Occupational cancer in Italy.

This article is a discussion of occupational cancer in Italy. The introduction provides the necessary context of Italian industrialization and occupational health regulation. This is followed by a review of Italian epidemiologic studies of occupational cancer risks considered in terms of relative measures of risk and attributable risk of carcinogenic agents or exposure circumstances. We attempt to establish the number of workers exposed to carcinogens in Italy and the intensity of their exposures. Finally, the Italian system of compensation for occupational cancer is discussed. Several cohort and case-control studies have addressed the issue of occupational risks, mostly among male workers. The results of these studies suggest that the growing incidence of and mortality by mesothelioma is explained by the widespread and intense exposure to asbestos in some Italian industrial settings. A high attributable risk of lung tumors among male populations in industrial areas of northern Italy is explained by occupational exposures. However, insufficient data are available for clear definition of the extent and intensity of occupational exposure to carcinogenic substances. In Italy, we must prioritize and maximize resources in occupational cancer epidemiology and revitalize the role of national institutions. Recent legislation has established new regulations on the handling of carcinogenic substances in industrial settings, a new list of occupational diseases, and a national registry of mesothelioma linked to asbestos exposure. These legislative changes are expected to have positive effects.


Industrial and Regulatory Background Industializaton
In Italy the Industrial Revolution began in the early nineteenth century, considerably later than in other European countries such as England, France, and Germany. It first affected the textile and metallurgical industries in the North. The production demands for the two World Wars accelerated industrial development, which progressed at a tremendous pace after the 1950s. Italy's entrance into the European Economic Community (EEC) in 1957 stimulated development of the automobile industry, mechanical and precision metalworking, and the chemical industry.
Thus, northern Italy developed an industrial structure comparable to those of other areas of northwestern Europe; southern Italy, however, retained its original agricultural orientation. This gap between the industrial development of the North and that of the rest of the country persisted, even though some large metallurgical enterprises and chemical facilities settled in the South. Italy is remarkable for good manufacturing based on unskilled labor and for exploiting few of her own raw materials, although there has been mining of asbestos (chrysotile), bauxite, coal, iron, lead, mercury, pyrites, sulfur, and talc. Italy also has a longstanding shipping tradition, and there are dockyards in both the North and the South.
In the 1960s the Italian Industrial Revolution got a second wind in the rapid growth of smalland medium-size industries. Small firms producing the same product tended to cluster in particular areas. Large firms broke up into networks of smaller specialized artisan or family-run enterprises. These specialized enterprises employed local labor, so that particular areas became known for producing products such as ceramics, tiles, eye glasses, shoes, silk dyeing, furniture, or tinned food.
Thus, by the end of the 1960s, most Italian firms were small or medium size; 40% of the work force worked in enterprises employing 11 to 500 workers, and 3 million workers were artisans. In the 1970s, 7 million Italians were industrial workers, 1.2 million were employed in agriculture, 3.3 million in tertiary activities, and 1.8 million in public administration. A further 7 million were believed to be involved in the underground ("black") economy, meaning they worked in unregistered firms or without paying for social security. One million of these workers were working two or even three jobs simultaneously. Women were an important part of this underground work force (1). Subsequent census data suggest that important changes in the distribution of the work force have taken place since the 1970s. Between the 1970s and the 1990s, the number of people working in office jobs doubled, and there was a 10% reduction in both the agricultural and the industrial work force (2,3) (Table 1).
Women worked in selected industrial and agricultural jobs but were virtually absent in other sectors. In some cases this was because these jobs were traditionally reserved for men and in others it was because of regulations. For example, women could not be employed in mining, in physically demanding jobs, or, during pregnancy, in jobs involving exposure to agents with a reproductive hazard.
In the early twentieth century and again during the autarchy era of the 1940s, poverty forced many workers and their families to migrate to other European countries or overseas. Between 1876 and 1985, 26 million Italians emigrated, mainly from the South and the mountainous areas of northern Italy. In addition, there was a great deal of seasonal and temporary migration of Italians workers, who thus became known as birds of passage.
Abroad, they were employed in dangerous jobs, where they often contracted occupational diseases. An example of this is the persistence of high mortality rates in Italy due to silicosis (4). The economic boom of Table 1. Enterprises and employees in Italy by type of economic activity at the 1981 Census (2), and changes recorded at the 1991 Census (3). No Italy because of their carcinogenic risk. In 1964 benzene was prohibited in the manufacture of glues and varnishes because of the high number of leukemia cases detected among shoemakers, printers, and rubber workers. Some aromatic amines were banned because of the high relative risk of bladder tumors among dyestuff workers exposed to o-toluidine and 4,4'methylenebis(2-methylaniline) (9).
