Benzene and Worker Cancers: “An American Tragedy“

Bloated and bed-ridden, his skin browned by blood transfusions, John Thompson succumbed to leukemia on November 11, 2009.

A carpenter by trade, Thompson, then 70, had spent much of his life building infrastructure for the petrochemical industry in his native Texas — synthetic rubber plants in Port Neches, chemical facilities in Orange. Throughout the 1960s and early 1970s, he often encountered benzene, stored on job sites in 55-gallon drums, which he used as a cleaning solvent. He dipped hammers and cutters into buckets full of the sweet-smelling liquid; to expunge tar, he soaked gloves and boots in it.

Thompson never figured the chemical could do him harm. Not when it stung his hands or turned his skin chalky white. Not even when it made him faint. But after being diagnosed with a rare form of leukemia in 2006, relatives say, he came to believe his exposure to benzene had amounted to a death sentence. Oil and chemical companies knew about the hazard, Thompson felt, but said nothing to him and countless other workers.

“They put poison on his skin and in the air he breathed,” said Chase Bowers, Thompson’s nephew. “He died because of it.”

Thompson died before a lawsuit filed by his family against benzene suppliers could play out in court, where science linking the chemical to cancer could be put on display. Over the past 10 years, however, scores of other lawsuits, most filed by sick and dying workers like Thompson, have uncovered tens of thousands of pages of previously secret documents detailing the petrochemical industry’s campaign to undercut that science.

Internal memorandums, emails, letters and meeting minutes obtained by the Center for Public Integrity over the past year suggest that America’s oil and chemical titans, coordinated by their trade association, the American Petroleum Institute, spent at least $36 million on research “designed to protect member company interests,” as one 2000 API summary put it. Many of the documents chronicle an unparalleled effort by five major petrochemical companies to finance benzene research in Shanghai, China, where the pollutant persists in workplaces. Others attest to the industry’s longstanding interest in such “concerns” as childhood leukemia.

Taken together, the documents — put in context by interviews with dozens of lawyers, scientists, academics, regulators and industry representatives — depict a “research strategy” built on dubious motives, close corporate oversight and painstaking public relations. They comprise an industry playbook to counteract growing evidence of benzene’s toxic effects, which continue to command the attention of federal and state regulators and be fiercely debated in court.

“The conspiracy exists, and the conspiracy involves hiding the true hazards of benzene at low doses,” said Robert Black, a Houston lawyer who represents plaintiffs in toxic tort cases. Since 2004, while handling dozens of lawsuits filed on behalf of workers sickened by leukemia, lymphoma and other diseases associated with benzene, Black has obtained 16,000 pages of internal records detailing the industry’s research tactics, which he shared with the Center. Other lawyers provided an additional 5,000 pages.

The documents may represent the tip of the iceberg. For decades, the petrochemical industry has employed what one litigation guide calls a “comprehensive strategy” for defending against workers’ legal claims. Penned by a senior attorney at Shell Oil, the undated document lays out a coordinated “industry response” aimed at shielding internal company records on benzene. It warns defense attorneys to “avoid unnecessary or inadvertent disclosure of sensitive documents or information,” for instance, and to “disclose sensitive benzene documents only on court order.”

“We don’t know what the health effects are because they’re not going to let us know,” Black said. “It’s an American tragedy.”

Five Million Americans at Risk

Benzene emissions in the United States have declined sharply since 1987, when federal regulators set the occupational exposure limit at 1 part per million. Over roughly the same period, there has been a 66-percent drop in releases of the chemical into the environment. Yet experts say it remains a formidable threat. “You’re still seeing elevated risks of leukemias and lymphomas among occupational groups exposed to benzene,” said Peter Infante, a former director of the office that reviews health standards at the Occupational Safety and Health Administration, who has studied the pollutant for 40 years, “as well as populations being polluted from these benzene sources.”

In May, the U.S. Environmental Protection Agency estimated that 5 million Americans — not counting those with workplace exposures — face heightened cancer risks from benzene and 68 other carcinogens spewed into the air by one such source: the nation’s 149 oil refineries. “We are concerned about benzene,” said Kelly Rimer of the EPA’s Office of Air Quality Planning and Standards, which has proposed a rule that would require refinery operators to monitor for the chemical along their fence lines.

