The Double-Edged Sword: Genomic Profiling in Drug and Chemical Litigation

Since completion of the Human Genome Project, genomic profiling and the related advent of personalized medicine have become a hot topic. Many predicted this new genomic information would lead to targeted drug treatments that work better with less risk. Along with this optimism came concern about new litigation, as plaintiffs seek and find novel liability theories. What have we seen? Translating scientific knowledge into clinical practice has presented difficulties, some expected and some surprising, but there have been exciting advances, too. Increasing computer power and data collection may drive major new breakthroughs. And while plaintiffs have found ways to use genomic data in products liability cases, defendants also have found ways to parry the novel allegations and riposte, with causation defenses supported by genomic evidence.

What Is Personalized Medicine?

The Human Genome Project began in 1990 and was completed in April 2003, culminating in a complete map of all of the human body’s genes. This enabled the medical community to better understand how genetic expression may predispose a person to develop a certain disease or affect the way a body metabolizes a drug. The science confirmed long-recognized associations between certain inherited traits and medical risks and helped develop the field of personalized medicine.

Broadly speaking, personalized medicine refers to the use of new molecular analysis methods to better discern a patient’s disease risk or to develop or select the medication most suited to the patient’s genetic profile. Genetic testing is a core tool. New technology does not always help target therapy: scientists have discovered many mutations in people who have the same disease, and it can be difficult to determine which is the driver.

The ability to identify patients based on genetic information can be crucial to the positive risk/benefit balance for a drug, however. When side effects may be severe, there is benefit to ruling out patients vulnerable to those effects or identifying people most likely to respond to the drug. Increasingly, targeted therapies are developed along with a diagnostic test to identify the right people. In August 2014, the FDA issued its Guidance on In Vitro Companion Diagnostic Devices. This reflects a focus on clinical importance, separate from the analytical validity of test results. When treatment decisions will be based on the results, FDA wants strong clinical evidence that these are the right decisions.

How Are Genetic Profiles Being Used in Litigation?

The ability to profile a human’s genes is being used as a sword by both sides of the “v” in lawsuits. The science can factor into everything from pleading requirements to discovery to expert opinions. Plaintiffs may allege a duty to discover and warn about differing drug reactions, based on genotype. Courts have screened these claims out when they are not backed up by actual evidence, showing the plaintiff has a material variant. In a twist that may have broader potential, vaccine plaintiffs have been blocked when the defense argues the patient’s underlying genetic profile is strongly correlated with the injury, so this relationship independently accounts for cause.

Manufacturers’ Duties Arising From Personalized Medicine

Plaintiffs may allege that a manufacturer should look for genetic variants that affect drug response and warn doctors about particular genetic susceptibilities.

In jurisdictions that follow the Restatement (Third) of Torts, manufacturers will benefit from its relative safe harbor, which helps insulate manufacturers from design liability except where the risk/benefit balance means doctors would not prescribe the drug or medical device “for any class of patients.” Products Liability § 6(c). Given that a medication designed to target certain diseases is likely both to increase the therapeutic benefit for the intended patient and reduce the risk for the general population, plaintiffs will be hard pressed to argue that the drug should not be prescribed for any class of patients.

At this time, all plaintiffs face major challenges in claiming that a manufacturer failed to test for potential adverse drug reactions or otherwise target therapies based on genetic profiles. In its 2013 Pharmacogenomics Guidance, FDA recognizes that “most pharmacogenomics research is exploratory at this stage and is often intended to discover relationships for which no prior hypotheses exist.” In addition, most jurisdictions have rejected the idea of an independent duty to test and instead consider failure to test allegations only as a component of design defect or failure to warn claims. Without clinical evidence in targeted populations, it would be hard to establish defective design or the need for warnings. The Newman v. McNeil Consumer Healthcare (N.D. Ill. 2013) case is illustrative, where the court excluded as “speculative and irrelevant” plaintiffs’ expert opinion about a potential genetic link that may be discovered in the future between ibuprofen and an adverse reaction.

Plaintiffs’ bread-and-butter theory of liability in pharmaceutical litigation is failure to warn. In personalized medicine lawsuits, plaintiffs may claim the drug manufacturer had a two-fold duty:

1. Identify on the label the potential genetic profiles at increased risk of adverse reactions, and
2. Warn that the patient should be tested for those profiles.

FDA stated in its 2013 Pharmacogenomics Guidance that drug labeling should include information on how genes may affect drug response “only if it is useful to inform prescribers about the impact (or lack of impact) of [the gene’s effect on the patient’s trait]; or to indicate whether a genomic test is available, and, if so, to indicate whether testing should be considered, is recommended, or is necessary.” Future litigation likely will focus on what is “useful” and reasonable, considering the potential known genetic risk.

