We Were There – Diacetyl

Just another WordPress site

We Were There – Diacetyl

[ Beeping ] [ Music ] >> Thank you so much for joining us for the ninth We Were There It, Not everything is better with butter flavoring: popcorn lung disease, and the dangers of diacetyl This is a fascinating story of phenomenal epi work with a little bit of luck About the location and a lot of fantastic lab work to back up the story Without further ado, it’s my pleasure to introduce Dr. Anne Schuchat, the Deputy Director for CDC Thank you [ Applause ] >> Thanks so much, Phoebe, and welcome everybody to today’s installment of We Were There There’s probably no symptom more concerning than the difficulty breathing My, probably one of the most memorable experiences I had as a intern Newly in the hospital on my first night on call in internal medicine Was going into the intensive care unit to follow a patient who was fighting for life Very, very difficult to ventilate Today he would be waiting for a lung transplant But in that evening, my job was just to try to keep him alive through the night As I was looking at the background for today’s installment I realized that even that individual could have potentially gotten that devastating disease through his job I think we take for granted that our jobs don’t make us sick But as we learn the mystery that was uncovered in the, in the sleuthing that was done on this front We realize that even today there are potential work-related serious conditions I don’t want to go through the story because it’s a very good one But in this cold weather where everybody’s huddling with their Netflix stream and eating their popcorn This is a kind of a cautionary tale So without further ado, I’d like to introduce our fabulous lineup of all-women speakers today And welcome to We Were There >> Thank you, Dr. Schuchat And without further ado, I’d like to introduce Dr. Kathleen Kreiss, our first speaker and Disease Detective >> Thank you, Phoebe Our story is how we discovered that diacetyl, which is the principle component of butter flavoring is not safe to breathe Despite the Food and Drug Administration’s assurance that it’s safe to eat The backstory is instructive In 1999, a 51-year-old worker at a Missouri microwave popcorn plant developed disabling skin disease And his fissured, bleeding hands and feet precluded him from working He lost a Worker’s Compensation claim for occupational skin disease At which point he acknowledged that he couldn’t breathe either His wife knew of four of her husband’s former coworkers who had respiratory difficulties And she suspected that the cause of their disease was occupational She gave names and contact information to her lawyer’s son The lawyer had no experience with work-related claims And he passed the information to a Worker’s Compensation lawyer in Kansas City two hours away By interviewing the five former workers with lung problems, he found three more workers with lung complaints And passed medical records for the eight former workers To an occupational medicine physician The eight former workers reported cough and progressive shortness of breath with any exertion They had developed symptoms at ages 27 to 51

Which was much earlier than customary for smokers at risk for emphysema or chronic obstructive pulmonary disease In any case, five of the eight had never smoked or had smoked very little The respiratory problems dated from 1992 to 2000 with no common accidental exposure at work Their symptoms started five months to nine years into their employment at the microwave popcorn plant None of them thought that they had a work-related cause of disease Because they had no pattern of symptoms that was worse at work Or better on vacation Or better after leaving work because of the respiratory disability The occupational physician, Dr. Allen Parmet noted that four of the eight former workers were listed for lung transplant Which was unusual coming from a small, rural community in southwest Missouri Who shared a common last employment at a microwave popcorn plant Although they had a variety of diagnoses, asthma, emphysema, bronchitis Their difficulty getting air out of their lungs and no response to asthma medications was suggestive of bronchiolitis obliterans Medical textbooks describe bronchiolitis obliterans or BO as a rare lung disease Workers with BO have a single, overwhelming exposure to a lung toxin such as nitrogen oxides in silos If they survive this respiratory emergency, they seem to recover completely Then several weeks later, they develop difficulty blowing air out of their lungs That is obstruction to airflow No medications helped their breathing difficulties Unlike asthma and chronic obstructive lung disease And for that reason, lung transplant appears to be the only option when their obstruction on breathing tests is bad enough Chest x-rays and diffusing capacity which reflects the lungs’ ability to transfer gases are not helpful in diagnosis None of the eight popcorn plants, cases reported such a textbook history But Dr. Parmet knew that the cases didn’t have asthma or smoking-related lung disease He reported a cluster of suspect bronchiolitis obliterans to a sanitarian at the Missouri Department of Health and Social Services And that sanitarian called CDC in May of 2000 CDC referred the report to Hector Ortega, EIS Officer in the Division of Respiratory Disease Studies in Morgantown, West Virginia At the National Institute for Occupational Safety and Health We told the sanitarian to obtain the medical records to confirm the diagnoses Dr. Ortega and I called back to the sanitarian at the Missouri Health Department every month to check on the medical records’ findings which never surfaced We eventually contacted the new state epidemiologist, Dr. Eduardo Samos And within three days, he had reviewed the records and requested our assistance In investigating under the state’s authority Spirometry, a type of breathing test can document normal and abnormal patterns but can’t differentiate among diagnoses A person with healthy lungs can take a deep breath and blast it out quickly, as indicated in the normal scarlet slide, line in the slide Persons with emphysema, asthma, or bronchiolitis obliterans have difficulty blowing out rapidly as shown in the two curves to your left They have obstruction to airflow which results in a scooped pattern They often trap air in the lungs which results in greater lung volumes in these plots of airflow against volume The other breathing test pattern’s called restriction And this pattern reflects fibrosis That is lung scarring When a person with scarred, inflamed lungs takes a deep breath, they can’t get air into their small lungs because they’re stretching their scars But they can get air out quickly So although textbooks at the time said that bronchiolitis obliterans had an obstructive breathing pattern We now know that BO can have obstructive, restrictive, or normal breathing tests This is the microwave popcorn plant at which the eight former workers worked Kernel popcorn is delivered to this plant, and workers package it in polyethylene bags And in microwave popcorn bags Inside the plant, there was a large room with many lines in which a machine operator, shown by the circle Filled microwave bags