“It’s a known human carcinogen,” Rimer said, “and it’s emitted from lots of sectors.”

One month later, California officials lowered the long-term exposure level for benzene from 20 parts per billion to 1 ppb — among the lowest in the country — setting the stage for further emissions cuts at refineries and bulk-oil terminals in that state. Melanie Marty, of the California Office of Environmental Health Hazard Assessment, said regulatory limits are now “getting lower and lower for [benzene’s] non-cancer risks” — dizziness, rapid heart rate, neurological problems, anemia — and not just its carcinogenic effects.

“We have to make sure we’re not exposing people to things we can do something about,” she said.

A naturally occurring component of crude oil, benzene is used to make household products such as plastics, pesticides and dyes. It remains a key ingredient in gasoline, a source of exposure for workers as well as the public: In 2006, the EPA found benzene to be such a “significant contributor to cancer risk from all outdoor air toxics” that it limited levels in fuel. Even oil executives acknowledge its ubiquity; in documents, they call it “universal” and “a basic petrochemical building block.” Benzene ranks 17th among the top 20 chemicals produced in the United States, according to the federal government.

The petrochemical industry’s decade-long research effort on benzene echoes those launched by other industries — asbestos, tobacco, plastics — that used science to create doubt. These industries have employed a host of tactics to try to convince courts and regulators that a chemical or product causes no harm. At times, they funded their own studies in an attempt to show the lack of adverse effects. Experts say the petrochemical industry has bankrolled more research — at greater cost — than anyone but Big Tobacco, which coined the phrase “manufacturing doubt.”

“The more they feel threatened by the outcome of independent research, the more they will quote-unquote invest in their own,” said Celeste Monforton, a public health researcher and lecturer at George Washington University, who has written about corporate corruption of science. Monforton considers the petrochemical companies’ study of workers exposed to benzene in Shanghai to be the most expensive and elaborate effort by any industry to try to refute damning scientific evidence.

The reason, in her mind, is clear: “Litigation is continuing and potential for environmental exposures is still significant,” she said. “They need to protect their economic interests.”

Underwritten by the biggest names in petrochemicals — British Petroleum, Chevron, ConocoPhillips, ExxonMobil and Shell Chemical — and administered by the powerful API lobby, the Shanghai Health Study purported to examine how benzene exposure affects workers’ health. It consisted of three inquiries: The first investigated the link between benzene and non-Hodgkin’s lymphoma as well as acute myeloid leukemia, or AML; the second, progression of diseases caused by the chemical; and the third, the exposure level at which such biological markers as lower blood cell counts indicate benzene’s toxicity.

But the study’s outcome seemed to some like a foregone conclusion. Documents suggest oil companies set out to counter U.S. government research tying benzene to more types of cancer and at lower exposure levels than previously known. They show how company executives and scientists plotted objectives and “expected” results before the study began, banking on conclusions that would play down health hazards.

“This is just appalling,” said Carl Cranor, a philosophy professor at University of California, Riverside, who has read some of the Shanghai documents. “This does not sound like a scientific inquiry where you’re not sure what the outcome will be.”

Infante, the former OSHA official, who now testifies for plaintiffs in benzene litigation, put it more bluntly: “It’s called potential bias.”

Study’s Authors: No Bias

Industry representatives and the scientists they paid to do the Shanghai work say such criticism is unwarranted. Some oil executives, they say, may have been seeking an alternative to government research; others may have wanted to better understand benzene’s connection to disease. Whatever the impetus, they argue, scientific integrity was not compromised.

“There could have been the best intentions or the worst intentions,” said Harvey Checkoway, an epidemiology professor at University of California, San Diego, who served on a scientific review panel created by the petrochemical companies to review the study. “We set that aside for the research.”

Richard Irons, one of the study’s two principal investigators and now head of a consulting firm that does research for the petrochemical industry, said that “if you’re ignorant, it’s a logical conclusion” to view the work as biased. “But it’s an accusation not founded in fact.” Irons acknowledged that he has never testified for a plaintiff in a benzene exposure case. The API has financed his work on benzene since the early 1990s, documents show. Irons said he’s no longer receiving money from the institute but has gotten $100,000 for a small benzene project from the American Chemistry Council, the chemical industry’s main lobby group.

