The limits of peer review ultimately make it a poor proxy for the validity tests posed by Rules 702 and 703. Published peer review articles simply do not permit a very searching evaluation of the facts and data of a study. In the wake of the Daubert decision, expert witnesses quickly saw that they can obscure the search for validity by the reliance upon published studies, and frustrate the goals of judicial gatekeeping. As a practical matter, the burden shifts to the party that wishes to challenge the relied upon facts and data to learn more about the cited studies to show that the facts and data are not sufficient under Rule 702(b), and that the testimony is not the product of reliable methods under Rule 702(c). Obtaining study protocols, and in some instances, underlying data, are necessary for due process in the gatekeeping process. A couple of case studies may illustrate the power of looking under the hood of published studies, even ones that were peer reviewed.
When the Supreme Court decided the Daubert case in June 1993, two recent verdicts in silicone-gel breast implant cases were fresh in memory.[1] The verdicts were large by the standards of the time, and the evidence presented for the claims that silicone caused autoimmune disease was extremely weak. The verdicts set off a feeding frenzy, not only in the lawsuit industry, but also in the shady entrepreneurial world of supposed medical tests for “silicone sensitivity.”
The plaintiffs’ litigation theory lacked any meaningful epidemiologic support, and so there were fulsome presentations of putative, hypothetical mechanisms. One such mechanism involved the supposed in vivo degradation of silicone to silica (silicon dioxide), with silica then inducing an immunogenic reaction, which then, somehow, induced autoimmunity and the induction of autoimmune connective tissue disease. The degradation claim would ultimately prove baseless,[2] and the nuclear magnetic resonance evidence put forward to support degradation would turn out to be instrumental artifact and deception. The immunogenic mechanism had a few lines of potential support, with the most prominent at the time coming from the laboratories of Douglas Radford Shanklin, and his colleague, David L. Smalley, both of whom were testifying expert witnesses for claimants.
The Daubert decision held out some opportunity to challenge the admissibility of testimony that silicone implants led to either the production of a silicone-specific antibody, or the induction of t-cell mediated immunogenicity from silicone (or resulting silica) exposure. The initial tests of the newly articulated standard for admissibility of opinion testimony in silicone litigation did not go well.[3] Peer review, which was absent in the re-analyses relied upon in the Bendectin litigation, was superficially present in the studies relied upon in the silicone litigation. The absence of supportive epidemiology was excused with hand waving that there was a “credible” mechanism, and that epidemiology took too long and was too expensive. Initially, post-Daubert, federal courts were quick to excuse the absence of epidemiology for a novel claim.
The initial Rule 702 challenges to plaintiffs’ expert witnesses thus focused on immunogenicity as the putative mechanism, which if true, might lend some plausibility to their causal claim. Ultimately, plaintiffs’ expert witnesses would have to show that the mechanism was real by showing that silicone exposure causes autoimmune disease through epidemiologic studies,
One of the more persistent purveyors of a “test” for detecting alleged silicone sensitivity came from Smalley and Shanklin, then at the University of Tennessee. These authors exploited the fears of implant recipients and the greed of lawyers by marketing a “silicone sensitivity test (SILS).” For a price, Smalley and Shanklin would test mailed-in blood specimens sent directly by lawyers or by physicians, and provide ready-for-litigation reports that claimants had suffered an immune system response to silicone exposure. Starting in 1995, Smalley and Shanklin also cranked out a series of articles at supposedly peer reviewed journals, which purported to identify a specific immune response to crystalline silica in women who had silicone gel breast implants.[4] These studies had two obvious goals. First, the studies promoted their product to the “silicone sisters,” various support groups of claimants, as well as their lawyers, and a network of supporting rheumatologists and plastic surgeons. Second, by identifying a putative causal mechanism, Shanklin could add a meretricious patina of scientific validity to the claim that silicone breast implants cause autoimmune disease, which Shanklin, as a testifying expert witness, needed to survive Rule 702 challenges.
The plaintiffs’ strategy had been to paper over the huge analytical gaps in their causal theory with complicated, speculative research, which had been peer reviewed and published. Although the quality of the journals was often suspect, and the nature of the peer review obscure, the strategy had been initially successful in deflecting any meaningful scrutiny.