Genomic evidence also could impact standard defenses. As to state of the art, the manufacturer is held to a standard of what was “reasonably scientifically knowable at the time of manufacture.” What is scientifically knowable takes on a whole new meaning with genomic profiling, but the current uncertainty in interpretation should provide plenty of room for reasonable scientific decisions not to test. Once a plaintiff is known, hindsight can suggest an “obvious” step. But reasonableness at the time should be viewed in light of the vast number of potential associations. A resistance to second-guessing was seen in Munro v. Regents of University of California (Cal. Ct. App. 1989), where the court in a medical malpractice case rejected a hindsight approach for determining the reasonableness of genetic testing recommendations, concluding that the specific test at issue “was not indicated by any facts which plaintiffs told to defendants, or even any facts of which plaintiffs were aware.”

When applying the learned intermediary doctrine, a key question will be whether the doctor was aware of a patient’s genetic predisposition for an adverse drug reaction — and whether the plaintiff’s genetic data are even relevant. For example, in Solomon v. Bristol-Myers Squibb Co. (D.N.J. 2013), the plaintiff claimed the defendant failed to warn physicians about the need to test for genetic susceptibility to excessive bleeding with Plavix use. But the prescriber was already aware of the risk for excessive bleeding for the general population at the time he prescribed Plavix for the plaintiff, and the court granted the defendants’ summary judgment motion in part on learned intermediary grounds, without reaching the merits of the genetics claim.

Plaintiffs’ Issues With Proof of Causation

Genetic information also will have a growing impact on proof of cause.

When genetic data has been associated with an adverse drug reaction, a plaintiff may be held responsible for determining if he or she has that profile. This starts at the pleading stage. In Mills v. Bristol-Myers Squibb Co. (D. Ariz. 2011), the court dismissed plaintiff’s allegations that Plavix was defective in warning and design, when plaintiff alleged only “on information and belief” that she was one of the poor drug metabolizers who may face an increased risk of adverse events. Plaintiff’s genetic information was solely in her own control and she had not taken the available test to determine her status. Because a significant majority of Caucasian people do not have the at-issue genetic variant, plaintiff’s claim “has not moved from the realm of possible to plausible.” Beyond the pleading stage, expert opinion may be screened out when there is no evidence that a person has the genetic profile linked with a claimed adverse reaction. One example is Agee v. Purdue Pharmaceuticals, Inc. (W.D. Okla. 2004), where the court excluded plaintiff’s expert’s opinion that “slow” OxyContin metabolizers may be more prone to overdose in part because the expert had no basis to conclude that the patient in question was a slow metabolizer.

Failure to obtain recommended genetic testing prevented recovery for parents who brought a vaccine claim for their daughter’s extensive but ambiguous neuromuscular condition in Simanski v. Secretary of Health & Human Services (Fed. Cl. 2014). The Special Master found the evidence supported a diagnosis different from the one the parents claimed, and the vaccine had not been shown to cause that condition. The order dismissing the case was affirmed, in part because the parents had not authorized genetic testing recommended by their daughter’s doctor to help determine the diagnosis.

An important discovery issue was presented in Kirk v. Schaeffler Group USA, Inc. (W.D. Mo. 2014), where the court allowed defendants to subpoena medical information from the plaintiff’s mother, a third party to the litigation, to determine whether a family history of autoimmune illness may have caused plaintiff’s condition. Because defendants could show that plaintiff’s mother had an autoimmune illness, the court was willing to order this discovery with safeguards for privacy. We have not researched the question of how unique this holding is. Given both the sensitivity of the information and the need for it, policy questions are raised.

A powerful argument is that plaintiff’s genetic profile alone accounts for her condition, or precludes reliable expert testimony linking a drug or chemical exposure to the condition. In vaccine cases, plaintiffs have been denied relief in part because they could not prove that the disease course would have been different without the vaccine. Two such cases are Godfrey v. Secretary of Health & Human Services (Fed. Cl. 2014), where plaintiff could not establish cause to recover for injury allegedly from the HPV vaccine given her genetic risk and family history, combined with a known trigger for her rheumatoid arthritis, and Waters v. Secretary of Health & Human Services (Fed. Cl. 2014), where plaintiffs could not establish that vaccines changed the course of their child’s seizure disorder, which was caused by a congenital genetic mutation. The cause standard varies by jurisdiction and claim, but this concept could be used in other contexts.

Framing the question matters and detail is relevant, as illustrated by Milward v. Acuity Specialty Products Group, Inc. (D. Mass. 2009), a case alleging that benzene exposure caused leukemia. Plaintiffs’ general cause expert was excluded after defendants focused on specific leukemia subtypes and treated each one as a different entity for which cause must be established. The district court found insufficient foundation to link benzene exposure reliably to the subtype plaintiff had. On appeal, the First Circuit reversed the exclusion, rejecting such a narrow focus and finding the totality of evidence linking benzene to leukemia sufficiently reliable. On remand, however, summary judgment was granted after plaintiff’s specific cause expert was excluded. Milward v. Acuity Specialty Prods. Grp., Inc. (D. Mass 2013).

Conclusion

Personalized medicine is designed to increase drug efficacy and reduce adverse events by using genetic information to more accurately predict individual response. Widespread success could reduce both injuries and claims. Even if such a golden age is possible, it is unlikely to emerge for some time. Drug manufacturers will need to understand plaintiffs’ theories and use genomic information affirmatively to dismantle unsupported claims.