Defenders of the Shanghai study stress the independence of its design. Scientists, they say, have operated under the guidance of not just the scientific review panel, but of two Chinese government ministries and two university boards, all ensuring a proper inquiry. Many of the results — positive and negative, they say — have been published in peer-reviewed journals.

“The results don’t support the presumption of bias,” Irons said, explaining that, so far, the research has confirmed benzene’s association with AML as well as myelodysplastic syndrome, or MDS, a cancer of the bone marrow.

The study’s co-principal investigator, Otto Wong, who directed the work on AML and non-Hodgkin’s lymphoma, said, “I was interested in doing a cutting-edge study and I was confident I had control over our [portion]. The rest is really not a concern.” Wong acknowledged that he has never testified for a plaintiff in a benzene exposure case. His ties to the oil industry date to the 1970s. Now retired, Wong said he has “no contact with the API people at all.”

Representatives of BP, Chevron and ConocoPhillips all declined to comment for this article, referring questions to the API, which did not respond to repeated interview requests. In a one-paragraph statement, Shell said the company’s financial “support for the study reflects our ongoing commitment to health, safety, and product stewardship,” stressing that “the study was wholly independent of Shell.”

ExxonMobil, whose scientists participated in the Shanghai study, said in a written statement that it “supports scientific research through funding and technical support,” painting its involvement in this project — as well as a 2012 ExxonMobil-sponsored study of benzene-exposed workers showing adverse health effects at levels below legal occupational limits — as part of a longstanding corporate commitment to better understanding the chemical. The 2012 results in particular prompted ExxonMobil to “voluntarily reduce allowable benzene exposure limits to one-half of OSHA legal limits” at its workplaces, it said — or 0.5 ppm for an eight-hour shift.

By contrast, the National Institute for Occupational Safety and Health (NIOSH), part of the Centers for Disease Control and Prevention, recommends that workers limit their benzene exposure to an average of 0.1 ppm during a shift.

“Our conservative approach to setting workplace benzene exposure limits is influenced by the most up-to-date scientific evidence, which includes the 2012 study,” ExxonMobil said, noting that its short-term limit is five times lower than OSHA’s.

As the Shanghai findings seep into the scientific literature and, ultimately, the courts, the petrochemical industry generally admits that benzene causes AML and MDS at higher doses. But other blood and bone marrow cancers continue to kill — at lower and lower exposures to the chemical.

On February 27, 2012, Michael Boley, 68, died of a disease Shanghai researchers say can’t be tied to benzene: chronic myelomonocytic leukemia, a form of MDS, combined with another bone-marrow condition known as “myeloproliferative disease.”

Strong and industrious, with a knack for avoiding even the flu, Boley spent 23 years at a Goodyear Tire & Rubber plant in St. Marys, Ohio, site of a seminal benzene study by NIOSH in the 1970s. The research quantified for the first time the leukemia risk for workers exposed to the chemical in the plant’s Pliofilm unit, prompting OSHA to work on the benzene standard that took effect in 1987.

Boley didn’t make the benzene-soaked rubber film for which the unit was named; rather, he was a plant electrician, supervisor and engineer who worked there for an hour or two daily. Still, he knew benzene was in the air: While in the unit he noticed “Authorized Personnel Only” signs for Pliofilm workers who, he testified in a deposition, “were monitored on a regular basis.” At times, he complained about the unit’s “sweeter-than-gasoline” smell. Once, Boley asked a supervisor if he could have the same blood test the company administered to Pliofilm workers.

“His comment was, no, we wouldn’t require monitoring,” Boley testified. “Our levels would be safe.”

Four decades later, after enduring the fatigue, feebleness and shortness of breath accompanying a diagnosis of chronic monomyelocytic leukemia, or CMML, Boley sued Goodyear and its benzene suppliers, including ExxonMobil. “He wanted them to know what had happened to him,” said his widow, Cheryl. But the suit went nowhere: In 2011, Boley settled the litigation in a confidential mediation. His still-pending workers’ compensation claim has seen little traction.