Many of the silicone cases were pending in a multi-district litigation, MDL 926, before Judge Sam Pointer, in the Northern District of Alabama. Judge Pointer, however, did not believe that ruling on expert witness admissibility was a function of an MDL court, and by 1995, he started to remand cases to the transferor courts, for those courts to do what they thought appropriate under Rules 702 and 703. Some of the first remanded cases went to the District of Oregon, where they landed in front of Judge Robert E. Jones. In early 1996, Judge Jones invited briefing on expert witness challenges, and in face of the complex immunology and toxicology issues, and the emerging epidemiologic studies, he decided to appoint four technical advisors to assist him in deciding the challenges.
The addition of scientific advisors to the gatekeeper’s bench made a huge difference in the sophistication and detail of the challenges that could be lodged to the relied-upon studies. In June 1996, Judge Jones entertained extensive hearings with viva voce testimony from both challenged witnesses and subject-matter experts on topics, such as immunology and nuclear magnetic resonance spectroscopy. Judge Jones invited final argument in the form of videotaped presentations from counsel so that the videotapes could be distributed to his technical advisors later in the summer. The contrived complexity of plaintiffs’ case dissipated, and the huge analytical gaps became visible. In December 1996, Judge Jones issued his decision that excluded the plaintiffs’ expert witnesses’ proposed testimony on grounds that it failed to satisfy the requirements of Rule 702.[5]
In October 1996, while Judge Jones was studying the record, and writing his opinion in the Hall case, Judge Weinstein, with a judge from the Southern District of New York, and another from New York state trial court, conducted a two-week Rule 702 hearing, in Brooklyn. Judge Weinstein announced at the outset that he had studied the record from the Hall case, and that he would incorporate it into his record for the cases remanded to the Southern and Eastern Districts of New York.
Curious gaps in the articles claiming silicone immunogenicity, and the lack of success in earlier Rule 702 challenges, motivated the defense to obtain the study protocols and underlying data from studies such as those published by Shanklin and Smalley. Shanklin and Smalley were frequently listed as expert witnesses in individual cases, but when requests or subpoenas for their protocols and raw data were filed, plaintiffs’ counsel stonewalled or withdrew them as witnesses. Eventually, the defense was able to enforce a subpoena and obtain the protocol and some data. The respondents claimed that the control data no longer existed, and inexplicably a good part of the experimental data had been destroyed. Enough was revealed, however, to see that the published articles were not what they claimed to be.[6]
In addition to litigation discovery, in March 1996, a surgeon published the results of his test of the Shanklin-Smalley silicone sensitivity test (“SILS”).[7] Dr. Leroy Young sent the Shanklin laboratory several blood samples from women with and without silicone implants. For six women who never had implants, Dr. Young submitted a fabricated medical history that included silicone implants and symptoms of “silicone-associated disease.” All six samples were reported back as “positive”; indeed, these results were more positive than the blood samples from the women who actually had silicone implants. Dr. Young suggested that perhaps the SILS test was akin to cold fusion.
