Thursday, May 06, 2010

Gadolinium and Spherical Error

Fritz Zwicky, the tart-tongued scientist (discoverer of, among other things, supernovae and neutron stars) was wont to label his critics in the astrophysical world (of whom there were many) “spherical bastards."  That was his shorthand for someone who was a “bastard, when looked at from any side.”

Hence the title of this post. We think that the recent decision in In re Gadolinium-Based Contrast Agents Products Liability Litigation, MDL No. 1909, slip op. (N.D. Ohio May 4, 2010), is a spherical error.  That is, it’s a decision that, no matter what direction we look at it, looks like error to us.

This goes beyond mere legal analysis, and encompasses a truly troubling disparity in the approach to defense and plaintiffs’ experts.  Leaving astrophysics for something less cosmic (but more interesting to us baseball fans) the plaintiffs’ experts got to pitch to a Kong Kingman strike zone.  E.g., Slip op. at 39-40. But when defense experts had to toe the same rubber, well it was Eddie Gaedel at the plate.  E.g., Id. at 52-53.

Read on, you’ll see what we mean.

That’s odd, because we looked at Judge Polster’s opinions to check his history was in product liability cases. We found nothing unusual in any past opinions. While Judge Polster doesn’t much like fraudulent misjoinder, he’s hardly alone in that.  He doesn’t have a long product liability track record, but he seemed okay in asbestos cases.

So we’re still scratching our heads at where this spherical error comes from.

We knew practically nothing about the Gadolinium MDL before the other day.  It had produced no opinions beside a few removal/remand decisions. Gadolinium itself is a “rare earth,” one of those oddballs that hang underneath the main periodic table, as Bexis found out about when his daughter told him she knew all the elements’ numbers by heart (it’s number 64, and, yes, she did know it).  Apparently, gadolinium’s magnetic properties make it a superior contrast agent when used in now ubiquitous resonance scans.

But gadolinium contrast agents are prescription drugs (at least the FDA appears to treat them as “drugs” rather than “devices”), and thus they have risks.  In this particular litigation, the alleged risk is of a new disease, as described by the FDA:

9. What is Nephrogenic Systemic Fibrosis (NSF)?

NSF was first described in the medical literature in 2000. The first case of NSF was identified in 1997. The cause of NSF is unknown but it has been reported only in patients who have severe kidney disease. NSF causes fibrosis of the skin and connective tissues throughout the body. . . . NSF usually starts in the lower extremities. Fibrosis can also develop in the diaphragm, muscles in the thigh and lower abdomen, and lung vessels. Over time, NSF becomes worse and can cause death.
Any “new disease” like NSF automatically gives us pause to wonder whether it’s a “made for litigation” disease like "pleural thickening" in asbestos cases or the purported “types” of autism dreamed up in vaccine litigation.  We were particularly suspicious given that from the opinion it appears the plaintiffs’ experts generated much of the medical literature.  But since the FDA treats NSF as a legitimate disease (at least in people suffering “acute or chronic severe” kidney disease, id. at FAQ 8), we will too.

A big problem in the Gadolinium litigation – as in so many other mass torts – is nobody really knows what causes the disease that the plaintiffs claim to have.  As the FDA says, “Whether the [gadolinium contrast agents] are the only agents or conditions that may be associated with NSF in patients with renal disease is unknown. The conclusions that can be drawn from the NSF reports are limited.”  Id. at FAQ 13. The medical literature “suggest[s] that they “play a role.” Id.

The plaintiffs’ experts, of course, are not nearly so cautious in their opinions, and that’s mostly what the Gadolinium opinion is about.

Gadolinium has a fairly long history of non-medical use, and is known to be toxic to humans in the environment.  Thus, when used for its approved medical applications it has to be “chelated” to become chemically inert.  That means, we gather, that it has to be stuck (chemically bound) inside of some much larger organic compound – something like iron in hemoglobin or magnesium in chlorophyll.  That’s the only form in which the FDA has approved it.  None of this is controversial.

The controversy starts with Gadolinium plaintiffs’ experts’ “free gadolinium” theory that these sneaky little atomic buggers can somehow escape their chemical prisons and run amok in the body, presumably causing the novel disease NSF.  See Slip op. at 5-6.  Sound familiar?  To us it does.  It sounds a lot like the junk science peddled by thimerosal plaintiffs that chemically bound and thus biologically inert lead is also somehow escape prone.  Just substitute “gadolinium” for “lead” and “NSF” for “autism,” and it would be very hard to tell them apart.