“He couldn’t prove it,” Cheryl said, alluding to the companies’ claims that benzene couldn’t have caused her husband’s illness. Those claims were supported by Irons and other scientists affiliated with the Shanghai study, who reported in a journal article last year that “benzene exposure does not appear to be a significant predictor of CMML.”

Only Safe Level “Is Zero”

The petrochemical industry has known about benzene’s dangers since the turn of the last century. As far back as 1948, the API’s toxicological profile of the chemical discussed “reasonably well documented instances of the development of leukemia as a result of chronic benzene exposure,” cautioning that “the only absolutely safe concentration … is zero.”

Later, as scientific evidence of benzene’s hazards accumulated and regulatory limits on workplace and environmental levels tightened, the industry took a different stance. By 1990, the API and member companies such as BP, Chevron, Mobil and Shell had launched a research program meant to keep further restrictions at bay — or, minutes from an API meeting in 1992 state, research “that will be most useful in improving risk assessment and influencing regulation.”

Within years, the catalyst for the Shanghai Health Study appeared. In 1995, company representatives turned their attention to work by the National Cancer Institute, which was repeating the Pliofilm study in China to examine the effects on workers exposed to benzene at levels below the OSHA limit. Exxon, which had yet to merge with Mobil, even sent company scientists on a fact-finding visit to interview government researchers.

“We are monitoring the NCI studies,” an Exxon memo explained, “because of their potential impact concerning the health risks at low benzene exposures.”

In 1997, the NCI published a landmark study on benzene-exposed workers in Shanghai. The results reinforced past research showing the chemical causes leukemia, said Richard Hayes, a former NCI epidemiologist and the study’s lead author, but “what moved the science forward” were two findings: That workers with chronic benzene exposures had an increased risk of developing MDS and non-Hodgkin’s lymphoma — i.e., diseases other than leukemia — and that such effects could be triggered by doses of the chemical as low as the OSHA limit.

In 2004, the NCI released the results of a second study. It found that Chinese shoe makers inhaling benzene in amounts below the OSHA limit had fewer white blood cells than unexposed workers, suggesting the chemical has no safe threshold.

“In general,” Hayes said, “we found benzene was a larger problem than we originally thought.”

The industry quickly attacked the NCI’s work. Documents show that the API commissioned a $25,000 “critical review” of the government research from California epidemiologist Wong. In a 10-page paper, Wong challenged the NCI study from every conceivable angle. “The findings,” he wrote, “are unreliable.”

Wong insists today that his corporate funding had no influence on this conclusion. “My critique of the NCI study was comprehensive and specific,” he said. “I was responsible for every comment.”

For the industry, the review had the desired effect: It cast enough doubt on the NCI’s first study to convince the EPA, in 2000, not to rely on the research for estimating benzene’s carcinogenic effects. “We thought there were methodological issues that might be questionable,” said Bob Sonawane, of the EPA’s Office of Research and Development, who has overseen agency assessments of benzene’s health risks. The agency did use the NCI work to assess non-cancer effects in 2002, Sonawane said.

By then, industry representatives were already thinking beyond conventional critiques. Wong remembers reaching out to Chinese scientists about a possible benzene study before broaching it to industry contacts. “I knew quite a few people at API and member companies,” he said. “We just started a conversation.” The campaign to finance an alternative study in China was kicked off in earnest in the late 1990s, when the API approached Irons, then a pathology professor at University of Colorado, Boulder. Irons said API officials asked him to visit Shanghai in 1999 and consider doing a study similar to the NCI’s, which examined workers’ diseases and estimated their benzene exposures after the fact and “had some provocative findings.”

Irons went to China. Upon his return, he urged the API to instead conduct what he calls “a real-time clinical study,” in which researchers examine workers’ diseases as they occur. Within a year, he, Wong and ExxonMobil scientists had drafted proposals for the Shanghai study, which the API circulated among its members to drum up financing.

API representatives went from company to company, giving what amounted to a sales pitch for the Shanghai study. They laid out just what executives might anticipate in return.

A 2001 document listed the following “expected” results:

Provide strong scientific support for a lack of a risk of leukemia … at current ambient benzene concentrations to the general population.

Establish … current occupational exposure limits do not create a significant risk.

Refute the allegation that Non-Hodgkin’s lymphoma can be induced by benzene exposure.