By the time counsel assembled in Judge Weinstein’s courtroom, in October 1996, some epidemiologic studies had become available and much more information was available on the supposedly supportive mechanistic studies upon which plaintiffs’ expert witnesses had previously relied. Not too surprisingly, plaintiffs’ counsel chose not to call the entrepreneurial Dr. Shanklin, but instead called Donard S. Dwyer, a young, earnest immunologist who had done some contract work on an unrelated matter for Bristol-Myers Squibb, a defendant in the litigation. Dr. Dwyer had filed an affidavit previously in the Oregon federal litigation, in which he gave blanket approval to the methods and conclusions of the Smalley-Shanklin research:
“Based on a thorough review of these extensive materials which are more than adequate to evaluate Dr. Smalley’s test methodology, I formed the following conclusions. First, the experimental protocols that were used are standard and acceptable methods for measuring T Cell proliferation. The results have been reproducible and consistent in this laboratory. Second, the conclusion that there are differences between patients with breast implants and normal controls with respect to the proliferative response to silicon dioxide appears to be justified from the data.”[8]
Dwyer maintained this position even after the defense obtaining the study protocol and underlying data, and various immunologists on the defense side filed scathing evaluatons of the Smalley-Shanklin work. On direct examination at the hearings in Brooklyn, Dwyer vouched for the challenged t-cell studies, and opined that the work was peer reviewed and sufficiently reliable.[9]
The charade fell apart on cross-examination. Dwyer refused to endorse the studies that claimed to have found an anti-silicone antibody. Researchers at leading universities had attempted to reproduce the findings of such antibodies, without success.[10] The real controversy was over the claimed finding of silicone antigenicity as shown in t-cell or the cell-mediated specific immune response. On direct examination, plaintiffs’ counsel elicited Dwyer’s support for the soundness of the scientific studies that purported to establish such antigenicity, with little attention to the critiques that had been filed before the hearing.[11] Dwyer stuck to his unqualified support he had expressed previously in his affidavit for the Oregon cases.[12]
The problematic aspect of Dwyer’s direct examination testimony was that he had seen the protocol and the partial data produced by Smalley and Shanklin.[13] Dwyer, therefore, could not resist some basic facts about their work. First, the Shanklin data failed to support a dose-response relationship.[14] Second, the blood samples from women with silicone implants had been mailed to Smalley’s laboratory, whereas the control samples were collected locally. The disparity ensured that the silicone blood samples would be older than the controls, which was a departure from treating exposed and control samples in the same way.[15] Third, the experiment was done unblinded; the laboratory technical personnel and the investigators knew which blood samples were silicone exposed and which were controls (except for samples sent by Dr. Leroy Young).[16] Fourth, Shanklin’s laboratory procedures deviated from the standardized procedure set out in the National Institute of Health’s Current Protocols in Immunology.[17]
The SILS study protocol and the data produced by Shanklin and Smalley made clear that each sample was to be tested in triplicate for t-cell proliferation in response to silica, to a positive control mitogen (Con A), and to a negative control blank. The published papers all claimed that the each sample was tested in triplicate for each of these three response situations (silica, mitogen, and nothing).[18] Shanklin and Smalley described their t-cell proliferation studies, in their published papers, as having been done in triplicate. These statements were, however, untrue and never corrected.[19]
The study protocol called for the tests to be run in triplicate, but they instructed the laboratory that two counts may be used if one count does not match the other counts, which is to be decided by a technical specialist on a “case-by-case” basis. Of data that was supposed to be reported in triplicate, fully one third had only two data points, and 10 percent had but one data point.[20] No criteria were provided to the technical specialist for deciding which data to discard.[21] Not only had Shanklin excluded data, but he discarded and destroyed the data such that no one could go back and assess whether the data should have been excluded.[22]
Dwyer agreed that this exclusion and discarding of data was not at all a good method.[23] Dwyer proclaimed that he had not come to Brooklyn to defend this aspect of the Shanklin work, and that it was not defensible at all. Dwyer conceded that “the interpretation of the data and collection of the data are flawed.”[24] Dwyer tried to stake out a position that was incoherent by asserting that there was “nothing inherently wrong with the method,” while conceding that discarding data was problematic.[25] The judges presiding over the hearing could readily see that the Shanklin research was bent.
