by Mike Kelly and Andrew Dupre
Insurance contracts are a language game that must be played very carefully, as AstraZeneca learned.
From 2003 to 2012, AstraZeneca Plc was locked in one of the largest mass tort litigations in the United States, defending its blockbuster second generation atypical antipsychotic product Seroquel from allegations of failure-to-warn of risks of metabolic injury such as diabetes mellitus Type II. AstraZeneca prevailed on the lone case to reach a jury, won an additional nine cases during pretrial proceedings nationwide–losing none, and then ended the litigation with a global settlement acclaimed in the business press as a “great deal,” at roughly 50% of what its competitors paid. Yet through a curious twist, AstraZeneca’s insurance claim for the cost of its winning defense was rejected to the tune of £133 million. AstraZeneca Insurance Company v. XL Insurance (Bermuda) Ltd. & ACE Bermuda Insurance Ltd. – a case of first impression under England & Wales law – has important implications for any company insuring US pharmaceutical mass tort risks.
Over 23,000 plaintiffs in multiple US jurisdictions alleged that AstraZeneca’s atypical antipsychotic (colloquially, “AA”) product Seroquel, along with competing products in the AA class such as Eli Lilly’s Zyprexa and Johnson & Johnson’s Risperdal, “caused” diabetes mellitus type II and related metabolic conditions, of which the relevant FDA product labels failed to adequately warn.
In 2005, Eli Lilly reportedly settled approximately 8,100 Zyprexa actions for $700 Million, long before any case was triable and after only a bulk first round of discovery. That settlement, however, required multiple reopeners to accommodate an ever-increasing number of new filings, and Eli Lilly reportedly settled approximately 18,000 additional Zyprexa actions for another $500 million in 2007. Its total settlement value (sans legal costs) increased to a reported $1.2 billion, or $45,997 per case in 2007, not including the additional $1.4 billion Eli Lilly paid to settle the various criminal investigations.
AstraZeneca took the opposite course in its defense of Seroquel, declining to settle early and forcing the major US plaintiffs’ firms to put proof on the table to value their claims. Key to that strategy was that the plaintiffs were unable to stick to a story for how AAs supposedly cause metabolic injury. The law firms driving the litigation first asserted that AAs cause diabetes through some unidentifiable direct mechanism; i.e. use the product, wake up with metabolic disease. As the cases crawled through expert discovery, it became clear that theory would be deemed empty speculation, not admissible in any US court. The claimant law firms thus changed horses midstream to allege that AAs cause weight gain and that the weight gain causes metabolic injury, a theory described as “indirect causation”. The problem, however, was that no specific plaintiff could possibly prove that any specific pound they gained was caused by an AA product rather than factors such as diet, and no specific pound can be proven to cause a metabolic condition such as diabetes mellitus. In other words, the claimant law firms could not prove their cases, and AstraZeneca knew it.
AstraZeneca won a series of court rulings while embarking on settlement negotiations. Between 2010 and 2011, under the aegis of mediator Professor Stephen Saltzburg of the George Washington University School of Law, AstraZeneca reported that it had settled approximately 28,461 Seroquel suits for $647 million or $22,732.86 per case. Moreover, the vast press about the Seroquel litigation, along with subsequent FDA clarifications, set a firm date for any reasonable patient to be on notice of their claims, establishing a rock-solid statute of limitations defense to subsequently filed cases.
The business press hailed AstraZeneca’s defense of the Seroquel mass tort actions. Media outlets including Bloomberg termed the settlement “a great deal,” while the Wall Street Journal called it “a pleasant surprise for shareholders.” AstraZeneca’s reported 2010 gross sales of Seroquel were $5.3 Billion, meaning that the company was able to end virtually all tort claims over the product for approximately 6.3 weeks of its sales value in 2010.
The Insurance Dilemma: Too Good For One’s Own Good?