Other carcinogenic agents were prohibited for agricultural uses (arsenic, mercury, 2,4,5-trichlorophenoxyacetic acid), after similar bans had been applied in other countries. Finally, in 1992 Italy banned the production, commercialization, and import of all asbestos fibers. Until 1978, a technical body of the Labour Ministry was in charge of inspecting factories and verifying the application of laws and regulations. At that time, the prevention of occupational diseases was a low priority because of limited resources and the prevailing emphasis on the prevention of accidents in the workplace. In the 1978 health reform legislation, the responsibility for the evaluation of occupational health was transferred to the Ministry of Health. Thereafter, a network of occupational health units was developed as part of the National Health Service. These units were aimed specifically at the detection and prevention of occupational disease and had the right to inspect workplaces. However, they were not adequately supported and maintained in southern Italy, and there was a lack of central coordination.

Review of Italian Epidemiologic Studies on Occupational Cancer Risk
Cohort and Case-Control Studies In this section, we present a systematic review of Italian studies on the association between cancer and agents or activities in the workplace. These studies appeared in peer-reviewed journals between 1975 and 1994. They were traced by a search through MEDLINE and CANCERLIST. Case-series and descriptive epidemiologic studies are excluded from this review. However, the results of several descriptive studies on mesotheliomas will be discussed because they are directly relevant to occupational cancer risks.
Most epidemiologic studies on occupational cancer risks were conducted from 1975 to 1994 because the discipline emerged in Italy after the mid-1970s. This review presents estimates of attributable risk among the exposed (ARe) and the general population (ARp).
A number of cohort studies (Tables 2-4) (10-78) and case-control studies ( Table 5) (79-122) have been published. These two types of studies share several characteristics: the populations studied were primarily in northern and central Italy, the studies were often conducted in cooperation with the occupational health units of the National Health Service; the studies tended to be small (33% of cohort studies were based on less than 20,000 person-years of follow-up).
The cohort studies ( Table 2) began earlier and are more numerous than the casecontrol studies. Because employee records are kept in factories, cohort studies are feasible and can attain a satisfactory level of completeness and accuracy, with municipalities cooperating in providing the vital status of subjects and, where appropriate, the cause of death. However, inadequate exposure data impaired exposure assessment (only 36% of cohort studies had hygiene data collected at the factories under investigation). Rarely has a cohort study been conducted in a large factory or in a number of factories exposed to the same risk. The notable exceptions are those conducted in cooperation with the International Agency for Research on Cancer on vinyl chloride producers and on workers exposed to styrene in reinforced plastic manufacturing plants. Often the agents under study were those for which the evidence of carcinogenicity in humans was already defined. Most studies were of male subjects.
Cohort studies often detected elevated risks for all cancer mortality-up to relative risk (RR) of 2.7-because of high risks of tumors of the respiratory tract (RR values from 1.5 to 6.9 for lung cancer; from 5.6 to 150 for primary pleural cancer; from 1.8 to 4.5 for cancer of the larynx). The highest RR values were detected for primary pleural cancer among workers exposed to asbestos and for cancer of the bladder (RR values up to 30.6) among dyestuff producers (Table 3). These risk estimates suggest that the intensity of occupational exposure to some carcinogenic agents has been very high in the past.
In cohort studies that use the general population as a reference, lack of data about patterns of exposure to risk factors and confounding factors in the general population can lead to spurious estimates of attributable risk (AR).
Cohort studies can be used, however, to compute estimates of AR among the exposed (ARe: RR -1/RR x 100) ( Table  4). It is not surprising, considering that only studies with significantly increased RR values are reported in Table 3, that high ARe were derived. An ARe of approximately 100% for pleural cancer was found among those exposed to asbestos and for bladder tumors among those exposed to dyestuffs. ARe values ranging from 33 to 8% for lung cancer and 23 to 58% for all cancers were observed among those exposed to asbestos, silica, and polycyclic aromatic hydrocarbons (PAHs).
For case-control studies, we report AR values. Occasionally, we estimated the AR percentage in the population (AR,p) from case-control studies on the basis of the following formula: AR% = (OR -1)/OR x Pe, where OR is the age-adjusted odds ratio and Pe is the proportion of exposed cases. This is the only approach that was feasible because direct access to the data was not possible.