At this point, the lead plaintiffs’ counsel, Michael Williams, sought an off-ramp. He jumped to his feet and exclaimed “I’m informed that no witness in this case will rely on Dr. Smalley’s [and Shanklin’s] work in any respect.” [26] Judge Weinstein’s eyes lit up with the prospect that the Smalley-Shanklin work, by agreement, would never be mentioned again in New York state or federal cases. Given how central the claim of silicone antigenicity was to plaintiffs’ cases, the defense resisted the stipulation about research that they would continue to face in other state and federal courts. The defense was saved, however, by the obstinence of a lawyer from the Weitz & Luxenberg firm, who rose to report that her firm intended to call Drs. Shanklin and Smalley as witnesses, and that they would not stipulate to the exclusion of their work. Judge Weinstein rolled his eyes, and waved me to continue.[27] The proliferation of the t-cell test was over. The hearing before Judges Weinstein and Baer, and Justice Lobis, continued for several more days, with several other dramatic moments.[28]
In short order, on October 23, 1996, Judge Weinstein issued a short, published opinion, in which he granted partial summary judgment on the claims of systemic disease for all cases pending in federal court in New York.[29] What was curious was that the defendants had not moved for summary judgment. There were, of course, pending motions to exclude plaintiffs’ expert witnesses, but Judge Weinstein effectively ducked those motions, and let it be known that he was never a fan of Rule 702. It would be many years later, before Judge Weinstein allowed his judicial assessment see the light of day. Two decades and some years later, in a law review article, Judge Weinstein gave his judgment that
“[t]he breast implant litigation was largely based on a litigation fraud. … Claims—supported by medical charlatans—that enormous damages to women’s systems resulted could not be supported.”[30]
Judge Weinstein’s opinion was truly a judgment from which there can be no appeal. Shanklin and Smalley continued to publish papers for another decade. None of the published articles by Shanklin and others have been retracted.
[1] Reuters, “Record $25 Million Awarded In Silicone-Gel Implants Case,” N.Y. Times at A13 (Dec. 24, 1992) (describing the verdict returned in Harris County, Texas, in Johnson v. Medical Engineering Corp.); Associated Press, “Woman Wins Implant Suit,” N.Y. Times at A16 (Dec. 17, 1991) (reporting a verdict in Hopkins v. Dow Corning, for $840,000 in compensatory and $6.5 million in punitive damages); see Hopkins v. Dow Corning Corp., 33 F.3d 1116 (9th Cir. 1994) (affirming judgment with minimal attention to Rule 702 issues).
[2] William E. Hull, “A Critical Review of MR Studies Concerning Silicone Breast Implants,” 42 Magnetic Resonance in Medicine 984, 984 (1999) (“From my viewpoint as an analytical spectroscopist, the result of this exercise was disturbing and disappointing. In my judgement as a referee, none of the Garrido group’s papers (1–6) should have been published in their current form.”). See also N.A. Schachtman, “Silicone Data – Slippery & Hard to Find, Part 2,” Tortini (July 5, 2015). Many of the material science claims in the breast implant litigation were as fraudulent as the health effects claims. See, e.g., John Donley, “Examining the Expert,” 49 Litigation 26 (Spring 2023) (discussing his encounters with frequent testifier Pierre Blais, in silicone litigation).
[3] See, e.g., Hopkins v. Dow Corning Corp., 33 F.3d 1116 (9th Cir. 1994) (affirming judgment for plaintiff over Rule 702 challenges), cert. denied, 115 S.Ct. 734 (1995). See Donald A. Lawson, “Note, Hopkins v. Dow Corning Corporation: Silicone and Science,” 37 Jurimetrics J. 53 (1996) (concluding that Hopkins was wrongly decided).
