The Deadly Dinner Party: and Other Medical Detective Stories Read online

  The magnitude of the problem has been recognized in Massachusetts for some time. In May 1989, the Massachusetts Special Legislative Commission on Indoor Air Pollution issued a report recommending that regulations be strengthened and new legislation be enacted to protect people from these unseen hazards. One expert from Harvard estimated that indoor pollution is the largest public health problem we have and that it could be responsible for as much as 60 to 70 percent of all illnesses.

  Dr. Edward Banaszak and his colleagues put it more poetically in the introduction to their report in the New England Journal of Medicine. “In classical mythology, the tale of Pandora’s box describes how Pandora, in her curiosity, released the evils of the world from a box. The evils were carried forth on winds and breezes to all the world. This early reference to air pollution correlates strikingly with the discovery of hypersensitivity pneumonia due to exposure to a thermophilic actinomycete contaminating a central air-conditioning system.”

  Once Philip Bradford’s problem had been accurately diagnosed, Robert Rubin had one last mystery on his hands. Why hadn’t Bradford’s fellow office workers also become sick?

  “We took serum samples from thirty or forty other workers in the office to look for precipitins, and also inquired about a history of respiratory tract illness,” Rubin recalls. “We found five other workers with a high level of antibody similar to Bradford’s, with a negative response to the broad-class antigen at the School of Public Health, but a strong [reaction] to the specific antigen derived from the organism isolated from the ductwork. Four of these five co-workers had unexplained respiratory tract disease, which had been diagnosed as ‘recurrent pneumonia,’ ‘weak lungs,’ ‘smoker’s lung,’ and a variety of other terms.”

  Why didn’t more workers fall sick? The full answer to this mystery remains elusive, but other building-related clusters of hypersensitivity pneumonia have shown the same phenomenon. Among the possible explanations include an individual’s degree of exposure, and differing genetic predispositions, just as some people will develop an allergy to a given exposure while others do not. In the outbreak of hypersensitivity pneumonia from an office air-conditioning system reported by Banaszak, only four of twenty-seven workers developed symptoms, even though all had the same exposure to the air. Of the remaining twenty-three, whose lungs were perfectly normal on pulmonary function testing, eight had positive precipitin tests, but they remained entirely asymptomatic.

  After Rubin’s investigation, the building’s HVAC ducts were thoroughly cleaned, at great expense to its owners. And neither Philip Bradford, who worked in the building for four more years, nor any of his coworkers ever had another bout of the mysterious pneumonia again.

  “I am impressed how, in hindsight, we can casually dismiss the possibility of a malignant tumor in this patient. . . . It is remarkable that he escaped a needle biopsy or an open biopsy,” said Dr. Wilkins during a clinical case conference that was later held at the Massachusetts General Hospital. “If I get any credit as the patient’s first physician at this hospital, it is for picking a detective as well as a physician to take over this diagnostic problem.”

  8 Monday Morning Fever

  Early one afternoon at work, chemical engineer George Melville broke into a sweat, developed pains in his chest, and felt profoundly weak, especially in his thighs. He took himself at once to the emergency room at Lawrence General Hospital in northeastern Massachusetts. A fortytwo-year-old male smoker with chest discomfort triggered the usual concern for patient and doctor alike—could this be a heart attack? The usual tests that were available in 1973 were performed. His electrocardiogram (EKG), though strictly speaking abnormal, did not show the clear-cut changes usually found after a heart attack. Blood tests were done. These enzyme tests, which offer supporting evidence in heart attack cases, were also slightly abnormal, but not diagnostic. So far, all of the evidence pointing to a heart attack was circumstantial.

