Cancer, Causation, and Toxic Torts

For toxic tort defendants, particularly those embroiled in asbestos litigation, the Holy Grail has been limiting liability to only those damages attributable to their conduct. This article will explore how these defendants came to be at the mercy of the regularity, frequency, and proximity (RFP) test and why that test is inappropriate for deciding causation in cancer cases.

With that background in mind the discussion will turn to the DES cases from 30 to 40 years ago and how they might be used to challenge the prevailing orthodoxy on causation.

Plaintiffs often sue all defendants responsible for exposing them to the toxin in question, but not all of the defendants are equally responsible for the plaintiff’s injury. Yet, in most jurisdictions these less culpable defendants would be forced into sharing fault equally with their more culpable co-defendants. Even worse, the less culpable defendants have no way of predictably limiting their liability. They are subject to each jurisdiction’s method of accounting for settlements. Because these methods place no upper limit on their potential liability, they are forced to settle for amounts that even the most jaded plaintiffs’ lawyer would admit are well beyond their degree of responsibility.

An example from the asbestos world makes the point. The plaintiff is diagnosed with mesothelioma and sues everyone who ever exposed him to asbestos, except for the most culpable defendants. The most culpable defendants, meaning those who exposed the plaintiff to the greatest amount of asbestos, were the insulation manufacturers. Because the insulation manufacturers are all bankrupt, the plaintiff pursues the so-called peripheral manufacturers whose products contained little or no asbestos. Even if the peripheral manufacturers could find out whether the plaintiff received any money from the insulation manufacturers’ bankruptcy trusts before they go to trial, they would only be entitled to a setoff of the trust money against the verdict, regardless of how big it is.

The problem is aggravated by the local rule in many jurisdictions that a non-settling defendant cannot find out how much his settling co-defendants paid until just before closing arguments. With no upper limit on the verdict and a settlement cushion of uncertain size, the remaining defendants inevitably settle, rather than risk a potentially ruinous verdict. While this is one way for courts to control their dockets, no one should pretend that justice is being done.

The peripheral manufacturers have developed a number of ingenious defenses over the years. There’s the chrysotile defense, or my type of asbestos can’t cause your type of cancer. There’s the low-dose defense, or you couldn’t have breathed enough of my asbestos fibers to cause your cancer. There’s also the replacement parts defense, or you weren’t exposed to asbestos from my product in the first place, it was a replacement part manufactured by somebody else. These defenses have all met with some success, but also some failure. They fail because the medicine doesn’t unequivocally support them or they rely too much on the plaintiff’s memory or that of his co-workers.

Plaintiffs’ memories for these events are notoriously fallible. This should come as no surprise. They are diagnosed with an incurable and always fatal disease. A generation ago similar plaintiffs could remember with uncanny accuracy the names and product descriptions of every insulation manufacturer they encountered. When their lawyers explain that these companies are now bankrupt and their lawsuit is the only way they have to provide for their families’ future, many will forget the names of the insulation manufacturers and remember the names of the peripheral defendants.

Perils of `But For’ Causation

To avoid relying so heavily on a plaintiff’s memory, some manufacturers have pointed to “but for” causation as a way out of the mess. They argue that a plaintiff must prove that but for exposure to one of their products he would not have become sick. This argument, however, can be unfair to plaintiffs. If, say, 30 defendants’ products caused the plaintiff’s disease, it may be impossible to prove that but for his exposure to any one of them he would not have become sick.

Yet, all 30 products caused the plaintiff’s disease. To deny recovery under this circumstance would mean that justice was not done again. The better approach focuses on “but for” causation and the timing of the plaintiff’s injury. To understand this point, particularly when it comes to cancer cases, one needs to look a little bit harder at cancer.

While a detailed discussion of carcinogenesis is beyond the scope of this article there are four uncontroversial points to make. First, all cancerous tumors begin as a single cell. Second, no one knows which toxic exposure or exposures caused the first malignant cell to develop.

