Cancer and occupational and environmental

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES Is pollution increasing your cancer risk? Foreward We live in a sea of carcinogens, exposed daily to carcinogens in the air we breath, the water we drink, and the food we eat. Carcinogens can also be absorbed through the skin while swimming or bathing. We often try to limit exposure by eating organic vegetables and fruits. Environmentalists warn us of the dangers of pollution to our health, neighborhoods have been abandoned because of polluted soils, and famous court cases have been made into movies. Should we eat only vegetables and fruits that have not been sprayed with pesticides? Should we drink bottled water? Should we move out of polluted cities? The answers require a quantitative risk assessment to put the potential risks in the context of the dominent risk factors of living. Around one third of us will become ill with cancer by age 75. Our risk is very low at age 20 , but then increases expodentially as we age. It is part of the aging process. There are things we can do to lower our risk, but not totally prevent it. The major preventive practices include eating a diet rich in fruits, vegetables and fish, exercising regularly, keeping our weight down, and of course not smoking cigarettes or indulging in unprotected sexual activity. This book uses two stories of environmental pollution to teach the basics of quantitative risk assessment. The first story is about water and sediment pollution and claims of navy divers that diving in such waters resulted in an increased risk for cancer. The second story is about similar claims from city air pollution. The ability to assess the risks of exposures and to put potential risks into perspective is important for those involved in public health, occupational health, for lawyers and for anyone worried that their disease was caused by exposures to carcinogenic materials. At the end of the book is a section of bare facts. It has not been checked for being uptodate in the last few years, so that readers should use the section at their own risk. Neverthess, I believe that it summarizes what is important and can be corrected by those who find it useful. Also I included a section on selected terms, that might be useful for the reader. Hopefully it will help students in their efforts to understand environmental and occupational exposures and risks. Furthermore I hope that this book will be a work in progress. Anyone who wants to rewrite a section is welcome and I will modify and give credit to the author. Also any corrections or comments are welcome. My email is [email protected] All the best,

Prof. Paul Froom MD, School of Public Health, University of Tel Aviv

Self-published and open access January 10, 2017

MEASURING RISK 2

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES

Chapters Title Introduction Chapter 1 Water and sediment pollution Chapter 2 Biological plausability and the limitations of animal studies in defining risk. Mice are not men. Chapter 3 Exposure scenario –quantitative analysis of body exposures to polluted water and sediment Chapter 4 The cancer slope and risk calculation Chapter 5 Quantitative risk assessment-modifying risk factors Chapter 6 Air pollution and risk for cancer Chapter 7 Risk communication Chapter 8 Bare facts-occupational and environmental exposures Chapter 9 Selected terms and definitions Appendix

MEASURING RISK 3

Pages 4- 20 21- 26 27- 42 43- 49 40- 54 55- 64 65- 69 70- 73 74-131 132-167 168-178

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES Introduction: Water pollution and cancer risk We will start with the story about Israeli Navy divers who were exposed to polluted waters and sediments of the Kishon River system to point out how quantitative risk assessment can prevent prevalent misconceptions. The divers, the former Chief Supreme Court Justice, and many scientists came to unwarranted conclusions because they did not use such methodology. Despite the facts, most of the Israeli population believes that the Navy divers became ill because of the diving activity in the Kishon River system. This is similar to the Love Canal story in the United States that we will mention later. Retired Lt Commander John just got the bad news that he was ill with a second cancer in the space of two years. He had just finished receiving chemotherapy for malignant melanoma, a potentially deadly skin cancer, and now he had to undergo additional treatment for colon cancer. He could not understand why he was so unlucky, and understandably looked back over his life to determine what went wrong. He was married with two children, just turned forty, did not smoke, did not drink alcohol excessively, exercised regularly, and was not obese. His parents, brothers and sisters were healthy. John had volunteered for a special forces unit of the Israeli Navy, and was chosen after rigorous and demanding physical testing over hundreds of other volunteers. He believed that he should be healthier then the general population and should not have become ill with cancer. John retired from the Israeli Naval forces at age 40. He trained and also commanded a unit whose job was to protect Israeli Navy ships from sabotage attempts. The job entailed diving in the murky polluted waters of the two main harbors, the Haifa Harbor and the Kishon Harbor to check out the sides and bottoms of the ships. He thought back to the more than 1000 hours he had spent in those waters over his 23-year career. He remembered the punishment dealt out during 18 months of basic training, the numerous times he had to drink a flipper full of smelly fowl tasting polluted waters. He remembered the oil slicks and the difficulty removing the putrid grime from his skin after leaving the water, and the use of the caustic soap resulting in a burning sensation on his skin. He remembered having to check out ships near pipes with sewage flowing out all around him, and the difficulty of looking for explosives because the water was so murky. If navy boats dropped equipment overboard, he had to search the harbor's mud bottom, arms covered in the polluted sediment. It seemed to him logical that years diving into these waters, inhaling fowl smelling air, absorbing the toxins through his skin and absorbing pollutants from drinking the dirty waters caused his illnesses, two cancers despite his young age. What could be clearer than a cause and effect relationship between exposure to definite carcinogens over a period of twenty years and an increased risk for cancer? Becoming ill with his first cancer raised his suspicions and after the second he set out in earnest to find a link between the exposures and cancer. One of those who came to the aid of John was Prof. S whose career included seeking justice for those exposed to occupational pollutants. In the past he had some notable successes, and together with John tried to determine the increased risk of cancer in the exposed Navy divers. The first step was to check for cancers in his fellow navy soldiers. After a lot of legwork, and with the help of others, over 50 divers with cancer were identified. Other divers were said to be suffering from "failure of the nervous system", from kidney failure, and from other strange illnesses. As more and more MEASURING RISK 4

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES divers with serious and strange diseases were found, John felt a confirmation of his suspicions.

1. Does the history of the diver's exposure suggest a cause for his disease? The first step in the analysis needs to consider what chemicals were actually present in the waters and sediments. Are these chemicals known to be carcinogenic and under what conditions? There are many organizations such as the IARC (International Association For Research on Cancer), the US-EPA (United States Environmental Protection Agency) and the NTP (National Toxicology Program) that convene groups of experts to decide whether or not there is evidence that a chemical has carcinogenic potential. They use the basic tools of epidemiology to evaluate exposure studies in man and animals. 2. What does the finding of 50 divers with cancer mean? It only has meaning if this is more than what is expected to occur normally in such a population. This is often difficult to assess, and estimates can be problematic, requiring vigorous scientific methods as we will see later. Small increased risks can be missed if the study population is too small, and bogus increased risks can be reported if the studies are done improperly. The evaluation of studies of workers exposed to chemicals and then followed-up over a long time period to assess the possibility of increased risk is part of the assessment used to determine if a chemical is carcinogenic. 2. What does the detection of “serious and strange” diseases do the probability that exposures caused the cancers? .3 The answer is that it doesn’t change probabilities unless the symptoms are evidence of increased exposure. For example, if someone complains of multiple protruding lesions (papules) on his palms and soles, this could indicate high exposures to arsenic in the drinking water. In other words, if non-carcinogenic effects are evidence of higher exposures then this can reflect on the probability of causality. The non-carcinogenic effects however need to have been observed for the exposure to high levels of chemicals in other circumstances. Unexplainable symptoms are common in the general population. Symptoms are also a consequence in some cases of anxiety after a person believes that he was exposed to drugs or dangerous pollutants. The claim of many other associated and strange illnesses was reported in the newspapers, and is a typical way newsman and others increase the impact of their story. We will deal with risk communication later on but suffice to say at this point that the unknown increases the anxiety of the reader and fear of the consequences of being exposed. It makes for good press. It is natural for patients with cancer to think hard why they became ill, and to believe that they became ill because of something in their past rather than just by MEASURING RISK 5

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES chance. This belief can impact negatively on others with similar exposures. There are divers who believe that it is just a matter of time before they will become ill with cancer. In epidemiological studies this phenomenon can lead to those with disease reporting more exposures than in those without disease even if there is no actual association between the exposure and the disease. This is called recall bias. After gathering all those with cancer, Dr. S. tried to compare the number of cancers in the divers to those that would be expected in the general population, and estimated that the risk of cancer was increased by at over two fold. He told reporters that it was clear that diving in the polluted water led to an increased risk of cancer and a risk of early death for all divers. He was quoted as claiming that the worst period was after 1980 when observed rates of cancer in the divers were about 700 times greater than would normally be expected. He said that the tip off for him was the short time that it took from the first time the divers were exposed to the polluted waters until becoming ill with cancer. He thought that these findings were plausible since the fish had died off in the Kishon River, and that in the Kishon River there was a toxic chemical cocktail of known cancer causing agents; organic chemicals, and heavy metals. He believed that cancer was not the only result of such exposures, and that non-carcinogenic effects included a weakening of the immune defenses with subsequent associated diseases and respiratory problems. 4. Is a risk of 700 times possible? Cancers are common diseases. If we consider the risk of cancer up to age 75 years, a 700-fold increased risk is mathematically impossible. Around 30% of white males in developed countries become ill with some type of cancer by age 75 years. This does not include skin cancer where the cumulative risk by age 75 in fair skinned individuals is 50-80%. Therefore, if 100% of the divers developed cancer the increased risk would be around 3 fold. If we considered males 20-30 years old, we would expect around a 2% risk of cancer over a 10 to15-year period. So if 100% of young men became ill with cancer over that time period, then the increased risk could be as high as 50-fold. Certainly an increased risk of 700 times is mathematically impossible. 5. Does a short time period from exposure until the development of disease (latency period) increase the probability of a connection between an exposure and a disease? Shorter latency periods decrease the probability of a causal connection between the exposure and the cancer. It takes time to develop disease after exposure to a toxin. The time from the beginning of the exposure until the discovery of the disease is called the latency period. This period can be a matter of days or hours if exposed to infectious agents or toxins present if food. From studies of workers exposed to high concentrations of carcinogenic materials we have learned that for cancer the latency period is usually 20 to 30 years or more of continuous exposure. There are exceptions. For example, in those exposed to ionizing radiation, the risk for leukemia is greatest 510 years after exposure and then the risk decreases. For cancer, there are no known carcinogens with a latency period of less than 5 years. MEASURING RISK 6