with popcorn kernels and flavored oil Which came from the holding tanks on the mezzanine above them The bags were sealed and placed in boxes for shipping in the production area This large packaging area also had a small quality control room Where three QC workers popped the bags to test for unpopped kernels and salt content Here you can see the popcorn being put into the bag by a machine Overseen by machine operators on each production line And this is the mixing room which opened to the packaging area For shelf stability in grocery stores, microwave popcorn needs a solid fat with flavorings at room temperature Artificial butter flavorings have many chemicals in proprietary recipes The predominant chemical is diacetyl Which is a four-carbon dicarbonyl It’s the same com– compound that’s found in real butter But it’s much more concentrated One mixer per shift dumped microfine salt and heated liquid butter flavoring into a heated tank of soybean oil Which was then pumped up to the mezzanine holding tanks Which you saw above the packaging lines The holding tanks fed the heated mixture of oil, flavorings, and salt that was squirted into the microwave bags of kernel popcorn The Missouri State EIS Officer, Joe Malone, and NIOSH EIS Officer, Rich Canjois Reported the first hint of risk in the factory Of the eight cases among former workers, four were mixers of butter flavoring and heated oil When we went to the factory, we found there was only one mixer working on each shift Obviously they had very high risk Since we estimated that there had only been about 13 former mixers since 1992 When the first Sentinel case became symptomatic The other four workers worked in microwave packaging And there were no cases seen among workers who packaged unflavored popcorn or worked in offices or the warehouse In contrast to the formal worker findings, none of the current mixers had abnormalities in our cross-sectional evaluation of 117 workers Among the 135 employees This likely reflects a healthy worker effect In which ill workers leave employment Nevertheless The employees had a burden of respiratory illness compared to the National Health and Nutrition examination survey of the U.S. population The prevalences of respiratory symptoms and respiratory diagnoses were two to three times that expected About a quarter of the workers had abnormal breathing tests, that is barometry The prevalence of the patterns barometry that we thought at the time was characteristic of bronchiolitis obliterans Was three times that expected And the abnormality did not respond to asthma medications Three workers had severe abnormalities in breathing tests Putting them into the range of the former workers listed for lung transplants Other screening tests such as x-rays and a measure of gas exchange were normal We concluded that there was a lot of unrecognized, work-related lung disease among current workers We have measured the average diacetyl exposures for each job category And we then estimated cumulative exposure for each worker on the bases of the average exposure And length of time in each job he or she held Here you can see that workers in the highest quartile of cumulative exposure had higher proportions of abnormal spirometry than other quartiles There was an exposure-response relationship Between flavoring exposure as reflected by diacetyl And risk of abnormality So collectively, we had satisfied several of the criteria listed by Sir Austin Bradford Hill for a causal relationship between flavoring exposure and occupational lung disease Exposure preceded the disease in the former sentinel cases The magnitude of risk among current workers was high With never-smokers having a 10.8-fold risk of abnormal breathing tests, compared to the U.S. population And exposure-response relationship existed And alternative explanations were unlikely for a rare disease In the meantime, our team was investigating other microwave popcorn plants And four of five other plants had cases consistent with bronchiolitis obliterans Especially in mixers This fulfilled the criterion of replication of our findings in other populations With this documentation that flavoring exposure risk existed,

our NIOSH engineers and industrial hygienists helped the plant to control flavoring exposures The company isolated the mixing room with airlock doors And the mezzanine holding tanks from the packaging area Flavoring and holding tanks were exhausted, temperatures of heated tanks were lowered And ventilation for dilution was increased Flavoring transfers were closed Instead of workers having to manually add flavors to the mixing tank And mixers began wearing supplied air respirators Exposures to diacetyl decreased As a result of these interventions over three years, average diacetyl concentrations fell on this logarithmic scale for these three job categories Mixers, machine operators on the packaging lines, and quality control workers who popped 100 microwave popcorn bags per shift Mixer exposures dropped from about 38 parts per million to less than one part per million In packaging and quality control, exposures at our last survey in 2003, were all below detectible limits Below about .005 parts per million Mixers still needed supplied air respiratory protection In that mixing room concentrations exceeded those associated with prior cases in packaging and quality control Although no current mixer had abnormal breathing tests in the initial cross-sectional survey Five months later, a 23-year-old mixer had had a dramatic fall of 1300 milliliters in the volume of air that he could blow into a spirometer in the first second of the FEV1 In the second follow-up survey, five months after that, he had lost an additional 700 milliliters For a total loss of two liters in ten months And he decided to leave employment at the popcorn plant His breathing tests then stabilized somewhat over the remaining surveys But he was left with severe lung disease This worker demonstrated that frequent medical screening with breathing tests in flavoring-exposed workplaces could identify those with rapid declines in their breathing test measurements Hopefully before workers became disabled These rapid decliners could be relocated to lower, or no-exposure positions And also indicated where more protective measures were needed As flavoring exposures were reduced in this microwave popcorn plant, our medical screening with breathing tests showed that annualized declines in the breathing test results, the forced expired air volume in one second Declined among the 27 workers who participated in all eight cross-sectional surveys In 2001, the first year of follow-up, average FEV1 fell by 142 milliliters In 2002, FEV1 fell in these same individuals by 46 milliliters In 2003, the third year of follow-up, the FEV1 fall was 22 milliliters And this last average decline per year is what we would expect in an unexposed population By reducing flavoring exposures, the company interrupted the average airway damage reflected in breathing tests decline in this plant Of course medical screening identifies individuals who require secondary prevention by decreasing exposure Or medical removal Ultimately, medical screening and exposure control requires standard setting with a permissible exposure limit promulgated by the regulatory agency, the Occupational Safety and Health Administration in the Department of Labor With dissemination of the microwave popcorn story, the chest physicians of California reported diagnoses of bronchiolitis obliterans In two workers from flavoring manufacturing plants in California in 2004 and 2006 Unlike the microwave popcorn industry exposures Flavoring workers may be exposed to about 2000 flavoring chemicals in batch operations That differ from hour to hour, and from day to day This variability in chemical exposures makes exposure characterization nearly impossible However, diacetyl is used to make many flavorings such as nut and fruit flavors Since flavors are generally regarded as safe to ingest by the Food and Drug Administration The flavoring industry had not considered chemical breathing hazards As shown in this picture of a worker weighing a poured ingredient in open containers