Other documents show that the industry was counting on such findings to combat stringent regulation and stave off “tremendous” costs that would come from having to cut benzene emissions. “Significant issues of concern” identified in a 2001 PowerPoint include potential requirements to reformulate gasoline and “control emissions from stationary sources.”

Liability also was a worry. Documents warned of “litigation costs due to perceptions about the risks of even very low exposures to benzene” and lawsuits “alleging induction of various forms of leukemia and other hematopoietic diseases,” including more commonly diagnosed lymphomas.

For some in the industry, the bait proved enticing.

“Given the magnitude of [health, safety and environment] issues surrounding benzene as well as the litigation claims we continue to see, I believe it would be worthwhile to participate,” a Shell executive wrote in an email in 2000, a year before the company and its four counterparts formed the Shanghai study’s official sponsor: the Benzene Health Research Consortium.

Later, as consortium members tried to plug an ever-increasing budget gap — boosting the research’s price tag from $19 million in 2001 to more than $35 million in 2008 — their argument turned repeatedly to economics. One 2003 script for a CEO-level phone call states, “This study will positively impact our global business concerns.”

Critics say such documents expose the Shanghai study for what it is: An industry attempt to buy scientific evidence. “It’s all about influencing science to get what industry wants,” said Myron Mehlman, formerly chief toxicologist at Mobil, who became a whistleblower in 1989 after the company fired him for complaining about benzene levels in its gasoline. He sued Mobil, winning a $7 million judgment.

Mehlman remembers hearing about the Shanghai study in 2005 and immediately firing off letters to 45 executives at sponsoring companies. “I knew the scientists would do whatever it takes and whatever the industry needs done,” he said. In response, he said, he got a consortium form letter that “just re-confirmed how the study is being done for a single purpose — to get desirable outcomes.”

Industry-funded researchers bristle at the science-for-sale accusations.

“I didn’t see refuting anything as my charge,” said Irons, the study’s co-principal investigator, “and I wouldn’t have responded favorably to that.”

In depositions, however, Irons conceded that oil companies had a vested interest in the project.

“The oil companies … expected … it would be used for regulation, litigation, and for understanding the health effects of benzene,” he testified in 2010, adding that he didn’t believe a “funding source or the amount of money necessarily impacts on the [study’s] objectivity.”

Wong sounds a similar note. “We didn’t know what the results were before the study began,” he said, claiming he wasn’t privy to “any discussion” among the sponsoring companies. He considers it “just unthinkable” that critics would suggest “all those outside scientists, together with us, tried to create some results.”

Noting that the study relied on actual cases from 29 Shanghai hospitals, Wong explained, “It’s very difficult to argue that we have influenced our data one way or another.”

Members of the Shanghai study’s scientific review panel echo this sentiment. They saw no signs of overt bias in the design, they say, no way to yield preconceived results. “That wasn’t my experience,” said John Cherrie, a former panelist now heading occupational health research at a British nonprofit that has worked for the petrochemical industry. “The studies were designed to investigate the true situation without any obvious bias.”

Some would-be funders weren’t so sure, documents show.

In 2002, consortium members landed a meeting with seven scientists from Dow Chemical Company to pitch the Shanghai study. The meeting came after the scientists had voiced what one described as “specific technical concerns” about its design. Dow eventually opted not to contribute.

In a deposition two years later, Dow’s head of epidemiology, James Collins, testified that the company feared the study could generate inaccurate risk estimates and thus “be biased.”

“Independent” Review Panels

Documents suggest the Shanghai study’s sponsors were keenly aware of such perceptions. To deflect criticism, they set up “independent” review panels consisting of 10 ethicists and scientists, reputable leaders in fields like epidemiology, clinical medicine and bio-statistics. By 2001, panelists were meeting investigators and reviewing protocols — “essentially quality control,” said Jerry Rice, who chaired the scientific panel. Advisory in nature, the boards have remained involved in nearly every aspect of research.

Industry representatives viewed the boards as essential for lending credibility to the study. “There are going to be people out there who will want to misinterpret and criticize the study,” one argued in a 2003 email. “It is important that ‘the integrity’ … be maintained” with the panels.

