Review of the Evidence from Epidemiology ... - ACS Publications

Subscriber access provided by Duquesne University - Gumberg Library

Review

A Review of the Evidence from Epidemiology, Toxicology and Lung Bioavailability on the Carcinogenicity of Inhaled Iron Oxide Particulates Camilla Kay Pease, Thomas Rücker, and Thomas Birk Chem. Res. Toxicol., Just Accepted Manuscript • DOI: 10.1021/acs.chemrestox.5b00448 • Publication Date (Web): 10 Feb 2016 Downloaded from http://pubs.acs.org on February 24, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Chemical Research in Toxicology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 75

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Chemical Research in Toxicology

A Review of the Evidence from Epidemiology, Toxicology and Lung Bioavailability on the Carcinogenicity of Inhaled Iron Oxide Particulates Camilla Pease1, Thomas Rücker2, Thomas Birk3.

1 Ramboll ENVIRON UK Limited, 1 Broad Gate, The Headrow, Leeds, LS1 8EQ, UK. 2 Ramboll ENVIRON Germany GmbH, Aschauer Straße 32a, 81549 München, Germany. 3 Ramboll ENVIRON Germany GmbH, Friedrich-Ebert-Strasse 55, 45127 Essen, Germany.

Address for Correspondence: Camilla Pease, Ramboll Environ UK Limited, 1 Broad Gate, The Headrow, Leeds, LS1 8EQ, United Kingdom. Tel: +44 1132 457552. [email protected]. Key words: pulmonary overload, particle overload, particle effect, poorly soluble particles, iron metabolism, bioavailability, reactive oxidative stress, Fenton reaction, tumorigenicity, mode of action.

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment

Page 2 of 75

Page 3 of 75

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Abstract Since the iron-age and throughout the industrial age, humans have been exposed to iron oxides. Here, we review the evidence from epidemiology, toxicology and lung bioavailability as to whether iron oxides are likely to act as human lung carcinogens. Current evidence suggests that observed lung tumours in rat results from a generic particle overload effect and local inflammation that is rat-specific under the dosing conditions of intratracheal instillation. This mode of action therefore, is not relevant to human exposure. However, there are emerging differences seen in vitro, in cell uptake and cell bioavailability between ‘bulk’ iron oxides and ‘nano’ iron oxides. ‘Bulk’ particulates, as defined here, are those where greater than 70% are >100nm in diameter. Similarly, ‘nano’ iron oxides are defined in this context as particulates where the majority, usually >95% for pure engineered forms of primary particulates (not agglomerates), fall in the range 1-100nm in diameter. From the weight of scientific evidence, ‘bulk’ iron oxides are not genotoxic/mutagenic. Recent evidence for ‘nano’ iron oxide is conflicting regarding genotoxic potential, albeit genotoxicity was not observed in an in vivo acute oral dose study and ‘nano’ iron oxides are considered safe and are being investigated for biomedical uses; there is no specific in vivo genotoxicity study on nano iron oxides via inhalation. Some evidence is available that suggests, hypothetically due the larger surface area of ‘nano’ iron oxide particulates, toxicity could be exerted via the generation of reactive oxygen species (ROS) in the cell. However, the potential for ROS generation as a basis for explaining rodent tumourigenicity is only apparent if free iron from intracellular ‘nano’ scale iron oxide becomes bioavailable at significant levels inside the cell. This would not be expected from ‘bulk’ iron oxide particulates. Furthermore, human epidemiological evidence from a number of studies suggests that iron oxide is not a human carcinogen, and therefore based upon the complete weight of evidence, we conclude that ‘bulk’ iron oxides are not human carcinogens. 3 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 75