And no respiratory protection In response to the case reports, the California Department of Health and Cal/OSHA Partnered in a unique, preventive program for workers in the flavoring industry from 2006 to 2009 Sixteen companies reported identifiable medical screening data to the Health Department in exchange for preventive consultation by the Cal/OSHA regulatory agency And avoidance of compliance fines Thomas Kim, the California EIS Officer, analyzed the aggregate breathing tests from the first cross-sectional screening reported by the companies He showed that the four companies using the most diacetyl per year each had four cases with abnormal breathing tests Consistent with bronchiolitis obliterans Because they were unresponsive to asthma medication In the California flavoring industry, more cases developed during longitudinal screening And the same four companies with four-person clusters of lung disease Had workers with excessive declines in their breathing tests with follow-up medical screening In both the sentinel plant and the flavoring plants in California, some cases had onset of chest symptoms ten years before these investigations In 1985, NIOSH had investigated a cluster of bronchiolitis obliterans in a flavoring plant Making flavorings, flavors for the baking industry In 1993 to 1996, Jim Lockey from the University of Cincinnati had studied a cluster of bronchiolitis obliterans in the flavoring plant with which he had a confidentiality agreement What were the barriers to recognition of the diacetyl risk in the 15 years before our popcorn plant investigation? Well, physicians usually diagnosed patients with cough, wheezing, shortness of breath with common illnesses such as asthma, bronchitis, and chronic obstructive lung disease The latter especially in people who have ever smoked Even most chest physicians have never seen a patient with bronchiolitis obliterans And if so, it was a textbook presentation of an overwhelming noxious exposure followed by recovery And then development of untreatable difficulty getting air out on exhaling Individual cases could only be attributed to preceding exposure if he or she had been healthy before surviving dramatic pulmonary edema Following an accident No one had reported indolent onset, onset of bronchiolitis obliterans from ongoing exposure With cases occurring one by one over several years Even for a rare disease, attribution of individual cases to an exposure is not possible without epidemiologic risk factors And as you know, epi investigation requires a cluster of disease in a defined population such as a workforce The clusters in two flavoring workforces before the microwave popcorn plant investigation were not epidemiologically investigated But the hundreds or thousands of chemicals in flavoring plants precluded identifying an unknown chemical cause The eight sentinel cases in the microwave popcorn plant investigation occurred in a rural factory outside of Jasper, Missouri There were two chest physicians locally who saw all the cases Referred several of them to national referral centers Which sometimes missed the diagnosis And knew of their common employment One of them contacted OSHA which reported that there were no respiratory hazards in the plant OSHA enforces regulations But no regulations exist for unknown hazards A public health agency would have been a better contact, as known by Dr. Parmet who contacted the Missouri Department of Health and Social Services The sanitarian was not likely to know what suspected bronchiolitis obliterans was And did the right thing calling CDC Luckily, the Atlanta respondent called NIOSH and specifically the Division of Respiratory Disease Studies There we knew that eight cases in a small food production factory with four severe enough to be listed for a lung transplant was something unknown We wanted to investigate but needed a Health Department invitation Since our only right of entry at NIOSH is a request by three current workers

A labor union, or management We had the tools, questionnaires, breathing tests, chest x-rays and exposure assessment expertise To conduct on-site characterization of the plant population We recognized that a population with 26% abnormal breathing tests and excess chest symptoms was anomalous And had ongoing risk Thus, we could establish epidemiological risk factors such as diacetyl exposure And rule out smoking as the cause Compared to flavoring manufacture, we were only dealing with components of butter flavoring And exposure of volatiles for specific job categories were the same over an eight-hour shift Although the predominant volatile was diacetyl We could only say initially that it was a marker of the causative agent To fulfill the last of Sir Austin Bradford Hill’s criteria for causal inference from epidemiological associations, we needed to demonstrate biologic plausibility from toxicologic studies And that is why we turned to Ann Hubbs who will share an exciting story of her continuing investigations of a brand-new area in toxicology And with that, I’d like to introduce Ann Who is a Research Veterinary Medical Officer at NIOSH in Morgantown, West Virginia >> Thank you [ Applause ] So in the Health Effects Laboratory Division, we look at causes of disease very often And this was not just my work I’m just up here speaking So there were many people involved in this work Into the biological plausibility from popcorn workers’ lung and its molecular mechanisms So butter flavoring vapors were originally associated with disease in workers But we thought the actual cause, not a marker, would be diacetyl Let me explain why This is a GC mass spec of the vapors So we’re talking the, the gas phase Of artificial butter flavoring Here we’re looking at the abundance of a series of molecules And as they come off, with time, from the GC mass spec And these are the names of some of the many compounds that make up those complex vapors But we were interested and concerned with diacetyl because one it was an agent of unknown toxicity, and we knew the safety or toxicity of some of the others And the other thing was its chemical structure told us it could be a significant risk So it was immediately a suspect and here’s why This is the structure of diacetyl It’s a four-carbon compound On each side is a methyl group, something that’s not going to be particularly reactive But in the middle of the molecule are two carbonyl carbons Double bonds here between carbon and oxygen And they’re adjacent carbon molecules And what that allows the electrons to do is to move around between those two carbon groups So I don’t like to show diacetyl this way Let’s show it the way the body sees it With its dancing electrons So the electrons move around And in moving around, they can react with other structures And biological molecules interact with diacetyl So it would likely be a reactive compound It could be causing problem But would it target the airways? Because these workers largely had airway disease So we looked at what happened in the airways of rats that were exposed to butter flavoring vapors or diacetyl And we saw a common pattern emerge When they were exposed to butter flavoring vapors, or diacetyl, we saw the death of the cells that lined the airways And so if the exposure was sufficient, there was significant damage And we also knew like any other structure