Some experts, particularly those on the plaintiff’s side, will opine that all previous toxic exposures through a variety of possible mechanisms caused the first malignant cell to appear. Because nobody knows exactly what causes a normal cell to reproduce as a malignant one, the merits of these explanations are debatable. Part of the reason these explanations can be so unsatisfying is the third point: Nobody can know when the first malignant cell appeared because nobody was peering inside the plaintiff’s body to see the first malignant cell start.

While some experts may opine about growth rates and then extrapolate backward to the time when the first malignant cell must have appeared, these remain opinions and not facts. Fourth, once a tumor starts growing subsequent exposures to the toxin will not change its rate of growth. (It is possible that subsequent exposures may result in the development of completely new tumors, but that is the rare case. Even so, the new tumors are of little importance because the original tumor will prove fatal before they do.)

With these principles in mind it is now possible to see how some toxic tort litigation, particularly asbestos, went off the rails and how litigants might bring it back on track. The original case discussing causation in asbestos cases, and which has expanded to all toxic tort cases, is Lohrmann v. Pittsburgh Corning Corp., 782 F.2d 1156 (4th Cir. 1986). Lohrmann was not just an asbestos case, it was an asbestosis case. Asbestosis, a scarring of the lungs from asbestos exposure, is a cumulative, progressive disease. Unlike cancer where subsequent exposures have no effect on the tumor once it starts growing, the same cannot be said of asbestosis, because each exposure causes more scarring and makes the condition worse.

Regularity, Frequency, and Proximity Test

Faced with a disease where every exposure to the toxin made it worse, the Lohrmann court devised the regularity, frequency, and proximity (RFP) test. Adapting the substantial factor analysis for causation from the Restatement (Second) of Torts to asbestos litigation, the court held:

To support a reasonable inference of substantial causation from circumstantial evidence, there must be evidence of exposure to a specific product on a regular basis over some extended period of time in proximity to where the plaintiff actually worked. Id. at 1162-63.

Given a fungible product and a cumulative, progressive disease, the Lohrmann RFP test makes a lot of sense. But both of these criteria must be met before the RFP test should be used. That is to say, the toxic element, whether it’s asbestos or something else, must be present in each potential defendant’s product. Furthermore, each exposure to the toxin must be capable of causing some additional injury. Heedless of these criteria, courts all across the country adopted the RFP test in asbestosis and non-asbestosis cases alike.

The RFP test worked well during the first wave of asbestos litigation when the majority of the cases were brought by asbestotics. Because the test worked so well, little thought was given to its application in cancer cases.

The RFP test is fundamentally a test for the admissibility of circumstantial evidence. Any toxic exposure satisfying the RFP test probably does contribute to further injury from a cumulative, progressive disease.

But the confidence one needs to draw the RFP conclusion about causation from circumstantial evidence doesn’t exist when it comes to cancer. Although cancer can be viewed as a progressive disease, meaning it gets worse over time, it is not a cumulative disease, meaning that each exposure makes it worse. Therefore, a toxic exposure satisfying the RFP test will not necessarily play any role in causing a tumor.

All cancers have a definite beginning as a single malignant cell. A toxic exposure satisfying the RFP test has no value as circumstantial evidence unless one knows when the exposure occurred. If a toxic exposure satisfying the RFP test occurred before the first malignant cell, it may have contributed to that cell’s development. But, if the toxic exposure occurred after the cell started growing, it could not have played a role in its causation. When it comes to cancer, the fatal flaw in the RFP test is that it does not relate the timing of the exposure to the timing of the tumor’s development. Despite the obvious disconnect between the RFP test and cancer, courts have nonetheless used it repeatedly.

The DES Cases

Although never applied to the toxic tort context, the DES cases from the 1970s and 1980s are squarely at odds with the RFP test. DES, or diethylstilbestrol, was a drug given to women from about 1940 to 1970 to reduce the risk of complications from pregnancy. As many as 300 companies manufactured and distributed it at one time or another. Two medical studies in 1971, however, suggested that DES could cause cancer in young women whose mothers took the drug during their pregnancy. Later that year the FDA banned the use of DES by pregnant women.