There’s another eerie similarity to thimerosal litigation – that even in people with serious kidney dysfunction, the vast majority do not come down with NSF:

Millions of people have received MRIs with [gadolinium contrast agents]. Not only have they had no complications, but they have avoided the risks associated with iodine, the magnetic contrast agent used prior to gadolinium. A small subset of patients who have been administered GBCAs have renal failure, and in recent years, a very small percentage of these renally-impaired patients have developed a new disease, NSF.
Slip op. at 15. This fact strongly reminds us of the “small subset” arguments with which vaccine/thimerosal plaintiffs have repeatedly attempted to justify untenable and speculative causation theories.  E.g., Mean v. Secretary HHS, 2010 WL 892248, at *28, 41, 112 (Fed. Cl. March 12, 2010); Dwyer v. Secretary HHS, 2010 WL 892250, at *105, 146 (Fed. Cl. March 12, 2010); King v. Secretary HHS, 2010 WL 892296, *68-70 (Fed. Cl. March 12, 2010).

Almost from the beginning, the Gadolinium opinion takes a rather unusual view of the Daubert inquiry – that “unknown” mechanisms of causation are okay because “establishing causation means providing scientific evidence from which an inference of cause and effect may be drawn.”  Slip op. at 8 (citing treatise). That’s because, we're told, “science is an evolving process, and there are no certainties in science.”  Id. at 7.

That's one way to look at it, we suppose, but it’s simply not the prevailing standard for addressing novel expert testimony under Daubert.  Rather, as Judge Posner eloquently put it, “the courtroom is not the place for scientific guesswork, even of the inspired sort. Law lags science; it does not lead it.”  Rosen v. Ciba-Geigy Corp., 78 F.3d 316, 319 (7th Cir. 1996); accord, e.g., McClain v. Metabolife International, Inc., 401 F.3d 1233, 1247 (11th Cir. 2005); Goebel v. Denver & Rio Grande Western Railroad Co., 346 F.3d 987, 1002 (10th Cir. 2003); In re Viagra Products Liability Litigation, 658 F. Supp.2d 936, 950 (D. Minn. 2009); Henricksen v. ConocoPhillips Co., 605 F. Supp.2d 1142, 1178 (E.D. Wash. 2009); Perry v. Novartis Pharmaceuticals Corp., 564 F. Supp.2d 452, 468 (E.D. Pa. 2008); Colon v. Abbott Laboratories, 397 F. Supp.2d 405, 416 (E.D.N.Y. 2005).  We thus count published case law from the second, third, seventh, eighth, ninth and tenth circuits – and that’s just one quick search of one nicely turned phrase.

Thus, from this angle, it looks like error to us.

Frankly, we’re not even sure why the opinion addressed that issue. We can’t tell exactly, but there may actually be epidemiological evidence supporting the causal link.  See Slip op. at 10 (discussing NSF emergence and decline).  But Gadolinium seems bound and determined to give plaintiffs everything – including the kind of stuff most courts consider scientific garbage:  animal studies, in vitro, in vivo, etc., slip op. at 10-11 –  that it doesn’t develop a coherent a coherent discussion of data (assuming it's there) that most courts would actually allow. As for the rest, see our various prior posts on some of the many cases whi8ch excluding that kind of stuff as unreliable.

For failing to separate the wheat from the chaff, we again see error.

One particularly egregious example of reliance on chaff, while ignoring the wheat, has to do with the use of drug experience reports (also called "AERs," "ADEs," “case reports” and various other things) that doctors voluntarily report to the FDA.  As we’ve discussed before, the FDA has repeatedly cautioned against misuse of its drug reporting statistics in a causation setting – from at least 1996 (“Accumulated ADE cases may not be used to calculate incidences or estimates of drug risk,” FDA Annual Adverse Experience Drug Report: 1996) until now (“AERS cannot be used to calculate the incidence of an adverse event in the U.S. population,” FDA, AERS Description). Way back in 2007, we collected 19 cases, each of which identified as bad science under Daubert any use adverse drug experience reports as a means of proving causation.

The Gadolinium opinion let AER-based causation opinions in anyway.  It didn’t really dispute any of the grounds on which practically every other case excludes them – their anecdotal nature, the explicit lack of any causation requirement for reporting, the susceptibility of reporting rates to press attention, litigation and other outside influences.  Instead, its reasoning was that, because the FDA did it, so could the plaintiffs’ experts:

The FDA [in 2009] was particularly interested in differential risk considerations among the various [gadolinium contrast agents] and any other considerations that should be addressed in labeling or other risk-reduction methods. In preparation for that meeting, the FDA asked its Office of Surveillance and Epidemiology (“OSE”) to review the evidence and reach a conclusion. The evidence reviewed by the OSE included AERs. . . .
Slip op. at 18.