When you damage the lining of an airway, it will scar Biologists call this fibrosis We often talk about pulmonary fibrosis when it’s out in the deep lung But the airways can also scar And that gives us fixed airways disease What this looks like to someone like me who is a pathologist Is that instead of this wonderful carpet of epithelium that lines the airway We instead see in the diacetyl-exposed airways that this carpet has been disrupted And so rather than having intact cells, we see cell loss And we see the products of cell necrosis and the response to that out here in the lining And we would expect a scar to develop And a scarred airway is fixed Colleagues at the National Institute for Environmental Health Sciences in Duke did repeated exposures to diacetyl And they found that indeed that was the case So with repeated exposures to diacetyl And this is the image from the airway of a rat We see this fibrous connective tissue here That is part of a scar And in fact this polypoid structure’s actually healed back to the airway, obstructing it But it wasn’t just diacetyl It was a related compound that caused similar damage With sufficient exposure This is another butter flavoring vapor This is. And here this should 2,3-pentanedione be an airway But in fact this trichrome stain is showing early fibrosis here In what should be the airway passage And this is very much an obliterative lesion Hence, bronchiolitis obliterans So we have these dancing electrons And we asked why is the airway epithelium damaged by diacetyl in 2,3-pentanedione? In its simplest form, it is just that both of these compounds share this reactive alpha dicarbonyl group And in diacetyl, there is one carbon here And in 2,3-pentanedione there are two So these are extremely similar structures But what is going on? They’re reactive They’re both alpha dicarbonyl compounds But really importantly, they react with amino acids, the building block of protein So go where the data takes you And the data tells us the hypothesis might have something to do with that And that protein damage underlies airway epithelial toxicity of diacetyl So why would we care about protein? Well, either one of them can damage protein function It’s not just they’re damaging those proteins But the proteins don’t act right So colleagues at the National Institute for Environmental Health Sciences in Duke University looked at the activity of some representative proteins Enzyme activity And exposure time in minutes to diacetyl or 2,3-pentanedione So looking at how much activity was left with time And in fact, both of them caused loss of protein function But okay, protein? Really? Why is protein damage important? That’s because protein is what determines everything, for the most part That happens in the cell So this is a cell And we all know genes are important, right? That’s why we got the genes we got from our parents And that is because of DNA But actually, I’m going to tell you that DNA is mostly important after you’re born Because it produces protein So DNA transcribes to RNA and that’s translated into protein And that’s why cells do what cells do, for the most part But there’s another side to proteins It’s not all about your genes and us thinking that DNA is important Protein quality has to be controlled at the other side, too We have to have control of the quality of proteins And that’s where research has been exploding in the past few decades So when a protein is damaged or if it’s a normal protein And it’s supposed to end its life What is supposed to happen is that protein is recognized

by a molecule called ubiquitin Sometimes more than one molecule and mostly at a specific site called K48 ubiquitin And K48 ubiquitin, sometimes multiple molecules like that Here as little orange balls, the color of butter They then go to a structure called the protyism And there they get recycled And what ends up happening is you get the amino acids back What a wonderful system But if we have too many damaged proteins there Something else can kick in And that is a process that’s mediated by K63-linked ubiquitin Usually multiple molecules this time linked as K63 ubiquitin which actually refers to the lysine group And then there’s another very important protein Which is sequestosome 1 Sequestosome 1 makes a scaffold that sequesters those damaged proteins So that they don’t have their toxic effects in the cell So it sequesters them in air and then you’ve got to get rid of them So you send them to the lysosome And this is a process called autophagy And it’s also really, really important It controls getting rid of those toxic proteins you can’t get rid of by recycling And it also has a number of other biological processes But instead of recycling, it’s throwing it out in the trash So our approach was to look at ubiquitin in autophagy after diacetyl exposure Because diacetyl damages proteins And here’s how And this is why it’s important Damaged proteins are toxic to cells Ubiquitin is the central mediator of protein quality control And the Nobel Prize was awarded for its discovery in 2004 This is very important And ubiquitin depletion is toxic Because you dysregulate all the proteins The cell doesn’t work right Some proteins live too long Damaged proteins accumulate Amino acids are depleted And ubiquitin identifies proteins that are no longer needed or damaged And that’s the big summary It’s important in controlling lifespan in normal proteins and damaged, toxic ones Proteins are recycled into their constituent amino acids in the proteasome And in the presence of overwhelming protein damage, ubiquitin can send damaged proteins for destruction in the lysosome As a process known as autophagy for which the 2016 Nobel Prize was awarded So when we look at an airway And in this case, we have stained the airway epithelium so that it is green And you can see the, the cell junctions here in bright green And we’ve also stained for ubiquitin And you really don’t see anything It’s kind of a pink haze because we stain for ubiquitin in red In this control airway from a rat Actually, this is a mouse But in the diacetyl-exposed airway from a mouse Instead, what you see is still the airway epithelium It’s a little damaged and shriveled up here But what you see in red is that ubiquitin has gathered together in discrete areas of the cell So we wondered if protein quality control had actually been disrupted And so we wanted to look at the scaffolding protein The sequestosome 1, that sequesters those damaged proteins Maybe it was doing something And this is again the airway of a diacetyl-exposed mouse Here we’ve stained in green sequestosome 1 And we’re looking at it under green fluorescent light If we look at this same airway section, this is the lining epithelium in red This is ubiquitin Now if they’re both in the same spot, we’re going to see red and green at the same time And that’s going to give us yellow There’s the yellow! The little ubiquitin is lining up right along the scaffold in, in autophagy-related process So we wanted to see hey, what happened to the lysosomes? That’s where your, the ubiquitin’s supposed to go in autophagy And so here’s the airway with the lysosomes in green Again, we’ve stained ubiquitin in red And now we’re going to look at the lysosome should be yellow If the ubiquitin is in the lysosome, bingo! There it is We have autophagy

