintiff’s physicians were aware of the risks associated with heparin and there is no evidence any physician would have changed his treatment method had additional warnings been provided.

Id. at *8.

Nor did Carter stop there. The plaintiff’s experts weren’t any good, either.  They had no competent opinions how any defect in the labeling could have caused the injury.  These opinions failed for three reasons:  (1) some described methods that were not the “standard of care,” either generally or according to the hospital where the defendant was being treated.  Id. at *9-10 (both of plaintiff’s experts); (2) there was no commercially available alternative to heparin flushes.  Id. (both experts); and (3) some of the opinions referenced information that already was “contained in the [defendant’s] label.”  Id. at *10 (one expert).

Finally, there was no causation because the relevant decision – about requiring heparin flushes of catheter tubing – was not made by the plaintiff’s doctors at all, but rather by the hospital.  The procedure in question “was authorized without a specific physician order.”  Carter, 2013 WL 5532767, at *10.

[U]sing heparin for arterial flushes to treat trauma patients was the common practice at [the hospital]. . . .  [T]he decision by [the hospital] to use a heparin flush to maintain patency of Plaintiff’s arterial line was not below the standard of care . . . . [T]he decision to use a heparin flush in Plaintiff’s treatment was not made by the physicians but by the hospital, other than on [one occasion].  Plaintiff has not established a causal link.

That’s an interesting causation point as well – one that we don’t see very often, no doubt due to the nature of the procedure in question. Some (more routine) uses of drugs may not be determined by physicians at all in hospitalization settings, but by the facility itself – and it would usually be rather difficult to pin an institution’s reliance on one particular company’s labeling for one particular drug.

Intriguing.