6. Does the fact that the fish died off increase the probability for causality in exposed humans? First of all, evidence presented by the media has to be taken with a grain of salt. Secondly there are many possible causes of a decrease in the fish population besides pollution, including over fishing, and thick sediment layers that lead to hypoxic conditions for developing fish eggs. Exposure to sediment can also damage fish gills. Lastly fish and other water life are more susceptible to water pollution than are humans. We will learn about the relative sensitivities of fish and other water life to pollution compared to what is calculated to represent a danger to humans later on. We will also learn how to determine concentrations in water and sediment thought to be a danger to public health, and the methodology needed to determine when recreational exposure such as during swimming and diving, should be prohibited. There were other claims made by those interviewed by the media. An environmentalist stated that we know that those chemicals cause cancer; we know that these men were exposed, and this group made up of men who are much stronger than most, now has a rate of cancers and illnesses that is much higher than that of the general public. There isn’t one main kind of cancer: they have all kinds. There have been brain cancers, blood cancers, skin cancers, and cancer of lymph nodes (lymphoma). That’s one of the strong points showing that they were exposed to an outside source that penetrated the body in different ways. If it is carcinogenic, then it’s carcinogenic at any level. “Don’t set levels you say are safe: there’s no such thing”; adding a note of urgency of cleaning up the Kishon saying that they are worried because people are dying now. Another navy diver, 48 years old caterer, told the press that he had undergone 20 skin cancer operations since 1998 and had to give up his job. He told reporters that during dives his skin would burn, eyes would be red and bulging.

7. Does the appearance of multiple cancers increase the probability of causality? Non-specificity will decrease the probability of causality, especially if an increased risk is reported for cancers with no reported association with chemical exposures. For example, a diver with colon cancer is unlikely to have become ill because of exposure to pollution, because even in the workplace with very high exposures to various chemicals no increased risk of colon cancer has been consistently reported despite an increased risk of other types of cancer. Even in heavy smokers exposed to high levels of many different carcinogens in inhaled smoke, an increased risk of colon cancer has not been unequivocally demonstrated. Most exposures increase the risk for only a few types of cancers.

8. What does it mean to state that a substance is carcinogenic at any level? This is a theoretical assumption used by environmental protection agencies meant to be public health protective. No one claims that this means that there is an actual real MEASURING RISK 7

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES risk. In the last half of the last century, people became worried that low exposures to carcinogens were a danger to public health. The United States Environmental Protection Agency among other agencies developed methodologies to provide a high degree of public safety. These agencies use studies in humans but more commonly in animals to determine how higher exposures result in a higher risk of disease. Then they extrapolate these findings to very low doses. There is an exposure level that is called the point of departure (POD), where the extrapolation is assumed to continue until very low and immeasurable risks. This method results in what is called a cancer slope, assuming that at any dose of a chemical it is possible to calculate the lifetime risk of getting cancer. An acceptable risk, that is defined as a “virtually zero” risk is usually an increased risk of 1 per 100,000 people over a lifetime. This can be compared to 30,000 cases of cancer per 100,000 that is expected to occur in the general white male population by age 75. Therefore, this is a theoretical and immeasurable increased risk. It was developed only so that the environmental protection agencies could fix acceptable pollutant concentrations in water, food and in the air. We will see later the limitations of the data used to fix such concentrations, and why the probability of a causal connection is virtually zero even if there is exposure to chemicals, 1000 fold higher than the recommended limits. 9. What can we learn by the story of the caterer? This worker who had dived in the past probably had skin cancers that are caused by sun exposure primarily in people with light skin. Here we find out that we need to consider other risk factors when assessing the probability of an association. It is unclear why the diver’s eyes would bulge and become red even if he didn't wear a diving mask. Irritation of the skin also has not been shown to be a cause of skin cancer unless there are resultant scars or chronic inflammation. The reason he had to give up his job is unclear, and is logical only if his job entailed continued sun exposure, not recommended for those with skin cancer. With this publicity other workers started worrying that they were also affected by exposure to the Kishon polluted waters. Haifa fishermen docked at the Fishermen's cove an area further away from the sea and closer to the polluting factories than the major harbors, and were exposed to the polluted waters while washing their nets, cleaning their boats, and entering the water to either check their boats or remove plastics and other materials tangled up in the propellers. They also spent hours in the docks before sailing and after returning to port, at which time they breathed the fowl smelling air that came from the polluted harbor waters. The media reported that fishermen who once worked the river are now contracting strange diseases. Some of their legs are scarred and look if they are rotting. Even propellers from their boats last only three or four months before the metal is eaten by the acids. A dog fell in the river one fisherman said: “when he got out he lay on the bank and died”. Many fishermen with cancer also looked to the Kishon as the reason for their illness. 10. What do these claims add to the probability of causality?

MEASURING RISK 8

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES This is evidence of cheap journalism. It is another example how to use risk communication to increase anxiety and convince people that the Kishon waters and sediments are dangerous. This type of journalism doesn’t require logic. Why would a dog falling in the water die suddenly when this didn’t happen to humans? Did the fishermen really have scarred legs that looked as if they are rotting? What disease might be the cause of that phenomenon? An elderly Attorney asked for the help of local experts to make the case for the fishermen in court. He recruited 5 experts in the field; an industrial hygienist, a specialist in internal medicine, an epidemiologist, and a biologist and an expert in water pollution. The epidemiologist attempted to compare the fishermen to the general population and to other fishermen who did not use the Fisherman's cove (polluted by the Kishon river waters) as a home base. He concluded that the risk of cancer to the fishermen was increased over 40 times. The expert in water pollution described the concentrations of carcinogenic and other pollutants in the waters and sediments of the Kishon River system. The industrial hygienist described the exposures to the carcinogens. The biologist theorized that these chemicals caused an increase in oxygen radicals leading to an increased risk of cancer. Finally, a specialist in internal medicine, examined the fishermen. He concluded that a toxic cocktail was ingested, absorbed through the skin, and inhaled resulting in the high risk calculated by the epidemiologist. 11. Is possible that a combination of chemicals can increase the risk of all cancers 40 times in people aged 50-60 years old? This was answered above. If from age 20 to age 60 we can expect that 9% of white males in the United States will develop cancer (not including most types of skin cancer) and that this is similar to what is found in Israeli males, then the maximal increased risk can be calculated to be 100% divided by 9% = 11 fold. Therefore, a 40-fold increased risk is impossible mathematically unless we are talking about very young individuals, all of whom become ill with cancer. The epidemiologist was told that there were only 2 fishermen with cancer out of around 300 fishermen who worked out of a harbor in another region. He compared this to 27% cancers found in the Kishon fishermen and concluded a 40-fold risk. This is an obvious error, since in men over age 20 followedup for tens of years such a low risk of cancer in the comparison group has never been described. The underestimation might be due in part to the fact that people generally don't talk about their cancer illnesses. In any case it is obvious that the epidemiologist didn't examine the fishermen from other regions in Israel. The biologist added a possible mechanism for the damage concluding that the toxic cocktail that was absorbed led to the creation of oxygen radicals in the bodies of the fishermen causing aggressive cancers, a suppression of the immune system, an increase in infections and an increase in cardiovascular diseases. Oxygen radicals are naturally occurring substances in the body, and can be increased by smoking and eating certain foods. Eating fruits and vegetables decrease the risk of both cardiovascular disease and cancer, and one of the predominant theories is that they do this by the vitamins in these foods, some of which have antioxidant effects. Oxygen radicals increase naturally with age and also are thought to be part of the aging process. The biologist claimed that the MEASURING RISK 9