DES was a fungible product. Because of the hundreds of different manufacturers of DES and the length of time between the mothers’ use of the drug and the daughters’ development of cancer, proof of causation as to any particular manufacturer or manufacturers was often hard to come by.

To overcome the difficulties with their proof of causation, the DES plaintiffs proposed a number of theories based on each defendant’s market share. The California Supreme Court accepted a version of this theory in Sindell v. Abbott Laboratories, 26 Cal. 3d 588, 607 P.2d 924 (1980). The Sindell court held that a DES plaintiff must join as defendants the manufacturers with a “substantial share” of the market for the DES which her mother may have taken. Once the plaintiff proved a prima facie case except for the identity of the actual tortfeasor, the burden shifted to each defendant to prove that it could not have manufactured the DES causing the plaintiff’s injury. If a defendant could not exculpate itself, it would be held liable for part of the judgment based on its share of the market.

The Sindell decision was met with widespread criticism, and the California Supreme Court revisited its position eight years later in Brown v. Superior Court, 44 Cal. 3d 1049, 751 P.2d 470 (1988). To remedy some of the Sindell inequities, the Brown court held that liability should be several, rather than joint and several. In those cases where a plaintiff did not join all of the manufacturers in the market, the plaintiff’s total damages would be reduced by the percentage of the unjoined defendants’ market-share, with the joined defendants paying their proportionate share of this reduced amount.

Outside of California almost no court agreed with the Sindell theory of market-share liability, because a principled effort had not been made to define the relevant market or what constitutes a “substantial share” of that market. Rejecting the Sindell theory, the Washington Supreme Court proposed its “market-share alternate liability” theory in Martin v. Abbott Laboratories, 102 Wash. 2d 581, 689 P.2d 368 (1984). According to the Martin court, the plaintiff need not bring suit against a “substantial share” of the market, a single defendant could suffice. As in Sindell, the Martin court permitted the plaintiff to prove a prima facie case against a defendant without identifying whether the defendant was the actual wrongdoer. The burden then shifted to the defendant to exonerate itself.

Like its California counterpart, the Washington Supreme Court was forced to clarify its theory in George v. Parke-Davis, 107 Wash. 2d 584, 733 P.2d 507 (1987). Under the George approach the defendants are presumed to have an equal market share. They may rebut this presumption, however, by proving their actual market share, which then leaves them liable only for that percentage of the plaintiff’s damages.

The Washington Supreme Court’s theory of market-share alternate liability also came under heavy fire. Smaller defendants who had the misfortune of being the only one named or whose records no longer permitted proof of their actual market share could end up paying a substantial judgment. Many critics pointed to this obvious unfairness as a reason for rejecting the Martin theory.

The Wisconsin Supreme Court tried its hand at fashioning an appropriate remedy in Collins v. Eli Lilly Co., 116 Wis. 2d 166, 342 N.W. 2d 37 (1984). The Collins court came up with what it called the “risk contribution theory.” As with the other theories, the plaintiff did not have to identify the actual tortfeasor. And, as was true in Washington, a DES plaintiff in Wisconsin could sue just one defendant. Wisconsin broke with Washington, though, by holding that a defendant could exonerate itself only by proving that it never marketed DES during the relevant time period or in the relevant geographical market. If the one defendant cannot exculpate itself, it will be liable for all the plaintiff’s damages.

The Collins court decided to impose liability on a defendant merely for creating a risk of harm. As the critics correctly pointed out, this theory of liability contravened one of the most fundamental tort principles, a mere possibility is not enough to prove causation.

Taking these decisions to their logical extreme, the New York Court of Appeals adopted pure market-share liability in Hymowitz v. Eli Lilly & Co., 73 N.Y. 2d 487, 539 N.E. 2d 1069 (1989). The plaintiff need only prove that she suffered an injury caused by DES. Any defendant will then be held severally liable based on its share of the national market for DES, but no defendant can exculpate itself. Like its predecessors the Hymowitz decision was severely criticized.