Table of Contents Introduction

4

Literature Search

6

Review of Human Epidemiology and Occupational Exposures to Iron Oxides

7

Toxicology Studies on Iron Oxide

12

Potential mode of action for tumorigenicity in rats

27

Bioavailability of iron oxide

37

Application of the WHO IPCS Mode of Action Framework

42

Summary & Conclusions

45

4 ACS Paragon Plus Environment

Page 5 of 75

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Introduction Iron oxides encompass a number of Fe, Fe (II) and or Fe (III) compounds, which can be naturally occurring or produced by man on an industrial scale. Naturally occurring mineral forms of iron oxides include hematite (mainly comprising Fe2O3), magnetite (mainly Fe3O4) and wüstite (mainly FeO). The largest source of iron oxides globally is in the mining of iron ore and the subsequent manufacture of iron and steel via the integrated iron and steel process. Iron oxide in the form of powder is also manufactured by industry for various uses such as polishing compounds, pigments, magnets, or catalysts, to name but a few.1 Apart from iron and steel production the predominant industrial use of iron oxides is in construction, paints/coatings, masonry and tiles. More specialist applications include use in plastics, rubber, asphalt, bitumen, artists’ colours and wood varnishes. Occupational exposure to iron oxides occurs mainly via the unintentional release of iron oxides from iron or steel production or, hot work activities such as welding or melting of steel and other hot ferrous metal workplaces such as foundries, casting and forging and also iron ore mining. Furthermore, workplace exposure to iron oxides occurs in underground railways as generated by interaction and friction of brakes, wheels, and rails and engineered ‘nano’ form metal oxides, including iron oxides that are used in the toner manufacturing industry. 2,3

It is important to understand the chemical and particulate form of iron oxide when attempting to interpret the mode(s) of action relevant to particular toxicological observations. The word ‘bulk’ is used in different contexts in the iron oxides industries, but in the context of this review ‘bulk’ is defined similarly to how it is defined in recent toxicological publications i.e. as ‘bulk’ particulates where greater than 70% by particle number are >100nm. Similarly, ‘nano’ iron oxides are particulates where the majority, usually >90% of pure engineered nanoforms of primary particulates (not agglomerates), fall in the range 1-100nm in diameter. 5 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Human exposures to ‘nano’ form iron oxides are now under investigation in a variety of biomedical applications, although administration is not usually via the lung.4

The toxicology and epidemiology data for iron oxides has been reviewed generically and previously by authoritative bodies and independent academics none of whom found convincing evidence for an association between iron oxide exposure and risk of cancer, specifically lung cancer.1,5,6,7,8 More recently, the German Commission for the Investigation of Health Hazards of Chemical Compounds in the Work Area (MAK Commission) published a report on the health effects of iron oxides.9 In this report (which did not consider recent epidemiological data from studies in iron and steel industry workers, the MAK Commission classified bioavailable iron oxides as suspected carcinogenic substances based primarily on observations in the rat together with considerations of putative mechanisms of tumourigenicity via free bioavailable iron leading to reactive oxygen species (ROS) formation.10 The rat is known to be more susceptible than other species to the effects of poorly soluble particle (PSP) lung overload. It has been thoroughly reviewed in a technical guidance document produced by ECETOC.11 Notably, differences in lung tissue distribution of PSPs have been observed between rats and humans, and pulmonary inflammation is a pre-requisite for tumourigenicity. In the ECETOC document, based upon a comprehensive review of evidence, it states ‘it can be concluded that the observed lung tumours in rats are not due to a substance-specific toxicity but the result of generic particle toxicity in a highly sensitive species’. We consider that this mode of action of lung PSP overload in the rat is at play for iron oxides.


Page 6 of 75

Page 7 of 75

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The hypothesis in the MAK commission report to suggest that airborne iron oxide particulates can be solubilised in the lung and that iron oxide could be a substance-specific human relevant carcinogen via a ROS-mediated mode of action remains speculative.9 The bioavailability of iron as released from ‘bulk’ iron oxides in the lung is expected, from recent evidence, to be very low, and a full review of the mechanistic context is warranted, before definitive conclusions can be drawn that iron oxides can act as human carcinogens.