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES toxic cocktail resulted in very high levels of oxygen radicals in the bodies of the fishermen, way above the levels that occur naturally or even from inhalation of cigarette smoke. Nearly all the fishermen smoked. 12. What does this story do to the plausibility of a connection between the exposures and the disease? And what do we know about oxygen radicals and is it biologically plausible that oxygen radicals lead to an increased risk of cancer in the fishermen? Biological plausibility is one of the criteria used when determining the carcinogenic potential of a chemical. Stories are also valuable in trying to convince people that there is a real risk, however one needs to take such theoretical stories with a great deal of skepticism. There were no laboratory tests showing that the fishermen had higher levels of oxygen radicals. Such stories are more properly used when considering if a substance can be carcinogenic under certain circumstances or not. They need some laboratory support to be considered as evidence. Also oxygen radicals are naturally occurring substances in the body. They increase with age, and help the body deal with damaged cells. It is not known what the ideal concentration of these radicals should be. There is one study for example that showed that taking antioxidant pills that decrease oxygen radicals actually increased the risk for lung cancer in smokers, and the study had to be stopped prematurely. Today experts in nutrition recommend that vitamins and fish oils cannot be substituted for a diet rich in fruits, vegetables and fish. In fact, we do not know why such a diet leads to less disease. Together these experts felt that their conclusions left no room for doubt. The fishermen were exposed to the same polluted waters as the divers that led to a very high risk of developing cancers, much higher than expected in the general population. Meanwhile there were further activities regarding the military divers. John got a group together and tried to get the military to conduct an investigation and to help those afflicted. The military establishment refused, and John felt "they had left us wounded in the battlefield". He was angry, and also felt guilty about his responsibility for sick soldiers who had served under his command. However, after continued public pressure, the military appointed a blue ribbon-panel to investigate the claims of the divers, headed by a well-known and highly respected judge. The Governmental Committee was made up of three members, the judge, one statistician, and an epidemiologist. 13. What experts were missing? No experts in occupational and environmental medicine, toxicology, or risk assessment were members of this committee. The courts have traditionally considered probabilities as important. Thus a plaintiff wins the case if the connection is shown to be more likely than not (>50% probability) although there are those who believe that the plaintiff should receive a proportion of his claim according to the probability of a causal connection. The epidemiologist and statistician can determine if there was a statistically significant increase in cancer rates in the divers, but they usually don’t have the expertise to consider if an association is plausible between the exposures and the cancers. Furthermore, a negative or positive study won't conclusively rule in or rule out MEASURING RISK 10

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES causality. There was no attempt by the committee to determine under what circumstances a chemical is carcinogenic, the possible absorption of the chemicals, and then to perform a “quantitative” assessment of risk. The question whether an increased risk is more probably than not (quantitative risk assessment) was not answered in the final Governmental Committee report. The judge and the two experts began their work in August 2000, and on Nov 21, 2002, the media reported that the panel’s interim conclusion was that there was no statistically significant association between the exposure of the soldiers to toxic chemicals around the stream during their service and the variety of cancers found. In the media, outraged opinions were expressed describing the outcome as a scandal and abandonment of the soldiers. John said he believed that the statistics used had been seriously manipulated. Others argued that the cohort of divers was a "particularly healthy group" and a causal connection was indisputable. In February, 2004 Haaretz reported the final conclusions of the Governmental committee. The epidemiologist, and the statistician concluded that there was no scientific proof of a causal link, whereas the judge concluded that the exposure to the pollution led to the diver's cancers. The judge wrote that the court is not needed to determine complete causality but only probable causality, and determine that the reason looks reasonable. It is not necessary to decide by considering the theoretical statistics when the committee has before it concrete evidence. Yet there should be a balance that is a logical discussion of the facts.

14. What does the judge mean by probable causality? What is a logical discussion of the facts? What is concrete evidence? It seems that the Judge based his opinion on the premise that exposure to known carcinogens in those with cancer is concrete evidence. Unfortunately, the judge did not get expert evidence to teach him that cancer is common, and that exposure to carcinogens is common, and he has no idea how to do quantitative risk assessment to calculate the probabilities. The judge as well as the public also has misconceptions on exposures. A recent report found that over 50% of the general population in the United States believes that there should be no carcinogens in drinking water. They don’t understand that there are low concentrations of multiple carcinogens in food and in drinking water and that the Maximum Contaminant Levels (MCL) of such substances set by the EPA and governmental agencies from other countries provides a very high degree of public health safety. We are all exposed daily to carcinogens in water, food and from breathing air. Some elements that are carcinogenic under certain circumstances are actually essential elements required for human health under other conditions. Therefore, it is not enough to say there is exposure in someone with cancer to use that as concrete evidence of a connection between the disease and the cancer. The fact that someone is exposed to a carcinogen is not enough to suggest that becoming ill with cancer was caused by that exposure. Estimation of the dose is essential since, the cells of our body repair thousands of damaging events to our DNA daily caused by natural and pollutant carcinogens. Acceptance of the judges conclusions means that anyone with cancer can sue the government for allowing exposure to carcinogens. MEASURING RISK 11


The history of the Kishon River and pollution The Kishon River is 63 kilometers long, starts in the Jezreel Valley and empties into the sea just south of Haifa. The source is south of Mount Tabor, the highest mountain in the lower Galilee. The mountain is 1800 ft above sea level, and from there the river runs southwesterly through the valley of Jezreel, then through a mountain ridge to the plains of Akko, near the foot of the Carmel Mountains, and finally falls into the Mediterranean Sea just North of Haifa. The Jezreel Valley marks the southern limit of Lower Galilee and historically was important as a line of east-west communication. It is mentioned numerous times in the bible. Mount Tabor was a meeting place of men before battle, and the Jezreel Valley was important in uniting the northern and southern tribes of Israel. In the War of Deborah “the torrent Kishon swept them away, the on rushing torrent, the torrent Kishon” (Judg. 5:21). Thus the Arabs called the river Nahr Mukata (the Stream of Slaughter). The Jezreel Valley has been predicted as the place where nations will assemble for the final conflict, taking place at the coming of the messiah. In the 1900s, the waters were reported to be clear and of a greenish color. In the summer there was very little flow, but in the winter the river became at times so broad that the whole valley of Jezreel was covered with water. The rich alluvial soil of the Valley produced fine crops of wheat and barley. Before 1930 the Kishon River flowed through a deep bed in marshy ground, dangerous to those who tried to cross it. Its depth sunk in places 15-20 feet below the level of the plain, and had many tributaries flowing from the mountains. The lower river is 7-8 km long starting after the entrance of the Ztipori River a major tributary (figure 1). For years it was known that the river was polluted by industry, municipal sewage, and from various other local sources as well. As early as 1951, research was done on pollution of Haifa harbor. In 1953 a joint committee included representatives from the Agriculture ministry; from the Navy, from the harbor authority and from a paint factory. They wanted to deal with the problem of corrosion and fouling in the harbor that led to an increased growth of organisms on the boats side and drag during sailing. As early as 1978, there was a directive that defined acceptable pollution levels. There were not however timetables to reach these levels, and they were not enforced. The industries polluting the river included an oil refinery, petrochemical plants, plastic manufacturers, fertilizer producers, various chemical biochemical production plants, and a large regional sewage plant. Over the years the flow of the river decreased (because of a dam and water use). Today the river might better be defined as a sewage ditch, with sewage increasing to the extent that during certain times of the year, 80% of the rivers total flow is from the municipal sewage plant and factory waste waters. Partially because of the decrease in flow, mud accumulates at the bottom of the river and fish nearly disappeared from the river, although fish were still found in abundance in the adjacent Haifa Harbor. To keep the Kishon Harbor open, periodic dredging was done. The fishermen claimed to be exposed to chemicals during the cleaning of nets and occasionally from entering the water to free the propellers of debris. There were also claims of additional absorption of chemicals from air pollutants that evaporated from MEASURING RISK 12

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES the waters. There were variable estimates of the duration and the amount of exposures by the divers who were questioned by the governmental committee. In the 1990s the Navy became aware of possible effects of diving in the Kishon and Haifa Harbors, and such dives were nearly stopped. But up until that time the Navy dived in the Haifa Harbor, the Kishon Harbor and sometimes even in the Fisherman’s cove. One diver said that the exposure was tens of hours and, they sometimes had to wait up to two hours before showering. They claimed that they swallowed water, estimated to be on average a half-cup per dive (100 ml). Figure 1. The Kishon river (D) flows into the Fisherman’s Cove (C), Kishon Harbor (B), and out to sea and into the Haifa Harbor to the left of the sea break (A). Maximally the Navy seals spent 2000 h in the Haifa Harbor and outside it, 400 h in the Kishon Harbor, and 100 h in the Fisherman’s Cove.