  Faced with this ambiguity, the attending physician took the prudent course of action: he ordered Melville (as I’ll call him) to be admitted to the intensive care unit for further testing. Over the next day, Melville’s doctors tracked down an older EKG taken years before. The “changes” that they were concerned about on his admission EKG had also been present on the earlier recording. And the blood enzyme tests quickly normalized in a pattern not suggesting a heart problem. Analyzing all the evidence that became available during the first two days of the case, his doctors were pretty sure that he did not have a heart attack. Melville also had a low-grade fever and some chills, and although these too are sometimes found after a heart attack, they suggested a whole host of other possible diagnoses as well. The prominent thigh weakness also did not fit with a heart attack. So other tests were done.

  This was before HMO medicine and bean counters’ scrutiny of “resource utilization.” In situations where a diagnosis was unclear, patients would often remain hospitalized simply to have diagnostic testing that in today’s medical environment would be undertaken on an outpatient basis. In 1973, Melville stayed in the hospital for two weeks. Even after two weeks of extensive testing, however, his doctors were baffled and the diagnosis remained elusive. Fortunately for Melville, he began feeling fine almost immediately after being admitted. And although they had excluded a heart attack as the cause of the symptoms, the doctors were troubled that they had not a clue about the true nature of his illness. Nevertheless, two weeks after he was admitted, they discharged him home feeling perfectly well.

  That might have been the end of the whole affair—except for one curious thing. In conversations around the water cooler, George Melville learned that several of his co-workers had suffered from remarkably similar symptoms. Most had some chest symptoms—usually cough— and many had low-grade fevers. One worker, a twenty-four-year-old supervisor, had three brief but distinct illnesses, each lasting only a day or two. Each episode was nearly identical. He would develop the abrupt onset of weakness, especially in the legs, and a cough, shortness of breath, and wheezing. Each occurrence started in the mid-morning. Twice he consulted his doctor. He too underwent multiple tests, even including a spinal tap; all the results were normal. Ultimately, he was referred to a neurosurgeon, who in turn suggested a psychiatric consultation. He had lots of doctors, but he had no diagnosis.

  Two other individuals at Melville’s place of work had similar symptoms that on some occasions were severe enough that they had to leave work. One forty-two-year-old woman was hospitalized twice with what was called a “severe viral illness,” which sometimes is doctor-speak for “we don’t know what’s wrong and can’t make a specific diagnosis.”

  George Melville was a manager of a division of Malden Mills, a textile plant with a venerable past situated on the Merrimack River in Lawrence, Massachusetts. The Merrimack, 110 miles long, snakes its way from its origin in Franklin, New Hampshire, to its mouth in Newburyport, Massachusetts, where it empties into the Atlantic. Its powerful current made it an ideal place for factories. In the nineteenth century, because the Merrimack River provided an unending source of power, numerous textile plants sprang up in the area. Malden Mills, which itself has a very colorful history, was one of them. The factory was founded in 1906 by Hungarian immigrant Henry Feuerstein, who worked his way up from sewing blouses in a New York sweatshop to a textile magnate. The mill’s name comes from the fact that the plant was originally located in Malden, Massachusetts.

  In 1956, Feuerstein moved Malden Mills to Lawrence, in a complex of red brick buildings along the river’s edge. The plant started out making wool sweaters and bathing suits. During World War II, it landed some large government contracts to make fabrics for military use. The textile factory housed all its processes, including dyeing, printing, and finish work, in different buildings on this complex. One early employee in the factory while it was still in Malden was Robert Frost, who grew up in Lawrence. He even wrote a poem, “A Lone Striker,” in 1933, that harkens back to his work in the old textile plants.
/>   Like many factories that dot the Merrimack River, Malden Mills houses several of its divisions in brick buildings with large mullioned windows. In its heyday, the plant employed more than three thousand workers. In the particular building where Melville worked, he and thirteen other employees produced imitation crushed velvet, in a process called flocking. The product is one that everyone is more familiar with when it is called velour, which is used in clothing, automobile and home furniture seat covers, novelty items, and other common household accoutrements.

  Flock is the industry term for the short fibers that are cut from longer cables of synthetic monofilaments (called tow) that are applied (flocked) onto an adhesive-coated material. In the case of the imitation crushed velvet, the fibers were made of nylon. Because Melville suspected that his illness might have been related to the worksite, he made a telephone call.