We think that's error.

The Gadolinium decision completely overlooked the rather fundamental fact that the FDA is an administrative agency and is not bound by legal causation standards in making regulatory decisions about the products it regulates.  Literally dozens of courts have discussed in depth the fundamental difference between the administrative standards used by agencies and the more stringent legal causation standards required in litigation.  We recently dealt with this issue in great detail here.  For now, we’ll content ourselves with just three of the dozens of cases we discussed in that post:  The “FDA is a regulatory agency whose mandate is to control which drugs are marketed in the United States and how they are marketed.  FDA ordinarily does not attempt to prove that the drug in fact causes a particular adverse effect.”  Soldo v. Sandoz Pharmaceuticals Corp., 244 F. Supp.2d 434, 543 (W.D. Pa. 2003). “Although evidence of an association may . . . be important in the scientific and regulatory contexts . . ., tort law requires a higher standard of causation.”  Newton v. Roche Laboratories, Inc., 243 F. Supp.2d 672, 677 (W.D. Tex. 2002). Administrative agencies “impose[] different requirements and employ[] different labeling and evidentiary standards” because a “regulatory system reflects a more prophylactic approach” than the common law.  In re Seroquel Products Liability Litigation, 601 F. Supp.2d 1313, 1315 (M.D. Fla. 2009).  In short, that the FDA might choose to review AERs to make a regulatory determination about a product in no way allows a experts in a civil matter to ignore Daubert causation standards with the excuse that "the FDA did it, too."

Gadolinium goes on to apply its Kong Kingman strike zone to the plaintiffs’ reliance on this sort of anecdotal voluntary reporting.  Slip op. at 12-15.  The view seems to be that, because nobody knows the cause, anybody wearing an expert tag is allowed to say anything, at least when offered by the plaintiffs' experts (more on that last caveat later):

[N]o one knows exactly how NSF develops or why only a tiny percentage of renally-impaired patients who have been administered [gadolinium contrast agents] have developed NSF.  Under these circumstances, the Court is reluctant to exclude from the jury’s consideration any expert theory on NSF causation.  While the free gadolinium theory advanced by Plaintiffs’ experts is hardly established, neither is any other theory.
Slip op. at 15.  So when nothing's "established" anything the plaintiffs' expert says is allowed – because the true cause is unknown.  “AERs” as well as “chemistry studies,” “in vitro studies,” and “animal studies” are all okay when the umpire is blind.  “Cross-examination” is sufficient to bring out the weaknesses in the plaintiffs’ expert’s causation opinions.  Id. at 16.

That's error too.  That kind of reasoning stands Daubert quite precisely on its head.  Daubert, as those of us who once handled Bendectin litigation vividly recall, was where the plaintiffs similarly relied upon pieces of scientific refuse – “'in vitro’ (test tube) and ‘in vivo’ (live) animal studies . . .; pharmacological studies . . . that purported to show similarities between the structure of the drug and that of other substances . . .; and the ‘reanalysis’ of previously published epidemiological (human statistical) studies.”  Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 583 (1993).  In the book of Daubert it is written that an unknown cause does not openeth the flood gates so that plaintiffs might take scraps of inadmissible junk science, throw them together, and come up with something admissible:

The scientific project is advanced by broad and wide-ranging consideration of a multitude of hypotheses, for those that are incorrect will eventually be shown to be so, and that in itself is an advance.  Conjectures that are probably wrong are of little use, however, in the project of reaching a quick, final, and binding legal judgment-often of great consequence-about a particular set of events in the past.  We recognize that, in practice, a gatekeeping role for the judge, no matter how flexible, inevitably on occasion will prevent the jury from learning of authentic insights and innovations. That, nevertheless, is the balance that is struck by Rules of Evidence designed not for the exhaustive search for cosmic understanding but for the particularized resolution of legal disputes.
Daubert, 509:597 (emphasis added).  Thus, an unknown cause does not authorize a free-for-all.  The point was developed further in Daubert after remand:

[P]laintiffs rely entirely on the experts’ unadorned assertions that the methodology they employed comports with standard scientific procedures.  In support of these assertions, plaintiffs offer only the trial and deposition testimony of these experts in other cases.  While these materials indicate that plaintiffs’ experts have relied on animal studies, chemical structure analyses and epidemiological data, they neither explain the methodology the experts followed to reach their conclusions nor point to any external source to validate that methodology. . . .  Under Daubert, that's not enough.
Daubert v. Merrell Dow Pharmaceuticals, Inc., 43 F.3d 1311, 1319 (9th Cir. 1995) (emphasis added).