Actually, when we, that being me, micro dissected the airways out from the diacetyl-exposed mice And we looked at a western blot in band intensity, by molecular weight The free ubiquitin is just little molecules out there before they’re complexed with protein And the process had actually depleted the free ubiquitin in the airway-enriched fraction So there was loss of free ubiquitin, the central mediator of protein quality control In the diacetyl-exposed airways So what do we think is happening to the pathways For proteins damaged by diacetyl? What we believe happens is the protein, when it’s been damaged and is now misfolded and damaged and toxic Is recognized by ubiquitin We probably have some K48 ubiquitin links We’re not quite certain of that yet But what we know is something goes wrong in terms of this proteasomal process It appears to have been overwhelmed or maybe the function has somehow been altered But it can’t keep up with the damaged protein Instead, the alternate pathway for handling overwhelming damage begins to play a role And we see a response to overwhelming protein damage With the accumulation of ubiquitin in the scaffolding protein, sequestosome 1 And we see its movement to the lysosome And autophagy taking over So proteins are going out in the trash And are being damaged So the mechanisms of flavorings-related lung disease appear to be that butter flavoring vapors damage airway lining Diacetyl and 2-3-pentanedione are major toxic components of butter-flavoring vapors And if we look at diacetyl itself and the mechanisms of its damage We know that it damages proteins And that diacetyl inhalation overwhelms the protein quality control in the airway lining And the data are consistent with protein damage As the fundamental mechanism for flavorings-related lung disease And we have additional mechanistic studies that are ongoing today And with that, I would like to introduce our next speaker, Kristin Cummings Who will talk about where this research is going in the field today [ Applause ] >> Well, thank you It’s a pleasure to be here today I do have a confession to make This series is called We Were There But I wasn’t actually there for these early investigation that Doctors Kreiss and Hubbs so eloquently described I joined CDC as an Epidemic Intelligence Service Officer in 2005 Assigned to NIOSH in Morgantown, West Virginia My EIS classmates were busy investigating things that CDC was famous for Like outbreaks of measles and fungal carotiditis in the U.S and HIV and malaria globally I on the other hand had been assigned to a group looking at butter One confused classmate asked, “Are you working on the obesity epidemic?” [Laughter] The idea that butter vapors could pose a health hazard was just not on people’s radar And yet the work that Doctors Kreiss and Hubbs and their many colleagues had done Made clear this was an important public health problem The cluster of lung disease cases at the microwave popcorn plant in Missouri served as a sentinel health event Warning that a larger community was at risk Initially that community comprised the former and current workers at the same plant Afterall, they had been working in the same environment as the case patients And were also exposed to butter flavorings And indeed the investigative team showed that other workers at that plant had been adversely affected by their workplace exposure Some with lung function deficits as severe as those seen in the sentinel cases As Dr. Kreiss described But this wasn’t the only microwave popcorn plant in the country If workers at one plant were developing lung disease related to butter flavorings The community at risk must also include workers at other microwave popcorn plants That use butter flavorings NIOSH investigators conducted extensive evaluations at five other microwave popcorn plants Finding workers with lung disease in all but one This finding indicated the risk did indeed go beyond the sentinel plant Exposures to diacetyl associated with disease were as low as 0.02 parts per million

Or 20 parts per billion Which is much lower than exposures measured in the sentinel plant The need for respiratory protection and engineering controls to reduce exposures was evident Of course microwave popcorn was not the only food product made with butter flavorings The community at risk potentially included other food production workers Making such things as candy, cookies, and snack foods like potato chips As well as the workers manufacturing the flavorings themselves Investigators asked if NIOSH had ever seen cases of lung disease in industries other than microwave popcorn And reviewed the archives containing records of prior investigations Well it turns out, there really is nothing new under the sun NIOSH had seen lung disease in flavoring-exposed workers before In 1985, 15 years before Dr. Kreiss and her colleagues were alerted to the cluster of lung disease cases at the Missouri microwave popcorn plant NIOSH had conducted a health hazard evaluation at a flavoring manufacturing facility The evaluation was requested by the facility’s management after two employees developed severe lung disease over a short time period I think it’s worthwhile to spend a moment reviewing these cases which are now more than 30 years old The first case occurred in a 28-year-old man who was a nonsmoker and was without respiratory complaints when he began working in the mixing room in 1983 He noticed the onset of shortness of breath with exertion after five months on the job He was found to have severe airways obstruction that was described as mildly responsive to bronchodilators He left employment after one year He had no improvement in symptoms or lung function despite treatment with bronchodilators And steroids and cessation of the workplace exposure Case two is remarkably and tragically similar In 1984, after the worker in case one became too ill to continue working in the mixing room A 30-year-old healthy man who was also a nonsmoker Replaced him in that job He developed shortness of breath four months later And was found to have severe airways obstruction Which was described as minimally responsive to bronchodilators He left employment after just five months And like the worker in case one, had no improvement in symptoms or lung function NIOSH investigators recognized these cases as bronchiolitis obliterans And they concluded that they were likely caused by an exposure in the mixing room of the facility Yet unlike at the sentinel microwave popcorn plant, they were unable to pinpoint the cause It’s tempting to attribute this difference to the superior technology, for air sampling available to the 21st century team Compared to their 20th century counterparts Or perhaps some unique combination of tenacity and brilliance among the more contemporary investigators [Laughter] It just wouldn’t stop until they had identified the culprit Unfortunately, the best explanation is one that probably plays a greater role in science than we would like to admit And that’s chance The team investigating the cases of lung disease at the sentinel microwave popcorn plant was absolutely outstanding in terms of their training, experience, insights and approach And they were also extremely lucky Let me explain what I mean On the next few slides is a list of ingredients used at the flavoring manufacturing plant which I’ve taken from the health hazard evaluation report Did anyone spot diacetyl? So it’s listed there a couple of times In different forms, along with more than 200 other ingredients In a batch operation like this one, exposures will vary day to day and hour to hour As workers fulfill customer orders That might range from essence of strawberry to steak flavor In contrast, here is an exhaustive list of the ingredients used to make microwave popcorn The process was the same from day to day, hour to hour Whereas the investigators at the flavoring plant in 1985 were searching for the proverbial needle in a haystack The investigators at the microwave popcorn plant in 2000 were dealing with just a handful of ingredients and a far more predictable exposure profile Meaning that they were in a much better position to zero in on the cause of the lung disease among workers Identifying the cause meant that the recommendation for worker protection Could be targeted to address diacetyl’s specific characteristics For instance, in 1985, the investigators recommended controls and at reducing exposures to dust Because they noticed that the mixing room was very dusty Yet such controls would not necessarily be effective against diacetyl Which was a volatile organic compound rather than a dust Fifteen years later, investigators at the microwave popcorn plant could recommend controls specific to diacetyl Which were likely to be more effective Well, by 2007, the issue of lung disease related to diacetyl was picking up traction As often happens when an occupational illness is identified Members of the general public began to ask if they were at risk