  Voltaire wrote, “Work keeps us from three great evils: boredom, vice, and need.” He was only partly right. Work itself can be an evil; it carries health risks of its own, whether you are a lumberjack in the Pacific Northwest, a fisherman on the Grand Banks off Nova Scotia, or a stock trader on Wall Street.

  In 1988, according to the Bureau of Labor Statistics, the private sector reported 6.2 million work-related accidents and nearly a quarter million non-traumatic occupational illnesses, figures that represented substantial increases over previous years. These types of studies often underestimate the frequency of work-related problems. First of all, some occupational illnesses are not detected or recognized as such. Also, exacerbations of chronic illnesses and preexisting diseases by hazardous exposures at work are often not recognized or reported. Furthermore, surveys may not include small businesses, which, combined, employ large numbers of workers. One statistic that is less ambiguous is jobrelated deaths; there were 3,270 of them in 1988.

  Occupational medicine was largely ignored by the medical community until the 1980s. Over the past decade, demand for expertise has risen sharply because of increasing governmental regulation, litigation stemming from toxic exposures, and mounting public recognition of environmental risks. Two federal agencies that were created in 1970, the National Institute for Occupational Safety and Health (NIOSH) and the Occupational Safety and Health Administration (OSHA), police American industry. NIOSH is the research and advisory branch, investigating potential work-related disease outbreaks and performing industry-wide surveys. It advises OSHA, which writes the standards, inspects workplaces for compliance, and enforces the standards by imposing fines.

  When Melville fell sick in 1973, however, expertise was hard to come by. Fortunately, he lived in an area where help was just a phone call away. Harvard’s School of Public Health, he thought, would surely be able to help find an explanation.

  “I received a call one day from this man, asking us to investigate what he thought was a problem with his plant,” recalls John Peters, then an associate professor of occupational medicine at the Harvard School of Public Health (and now a professor of occupational and environmental medicine at the University of Southern California School of Medicine). “He [Melville] told me about his hospitalization and then about a couple of other people with vague symptoms, one of whom had been hospitalized twice for what was thought to be a severe viral illness,” recalls Peters. “Another guy had a couple of attacks which he had attributed to the work setting. On that basis, he wanted me to get involved. So I did.

  “I visited the textile mill to figure out the process, understand all the elements and possible exposures.” This is a standard occupational health approach to a possible outbreak of disease in the workplace. The investigators generally follow a three-step plan. First, understand the details of the process—the workflow. Often a change has been instituted recently that may seem very minor to the employees or to the bosses, so it goes unnoticed. Second, using this detailed analysis of the workflow process, generate a list of possible exposures. Last, interview the workers and tabulate the information, in standard epidemiologist fashion, to see if any clues fall out.

  So this is what Peters, and his colleague Dr. David Wegman, did. They talked with the thirteen men and women, ages nineteen to sixtyone, who worked with Melville. Some had been employed in the velvet flocking division for as long as six years. The investigators learned from the managers of Malden Mills, who were very cooperative, that the process for making the crushed velvet had been changed recently, and that the workers’ health complaints had started after the introduction of the new manufacturing process.

  The odd thing was that only seven of the thirteen workers were experiencing problems. Dr. Peters conducted detailed interviews with all of them. A prevailing story emerged from these discussions.

  “It was so standard,” he says, “that I thought these people must have been talking to one another and rehearsing. Of course, they hadn’t been. They would report to work feeling fine, and then they would develop a typical sequence of symptoms.” Most had a cough with fever or chills. Many complained of weakness and achy pains throughout their body; a few suffered from shortness of breath. Melville had experienced chest pain. The symptoms would develop at work and last an average of twenty-four hours. The workers would never get the symptoms on weekends.