Among the Italian case-control studies, a few produced estimates of AR values for nasal cancer from wood and leather dust exposures and for pleural cancer from asbestos exposure. A more substantial number of case-control studies of lung cancer have been specifically aimed at computing AR values from occupational exposures ( Table 5). For unbiased estimates, the studies on lung cancer were based on population, and the probability and intensity of exposure was evaluated by hygienists from working histories derived from face-to-face interviews. These studies, carried out in northern Italy from 1983 to 1992, investigated the incidence or mortality in the male population from 1976 to 1986. We made no attempt to combine the results of these studies, even though this might have been useful (123). In the general male population (the study base), the ARp for lung cancer from occupational agents evaluated as carcinogenic to humans varied from 12 to 39%. Estimates increased to ARp 25.5 to 48.1% when exposure to agents "probably" carcinogenic to humans were included. Occupational exposure to asbestos explains the largest percentage of ARp in two studies (20% of lung cancer in the male population in the dockyard town of Trieste in northern Italy; 50% of respiratory cancer in the male population of a restricted area of northern Italy where most of the work force of a factory was recruited) (102,111,112). Among those exposed, the AR values varied from 31 to 84% for occupational exposure to agents that have been proven to be carcinogenic.
Exposure to leather dust resulted in a very high ARp (92%) for nasal cavity cancers in a population-based case-control study performed in an area where shoe manufacturing was the dominant industry (90). An ARe up to 80% was reported among subjects occupationally exposed to both leather and wood dust in a hospitalbased study involving several areas where employment was predominantly in shoe and furniture production (114). Only one population-based case-control study on bladder cancer has been done in Italy to date, although there have been several hospital-based studies. The population-based study was carried out in an industrialized area near Genoa. No estimate has been made of ARp due to exposures to risk factors of industrial origin; however, significant excesses due to dyestuff production and PAHs have been observed (99).
A limitation of ARs is that they are time-and place-specific, i.e., they express exposure circumstances of the location and the time in which each study was conducted. If studies are conducted in areas with a high prevalence of high-risk occupations, the estimates could overestimate the risks among other populations or at the national level. However, as previously mentioned several areas of Italy are highly specialized in similar productions, and estimates derived in any of these areas should be reasonably expended in others.
The studies provided information on 12 occupational categories indcluding farming, mining, and the chemical, metal, textile, leather and shoe, rubber, and plastics and 44 specific causes of death including several cancer sites, myocardial infarction, respiratory diseases, cirrhosis, accidents. Separate studies have been done on women (126) and farmers (127).
Manual workers were shown to have a significantly higher relative risk of mortality than nonmanual workers (125 Italy (1975Italy ( -1994, main characteristics (listed by year of publication).

Mesothelioma Due to Asbestos in Italy
The mining of asbestos in Italy began in 1870. The Balangero mine in the northern Italian Alps was a rich source of chrysotile, and this availability encouraged Italian consumption of raw asbestos. Both amosite and crocidolite have been extensively used in Italy. For example, amosite was used in pipe insulation and in shipyards as insulation applied in spray form; crocidolite (which was not banned until 1986) was used as insulation for railway carriages and was also applied in spray form. The consumption of raw asbestos increased sharply after the 1950s, when it became comparable to that of the United Kingdom and France. However, it was not until the 1970s that the intensity of exposure to asbestos in workplaces began to be controlled.
The cancer registries of Genoa and Trieste are now recording the world's highest incidence rates of mesothelioma (pleural and peritoneal combined) among males: 5 + 0.4 and 6.36 + 0.85 age-standard rate x 100,000 (world population as standard), respectively (128). High incident rates of mesothelioma have also been 266 observed in the female populations of both cities (0.92 + 0.17 and 0.56 + 0.25, respectively); that of Genoa's female population is the highest worldwide. Because women do not work in shipyards of either city, environmental exposures would appear to be the most plausible explanation for this high incidence of mesothelioma (128).
High rates of mortality from primary pleural tumors were observed in several shipyard areas of Italy: Trieste, Genoa, Gorizia, Venice, La Spezia, Taranto, Leghorn, and Naples. They were also observed in areas of Piedmont and Lombardy where asbestos has been used in textile production, automobile brakes, and in asbestos cement manufacture (129). A positive correlation was observed between areas with high mortality rates from pleural and peritoneal tumors (130) and areas in which numerous compensations have been awarded (131).
The high rates of mesothelioma detected in descriptive epidemiologic studies have been investigated and explained as being the result of high risks among shipyard, asbestos cement manufacturers, and textile asbestos workers (11,(132)(133)(134)(135).