Would popping microwave popcorn at home before watching a movie or the Super Bowl cause irreversible lung damage NIOSH and other experts emphasized that all known cases had occurred in workers Who had been exposed repeatedly and on a daily basis to relatively high levels of diacetyl The exposures at home were simply not comparable And then a physician contacted the Food and Drug Administration to report a possible consumer case The case reported to the FDA was that of a 53-year-old man who worked in sales An exhaustive, exhaustive clinical evaluation turned up no explanation for his disease It would have likely been tossed in the idiopathic basket had his physician not asked a seemingly odd question “Do you by any chance like microwave popcorn?” His response was something along the lines of, “How did you know that? I’m Mr. Popcorn.” And he went on to describe a decade-long habit of consuming multiple bags per day And a fondness for inhaling the vapors after opening each bag The press jumped on the story with major outlets like the New York Times, Reuters, and the big television networks running stories Their headlines varied from fairly staid to somewhat sensational But the upshot was the same A frightening disease that had been limited to workers was now a threat to the general public The next day, industry leaders collectively responsible for 80% of all microwave popcorn sales in the U.S. announced That they were removing diacetyl from their recipes for butter-flavored microwave popcorn Although there were comments that these plans had been in progress for months The timing of the announcement suggested that the need to address perceptions of consumer risk played at least some role In motivating implementation As it turns out, there has not been a widespread outbreak of lung disease in consumers And it’s likely that the risk to the general public is extremely low Mr. Popcorn was not the tip of an iceberg But whether undertaken for workers, the general public, or both This voluntary move by the industry was welcome After all, it was consistent with the hierarchy of controls a tenet of occupational health that emphasizes The effectiveness of steps such as elimination and substitution of a hazard over less effective measures Like administrative controls, and personal protective equipment Which put the burden of exposure reduction on the worker Getting rid of diacetyl and replacing it with something else seemed like the right thing to do At this point, our work shifted from a focus on diacetyl to broader questions about the inhalational risk of food-flavoring chemicals more generally We would learn that substitution does not always resolve an occupational hazard And that the presence of one particular inhalational hazard in a workplace does not preclude the existence of others In a sense, the community at risk was the same But the exposures were changing In 2008, I was the medical officer for a health hazard evaluation at a bakery mixed manufacturing facility Workers combined ingredients to produce dry bakery mixes such as buttermilk pancake mix Ingredients included liquid butter flavors which are added to large mixing tanks as pictured here The management informed us that they had asked their suppliers to remove all diacetyl from their flavorings When we asked what diacetyl had been replaced by They replied that they didn’t know As we would learn, flavoring manufacturers used related alpha diketones with additional carbon groups in place of diacetyl And we heard a little bit about this from Dr. Hubbs As you can see here, diacetyl is also known as 2,3-butanedione The substitutes have the same chemical structure, but a longer tail So 2, 3-pentanedione had one extra carbon group 2, 3-hexanedione two extra carbon groups And 2, 3-heptanedione three extra carbon groups Well my colleagues in the laboratory found that the predominant ingredient in the new flavoring was 2, 3-pentanedione Exposures to 2, 3-pentanedione at this bakery mixed facility were as high as 91 parts per billion At the time, we didn’t know whether 2, 3-pentanedione was toxic And if so, what a safe level would be But we were concerned about the structural similarities between diacetyl and 2, 3-pentanedione And the measured exposures which were comparable to the diacetyl exposures That had been associated with disease in the microwave popcorn industry These findings stimulated the subsequent experimental work that Dr. Hubbs described Confirming that 2, 3-pentanedione is also a respiratory toxin Well in 2011, I led a health hazard evaluation at a large flavoring manufacturing facility that was phasing out use of diacetyl Again, we found that the predominant substitute was 2, 3-pentanedione Yet it was notable that we detected 2, 3-pentanedione in a minority of air samples At least during our visit, most workers were not being exposed to diacetyl or its substitute But that didn’t mean that they didn’t have flavoring chemical exposures These photos, show the research and development area on the left

And the samples preparation area on the right The many bottles representing different flavoring chemicals are evident in each photo And this is really the 1985 list of ingredients in pictorial form This was clearly a very complex environment Where workers were potentially exposed to hundreds of flavoring chemicals Most of which have poorly-characterized inhalational toxicity So we offered questionnaire and lung function testing to the current workers Who numbered about 400 And we found that participants with higher flavoring chemical exposure from longer plant tenure Or more time spent in production areas Had significantly more respiratory symptoms and lower lung function Than their coworkers And we urged a proactive approach That acknowledged that in flavoring manufacturer, diacetyl and its substitutes were unlikely to be the only inhalational hazard In the absence of certainty, overall exposure reduction and medical surveillance are prudent More recently, we published a summary of NIOSH health hazard evaluations of microwave popcorn and flavoring manufacturing facilities From 2012 to — sorry, 2000 to 2012 Just over 1400 workers from nine facilities in eight states participated in NIOSH medical evaluations Including questionnaire and spirometry testing We examined respiratory abnormalities using several different mutually exclusive case definitions So the most stringent was pathologist-reported And just four workers or 0.3% of all participants met a biopsy-based definition of bronchiolitis obliterans The next two definitions were probable and possible And they were defined by abnormal lung function Together they accounted for about 20% of the participants Which is about twice as high A prevalence of abnormal lung function as we would expect to find Finally, symptoms only referred to participants with normal lung function But who reported cough or shortness of breath And over 400 over about 29% met this definition Which is also roughly twice expected So we concluded that the burden of respiratory abnormalities in these workers indicates flavoring-related health effects That go far beyond the few biopsy-proven cases of disease Well in 2013, our attention turned to the risk of flavoring-related lung disease in the coffee industry We reported two cases of bronchiolitis obliterans which is also called obliterans bronchiolitis Among workers at a coffee facility in Texas Five clinical cases would ultimately be diagnosed And this facility roasted coffee beans Producing both unflavored and flavored coffee A subsequent health hazard evaluation at the facility found that diacetyl and 2, 3-pentanedione levels were elevated in the flavoring room Where flavors such as vanilla and hazelnut were added to roasted coffee Exposures were also elevated in the separate grinding and packaging room where unflavored coffee was produced The source was not the flavoring room itself It had separate ventilation and was under negative pressure Compared to the rest of the plant So as it turns out, roasting and grinding unflavored coffee beans releases these volatile organic compounds Meaning that the health risk was not necessarily limited to areas Where flavorings were used That being said, all of the clinical cases at this facility occurred in workers who had spent at least some of their employment in the flavoring room So it was not possible to attribute any of the cases to unflavored coffee alone The risk to workers of roasting and grinding unflavored coffee is an area of ongoing investigation for NIOSH Well one of NIOSH’s responsibilities is to develop recommended exposure limits for specific substances which OSHA takes into consideration When setting legally binding permissible exposure limits In 2016, NIOSH released its criteria for recommended standard for diacetyl and 2, 3-pentanedione Setting its recommended exposure limits at five parts per billion and 9.3 parts per billion respectively The document included recommendations for engineering controls Respiratory protection Hazard communication and periodic medical surveillance of exposed workers To date, federal OSHA has not issued a permissible exposure limit Although Cal/OSHA in California issued a diacetyl standard in 2010 I mentioned hazard communication One of the primary means of communicating hazards to workers is the safety data sheet Yet these may not always include information about diacetyl and its substitutes You recall that when we asked the management of the bakery mix facility What had been substituted for diacetyl in their diacetyl-free flavorings They didn’t know That was in part because it hadn’t been declared on the safety data sheet Where constituents may be omitted if found in small amounts or considered proprietary Recently we looked at 26 flavorings used