Hence, in this paper we provide a critical evaluation of the available epidemiological and toxicological evidence to date, including some recent data regarding lung effects of iron oxides and we also discuss the chemistry, bioavailability and mechanistic aspects of iron oxides in the lung.12

Literature search As background work to the registration of several iron compounds under REACH, we conducted a comprehensive literature search for all evidence pertaining to the health effects of iron oxides, primarily in the United States National Library of Medicine PubMed and Toxline databases, supplemented by the evaluation of the reference lists of existing studies, reviews and key publications. The review included toxicology data available from industry, studies concerned with the human health effects (epidemiological and case studies in occupational settings, as well as controlled studies in human volunteers) and the toxicodynamics and toxicokinetics of iron, iron oxides and iron salts.

For this current consideration of the weight of evidence, we have built upon the previous data review performed by the Iron Platform for REACH dossier submission (http://www.iron-



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

consortium.org/substances.html) and included toxicology and epidemiology literature and in vitro studies published after 2009. We have reviewed the recent new evidence post-2009 covering the potential for oxidative stress mechanisms arising from iron oxide exposure in vitro and in vivo and the bioavailability of iron oxide in mammalian systems. The studies reviewed from recent years bring new information and add to the overall weight of evidence to assess the likelihood of whether ‘bulk’ iron oxides should be regarded as human carcinogens. It is clear from the literature that there are distinctly different chemical and physical forms of iron oxides that are best considered separately (see Figure 1). We begin by reviewing the epidemiology data in humans, followed by the animal toxicology data and then look at the available evidence on mode/mechanism of action of iron oxides in the lung. Figure 1 here

Review of Human Epidemiology and Occupational Exposures to Iron Oxides It was observed by Billings & Howard in a review of ‘welder’s lung’, that iron ore miners, foundry workers and welders suffer from an increased risk of lung cancer, and that this has been attributed to smoking, exposure to tars and radon etc.13 Their hypothesis was that siderosis of the lung (i.e. deposition of iron in the tissue) could also be a contributory factor in the pathology of lung disease and that welding fumes, containing iron, could lead to adverse lung pathology. However, in this review it was not possible to definitively attribute iron oxides exposures to the increased incidence of lung cancer due to many other confounders.


Page 8 of 75

Page 9 of 75

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


An assessment of cancer risk due to exposure specifically to iron oxides in occupational settings is problematical, since most industrial activities which may give rise to iron oxide exposure generate mixed exposures containing several different chemicals and dusts, or radiation. While excess risks of lung cancers have been reported in some epidemiological studies of iron ore miners, foundry workers, steel workers and welders, industrial settings where exposure to iron oxides were involved, the majority of the studies were not able to separate iron oxide exposure from other exposures to known or suspected carcinogens (e.g. radon, diesel exhausts, other metal fumes, crystalline silica, polycyclic aromatic hydrocarbons) often present at the same settings or same workplaces; or, iron oxide per se was simply outside the specific research interest and so not investigated at all.

However, there are a number of occupational epidemiological studies available which either have tried to separate iron oxide exposure from other exposures or which have been conducted under circumstances in which only low levels of exposure to other agents were present. These studies have been conducted in different industrial settings.

Iron and steel industry Moulin et al. conducted a nested case-control study within a cohort of 4897 workers in an alloyed and stainless steel producing French factory and which included mortality between 1968 and 1992 to evaluate the potential relationship between iron oxide and other exposures and lung cancer development.14 Fifty-four cases and 162 controls were selected for the casecontrol study. A job-exposure matrix (JEM) was created by a panel of five experts including epidemiologists, occupational physicians and industrial hygienists. The JEM was based on the job periods provided by the job histories from cases and controls, and the panel interviewed former and present workers about working conditions to derive semi-quantitative 9 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

exposure estimates for intensity and cumulative exposure due to the absence of any exposure measurements. No risk due to iron oxides exposure was detected in the crude and adjusted analysis, neither for increasing exposure levels, duration of exposure, frequency unweighted or weighted cumulative dose. The odds ratio (OR), adjusted for potential confounding factors, was less than 0.50.