The analysis that Gadolinium employs to justify admission of pseudo-evidence without any inherent scientific validity is simply a throwback to pre-Daubert times when experts were allowed to throw whatever they wanted against the “wall” of cross-examination and see what might stick.  It wholly ignores the teaching of Daubert that alchemy is no longer scientific – piling up different types of scientific garbage only creates a bigger pile of garbage, and does not transmute inadmissible lead into admissible gold.


Having thus established that the scientific playing field shall have no boundaries, the opinion addresses individual expert witnesses.  We note particularly that it does so without the benefit of any Daubert hearings. Slip op. at 1.  With this cast of characters, we think that was error in and of itself.  The leadoff witness is the very well-traveled Dr. Suzanne Parisian.  Just last week we reported on how the judge in the Trasylol MDL – after getting a snootful of Dr. Parisian’s testimony in a live Daubert hearing, decided that “expert” testimony Parisian style wouldn’t be “helpful” to a jury:

Plainly stated, Dr. Parisian is an advocate, presented with the trappings of an expert but with no expectation or intention of abiding the opinion constraints of Rule 702. She comes armed with a Report designed to be broad enough to allow her to gather and stack inference upon inference in order to offer her "takeaway" or "take home message" with respect to intent, knowledge, or causation in a manner unrelated to any regulatory expertise. Her testimony is unreliable and would not be of assistance to the jury.
In re Trasylol Products Liability Litigation, ___ F. Supp.2d ___, 2010 WL 1737107, at *20 (S.D. Fla. April 27, 2010).

Not so in Gadolinium, where total lack of familiarity with how experts actually perform on the stand has apparently bred an equally total suspension of contempt (at least as to the other side of the "v.").  But it’s not like there was no warning.  The opinion discusses similar conclusions reached by the judge handling the Fosamax MDL before it, and essentially pooh-poohed them.

As in Fosamax, Dr. Parisian may not provide a narrative history of [the product], which must be presented through direct evidence.  Nor may she testify as to the knowledge, motivations, intent or purposes of [the defendant], its employees, and the FDA.  Based on the analysis of this particular issue with regard to Dr. Parisian alone, the Court concludes that Plaintiffs’ experts may offer opinion testimony based on . . . internal documents, studies and regulatory filings to the extent it is relevant to the issues in this case and the experts are qualified to offer opinions on those issues.
Slip op. at 25.  Good luck – we strongly suspect that it would have been a much better idea to have a Daubert hearing before reaching this sort of conclusion about a witness that other courts have said so much about.

Actually, Gadolinium did a little worse than perhaps overestimating the ability to constrain the notoriously hard-to-control Dr. Parisian.  It also stated that whether the defendant “provided adequate information regarding the risks of [the drug] to the FDA during the approval process” was somehow “relevant.”  Slip op. at 23-24.

More error.

Fraud on the FDA claims have been preempted for close to a decade now.  We turn to the book of Buckman for our next reading, at

State-law fraud-on-the-FDA claims inevitably conflict with the FDA’s responsibility to police fraud consistently with the Administration’s judgment and objectives. As a practical matter, complying with the FDA’s detailed regulatory regime in the shadow of 50 States’ tort regimes will dramatically increase the burdens facing potential applicants-burdens not contemplated by Congress in enacting the FDCA and the MDA. Would-be applicants may be discouraged from seeking §510(k) approval of devices with potentially beneficial off-label uses for fear that such use might expose the manufacturer or its associates (such as petitioner) to unpredictable civil liability. . . .

Conversely, fraud-on-the-FDA claims would also cause applicants to fear that their disclosures to the FDA, although deemed appropriate by the Administration, will later be judged insufficient in state court. Applicants would then have an incentive to submit a deluge of information that the Administration neither wants nor needs, resulting in additional burdens on the FDA's evaluation of an application.
Buckman Co. v. Plaintiffs’ Legal Committee, 531:350-51 (2001).  In other words, whether a regulated manufacturer provided the FDA with “adequate” information isn’t anybody’s business but the FDA’s.  The completeness of regulatory submissions are not to be second-guessed in state-law actions.