at two coffee facilities to compare what was included in the safety data sheet Versus what was actually in the flavoring We found that none of these safety data sheets listed diacetyl Although it was in 81% of the samples Similarly, non-listed 2, 3-pentanedione although it was in 58% of the samples While we can’t be certain why this discrepancy existed We did note that all of the safety data sheets contained trade secret designations So under these circumstances, we concluded that a precautionary approach is warranted when working with flavorings Including exposure monitoring and exposure reduction strategies Well as you might imagine, the issue of flavorings and lung disease took on a litigious tone With workers and consumers suing employers, manufacturers, and even grocers And in one difficult episode, a manufacturer suing NIOSH and NIOSH employees Information that became available during tort proceedings Shed light on why the epidemic in the microwave popcorn industry occurred when it did In the 1990s, in the context of an ever-increasing national waistline The market for low-fat snacks was also growing To create low-fat popcorn, it’s necessary to lower the oil content Which negatively affects the taste To make up for the missing oil, manufacturers increased the diacetyl content of the butter flavorings And added more butter flavoring to their recipes Exposing workers to far greater concentrations of diacetyl than ever before So to the extent that flavoring lung disease came to light because of the low-fat food craze My EIS classmate had been right I was working on the obesity epidemic [Laughter] And getting a first-hand introduction to the law of unintended consequences I played a small part in this incredible story I’ve tried throughout the presentation to include the names of authors In order to acknowledge the many people whose scientific contributions were so critical Yet the scientists alone could not do this work without an incredible support team From the medical industrial hygiene and laboratory technicians Who collected the data To the travel preparers who made sure we got there and back on so many field studies Beyond those in the respiratory health division I want to recognize the invaluable contributions of our colleagues in other NIOSH divisions and offices In-state and local health departments across the country And in-state and federal OSHA sites I also want to acknowledge the astute clinicians who recognized that their patients had something more than routine asthma or chronic obstructive lung disease And took the time to reach out to public health authorities And of course this work could not have been, succeeded without the participation of workers, the support of labor unions where they existed And the cooperation of management and industry who are ultimately in the position to implement the recommendations we made To protect their workers Thank you very much [ Applause ] >> Thank you >> I think we can now take questions >> Thank, thanks for just a beautiful set of investigations and stories And the incredible presentations The comment that it’s okay to eat these things but not to in– inhale them, I guess Raises the question of they being and whether they’re– you have information about particular chemicals that are used in flavoring For e-cigarettes or for vaping That cause concerns for you or whether there’s a science of that or those exposures yet? >> You want to take that? >> NIOSH participated in the original workshop that was done by the FDA when they were determining whether or not they wanted to regulate the e-cigarettes And we did present data on diacetyl and 2, 3-pentanedione there And that is the extent to which we have been involved Because that is more an FDA issue and an indoor air pollution issue Which we’ve been less involved in However, there have been studies that were done for instance by Joseph Allen, I think, was one of the first people To publish on the composition Dr. Parcelionos in Greece has also published on the issue And there is significant diacetyl concentration or at least was in those original e-cigarette vape, its liquids And it is a concern I don’t know how much regulation has occurred But one of the huge concerns is the vaping of e-cigarette liquids that do contain diacetyl intentionally

bypasses the nose And would deliver those alpha dicarbonyl compounds to the deep lung So it is a concern for the most part In terms of the intentional inhaling of them That’s more of an FDA issue So we have not particularly gone there >> But I think the other comment I would make is that inhalation toxicology is very expensive and is not routine And so the designation by the Food and Drug Administration That flavorings that are currently in commerce are generally regarded as safe It is a label the FDA uses And yet that label is really, really is derived from the industry And industry experts It does not reflect toxicologic testing >> The, the trade group Flavoring and Extract Manufacturers’ Association has actually made a public statement regarding that Their, their designation by FEMA And that’s the flavoring FEMA Not the other one Is intended to mean that the substance is safe to eat There is a lot of controversy because I think the wording as it was defined is as intended to be used And you know if you ask some people It’s intended that the worker makes it In terms of the worker issue But when it comes to the flavors that people are vaping in e-cigarettes That is an intended use And FEMA does not approve that process >> And just in follow-up I thought you know you’re sort of looking through the biologic possibility was so beautiful And I think you made the case for being able to predict that certain chemicals are likely to be problematic I wonder whether the industry has sort of absorbed that In their you know research for better flavoring >> I, I don’t know where they’re at in regards to that That is part of where we are at Is, is can we predict which ones are going to be a hazard And also do we go beyond the compounds that are directly reactive? To those that may be metabolized within the airway epithelium to target that same ubiquitin proteasome pathway And so that is ongoing research >> Dr. [inaudible]? >> Thank you again, also for a great series of talks I was wondering if you’re in a position to comment on what the manufacturers knew, when they knew it And what do they communicate to the producers of butter-flavored products? And what responsibility they have to do some product stewardship, perhaps? And if they weren’t doing that well in the past, are they doing it better now? >> I’ll start on that The Flavoring and Extract Manufacturers’ Association has a lot of proprietary data from chemical manufacturers And in the early 1990s, one of the manufacturers of diacetyl actually conducted toxicologic studies In rodents, and noted that the rodents had respiratory disease And that was reported to the Flavoring and Extract Manufacturers’ Association In an unpublished monograph We became aware that such a monograph existed But that it was considered confidential in the courts And we eventually contacted the medical director of that chemical manufacturer Who sent us the entire monograph And we had been provided only the first three pages by the Flavoring and Extract Manufacturers’ Association Withholding the animal studies that, that basically documented That there was a respiratory hazard So you know I think that work had been done in Germany And I don’t know what the, the legal responsibilities would have been But it was an international chemical company You know with respect to the buttermilk bakery mix investigation that Kristin described and was involved in The baking mix company was precluded in their contractual obligation to the flavoring suppliers from reverse engineering