Panelists insist the boards weren’t simply for show. They say industry representatives routinely encouraged them to offer criticisms and recommendations, and they obliged. Documents show investigators incorporated so many panel suggestions that research costs soared $3 million in one year alone.

“I never felt there was any desire to muzzle or tone down criticism,” said Rice, formerly with the World Health Organization’s International Agency for Research on Cancer, who oversaw the agency’s evaluations of chemicals for carcinogenic risks from 1996 to 2002. At IARC, Rice had come to know oil industry executives, who recruited him for the panel. “If there had been any of that, we’d have all quit.”

Still, company executives maintained tight control over the study. Documents indicate the consortium operated like a corporation, replete with committees governing research, finances and communications. Once a year, it hosted a meeting of 50 or so participants, flying American and Chinese scientists as well as review panelists to a two-day retreat to discuss the work as executives observed. Twice a year, consortium members combed through detailed progress reports filed by investigators.

The consortium also required industry review of draft manuscripts until 2005, when Irons penned his first. In it, Irons announced preliminary results linking a previously unrecognized form of myelodysplastic syndrome — MDS — to benzene exposure. The draft set off debate within ExxonMobil and Shell, both of which alerted the EPA — as the law required — to what was vaguely described in a consortium email as “health findings reported in a draft publication.”

At the time, Irons expressed reservations over the manuscript-review requirement.

Consortium members dropped it but kept the review boards. For years, panelists kept circulating “ready-for-submission manuscripts” for “feedback,” documents show.

Consortium members have worked especially hard on controlling their message. Before the Shanghai study began, the communications committee was crafting its marketing strategy. Documents reveal reams of “if asked” statements and media-relations plans, listing objectives such as “counter activists’ negativity.” Among the committee’s tasks, according to a 2001 email:

perception needs to be that this is not being done to protect against litigation

use a consulting attorney to address these issues of perceived motivation

Consortium sales pitches to prospective contributors, expressed in recruitment briefs, “call sheets” and “adaptable” slideshows, epitomize this corporate spin. Publicly, members have claimed altruistic motives for backing the Shanghai study. During a presentation in 2002 at PetroChina, a Chinese oil firm targeted for sponsorship, one Shell scientist proclaimed, according to a company email, “We believe it is important to understand the hazards of the products we make and sell, and we believe it is the right thing to do!”

Privately, consortium members betrayed less charitable motives. A draft of a 2002 recruitment brief reminded potential sponsors that “there is continued concern with the potential health effects of benzene as it relates to worker exposure … and personal injury claims.”

In response, a Shell executive urged colleagues to “delet[e] the reference to legal liabilities” and emphasized that “the only reason we are doing this is in support of protecting workers.”

Widows like Carolyn Wright have trouble processing such statements. Her husband, Eric, was a consummate company man whose closets brimmed with Shell awards, buckles, hats and magnets. Wright spent 34 years, from 1976 to 2010, working on offshore vessels for five oil companies — 18 at Shell. A trained “gun shooter,” he repaired sonar equipment, breaking down air guns. He soaked parts in benzene-laced solvents in a diesel-engine room, breathing in the chemical at exposures estimated to be two to five times greater than the legal limit.

Wright died, at 63, of leukemia in 2010, exactly five months after receiving an MDS diagnosis. Most of that time he remained hospitalized, beset by body sores, eye infections and a failing gall bladder. Within weeks, he couldn’t feed or bathe himself; eventually, he couldn’t hold up his head. Carolyn remembers leaving her husband’s bedside twice — first, to get clothes and then to visit a funeral home.

“They took away my best friend,” she said, referring to Shell and eight other petrochemical companies she and her husband sued 14 days before his death. “They were responsible because they made the benzene.” She settled the case for an undisclosed amount, most of which went toward her husband’s outstanding medical bills and legal fees. Shell admitted no liability.

“I’d rather have my husband,” Carolyn said. “There is not another one like him.”

Seeding the Literature

The petrochemical industry’s research strategy had another key component: publication. Oil and chemical companies have long seeded the scientific literature, paying consultants to publish in peer-reviewed journals. They often use the published articles to advance their positions in regulatory and legal arenas. Infante, the former OSHA official, considers benzene “a good example of how the general scientific literature is being polluted by people working for industry.”