In a cohort study, Bourgkard et al. followed 17 701 (16 742 men and 959 women) French carbon-steel workers for 30 years looking at the potential effects of iron oxide exposure on lung cancer mortality.10 The workers included in the cohort had been employed at least one year between 1959 and 1997, with follow up from 1968 to 1998. The causes of death were taken from death certificates. Job histories and smoking habits were available for 99.7% and 72.3% of the subject male and females, respectively. Occupational exposure was estimated by a JEM developed by 8 experts that was validated by more recent air measurements (around 970 total dust samples) taken at the plant. Quartiles of total dust concentration were 1.78, 3.22, 8.48 mg/m3 and percentage of total iron in total dust estimated to be in a range of 10%-50%. Based on 233 lung cancer cases, relative risks (RR) were calculated for internal comparisons and standard mortality ratios (SMRs) were calculated for external comparisons. Among men, no lung cancer excess was found for iron oxide exposure (RR 0.80, 95% CI 0.55 - 1.17). Observed mortality for lung cancer was lower than expected compared to the local population (233 deaths, SMR 0.89, 95% CI 0.78 to 1.01) but higher than expected for the French population (SMR 1.30, 95% CI 1.15-1.48). For all subjects exposed at an intensity level >2, the RR based on 64 exposed cases, was less than one (RR 0.80, 95% CI 0.55 to 1.17). Importantly, no dose–response relationships were observed in the Poisson regression models with the highest exposure level in work history (RR per added level 0.98, 95% CI 0.87 to 1.10), duration of exposure at an intensity level >2 (RR 0.82 per duration period, 95% 10 ACS Paragon Plus Environment

Page 10 of 75

Page 11 of 75

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


CI 0.62 to 1.07) and by quartile of the frequency weighted cumulative index (RR 1.00, 95% CI 0.89 to 1.12). The authors also did not find a dose-response relationship with either intensity, duration of exposure or cumulative index. Overall, the authors did not find any relationship between iron oxide exposure and lung cancer mortality in this large cohort.

Sulfuric acid production from pyrite Axelson & Sjoberg conducted a case-control study in the City of Helsingborg, Sweden, with around 80 000 inhabitants and a plant producing sulfuric acid from pyrite since 1905 to evaluate the possible carcinogenic effects of high iron oxide exposure in the almost complete absence of exposure to other possibly carcinogenic agents.15 Cancer cases were identified through the Swedish National Cancer Register between 1958 and 1971 and controls (2 per case) selected from the local population register. The respective exposure status was determined by search in company records. A total of 771 cases of stomach, liver, lung, kidney, bladder and hematological cancers were identified. Exposure in the plant occurred to "mist" containing almost exclusively iron oxides (Fe2O3) impurities (identity and concentrations): 0.01-0.1% iron arsenate (FeAsO4), 1-2% copper, "small amounts of nickel and cobalt"- Other: 5-10% of dust particle size < 5µm. No measurements were available, but a semi-quantitative expert assessment based on knowledge of process changes over time was conducted and exposure levels estimated to be between 50-100 mg/m3. 183 cases of stomach cancer, 33 cases of liver cancer, 217 cases of lung cancer, 59 cases of kidney cancer, 150 cases of bladder cancer and 129 cases of hematological malignancies were detected with relative risks (RR) for stomach cancer RR = 1.0; liver cancer RR = 0; kidney cancer RR = 4.0; 90% CI 0.62-25.74; lung cancer RR = 1.0; bladder cancer RR = 1.0; hematological malignancies RR = 0.14; 90% CI 0.03-0.60. No trend for lung cancer was seen in an analysis



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

by exposure scores. In spite of the high exposure levels, no excess of cancer has been observed, either in the respiratory system or at other sites.