To find out what was in the flavoring And so the only way that they found out was because of NIOSH testing and telling them what was in their flavoring that they had bought So the whole trade secret regulations are very difficult from a public health point of view The other point I would make is that since industry often has much better information than the academic community or government agencies And since even the 1985 outbreak in the flavoring company supplying flavored materials to the baking industry Those cases did come to litigation We didn’t know that of course at the time But one of the thoughts that I had was maybe there ought to be surveillance of litigated cases Which would allow public health agencies to know about what might be arising >> Good luck [Laughter] >> Yeah, I just wanted to make a comment Maybe you have some thoughts It was Dr. Cummings was very interesting to see that the individual consumer case had such prompt response When you think about sort of the industry research and very careful work And I think you, you have this double-edged sword that industrial exposure’s likely to be higher And likely to be a cleaner case for exposure outcome But has its own set of complexity in terms of the effectors to change versus consumer concern And reputation and credibility and you know what really makes us try to lower risk You know or companies want to lower risk rapidly It made me wonder how closely NIOSH works with other parts of CDC or with others Who are engaged in the same exposures? But not for the workplace exposures Whether we have opportunities for more collective impact with that kind of collaboration? >> I would, I would agree with you that it does seem like the, the level of evidence needed for action is different for workers versus the public And I do wonder if you know case, paper in the New England Journal in 2002 That showed there was an association with diacetyl that couldn’t say it was diacetyl But you saw that beautiful [inaudible] response graph Had that been in consumers, we might have seen a different response It may not have taken five years as it did to, to have the industry announce That they were removing diacetyl In terms of collaboration with other groups at CDC I mean certainly NIOSH is, is open to that And is always looking for opportunities to collaborate >> Attributed that to, that the– [inaudible] [ Inaudible ] >> There, there hasn’t been continuing surveillance I mean, we did surveillance in the sentinel plant for almost three years But we have, we have little information about what’s happened after we left One of the issues that was very clear in California was when, when Cal/OSHA and the Health Department in California You know asked for voluntary cooperation The flavoring industry companies to surveil their workforce Only about one of the commercial suppliers of, of spirometry had adequate quality spirometry And you know it, it’s not difficult to do spirometry But quality control is, is really necessary And it was clear from looking at the physicians’ signings relative to reports That they had no idea even what was a completely unphysiologic response So I think industry across the country has probably not had the benefit of good surveillance Because it isn’t available from the private sector And — >> That’s another challenge is that many of these workforces have high turnover So particularly in the coffee industry that we’ve been studying more recently You know the average tenure, when we ask, “How long have you worked at this plant?” Is, is often less than a year or two So in order to follow the population over time

You’d have to try to track down people who might be pretty transient So that’s another challenge >> Sorry, and just a follow up Like did the plant implement like any engineering controls or have their workers use respirators? Can you comment briefly on that side of the [inaudible]? >> Well certainly in the sentinel plant We worked with them to implement the controls You know as industry has, as service jobs have replaced industrial jobs in this country The availability of industrial engineering has gone down markedly And we really, and we saw evidence of that even in the sentinel plant When we told them that they had to exhaust their tanks that held heated flavoring On the mezzanine, they exhausted them into the general plant air Which at the time was common to the packaging area And when we came and said well, you’ve exhausted them But you’ve exhausted them to create a worse situation You know it was only then, that those got exhausted to the outdoor air So you know I think it’s, it’s challenging And in the absence of regulation and enforcement I suspect we have a whole bunch of plants across the country that are still producing these cases >> So I’m wondering if you’ve considered or thought about whether there’s other lung disease might be present amongst the people who don’t respond? Over long term? Is it possible there could be some cancer risk over many years of exposure? Short of getting a, a obliterans-type of condition? >> The National Toxicology Program has recently completed lifetime carcinogenicity studies mice and in rats And they did have limited evidence Or it may have been equivocal evidence that, of carcinogenicity from diacetyl And I don’t remember the actual numbers They were respiratory tract adenocarcinomas, as I recall Of the nose, which in the rodent absorbs a lot of these things And there may, it may have been of the lung as well Nose often puts, having the same cell type of origin as the human lung does And because it does symmetry, maybe the targeting rodents But the workplace data, we don’t yet have human data And but it doesn’t appear to be a powerful carcinogen Due to the nature of how it kills cells I would expect it to be a powerful promoter But that work has not been done That’s just a hypothesis >> Yeah, I would, I would consider that similar to asbestos, the smoking exposure action might be critical >> Yeah, that kind of interaction is what one would predict Because the cells are not going to have their normal defenses And once they’re induced, they’re going to proliferate and fix any genetic damage done by a co-carcinogen But, but those studies have not been done That is a hypothesis >> Can you clarify with a biologic possibility why you think it, the disease continues to get worse after they’re out of the plants? >> There’s actually not much evidence that there’s progression after exposure ceases At least after about two years Many of the clinical cases had horrible coughs for about two years until after they left the plant And then their cough resolved somewhat But their shortness of breath, you know persisted But in general I think that, that the reason many of these cases have not come to lung transplant is that they have a very different natural history Than people who have emphysema, for example Who put, are put on lung transplant lists And equivalent pulmonary function deficits, the flavoring-related cases You know seem to not progress which is not the case with smoking-related lung disease >> From a point of view of individuals entering the, the practice Of public health today It’s a, a lot of good lessons in what you were presenting The, the, the things that you observed The tools that we had The insights we had at hand, going back to the 1985 study

Which, which was a you noted, many, many ingredients Only a couple limited cases And I think some of the rules of thumb that you were basically presenting to people were Think about the way that industry’s changing Consider the way that use of materials is changing Perhaps looking at the coffee industry today with so many small packaging and formulating companies coming into play I think these are some questions that, that everyone could keep in mind as they Either practice within, within formal public health or get trained at universities to go into business Or even through schools to educate students better So that when they go into the workforce, they’re recognizing some of these situations That have, have gone way beyond control Before they’re, before they’re recognized and protection is provided So thanks for the re-recapping the history And I think especially, emerging manufacturing and the changing, the changing workforce across the world Also climate change and how that changes the conditions of exposure And so thank you again >> Thank you Are there any other questions from Envision or from the audience? Thank you very much for your attention [ Applause ]