A 2002 Shell summary of the Shanghai project defined a “Key Measure/Indicator of Success” as a “Cost effective study reported in public literature … [that would] support … advocacy.”

By the time the research consortium disbanded in 2009, Irons and his colleagues had released results in 20 journal articles as well as at an “international benzene symposium,” in Germany. Industry-funded scientists have added to the literature since; according to Irons, the consortium’s dollars have yielded a total of 30 papers.

Such publications can play a critical role in benzene litigation. The Shanghai study has helped companies deny liability by casting doubt on causation, the central issue in a toxic-tort lawsuit. One tactic the industry has employed — with aid from the study — is to separate cancers into subsets, making it harder to prove a specific link between benzene and the disease. David Eastmond, a toxicology professor at University of California, Riverside, said that the petrochemical industry “gets sued on a fairly regular basis” over “a wide range of diseases.”

“If [it] could narrow down which subtypes are caused by benzene,” he explained, “the industry could eliminate a number of lawsuits.”

Industry-funded Shanghai researchers found that only certain subtypes of MDS and AML are significantly linked to benzene.

Now, in claims involving auto mechanics, gas station attendants and printers, plaintiff’s lawyers are seeing this claim play out in court. There is the case, for example, of a contract worker for southeast Texas refineries and chemical plants who washed his tools in benzene and later developed MDS. Citing a 2009 article by Irons, defense lawyers argued that the worker didn’t have the subtype of MDS associated with benzene exposure. The worker’s family resolved the case under terms that remain confidential.

In another case, a refinery technician who used benzene in laboratory experiments developed the very MDS subtype the 2009 article tied to the chemical. In testimony, however, defense experts used the Shanghai research not to affirm this link, but rather to debate whether his MDS was, in fact, properly diagnosed. The case settled in 2010 as part of a confidential agreement.

Some find such defenses scientifically disingenuous. Hayes, the former NCI scientist, who is now head of epidemiology at New York University’s medical school, explained that scientists cannot eliminate a disease subtype simply because a study doesn’t show an association with benzene exposure. “It doesn’t mean there isn’t any effect,” he said. “The end result is to sow the seeds of doubt.”

Wong said that all he and his fellow researchers can do is report their findings — positive and negative. “I’m sure some people use the results in their own way,” he said, referring to defense lawyers. “I can’t really speak to that.”

Acknowledging that “no study can prove a negative,” Irons defends the Shanghai research as “the best available evidence” benzene causes only certain disease subtypes. That “doesn’t reduce the fact that benzene is associated with MDS,” he said — or, for that matter, AML. And ultimately, some say, these findings aren’t as conclusive as oil executives may have wanted.

“In terms of something that would once and forever cause the petroleum industry’s legal liability problems to go away,” said review panelist Rice, “it didn’t do that.”

In fact, some outcomes have hurt defendants in court. In 2004, after 30 years fixing refrigerators at ice-skating rinks and meat-packing plants — receiving a daily dose of petroleum-based products — Brian Milward developed a rare condition known as acute promyelocytic leukemia, or APL. At the time, he was 47 years old.

“You don’t want to find out you have cancer at that age because of somebody else’s wrongdoing,” Milward said, explaining he had no idea the solvents and paints he used to clean and seal pipes contained benzene until after he was diagnosed.

In 2007, he filed suit against Rustoleum Paints and 20 other manufacturers, whose experts argued no epidemiological studies show a link between benzene and APL. Experts hired by Milward countered that evidence linking benzene to AML essentially links it to all subtypes, including APL. The judge disagreed, ruling for the defendants. But the debate has stymied the case, fueling appeals on both sides, and sparking a brief signed by 27 preeminent scientists affirming benzene can cause any leukemia.

Meanwhile, the Shanghai study affirmed the very link Milward’s lawyers tried to draw. In a 2010 article confirming benzene’s tie to “an increased risk of AML,” Wong discussed the relationship by disease subtype. APL was the “most strongly related” to benzene exposure, he wrote.

While the ruling against Milward eventually was struck down, he has yet to appear before a jury. After settling with every defendant but Rustoleum, he has watched the company challenge the science again. This time, defense experts question whether his workplace exposures could have caused his cancer. A second appeal is pending.

