First published in Drug Recall Litigation Reporter, March 2004.
Reprinted by permission.
Drug manufacturers do not have a legal duty to use pharmacogenetics or pharmacogenomics to reduce the risk of idiosyncratic drug-induced liver injury, according to Texas defense attorney Jack E. Urquhart and associates C. Kelvin Adams and Laura E. De Santos, who say the fields are simply too immature to serve as the foundation for a liability claim.
Pharmacogenetics and pharmacogenomics may eventually enable science to identify genetic populations at increased risk of drug-induced liver injury. At present, however, these disciplines have not developed scientifically reliable methods for identifying at-risk populations. Drug manufacturers, therefore, have no legal duty to use these unproven methodologies.
Idiosyncratic Drug-Induced Liver Injury
Idiosyncratic drug-induced liver injuries occur rarely with most drugs. W.M. Lee, Drug-Induced Hepatotoxicity, NEW ENG. J. MED., 349:474-485 (2003). "Most idiosyncratic drug reactions result from a succession of unlikely events, a Â‘multi-hit' process." Id. The infrequency of these injuries is evidence that genetic variants alone are an unlikely cause of severe drug-induced liver injury. Id.
Pharmacogenetics and Pharmacogenomics
The terms pharmacogenetics and pharmacogenomics are used imprecisely and, at times, inconsistently. The initial Food and Drug Administration industry workshop on these subjects was held May 16-17, 2002, and adopted working definitions that will be used here.
Pharmacogenetics is the application of a single gene sequence or a limited set of multiple gene sequences to study variation in DNA sequences related to drug action or drug disposition. Lesko et al., Pharmacogenetics and Pharmacogenomics in Drug Development and Regulatory Decision Making: Report of the First FDA-PWG-PhRMA-Drug Safe Workshop, J. CLIN. PHARMACOL., 43: 342-358 (2003).
Pharmacogenomics is the application of genome-wide single-nucleotide polymorphism maps and genome-wide gene expression analysis to study variations that influence drug action. Id.
The theoretical use of these disciplines in drug development is exciting. See K.A. Phillips et al., Potential Role of Pharmacogenomics in Reducing Adverse Drug Reactions, J. AM. MED. ASS'N, 286: 2270-2279 (2001).
"Eventually, drug-mediated injuries may be prevented by screening methods that can identify aberrant gene polymorphisms or RNA-expression profiles before a patient uses a drug." See Lee, supra. If genetic populations at particular risk of idiosyncratic drug-induced liver injury could be identified, drug labeling could contraindicate the drug for members of that high-risk genetic group. This would permit the use of some drugs that are now unmarketable by those without the genetic risk.
Imagination is the only limitation to the potential benefits of pharmacogenetics and pharmacogenomics. Focusing only on drug safety, these disciplines may one day play a significant role in preclinical toxicology. Already, a field referred to as "toxicogenomics" is developing. The FDA workshop recognized "the huge potential for these sciences to improve the drug development process and address future public health needs." With this advance in science, do drug makers have a duty to use it?
Reliability of Pharmacogenetics and Pharmacogenomics
Pharmacogenetics and pharmacogenomics, for all their promise, are still experimental. The use of these disciplines is not required by the FDA. Their methods are not yet generally accepted. Few published studies apply pharmacogenetics or pharmacogenomics to real-world issues of drug safety or drug efficacy. Those that do illustrate the current limitations of those disciplines.
Pharmacogenetic studies have been published that explore the relationship between genetic variations and drug-induced liver injury. An example is the study of genetic variations as a potential risk factor for patients taking anti-tuberculosis drugs. Huang et al., Polymorphism of the N-acetyltransferase 2 Gene as a Susceptibility Risk Factor for Anti-tuberculosis Drug-Induced Hepatitis, HEPATOLOGY, 35: 883-889 (2002); Huang et al., Cytochrome P450 2E1 Genotype and the Susceptibility to Anti-tuberculosis Drug-Induced Hepatitis, HEPATOLOGY, 37: 924-930 (2003).