Of those cases rejecting any form of market-share liability, the Illinois Supreme Court’s decision in Smith v. Eli Lilly and Co., 137 Ill. 2d 222, 560 N.E. 2d 324 (1990), is typical. Part of the problem, as the Smith court saw it, was the lack of good data to establish the defendants’ market share. Without good data, verdicts would be arbitrary and unjustly variable.

Setting aside the market-share proof problems, the Smith court raised some more fundamental objections. Market-share liability means that a defendant who has no connection with the product causing the harm can be held liable. The vice was “that some defendants wholly innocent of wrongdoing towards the particular plaintiff will shoulder part or all of the responsibility for the injury caused.” Id. at 258. As will be seen, this vice is inherent in rote application of the RFP test.

The absence of some defendants compounded the problem. The plaintiff named only eight defendants out of the hundreds who manufactured and sold DES. “To say that the true defendant has been located is speculative and improper conjecture.” Id. at 259. The same can be said for much of today’s toxic tort litigation.

As for the public policy reasons favoring adoption of market-share liability, the Smith court thought it unlikely that safety could be improved by holding a defendant responsible for decades-old conduct and after the offending product’s removal from the market. Again, this same rationale could be applied in the current toxic tort environment. The court feared that imposing liability could result in defendants paying damages “which rightfully belong[ed] to companies which are insolvent.” Id. at 267. Therefore, it rejected liability based solely on a breach of duty without proof of causation. Proof of causation in fact would remain a “vital part” of Illinois tort law.

`Evolution’ of RFP Test

Yet, just over two years later the court abandoned these principles in Thacker v. UNR Industries, Inc., 151 Ill. 2d 343, 603 N.E. 2d 449 (1992). The plaintiff’s husband died from lung cancer (not mesothelioma), hastened by asbestosis. In other words, he died from a combination of a cumulative, progressive disease, asbestosis, and a non-cumulative, progressive disease, cancer. This should have greatly complicated the court’s analysis, but it seemed unaware of the potential difficulties. The defense did little to help the court through them, as it argued only that the plaintiff was not exposed to its asbestos.

The defendant’s approach invited the court to ignore the difference between asbestosis and cancer by adopting the RFP test which it did. The court decided that satisfying the RFP test proved that the decedent was exposed to enough asbestos to cause his disease. Because that was how the defendant framed the argument, the court never considered when the exposure occurred. Did the RFP exposure actually occur before or after the fatal cancer started growing? The plaintiff did not bother to answer this question, so it was impossible to decide whether the decedent’s exposure to the defendant’s product could or could not have caused his cancer. It is not enough, as the court suggested, to show through the RFP test that the decedent inhaled some of the defendant’s asbestos. The plaintiff should also have proven that the decedent inhaled some of the defendant’s asbestos when it could have contributed to the cancer.

Despite the obvious application of the various market-share theories of liability to asbestos and other toxic torts, courts, even in those jurisdictions that have adopted the theory, refuse to do so. The reasoning of the Oklahoma Supreme Court in Case v. Fibreboard Corp., 743 P.2d 1062 (Okla. 1987), is a good example. The court began by noting that the DES cases were “decided in the context of a product that was truly fungible.” Id. at 1065. Unlike DES, however, asbestos is not fungible because each mineral in the asbestos family (chrysotile, crocidolite, tremolite, etc.) carries a different risk as does each form of the product (encapsulated, free, etc.).

If the market-share liability cases have it right and asbestos products are not truly fungible, what does that say about the continuing vitality of the RFP test? The RFP test presupposes that an exposure to any product satisfying the test supports at least an inference of substantial causation. Assuming asbestos products are not fungible, then satisfying the RFP test is meaningless because the exposure to one type of mineral in one product form may create a trivial risk of harm compared to another mineral in a different product.

Imagine two asbestos products. According to the market-share liability cases, they are not fungible. If a plaintiff were to prove an exposure to each product, thus satisfying the RFP test, what does that tell the fact finder about whether either product was a substantial factor in causing the plaintiff’s cancer? Surprisingly little. Without knowing the risks for the mineral varieties in the two products and the different risks for the products themselves, it is impossible for the fact finder to infer anything about whether either was a substantial factor in the cause of the plaintiff’s disease.