15. Is it probable that the divers drank 100 ml on average on each dive? Is it likely that the fishermen had daily exposure to the Kishon waters and sediments? It is likely that both stories are overestimates. People who are ill or believe that they have been endangered will remember only extreme cases of exposure. This is called recall bias as we have already mentioned. Estimates of gastrointestinal intake by both recreational and professional divers not ill at the time of answering the question has been reported to be 10 ml on average. Fishermen who were not ill claimed that they rarely or never had contact with water or sediment from the Kishon River system. As part of the quantitative risk assessment, after determining that there was exposure to carcinogens, the next step is to visit the workplace often called a walk-through survey commonly done by occupational medical physicians and hygienists. This is done to determine the exposure scenario. In the workplace appropriate information includes work conditions, the use of protective equipment, and actual measured concentrations of chemicals. Divers claims of exposure time to the polluted waters varied from tens to thousands of hours. They reported that the smell was bad. They claimed to suffer from various symptoms such as headaches, burning skin, cuts that didn’t heal, and recurrent ear infections. They had to scrub off the dirt that might have increased the absorption of the chemicals. Drops of water were inhaled while washing the breathing tubes of the diving equipment. The soldiers in training were forced occasionally to drink a flipper full of MEASURING RISK 13

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES the Kishon water as punishment. Evidence from divers included descriptions of the masks changing color, to green after diving in the fishermen's cove. 16. What do we do with such information when doing quantitative estimates of exposure? In risk assessments extreme circumstances are usually assumed in order to give conservative estimates that protect public and individual health. Therefore, we will see later that for risk assessment, for the divers a conservative exposure scenario was 1000 dives, 2500 hours of exposure with 95% of the body exposed to the polluted waters, and 50% of the body concomitantly smeared with sediment, and a total of 100 liters of the Kishon waters ingested (100 ml ingested on each dive). This is a very conservative estimate maximizing potential exposures, called the worse-case scenario, probably never or rarely occurs in real life. The navy dived in the Kishon river area from 1950 until the early 1990s, and according to the governmental committee they were exposed to carcinogenic materials, including chemicals, biological agents and radiation. The Governmental Committee concluded that these agents are known to damage and endanger life, and principally to cause cancer. Over the years there was deterioration with regard to water quality and health dangers from carcinogens from industries and sewage. The judge concluded that “There could be no doubt in the hearts of the factory directors what their pollutants contained, and what their possible effects are with all those who come in contact with the water”. They also believed that one should not de-emphasize the importance of the mixed exposures where the sum of their effects has not yet been studied in full. 17. How do you determine the importance of mixed exposures? There is a mathematical way to deal with mixed exposures assuming that the final effect is the multiple of each separate effect. This is a very conservative assumption. The effect of two substances can be greater than either one alone (synergism), or less than the sum of their effects (antagonism). In any case the “unknown effect of mixtures” is a powerful way to persuade the uninitiated that the exposure caused cancers, since “uncertainty” is one way to increase anxiety and persuade the doubters of cause and effect. As we will see later for low exposures it really doesn't matter if the model used to estimate risk is either synergistic or additive or antagonistic. The World Health Organization summarized the literature on synergistic effects of low exposures and concluded that synergistic effects of low exposures do not affect human health. In the report the Judge wrote “We have no scientific way to test the individual effect of the carcinogens on each individual diver”. The carcinogens included inorganic chemicals, chromium, zinc, copper, lead, nickel, cobalt, cadmium, mercury and arsenic and radioactive materials from phosgum, naturally occurring earth with high levels of phosphates used to make fertilizer. The water was overflowing with inorganic carcinogens and carbon based carcinogens (organic) in amounts that created a clear health danger. There were changes over time and therefore no determination can be made of exposed to a particularly dangerous situation for any specific diver. MEASURING RISK 14


18. Is there any way to test the effect of the carcinogens on each individual diver? Risks based on individual sensitivities are usually not possible to calculate. Uncommonly, there are cases where certain genetic make-ups can increase or decrease risks based on quantitative assessments. When we talk about anything regarding the individual we cannot predict what will happen, but only give it a probability that relates to groups of people. For example, we can answer the question of lifespan as a probability. We have around a 50% chance to reach age 80 years and perhaps a 1% chance to reach age 100 years. Still for the individual person we can’t say whether or not his genetic make-up, luck and the stresses of life will allow him to reach age 100. Therefore, the statement of the judge is philosophically correct but irrelevant in either the scientific or judicial realms where a probability is required and possible to estimate. This does not mean that there are not uncertainties that also should be discussed. It is based on group statistics, but then extrapolated to the individual. This is done all the time. For example, we know that smoking increases the risk for lung cancer. We recommend that people stop smoking even though we don't know if the individual will become ill with cancer if he continues to smoke. The worse case scenario can be estimated, absorption calculated, and quantitative risks determined. To show a “clear health danger” you need to quantify the exposures, define the cancer slopes and only then the risk can be quantified.

19. Did the evidence show that the water was overflowing with inorganic carcinogens and carbon based carcinogens in amounts that created a clear health danger? Concentrations and amount of absorption based on an exposure scenario is the correct terminology. The claim that there was a clear health danger was not supported in the Governmental Committee reports. There was no attempt to quantify the risk. The inclusion of non-carcinogenic materials in the list is evidence of a significant bias of the committee. It is unclear why the judge again mentions the effects on an individual diver when he emphasizes that a probabilistic approach is recommended. Most of the assessments of risk have concentrated on the sediment, since the sediment acts like a sink, and concentrations can be measured, whereas pollutant water concentrations are very low, unless measured right near the source. The committee report wrote that the international community has classified cadmium and chromium as definite carcinogens, and that other inorganic chemicals are dangerous even if not classified as definite carcinogens. Mercury is particularly dangerous. Lead was used to be added to gas and has been taken out because of its poisonous effects. There is also a possibility of the production of additional mixtures; of different and changing compositions whose effects are libel to be greater together then each by itself. The same is true for organic compounds. Even today there are hundreds of organic materials, and possibly even more previously, since in 1994 there was pressure to decrease pollution of the waters. There were periods where the oil covered the water. Compounds of organic aromatic and non-aromatic nature included MEASURING RISK 15

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES additions of brome and chloride. Most of these materials can damage one's health and some are known to be carcinogens; like benzene. The EPA requires swimming water to have less than 200 fecal coliform/100 ml, 126 E. coli/100 ml or 33 enterococci/100 ml. The only clear and present health danger is from high concentrations of bacteria even today. There are a lot of organic materials that would even today take years to define, some which were created (probably) by chemical reactions between the different materials in the river. It is possible that these materials damage health no less than the chemicals that were in the polluted afferents. 20. What is known about chemical reactions in polluted waters? Again the judge or the expert opinion provided to the committee quotes the unknown that increases credibility. Chromium is an essential element needed to maintain human health. Chemical reactions that create new and more toxic chemicals in sediment are very unlikely. The chemicals are in low concentrations in the waters, poorly reactive, and their resident time is short. The chemicals are nearly all hydrophobic, and therefore concentrate in the sediment, where they undergo degradation by bacterial action. There is no evidence and it is improbable that a supercarcinogen was produced by chemical reactions in the polluted waters. No expert provided any scientific studies supporting that hypothesis.

21. What about the international classification for carcinogens? The classification takes into consideration the whether the exposure is oral, dermal or pulmonary. For example, cadmium and chromium are not considered carcinogenic by oral or dermal absorption. It is unclear why the committee talks about noncarcinogenic materials such as mercury. No symptoms of mercury poisoning were reported. It is true that bacterial organisms often exceeded those recommended for recreational swimming exposure. This is one criterion that might have lead to the government putting off limits recreational swimming in the Kishon River system. However, in the context of decisions made in the army and in a factory, exposures can be warranted if the risks are low, if all is done to prevent the exposures and if the workers are informed of the risks. For example, if the training was essential to prevent loss of life during combat, or due to placing explosives on the sides of the ships, then certain unavoidable risks are warranted. Other expert opinion presented to the governmental committee included the following opinions. 1. Prof. biological sciences: a. "there are some materials which even one exposure will cause damage that is irreversible". b. You can extrapolate from the reaction of organisms to pollution to man because most of the physiological mechanisms are the same. c. there were no non-vertebrates in the lower river, proving the poisonous nature of the water. He was quoted as saying that this is like no other

MEASURING RISK 16

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES river in Israel, and no life forms were found as low as the fishermen's cove. 2. Epidemiologist: a. if there is doubt then preventive action should be taken. He had published an article showing that the risk of cancer in the navy divers was 2.29 fold higher than in the general public. b. He could not explain why the conclusions of the committee on the calculated epidemiological risks differed from his publication, nor how he found more cancers in less divers. c. He also presented tables showing massive exposures that he published in a scientific journal. d. He claimed that the short latency period, less than 10 years and cancer in young soldiers are evidence for a high risk that is increasing. 22. Is it possible that one DNA hit can cause cancer? Anything is possible, but the probability is extremely low. Multiple genetic defects are found in cancerous cells. Today the predominant theory is that it takes multiple genetic defects accumulated over time for cancers to develop. There are thousands of daily events that cause DNA damage, and most of this damage is repaired. The risk of cancer is due to a cumulative increase in unrepaired DNA and the decrease in our ability over time to repair such damage. That is one reason that the risk of cancer increases as we age. The risk of cancer from long-term chronic exposures is much higher than that from a single exposure, unless that exposure caused scaring or a chronic disease. The question is probability rather than possibility. 24. Is it possible to extrapolate from other organisms to man? It is uncertain if exposures of mice can be extrapolated to rats or to man. Certainly the extrapolation from organisms living in the sediment (benthic) or from fish to men should be done with the highest degree of caution. We will present later how experiments for carcinogenicity are done in animals and what conclusions can be made from such data. 24. What is the significance of the lack of benthic (sediment living) organisms? This shows that the polluted sediments were toxic to the ecosystem. We will see later that organisms living in the sediment are sensitive to pollutants, and can be poisoned at concentrations that do not represent a danger to people using the river for recreational or occupational purposes. 24. Is the study showing a 2 fold increased risk enough in itself to prove causality? Epidemiological studies are important in determining the possibility of a causal connection according to accepted criteria. Even if accepting the findings of the study over that reported by the committee, such studies are only the first step in demonstrating causality. An increased risk compared to the general population does not MEASURING RISK 17

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES isolate the factors that are responsible for the increased risk as we will see later. Therefore, quantitative risk assessments are essential and should be the first step. The governmental committee concluded that despite the gradual improvement “there is no place to minimize the health danger from the Kishon river and harbor” and that even today it is a “public health danger”. Diving and swimming in the water should be prohibited. And in the opinion of the committee sport boating should also be prohibited. The epidemiologist and statistician in the committee did not find a statistically significant increase in cancers in the divers compared to the general population or of exposed divers to non-exposed divers. The Judge however believed that the results were biased because they did not include 8 female course instructors who dove and swam 150-180 hours in the region of the Kishon, with some of them also becoming ill with cancer. Four of them had up to 500 hour's exposure in the Kishon Harbor. The committee agreed that further diving in the Kishon Harbor should be prohibited. As opposed to the two scientists, the judge concluded that the probability that the exposures caused the cancers was more than 50%. The judge based his decision on the fact that no one disputes the presence of carcinogens in the water, and that it appeared that those with the heaviest exposures had an increased risk of developing cancer. All should benefit from this decision since whether cancer occurred in an individual depends on individual sensitivity in part due to the immune system of the individual diver. There are however significant side effects to his decision that we will show later.