AstraZeneca insured its products liability risk through an eponymous Bermuda captive for the first £365 million of losses. It subsequently reinsured a £133 million excess layer of coverage immediately above £365 million with ACE and XL, subject to a $100 million “per occurrence” cap. The intended result was for ACE and XL to separately insure up to $100 million of AstraZeneca’s product liability exposure for Seroquel, among other products, in excess of losses of £365 million.
Unusually, AstraZeneca and its captive agreed to have their insurance contract governed by the law of England & Wales. Bermuda form insurance contracts typically incorporate New York law with a London arbitration choice of forum. For reasons unclear on the face of the opinion, these otherwise form contracts were modified to English law with High Court, Queen’s Bench Division Commercial Court as the choice of forum. That choice carried through to the reinsurance policies with ACE and XL, who as reinsurers were liable to indemnify only a category of losses for which the captive insurer was originally liable.
ACE and XL rejected the indemnity demand. Their principal argument was that English law mandates coverage only for actual liabilities and not for settlement of disputed claims, however logical and winsome such a settlement may be. Thus, the insurers argued that they owed no duty to cover the amounts AstraZeneca agreed to pay claimants in settlement. Further and perhaps more disputatiously, the insurers argued that the right to recover cost of defense is contingent upon the legitimacy of the underlying claim. Because AstraZeneca settled the Seroquel tort litigation having never been proven liable, i.e. from a position of negotiation strength, the insurers argued that they owed no duty even to cover AstraZeneca’s legal defense costs. In sum, the insurers refused to pay anything at all because AstraZeneca had won the Seroquel mass tort litigation “too well.”
The insurers prevailed Justice Flaux rejected AstraZeneca’s arguments that the reinsurance contracts must be interpreted within the context of the US pharmaceutical product liability litigation risk that they were meant to cover, and through norms of Bermuda insurance law. Instead, the Court held that the incorporation of English law into the reinsurance contracts evinced a conscious choice to invoke a far narrower scope of coverage. Because English law does indeed limit an insurers’ coverage obligation to only “actual” – rather than disputed – liabilities, the insurers owed no duty to cover the Seroquel settlement. Moreover, because English law does not recognize the common law duty of an insurer to defend its insured present in all US jurisdictions, English insurers owe a duty to cover defense costs only upon an express undertaking to “sue and labour,” which was not present in these Bermuda form contracts.
Should AstraZeneca v. XL stand on appeal, it will mean that an English law insurance contract, and all reinsurance policies derived from it, are worthless for settling US pharmaceutical mass tort liabilities not already lost in a trial. To attain coverage for a global settlement, a manufacturer would have to prove that it would more-likely-than-not have lost the litigation underlying the coverage demand. In other words, the company would have to prove – publicly – that its own product causes injury, a tack no rational pharmaceutical company would pursue. Even to attain defense cost coverage would require either at least one straight trial loss or an arduously worded sue-and-labor rider that is otherwise foreign to standard Bermuda form insurance contracts.
Moreover, the right to coverage under English law, even in the face of a lost trial, is now dubious. Assume for the sake of argument that a single plaintiff had won a trial in which she alleged that an AA product caused her diabetes mellitus. That loss would not necessarily create a coverage obligation for a global settlement. Diabetes has many causal factors, and an insurer could reasonably argue that the ability of Claimant 1 to prove causation does not extend to Claimants 2 through 23,000, rendering a global settlement a non-covered event.
The critical lesson learned for the pharmaceutical industry is the importance of not having form insurance contracts modified to English law. In most cases, the problem faced by AstraZeneca should be avoidable by a simple default to Bermuda form insurance contracts incorporating New York law with a London arbitration forum. At the very least, pharmaceutical companies should be aware when entering into their insurance agreements that they must be careful to avoid coverage that will punish them for winning their litigation “too well.”
Michael P. Kelly is the Chairman of McCarter & English LLP, and was formerly AstraZeneca’s Delaware counsel who won the four dismissals that led the Seroquel products liability settlement, and was a lead negotiator of that settlement. Andrew Dupre, also of McCarter & English LLP, is a solicitor of England & Wales who worked extensively on the Seroquel mass tort litigation.