Iron ore miners Lawler et al. followed 10 403 Minnesota iron-ore (hematite) miners for 32 years looking at the potential effects of hematite ore exposure on mortality.16 There was an apparent absence of significant radon daughter exposure, a smoking prohibition that was strictly enforced on all underground miners while working, an aggressive silicosis control program, and the absence of diesel motor use underground. The authors found no excesses of lung cancer in either the underground (SMR = 100) or above ground (SMR = 88) miners. In Yugoslav-born miners however, there was a significant excess mortality from lung cancer that did not appear to be associated with mining exposures. A significant excess of stomach cancer was found in both above ground and underground miners compared to the US population, but this increase disappeared in all but Finnish-born miners when compared to the local county population. The apparent absence of significant radon exposure, the strict smoking prohibition underground, the aggressive silicosis control program, and the absence of underground diesel fuel use may explain why these underground miners did not appear to incur the lung cancer risk reported in other studies of iron ore miners where the confounders cannot be separated. However, quantitative exposure information was not available.

In summary, those epidemiological studies specifically addressing the effect of iron oxide on lung cancer did not provide any evidence for an increased risk of lung cancer associated with exposure to iron oxides.


Page 12 of 75

Page 13 of 75

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Toxicology Studies on iron oxide In vitro cell-based assays A series of in vitro studies with iron oxides are available in which the effects of iron oxide particles are examined in different biological systems. With one exception, ‘bulk’ iron oxide particles did not induce any cytotoxic effects when tested on different cell cultures.17

Hanawa et al. examined the effects of Fe2O3 and Fe3O4 in human gingival fibroblast cells. The fibroblasts (5 ml of 1 x 105 cells per ml in Fagle’s MEM with 10% FCS) were evenly seeded on a cover glass (24 mm X 40 mm) in a tissue culture dish (60 mm indiameter and 15 mm in depth). Cultured cells were exposed to 5g of test substance powder in contact with the cells for 24 hours.18 This investigation was part of a study to investigate potential cytotoxicity of metal oxides as used in dental alloys. Individual metal oxides were milled into the particulate size range 0.5-3 µm as verified by x-ray photoelectron microscopy. Copper was used as a positive control, chromium-, copper-, silver- and zinc oxides all showed signs of cytotoxicity in this assay. Iron oxides showed no cytotoxicity compared to controls.

Warshawsky et al. isolated viable alveolar macrophages from hamster or rat trachea and exposed them to iron(III) oxide particles (particle size: 98.9% ≤5 µm; 91.5% ≤ 1 µm; median 0.32 µm) at the concentrations of 0, 0.05, 0.1 and 0.5 mg/ 1 x 106 cells, for 24 or 48 h.19 No effects were observed on the phagocytic activities of the alveolar macrophages. A small effect on cell viability was observed at the highest dose of 0.5 mg/1 x 106 cells; in hamster 81.1% viability compared to 89.3% control, after 24 hours and in rat 85.1% viability compared to 89.6% in control. Peritoneal macrophages obtained from mice by lavage were exposed to Fe3O4 particles for 18h at concentrations up to 150 µg/ml. Lactate dehydrogenase



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(LDH) and beta-glucuronidase (BGLU) levels were measured in the cell lysates. Fe3O4 had little effect on the release of these factors.

Cytotoxicity was assessed by Könczöl et al, who studied magnetite particles in human alveolar epithelial-like type-II cells (A549); the four forms studied were magnetite (0.2-10 µm), respirable fraction (2-3 µm), alveolar fraction (0.5-1.0 µm), and ‘nano’ particles (20-60 nm).17 Although the composition of magnetite mineral is mainly Fe3O4, (72% wt in ‘bulk’ forms, and 71% wt in ‘nano’ form), other chemical components of magnetite comprise Al, Cr, Mg, Mn and Ni. Therefore, it is not a pure assessment of iron oxide per se, but of the magnetite mineral. After 24 h of exposure, the A549 cells were investigated by transmission electron microscopy (TEM) to study particle uptake. Cells exposed to ‘bulk’ magnetite, took up only a few particles

Review of the Evidence from Epidemiology ... - ACS Publications

Recommend Documents