We have our reservations about so-called “FDA experts” in any event, but that’s because we've seen a lot of precedent holding that experts are not supposed to offer opinions on questions of law, such as how the FDA’s regulations apply to particular cases and whether or not those regulations were complied with.  As we discussed here, and here, among other places, instructing the jury on the law is job of the judge, not hired litigation experts.

The next witness, Laura Plunket, Ph.D. was prepared to testify that gadolinium contrast agents “should have been contraindicated in patients with significant renal impairment as early as 1996,” slip op. at 27 – a neat trick, since the first reports of NSF did not happen until 1997.  Id. at 6.

That physical impossibility, and Dr. Plunket's questionable qualifications – she’s not a doctor, and thus not legally allowed to prescribe any drug, slip op. at 25 – didn’t cause very much pause, with the Kong Kingman strike zone in effect.  Dr. Plunket was allowed to use the same scientific garbage as in the Bendectin litigation (“in vitro assessments of stability, in vivo data on biodistribution in animals, pharmacokinetic studies”) as well as “regulatory filings” to attack the “accuracy and adequacy of the toxicology, pharmacology and pharmacokinetics data appearing on the [product] label.”  Slip op. at 25-26.  In plain English, this witness is being allowed to use exact evidence that Daubert prohibited to give the sort of opinions that are preempted under Buckman.

Obviously, we think that's error.

Next up is Cheryl Blume, Ph.D. – another non-physician.  Slip op. at 27-30.

Parisian and Plunkett and Blume, oh my!  These three have kept many legal fires burning and litigation cauldrons bubbling over the years.  Without knowing anything else about Gadolinium, we’d be suspicious of the scientific basis of this litigation just from the cast of characters.

But in the Gadolinium MDL, ‘tis time, ‘tis time!

The opinion allows Dr. Blume to testify about all of the plaintiffs’ causation theories and all of the defendant’s scientific studies.  Id. at 28-29.  The earlier error of allowing admission of AERs into evidence is compounded, as Dr. Blume also gets the green light to interpret a 2005 FDA guidance (which may not have existed during the FDA interactions that supposedly established her expertise) as providing that mere four AERs – out of literally millions of uses – can constitute a “safety signal.”  Slip op. at 29-30.  Heck, why require four; “one or more” might be enough.  Id. at 30.

To top it all off, the opinion refuses even to preclude Dr. Blume from testifying “about foreign regulatory events and foreign law.”  Id. at 30.  Instead, “foreign regulatory events” “may be admissible and relied upon by Dr. Blume in forming her opinions.”  Id.

No law was cited in support of that singular proposition.

That's an entirely new dimension of error.

Specifically, foreign regulatory actions have never been admissible in American product liability cases.  As the Sixth Circuit held, in ruling European regulations irrelevant in a prescription drug mass tort:

Plaintiffs also allege that the warning label for [the drug’s] European equivalent contains more detailed instructions for the treating physician.  Citing no authority, Plaintiffs argue that the difference in instructions creates a triable issue of fact.  We disagree.  American regulators have different priorities and deal with often more diverse populations than their European counterparts.  The issue is whether the United States label . . . provides adequate instructions upon which a physician may safely base her treatment strategy.  Plaintiffs have failed to make a showing of inadequacy such that a reasonable jury could find for the nonmoving party.
Meridia Products Liability Litigation v. Abbott Laboratories, 447 F.3d 861, 867 (6th Cir. 2006).  Back in 2007, we collected a bunch of precedent excluding foreign regulatory evidence here, and the consensus for exclusion has only grown since.  See, Trasylol, 2010 WL 1737107, at *7; Seroquel, 601 F. Supp.2d at 1318.

As to the next witness, Dr. Fine, the primary question was whether, as a nephrologist, he was competent to offer his wide-ranging opinions.  See Slip op. at 32.  Kong Kingman was still at the plate, so Dr. Fine got “wide latitude.”  Id. at 34.  “Dr. Fine does not have to have to be a toxicologist, chemist, radiologist and/or have personally conducted animal studies . . . in order to offer an opinion on the cause of NSF, the instability of [the product], or the free gadolinium theory.”  Id.  He could even opine on the adequacy of product warnings.  Id. at 35 (“may offer opinions on whether [the] labeling information or Dear Doctor letters contained adequate information”).