[4] See David L. Smalley, Douglas R. Shanklin, Mary F. Hall, and Michael V. Stevens, “Detection of Lymphocyte Stimulation by Silicon Dioxide,” 4 Internat’l J. Occup. Med. & Toxicol. 63 (1995); David L. Smalley, Douglas R. Shanklin, Mary F. Hall, Michael V. Stevens, and Aram Hanissian, “Immunologic stimulation of T lymphocytes by silica after use of silicone mammary implants,” 9 FASEB J. 424 (1995); David L. Smalley, J. J. Levine, Douglas R. Shanklin, Mary F. Hall, Michael V. Stevens, “Lymphocyte response to silica among offspring of silicone breast implant recipients,” 196 Immunobiology 567 (1996); David L. Smalley, Douglas R. Shanklin, “T-cell-specific response to silicone gel,” 98 Plastic Reconstr. Surg. 915 (1996); and Douglas R. Shanklin, David L. Smalley, Mary F. Hall, Michael V. Stevens, “T cell-mediated immune response to silica in silicone breast implant patients,” 210 Curr. Topics Microbiol. Immunol. 227 (1996). Shanklin was also no stranger to making his case in the popular media. See, e.g., Douglas Shanklin, “More Research Needed on Breast Implants,” Kitsap Sun at 2 (Aug. 29, 1995) (“Widespread silicone sickness is very real in women with past and continuing exposure to silicone breast implants.”) (writing for Scripps Howard News Service). Even after the Shanklin studies were discredited in court, Shanklin and his colleagues continued to publish their claims that silicone implants led to silica antigenicity. David L. Smalley, Douglas R. Shanklin, and Mary F. Hall, “Monocyte-dependent stimulation of human T cells by silicon dioxide,” 66 Pathobiology 302 (1998); Douglas R. Shanklin and David L. Smalley, “The immunopathology of siliconosis. History, clinical presentation, and relation to silicosis and the chemistry of silicon and silicone,” 18 Immunol. Res. 125 (1998); Douglas Radford Shanklin, David L. Smalley, “Pathogenetic and diagnostic aspects of siliconosis,” 17 Rev. Environ Health 85 (2002), and “Erratum,” 17 Rev Environ Health. 248 (2002); Douglas Radford Shanklin & David L Smalley, “Kinetics of T lymphocyte responses to persistent antigens,” 80 Exp. Mol. Pathol. 26 (2006). Douglas Shanklin died in 2013. Susan J. Ainsworth, “Douglas R. Shanklin,” 92 Chem. & Eng’g News (April 7, 2014). Dr. Smalley appears to be still alive. In 2022, he sued the federal government to challenge his disqualification from serving as a laboratory director of any clinical directory in the United States, under 42 U.S.C. § 263a(k). He lost. Smalley v. Becerra, Case No. 4:22CV399 HEA (E.D. Mo. July 6, 2022).
[5] Hall v. Baxter Healthcare Corp., 947 F. Supp. 1387 (D. Ore. 1996); see Joseph Sanders & David H. Kaye, “Expert Advice on Silicone Implants: Hall v. Baxter Healthcare Corp., 37 Jurimetrics J. 113 (1997); Laurens Walker & John Monahan, “Scientific Authority: The Breast Implant Litigation and Beyond,” 86 Virginia L. Rev. 801 (2000); Jane F. Thorpe, Alvina M. Oelhafen, and Michael B. Arnold, “Court-Appointed Experts and Technical Advisors,” 26 Litigation 31 (Summer 2000); Laural L. Hooper, Joe S. Cecil & Thomas E. Willging, “Assessing Causation in Breast Implant Litigation: The Role of Science Panels,” 64 Law & Contemp. Problems 139 (2001); Debra L. Worthington, Merrie Jo Stallard, Joseph M. Price & Peter J. Goss, “Hindsight Bias, Daubert, and the Silicone Breast Implant Litigation: Making the Case for Court-Appointed Experts in Complex Medical and Scientific Litigation,” 8 Psychology, Public Policy & Law 154 (2002).
[6] Judge Jones’ technical advisor on immunology reported that the studies offered in support of the alleged connection between silicone implantation and silicone-specific T cell responses, including the published papers by Shanklin and Smalley, “have a number of methodological shortcomings and thus should not form the basis of such an opinion.” Mary Stenzel-Poore, “Silicone Breast Implant Cases–Analysis of Scientific Reasoning and Methodology Regarding Immunological Studies” (Sept. 9, 1996). This judgment was seconded, over three years later, in the proceedings before MDL 926 and its Rule 706 court-appointed immunology