Attempts to Estimate the Number of Workers Exposed to Carcinogens and the Intensity of Their Exposures Systematic data on the number of workers exposed to carcinogens are not available in italy. Estimates have been made in the context of population-based case-control studies (Table 4). These studies, however, have been conducted for the most part in northern Italy, and the data they provide are rather heterogeneous because of the various methods of exposure assessment used. It is difficult, therefore, to judge whether results of these studies reflect genuine differences or merely differences of approach.
To overcome this difficulty, we have chosen three recent studies in which study designs were similar (population-based case-control studies, with control groups that are random samples of the general population) (112,136) and whose common methodology led to a particular accuracy in the assessment of exposure. We have reconsidered the original data of these studies, one of which was on cancer of the larynx, one on hematolymphopoietic malignancies, and one on lung cancer. The study on cancer of the larynx used the  (137) and/or the 4-digit International Standard Classification of all Economic Activities (138,139). A team of industrial hygienists (or agronomists, in the case of the first study) assessed the probability of a person being exposed to various agents on an ordinal scale (i.e., unexposed, possibly exposed, certainly exposed, certainly exposed at high levels). There has been a historical development in the methods for exposure assessment in the context of population-based studies. The study of hematolymphopoietic malignancies, in particular, reflects this development, which consisted of a) the use ofjob-specific forms (in addition to the general occupational history) in occupational interviews, and b) the standardization of exposure assessment procedures by experts, including the evaluation of inter-rate agreement (136).
We used only a fraction of the information available in these studies for our estimates. In our preliminary analysis of the study of hematolymphopoietic malignancies, for example, we only considered the frequency of having been certainly exposed. More detailed information on intensity and probability will be forthcoming in future studies.
In the study of cancer of the larynx, 7.8% of subjects were certainly exposed to asbestos in their work histories. This is because the study was carried out in heavily industrialized areas. Exposure began around 1944 and was mainly in the metal products industry.
In the study of hematolymphopoietic malignancies, 35 men (4.8%) and 6 women (1 %) were certainly exposed to asbestos. Since about 1954, the construction industry was the employment sector in which more frequently exposure occurred; 26 men (3.5%) and 4 women (0.6%) were certainly exposed to benzene. Among the men, most of these exposures occurred before 1964 (when the law was changed), and only four exposures began after that date (three in construction and one in the metal industry).
The lung cancer study in Trieste assessed the cancer risk from exposure to asbestos (used in shipbuilding, Trieste's largest single industry) and PAHs (used mainly in steel foundries and metal working, both of which are also large employers in the area). Ninety-four of the 755 controls (12%) were certainly exposed to asbestos and 36 (3.4%) were certainly exposed to PAHs.
These three studies have limitations: they considered only a few carcinogens, they studied a relatively small number of subjects, and the areas where they were conducted were not selected to be representative of the general Italian population. It is possible nevertheless to make some comments on asbestos exposure in Italy in the light of the results of these studies.
The proportion of controls exposed to asbestos varies between 12% in an area dominated by a single industry (shipbuilding) and 4.8% in the combined results from nine areas with different levels of industrialization (induding areas in central and southern Italy). The intermediate value of 7.8% in Varese and Turin makes sense, when considering the previously mentioned extremes. Furthermore, the results suggest that the construction industry is the source of workers' more recent exposure to asbestos.

Compensation for Occupational Cancer
Italy has had a system for compensation for injuries in the workplace since 1898 and for occupational diseases since 1934. This system is based on the principle that the workplace can be hazardous and employees must, therefore, be insured under a national insurance program at the employer's expense, the cost to be determined by the extent of the estimated risk. If a worker has an accident or contracts an occupational disease, the cost of treatment will be reimbursed to the National Health System by the insurers. The worker will receive compensation if he is disabled to the extent that his original capacity to work is decreased by 1 1 % or more; he is also compensated if his employer did not pay the insurance. However, the system has always differentiated between compensation for accidents and compensation for diseases. The category of accidents at work is rather broad, including those in itinere, whereas diseases are more strictly defined as those caused by work. Furthermore, the percentage of disablement as a result of an accident is easier to calculate than that caused by diease.
Since 1988, Italy has had a mixed system for compensating for occupational diseases. A list has been compiled of occupational diseases and the right to compensation if a causal link is proven between an exposure and a disease (140). Items on the list fall into two distinct group-one in which the disease can vary but the etiologic agent is clearly defined (e.g., diseases caused by lead); and one in which both the disease and the agent are clearly defined (e.g., skin diseases caused by cement). A similar distinction is made among occupational cancers, for example, diseases caused by chromium (lung cancers among those exposed to hexavalent chromium could be included in this category) and mesothelioma caused by asbestos (in which both the cancer and the agent are specified).