The next witness, a rheumatologist, had published a study on NSF.  The defendant challenged study because the gold standard diagnostic test for NSF, a skin biopsy, was not done in over 80% of the study subjects.  Slip op. at 37.  That objection was unavailing because “performing a biopsy would subject them to unwarranted health risks.”  Id. at 38.

More error.  This is a skin biopsy, not a lung biopsy or something else invasive.  Some skin biopsies don’t even require stitches.  For the diagnosis of a potentially fatal disease, we hardly think that removal of a “small piece” of skin presents an “unwarranted” health risk.

But in Gadolinium, plaintiffs’ experts can do no wrong.

A case in point:  Dr. Abraham, the plaintiffs’ pathologist.  Should he be limited to opinions on pathology? Well, not if one interprets his specialty “broadly.”  A “narrow” reading “does not account for what appears to be a rather broad field of study and expertise, and it is certainly too narrow to describe the work of research pathologists.”  Slip op. at 40.

Plaintiffs also call a statistician, Dr. Ix (it's sort of apt, we guess, for a statistician to have a Roman numeral for a last name).  He proposed to offer causation testimony “[b]ased on a meta-analysis of five studies selected from an original group of thirty-seven.”  Slip op. at 41.  He claimed that that gadolinium-caused NSF occurs “among patients with moderate to severe kidney disease,” id. – which, of course, contradicts the FDA’s position that only "severe" renal impairment leads to NSF.
8.  The information FDA released in December 2006 said that patients with moderate renal insufficiency are at risk for developing NSF. Why has this changed?

The December information was based upon reports of NSF among patients with purportedly moderate renal insufficiency. Since issuing the information in December 2006, FDA has received new information regarding these patients. Additional details have clarified that the patients actually were in acute renal failure at the time they received a GBCA. Considering this clarification, FDA has not received reports of NSF among patients with normal renal function or moderate renal insufficiency.
Emphasis added.  Dr. Ix’s opinion was generated “solely” for litigation.  Slip op. at 41.

But it didn’t seem to matter that Dr. Ix deviated from his usual research practices to come up with a litigation opinion. “[W]hether Dr. Ix followed procedures he uses in his own independent research is irrelevant.”  Slip op. at 42.

That's big time error.

Let us open the Holy Writ to the Book of Kumho, chapter 526, verse 152, wherein it is written. “The objective of that [gatekeeping] requirement is to . . . make certain that an expert, whether basing testimony upon professional studies or personal experience, employs in the courtroom the same level of intellectual rigor that characterizes the practice of an expert in the relevant field.”  Kumho Tire Co. v. Carmichael, 526 U.S. 137, 152 (1999) (emphasis added).

So yes, we do see error, because under the express terms of Kumho Tire, Dr. Ix’s conceded failure to employ in his litigation-generated opinion, the same “procedures” used in his “independent research” is not only relevant but probably controlling – especially given the “greater rigor” that the opinions of “a quintessential expert for hire” are supposed to undergo.  Johnson v. Manitowoc Boom Trucks, Inc., 484 F.3d 426, 435 (6th Cir. 2007).

That sure didn’t happen here. Why didn’t Dr. Ix include the other 32 of the 37 studies?  Dunno.  The issue of selection bias didn’t seem to register in Gadolinium.  Nor did anything else.  All we know is:  (1) he cherry-picked the data (“failure to consider studies not reporting an association between [gadolinium contrast agents] and NSF also does not render Dr. Ix’s meta-analysis unreliable,” slip op. at 45) and he didn’t consult anyone else when he did cherrypick the data.  Id. at 42-43.  That didn’t matter either.  Id.  Oddly (or perhaps not), the five chosen studies had unusually “wide confidence intervals.”  Id. at 43-44. Fuggedaboudit.  That’s just for cross-examination.  Id. at 44.  With a new disease, anything goes. “Given how recently NSF was identified as a disease, the causation opinion of any expert is still just a theory.”  Id.

Sitll more error.   See our prior “litigation lags science” discussion.

Finally, plaintiffs offered a radiologist, Dr. Semelka, who proposed to testify that gadolinium contrast agents caused NSF, even though his peer-reviewed publications on the topic stated that “[t]he exact mechanism for the development of NSF remains uncertain.”  Slip op. at 47.  Once again, “uncertain” causation was no cause for concern:

[A]ll parties agree . . . that the exact mechanism for NSF causation is uncertain. A corollary of [defendant’s] argument would therefore be that nobody could opine about what causes NSF, and that [defendant] should be granted summary judgment in this MDL. This is not going to happen.
Slip op. at 48.