  In this kind of analysis, what does not happen can be just as important as what did happen, like the dog that didn’t bark in the Sherlock Holmes story The Hound of the Baskervilles. Why did six of the workers never get sick? The reason could not be explained on the basis of age or sex or any other obvious variables. They all worked in the same division in the same building and therefore had the same exposures. So Peters examined the manufacturing process in detail, hoping to find a clue about why some workers were affected and others were not.

  In the flocking process at the plant, nylon fibers were first dyed and then cut into short strips measuring about an eighth of an inch. Next, they were treated with a clay slurry to facilitate their forming a velvety pattern when fixed to a rayon cloth. The flocking step involved treating the cloth with an adhesive and a formaldehyde resin to fix the nylon fibers onto the rayon surface. The material was then cured at 300 – 330 degrees Fahrenheit, using ammonia to stabilize the acidity.

  After flocking, some of the material, about 20 percent, was used in a different process that involved different employees, so this part of the routine fell outside the investigators’ analysis. The remaining 80 percent was treated with a fluorocarbon polymer to make it stain resistant and waterproof. Next it was dipped, rolled, and squeezed, then cured at 300 – 310 degrees, and finally crushed and steamed at 275 degrees. Last, the crushed velvet was cooled, back-steamed at 190 degrees, and then treated to eliminate static electricity. This last step could potentially produce ozone.

  “So there were several potential exposures,” says Peters, “but there were problems with all of them. We looked for ozone but didn’t find any. And both the ammonia and the formaldehyde were present in too low a concentration to be toxic. Plus, the nylon fibers were too large to be inhaled into the lungs.” That left the fluorocarbon polymer. There were several hints that it might be the source of the problem. For one, none of the workers who handled the cloth that had not been treated with the polymer ever got sick. Second, the rest of the workers rotated jobs, and these workers got sick only when they did handle the polymer-treated cloth. And last, Peters was aware of a syndrome called polymer-fume fever, an illness that matched the textile workers’ symptoms perfectly.

  The word “polymer” derives from the Greek poly (many) and meros (part). Polymers consist of simple substances linked chemically in long chains, with the resulting structure having properties different from those of the individual components. The polymer used in the crushed velvet flocking process is called polytetrafluoroethylene, mercifully referred to as PTFE.

  The history of PTFE in itself is remarkable. Like many other important discoveries, PTFE was stumbled upon completely by accident. A research chemist named Roy Plunkett, working for the DuPont Company in the late 19
30s, was searching for a better refrigerant. Early refrigerants, such as ammonia and sulfur dioxide, regularly poisoned food-industry workers and even people in their homes. In his search for a safer method, he was working with tetrafluoroethyelene (TFE). At normal temperatures, TFE is a gas, and Plunkett had prepared hundred-pound cylinders of TFE in preparation for adding chlorine. To keep the gas stable overnight, he froze some of the metal tanks in dry ice. The next morning, April 6, 1938, he opened one of the tanks, but nothing came out.

  Thinking that the gas must have escaped, he weighed the tank to test that hypothesis (since a full tank would weigh more than an empty tank). The weight showed that none of the gas had escaped. Plunkett and his assistant sawed open the tank and found that it contained a white waxy solid that on further analysis turned out to be polymerized TFE—or PTFE. Curious about this new substance, Plunkett experimented with it further and found that it had amazing heat resistance and was chemically inert; that is, it would not react with other substances. In addition, one of its most important qualities was its low surface friction, so low that almost nothing stuck to it. The first scientific report on PTFE was published in 1941.

  For all of these reasons, Plunkett and DuPont realized that there were probably some commercial uses for PTFE. They began using it in many other applications, but one of the first general uses for it was in 1946 when it was used to coat cookware to prevent food from sticking. The company needed a trade name and came up with a rather catchy one: Teflon.

  Within ten years of the 1941 report however, the first cases of toxicity surfaced. D. Kenwin Harris was a divisional medical officer in the plastics industry in England. He wrote an article that was published in the medical journal The Lancet, titled “Polymer-Fume Fever,” in 1951. This terse report contains quite a bit of information that had a direct bearing on the investigation at Malden Mills.