25. Was the decision of the statistician and epidemiologist correct not to include female course instructors? Was it biasing the results not to include the female course instructors? This shows the bias of the judge who doesn’t understand epidemiological methodology. Any analysis of a group of 8 lacks the precision to draw any conclusion. Therefore, in the analysis of the exposures of thousands of male soldiers it was methodologically correct to focus only on the male soldiers. There might have been other groups exposed. Here one needs to be careful when adding other groups to the major cohort, since they may be selected because of their known morbidity. This is called selection bias. This is a problem for cluster analysis when people were chosen because an increase in cancer was noted in some group. The judge might have accepted this claim because of the testimony of an epidemiologist who incorrectly included these female soldiers in his analysis. I personal took care of one of these female instructors who became ill with Hodgkin’s lymphoma, a cancer of the lymph nodes 1-2 years after finishing her army service. A latency period of less than 5 years in itself for Hodgkin’s lymphoma decreases the probability of causality. Furthermore, studies on Hodgkin’s disease in workers exposed to very high levels of carcinogens have not shown increased risks. Even smoking has not been shown to increase the risk of Hodgkin’s disease. Including female instructors in the cohort because they had one or two who became ill with cancer is a positive directional selection bias. They are not defined as divers and their inclusion was done only after discovering that 2 of the instructors became ill with cancer. There were 1000s of others exposed to the Kishon water system during MEASURING RISK 18

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES their army service. It is a common error to define a cohort after observing the diseases. The result of such practices is like shooting at a barn, and then drawing a target around the spot of bullet entry. It is extremely important first of all to define the cohort and then determine the incidence of disease (how many people become ill over a defined follow-up period). 26. Should diving, swimming and sport boating be prevented because of the possibility of dangerous concentrations of chemicals in sediments and waters of the Kishon River System? There are Interventional Levels that are recommended, based on exposure scenarios. Interventional levels are those where humans should not be exposed until the recreational waters are cleaned up. No such assessment was done, so those recommendations are unfounded. For swimming and recreational diving however fecal bacterial contamination might limit the use of the river for recreational use, and special masks can be used by professional divers to prevent exposure to infectious material. I was paid for my opinions against the claims that the Navy divers and others became ill because of their exposures to the sediments and waters of the Kishon River system. The atmosphere is highly charged. Arguments for and against however are irrelevant to the basic understanding of quantitative risk assessment. I have presented this story to put quantitative risk assessment into a real life context. The actual outcome of the court cases is irrelevant to the subject at hand. Nevertheless, the reader will have to assess on his own the possible bias of this book. Quantitative risk assessment requires answers to the following questions: 1. What were the potential chemical exposures? 2. What were the observed cancer types? 3. Are the potential chemical exposures (route of exposure specific) causally connected to an increased risk of cancer? 4. What is the exposure scenario? 5. What is the dose-response (cancer slope) estimated for the exposure? 6. Using the previous answers, what is the quantitative cancer risk? 7. Finally, is this risk modified by the type of cancer, the latency period, time from last exposure, or by other competing risk factors?

We need to understand how to determine if a certain group of people have an increased risk of cancer as we reviewed already (Epidemiology). We have to understand how the body absorbs toxic chemical in the environment and in the work place (toxicology). We need to learn about exposure to carcinogens in our food, water, sediment, and in the air we breathe, and how environmental protection agencies fix acceptable limits for chemicals in food, water and in the air (industrial hygiene and risk assessment). Also we need to know how to assess the danger of exposure to water and air over the acceptable limits and how to determine interventional values. We need to learn about carcinogenesis, about carcinogens, and how the International Agency for Research on Cancer (IARC) determines if a certain chemical is carcinogenic. We will MEASURING RISK 19

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES also learn about risk communication including communicating with the general population, with workers, and with courts of law. Only then can we understand how so many "experts" and a well respected judge made so many errors, and came to the wrong conclusion with negative consequences. The way to determine the probability that a chemical caused a patient's illness called personal quantitative risk assessment. In 1983 the National Research Council published Risk Assessment in the United States Federal Government: Managing the Process, commonly known as the “Redbook.” This established the 4-step process that has become the dominant paradigm for risk assessment, also used today by the major textbook of Cancer Epidemiology (Schottenfeld D, 2006). A 5th step is added that considers final modifications including the effect of other personal risk factors (like cigarette smoking). Risk assessment integrates the disciplines of toxicology and exposure assessment to attempt to understand and measure what types of harm humans or ecosystems might experience from exposure to a chemical or pollutant. It uses available scientific evidence as well as assumptions, mathematical modeling and policy judgments, to attempt to estimate risk. In human health terms, risk is a measure of the chance that a person or population will experience injury, disease or death (a hazard) under certain circumstances or exposures. It is a combination of: the probability that an undesired event (exposure to a toxic chemical) will occur, and the consequences that occur as a result of that event (injury, disease or death).

MEASURING RISK 20


Chapter 1. Water and sediment pollution Water pollution is defined as the impairment of water for its intended use, actual or potential, by man-caused changes in the quality of water. It may be a natural substance or a toxic synthetic compound. It may be from a point source such as discharges from a specific location through pipes or ditches as from factories or municipal wastewater treatment facilities. Non-point sources include runoff from farmlands, the largest source of water pollution in the U.S. (over half of pollutants into streams, entering lakes and into coastal areas). This leads to pollution of both water and sediment. In the Kishon River system, the point sources of pollution are from the factories and the municipal sewage plant and from gasoline stations in the Harbors. Non-point sources are primarily from human activities in the Harbors, including inappropriate dumping by the fishing, other commercial, and military fleets, the chemical treatment of the boats, and from gasoline or diesel motors. Question 1: What makes up most of the exposure to benzene, a definite carcinogen in the Harbors fed by the Kishon River system? Answer; Non-point sources are the dominating source, Although, benzene is an industrial pollutant, it rapidly volatizes or absorbs to sediment. Up to 30% of outboard boat motor gasoline including benzene pollutes the waters. Therefore, it is not surprising that actual water measurements show that the river above the harbors has low levels of the combination of benzene-toluene-xylene (called BTX), that increases precipitously in the Harbors due to non-point sources. US EPA ‘s mission is to protect human health and the environment. For human health this includes exposure through direct contact or through the aquatic food chain. The National Sedimentation Inventory (NSI) examines 8.8% of the US territory overall, and over 50% of the country's watersheads, monitoring around 19,000 sites where sediment might represent an ecological or human health risk (SEDIMENT from EPAcontaminated sediments science priorities. December 2004. Science Policy Council, US EPA , Washington DC). They found that 43% of the sites had probably adverse effect on the ecology and/or human health (Tire 1), 30% with possible effects (tier 2) and only 27% no effects seen (Tire 3) . There are fish advisories for 23% of lake water areas and over 9% of the nation’s river areas mostly due to mercury and polychlorinated biphenyls (PCBs)(a banned chemical that originated from transformers, with 75% remaining in the sediment even after 10 years). Most of the risk to human health from lakes and rivers is generally estimated to come from the ingestion of fish. For fish it is important to differentiate between bioconcentration and bioaccumulation. Bioconcentration is the direct uptake from the medium, and is leads to less exposure compared to bioaccumulation that defines a process where the concentrations of the chemical builds up along the food chain. This is because the mass of the herbivores is less than the mass of plants they feed on. Each step upward in the food chain results in a reduction of the biomass, called the pyramid of the biomasses. The loss of pollutants does not follow the same pattern resulting in bio-magnification. These chemicals are often called persistent bioaccumulative toxins MEASURING RISK 21