The Lohrmann court itself recognized the limitations of its test when it said:

Such a rule is in keeping with the opinion of the plaintiff’s medical expert who testified that even thirty days exposure, more or less, was insignificant as a causal factor in producing the plaintiff’s disease. (Lohrmann, 782 F.2d at 11621-63.)

Under the Lohrmann court’s original formulation, a plaintiff’s exposure to, say, asbestos-containing gaskets and packing in a valve on a handful of occasions would not satisfy the RFP test.

But that’s not how courts interpret the RFP test today. Over the 25 years since Lohrmann, plaintiffs’ experts gradually changed the definition of regularity and frequency. Beginning with the idea that there is no known minimum threshold for asbestos exposure when it comes to causing mesothelioma, they argued with a great deal of success that any exposure was enough. What the Lohrmann court had described as a de minimis rule became exactly that. Any exposure to asbestos permitted an inference of substantial causation. And the RFP test turned into just a proximity test with no need to establish frequency and regularity.

Regardless of how it is applied, though, there is a more fundamental problem with the RFP test. It is in reality a market-share liability theory with none of the protections defendants usually enjoy. As the district court in Hardy v. Johns-Manville Sales Corp., 509 F. Supp. 1353, 1358 (E.D. Tex. 1981), said, “in the [last] ten years … we have been imposing a kind of Sindell liability.” The crucial difference, though, is that today’s defendants are tagged with Sindell liability without the market-share limitation on the amount of damages. Rather than facing only several liability limited by the amount of their market share, they face devastating liability subject only to setoffs from the settling defendants.

Practical Tips for Litigators

To be sure, asbestos and other toxic tort cases have not dispensed with proof of causation altogether. The plaintiff must still prove that he worked in proximity to a defendant’s product regardless of whether he did so with the kind of regularity and frequency the Lohrmann court envisioned. The problem is that no court requires a plaintiff to prove that the exposure occurred before the cancer starting growing. Without this crucial piece of evidence, it is entirely possible, as the Illinois Supreme Court feared, that wholly innocent defendants will be found liable.

To set up these arguments, the defense lawyer needs to establish a few facts with a plaintiff’s experts. Most importantly, the plaintiff’s experts should agree that the toxin is genuinely fungible.

Next, none of the plaintiff’s experts knows when the plaintiff’s tumor started growing. While the expert may have an opinion on the subject, he cannot know because he was not there at the time.

Assuming that the plaintiff’s expert has an opinion about when a tumor started growing, he must be questioned about the tumor’s growth rate. If he does not have an opinion about the growth rate, his opinion about when the tumor started growing loses credibility. If he has an opinion, he should be questioned about the size of the tumor at diagnosis, measured in terms like the number of doubles or something specific that can be tied to the growth rate.

Armed with this information, the defense lawyer can extrapolate back in time to when the tumor started growing. If the plaintiff’s expert gives a broad range for the doubling time, the same exercise can be used to develop end points for the period when the tumor started growing. Either way, the lawyer for a peripheral defendant may uncover evidence where the exposure to his client’s product occurred after the tumor started growing.

Once these ground rules have been established through the plaintiff’s experts, the defendant’s lawyer can then take advantage of the DES line of cases in that jurisdiction. If the jurisdiction recognizes market-share liability and the plaintiff’s experts concede fungibility, he can argue that the defendant should have its percentage of responsibility limited to its share of the asbestos market.

Or, if the jurisdiction follows the RFP test and rejects the application of market-share liability to toxic tort cases, and the plaintiff’s experts deny fungibility, he can argue that the RFP test should be reconsidered because his product is not truly fungible. Because the jurisdiction does not recognize market-share liability, he now has the facts for a dispositive motion based on the DES cases, arguing that the plaintiff’s proof could subject an innocent defendant to liability.

The larger point is that the defendant can use the conflict between the DES cases, one way or the other, and a jurisdiction’s standards for proving causation in toxic tort cases to push for a new standard of proof in cancer cases, one more closely aligned with reality.