Now in remission, Milward must grapple with cancer’s lasting effects. Nearly a decade of chemotherapy, along with diabetes and a rare bowel disorder, have left him battling what he calls “absolutely ridiculous” fatigue. Retired and on disability, he remembers returning to work twice. First, he resorted to napping to endure an eight-hour shift. When his boss assigned him to office duty, pushing paper and making calls, he still fell asleep at his desk.

“I can’t really do anything,” said Milward, 57 — at least, not what he loves: repairing race cars, working in his yard, playing with his grandchildren. “It just sucks when you get a cancer like this.”

“He Never Complained”

To family and friends in Deweyville, Texas, a pine-shaded town along the Louisiana border, John Thompson was “John” — a life-long resident, church deacon and carpentry teacher who reinforced souls as much as structures. To Irons, the Shanghai study investigator who charged $600 per hour as an expert witness, he was “John H. Thompson” — deceased, one of three case referrals from a corporate-defense firm, all in lawsuits alleging cancers caused by benzene.

Thompson had built scaffolding for refineries and chemical plants in the 1960s and early ’70s. At the end of most workdays he’d be covered in a thick adhesive known as mastic and would use pure benzene to get it off his hands, clothes and tools. In 2010, three months after Thompson’s death, Irons flew to Houston from China to testify in the benzene litigation, armed with what he has called “the largest single study of AML in history” —the Shanghai study. In a deposition, the scientist touted the research, saying it “inform[ed] with respect to disease specificity to a much greater degree than other previous epidemiological studies.”

“Assumptions and presumptions,” Irons testified, “have to be modified and re-assessed.”

The comment proved prescient in Thompson’s case.

In 2006, the former contract worker was mowing his lawn when he mentioned feeling fatigued. A pillar of a man, Thompson, then 66, had gotten two heart stents without uttering a gripe. “He never complained,” said Chase Bowers, his nephew. Not when he broke both arms on an oil rig, or lost his left eye in a shop accident. Bowers, who, as a teenager, was raised by Thompson and his wife, Carol, remembers watching a tree branch rip open his uncle’s ear. One emergency-room visit and Thompson returned to work.

“That’s the kind of guy he was,” Bowers said.

Within days of the lawn-mowing episode, Thompson learned he had acquired AML — specifically, a subtype of AML known as “inversion 16.”

In the contentious world of toxic-tort litigation, Thompson’s case seems like a classic: The only leukemia the petrochemical industry has admitted benzene can cause is AML — especially at higher doses. According to the Shell benzene-litigation defense guide, the industry might have classified Thompson’s case as “high risk”: He not only suffered from AML, but, as the guide states, “pure benzene is involved,” and his “exposure took place prior to OSHA involvement.” Experts hired by the family estimated Thompson’s cumulative benzene exposures were five times greater than the legal limit.

At trial, however, the case turned out to be anything but a slam dunk. Citing the Shanghai study, Irons worked to debunk a link between benzene and the type of AML afflicting Thompson. The Shanghai research has shown approximately two of the 20 AML subtypes are significantly associated with the chemical, he testified — neither of them “inversion 16.”

“We found many cases of inversion 16,” Irons said in court. “We did not find an association between benzene exposure and inversion 16.” Other scientists disagree, arguing that if benzene causes one subtype of AML, it likely causes all of them.

Keith Hyde, a Beaumont, Texas, lawyer who represents plaintiffs in toxic-tort cases, including the Thompsons, remembers Irons highlighting one Shanghai article after another at trial, raising doubt. “You have doubt here and there through all these industry studies,” Hyde said. “Let’s just say they do their job.”

After two weeks of testimony, the Thompson case ended in mistrial. The suit ultimately was resolved under terms that remain confidential.

In April, Carol, Thompson’s widow, died of liver cancer. In the years after the benzene trial, she rarely mentioned the ordeal — too many painful memories. She had to relive her husband’s death daily in court, watching videos and hearing testimony on his decline. “It was devastating for my aunt,” Bowers said. In his view, she never understood the extent to which oil and chemical companies hid benzene’s hazards and tried to spin the science.

“There’s obvious vested interest here,” Bowers said of the Shanghai study. “If oil companies are willing to spend $36 million to fund research, how much are they afraid of losing?”

Ashley Schwartz contributed to this story.