Umm ... We’ve figured that out. The plaintiffs could put up a voodoo doctor and, since no theory’s any more “established” than any other theory, voodoo would be allowed to testify.  Slip op. at 15.


Now we turn to the defense experts.

Time out!

Kong’s being lifted for a pinch hitter.  Now batting in the one-eighth position is Eddie Gaedel.

The defense sought to offer evidence, through a Dr. Newton, that NSF occurs in the absence of any exposure to gadolinium. The published literature apparently includes two case studies involving three patients, stating that this occurred.

They’re “fundamentally flawed,” holds GadoliniumSlip op. at 51.

But what about the holding, id. at 29-30 hold that four – and maybe even one – anecdotal studies were enough to constitute a “signal”?

Doesn’t matter.

What about any study being admissible where everything’s uncertain?

You're confusing Eddie Gaedel with Kong Kingman.

The studies are “fundamentally flawed” – for one thing they “did not confirm its findings by testing these patients’ tissue.”  Slip op. at 51. 

But … but … the holding that “a biopsy is not the only means of diagnosing NSF.”  Slip op. at 38 …?  The finding that skin biopsies posed "unwarranted" health risks?  Id.

Once again, you're confusing Eddie Gaedel with Kong Kingman.

All the error has our heads are spinning.  An actual live Daubert hearing is looking really good right about now.

Dr. Newton was also prohibited from testifying about an animal study – not because animal studies are inherently unreliable (a position we would agree with) – but because the witness was only a pharmacologist and cell biologist, not a dermatologist or a pathologist.  Slip op. at 51-52.

After holding that Dr. Fine, a nephrologist, could testify as far afield as to the adequacy of FDA-approved warnings, slip op. at 35, we would have thoughtthat cell biology would get Dr. Newton at least as far as dermatology or pathology – both of which involve examining cells.

We would have thought wrong. The Eddie Gaedel strike zone is a whole lot different than the Kong Kingman strike zone.  Slip op. at 51.

Needless to say – looks like more error to us.

Equally restrictive is the Gaedel strike zone for defense witness Dr. Watson – forbidden “to testify on matters outside the field of bioinorganic chemistry.”  Slip op. at 52.

Dr. Watson may not provide an opinion on the FDA’s review of Omniscan, his assessment of [the product] or [gadolinium contrast agent] toxicology or toxicology studies, any preclinical or clinical studies of [the product] . . ., analysis of relevant publications on [gadolinium contrast agents], or foreseeability of gadolinium poisoning.

In the abstract, we have nothing against restrictive interpretations of the scope of a witness’ expertise.  But after the “broad” rulings concerning Drs. Plunkett (slip op. at 26-27), Blume (id. at 28-29), Fine (id. at 33-34), Abraham (id. at 39-40), and Semelka (id. at 47), this turnabout is inexplicable except as a matter of error.

The Gadolinium beat goes on with respect to defense witness Dr. Gasparini, who is not allowed to offer a contrary opinion on what we think are (it’s not 100%) clear the same “four Adverse Event Reports” that Dr. Blume was allowed to call a “signal.”  Slip op. at 53.  Cf. id. at 28-30 (discussion of Dr. Blume).  Why?  Because he did not purport also to opine on other data such as ‘animal studies” or “medical literature” – his opinion was therefore based upon “based on incomplete information.”  Id. at 54.

That's peculiar.  Dr. Blume can testify that four AERs are, by themselves, sufficient to be a “signal” under FDA regulations, but no defense witness can rebut that testimony with why the same four reports should not be considered notice.

And it’s perfectly all right for the plaintiffs’ statistician to opine based upon a meta analysis that ignores 32 of 37 studies – where the five just happen to be those “not reporting an association” between the product and the disease.  Slip op. at 44.

The error is getting very spherical, now.

The next defense witness gets trimmed, too.  Dr. Waikar won’t be allowed to testify about the two articles reporting that NSF occurs in the absence of gadolinium exposure (see Dr. Newton) or about the four AERs (see Dr. Gasparini). That was a error before, so it’s still error.

And finally the Gadolinium pièce de résistance – the defense FDA expert won’t be allowed to testify at all.  And to think - no Daubert hearing is "appropriate" prior to that sort of ruling.

What happened?  Because his testimony deviates from the Holy Writ of the Book of Levinicus.  That’s right.  During his deposition the witness, Dr. Paul Waymack, a former “full-time medical officer at the FDA,” slip op. at 55, was asked whether he agreed with certain quotes excerpted from Wyeth v. Levine, 129 S. Ct. 1187 (2009).