Patients taking anti-tuberculosis drugs are at increased risk of hepatitis. Id. The hypothesis was developed that genetic factors made some individuals more susceptible to injury from anti-tuberculosis drugs. The Huang studies focus on genetic variations in the drug metabolizing proteins N-acetyltransferase and cytochrome P450 2E1.
In the second Huang study from 2003, which focused on CYP2E1, 318 tuberculosis patients had their CYP2E1 genotyped. The genotyping method used was a polymerase chain reaction with restriction length polymorphism. Three genotypes were revealed, and the patients were divided into groups genotype c1/c1, genotype c1/c2 and genotype c2/c2.
Forty-nine of the 318 patients in the study developed drug-induced hepatitis. The patients with genotype c1/c1 had a higher incidence of hepatitis than the other two groups. The authors concluded that, after adjusting for all other known risk factors, the c1/c1 genotype remained an independent risk factor.
With this information, however, the authors did not reach a causation conclusion. They wrote that the patients with genotype c1/c1 "may be at increased risk for anti-tuberculosis drug-induced hepatitis." The authors did not suggest contraindicating anti-tuberculosis drugs for this genotype. Instead, they wrote that regular liver enzyme monitoring for this group "may be considered."
Referring to the first Huang study from 2002, Lee distinguished between the ability of pharmacogenetics to identify genotypes that result in different plasma levels of a drug and the ability to actually identify specific markers of very rare idiosyncratic reactions. The Huang studies do not identify such markers, and Lee expresses doubt that this can be done, "particularly if the reaction involves multiple steps."
The FDA workshop was first convened in 2002 to address, among many related topics, the current utility of pharmacogenetic and pharmacogenomic methods in drug development and drug regulation. Recognizing the great possibilities of these emerging fields, the workshop, nevertheless, revealed significant questions about genomic testing and data quality.
The workshop, for example, questioned the ability of a genotyping assay to study some rare gene variants because too few people have them. It reported a consensus on the need for "standardized reference materials, standards for assay validation and specific regulatory guidance for validation criteria of methods."
It also reported the lack of data from well-controlled clinical trials. The workshop reported that for a variety of reasons, including the lack of uniformity of data collection and the use of differing technologies, the interpretation of data lacked certainty. Future workshops, of course, will continue to pursue the use of pharmacogenetics and pharmacogenomics. The prevailing sentiment is that scientific reliability of these sciences requires only "evolution" and not a "revolution."
The Duty to Use Pharmacogenetics or Pharmacogenomics to Identify At-Risk Patients
The FDA does not require the use of pharmacogenetics or pharmacogenomics for drug approval. Reserving a discussion of preemption, does a drug maker have a common-law duty to use these technologies?
A patient who allegedly suffered an idiosyncratic drug-induced liver injury could claim that the drug maker had a duty to use pharmacogenetics and pharmacogenomics. If the drug maker had used those techniques, the patient could argue, they would have revealed his or her genetic vulnerability to drug-induced liver injury. A claim of this nature should fail as a matter of law.
The cause of action for such a claim is a marketing or design defect. The better-reasoned cases reject "failure to test" as an independent cause of action. See Akee v. Dow Chem. Co., 2003 WL 21697345 (D. Haw. 2003); Stitt v. Philip Morris Inc., 245 F. Supp. 2d 686 (W.D. Pa. 2002); Cf., American Tobacco Co. v. Grinnell, 951 S.W.2d 420 (Tex. 1997).
Expert testimony is required to support the claims of product defect in this highly complex field. An expert must testify that the failure to use pharmacogenetic or pharmacogenomic techniques resulted in a marketing or design defect. This opinion, regardless of who gives it, should not make it past the gatekeeper. The opinion cannot currently meet the scientific reliability requirements of Daubert v. Merrell Dow Pharms., 509 U.S. 579 (1993).