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES (PBT). This is the reason that the major effects of water pollution on human health is estimated to be due to eating contaminated fish. The number one cause of ecological toxicity is habitat destruction with a build up of sediment and polluted sediment being a major contributing factor. Fish as well as their food (plankton, insects, crawfish, etc.) need a specific habitat to live in, and clear water for sunlight to promote plankton growth, the start of food chain. Many insects and other aquatic animals need rocks to live on or under, not mud, and they need flowing water. Fish need depth to over-winter and cover from rocks and trees, and if the habitat is not appropriate then fish can’t feed or reproduce. For example, bass need gravel areas to spawn. They need to build nests and cannot do that in mud and silt/sediment. A large amount of sediment in water can cover eggs depriving them of oxygen. High sediment levels can also injure fish by damaging gills and suspended sediment is very abrasive. The Kishon river system has a problem with the build up sediment, and dredging is required to keep the Fisherman's cove open. This is due to a decrease in water flow, and problems with run-off from adjacent lands. Parameters for water quality include physical parameters include temperature, presence of mud and sediment, water color, turbidity, odor and pH. Another important parameter is biological contamination that can vastly change the quality of the water, and defines fresh water into zones. In the U.S. about 35% of municipal sewage is discharged virtually untreated into marine waters. The Kishon water system has municipal sewage that is mostly treated before discharge, but overflow situations occur during the year. The 5 zones are as follows. Zone 1 - Clean Water The area above the discharge of bacteria. Typically, there is no offensive colors or odors and the temperature is right for the climate. There are no chemical contaminants and the biological activity is in balance. The dissolved oxygen and species variety are generally high and steady. Zone 2 - Active Degradation Zone 2 is the area of discharge of untreated wastes. Wastes are mixed with the receiving waters. The water becomes turbid, and there is bacterial growth. The dissolved oxygen (DO) (Table 1) starts to decrease as bacteria go to work on the organic materials. There are many coliform bacteria present. Organisms present in this zone include midge larvae, sludge worms, and some fish (like carp, bowfin, etc) can persist, but species variety drops. Zone 3 - Active Decomposition Waste products are decomposed by bacteria. DO is at the lowest point, and BOD is decreasing along with the increase in the quantity of organic matter in the stream. Bacteria population is at its maximum. Very few fish, if any, are to be found in these waters. The predominant species are the rat-tailed maggot, mosquito larvae, and sludge worms. Zone 4 - Recovery

MEASURING RISK 22

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES The stream is returning to normal. The bacteria count fall back toward normal levels. Dissolved oxygen increases and species variety rises. There is an increase in the leech population, because there are still few fish to eat them. The more persistent fish may be found. Nitrites are oxidized to nitrates found at high concentrations. Zone 5 - Clean Water (Secondary) Parameters back to pre-pollution load status except for nitrates, which may remain high for a time. The Zone Lapse Time refers to the time it takes to go from clean water to clean water. It depends upon the concentration and amount of waste load and the speed of complete mixing with receiving waters. Stream mixing is effected by stream velocity, depth, temperature gradient, and aeration.

The Kishon fresh water system after the municipal sewage plant efferents can be defined as Zone 2 or 3, but the harbors are difficult to define because of the tides. River waters are mixed with marine salt water since the river bottom is lower than sea level for over a kilometer inland from the river mouth. Table 1. Dissolved oxygen (D0) and an indicator of water quality Water Quality Dissolved oxygen (ppm) Good 8-9 Slightly polluted 6.7-7.9 Moderately polluted 4.5-6.6 Heavily polluted 4-4.4 Gravely polluted 100 mg/m3 SEDIMENT Sediment is both a repository and source of pollutants. In the United States, The Clean Water Act stipulated that wherever possible the standards should be met that protect wildlife and human health during recreation should be met by July 1983. TMDL – total maximal daily loads is defined as levels that meet the standards. To understand the pollution of sediment and it’s effects one needs to study sediment transportation; diffusion, dispersion within the sediment bed, water movement, bulk sediment movement, movement of suspended sediment, and dissolution into overlying water. Sediment can be capped by cleaner sediment-dependent on the disturbance of the sediment bed, by boat traffic and dredging. Terms include biosediment accumulation factors (BSAF), and bioaccumulative factors (BAF). For the quality of sediment, besides the measurement of contaminents, one also uses the presense of benthic organisms in both fresh and marine sediment. The EPA has used the term sediment quality triad (SQT) that includes sediment chemistry, sediment toxicity (defined in the laboratory) and benthic community composition. Interstitial water concentrations are concentrations of chemicals in the sediment that are free from the solids. They are often a better predictor of toxicity than the sediment concentrations because pore (interstitial) water concentrations reflect the concentration of chemical available to the living organisms rather than the binding phase where the chemical is not available. The major limitation of using the benthic assessment of pollution is the difficulty of relating the assessment to individual chemicals or other stressors. The major potential chronic human health hazard is from exposure to polluted sediment and not from polluted lake water because the pollutants of concern are not very soluble in water and concentrate in the sediment that acts like a sink and can pollute fish through the food chain. Sediment is also the number one cause of all problems in habitat destruction. Fish need clear water from sunlight to promote plankton growth and start of food chain. Sediment can cover eggs and then the eggs die of hypoxia. High sediment can injure fish gills by abrasion. The most important factor in decreasing the quantity of fresh water sediments is proper runoff control such as planting trees. Because the sediment acts like a sink, it is much easier to measure pollutant concentrations in the sediment. Water levels are very low in comparison. To understand the fate of various organic compounds, one needs to understand the equilibrium partitioning, the interaction between water and sediment for absorption and desorption (chemicals reentering the water from the sediment), metabolism of the chemical, and layering of the sediment. There is the active sediment layer that undergoes most of the changes, and the buried sediment that is much more stable, and less interactive. For example, to demonstrate the complexity of chemical behavior in sediment, let's consider one important pollutant, poly aromatic hydrocarbons (PAH). Their source is the incomplete combustion of fuels, urban runoffs, automobile exhaust and discharges, and accidental oil spills. PAH are harmful to fish and other aquatic life, especially to organisms that lack enzyme systems to break down PAH. MEASURING RISK 24

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES The individual chemicals are made of benzene rings, the higher the number of rings the more carcinogenic. The turn-over rate is high for the 1-4 ringed PAH, but very slow for the more carcinogenic PAH with 5-6 rings or more. The desorption for such larger ringed chemicals is close to zero, 50% is degraded and none is volatilized. The sediment half-life for these chemicals increases as the ring number increases; 2 rings- 3 weeks,, 4 rings- 1 year, and 5 rings – 6 years. Ecological and Human health Sediment levels requiring intervention and thought to be a significant risk to human health are called maximal tolerated risk (MTR) levels. The MTR is based on potential adverse health effects that are dependent on the degree of exposure, either by ingestion or skin exposure to the sediments as well as the biological response to various concentrations that can result in a risk to human health. Lake dwelling animals are affected at pollutant concentrations that are less that the MTR levels (Table 2). The terms for chemical pollution of sediments are as follows for effects on organisms; rare effect levels (REL), threshold effects limit (TEL), no observed effects limits (NOEL), probably effects limit (PEL), median effects level (ERM), and severe pollution, effecting most benthic organisms. Following is a typical chart (Table 2).(next page). The human toxicological definition for “serious soil contamination” is taken as the soil quality resulting in exceeding of the Maximum Permissible Risk for intake (MPRhuman). The MPR-human along with exposure modeling is the basis for what is called the serious risk concentration in humans (SRC-human). For genotoxic carcinogens the acceptable excess lifetime cancer risk was set at 1 per 10,000 individuals; for all other compounds the MPR-human does not result in any adverse health effects during a lifetime exposure (70 years). Intervention Values used to classify historically contaminated soils were extrapolated to sediments. For deriving human-toxicological risk for sediment the human-toxicological Maximal Permissible Risk (MPR) level was used in combination with the SEDISOIL exposure model (exposure to contaminated sediment), that assumes ingesting 350 mg of sediment per day, with 50% of the body smeared in sediment for 8 hours per day for a total of 30 days per year over a lifetime. In this exposure model it was also assumed that 95% of the body’s surface area is covered with water during exposure, and 50 ml are ingested per day of exposure (1.5 liters per year). The surface area of the body of course is dependent on the size of person. It takes into consideration the weight and height of the subject. For example, a person 1.70 meters weighing 70 kg would have a calculated surface area of 18,300 cm2. There are many internet sites that will do the calculation for you according to various methods. One such site is www.halls.md/body-surface-area/bsa.htm. . The formula usually recommended is called the Mosteller¹ formula; BSA (m²) = ( [Height(cm) x Weight(kg) ]/ 3600 )½ or for pounds and inches; BSA (m²) = ( [Height(in) x Weight(lbs) ]/ 3131 )½ The results of other equations such as the Dubois formula yield similar results. Mosteller RD: Simplified Calculation of Body Surface Area. N Engl J Med 1987 Oct 22;317(17):1098 (letter)

MEASURING RISK 25


Table 2. Sediment concentrations that negatively affect human and other organisms in ppm (mg/kg) Exposure REL TEL PEL Severe** MTR Target As 4.1 5.9 17 30 55-3300**** 29 Hg 0.09 0.13 0.49 0.7** 10- 6700 0.3 Cd 0.33 0.6 3.5 10 12-1800 0.8 Cr 25 37 90 110 380-17600 100 Cu 22 16 108 110 73->100000 35 Pb 25 30 91 250 530-3210 85 Ni 16 43 75 210-6700 35 Zn 124 270 410** 620->100000 140 DDT 0.0003 0.001 0.005 0.046** 4-7.3 0.01* PCB 0.025 0.023 0.18 5.3** 1 Benzene 1- 5.5 PAH 1 1.6 40 40*** Oil 1000 from 50 100-100000 REL- rare effect level, TEL – threshold effect level, PEL – probable effect level, Severe- severe effect level, ERM- median effect level, MTR- maximal tolerated risk, Target – level that should be achieved after remediation, As- arsenic, Hg- mercury, Cdcadmium, Cr- chromium, Cu- copper, Pb- lead, Ni- nickel, Zn- zinc, DDT- includes also DDE and DDD , PCB- , PAH- poly aromatic hydrocarbons. *includes DDT, DDE and DDD **is ERM rather than severe, ***includes the sum of only 10 carcinogenic PAHs ****The higher values are called the serious risk concentration which is meant to be equivalent to the MTR, the maximal tolerated risk, for a carcinogenic risk of up to an additional lifetime risk of cancer of 1 per 10,000 and with the following exposure to sediment: eating 350 mg per day, with 50% of the body smeared in sediment for 8 hours per day for a total of 30 days per year over a lifetime. Water drank per time is 50 ml, and exposure of 95% of the body’s surface area to the water.