An aside:  wasn't that sort of questioning objectionable?  We'd be jumping up and down - or better yet, have prepared our expert for this.

Needless to say, as a defense witness, he didn’t.  As the Gadolinium opinion put it, he “demonstrate[d] a . . . blatant disregard of the Supreme Court’s pronouncements in Wyeth.”  Slip op. at 57.  For that offense against Holy Writ, Dr. Waymack – without any live Daubert hearing – was excommunicated from the litigation.  Id. at 57.

That’s right. In an admittedly unprecedented ruling (“the Court was unable to find, nor did either party cite a Sixth Circuit case on this issue,” slip op. at 57) an expert witness was excluded solely because of the witness’ substantive legal opinions.  Gadolinium is fundamentalist territory.  The Book of Levinicus (if not the books of Buckman, Daubert, or Kuhmo) has been declared to be the literal Truth:

[C]ommon sense dictates that the Court prohibit a witness from offering opinions in direct conflict with Supreme Court holdings or observations made by the Supreme Court that serve as a foundation for reaching a conclusion (i.e. that the FDA’s limited resources mean the manufacturer has superior access to information about its own product).
Slip op. at 57.

We could abide a result – heck, we prefer a result – to the effect that no expert witnesses on either side should be allowed to offer any legal opinions.  But to elevate Levine to the status of litigation Holy Writ, as Gadolinium has done is serious error.

Levine (like any case) was a product of its facts – very disputed facts, if the majority and dissent are compared – and those facts don’t necessarily exist in other litigations.

To take the example in Gadolinium, Levine holds that “manufacturers have superior access to information about their drugs, especially in the postmarketing phase as new risks emerge.”  129 S. Ct. at 1202.  We can see why the Court may have thought so in Levine itself, since that case involved a single drug.  But the Gadolinium MDL appears to involve an entire class of drugs.  Where more than one drug is at issue, then the FDA has the power to order every manufacturer to turn over its post-marketing studies, adverse events, etc., if the FDA so chooses.

That's a governmental power beyond the ability of any private manufacturer.  Our clients simply don’t have access to this sort of information about a competitor's drugs. So if there’s a situation where the FDA orders a class-wide review (and we don’t know enough about the regulatory history of these gadolinium products to say whether or not that happened), there could very well be a situation where the FDA has superior knowledge about this or that drug because it in fact obtained access to more information than any single manufacturer could hope to have.

Courts, even the Supreme Court, draft their opinions based upon the facts of the case before them. That’s what the common law is all about. Thus we think it’s error for Gadolinium to demand literal adherence to Levine insofar as any and all expert opinions are concerned.

And that might be the biggest error of all.


Anonymous said...

Jim, you may not be aware of even less scientific testimony based upon differential etiology, on a closely-related subject.

Bruce C. Hamlin
Martin Bischoff, et al
888 SW Fifth Avenue, Suite 900
Portland, OR 97204
503-224-3113 (phone)
503-224-9471 (fax)
503-382-4204 (direct dial)
503-501-7467 (cell)

mail to:

Bexis said...

Jeez - that's as bad as New Jersey.

Ted Heise said...

Ugh. Allowance of any potential harms as evidence against the drug seems an incredibly excessive application of the Precautionary Principle.

One small clarification: chelation does not involve a pure chemical bond, it's ionic in nature, albeit multi-coordinate. As such, release of the chelated substance is generally easier than cleaving a chemical bond. That said, this particular binding is probably adequate in vivo, or it's doubtful FDA would have approved it.

Anonymous said...

As Ted Heise points out, chelation does not typically involve a covalent (~"permanent") bond. Attachment between gadolinium and its "chelation agent" is likely in equibrium (rapidly shifting between bound and unbound states). It's likely that the "bound state" dominates, but that doesn't mean that there is no "unbound state".

Where Ted Heise is wrong, however, is in stating "this particular binding is probably adequate in vivo, or it's doubtful FDA would have approved it."

I doubt that anyone knows the true ratio (and kinetics) of bound and unbound states of gadolinium in vivo. There are many examples of FDA approved drugs that appear to be bound/chelated under most in vivo tests. Under these circumstances, one would predict that the compounds would not be efficacious in vivo. Yet they are efficacious in vivo, implying that the small degree of "unbound" drug is sufficient for efficacy.

If a small fraction of unbound drug can be sufficient for efficacy in vivo, then it's reasonable to think that a small amount of unbound gadolinium could be harmful in vivo.