expert witness. See Report of Dr. Betty A. Diamond, in MDL 926, at 14-15 (Nov. 30, 1998). Other expert witnesses who published studies on the supposed immunogenicity of silicone came up with some creative excuses to avoid producing their underlying data. Eric Gershwin consistently testified that his data were with a co-author in Israel, and that he could not produce them. N.A. Schachtman, “Silicone Data – Slippery and Hard to Find, Part I,” Tortini (July 4, 2015). Nonetheless, the court-appointed technical advisors were highly critical of Dr. Gershwin’s results. Dr. Stenzel-Poore, the immunologist on Judge Jones’ panel of advisors, found Gershwin’s claims “not well substantiated.” Hall v. Baxter Healthcare Corp., 947 F.Supp. 1387 (D. Ore. 1996). Similarly, Judge Pointer’s appointed expert immunologist Dr. Betty A. Diamond, was unshakeable in her criticisms of Gershwin’s work and his conclusions. Testimony of Dr. Betty A. Diamond, in MDL 926 (April 23, 1999). And the Institute of Medicine committee, charged with reviewing the silicone claims, found Gershwin’s work inadequate and insufficient to justify the extravagent claims that plaintiffs were making for immunogenicity and for causation of autoimmune disease. Stuart Bondurant, Virginia Ernster, and Roger Herdman, eds., Safety of Silicone Breast Implants 256 (1999). Another testifying expert witness who relied upon his own data, Nir Kossovsky, resorted to a seismic excuse; he claimed that the Northridge Quake destroyed his data. N.A. Schachtman, “Earthquake Induced Data Loss – We’re All Shook Up,” Tortini (June 26, 2015); Kossovsky, along with his wife, Beth Brandegee, and his father, Ram Kossowsky, sought to commercialize an ELISA-based silicone “antibody” biomarker diagnostic test, Detecsil. Although the early Rule 702 decisions declined to take a hard at Kossovsky’s study, the U.S. Food and Drug Administration eventually shut down the Kossovsky Detecsil test. Lillian J. Gill, FDA Acting Director, Office of Compliance, Letter to Beth S. Brandegee, President, Structured Biologicals (SBI) Laboratories: Detecsil Silicone Sensitivity Test (July 15, 1994); see Gary Taubes, “Silicone in the System: Has Nir Kossovsky really shown anything about the dangers of breast implants?” Discover Magazine (Dec. 1995).
[7] Leroy Young, “Testing the Test: An Analysis of the Reliability of the Silicone Sensitivity Test (SILS) in Detecting Immune-Mediated Responses to Silicone Breast Implants,” 97 Plastic & Reconstr. Surg. 681 (1996).
[8] Affid. of Donard S. Dwyer, at para. 6 (Dec. 1, 1995), filed in In re Breast Implant Litig. Pending in U.S. D. Ct, D. Oregon (Groups 1,2, and 3).
[9] Notes of Testimony of Dr. Donnard Dwyer, Nyitray v. Baxter Healthcare Corp., CV 93-159 (E. & S.D.N.Y and N.Y. Sup. Ct., N.Y. Cty. Oct. 8, 9, 1996) (Weinstein, J., Baer, J., Lobis, J., Pollak, M.J.).
[10] Id. at N.T. 238-239 (Oct. 8, 1996).
[11] Id. at N.T. 240.
[12] Id. at N.T. 241-42.
[13] Id. at N.T. 243-44; 255:22-256:3.
[14] Id. at 244-45.
[15] Id. at N.T. 259.
[16] Id. at N.T. 258:20-22.
[17] Id. at N.T. 254.
[18] Id. at N.T. 252:16-254.
[19] Id. at N.T. 254:19-255:2.
[20] Id. at N.T. 269:18-269:14.
[21] Id. at N.T. 261:23-262:1.
[22] Id. at N.T. 269:18-270.
[23] Id. atN.T. 256:3-16.
[24] Id. at N.T. 262:15-17
[25] Id. at N.T. 247:3-5.
[26] Id. at N.T. at 260:2-3
[27] Id. at N.T. at 261:5-8.
[28] One of the more interesting and colorful moments came when the late James Conlon cross-examined plaintiffs’ pathology expert witness, Saul Puszkin, about questionable aspects of his curriculum vitae. The examination was revealed such questionable conduct that Judge Weinstein stopped the examination and directed Dr. Puszkin not to continue without legal counsel of his own.
[29] In re Breast Implant Cases, 942 F. Supp. 958 (E.& S.D.N.Y. 1996). The opinion did not specifically address the Rule 702 and 703 issues that were the subject of pending motions before the court.
[30] Hon. Jack B. Weinstein, “Preliminary Reflections on Administration of Complex Litigation” 2009 Cardozo L. Rev. de novo 1, 14 (2009) (emphasis added).