The current list was introduced in 1994. It includes 58 categories for industrial and 27 for agricultural activities (140). A distinction is thus made between industrial and agricultural claims. Agricultural workers get less compensation; they have only been compensated for occupational diseases since 1958. This difference in compensation is only partly explained by genuine differences in exposure situations. The current list devotes far more space to occupational cancer than the previous list. It includes all occupational exposures recognized by the International Agency of Research on Cancer and the Italian National Toxicology Committee as having "sufficient evidence" of carcinogenicity.
Limitations of the Italian system of compensation for occupational cancer are the following: it is extremely rigid, it allows few workers actually to receive compensation, and it tends to discourage daims. We consider each of these problems individually.
The Italian system is rigid because it has shown little ability to evolve over time in response to new knowledge and conditions. The list of diseases for which one can be compensated has been revised only 6 times since its introduction in 1934. The 1994 list replaced a list on the books since 1975. Italian legislation does not require regular updating of the list, and there is no permanent independent committee of experts to suggest change whenever the need arises. Lack of consideration on new epidemiologic findings linking exposures and diseases has meant that new diseases are not added to the list when they should be. A striking example of this is mesothelioma due to occupational asbestos exposure: Environmental Health Perspectives * Vol 107, Supplement 2 a May 1999 mesothelioma, in the absence of asbestosis (which was listed), was not included in the list until 1994 (141). For decades before that, studies on mesothelioma in Italy clearly argued for the need to update the compensation system for cancer caused by asbestos.
Few Italian workers actually receive compensation for occupational cancer (Table 6) (142). Since 1988 when the mixed system was introduced, 53% of the 466 daims for cancer have been rejected by the insurer in charge of the largest number of industrial workers. Fortunately, the 1994 list could be applied to past claims still under scrutiny, so the number of awards has recently increased. Nevertheless, many claimants have been turned away. The long latency period of occupational diseases, the low profile that the National Health System gives to the link between diseases and occupational exposures, and the lack of an active search for occupational diseases and of proper registries for some of those already known all help explain the low number of requests for compensation.
Finally, the Italian system of compensation actually discourages claims. No effort is made to inform workers of their rights. Instead, the system maintains a passive attitude of replying to requests if there are any. If a claim is made, a worker cannot approach the system as an individual. The assistance of trade union institutes is usually needed to follow the overall process from daim to the final decision on compensation. Finally, some working populations are dealing with special insurers. Recently, railway workers affected by mesotheliomas due to occupational exposures to asbestos reacted energetically against this state of affairs. When, even after 1994, all requests for compensation were denied, railway workers and their relatives went to court to overturn the rejections of their insurance claim. Conclusions The extent and severity of occupational cancer in Italy cannot be adequately described because of lack of information on the number of workers exposed to carcinogenic substances, the intensity of these exposures, and the variation over time. No inquiry has been made at the national level, and any such inquiry would encounter difficulties in finding, compiling, and interpreting information collected at the local level by the occupational health units of the National Health Service. The National Institute of Occupational Health has a low profile, and funding of projects on occupational cancer risks has been too modest. Some epidemiologic research has been done nonetheless (Tables 2-5) to investigate causal associations between exposure and disease where an agent is suspected of being a carcinogenic risk or of quantifying the impact of exposures to known carcinogens. However, few of these studies investigated the same risk in various industries and few attained sufficient statistical power and/or a gradient in exposure intensities.
The largest Italian industries-producers of cars, chemicals and petrochemicals, pharmaceuticals, etc.-have not been the subject of epidemiologic studies on cancer risk because they did not wish to participate in such studies. Companies known to use carcinogens often avoid or impede epidemiologic studies because they do not appreciate the public health benefit and fear liability.
The planned extension of record-linkage studies may constitute an alternative way to obtain new epidemiologic data. Italy should follow the example of other countries that require firms upon liquidation to release their employee records to national institutes.
Italy needs to prioritize and maximize resources in occupational cancer epidemiology and to revitalize the role of national institutions. A national registry of mesotheliomas linked to asbestos exposure has recently been established on the suggestion of a European Directive. A national registry of cancer cases suspected to be of occupational origin has been recommended in the legislation on carcinogenic substances at work. If these projects are pursued, we can expect a positive effect on the prevention and diagnosis of and compensation for occupational cancers.
ACKNOWLEDGMENTS. We thank F. Berrino and F. Barbone for providing original data on their studies, J. Vena for helpful suggestions and comments, D. Barret for editing the manuscript. E. Merler was at the Unit of Environmental Cancer, International Agency for Research on Cancer, Lyon, France, at the time this article was written.