MEASURING RISK 26

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES Chapter 2. Hazard Identification: General facts about Occupational and Environmental Cancer Most of what we know about the connection between chemical exposures and cancer are from observations in the workplace where is exposure is massive. In 1775, Dr. Pott observed a more than expected number (called a cluster) of scrotal cancer among children who slid down chimneys during their work as chimney sweeps. Scrotal skin cancer otherwise is extremely rare. The children didn’t wear clothes during their work, and had skin exposure to coal tar pitch as well as to other carcinogens. Another classic observation was by Creech and Johnson who took care of workers exposed to vinyl chloride monomers during plastic production. The workers had high exposure to vinyl chloride during hand cleaning of huge vats. They observed a cluster of a very unusual liver cancer with a proliferation of blood vessels (angiosarcoma). The cancer is so unusual that the probability that even 3-4 workers in a factory of 500 men will become ill with this cancer is virtually zero unless there is something that is increasing the risk. In the two examples, causality was relatively easy to prove since the cancers are so unusual. Most cancers however, are common, and causation multi-factorial. Also there is evidence that most cancers occur randomly, in other words are due to bad luck. The International Agency for Research on Cancer (IARC), the U.S. Environmental Protection Agency (EPA) as well as other agencies convenes experts to review the evidence and assign the substances to various Groups. The IARC assigns potential carcinogens to one of four major groups; group 1-4 (Table 1). Table 1.The IARC grouping of potential carcinogens Group Evidence as a carcinogen 1 Sufficient evidence 2 Limited evidence a. probable carcinogen b. possible carcinogen 3 Insufficient evidence 4 Evidence suggesting lack of an association

This is called hazard identification, the first step in a risk assessment. The evidence is substance and organ specific and is very conservative. It takes into account both data in humans and in experimental models (animals, bacteria, etc., Table 1). Included in Group 1 are usually substances that have been shown repeatedly in humans to increase the risk of cancer, as well as being a risk for cancer in animal models. Group 2 requires evidence in animals (experimental models) to be included in this group, and has two subgroups with 2a (a probable carcinogen) requiring in addition to animal evidence, some limited evidence in humans. Group 2b with evidence in animals only signifies a possible carcinogen. Group 2b however is a problematic group since as we will see later nearly 50% of all drugs and even natural chemicals found in fruits and vegetables test positive in animal models. Recently the demonstration that substances can bind and damage DNA under certain circumstances had led to upgrading of probably carcinogens to definite carcinogens even without evidence in humans. This decision is MEASURING RISK 27

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES controversial. A summary of well established group 1 carcinogens is found in the appendix (Tables A1, A2, and A3). Note that common target organs are the bladder, lung, the nasal cavity, and skin; organs with direct contact with carcinogens. This suggests that irritation effects might be the carcinogenic mechanism in some cases. An exception to this rule is exposure to benzene that increases the risk for certain types of acute leukemia. The first step in quantitative risk analysis Hazard Identification, aims to determine the qualitative nature of the potential adverse consequences of the contaminant (chemical, radiation, noise, etc.) and the strength of the evidence for causality. This is done, for chemical hazards, by drawing from the results of the sciences of toxicology and epidemiology. The IARC classifies chemicals at this stage. A chemical might be shown to cause cancer by both inhalation and oral exposure. On the other hand a chemical might cause cancer by inhalation but not by ingestion (Table 2). Table 2 Inorganic and organic carcinogens commonly found in polluted waters and sediments and in air pollution. Chemical

Metals Arsenic (As) Cadmium (Cd) Chromium (Cr) Lead (Pb)inorganic Organic lead Nickel (Ni) Mercury (Hg) Copper (Cu) Zinc (Zn) Organic Benzene*theoretical Poly aromatic hydrocarbons DDT Dioxins Aldrin Chlordane Air pollution SO2 NOx PM2.5 Ozone

Carcinogen Route causing cancer classification Respiratory GI Dermal

Type of cancer associated with exposure

I I I 2A

Yes Yes Yes Yes

Yes No No Yes

No No No No

Skin, bladder, Lung Lung Lung, Nasal septum Kidney (animals only) Limited evidence-man

Yes

No

No

Lung , nasal septum None None None

I

Yes

Yes*

Yes*

2A

Yes

Not Yes shown

2B 2B 2B 2B

Yes Yes Yes Yes

Yes Yes Yes Yes

Yes Yes Yes Yes

Acute non-lymphocytic leukemia Lung by respiratory route Skin by dermal exposure Liver (animals only) Liver (animals only) Liver (animals only) Liver(animals only)

III

No

-

-

III

No

III I IV IV IV

Lung -

MEASURING RISK 28


For example, chromium has been classified as causing lung cancer when inhaled but not when absorbed by the dermal or gastrointestinal routes. Chromium is an essential element for human health, and a deficiency will result in very high blood glucose levels in humans. This information might be relevant but was not mentioned by Julie Roberts in the film Erin Brockovich nor in the governmental Committee report. Therefore, the statement that exposure to a carcinogenic substance is dangerous at any level is not true, and points out the need to consider quantitative and not only qualitative exposures. Nevertheless, the overall decision by the IARC committees, stipulates only that a certain substance is or is not carcinogenic under certain circumstances, without defining the circumstances. This process is only hazard identification, and not as is often assumed the identification of actual risk. This quote was missed by the committee and is a common misconception. “A cancer 'hazard' is an agent that is capable of causing cancer under some circumstances, while a cancer risk is an estimate of the carcinogenic effects expected from exposure to a cancer hazard. The Monographs (IARC reports) are an exercise in evaluating cancer hazards, despite the historical presence of the word 'risks' in the title. The distinction between hazard and risk is important, and the Monographs identify cancer hazards even when the risks are very low at current exposure levels, because new used or unforeseen exposures could engender risks that are significantly higher"(IARC, 2006). In other words, the fact that a hazard is identified in a patient with cancer doesn't mean that the cancer was caused by the hazard. If such a simplistic approach is adopted, then all cancers are caused by chemical exposures. This is because we are all exposed to such chemicals every day in our food, in our water, in the air we breathe, and from physical exposures such as to the sun and to radiation which is all around us. The IARC committees identify the end-organ at risk in animal and in human models, but this doesn’t rule out the possibility that other end-organs are also at increased risk. Their expert decisions take into consideration the available studies in exposed animals and humans, and sometimes in experimental in-vitro systems (outside the body). They determine if the evidence supports causality. In the Kishon water system the carcinogenic chemicals found in the sediments included the metals listed in the table, as well as benzene and poly aromatic hydrocarbons. The sediment acts as a sink because the chemicals are hydrophobic, and concentrations in the water therefore are very low. Only benzene (group 1) and poly aromatic hydrocarbons (group 2A) are considered to be carcinogenic by skin or oral exposure. CAUSALITY (criteria for a causal relationship) (used by the IARC for hazard identification)

MEASURING RISK 29

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES Causality is an attempt to define a possible connection between some factor and the resultant disease. This is a qualitative assessment that attempts to determine if there is a causal relationship between an exposure and a disease under certain circumstances, as opposed to risk assessment, where a quantitative probabilistic assessment is required. The court of law usually talks in terms of more probable than not, or over a 50% probability that the cancer was caused by the exposure. The quantitative probability of causality is decreased if the type of cancer of the plaintiff is not the same type as shown in animals or in humans. This qualifier does not rule out an association but decreases the probability. If there is no qualitative association, then there is no need for a quantitative assessment and the risk assessment can stop here. If the exposure to a chemical is by a route not shown to increase the cancer risk, then the risk assessment can stop at this point. However, if the target organ has not been shown to be involved with the exposure, a quantitative risk assessment can still be done and the final probabilities downgraded after the calculations. Criteria for causality used by agencies assessment of potential carcinogenic agents are listed in table 3. These are often called Hill's criteria and are the accepted methodology used by epidemiologists to determine causality. Table 3.Criteria for causality 1. Strength of association 2. Methodology 3. Consistency 4. Dose-response 5.Temporality 6. Biological plausibility 7. Experimental studies 8. Specificity The qualitative association of a chemical with cancer is based on studies done. The studies should have been done properly (methodology), the disease should have occurred after the exposure (temporality), multiple studies should show a consistent association (consistency) and furthermore the strength of association, or proportional increased risk needs to be high; The strength of association is a measure of how strong a connection is found between the exposure and the disease. For a cohort follow-up study the strength of association is determined by comparing the incidence (I) of disease in the study group (number of cases per a certain group of people over a defined time period) and control group over time, generating a relative risk (RR). A finding of at least 2-fold supports causality. However, if there is only a small increase in risk the observation is more likely to be due to confounders. The lower the increased risk (lower strength of association) the higher the probability that some other factors lead to the differences observed between the two groups. The reason for this is that there are strong risk factors for cancer such as smoking, diet, and lack of exercise that need to be equivalent in the exposed and unexposed groups. Small differences in such factors can lead to increased disease in the exposed group than is not due to the exposure. In the general population we can MEASURING RISK 30