The hypothesis that pharmacogenetics or pharmacogenomics can identify specific genetic markers of very rare events of idiosyncratic drug-induced liver injury has not been tested. Peer-reviewed publications do not support the hypothesis with well-controlled studies. Needless to say, an error rate is unknown. The use of pharmacogenetics and pharmacogenomics for this purpose is not generally accepted and seldom used. Therefore, none of the Daubert indicia of scientific reliability currently are met. Testimony about pharmacogenetics and pharmacogenomics should be excluded in a trial about an alleged idiosyncratic drug-induced liver injury.
If the expert testified in a case involving an anti-tuberculosis drug, for example, he or she may attempt to use the Huang studies. If the patient-plaintiff was a c1/c1 genotype, the expert may attempt to opine that the drug maker should have done this testing and, at a minimum, should have warned that c1/c1 genotypes are at increased risk. This testimony must also be stopped at the gate. It fails to meet Daubert reliability requirements. Granted, the Huang studies are published in one of the finest peer-reviewed journals, but, as Daubert teaches, the studies must "fit" the case.
These studies are valuable for many reasons, but they are not a basis for testifying that a drug maker should have performed similar studies. Nor do they support the opinion that the failure to perform such tests resulted in a product defect. The Huang studies, as noted by Lee, do not identify specific markers of liver toxicity. They are experiments with a small group of patients that shed light on a possible mechanism of drug toxicity. Even the authors of the study are extremely reserved in the conclusions they draw from the studies.
The assertion that a drug maker should have used pharmacogenetics or pharmacogenomics is not relevant in a case alleging an idiosyncratic drug-induced liver reaction. The drug maker has no regulatory or common-law duty to use these technologies. If admitted, the testimony that the drug maker did not use these techniques will serve only to confuse or prejudice the jury, but probably both.
Daubert speaks of the power of expert testimony and the court's gatekeeping obligation to carefully tend the gate. As the Supreme Court cautioned in Daubert, science is a process that is constantly adjusting and correcting its course. A trial is about achieving closure and finality. Pharmacogenetics and pharmacogenomics are in the infancy of their scientific journey and poorly suited for courtroom opinions.
Pharmacogenetics and pharmacogenomics offer great hope for the future. As these disciplines evolve, they may unravel important genetic mysteries about idiosyncratic drug-induced liver injuries. They may locate specific genetic markers for these rare reactions. Prematurely imposing an actual or de facto duty on drug makers to use these technologies will discourage the cooperative atmosphere needed to overcome the current limitations of pharmacogenetics and pharmacogenomics.
Exposing drug makers to "evidence" that they "failed" to use these experimental techniques or used them incompletely serves no valid purpose. In a trial it only introduces irrelevant, confusing and prejudicial testimony. In the arena of drug development and drug regulation, it will impede the progress that the FDA workshop has begun.
The workshop, in fact, expressly addressed the need for a "safe harbor" provision in any federal regulations involving pharmacogenetics and pharmacogenomics. Without this safe harbor, the FDA recognizes that needed experimentation in these sciences will be discouraged. A regulatory safe harbor, however, will be undermined if courtrooms turn an experimental science into an industry duty.
Pharmacogenetics and pharmacogenomics raise a host of new ethical considerations that are not completely identified, and certainly not resolved. These disciplines assume the creation of extensive genetic databases on all people who use prescription drugs. Not all people will be thrilled by that prospect.
If the goal of identifying the very few who should not take a drug because of an increased safety risk is achieved, what medical treatment will those excluded receive?
If risk factors are related to race or gender, what are the implications of, for example, a blood pressure drug suitable for white males only? Perhaps, more fundamentally, how will the difficult issue of genetic studies based on hypothetical ethnic differences be resolved?
The law is also a process, just a very different process from science. The best available relevant scientific information should be the basis of courtroom testimony. Pharmacogenetics and pharmacogenomics are too new, too experimental, and too untested to support a legal duty or even to support courtroom testimony.
Jack E. Urquhart, C. Kelvin Adams, and Laura E. De Santos are civil litigation defense attorneys practicing in the health care, pharmaceutical, and biotech area of the law firm Beirne, Maynard & Parsons. They represent pharmaceutical and biotech companies and have amassed a significant win record in this practice.