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES explain up to 50% of the risk for cancer by personnel factors and there are estimates that up to 2% of cancers are due to carcinogenic exposures, mostly from natural radiation, the sun, and from air particle pollution. Thus we are often looking for a small boat in a sea of risk. Support for an association can include studies in animals and biological tissues (experimental studies), and the association should be plausible. For cancer this usually means binding and damage caused to the DNA; although some agents might act through hormonal stimulation, by increasing cell turnover and leading to an increased chance of DNA error. We already mentioned that most cancers have many risk factors (lack of specificity). Nevertheless, if the cancer is rare without other known risk factors, even one study showing that those exposed have an increased risk for disease is strong evidence for causality. Strength of association (types of studies) This is a measure of the association (connection) between the factor and the disease and its strength is shown through various types of studies found in the medical research literature. In the qualitative assessment used to assess the possibility of causality, the various types of studies done are assessed. Following is the types of studies in order of the strongest to the weakest evidence. Meta-analysis is an extensive literature review combining the results of individual studies. This is the first step done by the IARC when considering if a chemical is carcinogenic. This is the strongest type of study, and there are those who claim that each individual study is done in order to contribute to a meta-analysis. A single study is not usually adequate in itself to prove causality. However, under circumstances that are unique to a certain setting other studies might not be available and if the findings are very striking with a high degree of association found, with a substantial risk for a rare disease, an association might be considered definite. For example, in Turkey, the general population was exposed to an asbestos-like dust found naturally in one geographical area. The residents had a high incidence of cancer of the lining of the lungs (mesothelioma). Mesothelioma is a very rare disease and it is also not caused by smoking or other known exposures, so that the specificity of the exposure for the disease (not a common situation) and the rarity of the disease in the general population make this single observation strong evidence for a causal association. Cohort, follow-up study with a direct control group: Those exposed are compared to a direct control group, a convenient control group found in the same factory or other available factories where the workers are not exposed to the chemical under study. This is preferable to a study that doesn't have such a control group available because one can get a history of personal confounding factors such as a smoking history. A cohort study is one where the study group and control group are defined at one point in time and then followed up over a certain time period. We try to find out whether the workers became ill in both groups after that time period, and stipulate how many workers were lost to follow-up. A rule of thumb is that one can’t depend on the results if over 30% of the cohort was lost to follow-up. Cohort, follow-up study without a direct control group (compared to a standard population): This is done if we don't have a direct control group, and those exposed are MEASURING RISK 31

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES compared to the general population. Of course here the expected number of cases of disease takes into account gender, ethnicity and especially age. The workers are divided into groups according to age (usually 5- year periods, gender and ethnicity) and then from cancer registries an expected number of cases can be ascertained. This is then compared to the observed number of cases in the study group over the follow-up period. The comparison gives us what is called the standard incidence ratio (SIR) or if we are looking at deaths the standard mortality ratio (SMR). It was found in various countries that certain areas have more cancers than certain other areas. This is a type of cohort follow-up study without a direct control group since we don’t have smoking histories and other personal risk factors other than ethnicity, gender and age. We will see below that this is a common finding, and so far in the United States, expensive studies (e.g. the study of an increased risk of Breast Cancer in an area on Long Island) have not revealed an environmental cause. It is likely that such findings are due to either chance or an uneven distribution of important personal risk factors. This type of study is called an ecological study. One important factor that explains the variation of morbidity and mortality in ecological studies is socio-economic class. For example in 2004-2006 in England and Wales it was found that in Manchester 48% of males die by age 75 compared to 34% for the entire population, thought to be due to lower socio-economic classes in Manchester. There are those who suggest that the best health intervention in third world countries would be to increase the amount of money available to the population. Case-control study: A study where a group a sick people are compared to a group of well people who don't have the disease being studied. The sick people are usually found in a hospital or in a cancer clinic, and the well people are variously chosen, for example from other hospitalized patients, or from other clinics. This type of study is used when it is difficult to study large number of ill patients because the disease is not common. We have already talked about the biggest problem with this type of study, the recall bias, where sick people remember and exaggerate exposures when compared to well people; leading to an erroneous conclusion that the exposure leads to an increased risk of disease. Cross-sectional study: Studies the exposure and the disease at one moment in time, checking if those with disease have a past history of exposures that are different from those without disease. For cancer's this type of study is nearly impossible to do because both the disease and the exposure have to be common. Case study: An observation that a patient became ill after an exposure. A further discussion of epidemiological evidence for causality with examples is presented in the Appendix. Limitations of studies The meta-analysis summarizes the results of cohort studies and experimental studies in animals. In humans it is not ethical to divide a group at random into two groups and expose one group to a possible carcinogen and the other leave unexposed (the control group). So we have available studies of workers, groups of individuals defined at a point in time (cohort) exposed to a chemical or to chemicals sometime during the follow-up period. A follow-up study comparing a cohort of smokers to non-smokers MEASURING RISK 32

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES found that lung cancer was ten times more frequent in those who smoke compared to those who don’t smoke. The ideal study would be one where the exposure was assigned at random. In the smoking versus non-smoking example, the decision to smoke was not a random event. There are many social-economic factors influencing the individual decision to smoke. Theoretically it could be these factors rather than the smoking itself that result in an increased risk of lung cancer. Perhaps those who smoked were of lower social-economic status and therefore worked in factories where exposure to carcinogens in the work place was common. Such factors are called confounding variables, and cannot be ruled out unless the exposure is administered at random. However, attempts can be made to try to control for such factors using statistical models. This problem is the reason that many experts consider as positive evidence, only studies that show at least a two-fold risk. Very high risks such as smoking (10 fold or more increased risk in multiple studies) make an association with lung cancer very probable, despite the fact that historically there were those who claimed that lung cancer was not due to smoking but rather due to other factors that influenced the person to smoke. Other possible biases are summarized in the dictionary under the word "bias". The finding of an association does not in itself prove causality. When there is a lack of understanding of the biology of a disease, then often one can come to the wrong conclusion. For example, in the middle of the 19th century, the cause of Cholera was controversial. John Snow thought it came from drinking contaminated water, whereas William Farr believed it came from contaminated air that was dependent on altitude. In fact Farr demonstrated that in those living less than 20 ft above sea level the mortality rate from Cholera was 120 per 10,000 inhabitants per year, whereas at 340-360 ft the number of deaths from cholera decreased to 8 per 10,000. He also demonstrated a dose –response relationship, a decrease in the death rate for every increase in elevation. Of course at that time it must be remembered that the fact that Cholera was an infectious disease was unknown. Nothing was known of the biology of the disease. The erroneous conclusion of a causal connection between elevation and risk for cholera was due to an important confounding factor, the water supply. John Snow showed subsequently that at higher elevations, the water was less contaminated, and it was the contamination of the water and not the air that resulted in the observed association. Therefore, understanding of how a factor increases the risk of disease (biological plausibility) is of the highest importance and biological plausibility is an essential requirement for determining causality (see below). Confounding variables; the major risk factors for cancer People are often surprised and concerned to find out that many of their acquaintances have become ill with cancer. They don’t realize that cancer is very common. Today human beings in developed countries have a life expectance today of around 80 years. This is 40 years more than in the first part of the 20th century, and 5 years more since the 1970s, Since the 1970s this increase appears to be due related to an increase in the standard of living, since this has not occurred in developing countries or in those where there has been a decrease in the standard of living and increased differences between the rich and the poor such has occurred in the Soviet Union. In any case this increase in life expectancy in the past 40 years is related primarily to a decrease in blood pressure in the general population despite an increase in body weight, and in the incidence of diabetes mellitus leading to decrease in mortality from MEASURING RISK 33

CANCER AND OCCUPATIONAL AND ENVIRONMENTAL EXPPOSURES cardiovascular disease (myocardial infarction and stroke). There are those however who claim that the increase in lifespan is due to better medical care. Smoking has decreased also, and some medical interventions might explain part of the increase in life span that occurred over the last 30 years. Cancer has remained fairly stable over that time period and remains the second most common cause of death. For men, the life-time risk of developing cancer approaches 40%, with half of those with cancer dying of the disease. The risk for cancer increases dramatically with age, the most important risk factor for most diseases. It is uncommon under age 30, becomes frequent in the 60s and in very common over age 80 (Table 4). Most common cancers in males are prostate and lung cancer. Table 4 The total cumulative risk (%) and the risk during various 10-year periods in white males in the US during 2003 (values in percentage)(SEERS) Type age